AN ABSTRACT OF THE THESIS OF Michael James Thompson for the degree of Master of Science in Marine Resource Management presented on February 4, 2005. Title: Integrating Traceability with Onboard Handling to Enhance Product Quality and Marketability of Eastern Pacific Troll­caught Albacore Tuna (Thunnus alalunga). Abstract Approved:______________________________________________________ Gilbert Sylvia The fishing industry in the Pacific Northwest is a dynamic and highly competitive industry that must constantly respond to rapidly changing resource, regulatory and market forces. These forces have already had a significant impact on our fisheries and will continue to do so as we learn more about the biology and ecology of harvested species, and how this information can be used to effectively manage our fisheries. In addition, market forces, both here and abroad, including the push by some governments to improve food safety and consumer confidence, are also having an impact on the U.S. seafood industry. In an attempt to address these issues, the European Union (E.U.) recently enacted traceability legislation, now mandatory for seafood, which will require all U.S. businesses exporting to the E.U. to provide adequate traceability documentation on their products. This legislation will have an impact on many U.S. fisheries, including the troll­caught albacore fishery in the Eastern Pacific Ocean, which, after losing its traditional market, has recently relied heavily on exporting. Although the term traceability has been around for some time it is relatively new to the U.S. food industry and many in the seafood sector, who now face the possibility of mandatory traceability, have not yet considered all its implications. Another issue facing the Eastern Pacific albacore fishery is that of product quality. Traditionally, this fishery has concentrated on producing albacore for the high­volume cannery market, which does not require the same stringent quality control standards as other market sectors. Albacore tuna (Thunnus alalunga),
members of the Scrombroid family, are capable of thermoregulation and can exhibit internal temperatures of more than 15º C above ambient seawater. High internal temperatures, if sustained, and residual blood content can stimulate bacterial growth, which can eventually lead to scrombroid poisoning in humans. This makes the time immediately after landing critical in attaining and preserving important quality traits. In order to develop new markets for Eastern Pacific troll­caught albacore, quality needs to be improved and/or maintained within the industry at a sufficient level to supply market demands. This would require at least a portion of the industry to implement a set of onboard handling procedures designed to preserve and maintain quality. Although guidelines for handling albacore are available there is no industry “standard” for producing high quality products and only limited research presently exists into preserving quality traits of albacore onboard a vessel. The first paper in this thesis explores the concepts of traceability, current traceability legislation, and the potential applications of traceability systems to improve quality and marketing of U.S. seafood products. The second paper evaluates different onboard bleeding and handling techniques, which can affect product quality, currently used in the albacore industry and their ability to remove blood from the muscle tissues. Thirty­two different combinations of handling techniques and their affect on residual blood content were investigated by examining blood concentrations within the muscle tissue, using spectrophotometry methods, and cutaneous blood vessel coverage, by computer­aided digital color analysis. The results of the statistical analysis indicate that three of the five factors investigated had some influence on residual blood content. The third paper in this thesis evaluates the design and development of a computer­based onboard traceability system with integrated barcode technology, capable of efficiently recording capture and handling data. Two system trials were conducted, during normal fishing operations, where a total of 450 albacore tuna were successfully entered into the traceability database through the use of the Fishery Data Interface application. Documentation supplied by this system, which includes both spatial and temporal data, as well as information on product activities, will enable
businesses to meet existing traceability requirements while also providing a new tool for the marketing of seafood products produced in the United States.
©Copyright by Michael J. Thompson February 4, 2005 All Rights Reserved
Integrating Traceability with Onboard Handling to Enhance Product Quality and Marketability of Eastern Pacific Troll­caught Albacore Tuna (Thunnus alalunga) By Michael J. Thompson A THESIS submitted to Oregon State University in partial fulfillment of the requirements for the degree of Master of Science Presented February 4, 2005 Commencement June 2007
Master of Science thesis of Michael J. Thompson presented on February 4, 2005 APPROVED: Major Professor, representing Marine Resource Management Dean of College of Oceanic and Atmospheric Sciences Dean of Graduate School I understand that my thesis will become part of the permanent collection of Oregon State University libraries. My signature below authorizes release of my thesis to any reader upon request. Michael J. Thompson, Author
AKNOWLEDGEMENTS First and foremost I would like to thank my wonderful mother for putting up with me and always being there when I needed her. Without her support and love throughout my life, many of my accomplishments would have been much more difficult. My gratitude also goes out to my friends and advisors, Gil Sylvia and Michael Morrissey, who gave me the opportunity to achieve this lifetime goal. They were there to give me plenty of good advice (well sometimes?), support (definitely) and more work than I thought I could handle (always). Thanks also to OSU, COAS and the Marine Resource Management crew (in particular Jim Good, Laurie Jodice, Joy Burck, Ronda Bullis) and all the others for providing an excellent atmosphere and program. I greatly appreciate the help of my research assistant, and more importantly my friend, Sean for braving the open oceans and feeding many fish on our adventures ­ not to mention keeping me sane or insane depending on the need. Thanks also go out to the vessels, F/V Hans Halvor and F/V EZC, which were vital to accomplishing this research. I greatly appreciate the funding from the Community Seafood Initiative and the other members of this dynamic and enjoyable team (Diane, Michael and Dave). Recognition definitely goes out to all the people at the Coastal Oregon Marine Experiment Station and the Astoria Seafood Laboratory for their kindness and support. And last, but certainly not least, I would like to acknowledge all my friends that made it a pleasure throughout ­ especially for those who were so vital in destroying my house and keeping the fire burning. Without you this would not have been as enjoyable as it was. Thanks. THAT’S ALL FOLKS!
CONTRIBUTION OF AUTHORS Dr. Gilbert Sylvia and Dr. Michael Morrissey provided the funding and assisted in the initial re­write and final proof of the first paper in this thesis.
TABLE OF CONTENTS Page Introduction……………………………………………………………………... 1 Seafood Traceability in the United States: Current Trends, System Design, and Potential Applications……………………………………………………… 7 Introduction………………………………………………………………. 8 Legislation………………………………………………………………... 11 Traceability………………………………………………………………. 14 Software Solutions……………………………………………………….. 21 Discussion………………………………………………………………... 24 References………………………………………………………………... 26 Onboard Handling of Albacore Tuna (Thunnus alalunga): Bleeding Methods to Improve Quality in the Pacific Northwest Troll Fishery…………... 30 Introduction………………………………………………………………. 30 Materials and Methods…………………………………………………… 37 Sample Collection……………………...………………………….. 37 Sample Preparation…………...…………………………………… 42 Statistical Analysis…………………...……………………………. 46 Results……………………………………………………………………. 48 Discussion………………………………………………………………... 58 References………………………………………………………………... 67
TABLE OF CONTENTS (Continued) Page Design and Development of an Onboard Electronic Traceability System in the Eastern Pacific Albacore Tuna (Thunnus alalunga) Fishery……………. 72 Introduction…………………………………………………………....... 72 Materials and Methods………………………………………………….. 78 Results…..………………………………………………………………. 85 Discussion…………………………………………………….................. 90 References……………………………………………………….............. 99 Conclusions……………………………………………………………………. 101 Bibliography…………………………………………………………………… 104
LIST OF FIGURES Figure 2.1 Page Digital image of tuna loin with reference tile before cropping (A) and after (B)…………………………………………………………. 44 2003 Digital image visual analysis with medians (●) and standard deviations (in parentheses)…………………………………………... 49 Calibrated digital images of loins before color contouring (A,B) and after (C, D) with residual blood is shown in black………………….. 50 2003 Spectrophotometry results with median (●) and standard deviations (in parentheses)…………………………………………... 53 2004 Digital image analysis results with medians (●) and standard deviations (in parentheses)…………………………………………... 56 2004 Spectrophotometry results with medians (●) and standard deviations (in parentheses)…………………………………………... 57 3.1 Locations of albacore captured onboard the F/V Hans Halvor……… 86 3.2 Locations of albacore captured onboard the F/V EZC……………… 87 3.3 Locations of all 450 albacore captured during onboard trials……….. 88 3.4 Capture location (A) and traceability data collected (B) from one albacore (with extrapolated data underlined)………………………... 89
2.2 2.3 2.4 2.5 2.6 LIST OF TABLES Table Page 2.1 2003 Experimental variables and levels with procedure notation…… 38 2.2 Experimental groups with procedures in notation (see Table 2.1)…... 39 2.3 2003 Digital image analysis linear regression results……………….. 51 2.4 2003 Spectrophotometry analysis linear regression results…………. 54 2.5 2004 Digital image analysis linear regression results……………….. 56 2.6 2003 Spectrophotometry analysis linear regression results…………. 58 3.1 Traceability data collected on vessel and product activities………… 82 3.2 Traceability data collected on target and by­catch species………….. 83 3.3 Traceability data collected for cold storage and off­loading………… 85
Integrating Traceability with Onboard Handling to Enhance Product Quality and Marketability of Eastern Pacific Troll­caught Albacore Tuna (Thunnus alalunga) Introduction Many fisheries in the Pacific Northwest are experiencing frequent challenges, and are in a period of transition as regulatory, resource and market conditions rapidly change in the growing global economy. Concerns about resource sustainability, environmental health and new management practices are increasing and have many small seafood businesses and commercial fishermen apprehensive about their viability in the future. In addition, trepidations over food safety issues in the European Union (E.U.), and elsewhere, including the rising threat of bio­terrorism in the United States, have governments contemplating or enacting new legislation to improve food safety and consumer confidence in the food supply. The E.U., in response to several disastrous food safety issues, has recently enacted legislation requiring traceability documentation on numerous food items, including all seafood products. Traceability, defined in relation to the food industry, implies the ability to trace and follow feed, food and food­producing animals through all stages of production, processing and distribution (Food Standards Agency 2002). Although this may sound like inventory tracking, traceability is more than simply being able to locate a product. The fundamental basis for a traceability system is its ability to trace both products and the activities associated with their production (Moe 1998). The potential affects of mandatory traceability on the U.S. seafood industry, and the Pacific Northwest, are impossible to predict but one thing is readily apparent: the seafood industry is one commercial food sector in which traceability is quickly becoming a legal and commercial necessity (Borresen 2003). The Eastern Pacific albacore tuna (Thunnus alalunga) fishery is one example of a U.S. fishery confronting several major issues, including traceability, which presently finds itself in a period of transition. Albacore tuna are a highly migratory species that are accessible to the Northwest fishery for approximately six months a year, during the late spring and summer, along the Pacific Northwest coast. Currently
2 albacore stocks, which are unpredictable and hard to quantify due to their highly migratory nature, are not managed under NOAA Fisheries and are one of the few remaining open access fisheries within the U.S Exclusive Economic Zone ­ although this may change in the near future as albacore are now being considered for management under international agreements on straddling and highly migratory fish stocks. Open access makes this fishery extremely competitive, especially when fishing effort is easily reallocated to albacore from other marine fisheries where harvesting may be restricted or closed due to low stock estimates; a situation that is becoming more frequent along the U.S. Pacific coast. Additionally, this fishery has lost its traditional market as the large canneries, which once purchased the majority of albacore caught along the Pacific Coast, have relocated overseas. This loss has caused considerable financial hardships for many in the fleet and has made the industry aware of the need to diversify their market opportunities. One approach to reduce industry risk in the marketplace and compensate for lost earnings is to increase product quality in order to diversify products and markets (Sylvia and Peters 1992). Unfortunately, efforts to expand into new markets have met with limited success due to problems with product quality; a result of their past reliance on the cannery market, which typically accepted “lower” quality products, including albacore which were not bled. Product quality issues in albacore tuna are primarily related to time and temperature abuses; however, quality and appearance can be influenced by other factors as well, including onboard immobilization and bleeding techniques (Craven et al. 1997, Price and Melvin 1994). Quality loss is cumulative, increases with time, and cannot be reversed once it is initiated; therefore, it is critically important to begin quality preservation immediately after capture. Currently, many different methods of handling albacore are used and recognizing what onboard handling procedures are most effective in preserving quality is essential if the industry wants the ability to consistently produce high quality albacore products. To facilitate the development of new markets, which may require improved quality and quality assurances, the industry should: determine which onboard handling practices are most effective in preserving important quality attributes,
3 implement some form of industry wide quality guidelines or standards, and promote the important intrinsic, extrinsic and credence attributes of albacore tuna in the marketplace. Albacore tuna caught in the Pacific Northwest, which are not yet old enough to spawn, have many attributes which may be important to consumers, including minimal by­catch, low mercury concentrations (Morrissey et al. 2004) and high levels of omega­3 fatty acids (Wheeler and Morrissey 2002). One problem with promoting some of these attributes, however, is the industries ability, or lack thereof, to document and verify marketing claims about a product. Documenting handling procedures and storage parameters, in conjunction with other product attributes, are also vital in providing consumers with a high level of quality assurance on the products they purchase. With the possibility of traceability requirements in the near future, in view of the fact that a significant portion of recent harvests have been purchased for E.U.­ based canneries, the albacore industry, as well as other U.S. fisheries, may profit by being pro­active regarding traceability. Traceability systems which have the capacity to record capture and handling data onboard can provide sufficient documentation to meet traceability and Country­of­Origin requirements. Information collected by a traceability system can also be utilized by a seafood business to help improve product quality, streamline recall procedures, enhance supply­side management, and preserve desirable product attributes ­ all elements of traceability that can be used for marketing and creating a recognizable product brand. In addition, traceability information collected over time can be integrated with scientific information on oceanographic conditions, such as sea surface temperatures, to help improve fishery science and management. The Eastern Pacific albacore industry can improve quality by understanding the influence of onboard handling practices and implementing quality guidelines or standards designed to encourage the production of high quality products. Once quality is improved throughout the fleet the industry should concentrate on identifying and preserving important product attributes that can be beneficial when used for marketing purposes.
4 In the future, it is imperative that the industry diversifies into new product forms and develops alternative markets in order to maximize the economic benefits of the albacore fishery in the Eastern Pacific. This research attempts to address the legal and commercial ramifications of recently enacted traceability legislation and how traceability can be integrated into a comprehensive onboard system to improve quality and marketing. The specific objectives of this project were to: 1) provide the U.S. seafood industry with the latest information and research on traceability and traceability legislation in order to help them understand the implications of traceability and what benefits it may provide; 2) investigate what onboard handling and bleeding techniques are the most effective in reducing residual blood content in tuna muscle; and 3) design and develop an electronic traceability system integrating capture, handling and storage parameters which can be used to enhance product qualities, marketability and fishery science. To achieve these goals, comprehensive research was done to understand the concepts of traceability, which is a relatively new concept in the food industry, and how it can be achieved using today’s computer technologies. To investigate which onboard handling and bleeding methods are the most effective in removing blood, 32 different combinations of handling techniques currently employed in the tuna industry where analyzed for residual hemoglobin content using both spectrophotometry and computer­aided digital image analysis methods. The final part of the project involved designing an onboard computerized traceability system, able to withstand the harsh ocean environment, that is capable of recording and exporting both capture and handling data in an efficient and effective manner. The three manuscripts in this thesis provide an analysis and discussion of research, laboratory and field trial results. The first paper focuses on the concept of seafood traceability and includes research into current traceability legislation and computerized traceability systems. The second paper investigates different onboard handling techniques and their effectiveness in reducing the amount of residual blood in albacore muscle tissue. And the third paper, integrating what was learned in the first two papers, evaluates the design and field trials of an onboard computerized traceability system. The conclusions section integrates the findings of these three
5 manuscripts and discusses the implications that providing traceability and improved quality assurances to consumers may have on the Eastern Pacific albacore industry.
6 References Borresen, T. 2003. Traceability in the fishery chain to increase consumer confidence in fish products ­ application of molecular biology techniques. First Joint Trans­Atlantic Fisheries Technology Conference­ TAFT 2003. 11­14 June 2003, Reykjavik, Iceland. K4. Craven, C., Kolbe, E., Morrissey, M., and Sylvia, G. 1997. Onboard factors affecting chilling and freezing rates, and quality of albacore tuna. A Report to the American Fishermen’s Research Foundation. Oregon State University, November 1997. Food Standards Agency, Food Chain Strategy Division. 2002. Traceability in the food chain: A preliminary study [online]. Accessed on 15 November 2003. URL: http://www.foodstandards.gov.uk/multimedia/pdfs/traceabilityinthefoodchain. pdf Moe, T. 1998. Perspectives on traceability in food manufacture. Trends in Food Science and Technology. 9 (1998): 211­214. Morrissey, M.T., Rasmussen, R. and Okada, T. 2004. Mercury content in Pacific troll­caught albacore tuna (Thunnus alalunga). Journal of Aquatic Food Product Technology. 13(4):41­52. Price, R.J. and Melvin, E.F. 1994. Recommendations for on board handling of albacore tuna. University of California Cooperative Extension, Sea Grant Extension Program publication, UCSGEP 94­1. July 1994. Sylvia, G. and Peters, G. 1992. Product characteristics and market demand for Pacific whiting. In the Proceedings for the Pacific Whiting Harvesting, Processing and Quality Assurance Workshop, March 30­31, Newport, Oregon. Edited by Gil Sylvia and Michael Morrissey. pp 82­86. Wheeler, S. and Morrissey, M.T. 2003. Quantification and distribution of lipid, moisture and fatty acids in West Coast albacore tuna ( Thunnus alalunga ). Journal of Aquatic Food Product Technology. 12(2):3­16.
7 Seafood Traceability in the United States: Current Trends, System Design, and Potential Applications. by Michael Thompson, Gilbert Sylvia and Michael Morrissey Comprehensive Reviews in Food Science and Food Safety [online] 525 W. Van Buren, Ste. 1000 Chicago, IL 60607 January 2005 – Vol. 4, Issue 1.
8 Seafood Traceability in the United States: Current Trends, System Design, and Potential Applications. Introduction “Traceability” is defined as the ability to trace the history, application or location of that which is under consideration (ISO 9000:2000). The concept of tracing products from their origin to the consumer is not a contemporary idea. Many industries have incorporated product tracing into their internal operations for decades. Most of us have purchased items, from cars to electronics, which are labeled with unique serial numbers, allowing manufacturers and government authorities to identify and locate individual products. However, the introduction of traceability into the food supply sector is a relatively new concept which continues to gain momentum, particularly in the European Community. The seafood industry is a commercial food sector in which traceability is becoming a legal and commercial necessity (Borresen 2003). Globalization of trade and the lack of international standards have made identifying the origin and history of seafood products difficult, raising concerns from retail, food service, and consumers about the safety of their seafood supplies. These concerns have recently been heightened by the food safety problems experienced in Europe that have made traceability a prominent topic in the food industry. Driven largely by growing food safety issues, including bio­terrorism (Bledsoe and Rasco 2002), and demands by the consumer for detailed information on the nature, origin and quality of the food they are purchasing, traceability will make an impact on the seafood industry. Whether this impact is perceived as positive or negative within the seafood industry will depend on the potential market benefits and the design, management, and marketing of traceability concepts (Thompson et al. 2003). Within the food industry, traceability implies the ability to trace and follow feed, food and food­producing animals through all stages of production, processing and distribution (Food Standards Agency 2002). The fundamental basis for a traceability system is its ability to trace both products and activities (Moe 1998). This
9 requires a system capable of: 1) tracing products through the distribution chain, 2) providing information on product ingredients, and 3) understanding and communicating the effects of production practices and distribution on product quality and safety. While traceability by itself does not provide quality assurance, it has important aspects that relate to food safety, quality and product labeling (Kim et al. 1995). An effective traceability system also provides for an efficient flow of information through the entire market channel. Limited traceability is not new to the U.S. food industry, particularly with respect to food safety. Mandatory procedures have been established to reject or recall products that present a food safety issue. Good manufacturing practices (GMP), ISO 9000 quality management, and hazard analysis and critical control point (HACCP) procedures are growing in use and broadening the scope of traceability in accommodating this information (Moe 1998). Inspection and data systems such as HACCP, which is mandatory for all seafood, are designed to control biological, chemical, and physical hazards during processing. HACCP, however, does not require a traceability system since most of the collected data is not communicated to other market channel members in the supply chain (Hernandez 2001). Currently, product recall procedures and mandated documentation are the only form of product traceability in the U.S. Despite the apparent success of this system in preventing food borne illness, it still costs the U.S. food industry approximately seven billion dollars in 2000 for food borne disease (Economic Research Service 2003). The costs of product recalls associated with contamination are also increasing in the U.S. Class I product recalls, which are considered high health risk cases, have grown from 24 cases per year and 1.5 million pounds of affected food during 1993 to 1996, to 41 cases per year and 24 million pounds between 1997 and 2000 (Ollinger and Ballenger 2003). These trends may indicate that current regulations concerning food safety may be inadequate and that traceability could be an important strategy for reducing costs of food borne disease and product recalls, while also addressing consumer concerns over quality, the environment, and resource sustainability. Some firms have voluntarily begun to offer traceable products to their customers. Although largely limited to niche markets, these actions highlight the
10 growing demand, by food service, retailers, and consumers, for more information on food products. The use of informational labeling on food products is becoming a regulatory tool used to inform customers and influence markets for food quality (Caswell 1998). Currently, consumers are limited in the information about the origin and history of the food they purchase. The Commission of the European Communities believes that consumers have the right to receive information on the quality and constituents of their food so they can make informed decisions (Food Standards Agency 2002). Surveys have shown that a large majority of consumers both in the EU and U.S. were willing to pay a premium price for products which include Country­of­Origin­ Labeling (COOL) and geographical labeling and certifications (Roosen et al. 2003; Umberger et al. 2003; Clemens and Babcock 2002; Loureiro and McCluskey 2000; Wessels et al. 1999). Informational labeling requirements are likely to have a significant impact on the food market, helping to prevent fraud by providing more information to the consumer. Labeling, by itself, does not provide traceability, however, it is an important aspect of traceability which allows the physical tracking of the product and can be used as an effective means of differentiating products and creating brand recognition. Recent food safety concerns in Europe including Bovine Spongiform Encephalopathy (BSE), hoof and mouth disease, dioxin poisoning in chicken feed, and the growing anxiety over the proliferation of Genetically Modified Organism (GMO’s) foods, have increased attention in Europe, Japan, the U.S., and their trade partners on food traceability (Borresen 2003). In addition, the events of September 11, 2001 have highlighted the need to protect the national food supply from bio­ terrorism. Simply claiming that a product has been tampered with is sufficient enough to precipitate a full product recall, which may cost a firm, not only monetarily, but also its reputation (Bledsoe and Rasco 2002). Despite more stringent controls on food safety, confidence in the global food supply has continued to decline. Consumer confidence of food safety in the U.S. fell from 83% in 1996 to its current level of 74% (Economic Research Service 2002). This has resulted in increasing attention on traceability by policymakers in the U.S.
11 and other nations as a means to reduce uncertainty about food safety and to regain consumer confidence. Policy makers in the United Kingdom responded rapidly to the recent outbreak of BSE by enacting the Compulsory Beef Labeling Scheme (CBLS) in September of 2000. This law required that all producers of beef and beef products conform to a strict set of traceability and labeling guidelines. In January 2003, Japan passed legislation developed by the Ministry for Agriculture, Forestry and Fishery requiring domestic producers of beef to register all cattle into a centralized database. Other countries including Australia, New Zealand and Canada are considering or have implemented new traceability requirements for their meat products. Emerging mandatory requirements have made traceability an international trade issue which may strain relations and result in establishment of trade barriers. Traceability is also being debated within international forums. For example, during the most recent Codex Alimentarius committee meeting in December 2002, the U.S. strongly opposed the implementation of mandatory traceability. The U.S. argued that government should not be involved in the day­to­day operations of private industry, and should confine their role to issues of food safety. This debate is likely to continue as governments 1) attempt to increase consumer confidence in food safety, 2) counteract a heightened threat of bio­terrorism, and 3) confront a global economy increasingly influenced by consumer demands for more information on the origin and history of their food. The seafood industry, already confronted by inherent safety liabilities including scrombroid poisoning, ciguatera, shellfish poisoning, and mercury contamination, must address existing and emerging legislation and its effect on trade. In addition to food safety, concerns over declining fish populations and growing pressure from consumers to produce sustainable food will impact the role of food traceability in domestic and international markets. Legislation Although the specifics of a universal traceability system are yet to be defined, some nations have already passed legislation requiring comprehensive labeling and, in some instances, complete traceability of all food products. These regulations are
12 not only directed toward domestic products, but will be required on all imported products as well. Food safety is already a major priority in the U.S. and although there is currently no general legal requirement for the establishment of traceability in our food chain, existing law obliges any entity that chooses to sell or market food products to provide assurance that only safe food is sold. The following list includes new labeling and traceability standards that will affect the way that U.S. companies do business both here and abroad. European Union (EU), Article 4, regulation 104/2000 In effect since January 1, 2002, this regulation requires that all fishery products be labeled with commercial designation of the species, the production method (caught at sea, inland waters, or farmed), and if farm raised, the catch area or production location. European Community Commission Regulation 2065/2001, Article 8 This regulation pertains to detailed provisions for the application of EU regulation 104/2000 and requires that all chilled, frozen and smoked fish or fillets and shellfish, when offered for retail sale be labeled in accordance with EU 104/2000. In addition to these requirements, this information must be provided at each stage of the marketing chain either by direct labeling or acceptable commercial documentation. EU General Food Law Regulation 178/2002, Article 18 This article, which does not come into effect until January 2005, will require that traceability of food or food­producing animals or any other substance intended, or expected, to be incorporated into food or feed shall be established at all stages of production, processing, and distribution. It also requires the identification of supplier and customer for each market transaction, and provisions of all relevant documentation.
13 US Farm Security and Rural Investment Act of 2002 This act requires “Country­of­Origin” labeling on all beef, lamb, pork, fish, perishable commodities and peanuts. Voluntary guidelines were established in October of 2002 and will become mandatory September 30, 2004. US Bio­Terrorism and Response Act of 2002 Effective as of June 12, 2002 this law requires the registration of all food facilities, domestic and foreign, supplying food to the U.S. It addition it mandates records to identify the suppliers and recipients of all food products. The events which triggered this recent round of regulations will continue to have a dramatic effect on our U.S food industry (Sporeleder and Moss 2002). The EU has made the first move in requiring complete traceability on all fish and fish products by the year 2005, including imports. Although recent U.S. regulations do not mandate traceability, they contain some of the key concepts of traceability systems. Mandatory implementation of COOL will require all suppliers of food to the U.S. to clearly label the origin of all seafood products. Many sectors of the U.S. food industry support COOL legislation; however, there is also strong opposition to mandatory labeling. On January 27, 2004 President Bush signed public law 108­199 which effectively delays the implementation of mandatory COOL requirements for all covered commodities except wild and farmed raised fish and shellfish until September 30, 2006. Fish and shellfish suppliers will still be required to provide country­of­origin labeling on all products by September of 2004. Lobbying by both sides continues and debates are ongoing at this time whether or not mandatory labeling should be required and how it will be funded. A cost­benefit analysis of COOL, completed by the Agricultural Marketing Service, estimates the costs for the first year of implementation to be 3.9 billion dollars to create and maintain COOL information systems (Krissoff et al. 2004). A final ruling on COOL by the United States Department of Agriculture is expected soon, once all comments received are reviewed, the majority of which were related to the designation of wild or farm­raised fish and shellfish. In addition to COOL, the Bio­Terrorism Act clearly requires
14 certain traceability­related information including the registration of all food­related businesses with the U.S. government. However, this legislation does not specify, beyond the identity of the entity, what and how much information should be collected. There is increasing pressure to develop standardized traceability systems worldwide. During the 11 th session of the Codex Committee on Food Import and Export Inspection and Certification Systems (CCFICS) held in December 2002 in Adelaide, Australia, a discussion paper highlighted standardized elements to be included within a definition of traceability, and presented a framework for the future analysis of CCFICS texts with respect to product tracing (CCFICS 2003). The U.S., however, argues that this issue should only be addressed in Codex, as the global science­based international standards setting body for food safety, and not other international forums. These developments make it clear that traceability has become an international issue and is being discussed as a means to provide increased food safety and quality assurance to consumers. Traceability Traceability can essentially be described as a record keeping system designed to identify and track products from origin to consumption, while also providing the ability to quickly trace back products at any point in the food chain. The terms “product tracking” and “product tracing” have different meanings in the context of traceability. “Product tracking” refers to the recording of information as the product makes it way through the food chain, and the ability to identify in real time where the product is and what processes it has undergone. “Product tracing” refers to the ability to follow a product back through these processes from the consumer to their origin. Traceability appears to be a relatively simple concept; however, the actual process of creating an informational link between the origin of materials and their processing and distribution can be extremely complicated, especially given the quantity of food that makes its way into the global marketplace. Achieving traceability throughout the food supply chain requires the building of strong relationships in both directions along the food chain and a level of vertical
15 integration surpassing what is currently found within the industry. Integrating information flow between seafood companies in the U.S. and internationally will be complicated by the diversity within the industry, and the requirement of greater transparency. Although challenging, vertically integrating can result in more cooperative relationships, greater efficiency, and longer term market success by increasing consumer knowledge and satisfying their need for safe and quality products (Todd 2000). Information and knowledge management, which is enhanced by traceability, can help firms respond rapidly to internal challenges and external market opportunities (Peterson 2002). Two important motives for the formation and coordination of information in vertical supply chains is to manage liability associated with adulteration or contamination, and identifying and preserving quality traits (Westgren 1999). Traceability systems can be defined by these motives. Segregation systems attempt to separate batches of food and ingredients from each other during processing, while identity preservation systems identify the source and nature of each batch, requiring considerably information to guarantee that the traits and qualities of the product are maintained throughout the supply chain (Golan et al. 2002). The type of system to be used will depend on what the producer wants to accomplish and how much information they want to make available to other firms in the supply chain. Information on products and production practices must remain in the control of the entity responsible for these processes. Arrangements will need to be made between individual companies on how and what information will be shared to protect confidentiality and limit access to only legally entitled entities. Knowledge is a vital asset of all companies and the dissemination of proprietary knowledge will always remain an issue in traceability; thus knowledge based integration will always, by the nature of the free market, be restricted. Traceability Concepts The Food Standards Agency of the European Community recognizes two levels of traceability within the food industry. The first level, referred to as “internal
16 traceability,” takes place within one link of the chain (Moe 1998; FSA 2002). Considerable internal traceability already exists within the food industry providing individual firms the ability to track product through their internal operations; however, only very limited information actually follows the product to the next step (Golan et al. 2003). The real difficulty in designing and implementing a traceability system lies within the complexity of the next level, referred to as chain traceability (Moe 1998; FSA 2002). Chain traceability, which provides traceability between individual entities throughout the entire food chain, cannot be achieved without considerable knowledge­based vertical integration and may entail any number of entities in the seafood industry including fishers, buyers, processor, wholesalers, transporters and retailers. Achieving chain traceability requires comprehensive planning during the initial stages of development, particularly when addressing the three issues most crucial to the success of any traceability system: 1) compatibility, 2) data standardization, and 3) the definition of a Traceable Resource Unit (TRU) (Kim et al. 1995). Defining a TRU may be one of the most difficult steps involved in the design of a traceability system. A TRU is simply defined as a unit of trade, such as a whole fish or a batch of fish at the initial stage. However, this will invariably change during processing as new TRU’s are being assigned at each step within the food chain. The initial TRU must follow each fish or lot, through all steps of processing, distribution and retail. This process can become very complicated, especially during processing and it may be difficult to keep from mixing fish from several batches, especially when processing may include portioning, additional ingredients, processes, storage and transportation. Mixing of batches can occur between resource units, which may cause problems in identifying individual batches. Each firm must develop a system of assigning new TRU’s during processing, distribution and retail. Compatibility is the first component in a successful tracing system; it must be possible to trace products from one entity to another. This requires that all entities within the chain are able to communicate and transmit data efficiently. Having the ability to transmit and receive data does not, in itself, ensure traceability, it only provides a means. Rapid advances in information technology (IT) and increased compatibility between available operating systems have provided the necessary tools
17 to improve knowledge­based vertical integration. Standardized data transmission protocols and new computer applications are available with the ability to upload and download data between different operating systems and databases. Once compatibility has been established, data requirements must be identified and standards implemented. For traceability to work on a national or international level, standard data transmission protocols must be established. Without industry standards, close­knit supply chains may be able to integrate knowledge based operations, but more diverse and extensive food chains may find it difficult to implement traceability without carefully selected set of cost effective minimum standards (Wagner and Glassheim 2002). Standardization of data requires identifying which parameters during handling, processing and storage are important in preserving the identity of the product and its quality attributes. Once these parameters have been determined, standardized data formats are established at each step within the chain. Standardizing the content and quantity of information to be transmitted alleviates problems that may arise due to inconsistencies in data transmission protocols. The desired degree of detailed information will invariably change according to the purpose and entity (Moe 1998). Some firms may require more information than others. If an entity is conducting business with several firms, each with different data requirements, this can ultimately lead to considerable confusion and inefficiencies. In addition to data requirements within a sector, requirements will differ between sectors of the industry; processors will require information that may differ in content and quantity from that required by retailers. Another complex factor is the addition of new product information which occurs as the product moves through the food chain. Products may undergo additional handling and processes, including transformation, value addition, packaging, transport, and storage. Hernandez (2001) conducted a study of quality management and traceability in a fish processing facility and concluded that the key to complete traceability lies in the ability to follow products accurately through both mixing and transformation. The amount of information that becomes available for a given product may be significant. When one multiplies this information by the quantity of products produced daily, it becomes easy to understand how traceability systems may become too expansive, complex, and inefficient. For
18 this reason it is vital that any chain traceability system set minimum data requirements to constrain the number of variables that must be recorded for transmission to the next entity to only those critically necessary for identification, quality and safety purposes. Supplementary information may also be collected at any step within the food chain to provide data for analysis and optimization of production practices. In addition, it is necessary to establish limits on the length of time this information must be available to the food chain, government, or consumers. In an attempt to address these complex issues the European Commission funded a program from 2000 to 2002 entitled “Traceability of Fish Products,” or TraceFish, a consortium made up of 24 companies and institutes including representatives from exporters, processors, importers and research institutes (www.Tracefish.org). The goal of Tracefish was to identify informational requirements for chain traceability and formulate voluntary industry standards for the electronic collection and dissemination of traceability data. This program led to the development of standards for both captured fish and farmed fish chains as well as establishing standards for data transmission protocols. In their report they attempted to identify all variables that can be recorded at each step and divided these into three categories: information that “shall be recorded” (required), information that “should be recorded” (preferred) and information that “may be recorded” (optional). These voluntary standards may form the basis for Europe’s 2005 mandatory traceability requirements and have recently been adopted by the European Committee for Standardization (CEN 2002), an organization designed to promote voluntary technical harmonization within Europe. Despite the development of these standards a complete system for the collection and transmission of traceability data, including software to meet these standards, was not created by the Tracefish consortium. However, a traceability system has already been developed for the Danish fresh fish chain (Frederiksen et al. 2002), which was in development prior to the Tracefish project. This research focused on all aspects of the fresh fish chain, by utilizing barcodes and serial shipping container codes, to identify each resource unit and track each delivery. This research was successful in showing that traceability could be achieved and recognized the fact that system costs for vessels and small firms need to be addressed
19 and more friendly user­interfaces must be developed in order to promote efficiency. In addition, trials on traceability systems are currently being conducted in Japan (Hashimoto et al. 2003) and the Shetland Islands in Scotland, which are moving towards chain traceability by installing systems on ten vessels as part of Seafood Scotland project (MacDubhghaill 2000). Once the required information is defined for each link of the chain, a decision must be made on how this information will be transmitted and archived. The system must be developed to insure that each individual in the chain receives the data they require. If additional data is required by an entity further down the chain it must be determined if the information is transferred through the previous entity or is acquired from the source. The amount of information can be enormous, especially as one travels further along the food chain, where companies may have hundreds of different vendors. For transparency this information should be stored in databases that are easily accessible by entities within the food chain, including additional information desired by consumers and required by governmental organizations. These databases, owned by the producers of the information, must be designed to allow access, possibly by passwords, to those who have a right to the information. This method may be the most desirable for those firms which want to have more control over who has access to their information but may also become extremely difficult to manage. Another option is the use of a centralized database where all the traceability information is stored and disseminated. Access would need to be controlled if any propriety information is stored on the database; however, this would allow firms to query information that they desire without receiving unnecessary information. This would help to streamline the system and would away responsibility from firms that would otherwise have to transmit data from a previous source forward. One serious concern is system cost. Individual databases would require each firm to design and maintain their own resources. In addition to suitable computer hardware and software, each firm would need to develop their own accessible websites for data dissemination. A centralized system would require that firms send relevant traceability documentation, when it becomes available, to the central database. This information would be available through a single website and would provide
20 simplified access to information. The major problem facing the implementation of a central system is deciding who will pay for the design, development, and maintenance of the system. Each system will need to be designed to take into consideration the quantity and content of the information as well as the accessibility of the system. Access must be controlled yet provide for the needs of the consumers. The system itself must be easy to use and navigate and computer or database problems cannot hamper the day­to­day operations of the industry. Once each entity defines its resource units it must decide on an appropriate labeling scheme. Most food manufacturers and retailers around the world already use Universal Product Code (UPC) labels to identify their products. UPC labels are recognized internationally and are based on standards set by the Uniform Code Council (UCC) of the U.S. and European Article Numbering (EAN) system. EAN­ UCC labels come in a variety of configurations depending on the informational needs of the firm. One­dimensional UPC labels are limited in the amount of information that they can carry (up to 50 bytes), but are the most widely used UPC labels. They generally contain information that only identifies the entity and the product type; however, there are a variety of one­dimensional labels that can be used to provide more information, including a unique product identification number. Two­ dimensional UPC labels are also available and are able to hold significantly more information (up to 3000 bytes). Both one­dimensional and two­dimension labels have the ability to pass on limited traceability information but this information is only accessible by manually scanning the label with a bar­code reader. Bar­codes can present some difficulty when attempting to read UPC labels in cold and wet environments which are common in the seafood industry. New technology is becoming available that may eventually make paper barcode labels and manual scanning obsolete. Radio Frequency Identification (RFID) tags are becoming more widely accepted. RFID tags have the ability to hold large amounts of data, up to several megabytes, and can be custom tailored to suit individual needs including time and temperature readings and tracking product movement. These tags come in two types: passive and active. Passive tags do not contain their own power source and are dependant on a signal from a RFID reader to
21 start downloading their information. Active tags come equipped with their own supply of power and actively send out radio signals that are received by RFID readers as they move into range, automatically downloading their information without need to wait for a signal. RFID tags are more expensive than UPC labels but have several additional features that make them appealing. They do not require manual scanning and hundreds of RFID tags can downloaded into a computer at one time. One potential liability is that humans cannot read RFID tags without the use of machines, making the reliability of an RFID system of utmost importance. The ability to read numerous tags at once will save considerable time and man­hours as well as decreasing the number of errors that can occur with manual systems. RFID tags are currently being used in the National Livestock Identification Scheme of Australia to track cattle and other livestock and are also becoming more widely used in the EU meat industry. One factor complicating the use of RFID technology is compatibility as they are available in four different frequencies and there is no standardization. This discrepancy is currently being addressed by the UCC­EAN partnership, which is in the process of making an international standard for RFID technology called the Electronic Product Code (EPC) (EAN International 2003). RFID tags can be re­used which can reduce overall costs. They can be attached to a box of products, track those products until they are removed from the container, and once the container is returned, the tag can be reprogrammed and reused. Software Solutions As traceability gains momentum throughout the global food chain it has created economic incentive for computer software designers to develop software capable of tracking seafood from “fish to dish.” Advances in IT have made knowledge management and data transfer affordable, reliable, and efficient. Coupled with rapid advances in data capture devices and secure data transmission, traceability is attainable, not only for large firms, but also for small and mid­size enterprises. One major problem confronting traceability in the seafood industry isn’t technology related, but rather the heterogeneous nature of the fishing industry. Traceability
22 within a major corporation that encompasses all aspects of the supply chain, from the fisher through processing and transportation, can achieve traceability by incorporating Enterprise Resource Planning (ERP). ERP involves the integration of all functions of each department of a company into a single computer system and program. It requires that the needs of each department, from finance to processing, be incorporated into a single database, making information sharing and communication seamless, which enhances the ability of the company to achieve internal and chain traceability. ERP solutions are available for both small and large companies but require that each system utilize the same software package. This may be impossible, however, for many small businesses operating with a myriad assortment of hardware and software systems, which complicates vertical integration and makes true ERP solutions more difficult. The problem of data transmission between different operating systems and programs has led to the development of protocols and programs to facilitate communication. Electronic Data Interchange (EDI) allows computers to exchange information in a standardized format facilitating communication between entities. The exchange of data through EDI formats may require some manipulation for individual uses but has become a growing application for data transfer and e­ commerce. Data exchange can also be accomplished by utilizing programs capable of Open DataBase Connectivity (ODBC). Developed by the Microsoft Corporation, ODBC applications are designed to make it possible to access data from any application, regardless of which database management system (DBMS) is operating. OBDC is an open standard application programming interface (API) which is based on standard Structured Query Language (SQL) and can be used to access files in a number of different database formats including Access, Excel, dBase, Lotus, Oracle, FoxPro, SQL servers and Text. ODBC is accomplished through the use of ODBC driver, which is capable of translating data queries received from a sending application into SQL requests that are converted by the receiving system into requests recognized by its DBMS. Both the sending and receiving systems must have programs that are ODBC compliant and are equipped with an ODBC driver. ODBC versions also exist for UNIX, OS/2 and Macintosh platforms.
23 Rapid advances in IT and database connectivity has made it possible to vertically integrate the knowledge­based operations of various entities involved in the seafood industry. A large majority of small businesses utilize Microsoft windows operating systems and many of the programs, which advertise complete traceability, have been designed to operate in a Windows environment. Many software packages designed for traceability offer companies, both small and large, management tools in order to integrate their business operations. Modular options and different levels of customization allows small businesses the ability to design systems for their specific management needs while offering larger firms the ability to design completely integrated ERP solutions. Incorporating purchasing, sales, accounts receivable, accounts payable, general ledger, and inventory control into a single accessible database offers businesses both flexibility and control. ERP solutions for the seafood industry also have the ability to integrate with electronic data capture devices (including hand­held computers), various devices including bar­code scanners and printers, RFID devices, electronic scales, fish grading devices, and data loggers for time and temperature control. Scanvaegt and Marel, manufacturers of fish processing machinery, have recently incorporated traceability software that seamlessly integrates their line of processing equipment offering turnkey solutions to seafood processors wanting to provide traceability. ERP software is primarily designed to track products in lots or sub­lots through food processing including tracking of additional ingredients, portioning, and transformation. However, the recent push for traceability and subsequent legislation have made several software suppliers to the fishing industry including Wisefish and C­trace, incorporate traceability into vessel software, which either integrates with their land­based solutions or with ODBC compatible systems, allowing the exporting and importing of data between the various ODBC spreadsheet formats. Immediately after harvest, data can be transmitted from the vessel to shore­based processors and buyers. This integration allows for traceability requirements to be met from catch through processing. Utilizing ODBC capabilities makes traceability possible without the advantage of using the same software. These features will help to simplify integration; however, additional steps may be needed to ensure seamless data transfer
24 and incorporation of this data into inventory modules and customer accessible databases. Discussion It is clear that traceability will have an impact on the U.S. seafood industry. Governments around the globe are preparing or enacting legislation requiring seafood traceability. The reason for this seems clear: traceability is a policy designed to increase consumer confidence in the food supply. This decline in consumer confidence can mainly be attributed to global food safety issues, including the BSE crisis and bio­terrorism and recent concerns about mercury in seafood. These safety and liability concerns will encourage the development of effective traceability systems with time and temperature monitoring to insure that quality parameters are maintained (Merlmelstein 2002). The potential benefits of traceability on the seafood industry will depend on the design and development of cost effective IT based traceability systems, the willingness of consumers to pay for improvements in access to information about quality, history and origin of their seafood, its’ ability to improve efficiencies in supply chain management and provide brand protection. A well­designed traceability system may benefit many in the seafood industry. Dickinson and Bailey (2002) suggest that traceability could become a valued public good, especially for food safety. Companies that have accredited and verifiable traceability systems may be able to command favorable premiums from their insurance providers by reducing liability in food borne illness cases (Gledhill 2002). By preserving the identity of favorable attributes throughout the entire food chain, seafood producers can provide quality assurance securing the firm’s reputation (Unnevehr et al. 1999). Traceability may also provide information to the consumers about the sustainability of the resource and whether it is harvested in accordance with national and international management. These types of consumer concerns, which are not related to food safety, are becoming more important in domestic and global seafood markets.
25 American companies that export seafood to the European Community will have to implement traceability systems that meet E.U. requirements by the year 2005. The U.S. government seems unlikely to implement mandatory traceability requirements for the U.S. seafood industry. It is apparent, however, that traceability is gaining ground internationally as a food safety issue. This makes it important that industry stays abreast of the current developments and takes a proactive stance in the design and development of traceability systems which can meet the requirements of importers. The costs associated with these systems have not yet been determined. In the absence of mandatory traceability and in the case of seafood producers that do not export, the implementation of traceability systems will depend on the needs of the individual firm and the benefits that may be gained relative to potentially significant costs. .
26 References Bailey, D.V., Jones, E. and Dickinson, D.L. 2002. Knowledge management and comparative international strategies on vertical information flow in the global food system. American Journal of Agricultural Economics. 84(5):1337­1344. Bledsoe, G.E. and Rasco, B.A. 2002. Addressing the risk of bio­terrorism in food production. Food Technology. 56(2): 43­47. Borresen, T. 2003. Traceability in the fishery chain to increase consumer confidence in fish products ­ application of molecular biology techniques. First Joint Trans­Atlantic Fisheries Technology Conference­ TAFT 2003. 11­14 June 2003, Reykjavik, Iceland. K4. Caporale, V., Giovannini, A., Franscesco, D. and Calistri, P. 2001. Importance of the traceability of animals and animal products in epidemiology. [online] Accessed on 3 November 2003. URL: http://www.ole.int/eng/publicat/rt/2002/caporale.pdf Caswell, J.A. 1998. How labeling of safety and process attributes affect markets for food. Agricultural and Resource Economics Review. October 1998: 151­158. CEN Workshop Agreement. 2002. Traceability of fishery products­ specifications of the information to be recorded in captured fish distribution chains. CEN, 2002. Accessed 5 May 2003. URL: http://193.156.107.66/ff/po/EUTrace/WGCaptured/WGC_StandardFinal.doc Clapp, S. 2002. A brief history of traceability. Food Traceability Report: Response for Institute of Food Technologists Annual Meeting, Anaheim, California. 2002. Accessed 12 July 2003. URL: http://ift.confex.com/ift/responses/2002/16.doc Clemens, R. and Babcock, B.A. 2002. Meat traceability: Its effect on trade. Iowa Ag Review. 8(1): 8­9. CCFICS (Codex Committee on Food Import and Export Inspection and Certification Systems). 2003. Discussion paper on traceability / product tracing in the context of food import and export inspection and certification systems. Accessed on 17 November 2003. URL: http://www.bag.admin.ch/verbrau/Codex/CCFICS%20WG%20­ %20Discussion%20Paper.pdf Committee on Resources and the Committee on Commerce, Science and Transportation. 1997. Proposed implementation of a fishing vessel registration and fisheries information system [online]. Accessed on 21 May 2003. URL: http://www.nmfs.noaa.gov/sfa/401.pdf
27 Department of Agriculture, Fisheries and Forrestry, Australia. 2000. From plate to paddock: Turning the tables. Consumer driven demands on global food chains and the implication for Australia [online]. Accessed on 11 September 2003. URL: http://www.affa.gov.au/content/publications.cfm?ObjectID=E1975734­ DD8C­4277­9F389FD4E9E26BB3 Dickinson, D.L. and Bailey, D.V. 2002. Meat traceability: Are U.S. consumers willing to pay for it? Journal of Agricultural and Resource Economics. 27(2): 348­364. EAN International. 2003. EPCglobal [online]. Accessed on 15 March 2004. URL: http://www.ean­int.org/index800.html Economic Research Service, U.S. Department of Agriculture. 2002. Consumer food safety behavior: Consumer concerns [online]. Accessed 3 November 2003. URL: http://www.ers.usda.gov/Briefing/ConsumerFoodSafety/ Economic Research Service, U.S. Department of Agriculture. 2003. Economics of foodborne disease [online]. Accessed 3 November 2003. URL: http://www.ers.usda.gov/Briefing/FoodborneDisease/ Food Standards Agency, Food Chain Strategy Division. 2002. Traceability in the food chain: A preliminary study [online]. Accessed on 15 November 2003. URL: http://www.foodstandards.gov.uk/multimedia/pdfs/traceabilityinthefoodchain. pdf Frederiksen, M., Osterberg, C., Silberg, S., Larsen, E. and Bremner, A. 2002. Info­ Fisk. Development and validation of an internet based traceability system in a Danish domestic fish chain. Journal of Aquatic Food Product Technology. 11(2): 13­34. Gledhill, J. 2002. Tracing the line. Special to Food Processing, March 6. 2002. Accessed 3 November 2003. URL: http://www.foodprocessing.com/Web_First/FP.nsf/ArticleID/LKIE­57YN2X Golan, E., Krissoff, B. and Kuchler, F. 2002. Traceability for food marketing and food safety: What’s the next step? Agricultural Outlook. Jan­Feb 2002: 21­25. Golan, E., Krissoff, B., Kuchler, F., Nelson, K., Price, G. and Calvin, L. 2003. Traceability in the US food supply: Dead end or superhighway? Choices. Second Quarter 2003: 17­20.
28 Hashimoto, T., Tanaka, K. and Niwa, H. 2003. Trial of farmed fish traceability in Japan. First Joint Trans­Atlantic Fisheries technology Conference­ TAFT 2003. 11­14 June 2003, Reykjavik, Iceland. Hernandez, M.R.P. 2001. Study of the quality management system and product traceability in a fish processing company. United Nations University, Fisheries Training Programme, Final Project 2001. Accessed 11 September 2003. URL: http://www.unuftp.is/proj01/MariaRitaPRF.pdf Hooker, N.H., Nayga, R.M. Jr. and Siebert, J.W. 1999. Preserving and communicating food safety gains. American Journal of Agricultural Economics. 81(5): 1102­1106. ISO 8402. 1994. International Standards. Quality management and quality assurance: vocabulary. URL: http://home.netvigator.com/~timwwh01/ISO­ 8402.HTM. Kim, H.M., Fox, M.S. and Gruninger, M. 1995. An ontology of quality for enterprise modeling. In Proceedings of WET­ICE, Los Albamitos, CA, USA. Pp. 105­ 116, IEEE Krissoff, B., Kuchler, F., Nelson, K., Perry, J. and Somwaru, A. 2004. Country­of­ Origin labeling: Theory and observations. Outlook, Economic Research Service, United States Department of Agriculture. WRS­04­02, January 2004. Accessed 21 June 2004. URL: http:\\www.ers.usda.gov/publications/WRS04/jan04/wrs0402/ Loureiro M.L. and McCluskey, J.J. 2000. Assessing consumer response to protected geographical identification labeling. Agribusiness. 16(3): 309­320. MacDubhghaill, U. 2000. The outlook for the UK fishing industry [online]. Accessed 8 February 2003. URL: http://www.intrafish.com/intrafish­ analysis/UK_2000_48_eng/ Mermelstein, N.H. 2002. A look into the future of food science and technology. Food Technology. 56(1): 48­55. Moe, T. 1998. Perspectives on traceability in food manufacture. Trends in Food Science and Technology. 9 (1998): 211­214. Ollinger, M. and Ballenger, N. 2003. Weighing incentives for food safety in meat and poultry [online]. Amber Waves, April 2003 1(2):34­41. Accessed 11 November 2003. URL: http://www.ers.usda.gov/AmberWaves/April03/Features/WeighingIncentives. htm
29 Palsson, P.G., Storoy, J., Frederiksen, M. and Olsen, P. 2000. Traceability and electronic transmission of qualitative data for fish products. Icelandic Fisheries Laboratory, 2001. Accessed 11 Sept. 2003. URL: http://www.norden.org/fisk/sk/sluttrapport_Traceability.pdf Peterson, H.C. 2002. The “learning” supply chain: Pipeline or dream? American Journal of Agricultural Economics. 84(5): 1329­1336. Roosen, J., Lusk, J.L. and Fox, J.A. 2003. Consumer demand for and attitudes toward alternative beef labeling strategies in France, Germany and the UK. Agribusiness. 19(1): 77­90. Sporleder, T.L. and Moss, L.E. 2002. Knowledge management in the global food system: Network embeddedness and social capital. American Journal of Agricultural Economics. 84(5): 1345­1352. Tall, A. 2001. Traceability procedures based on FDA and CFIA regulations­ an understanding. INFOFISH International, 5: 49­51. Thompson, M., Sylvia, G., and Morrissey, M.T. 2003. Seafood traceability in the U.S. Presented at the annual meeting of the Institute of Food Technologists, July 12­16, Chicago, IL. Umberger, W.J., Feuz, D.M., Calkins, C.R. and Sitz, B.M. 2003. Country­of­Origin labeling of beef products: U.S. consumers’ perceptions. Presented at 2003 FAMPS Conference, Washington, D.C. 20­21 March 2003. Accessed 16 July 2003. URL: http://www.competitivemarkets.com/whats_new/2003/4­7.pdf Unnevehr, L.J., Miller, G.Y. and Gomez, M.I. 1999. Ensuring food safety and quality in farm­level production: Emerging lessons from the pork industry. American Journal of Agricultural Economics. 81(5): 1096­1101. Wagner, G.L. and Glassheim, E. 2002. Traceability of agricultural products [online]. Accessed on 11 September 2003. URL: http:// www.ngplains.org/ArticleView~article_id~50~s_article_category_id~2~.asp Wessells, C.R., Johnston, R.J. and Donath, H. 1999. Assessing consumer preference for ecolabeled seafood: The influence of species, certifier, and household attributes. American Journal of Agricultural Economics. 81(5): 1084­1089. Westgren, R.E. 1999. Delivering food safety, food quality, and sustainable production practices: The label rouge poultry system in France. American Journal of Agricultural Economics. 81(5): 1107­1111.
30 On­board Handling of Albacore Tuna (Thunnus alalunga): Bleeding Methods to Improve Quality in the Pacific Northwest Troll Fishery Introduction The albacore tuna (Thunnus alalunga) troll fishery conducted off the West Coast of the United States is currently facing a period of transition. Traditionally, catches of albacore tuna have been destined for cannery markets focused more on quantity then quality. Closures of the major canneries on the West Coast in the 1980’s, and their subsequent relocations overseas, in combination with the development of the Eastern Pacific long­line albacore fleets have left many of the West Coast albacore fishermen without adequate markets. Markets that currently exist lack sufficient competition and have led to some of the lowest prices in the history of the fishery creating a need to expand into new, non­traditional markets. Absence of a common strategy within the industry, which is largely comprised of independent family operated vessels, and inconsistent quality have hampered the development and growth of potential new markets. By the late 1990’s, however, there has been an increased interest in albacore caught in the Pacific Northwest. Media coverage of high mercury levels in canned albacore, primarily from older and larger albacore caught by long­line fisheries in the Eastern Pacific, has negatively impacted the entire fishery. New research now indicates that smaller, two to five year old albacore, targeted by the U.S. trolling fleet have mercury concentrations well below the national safety guidelines established by the Food and Drug Administration (Morrissey et al. 2004). Recent studies have also revealed high concentrations of healthy unsaturated fatty acids, or omega 3’s, in albacore caught along the Pacific West Coast (Wheeler and Morrissey 2002). These findings, in conjunction with a fishery that is believed to be comprised of healthy stocks by the National Marine Fishery service and an extremely low by­catch rate, have contributed to renewed interest in improving the quality and marketability of troll caught albacore tuna.
31 In order to exploit potential national and international markets for high quality fish on­board handling techniques for the albacore tuna industry must be improved and to some extent standardized. Seafood quality has many dimensions and is made up of both intrinsic and extrinsic characteristics including appearance, taste, odor, texture and price. Characteristics which define a product are determined by the buyer’s knowledge, needs and the intended use of the product (Anderson and Anderson 1991). Traditionally the majority of the albacore fleet has neglected many of these factors, both intrinsic and extrinsic, which are not as discernable in a canned, low price product, instead choosing to concentrate primarily on quantity. Seafood processors, large canneries in particular, have not demanded improvements in quality, instead focusing on providing the consumers with inexpensive products which meet minimum industry standards. Opportunity currently exists in the albacore fishery to expand market opportunities; however, before this can be accomplished the albacore fishing industry needs to focus on instituting more comprehensive quality standards to improve the quality of their product. Albacore tuna, members of the Scrombridae family, are highly advance teleost’s possessing physiological and biological characteristics that make proper handling procedures on­board the vessels of vital importance. They are capable of thermoregulation by way of counter­current heat exchange, have advanced cardiovascular systems, distinctive enzyme systems and high metabolic rates; all of which make the time immediately after landing crucial in preserving quality throughout processing, storage, distribution and retail (Craven et al. 1997, Price et al. 1991, Price and Melvin 1994). In order to maintain the highest level of quality and shelf life proper handling must begin immediately after the fish is landed and methods should be put in place that effectively reduces internal temperatures within FDA guidelines and preserves the intrinsic and extrinsic characteristics that consumers demand. Tuna muscles consist primarily of two types, red muscle for extended activity (slow twitch) and white muscle for bursts of activity (fast twitch). Red muscle comprises only a small fraction of total body mass, approximately 6.5%, while white muscle takes up the majority of total mass, about 54%, accounting for 89% of the
32 entire muscle mass (Korsmeyer et al. 1996). Both muscle types have extensive circulatory systems and large capillary­muscle fiber interfaces with mitochondrial volumes similar to mammals, though, white muscle, which has larger capillary size also has about 5 times lower capillary density than red muscle (Mathieu­Costello et al. 1992, Mathieu­Costello et al. 1996). This vast capillary system contributes to high rates of aerobic activity in red muscles and allows for the rapid recovery from exhaustive exercise, including the quick removal of lactic acid and other waste products resulting from anaerobic metabolism in white muscles (Brill 1996). Tuna white muscle anaerobic capacity is significantly greater than that found in other ectothermic fishes indicating an ability to generate more energy for burst swimming (Dickson 1996). Secondary circulatory systems have also been identified in tuna muscles (Dewar et al. 1994, Brill et al. 1998) increasing capillary capacity and contributing to high blood volumes. Laurs et al. (1978) found albacore to have particularly high blood volumes well above those found in many other fish species (Duff et al. 1987, Gingerich and Pityer 1989, Gingerich et al. 1990, Tort et al. 1991). Increased blood volumes are believed to be physiological adaptations in tuna for increased metabolic rates (Brill 1996), blood pressures and circulatory times (Dewar et al. 1994) but may also be related to functions other than oxygen delivery including thermoregulation (Mathieu­Costello et al. 1992). Tuna also exhibit high blood pressure (Jones et al. 1986, Jones et al. 1993, Tibbits 1996), increased blood hemoglobin concentrations and elevated muscle myoglobin concentrations (Dickson 1996). Scrombroid fishes, specifically tuna species, are also susceptible to histamine formation. Histamine is produced in the presence of histamine­producing bacteria, which possess histidine decarboxylase, a chemical that converts histidine to histamine. Tuna have high levels of the protein histidine, vital component of their blood pH buffering capacity (Abe et al. 1985). The formation of histamine is largely induced by high temperature abuse and can be controlled by proper handling techniques including bleeding, rapid chilling and maintaining consistent cold temperature storage (Ben­Gigirey et al. 1999, Kim et al. 1999, Kim et al. 2000). These physiological adaptations of albacore tuna, which make this highly advanced fish so successful in surviving their yearly trans­pacific migration, also
33 require that significant care be taken on­board the vessel to ensure quality remains high, especially for higher­end niche markets. During periods of high activity, like that found during capture, tuna utilize their muscles high aerobic oxidation rates and extremely high glycolytic capacity, due to enhanced concentrations of glycolytic enzymes, to produce adenosine tri­phosphate (ATP) for energy (Bushnell and Jones 1994),. As oxygen becomes depleted anaerobic oxidation of intramuscular glycogen (Weber and Haman 1996) takes place within the white muscles, which also have a high capacity to produce ATP through these anaerobic pathways, producing unusually large amounts of lactate and intracellular and extracellular acidification (Dickson 1996). Under normal circumstances tuna are able to minimize both the lactate buildup and resultant acidification with high buffering capacity, principally histidine­based buffers (Abe et al.1985), which suggests that re­synthesis of lactate occurs within white muscle (Arthur et al. 1992), allowing for the simultaneous removal of lactate, restorations of intracellular pH to normal levels, and replenishment of glycogen stores within the muscles. Lactate levels in skipjack tuna after exhaustive exercise can reach normal levels in as little as two hours, which is comparable to mammals and as little as one­twelfth the time it takes rainbow trout (Arthur et al. 1992). Tuna white muscle fibers also have high lipid content, a primary source of energy in a species that has limited carbohydrate intake from prey items, using proportionally more proteins and lipids for energy than other, low­aerobic species (Weber and Haman 1996). All of these factors contribute to the degradation in quality that can occur, quite rapidly in tuna species, if proper on­board handling procedures are not followed. After capture and killing, these biochemical pathways are terminated, leaving waste products accumulated during struggle within the muscles and blood. Blood lactate levels have been shown to be significantly higher in fish after struggling on the line (Swift 1983). Decreases in muscle pH due to acidosis, primarily caused by accumulation of lactic acid, have also been observed in fish which have been stressed by capture (Nakayama et al.1996, Terayama and Yamanaka 2000, Wood et al. 1977). The accretion of metabolic by­products within tuna muscles and blood can severely affect the quality of the flesh if proper handling procedures are not performed
34 immediately after capture. Tuna also have high levels of hemoglobin and lipids in their blood which, when improperly handled, can lead to lysis of the erythrocytes causing oxidation of the lipids after death. Lipid oxidation is the main factor attributed to the development of rancidity and off odors (Richards and Hultin 2001) and is usually the first extrinsic factor checked when determining quality. The effects of various handling procedures have been investigated in the past on pH, rigor mortis, flesh quality and appearance, lipid oxidation, ATP concentrations, histamine production and hemoglobin auto­oxidation in order to provide methods of improving quality. Both killing and bleeding methods have been studied in relation to their effects on flesh quality and storage. Killing procedures which cause instant mortality, including spiking or destruction of the spinal bulb, can delay the on­set of rigor mortis by delaying post­mortem ATP consumption and degradation, which helps to preserve both intrinsic and extrinsic properties associated with quality (Ando et al. 1996, Mochizuki and Sato 1996, Mochizuki et al. 1998, Nakayama et al. 1996, Oka et al. 1990, Terayama and Yamanaka 2000). Fish which have their spinal bulbs destroyed, either by spiking or stunning, immediately after capture also exhibit higher pH values than those left to die by asphyxiation and maintain elevated values over time (Amano et al. 1953, Nakayama et al. 1996, Terayama and Yamanaka 2000). Spiking has also been shown to produce products with firmer flesh, lower k­values 1 (Mochizuki and Sato 1996), and higher sensory values then fish which were not killed immediately after capture (Oka et al. 1990, Terayama and Yamanaka 2000). The results of bleeding studies show similar results to those found for spiking. Many fish species, including tuna and other pelagics, which were bled had firmer flesh and higher quality attributes than those which were not bled. Tretsven and Patten (1981) found that trout which were bled by a tail incision, removing approximately 37% of blood volume, were of better quality than un­bled trout and that fish bled immediately after removal from the water showed greater blood loss than those where bleeding was delayed by 20 minutes after capture. Bleeding decreases the rate at which rigor­mortis progresses preserving the firmness of the 1 K­values are used as an objective quality indicator which measures ATP degradation. Low values indicate freshness and high values spoilage (Voigt and Botta 1990).
35 flesh by delaying collagen fibril degradation (Ando et al. 1999, Oka et al. 1990), lowers the observed k­values (Mochizuki et al. 1998), helps maintain higher pH values (Amano et al. 1953), decreases rancidity by reducing lipid oxidation (Richards and Hultin 2001, Tretsven and Patten 1981) and improves flesh appearance (Terayama and Yamanaka 2000). Although prior research on the effects of various handling methods on product quality traits do exist they are limited to the effects of bleeding versus not bleeding (Ando et al. 1999, Price et al. 1991, Richard and Hultin 2002, Terayama and Yamanaka 2000) and the effects of immobilization by spiking and/or spinal cord destruction versus no immobilization (Amano et al. 1953, Ando et al. 1996, Mochizuki and Sato 1996, Oka et al. 1990, Nakayama et al. 1996). There is currently a lack of research addressing the effectiveness of different bleeding methods and handling techniques employed as a on­board quality system. At present albacore fishermen utilize various combinations of bleeding and handling methods which depends solely on the discretion of the vessel operator. Industry wide standards for producing high quality albacore are currently not in place which makes producing consistent quality in sufficient quantity to meet the consumer demands challenging. Without consistency in quality the barriers that albacore fishermen have traditionally faced in developing high end niche markets will remain. The variety of techniques used on­board include several different bleeding methods (gill, throat and pectoral), immobilization methods (stunning and spiking), seawater cooling (spray and submersion), bleeding position (lay­down and head­down), and time of bleeding before storage (20 minutes or longer). Suggested on­board handling guidelines for albacore exist (Price and Melvin et al. 1994) but no scientific research has been conducted to determine which of these factors or combination of factors are the most effective in removing blood. Interviews with fishermen during the initial stages of this project raised some interesting questions about on­board handling methods and their effects on product quality. From this interview process it was apparent that fishermen were quite interested in determining which of the bleeding methods are the most effective, how long bleeding should proceed and whether or not placing the fish in a head­down position promotes more effective blood loss. Questions were also
36 raised about the affects immobilization may have on bleeding and whether spiking or stunning would inhibit effective bleeding by stopping the heart from pumping blood prematurely. Another question that was raised by numerous fishermen concerns the use of seawater spray or submersion and if it helps to increase blood loss by preventing coagulation of the blood. These questions along with prior scientific research were used to design this research project and led to the development of several hypotheses’ regarding the various handling techniques used in this study. These included: 1) no difference would be found in the amount of hemoglobin removed between the gill and throat bleeding methods; 2) utilization of the pectoral cut in conjunction with other bleeding methods would reduce the amount of residual blood visible on the external surfaces of the loins; 3) immobilization by spiking or stunning would not impede blood loss and; 4) that the amount of time necessary for complete bleeding would be lower than what fishermen typically employ, which is approximately one hour. Given the problems with of product quality facing the albacore tuna industry, this collaborative research was designed to evaluate how the different onboard handling techniques currently used in the industry affect the amount of blood loss in fish bled immediately after capture. Two methods of analysis will be used to determine blood loss. The first method uses digital photo color analysis to investigate whether the different handling techniques affect the amount of residual blood on the external surface of the loins. Residual blood on the loin surfaces can influence both product appearances, which is a primary product attribute, and recovery rates during processing. Although digital photo color analysis is a relatively new field of study, research has shown that it is an effective tool in determining colors objectively (Wallat et al. 2001). The second analysis involved the utilization of spectrophotometry methods to determine the amount of residual blood content internally within the muscle tissues. By investigating how much residual blood remains in the tissues we can compare the ability of the different handling methods to remove blood from the vessels and capillaries which supply the muscles with oxygen and nutrients. Specific research objectives include: 1) determining which of the various onboard handling methods are the most effective in removing blood from the
37 muscle tissues and 2) the external loin surfaces. Results are discussed within the context of improving albacore product quality and will be used in conjunction with other available research on killing and bleeding methods, chilling and storage to outline a series of on­board handling recommendations for the industry to improve the quality of albacore tuna products. Materials and Methods Sample Collection Summer 2003 Albacore tuna samples were collected in August of 2003 from the 24 th to the 29th on­board the F/V Heide. These samples were caught during regular fishing operations in waters off the coast of Oregon located between 43º27’ and 44º42’ north latitudes and 125º05’ and 125º45’ west longitudes. Ninety­nine samples were collected with three samples each for the thirty­two experimental procedures and one control group. Albacore tuna were chosen immediately after striking the fishing jig, before they were landed on­board, for inclusion in the experiments. They were landed during normal fishing operations, which sometimes meant that they were allowed to remain struggling in the water until fish on shorter lines, which have priority, were landed and processed. Internal temperatures were taken by inserting a Digi­Sense® type K electronic thermocouple (made by Cole­Palmer Instrument Company, Vernon Hills, Illinois) through the dorsal surface directly adjacent to the dorsal fin into the thickest region of the fish until the probe encountered the spinal column. This was done immediately after landing before any experimental procedures were conducted. Numbered tags were attached at this time to the peduncle of each fish for identification throughout the experiment. Five variables were investigated including bleeding method, immobilization techniques, bleeding position, seawater cooling and time parameters. Each variable had from two to four levels and are shown in table 2.1. These variables were chosen
38 after numerous conversations with fishermen to determine which handling techniques they employ onboard and to address some of their questions as to which methods are the most effective in removing blood and improving quality. Table 2.1. 2003 Experimental variables and levels with procedure notation. Experimental Variables Bleeding Method Immobilization Bleeding Position Seawater Cooling Time (minutes) Levels Gill Gill and Pectoral Throat Throat and Pectoral None Spiking Stunning Lay­down Head­down None Sprayed Submerged 20 30 40 50 Procedure Notation G GP TH TP N SP ST LD HD NC SPR SUB 20 30 40 50 In order to study these various techniques and their effects on bleeding thirty­ two experimental procedures (Table 2.2) were chosen from the 288 possible by Conjoint Designer© version 3, a computer program by Bretton­Clark (New York, New York). The experimental design was based on a linear model with no­replicate procedures and was a partial­factorial design that imposed orthogonality 2 between the attribute levels. Each experimental procedure represented a different combination of the 5 handling techniques and was assigned a condition number 3 of 7.32. This program was used primarily to reduce the amount of experimental procedures due to vessel time constraints while permitting adequate data for statistical analysis of the five factors and their various levels. However, due to the limited number of 2 3 An orthogonal design indicates that all inter­correlations are 0. Conditions numbers above 25 can create problems during analysis.
39 experimental groupings imposed by time constraints (32) interactive effects between variables could not be adequately estimated. Table 2.2. Experimental groups with procedures in notation (see Table 2.1). Experimental Group Control 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Experimental Procedure Control G­N­HD­SUB­40 G­N­LD­SPR­20 G­SP­HD­NC­30 G­SP­HD­SUB­20 G­SP­LD­NC­50 G­SP­LD­SPR­40 G­ST­HD­NC­50 G­ST­LD­NC­30 GP­N­HD­SPR­30 GP­N­LD­SUB­50 GP­SP­HD­NC­40 GP­SP­HD­SPR­50 GP­SP­LD­NC­20 GP­SP­LD­SUB­30 GP­ST­HD­NC­20 GP­ST­LD­NC­40 TH­N­HD­NC­50 TH­N­LD­NC­30 TH­SP­HD­NC­30 TH­SP­HD­SUB­20 TH­SP­LD­NC­50 TH­SP­LD­SPR­40 TH­ST­HD­SUB­40 TH­ST­LD­SPR­20 TP­N­HD­NC­20 TP­N­LD­NC­40 TP­SP­HD­NC­40 TP­SP­HD­SPR­50 TP­SP­LD­NC­20 TP­SP­LD­SUB­30 TP­ST­HD­SPR­30 TP­ST­LD­SUB­50 After landing and recording of the internal temperature each fish was measured for length and circumference then processed according to their experimental groupings. First they were immobilized, if required, either by stunning (ST) the fish
40 with a blow to the head slightly behind the mid­line of their eyes with a wooden club or by spiking (SP), which was done by placing a splicing awl at the mid­line between their eyes and inserting it at a 45° posterior angle destroying their spinal bulb. Both of these methods produced the same observed results, which included a cessation of movement except for a slight twitching of the tail, which subsided within a few seconds. After immobilization the fish were bled by one of the four bleeding methods; gill (G), gill and pectoral (GP), throat latch (TH) or throat latch and pectoral (TP). The gill cut was performed by lifting the operculum and cutting the aortic arteries, which come from the heart and supply blood to the gill arches, directly posterior to the gill arches. Throat latch bleeding involved the severing of the throat latch, about an inch behind the area where the throat latch meets the lower jaw. Pectoral cuts were made in conjunction with either the gill or throat latch methods by inserting a short triangular blade about one inch, or the width of two fingers, directly posterior to the pectoral fin on the lateral line of the fish. The knife was inserted approximately three quarters of an inch in depth severing the paired lateral cutaneous blood vessels directly under the skin. This method of bleeding was investigated to determine it’s effectiveness in removing blood from the cutaneous blood vessels, located directly under the skin, which are supplied by the paired lateral vessels. After bleeding, each albacore was placed in bleeding bins depending on which bleeding position they were assigned. Fish which were assigned lay­down (LD) position were placed in a large 30 gallon ice chest located on the deck of the vessel. Fish assigned the head­down (HD) position were placed with their heads down inside either eight inch or ten inch diameter PVC pipes, depending on the size of the fish. The PVC pipe sections, which were about forty inches long, were placed inside (one of each size) a thirty gallon rectangular plastic garbage can. Both the lay­down and head­down bleeding receptacles were equipped with drain plugs which could be removed depending on which sea water cooling methods was assigned. Cooling methods used ambient temperature sea water supplied by on­board pumps and consisted of either spray (SPR) or submersion (SUB) during bleeding. Sea water temperatures ranged from 16° C to 18.5° C and were measured several times a day, however, the constant movement by the vessel made it difficult to measure accurately
41 while fish were being bled. Spraying was accomplished by using garden sprinklers placed above the fish, within the bleeding receptacles, and the water was allowed to drain out the drain holes to prevent submersion. Sea water submersion was done by closing the drain holes and filling the bleeding receptacles prior to placement of the albacore within the receptacles. Water was allowed to run during the entire bleeding time for both methods with excess water in the submersion method pouring over the top of the bleeding receptacle. Fish were allowed to bleed for 20, 30, 40 or 50 minutes depending on the experimental groups and were timed by count­down timers with alarms. When bleeding was finished the fish were removed from the bleeding receptacles measured again for internal temperatures using the same location as the previous measurement. After recording internal temperature the samples were rinsed in sea water to remove excess blood, which was required by the vessel operator to prevent excess cleanup, and placed into the air­blast freezer set to maintain a temperature of ­28º C. The samples were offloaded on September 23 rd , 2003 in Ilwaco, Washington and transported directly to the Astoria Seafood lab where they were place in the ­28º C freezer until dissection. Albacore samples were photographed for color photo analysis first after which samples of muscle tissue were taken, vacuum packed in freezer bags and stored at ­28º C until spectrophotometry analysis was conducted. Summer 2004 In 2004 albacore tuna samples were collected on­board the F/V Hans Halvor on August 15 th and 16 th off the coast of Oregon. These samples were collected between 43º 24’ and 44º 21’ degrees north latitude and 125º 05’ and 125º 16’ degrees west longitude. Experimental procedures for this portion of the research were designed by using the results from the 2003 albacore bleeding study. The main purpose of this additional research was to further investigate the two most effective bleeding methods, which included the gill and throat latch, and determine if one is more effective at removing blood if the other factors (immobilization, positioning and cooling) remained constant. The only other variables that were investigates included
42 time parameters, which were reduced from the first experiment in order to establish if bleeding times less than 20 minutes would show any significant difference in blood loss, and the size of the fish as measured by length. Fish were collected and handled using the same techniques outlined for 2003, except internal temperatures were not recorded due to equipment malfunctions. Fish in this study were divided into six groups and a control group comprised of four fish each. The six groups included three fish from each bleeding method, gill and throat latch, at times of 5, 10 and 15 minutes. The control group consisted of un­bled fish which were landed and placed into storage after struggling ceased. All fish in the experimental procedures were measured for length and circumference then stunned by clubbing prior to bleeding. They were then placed head down in the bleeding receptacles without sea water cooling. Head­down positioning was used during this second analysis after it proved to be effective in 2003. Bleeding was accomplished by the same methods as 2003 samples, using either the gill or throat latch method, depending on their grouping. Immediately after bleeding they were rinsed with sea water to remove external blood residue and lowered into the hold set to maintained a temperature of ­28º C. They remained in the vessel’s hold until they were off­loaded on August 19 th , 2004. After off­loading in Ilwaco, Washington they were transported directly to the Oregon State University Astoria Seafood laboratory and placed into a ­ 28º C freezer until processing for color photo and spectrophotometry analysis. Fish were photographed for digital color analysis and then muscle tissue samples were removed and vacuum packed in freezer bags for spectrophotometry analysis. Sample Preparation Digital Photo Color Analysis Albacore samples were removed from frozen storage and placed in a 1º C cooler and warmed to approximately ­5º C internal temperature before dissection. The skin was carefully removed from the right side of a specimen using a scalpel to remove any connective tissue in order to prevent the removal and damage of cutaneous blood
43 vessels, located directly under the skin surface. Once the right side of the skin was removed the sample was placed inside a custom made light box (4’x 2’x 3’) in order to insure that lighting stayed consistent for each sample. Two full spectrum florescent light bulbs, rated at 5000 Kelvin, were used for illumination. Digital photographs were then taken using a Canon A300 (Canon Inc., Tokyo, Japan) at 640 X 480 pixel resolutions at 3X zoom and converted from jpeg format into bitmap for analysis using Dell Image Expert® 3.4.1.(Corel Corporation, Ottawa, ON, Canada). A red cardboard reference tile was included in each photograph in order to provide a reference color for calibration of the photos (see Figure 2.1.A). After photographing one side, the loins were removed and samples collected for spectrophotometry analysis, which will be addressed in the following section. This procedure was then repeated for the remaining intact left side of the fish. Computer­aided color photo analysis was done using Color Expert© color analysis software (Engineering and Cyber Solutions, Gainesville, FL) for Windows XP© operating systems. Each bitmap digital photograph was cropped utilizing Dell Image Expert® 3.4.1 leaving only the red reference tile and the albacore, all background images being removed (Figure 2.1.B). Cropped photographs were then saved and one photo was chosen at random to provide a reference image (image 57 right side for 2003 and image 15 left side for 2004). The reference image was then analyzed for color and the L, a and b values of the red tile from the analysis were used in order to calibrate the remaining photos. Digital photo analysis utilizes three main colors of the spectrum in different concentrations to create the sixty­four representative colors used for the color analysis in the Color Expert© program (Wallat et al. 2001). These are red, blue and green, also referred to as the RGB system (Luzuriaga et al. 1997). Color Expert© utilizes this system of color analysis to define the three color reference values used for this experiment, L, a, and b. The L­value refers to the luminance or “lightness” which is based on a non­linear function and describes the percent reflectance or transmission of light from the object (Harold 2001). The a­value refers to the red responses which are compared with green to generate a red­to­green color dimension and the b­value compares the green response to the blue to generate a yellow­to­blue dimension (Harold 2001). These three
44 components make up the CIELAB scale, which is the current recommended color scale in most color measurement methods and specifications, both domestically and internationally (Harold 2001). A B
Figure 2.1. Digital image of tuna loin with reference tile before cropping (A) and after (B). Each digital photograph, excluding the reference image, was run through a calibration process using Color Expert©. This process utilizes the L, a and b color values of the reference photo’s red tile and calibrates each image so that the color values of each red tile are identical to those on the reference image. All calibrated digital images were then saved to disk and these images were then used for contour analysis and generation. Contour analysis was then conducted using Color Expert©. This process involves identifying the number of pixels which are below a specified 45 luminance reference value. Once these pixels are computed a red contour is generated on the image, highlighting the areas where residual blood is present. Luminance values were used to create the contours due to the high contrast between the albacore’s white flesh and dark residual blood. An L reference value of 45, which creates a contour containing all pixels with luminance value of 45 or below, was used to compute the contours on each image. This value was chosen based on its ability to created visually accurate contours on the digital images which contain residual blood. Both the left and right side photos were analyzed for each sample using the same experimental procedures. After the contour analysis was complete the amount of surface area affected by residual blood was calculated for each image by dividing the number of pixels within the contour by the total area in pixels. These ratios were used as the basis for statistical analysis. Spectrophotometry Analysis Immediately after photographs were taken for visual analysis muscle samples were removed and placed in freezer bags, vacuum packed and placed back into the ­ 20º C freezer for later analysis. Approximately one­quarter inch cross sections were removed from each fish and combined. Cross sections of white muscle were taken from three locations (Figure 1) on each fish: 1) directly posterior to the pectoral fin, 2) at the vertical mid­line and 3) and directly below the third finlet on the posterior end. For analysis each sample was removed from the ­20º C freezer and placed into a 1.5º C cooler for approximately four hours or until thawed. Methodology for Heme extraction from Richard and Hultin (2002) was adapted and modified (Kristinsson 2003) for this procedure to obtain a crude extract. Modifications of the procedure included using phosphate buffer at pH 8.0 during homogenization instead of distilled deionized water and the centrifuge settings were modified due to equipment limitations. All red or dark muscle tissue was removed and discarded from the samples prior to mincing in a Kitchen Aid™ (Kitchen Aid Inc., St Joseph, MI) mincer. Two grams of each minced sample were homogenized at speed 5 for one minute with a hand­held Polytron CH­6010 (Kinematica AG, Kriens/Luzern,
46 Switerland) homogenizer in 60ml of 20mM phosphate buffer (Na2HPO4) at pH 8.0. Homogenized samples were centrifuged for 50 minutes at 10,000 RPM (11950g) using a Sorvall® Instruments RC­5B refrigerated super­speed centrifuge (DuPont Company, Newtown, CT) . The resultant supernatant was removed from the centrifuge vials and filtered through Whatman #1 filter paper. The extract was then divided into two 25ml sub­samples for spectrophotometry measurements. Prior to taking the spectrophotometry absorption readings each sub­sample had 1ml of 0.06M sodium dithionite/phosphate buffer solution added and then was bubbled with 99.9% pure carbon monoxide for 30 seconds. Five milliliters of each of these the sub­ samples, which were from the same 2g of muscle tissue, were scanned with a reference sample of 8.0 pH phosphate buffer using a UV­2401 PB double beam UV­ Vis spectrophotometer (Shimadzu Corp., Kyoto, Japan) at 420 nm for hemoglobin (Richard and Hultin 2002). The spectrophotometry absorbance readings were then converted to heme concentrations (μmol/kg) using a standard curve created from bovine hemoglobin. This extraction method, which was used because of limited resources, does not completely separate myoglobin and hemoglobin subunits so both may be present in the reported results. It has been assumed that since myoglobin occurs primarily within the muscle cells themselves, the concentration of myoglobin should not be greatly affected by bleeding, and any observed differences between residual blood content of the control and experimental groups would represent hemoglobin concentration changes as a result of bleeding. These procedures were duplicated for each albacore collected and the two absorbance readings from each fish were averaged for statistical analysis. Statistical Analysis A one­way analysis of variance (ANOVA) F­test with simple contrasts was conducted using a separate means model (Ramsey and Schafer 2002) to determine if there were differences between the means of the control and experimental groups for the computer­aided visual analysis, spectrophotometry analysis and internal temperature loss (2004 only),. If the results of the ANOVA test indicated that there
47 was a significant difference between the means then a linear regression was conducted using only the experimental groups to determine which, if any, of the variables tested had significant effects on the results. Linear regression models were used for this analysis based on the design of the experimental procedures, which were a linear partial­factorial design, and their ability to pull out the individual factors which significantly contributed to blood loss (Ramsey and Shafer 2002). The factors used for the linear regression models in 2003 included the four handling variables (bleeding method, immobilization, bleeding position, seawater cooling) and their levels (see table 2.1), time and length. A set of reference variables were assigned which included a level of each factor and indicator or “dummy” variables were assigned to the remaining levels of each variable. The reference variables were assigned to the gill bleeding method, no immobilization method, lay­down bleeding position and no seawater cooling. Time and length were included in the model as continuous variables, not as indicators. Length was added to the model to represent size instead of weight, which could not be accurately measured on­board and was recorded only prior to dissection in the lab. Both bled weight and circumference were not used in the analysis because of both were found to be closely correlated with length, having r­squared values of 0.918 and 0.873 respectively. Internal temperatures at landing were also included in the model in order to determine if temperature may have an effect on bleeding. A simple linear regression was also used to determine if a correlation exists between the time of struggle and landing temperatures (Ramsey and Shafer 2002). Samples collected in 2004, which included the control group and six experimental groups, were analyzed in the same manner as those collected in 2003. An ANOVA with contrasts was initially conducted to determine if a difference in group means existed followed by a linear regression analysis of the individual variable effects. Only the gill and throat bleeding methods were used in 2004 with the gill method assigned as the reference variable. Time was input as a continuous variable as was length. These samples lacked internal temperature measurements due to equipment failure so it was not included in the analysis. Both ANOVA and linear
48 regression models were conducted using S­Plus© 7.0 (Insightful Corp. Seattle, WA) statistical analysis software for Windows®. Results Summer 2003 2003 Digital Image Color Analysis Differences in means of the average percent area showing residual blood between the control and experimental groups collected in 2003 did show a significant difference (p­value= 0.018, ANOVA F­test). The mean of the control group was considerably larger than those in any of the experimental groups indicating that any bleeding procedure and combination of experimental variables found in the 32 experimental groups significantly reduced the amount of residual blood appearing on the external surface of the loins. Results of the digital color photo analysis are shown in Figure 2.2. The mean of percentage of area covered with residual blood in the control group was slightly above 44%. Almost half of the total loin area in the control group that is visible after removing the skin had blood residue visible, either contained within the cutaneous and segmentary blood vessels or blood that has escaped from vessels that were damaged during handling, either on the vessel sustained during struggle or from the process of removing the skin 4 . 4 Every precaution was taken to prevent damage to cutaneous and segmentary vessels during the removal of the skin, however, due to fat deposits directly under the skin some minor damage did occur on some samples and was usually confined to the lower loin region where fat deposits are most prevalent.
49 Residual Blood Cover (%) 60 40 (7.66) (2.06) (6.55) (9.79) (2.33) 20 (4.20) (6.47) (3.94) (5.66) (4.02) (9.97) (2.45)
(7.49) (2.87) (4.76) (12.83) (5.39) (2.88) (7.25) (3.88) (2.96) (22.01) (2.27) (6.60) (2.97) (5.87) (11.87) (11.66) (8.96) (4.45) (11.62) (10.26) (8.24) TP/ST/HD/SP/30 TP/ST/LD/SUB/50 TP/SP/LD/N/20 TP/SP/LD/SUB/30 TP/SP/HD/SP/50 TP/N/LD/N/40 TP/SP/HD/N/40 TP/N/HD/N/20 T/ST/LD/SP/20 T/ST/HD/SUB/40 T/SP/LD/N/50 T/SP/LD/SP/40 T/SP/HD/SUB/20 T/N/LD/N/30 T/SP/HD/N/30 T/N/HD/N/50 GP/ST/LD/N/40 GP/ST/HD/N/20 GP/SP/LD/N/20 GP/SP/LD/SUB/30 GP/SP/HD/N/40 GP/SP/HD/SP/50 GP/N/LD/SUB/50 G/ST/LD/N/30 GP/N/HD/SP/30 G/ST/HD/N/50 G/SP/LD/N/50 G/SP/LD/SP/40 G/SP/HD/SUB/20 G/N/LD/SP/20 G/SP/HD/N/30 CONTROL G/N/HD/SUB/40 0 Experimental Procedure Figure 2.2. 2003 Digital image visual analysis with medians (●) and standard deviations (in parentheses). Experimental group means ranged in values from just over 14 to 35%. This difference can easily be detected by a visual inspection of the photographs (Figure 2.3), which clearly show fish in the control group having greater surface area affected by residual blood content then fish in any of the thirty­two experimental groups. Some individual samples within the groups had higher percentages but this can be expected due to variability between individuals in the population and bruising that may have occurred during landing and on­board handling, although bruising was much less prevalent in bled fish. One sample in experimental group 29 (fish #121) had an unusually high mean due to significant bruising. This was probably caused by rough handling on­board which can occur during normal fishing operations. 50 A Control B Exp. Group 17 C Control D Exp. Group 17
Figure 2.3. Calibrated digital images of loins before color contouring (A,B) and after (C, D) with residual blood is shown in black. With the outcome of ANOVA analysis indicating a difference between the control and experimental group means a linear regression model was created and analyzed to determine if any of the experimental variables contribute significantly to 51 reducing the amount of loin area affected by residual blood. The reference variables used in this regression model were gill bleeding method, no immobilization, lay­ down bleeding position and no seawater cooling. The results of the linear regression model are shown in Table 2.3 and were significant (p­value = 0.046, F­test). Table 2.3. 2003 Digital image analysis linear regression results. Variable Coefficient Length 0.324 Landing Temp. 0.243 Time ­0.079 TH ­1.772 GP 0.164 TP 3.717 ST ­2.364 SP ­1.307 HD ­3.481 SPR ­1.692 SUB 1.824 P value (F­test) = 0.046 Number of samples = 96 Multiple R 2 = 0.2024 * 95% Significance level Std. Error t­value P­value (t­test) 0.134 0.317 0.071 2.297 2.254 2.278 2.260 1.964 1.586 1.980 2.053 2.422 ­0.769 ­1.110 ­0.771 0.073 1.632 ­1.046 ­0.666 ­2.193 ­0.854 0.888 0.017* 0.444 0.270 0.443 0.942 0.107 0.298 0.507 0.031* 0.395 0.377 The results of this analysis indicate that out of the four bleeding methods only the throat and pectoral method approached significance (two­sided p­value=0.107, T­ test) at the 90% confidence level. Although not significant, samples collected with this method did, on average, have a higher percentage of residual blood cover then the other three bleeding methods. Pectoral cuts, in general, had a tendency to cause bruising and tissue damage around the incision sites and were not shown to improve blood loss from the cutaneous and segmentary vessels when used in conjunction with either the gill or throat methods. Bleeding by throat incision did show the lowest visual blood content overall but was not shown to be significantly better than the gill method.
The two immobilization methods, spiking and stunning, averaged around 1 to 2% less residual blood coverage but did not show a significant difference from those samples which were not immobilized indicating that immobilization of albacore by
52 spiking or stunning prior to bleeding does not have a detrimental effect on the amount of residual blood visible on the loins. Positioning of the fish during bleeding was shown to have an effect on residual blood cover of the loins. Fish placed in the head­ down position during bleeding had a lower residual blood coverage (two­sided p­ value=0.031, T­test) than those placed in the lay­down position. Head­down positioning resulted in a decrease in the amount of coverage by 3.481% ± 3.152% (95% CI). An ANOVA analysis was also conducted to determine if fish placed in the lay­down position tended to have a greater right side to left side difference in residual blood content than those place head­down. No significance was found (p value = 0.903, F­test) indicating that blood did not tend to pool in the side placed down despite the pressure of contact with the bleeding tote. The use of ambient temperature seawater for on­deck cooling with spray or submersion methods did not result in any significant differences from those which were not cooled (NC). Time of bleeding was also not a factor that contributed to blood loss with no differences found between bleeding times, which ranged from 20 to 50 minutes. This indicates that the majority of blood loss occurs within the first 20 minutes after making the incision. Internal temperature of the fish, which were recorded immediately after landing, did not The size of the albacore, as measured by length, did have an effect on the amount of residual blood visible on the surface of the loins (two­sided p­value=0.018, T­test). Since length, which varied between 57.5 cm and 87 cm, was included in the model as a continuous variable each increment (1 cm) larger would increase the amount of residual blood coverage by 0.324% ± 0.266% (95% CI). However, due to the large variability encountered in living organisms only a relatively small amount of the variance found within the samples could be explained by the linear regression model (R 2 = 0.202). 2003 Spectrophotometry Analysis Spectrophotometry absorbance measurements of the samples were used to determine the amount of residual blood within the muscle tissues. Absorbance values
53 for each sample were converted into heme concentrations using a standard bovine hemoglobin curve (µmol/kg) prior to calculating group means. The outcome of the ANOVA analysis shows that a significant difference exists between the spectrophotometry absorbance means of the control and experimental groups (p­value = 0.004, F­test). This indicates that any experimental procedure, which included bleeding as a factor, used in this experiment significantly reduces the amount of residual blood in the muscle tissues of albacore tuna. The results of the spectrophotometry absorption readings are shown below (Figure 2.4). The un­bled or control group had the greatest amount of heme subunits with a mean concentration of 28.846 μmol/kg. Experimental group means showed a wide range of variation, with heme concentrations from 9.5 to 24.1 μmol/kg, which can be expected when working with complex living organisms. Hemoglobin subunits (mmol / kg) 30 (0.69)
25 20 (0.57) (1.62) (2.19) 15 (1.71) (7.28) (4.03) (3.27) (3.87) (1.70) (1.95) (1.12) (0.79) (3.36) (3.06) (1.94) (4.95) 10 (1.10) (4.30) (8.27) (2.30) (4.83) (5.17) (5.55) (4.67) (4.39) (4.63) (3.66) (4.39) (7.61) (1.05) (5.81) (5.34) TP/ST/HD/SP/30 TP/ST/LD/SUB/50 TP/SP/LD/N/20 TP/SP/LD/SUB/30 TP/SP/HD/SP/50 TP/N/LD/N/40 TP/SP/HD/N/40 TP/N/HD/N/20 T/ST/LD/SP/20 T/ST/HD/SUB/40 T/SP/LD/N/50 T/SP/LD/SP/40 T/SP/HD/SUB/20 T/N/LD/N/30 T/SP/HD/N/30 T/N/HD/N/50 GP/ST/LD/N/40 GP/ST/HD/N/20 GP/SP/LD/N/20 GP/SP/LD/SUB/30 GP/SP/HD/N/40 GP/SP/HD/SP/50 GP/N/LD/SUB/50 G/ST/LD/N/30 GP/N/HD/SP/30 G/ST/HD/N/50 G/SP/LD/N/50 G/SP/LD/SP/40 G/SP/HD/SUB/20 G/N/LD/SP/20 G/SP/HD/N/30 CONTROL G/N/HD/SUB/40 5 Experimental Procedure Figure 2.4. 2003 Spectrophotometry results with median (●) and standard deviations (in parentheses). 54 A linear regression analysis of the spectrophotometry results for the 32 experimental groups was found not to be statistically significant (p value = 0.395, F­ test) indicating that the regression model did not adequately explain the variability between handling methods. The results of the linear regression model (Table 2.3), although not significant, showed some similarities with the computer­aided visual analysis linear regression. With length, head­down positioning and throat and pectoral bleeding having some influence on the regression model. Table 2.4. 2003 Spectrophotometry analysis linear regression results. Variable Coefficient Length ­0.124 Landing Temp. 0.115 Time 0.001 TH 1.264 GP 0.199 TP 2.324 ST 0.242 SP 0.918 HD ­1.375 SPR ­0.435 SUB 0.219 P value (F­test) = 0.395 Number of samples = 96 Multiple R 2 = 0.2024 * 90% Significance level Std. Error t­value P­value (t­test) 0.074 0.175 0.039 1.265 1.242 1.255 1.245 1.082 0.875 1.091 1.131 ­1.678 0.659 0.028 0.999 0.160 1.852 0.194 0.849 ­1.572 ­0.399 0.194 0.097* 0.512 0.978 0.321 0.873 0.066* 0.846 0.399 0.120 0.601 0.847 The results of both the visual and spectrophotometry methods were then analyzed using a simple linear regression which found that almost no correlation existed within the experimental procedures (R 2 = 0.063). A strong correlation was not anticipated given the difference in significant results between the two methods. The results of this component of the study indicate that all bleeding methods used in this experiment significantly reduce the amount of residual blood, both externally and internally.
55 2003 Internal Temperature Analysis Internal temperatures were taken immediately after landing, before any experimental procedures were performed, and again just prior to storage, after the experimental procedures were completed. Temperature loss was used to analyze if any of the experimental procedures increased temperature loss from those of the un­ bled control group. An ANOVA with contrasts was conducted which found no difference between the control group and the 32 experimental groups (p­value = 0.440, F­test). The average temperature loss was just 1.6° C for both the control group and the experimental groups indicating that bleeding by any of the various methods employed in this study does not increase the rate of temperature loss in albacore tuna. The effects of temperature on bleeding were also investigated by including the initial internal temperatures recorded after landing in the linear regression analysis. It was found not to be a significant factor (p­value = 0.512, F­test) and had no discernable effect on the amount of blood removed from the cutaneous vessels. 2004 Digital Image Visual Analysis The results of the 2004 digital image visual analysis for the control and six experimental groups are shown below (figure 2.5). ANOVA results indicate that the group means between the control group and the experimental groups were significantly different (p­value = 0.009, F­test). The un­bled control group had a mean of just below 40% of loin area affected by residual blood content. This was above the means found in the experimental groups, which ranged from 18.6 to 30.4%.
56 Residual blood cover (%) 50 40 (8.70) 30 (1.55) 20 (6.61) (2.66) (7.82) (2.69)
(11.84) T/ST/HD/N/5 T/ST/HD/N/15 T/ST/HD/N/10 G/ST/HD/N/5 G/ST/HD/N/15 G/ST/HD/N/10 CONTROL 10 Experimental Procedure Figure 2.5. 2004 Digital image analysis results with medians (●) and standard deviations (in parentheses). Given the significance of the ANOVA test a linear regression of the experimental procedures was then conducted, using the gill bleeding method used as the reference variable. The results of this analysis (Table 2.5) indicate that there is a significant difference between the methods used (p­value = 0.047, F­test). However, only one of the three variables investigated in this linear regression model showed significant affects on the amount of blood visible on the loins. Table 2.5. 2004 Digital image analysis linear regression results. Variable Coefficient Length ­0.555 Time 0.486 TH ­2.246 P value (F­test) = 0.047 N=24 Multiple R 2 = 0.322 * 95% Significance level Std. Error t­value P­value (t­test) 0.220 0.310 2.574 ­2.520 1.569 ­0.873 0.020* 0.132 0.393 57 Out of these three both the throat bleeding method, as compared to the gill bleeding methods, and the time of bleeding was not shown to influence the amount of blood loss. These results suggest that the two bleeding methods used in 2004, gill and throat cuts, are similar in their ability to remove blood from the cutaneous and segmentary blood vessels. Time of bleeding was also found not to influence the amount of residual blood visible on the external surface of the loins. Size, as measured by length, did have an affect on the amount of residual blood, however, this result is opposite to that found in 2003, with larger fish having less residual blood by area then smaller ones (p­value = 0.020, T­test). 2004 Spectrophotometry Analysis The results of the 2004 spectrophotometry analysis can be seen in figure 2.6. ANOVA indicates that the groups means between the control group and the experimental groups was significantly different (p­value = 0.037, F­test). There was a wide fish to fish variation in the spectrophotometry absorption readings within the experimental groupings; however the results are in the range of those found in 2003. Hemoglobin Subunits (mmol / kg) 40 30 20 (2.29) (10.03) (3.07) (7.95) 10 (8.38)
(4.64) (3.59) T/ST/HD/N/5 T/ST/HD/N/15 T/ST/HD/N/10 G/ST/HD/N/5 G/ST/HD/N/15 G/ST/HD/N/10 CONTROL 0 Experim ental Procedure Figure 2.6. 2004 Spectrophotometry results with medians (●) and standard deviations (in parentheses). 58 A linear regression analysis (Table 2.6), using the same reference variables as the 2004 visual analysis, of the experimental groups indicates that out of the three variables (length, time and throat bleeding) investigated none of them appeared to have any significant affects on the amount of residual blood remaining in the muscle tissues (p­value = 0.195, F­test). Table 2.6. 2003 Spectrophotometry analysis linear regression results. Variable Coefficient Length ­0.204 Time ­0.341 TH 3.573 P value (F­test) = 0.195 N=24 Multiple R 2 = 0.205 Std. Error t­value P­value (t­test) 0.205 0.289 2.402 ­0.992 ­1.179 1.488 0.333 0.252 0.152 This indicates that both the gill and throat bleeding methods are effective in removing blood from the tissues, as shown by the results of the ANOVA, and that length and time were not important factors in the removal of residual blood from the muscle tissues. Discussion Bleeding fish immediately after capture has been found to be an important factor in producing high quality seafood products. Prior research has shown that bleeding helps to delay the onset of rigor and the softening of the flesh producing seafood products with firmer texture than fish which were not bled (Mochizuki et al. 1998, Ando et al. 1999) while also enhancing their appearance (Terayama and Yamanaka 2000). In addition to producing products with better texture and appearance, bleeding helped to postpone the onset of rancidity effectively extending product shelf life by reducing the amount of lipid oxidation that occurs by components in the blood including hemoglobin (Tretsven and Patten 1981, Mochizuki et al. 1998). Richards and Hultin (2002) found that there was a good
59 correlation between hemoglobin content and lipid oxidation in fish muscle. Fish hemoglobin is more sensitive to auto­oxidation than mammalian hemoglobin and as temperature increases so does the rate of auto­oxidation (Jensen 2001), an important factor to consider for tuna quality given their ability to thermoregulate. The demonstrated ability of blood components to oxidize lipids makes bleeding of albacore tuna especially important given the high amounts of lipids present in juvenile fish targeted by the West Coast troll fleet (Wheeler and Morrissey 2003, Rasmussen et al. [in press]). Instant killing or immobilization techniques, which include spiking, have also been shown to improve both the quality and appearance of tuna products when using in conjunction with bleeding. Spiking helps to impede ATP consumption after death, delaying the process of rigor (Ando et al. 1996, Nakayama et al. 1996), which effectively increases product shelf life (Mochizuki and Sato 1996). Oka et al. (1990) reported that spiked fish maintain firmer texture than fish which were not spiked producing higher quality sashimi products. This is believed to be due in part to spiking’s ability to maintain higher flesh pH values which help to reduce proteolytic breakdown of pericellular tissue (Nakayama et al. 1996, Terayama and Yamanaka 2000). In addition, spiking is important for quality because it stops muscle contraction completely. This helps to limit the amount of waste products produced within the muscle tissue after capture, especially the build up of lactic acid, which can increase if fish are allow to struggle until death after capture. Spiked fish exhibited low levels of lactic acid in muscle tissues as compared to fish which were allow to die by asphixiation (Amano et al. 1953). In addition to improving quality it has also been found that bleeding and bleeding/spiking albacore can substantially increase the rates of recovery during processing (American Fishermen’s Research Foundation 2000). This research study found that recovery rates can be increased by around 3% in bled fish versus un­bled fish and that the amount of bruising that occurs is diminished in bled fish. The importance of bleeding and spiking, used in conjunction, has been demonstrated to improve the quality and appearance of seafood products. However, the effectiveness of different bleeding and immobilization techniques on albacore
60 tuna, as well as other handling procedures including position of fish during bleeding, the use of ambient sea water to promote bleeding, and the amount of time required for effective bleeding has not been previously studied. Various on­board handling methods are used by fishers in the US West Coast albacore troll fleet and in order to determine which are the most effective in removing blood these various methods were compared in order to provide the fleet with a set of handling guidelines to improve quality and maximize the economic benefits of this fishery. Our results show that bleeding, by any method and combination of handling techniques used in this experiment significantly reduces the amount of both, the residual blood visible in the cutaneous blood vessels on the external loin surface and the amount of residual hemoglobin within the muscle tissues themselves. Both the 2003 and 2004 digital color analyses show similar results with the un­bled control group having greater than 9% more area covered in residual blood than that of the highest experimental group mean, showing at least a 25% reduction in the amount of blood visible on the loins. Residual hemoglobin with the muscle tissues also decreased significantly within the experimental groups. Un­bled fish in 2003 and 2004 had around 4 μmol/kg more residual hemoglobin within the tissue than that of the experimental group with the highest heme subunit concentration. When all experimental groups are averaged together and compared with the controls, fish collected in 2003 showed a reduction in the amount of visible blood and hemoglobin concentration of 49% and 46% respectively. These finding are similar to the results reported by Ando et al. (1999) which found that effective bleeding methods can significantly reduce the amount of hemoglobin within fish muscle tissues by as much as 50 percent from those of un­bled fish. In the 2004 samples the loss of blood was not as great but was still around 39% in both the visual and tissue analyses, results which were similar to those found for trout (Tretsven and Patten 1981). Although the reduction in residual blood in both the visual and tissue analyses was significant when compared to the controls a considerable amount of blood still remained within both the cutaneous blood vessels and the muscle tissues, a finding that is consistent with prior research conducted by Richards and Hultin (2002).
61 Linear regression analysis on the 2003 results of the digital color images for the individual factors (Table 2.3) found that several of these factors did significantly affect the amount of blood visible on the external surfaces of the loins. Out of the 11 factors investigated only two were found to have a significant influence on the amount of blood removed from the cutaneous blood vessels. These included length and the position of the fish when bleeding. The results indicate that as size increases the amount of blood removed from the cutaneous vessels decreases. Although no prior research exists to help explain this we believe it is due to the distance the blood has to travel before it can be released at the incision site and the viscosity of the blood. Tuna blood coagulates rather quickly once circulation stops and as time passes after the incision is made it is reasonable to believe that the longer the blood has to travel to the incision site the more will coagulate, increasing the viscosity, thus preventing additional bleeding. Clotting times of fish blood has been reported to occur in as little as 5 seconds after bleeding (Connell 1975). Linear regression analysis of the 2004 digital images (Table 2.5) also indicates that the size of the fish impacts the amount of blood removed from the cutaneous vessels, however, the effects are opposite to those found in 2003 with larger fish showing less residual blood content. The reason for this in unclear but may be attributed to interactive effects between the factors including time, which were not analyzed because of the limited number of experimental procedures that could be conducted onboard, or the difference inherent in the organisms themselves. Despite these conflicting results this is not an issue that should be of concern to the fishermen since the size of the fish cannot be influenced. The other factor that had significant effects on the amount of residual blood visible on the loins in the 2003 digital image analysis was the positioning of the fish during bleeding. Fish which were placed in a head down position during bleeding showed a decrease in the amount of visible blood coverage of about 3.5% over those which were bled while lying on their side. One possible explanation for this is the effect that gravity has on the flow of blood. When fish are placed in a head down position the bleeding incision is located below the cutaneous vessels while fish that are position on their sides do not the advantage of gravity to help remove blood. The
62 cutaneous vessels also have a large diameter compared to the vessels and capillary systems supplying the muscles which would make it easier for the blood to move towards the incision site when the fish is positioned head down. In 2004 all the fish were placed in a head down position based on the results from 2003 so additional comparisons between lay down and head down could not be made. The remaining factors in the linear regression analyses of the 2003 and 2004 digital images were found not to have a significant effect on the amount of blood loss. Despite the lack of significance the results still tell us something about these methods. Of the four bleeding methods used in the 2003 experiments we found that the gill and the throat methods were similar in their results, supporting our initial assumptions, and can be used interchangeably depending on the demands of the market. The throat method is preferred by buyers of frozen fish because it’s easy to see if the fish have been bled by a quick visual inspection where it is difficult to tell if the gill method was employed because the operculum cannot be lifted when the fish has been frozen. However, for fresh markets it may be more desirable to use the gill method as it does not alter the outward appearance of the fish. Utilizing the pectoral method in conjunction with the gill or throat cuts was not as effective as using only the gill or throat method. These results were contrary to our initial assumptions that by severing the paired lateral vessels blood loss from the cutaneous vessels would be improved. Although the differences were not significant, fish which were bled using the gill/pectoral and throat//pectoral techniques tended to leave more blood in the cutaneous vessels than the gill or throat only methods. In addition, the area around the pectoral incision sites tended to show a considerable amount of hemorrhaging. This would tend to lower the recovery rates during processing. Given the small size of the albacore caught in this fishery, usually between 5 and 9 kg, any loss of flesh due to hemorrhaging or bruising could be substantial. The two immobilization methods investigated in 2003 were also found not to have a significant affect on the amount of blood removed from the cutaneous vessels. These results are what were expected prior to beginning the experiments. They were investigated because of concerns expressed by fishermen during the design stage. Fishermen were concerned that instant killing methods would stop the heart from
63 pumping blood which they believe is vital for effective bleeding. Some also believe that the muscle activity of fish that are allowed to struggle until death helps to promote additional blood loss. This was found not to be the case as fish immobilized immediately after capture could still be bled effectively. Even though no difference was found in terms of the amount of blood removed between the two immobilization methods, spiking is preferable to stunning because it does not cause hemorrhaging and bruising. We found that stunning a struggling tuna can lead to considerable bruising around the site of the blow and reduce processing recovery rates. Therefore it is recommended that albacore be spiked immediately after capture to prevent additional physical damage, including bruising, and reduce the amount of waste products produced within the muscles after landing which have an adverse affect on product quality (Wood et al. 1977, Swift 1983, Nakayama et al. 1996). The effects that seawater cooling has on bleeding and temperature loss were investigated based on the belief of many fishermen that it helps to promote blood loss by helping to prevent clotting of blood around the incision site. As mentioned earlier coagulation in fish blood can occur in as little as 5 seconds after the incision is made. Our results found that the use of seawater, both sprayed or submersion methods, did not improve the loss of blood from the cutaneous vessels. In addition, it was not shown to be effective in increasing internal temperature loss over fish that did not have seawater cooling methods applied. This is most likely due to the fact that the ambient temperature of the seawater was above 16°C. The average internal temperature immediately after landing of albacore collected in 2003 was 31°C which is consistent with prior reports (Price et al. 1994). Despite the average temperature difference of around 15°C internal cooling that was observed was minimal, an average of only1.6°C. This is most likely a result of the insulating properties of the flesh and skin and the high initial internal temperatures. Even though seawater spray and submersion did not significantly increase temperature loss over fish left on deck it was found to save time during additional handling by rinsing the fish clean prior to storage. The last factor that was investigated as a handling procedure was the time that the albacore were allowed to bleed before being placed into storage. In both the 2003
64 and 2004 digital image analysis time was not found to be a significant factor. Many fishermen, prior to the start of this project, believed that proper bleeding should take approximately 1 hour to complete. Fish were often left on deck for longer periods when blood was still observed coming from the incision site. Our results found that the majority of blood is removed almost immediately after the incision is made and that bleeding times from 5 to 50 minutes showed no discernable differences on the amount of blood that was removed from the cutaneous vessels. The reason for this is believed to be due to the high pressures that are found in the circulatory systems of tunas. Tuna have been found to have increased heart rates (Jones et al. 1993) and high blood pressures ranging between 70 to 100 mmHg (Bushnell and Jones 1994). They also found that tuna have very high ventral aortic pressures that are approximately 3 times higher than pressures found in trout. Once the bleeding incision is made the blood is rapidly expelled due to the high pressure and heart rates, which are elevated even further after struggling during capture. Once fish are bled the circulatory system is no longer a closed system and is no longer under pressure which would tend to inhibit further bleeding. This finding could be beneficial to the fishermen because they now have the option of placing fish into on­deck chilling tanks or into cold storage after only limited bleeding times and no longer should feel required to leave tuna bleeding on deck for an hour or more. This should improve overall product quality since rapid cooling is believe to be the most important factor in improving tuna quality (Price and Melvin 1991, Craven et al. 1997). Both the 2003 and 2004 spectrophotometry linear regression analyses for hemoglobin content in the muscle tissues were found to not be significant. Significant differences were found by ANOVA in both years between the amount of hemoglobin present in the control groups and the experimental groups when only looking at bled versus un­bled fish. The individual factors investigated in the linear regression models were not shown to have an influence on the amount of blood removed from the tissues. As mentioned earlier we found that about 39 to 50 percent of the blood is removed when fish are bled in both the visual and spectrophotometry analyses. Laurs et al. (1978) reported that albacore tuna have blood volumes between 82 and 197ml/kg. Studies on other species of tuna including yellowfin and skipjack
65 have been reported to be around 48ml/kg and 50ml/kg respectively (Brill et al. 1998), value much lower than those found for albacore. These values are above those found for trout, which has reported blood volumes of around 33ml/kg (Duff et al. 1987, Gingerich et al. 1987), and dogfish with blood volumes of 40ml/kg. The concentration of hemoglobin within the white muscle of rainbow trout is reported to be around 11.10μmol/kg and 7.39μmol/kg for un­bled and bled fish respectively (Richards and Hultin 2002). Our hemoglobin concentration results are consistent with prior research based on the blood volume estimates for albacore tuna, which are more than 2 times higher than those found in trout, with un­bled fish averaging around 28 μmol/kg and bled fish around 16 μmol/kg. Tuna have highly vascularized muscle tissues with red muscle reported to have 3 to 4 capillaries per muscle fiber (Mathieu­Costello et al. 1992, Mathieu­Costello et al. 1996). When the major blood vessels are cut during bleeding some blood will flow out at the incision site. We found that between 39 and 46% of the hemoglobin was removed from the muscles of bled fish when compared to un­bled fish. Given our results and those of other researchers it seems reasonable to assume that a substantial fraction of the hemoglobin will remain in the small capillaries that supply blood to the muscles, especially considering the loss of blood pressure that occurs immediately after the incision is made. Another factor that may contribute to hemoglobin remaining in the tissues is the viscosity of the blood which would tend to increase rapidly due to coagulation (Connell 1975). Based on the results of both the digital image visual analysis and the spectrophotometry it was found that bleeding by any of the methods utilized in this experiment significantly reduces the amount of blood, both in the cutaneous vessels and the muscle tissues. Only one individual handling factor, placing fish in a head down position when bleeding, was found to be effective in improving blood loss from the cutaneous vessels, which helps to improve the appearance of the loins. The remaining individual factors of bleeding method, immobilization, seawater cooling and time were not found to contribute significantly to blood loss. The results of this study have helped to answer many of the questions and concerns of fishermen about on­board handling methods currently employed in the fishery. Based on these results
66 and other research it is recommended that a comprehensive on­board handling system for albacore include the following procedures: A. Albacore tuna should be immobilized by spiking immediately after landing to prevent physical damage to the fish and the build up of waste products in the tissue, and; B. They should be bled by either the gill or throat method, depending on market demands, and; C. They should be placed in a head­down position during bleeding to improve their visual appearance to the consumers, and; D. Fish should be bled for 20 to 50 minutes before placement into cold storage. Future work on albacore tuna should include a market analysis of sensory attributes to determine what product attributes are most desirable to consumers and how different onboard handling techniques affect overall product quality.
67 References Abe, H., Dobson, G.P., Hoeger, U. and Parkhouse, W.S. 1985. Role of histidine­ related compounds to intracellular buffering in fish skeletal muscle. American Journal of Physiology. 249:R449­R454. Amano, K., Bito, M., and Kawabata, T. 1953. Handling effect upon biochemical change in the fish muscle immediately after catch­ I. Difference of glycolysis in the frigate mackerel killed by various methods. Bulletin of the Japanese Society of Scientific Fisheries. 19(4):487­498. American Fishermen’s Research Foundation. 2000. Albacore bleeding and spiking trials at Starkist Caribe, August 16­18, 2000. American Fishermen’s Research Foundation Final Report, August 28, 2000. Anderson, J.G. and Anderson, J.L. 1991. Seafood quality: issues for consumer researchers. The Journal of Consumer Affairs. 25(1):144­163. Ando, M., Banno, A., Haitani, M., Hirai, H., Nakagawa, T., and Makinodan, Y. 1996. Influence on post­mortem rigor of fish body and muscular ATP consumption by the destruction of spinal cord in several fishes. Fisheries Science. 62(5):796­799. Ando, M., Nishiyabu, A., Tsukamasa, Y., and Makinodan, Y. 1999. Post­mortem softening of fish muscle during chilled storage as affected by bleeding. Journal of Food Science. 64(3):423­428. Arthur, P.G., West, T.G., Brill, R.W., Schulte, P.M. and Hochachka, P.W. 1992. Recovery metabolism of skipjack tuna (Katsuwonus pelamis) white muscle: rapid and parallel changes of lactate and phosphocreatine after exercise. Canadian Journal of Zoology. 70:1230­1239. Ben­Gigirey, B., Vieites Baptista de Sousa, J.M., Villa, T.G., and Barros­Velázquez, J. 1999. Histamine and cadaverine production by bacteria isolated from fresh and frozen albacore (Thunnus alalunga). Journal of Food Protection. 62(8):933­939. Brill, R.W. 1996. Selective advantages conferred by the high performance physiology of tunas, billfishes, and dolphin fish. Comparative Biochemistry and Physiology. 113A(1):3­15. Brill, R.W., Cousins, K.L., Jones, D.R., Bushnell, P.G., and Steffensen, J.F. 1998. Blood volume, plasma volume and circulation time in a high­energy­demand teleost, the yellowfin tuna (Thunnus albacares). The Journal of Experimental Biology. 201:647­654.
68 Bushnell, P.G. and Jones, D.R. 1994. Cardiovascular and respiratory physiology of tuna: adaptations for support of exceptionally high metabolic rates. Environmental Biology of Fishes. 40:303­318. Connel, J.J. 1975. Control of fish quality. Whitefriars Press Ltd., London. pp. 272. Craven, C., Kolbe, E., Morrissey, M., and Sylvia, G. 1997. Onboard factors affecting chilling and freezing rates, and quality of albacore tuna. A Report to the American Fishermen’s Research Foundation. Oregon State University, November 1997. Dewar, H., Brill, R.W., and Olson, K.R. 1994. Secondary circulation of the vascular heat exchangers in skipjack tuna, Katsuwonus pelamis. The Journal of Experimental Zoology. 269:566­570. Dickson, K.A. 1995. Unique adaptations of the metabolic biochemistry of tunas and billfishes for life in the pelagic environment. Environmental Biology of Fishes. 42:65­97. Dickson, K.A. 1996. Locomotor muscle of high­performance fishes: what do comparisons of tunas with ectothermic sister taxa reveal? Comparative Biochemistry and Physiology. 113A(1):39­49. Duff, D.W., Fitzgerald, D., Kullman, D., Lipke, D.W., Ward, J., and Olson, K.R. 1987. Blood volume and red cell space in tissues of the rainbow trout, Salmo gairdneri. Comparative Biochemistry and Physiology. 87A(2):393­398. Gingerich, W.H. and Pityer, R.A. 1989. Comparison of whole body and tissue blood volumes in rainbow trout (Salmo gairdneri) with 125I bovine serum albumin and 51Cr­erythrocyte tracers. Fish Physiology and Biochemistry. 6(1):39­47. Gingerich, W.H., Pityer, R.A., and Rach, J.J. 1990. Whole body and tissue blood volumes of two strains of rainbow trout (Oncorhynchus mykiss). Comparative Biochemistry and Physiology. 97A(4):615­620. Harold, R.W. 2001. An introduction to appearance analysis. [online] Secondsight. A reprint from GAFTWorld. 84:1­7. http://www.color.org/ss84.pdf#search='Richard%20%20W.%20Harold%20an %20introduction%20to%20appearance%20analysis' Jensen, F.B. 2001. Comparative analysis of autoxidation of haemoglobin. Journal of Experimental Biology. 204: 2029­2033.
69 Jones, D.R., Brill, R.W., and Mense, D.C. 1986. The influence of blood gas properties on gas tensions and pH of ventral and dorsal aortic blood in free­ swimming tuna, Euthynnus affinis. Journal of Experimental Biology. 120:201­ 213. Jones, D.R., Brill, R.W., and Bushnell, P.G. 1993. Ventricular and arterial dynamics of anaesthetized and swimming tuna. Journal of Experimental Biology. 182:97­112. Kim, S.H., An, H., and Price, R.J. 1999. Histamine formation and bacterial spoilage of albacore harvested off the U.S. Northwest coast. Journal of Food Science. 64(2):340­ 343. Kim, S.H., Ben­Gigirey, B., Barros­Velázquez, J., Price, R.J., and An, H. 2000. Histamine and biogenic amine production by Morganella morganii isolated from temperature­abused albacore. Journal of Food Protection. 63(2):244­251. Korsmeyer, K.E., Dewar, H., Lai, N.C. and Graham, J.B. 1996. The aerobic capacity of tunas: adaptations for multiple metabolic demands. Comparative Biochemistry and Physiology. 113(A):17­24. Kristinsson, H.G. 2003. Personnal commincation . Monday, November 17 th , 2003. Laurs, R.M., Ulevitch, R., and Morrison, D.C. 1978. Estimates of blood volume in the albacore tuna. In The Physiological Ecology of Tunas. Edited by G.D. Sharp and A.E. Dizon. Academic Press, New York. pp. 135­139. Luzuriaga, D.A., Balaban, M.O., and Yeralan, S. 1997. Analysis of quality attributes of white shrimp by machine vision. Journal of Food Science. 62:113­118. Manning, J.M., Dumoulin, A., Li, X. and Manning, L.R. 1998. Normal and abnormal protein subunit interactions in hemoglobins. Journal of Biological Chemistry. 273:19359­19362. Mathieu­Costello, O., Agey, P..J., Lodgemann, R.B., Brill, R.W. and Hochachka, P.W. 1992. Capillary­fiber geometrical relationships in tuna red muscle. Canadian Journal of Zoology. 70:1218­1229. Mathieu­Costello, O., Brill, R.W. and Hochachka, P.W. 1996. Structural basis for oxygen delivery: muscle capillaries and manifolds in tuna red muscle. Comparative Biochemistry and Physiology. 113A(1):25.31. Mochizuki, S. and Sato, A. 1996. Effects of various killings procedures on post­ mortem changes in the muscle of chub mackerel and round scad. Nippon Suisan Gakkaishi. 62(3):453­457.
70 Mochizuki, S., Norita, Y., and Maeno, K. 1998. Effects of bleeding on post­mortem changes in the muscle of horse mackerel. Nippon Suisan Gakkaishi. 64(2):276­279. Morrissey, M.T., Rasmussen, R. and Okada, T. 2004. Mercury content in Pacific troll­caught albacore tuna (Thunnus alalunga). Journal of Aquatic Food Product Technology. 13(4):41­52. Nakayama, T., Toyoda, T., and Ooi, A. 1996. Delay in rigor mortis of red sea­bream by spinal cord destruction. Fisheries Science. 62(3):478­482. Oka, H., Ohno, K., and Ninomiya, J. 1990. Changes in texture during cold storage of cultured yellowtail meat prepared by different killing methods. Nippon Suisan Gakkaishi. 56(10):1673­1678. Price, R.J., Melvin, E.F., and Bell, J.W. 1991. Postmortem changes in chilled round, bled and dressed albacore. Journal of Food Science. 56(2):318­321. Price, R.J. and Melvin, E.F. 1994. Recommendations for on board handling of albacore tuna. University of California Cooperative Extension, Sea Grant Extension Program publication, UCSGEP 94­1. July 1994. Ramsey F.L. and Shafer, D.W. 2002. The statistical sleuth: a course in methods of data analysis. Duxbury, Pacific Grove, California. p.742. Rasmussen, RS, Morrissey, MT, and Carroll, S. [In Press]. Effect of Seasonality, Location, and Size on Lipid Content in North Pacific Troll­Caught Albacore Tuna (Thunnus alalunga). Journal of Aquatic Food Product Technology. Richards, M.P. and Hultin, H.O. 2001. Rancidity development in a fish model system as affected by phospholipids. Journal of Food Lipids. 8:215­230. Richards, M.P. and Hultin, H.O. 2002. Contributions of blood and blood components to lipid oxidation in fish muscle. Journal of Agricultural and Food Chemistry. 50:555­564. Swift, D.J. 1983. Blood component value changes in the Atlantic mackerel (Scomber scombrus L.) subjected to capture, handling and confinement. Comparative Biochemistry and Physiology. 76A(4):795­802. Terayama, M. and Yamanaka, H. 2000. Effects of bleeding on the quality of skipjack. Nippon Suisan Gakkaishi. 66(5):852­858. Tibbits, G.F. 1996. Towards a molecular explanation of the high performance of the tuna heart. Comparative Biochemistry and Physiology. 113A(1):77­82.
71 Tort, L., González­Arch, F., Torres, P., and Hidalgo, J. 1991. On the blood volume of the Mediterranean dogfish, Scyliorhinus canicula. Fish Physiology and Biochemistry. 9(2):173­177. Tretsven, W.I. and Patten, B.G. 1981. Effect of arterial incisions on the amount of bleeding and flesh quality of rainbow trout. Marine Fisheries Review. 43(4): 16­18. Voigt, M.N. and Botta, J.R. 1990. Advances in fisheries technology and biotechnology for increased profitability. Technomic Publishing Company Inc., Lancaster, Pennsylvania. pp 481­485. Wallat, G.K., Luzuriaga, D.A., Balaban, M.O., and Chapman, F.A. 2001. Analysis of skin color development in live goldfish using a color machine vision system. North American Journal of Aquaculture. 64(1):79­84. Weber, J. and Haman, F. 1996. Pathways for metabolic fuels and oxygen in high performance fish. Comparative Biochemistry and Physiology. 113A(1):33­38. Wheeler, S. and Morrissey, M.T. 2003. Quantification and distribution of lipid, moisture and fatty acids in West Coast albacore tuna ( Thunnus alalunga ). Journal of Aquatic Food Product Technology. 12(2):3­16. Wood, C.M., McHanon, B.R., and McDonald, D.G. 1977. An analysis of changes in blood pH following exhausting activity in the starry flounder, Platichthys stellatus. Journal of Experimental Biology. 69:173­185.
72 Design and Development of an Onboard Electronic Traceability System in the Eastern Pacific Albacore Tuna (Thunnus alalunga) Fishery Introduction Food safety scares in recent years have many in the food industry considering the implementation of traceability systems to address consumer concerns about food safety and to increase their confidence in the food supply. Traceability is a term that has been used in manufacturing for more than twenty years, primarily within the high technology sector, and refers to a firm’s ability to track products as they move through production and into the marketplace. Traceability also makes it possible for a product to be traced back from the consumer to the producer in an efficient manner, a feature that can be extremely useful for marketing and instances when a product must be recalled. In their 9000:2000 guidelines the International Organization for Standardization (ISO) defines the word traceability as the “ability to trace the history, application or location of that which is under consideration”. This definition is extremely vague and at this time the U.S. food industry does not have a specific standard for designing or implementing traceability systems. Lacking any specific guidelines seafood businesses with the desire or need to institute traceability programs are often faced with considerably more questions than answers. Although traceability software is currently available for the seafood sector it can be very expensive, beyond the financial resources of many small family­owned fishing businesses operating in the Pacific Northwest, and difficult to implement. Most commercially available business software solutions that offer product traceability are tailored towards the processing and distribution sectors and are not designed to collect data onboard fishing vessels. The key to traceability, however, is capturing product information from its origin onward. Therefore, providing traceability information for wild­caught species, which includes albacore tuna, requires data collection to start at the time of capture. Designing an efficient and inexpensive system to collect catch information onboard fishing vessels is a vital step towards achieving seafood
73 traceability and a need that has to be addressed by the industry as new mandatory requirements for traceability emerge around the globe. Defined in relation to the food industry traceability is essentially a detailed recordkeeping system that provides entities within a distribution or food chain the capability to follow food and food producing animals through all stages of production, processing and distribution (Food Standards Agency 2002). Inherent in this definition, although not readily apparent, is the fundamental role of traceability to track not only products but the activities associated with their production and distribution (Moe 1998). As the push for traceability gains momentum around the world many in the food industry are beginning to realize that traceability is increasingly becoming both a commercial and legal necessity (Borresen 2003). Faced with the possibility of mandatory traceability requirements in some overseas markets fishermen in the Pacific Northwest need access to onboard solutions that can provide sufficient traceability in order to meet regulatory requirements without excessive implementation and operating costs. Prior to designing and implementing a traceability system within a food business one must first consider their motivations for providing traceable products and whether or not a traceability system will provide benefits to their business. Three main reasons for implementing traceability systems within the food industry have been identified and include: 1) increasing food safety and quality controls, 2) improving supply side management, and 3) product differentiation in the marketplace (Golan et al. 2003, Thompson et al. 2003). While the second and third reasons are solely the responsibility of the private firm the first is a concern, not only to seafood businesses, but to governments and others entities responsible to the public for food safety. As our food supply continues to globalize, concerns over food safety and product origins will most likely increase and could eventually lead to the institution of mandatory traceability requirements on some food products sold in the U.S. Though gaining momentum within the seafood industry as a means to differentiate products and improve supply­side management the predominant factor leading to the implementation of mandatory traceability was a decline in consumer confidence brought about by food safety issues in the European Union (EU).
74 Traceability came to the forefront in the food industry after mad cow disease, or bovine spongiform encephalitis, decimated the United Kingdom beef industry. Additional food safety issues have augmented Europe’s drive towards traceable food products including dioxin found in animal feed, a proliferation of hoof and mouth disease, and consumer concerns over genetically­modified organisms (GMO) in their food products. Responding to the need for improvements in food safety, recall procedures, and consumer confidence the E.U. drafted the General Food Law regulation 178/2002 which came into effect on January 1, 2005. This law institutes mandatory traceability for all animal protein products sold in the E.U., including all products imported from the U.S. Following the E.U.’s lead Japan, Australia, Canada and other countries are now contemplating or enacting new legislation which will require traceability on some animal protein products. The U.S. government has not legislated for mandatory traceability but has followed other nations by enacting Country­of­Origin Labeling (COOL) requirements. These new labeling requirements were scheduled to take affect in April of 2005 for all seafood products sold in the U.S. Although COOL does not mandate traceability to the extent of the E.U. legislation it does require additional product documentation (i.e. – catch location) from the producer. Another concern on the mind of many governments is the increased threat of bio­terrorism and its possible effects on the food supply system, a system which has been identified as vulnerable to terrorist attacks (Bledsoe and Rasco 2002). In response to this growing threat the U.S. government passed the Bio­terrorism and Response Act of 2002. This legislation requires all entities, both foreign and domestic, providing food products to the U.S. to register their business with the government. This act establishes some degree of traceability by requiring all businesses to document the immediate previous source and immediate recipient, or one step forward and one step back, for all food products sold in the U.S. Food safety is not the only motivating factor behind the growing interest in food traceability around the world. Market dynamics, including a rising demand by consumers for more detailed information about the products they purchase, are making many within the seafood industry aware of the benefits that providing
75 traceability can have on their marketing strategies. In today’s marketplace consumers have a much larger influence on how firms produce and market their products then they ever have in the past. Quality and safety assurances above those mandated are in demand by some consumers and can help capture a premium price in a niche market (Unnevehr et al. 1999). Many high­end consumers are also demanding more from their food than just good taste. Conscientious consumers, especially those interested in both health and the environment, are looking for attributes (i.e. – Omega­3 fatty acids, eco­friendly production practices) that have no affect on quality and are not discernable when examining the product. Creating a successful brand identity requires a firm to promote those product attributes most important to their consumers. Extrinsic attributes, like color and odor, are readily discernable and are what most of today’s consumers are looking for when purchasing seafood. Intrinsic attributes, which include attributes such as fat content, are maintained by internal quality control programs within each firm and are not usually recognized until after the product is consumed. Besides looking at intrinsic and extrinsic attributes many consumers purchase their seafood based on certain credence values. Credence values or attributes include safety and nutritional characteristics as well as consumer defined values such as resource sustainability and eco­friendly production practices (Caswell and Mojduszka 1996). While rarely affecting quality they can be as important to some consumers as a product’s other attributes and can help influence success in the marketplace. Recent studies have shown that consumers are willing to pay more for food products that advertise distinctive intrinsic attributes and credence values including products which are certified in some fashion (i.e. – organic) or products identified by a geographical region (Roosen et al. 2003, Umberger et al. 2003, Loureiro and McCluskey 2000, Wessells et al. 1999). Marketing intrinsic and credence attributes can be very effective in creating a brand identity. However, without comprehensive documentation and verification protocols claims about a products attributes can often be difficult to support. An effective traceability system can supply accurate product documentation to support marketing claims. It can also offer the consumer a “story” to go along with their
76 seafood by creating a link between the harvester and the end consumer. In addition, traceability can improve food safety by streamlining product recall procedures and reduce the amount of time defective products remain in the marketplace. The ability to quickly and accurately locate defective products will help moderate the costs of a product recall and any potential liability to a business. When fully integrated into a firm’s food production practices a traceability system can contribute additional information necessary for Hazard and Critical Control Point (HACCP) regulations, Good Manufacturing Practices (GMP) and any product certification requirements. Over time the collection and use of this data will also help improve a businesses capacity to manage knowledge. Knowledge management, which includes both brands and reputations, is becoming recognized as a vital business asset and when combined with vertical integration and transparency can be a potential source of competitive advantage (Sporleder and Moss 2002). Another important factor to consider when promoting a product is that of quality assurance. Traceability can play an essential role in improving quality assurances; however, it is important for businesses to realize that traceability by itself does not provide a business with a quality assurance program. Moe (1998) describes traceability as an essential sub­system of quality management and an integral part of a quality assurance program. Although implementing a traceability system cannot automatically provide a product with differentiation from its competitors, it can be a beneficial tool by which many business facets can be linked including those of food safety, quality assurance, and marketing. Some seafood businesses and co­ops are already realizing the benefits of offering consumers more information and improved quality assurance on their products by promoting high quality traceable products with a unique brand identity (i.e. – Copper River Salmon). A recent study by Dickinson and Bailey (2002) found that U.S. consumers would be willing to pay more for products that provide traceability, transparency and enhanced quality assurances. Many U.S. fisheries, especially those facing challenges in the marketplace, may benefit by implementing traceability into their production practices as a management and marketing tool.
77 The troll­caught albacore industry in the Pacific Northwest is one example of a commercial fishery that can potentially benefit from the implementation of traceability. Historically this fishery has focused on producing lower quality product for large canneries and the industry has put little effort over the years into developing alternative markets for albacore troll­caught in the Eastern Pacific. The expansion of the Western Pacific fleets during the 1990’s provided a new and inexpensive supply of larger albacore to the major canneries. This development prompted the canneries to move overseas which resulted in the loss of the traditional market for albacore fishermen from Oregon, Washington, and California. A resulting decrease in demand led to very low ex­vessel prices and in many instances fishermen were simply unable to sell their catch before it spoiled. Because of circumstances like these many in the industry now realize that a need exists to develop alternative markets and help fishermen capitalize on normal market mechanisms, especially that of competition between buyers. In order to expand into non­traditional higher end niche markets, however, more emphasis must be placed on assuring the production of high quality albacore with marketable intrinsic and credence attributes. The industry has slowly been adjusting to the changing market conditions. Many vessels have now begun installing air blast freezing systems and are becoming increasingly aware of the need for onboard handling practices that preserve quality. Many within the industry are also becoming interested in developing new techniques to preserve important intrinsic and credence attributes of the product as it moves through the food chain and on to the consumer. As the fleets’ capacity to produce high quality albacore increases they can begin to focus on marketing the desirable market attributes of albacore tuna in order to differentiate their products from those of their competitors. The first step in providing traceable products within the albacore industry, as with any wild caught fish, is onboard the vessel. Without adequate data collection onboard during capture traceability of the resultant product cannot be achieved. Product quality also starts on the vessel and having the ability to capture information on how a fish was handled onboard, not only when and were it was captured, can provide valuable marketing tools for the seafood industry. Currently traceability
78 solutions specifically designed for use at sea are limited and can be very expensive. With legislation already in place in the E.U. and new traceability requirements being instituted elsewhere around the world, fishermen in the U.S. must have the capacity to meet traceability requirements if they desire to continue supplying many overseas markets. Instituting a traceability program will not only allow a company to meet any legal requirements it will also provide a new tool that can help improve numerous management and marketing aspects of small to large businesses. This paper focuses on research being conducted into the design and development of a computer based onboard traceability system in the Pacific Northwest albacore fishery. This system utilizes a custom web­interface application operating on common, commercially available, computer hardware and software that has the ability to record biological, environmental and handling data on individually tagged fish. Information about the vessel, catch, and handling procedures are collected by utilizing a system of two computers including an environmentally­sealed handheld computer, designed to withstand adverse conditions, on the weather deck of the vessel. A complete onboard trial of an experimental traceability system was conducted during the 2004 albacore season, under normal fishing operations, in the Pacific Northwest. The system’s reliability, performance and ability to withstand the hazards of fishing will be addressed. Materials and Methods Traceability System The onboard electronic traceability system tested during the 2004 albacore season consisted of three main components. These included a Dell Inspiron© 5100 laptop computer (Dell Inc., Round Rock, Texas), an LXE MX3X handheld computer (LXE Inc., Norcross, Georgia) and a Garmin 17 GPS receiver (Garmin International Inc., Olathe, Kansas). This system was designed to be versatile and easy to install with a total installation time of around ten minutes. The system was installed in the same basic configuration on each vessel with the Dell computer (laptop) placed in the
79 wheelhouse of the vessel, away from the harsh ocean environment. It was connected to the Garmin Global Positioning System (GPS) receiver using a DB9 serial port connector. The GPS receiver was mounted on the roof of the wheelhouse away from any obstructions that could prevent satellite reception. A simple computer application was used to convert and import the National Marine Electronics Association (NMEA) protocol data from the GPS receiver into the traceability database. The environmentally­sealed LXE computer (handheld), which has an Ingress Protection (IP) rating of 65 5 , was positioned on the weather deck of the vessel. It was secured to the vessel using an LXE vehicle mount specifically designed for this unit. Placement of the handheld in close proximity to where the fish were landed enabled data collection to take place directly after capture. This allows the vessel to record exceptionally accurate information on time and location of capture. Factory upgrades to the handheld computer included an optional monochrome screen designed for use outdoors with touch screen technology, an integrated barcode scanner, and an internal wireless networking card. The system was designed to allow data to be entered into the handheld on the weather deck of a vessel, where conditions can be extreme and loss or damage to equipment is not uncommon. Considering this distinct possibility the system was configured to transmit data from the handheld to the laptop computer, located in the wheelhouse. Once entered into the handheld data is immediately exported to the laptop for storage in the vessels database. Communication between the laptop and handheld computers was accomplished by using a peer to peer 802.11 b/g wireless area network (WAN) connection. Both the laptop and handheld were equipped with wireless network cards utilizing the 2.4 GHz bandwidth with a data transfer rate of up to 54 megabytes per second. A primary concern during development of this electronic traceability system was cost. In order to reduce costs to the fishing industry it was developed utilizing commercial software common on many personal and business computers. The Fishery Data Interface (FDI) application was programmed specifically for this 5 Ingress Protection (IP) ratings are used to specify the environmental protection of electrical equipment. An IP rating of 65 denotes total protection from dust (6) and limited protection against low pressure jets of water (5).
80 research and is not yet commercially available. The laptop was equipped with a Windows XP © operating system and Microsoft Office © Professional software, which comes standard with a Microsoft Access © database. Internet Information Services © (IIS), a common web server application, was installed to allow the laptop to act as a web server. The LXE handheld unit came from the manufacturer with a Windows CE © operating system and only required the installation of two additional programs: one for the management of the barcode scanner and one to operate the WAN card. Fish data entry was done through the use of the FDI program installed on the laptop. It was programmed using Microsoft Visual Studio.NET © and can be accessed through the use of an internet browser, in this case Microsoft’s Internet Explorer © , on either the laptop or the handheld. Once entered into the FDI data is sent to a Microsoft Access database for storage. Data stored in Access can be transferred to other commercial database applications like Oracle © or SQL Server © through the use of Open Database Connectivity (ODBC) tools. In order to keep the FDI program as versatile as possible it was designed to allow the user to modify or change the type of information recorded. FDI was installed on the laptop as a locally hosted website application, utilizing IIS, and is accessed through the handhelds’ internet browser using the peer to peer WAN. A wireless network was chosen for this project instead of a hardwired connection to make the system easy to install. This enables the traceability system to be installed and removed from a vessel in under twenty minutes and does not require the installation of power and/or data cables from the wheelhouse to the weather deck for connectivity. Alternating current and direct current power accessories are available and both were used to power the equipment during the pilot tests depending on the power options of each vessel. The laptop and GPS receiver require continuous power during fishing operations while the handheld was operated on battery power when in use on the weather deck. An LXE vehicle mounting cradle, used to dock the handheld computer, was attached in a secure location on the back of the weather deck. Since the vehicle cradle was not hardwired for power the handheld was brought into the wheelhouse and charged intermittently throughout the day, when catch rates were low, and overnight.
81 In order to record traceability data on individual fish every albacore captured during the fishing trip was marked with an aluminum barcode tag attached around the peduncle. The type of barcode used was Code 128, a variable length numerical code allowing the use of up to 14 digits. An eight digit barcode number was chosen for these trials with the first three digits being used for vessel identification and the remaining 5 for identification of the individual fish. Aluminum barcode tags were utilized because of their ability to withstand harsh conditions and rough handling and still be readable by a scanner. Electronic Data Collection The traceability system was installed and tested on two commercial fishing trips during the 2004 albacore season. The first system trial was conducted onboard the F/V Hans Halvor between the 14 th and the 18 th of August. The second trial took place onboard the F/V EZC from August 31 st to September 6 th . After installation onboard was complete the traceability system was powered up and allowed to initiate its wireless network and calibrate the GPS satellite signal. The handheld was then moved to the landing area on the weather deck and placed in the vehicle docking cradle. Once a secure wireless connection between the laptop and the handheld was established the system was put through a series of tests to make sure data imported into the database accurately. After completing these procedures the database was emptied and the system was prepared to record data. Data entry was accomplished through the use of the FDI program running as a web application on the handheld’s internet browser. The FDI program has a series of entry screens that can be accessed through a main interface screen. Data entry on the handheld is done either through the use of the keyboard, the barcode scanner, or through the integrated touch screen, depending on the type of data required. As you go through the FDI’s various data entry screens the program will prompt the user when to enter data then automatically uploads any related entry screens. Before entering data on individual fish the system requires the operator to enter information on the vessel and its onboard handling methods (Table 3.1).
82 Table 3.1. Traceability data collected on vessel and product activities. Data Field Vessel Identification Trip Identification Target Species Onboard Handling Methods Onboard Cold Storage Methods Sub­Categories Albacore Not bled Bled by gill cut Bled by gill cut w/ pre­chilling Bled by throat cut Bled by throat cut w/ pre­chilling Ice Brine chiller Air­blast freezer This information is attached to every fish entered into the system, whether bar­coded or not, and shows who harvested the product and the traceable onboard “activities” associated with it. All data fields listed in Table 3.1 that have sub­ categories can be modified by the user through the Access database, letting them change both the number of sub­categories and their descriptions. Once the required information concerning the vessel and its onboard handling was entered, which is only necessary at the start of each trip, data on every albacore captured by the researcher during each of the two trips was entered into the traceability system. The type of data collected on every target species entered into the system during these trials are shown in Table 3.2 (below). Each albacore landed was entered into the handheld computer immediately after onboard processing was completed. The type of onboard processing was done in accordance with the onboard handling methods defined by the vessels operator. To test the functionality and versatility of this traceability system all albacore landed were entered into the system in one of the following three methods: 1) bar­ coded and measured for length, 2) measured for length only or, 3) entered without a barcode or length. Without a barcode it is impossible to attach any specific data to an individual fish; however, it can be useful for keeping track of the vessels total catch and catch area (or region) and so was included as a part of these trials.
83 Table 3.2. Traceability data collected on target and by­catch species. Vessel Catch Databases Target Species By­catch Species Data Entered Common name Barcode number (optional) GPS location Time Date Length (optional) Weight (optional) Common name GPS location of catch Time of capture Date of capture The method of data entry used at any particular time depended primarily on current catch rates and processing. The two vessels in this trial used different onboard processing methods, each one reflected in the vessels own database. Onboard the Hans Halvor albacore were bled by an incision to the throat latch and not immobilized. On the EZC they were immobilized by spiking (destroying the brain) prior to bleeding by incision to the throat latch. For fish entered using method one (barcode and length) the barcode tag was scanned through the FDI barcode interface using the handheld’s integrated laser scanner. Once scanned into the handheld the tag was securely attached around the peduncle using a standard cable tie. After entering the barcode the FDI automatically downloads and displays the length and weight entry screen. Each fish was measured for a fork length, to the nearest half centimeter, that was input through the numeric key pad. Weight was not taken due to the difficulty of obtaining accurate measurements onboard these small vessels in rough sea conditions. The fork length was recorded because of the close relationship that exists between fork length and weight in albacore tuna (Uchiyama and Kazama 2003). This feature will allow the vessel to accurately estimate the weight of its catch prior to off­loading without the need for expensive scales. Every fish entered into the traceability database using this method has all the data listed in tables 3.1 and 3.2, except for weight, linked to that individual through the attached barcode number.
84 Method two (length only) was done by using the auto­entry application on the FDI which by­passes the barcode scan. Each fish was measured and the fork length entered through the numeric pad into the system. Every fish documented in this manner has the same information as method one fish with the exception of the barcode number. As mentioned previously, fish entered without a barcode number cannot be tracked individually; however, this method, as well as method three, can still be used to keep track of catch location and onboard handling for the entire catch or smaller segregated batches. Method three also required using the FDI auto­entry application; however, no length was entered before sending the data to the laptop. This is the quickest method and only requires the operator to touch the handheld screen twice, once to access the auto­entry (length/weight) screen and once to enter the data into the database. When conditions allowed, the time it took to complete each of the three methods was recorded using a water­proof stopwatch. This traceability system also has the capability to record landing data on by­catch (similar to method three) though this feature was not utilized during these trials due to a deficiency of by­catch. Once entered into the traceability system the albacore were allowed to bleed on deck until the being placed into cold storage. All albacore with barcodes were scanned into the FDI a second time just before placement into the vessels cold storage. This was accomplished through the use of the storage application within the FDI program. Recording “time into storage” provides the necessary data to calculate how much time a fish was allowed to remain on deck prior to placement into cold storage. All albacore recorded using methods two and three were entered into cold storage in various sized lots depending on how many were currently on deck at the time. Table 3.3 (below) shows the information recorded for each fish when entered into storage using the FDI applications. This information was used to compute additional handling data including the time each fish remained on deck and how long they remained in cold storage on the vessel before off­loading at the processor.
85 Table 3.3. Traceability data collected for cold storage and off­loading. Handling Data Into Cold Storage Off­load Data Entered Time Date Customer Identification Time Date In addition, table 3.3 also shows the type of information recorded at the conclusion of each trip, entered during off­loading at a shore­side processor. The off­ load application within the FDI requires the vessel to enter a customer identification prior to off­load. This step is vital when identifying the next step in the food chain, in this case a processor, and providing full chain traceability. Since each vessel sold and off­loaded their entire catch at the same time and facility all albacore were entered as a single unit for off­loading in the FDI. Bar­coded fish were not scanned again at this time. Scanning the barcodes at off­load is an option and was included for instances when the vessel segregates the catch based on quality parameters and/or divides the catch between more than one buyer. Once the catch was entered for off­load a final date and time stamp, which marks when the product officially changed hands, was made by the FDI for each fish in the database. This feature enables the vessel and/or buyer to calculate the amount of time each albacore spent in cold storage. Both the FDI “into storage” and “off­load” applications are optional and were included to provide additional quality assurance tools to the industry. At the end of each trip the vessel’s databases were downloaded into a desktop computer and checked for accuracy. Once this was completed each database was then exported to a Microsoft Excel spreadsheet, a Microsoft SQL database, and a ArcGIS geo­database using ODBC applications. Results The first two onboard trials of this electronic traceability system were very successful and the overall performance of the system was exceptional. Several easily
86 remedied issues that affected the system’s efficiency were encountered during these tests that will be addressed in the following discussion. During the first system trial, onboard the F/V Hans Halvor, a total of 156 albacore tuna were entered into the traceability system through the use of the FDI program (Figure 3.1). Of these fish, 68 were recorded using method one (barcode and length), 53 with method two (length) and 35 using method three (location only). Figure 3.1. Locations of albacore captured onboard the F/V Hans Halvor.
87 The second trial, which met with better ocean conditions, took place onboard the F/V EZC. During this trip a total of 294 albacore were processed into the traceability system (Figure 3.2). Out of these fish 82 were recorded with method one, 97 using method two and 115 with method three. Figure 3.2. Locations of albacore captured onboard the F/V EZC. Throughout these trials a total of 450 albacore tuna were entered into the traceability system through the Fishery Data Interface and all relevant data, depending on the entry method, was stored successfully in the systems’ Access database (Figure 3.3). One­hundred and fifty albacore tuna were entered into the system for each of the three data recording methods.
88 Figure 3.3. Locations of all 450 albacore captured during onboard trials. When possible the approximate time it took to process fish into the traceability system for each of these three methods was recorded. Procedure times could not be recorded for every fish because of rough sea conditions and/or periods of high catch rates. A total of 156 procedure times were recorded. These included 67, 47 and 42 for methods one through three respectively. These are considered approximate times because a small amount of time was required to turn off the stop
89 watch after finishing a procedure. Method one took an average time of 55 seconds to complete. Method two, which did not require scanning a barcode, took an average of 19 seconds to complete. Both methods one and two required extra handling time for each albacore, beyond what would be considered as standard onboard practice, in order to measure the length and/or scan the barcodes into the system. Method three, which did not require any additional handling, took an average of 3 seconds for each fish. Figure 3.4 (below) shows the location of a single albacore captured during these trials and all relevant traceability data stored for that individual fish, including vital product quality information extrapolated from the onboard records. Barcode #: Species: Date Captured: Time Captured: Latitude: Longitude: Length: Vessel ID: Vessel Name: Trip ID: Cold Storage: Onboard Handling: Date into storage: Time into storage: Time on deck: Offload date: Offload time: Customer: Time in storage: (A) 001­00001 Albacore Tuna 8/14/2004 6:51 A.M. 44.3612 ­125.2748 67 cm 001 F/V Hans Halvor 81404 Air­blast freezer Bled only (throat) 8/14/2004 8:05 A.M. 1 hr 14 min 9/11/2004 3:47 P.M. Jesse’s – Illwaco 27.5 days
(B) Figure 3.4 – Capture location (A) and traceability data collected (B) from one albacore (with extrapolated data underlined). During the 12 days of these initial trials the system functioned as expected and performed without a single system failure or program application error. All the data listed in the methods section (tables 3.1, 3.2 and 3.3) was accurately stored in the Access © database for each individual albacore captured. At the conclusion of each of 90 the two fishing trips each vessels Access database, with all captured fish data, was exported successfully into both an Excel © spreadsheet and an SQL © server database through the use of ODBC applications. In addition, the data was imported into an ArcGIS geo­database, from the Access © database, for use in a Geographic Information System (GIS). This database versatility allows the capture locations and all onboard handling data for any fish entered into this traceability system to be displayed in an interactive geographical format. Discussion The desired outcome of this research project was to design a versatile and cost effective electronic system that could capture product traceability data, including information about the origin and any associated activities, on working fishing vessels. Trials of this traceability system proved to be very successful and clearly demonstrate the feasibility of collecting capture and handling data onboard a fishing vessel, in this case with individual albacore, in an accurate and efficient manner. Detailed information collected by this system could be utilized by the seafood industry to meet food traceability requirements, which are being imposed around the globe, and facilitate improvements in both product quality and quality assurance. In addition, it will provide a useful new tool for marketing high quality Pacific Northwest seafood products. Throughout the initial 12 days at sea the system worked as anticipated and functioned without any hardware malfunctions of software application failures. All the traceability and handling information, as outlined in the methods section, was recorded accurately for each of the 450 albacore captured during the two fishing trips and successfully stored in the appropriate vessels database. Overall the performance of the system during these first trials was exceptional and the experience gained onboard essential for further development and refinement of this onboard traceability system. Although the system performed exceptional during these initial tests several aspects need to be addressed before the system is ready for operation by the
91 commercial fishing fleet. This traceability system was designed to be versatile, simple to install, and easily modifiable so it is capable of being used in a wide variety of fisheries and vessel configurations. In order to make it possible to install and remove the entire system in a manner of minutes, an attribute beneficial during testing, the handheld was operated on battery power and a wireless network was used for communication between the two computers. Operating the handheld on battery power for an entire day of fishing without recharging required setting the unit to go in standby mode when not is use to conserve energy. In standby mode the internal wireless network card is disabled to conserve energy. When the handheld is brought out of standby it may take the system up to several minutes for the wireless network to come back online. Data entry cannot begin again until the network is functioning correctly. If data entry is commenced before the network comes fully back on line a network error will occur which requires the operator to refresh the FDI entry page. This delay, although minor, may cause some confusion for those who are not familiar with wireless networks or slight interruptions during onboard processing, particularly when catch­per­unit effort is high. Hardwired options for a direct connection between the laptop and handheld are readily available for both power and networking which will alleviate this delay and permit for data entry without interruptions. Hardwiring the system will add no more than several minutes to the installation time but may limit some of the mobility of the handheld computer while on deck. The versatility of this system also allows the vessel operators to have the option of using both types of connections simultaneously. The power and network cables can be run through the docking station instead of directly into the handheld. In this configuration, when the handheld is on the dock, it has a direct network connection with the laptop and the battery is continuously charged. When removed from the docking station and running on battery power it will automatically connect to the laptop via the wireless network, although there may still be a slight delay, and can then be moved freely around deck. System installation can be configured onboard in any one of these ways according to the preferences of the fishing vessel and can be quickly changed to an alternate configuration if desired. Another benefit of this system is that once purchased it can
92 be set­up for use in less than hour, which including installing all required software and initiating the network connections. Once on a vessel it can be removed and re­ installed in a manner of minutes and takes very little space, a commodity that is in high demand on most small fishing vessels. The LXE MX3X handheld computer was sealed to the environment and has a maximum environmental rating of IP65. This protection was sufficient for these trials; however, for longer term use in the harsh ocean environment a unit with an IP rating of 67 or higher would be recommended and are available. A monochrome screen, intended for outdoor use, was ordered for the handheld in view of the fact that it’s primarily used on the weather deck of the vessel. This screen configuration worked well in all available light conditions but was hard to read while wearing polarized sunglasses. Various screen options are available from the various manufacturers, including color screens, which may work better for some vessels and situations. The key pad on this LXE model, used primarily to enter length and weight data, has small buttons that are sometimes difficult to access while wearing gloves. At least one manufacturer produces a unit with larger alpha and numerical buttons that can be readily adapted to this system. The integrated touch screen, by which the majority of data is entered, worked exceptional throughout these trials and can be operated efficiently with or without gloves. One issue that reduced efficiency of the system when entering data occurred intermittently when scanning the barcode tags. The barcode reader used for these trials is integrated with the handheld and though it operated fine throughout these trials the actual position of the reader window, which protects the laser, did cause some delays when scanning. This window is recessed and located on top of the handheld unit which allows water to collect on its surface. Water drops on the window tend to refract the laser away from the barcode making it difficult to get a successful scan. This situation occurred frequently during rough seas and is the reason why it took an average of 55 seconds to enter fish using method one (barcode and length). Because extra time was required to clean and dry the reader window before most scans could be completed the use of a more efficient barcode reader would significantly reduce the amount of time it takes to enter barcodes. Peripheral
93 barcode readers are available that can be attached directly to the handheld. They come in a variety of configurations including models that do not require manually activation. One common limitation found in many peripheral barcode readers is their lack of sufficient environmental protection and readers with an IP rating of 65 or above are often difficult to locate. The cost of hardware and software for this entire system is approximately $5000 with all necessary attachments, cables and the LXE vehicle dock station. The greatest expenditure for the system is the LXE handheld computer which includes additional upgrade costs for memory, an integrated barcode reader and a wireless networking card. A compatible laptop computer with wireless capabilities can be purchased for as little as $500 from a variety of manufacturers and the GPS receiver can be purchased for approximately $150. Although a GPS receiver was purchased for this project most fishing vessels already have GPS receivers that can be easily integrated into this system. The FDI program is still under development and is currently not available to the public. Another cost that should be addressed in the future is that of the additional labor required to operate this system. The time required to enter a fish into the system ranged from 3 to 55 seconds, depending on which method was utilized. With several modifications to the system, mentioned above, the time to complete method one could be reduced by more than half and barcode tagging would be more efficient and cost effective than was the case during these tests. Other entry methods that require less time and labor, including bar­coding without entering a length or weight, can be utilized by the vessel depending on their needs and those of their customers. Motivation for this research came from several important issues currently facing the albacore industry in the Pacific Northwest. One major issue is the loss of markets and buyers facing the industry since the major canneries moved overseas in the 1990’s. These circumstances have caused considerable turmoil with the industry and, unlike the past, many fishermen find themselves without a competitive market in which to sell their albacore. Because of this lack of markets many fishermen have been forced into direct marketing to sell their fish. This usually requires spending long periods of time tied to the docks searching for buyers and greatly reduces the
94 amount of time they can fish during the season. Lack of competition within the albacore market has also resulted in lower prices being offered to the vessels with current prices roughly equal to below those offered in the 1980’s. A related issue, that has a considerable impact on marketing, is that of product quality within the albacore industry. Traditionally the troll fleet has concentrated on producing albacore for large­scale cannery markets where supplying high quality products was secondary to providing high volume. Without the major canneries buying Northwest troll­ caught albacore the future of the industry lies in the development of new markets, including high end niche markets, which depend mainly on producing high quality products with desirable intrinsic, extrinsic, and credence attributes in demand by the public. The second motivating issue behind this research is the legislation of traceability requirements for seafood products that are rapidly appearing around the globe. In 2002, the E.U. passed the General Food Law Regulation 178/2002 (article 18) requiring all food and food producing animals have traceability at all stages of production, processing and distribution. This is of critical importance to the U.S. albacore fleet since a significant portion of the last few seasons catch has been purchased by Spanish buyers for E.U. markets. Other countries currently considering traceability legislation on food products include Canada, Japan and Australia. Even the U.S. has limited trace back requirements for food producers incorporated into the Bio­terrorism and Response Act of 2002, which makes one­step forward and one­step backward documentation mandatory. Many countries, including the U.S., already have COOL legislation that requires the producer to identify and display the product’s region of origin, production methods and species. Any new traceability requirements for seafood products have the potential of further limiting market access to those without the ability to supply traceability information and should be addressed proactively by the fishing industry, before they have a detrimental effect on the industry and the many small coastal communities that still depend on revenues from fishing. It is important to remember that traceability for wild seafood cannot be accomplished without beginning the process onboard, where the product first enters
95 the food chain. In order to provide accurate documentation onboard an internal traceability system, as described by Moe (1998), must be instituted on the vessel. Although other studies are currently underway with regards to traceability very limited work has actually occurred onboard vessels. Research into onboard traceability has been started in Denmark (Frederiksen et al. 2002); however, the circumstances and fishery are vastly different than those found in the Northwest troll fisheries, where they do not process and package fish onboard. That’s why designing an economical and efficient onboard electronic traceability system for U.S. fisheries is important, especially given the large numbers of small family­owned vessels that still operate in U.S. waters. Not only will it provide the industry with the means to meet any traceability regulations imposed by foreign markets, it will also provide those who are interested with new marketing opportunities for their seafood products. The next step after internal traceability onboard the vessel is the incorporation of the data into a chain traceability system, which encompasses all of the firms that handle the product, from producer to consumer (Moe 1998, Food Standards Agency 2002). Making the step from internal traceability to chain traceability requires a system that can export any required information to the various operating systems and database platforms used within the industry. This is of vital importance in order for the information to be transmitted seamlessly between all entities in the food chain. Without this ability chain traceability cannot be achieved. This system utilizes Microsoft’s Access, a common and inexpensive database application, which comes standard with Microsoft’s professional office software package. Access has all the tools necessary to accurately capture data onboard through the use of the FDI program and can be exported to other database formats using ODBC compliant programs. This versatility allows product information access between businesses operating with different business software solutions and does not require businesses to purchase the identical software to become vertically integrated, a prospect that would be quite impossible in this diverse industry. For those who are interested in providing traceability on a voluntary basis the major question needs to be addressed is whether traceability can provide enough benefits to offset the cost of purchasing and operating a traceability system. An
96 appropriately managed traceability system can provide a considerable array of benefits to companies that use them. Since this system is designed to capture data both on a product’s location and any associated activities that occur during handling and processing the information collected by this traceability system can be utilized for a wide variety of purposes including marketing, business management, and science. Westgren (1999) identified two important benefits of establishing traceability systems within food production. The first is their ability to help identify and preserve important quality traits of the product. Desirable product attributes, when properly documented by a traceability system, can be used for marketing and creating a brand identity for the business. By preserving the identity of the product and its favorable attributes producers can provide improved quality and assurances which will help secure a companies reputation in the marketplace (Unnevehr et al. 1999). Although many factors are involved in creating a successful brand identity one of the most important is a firm’s ability to provide excellent quality assurance. Although traceability by itself does not ensure an effective quality assurance program the information gathered by an appropriately managed system has many important aspects that relate to food safety, food quality and effective product labeling (Kim et al. 1995); aspects that can be used as the backbone of a successful quality assurance and product branding program. In addition, Good Manufacturing Practices (GMP) and Hazard and Critical Control Point (HACCP) programs can be improved by the use of traceability systems. As these manufacturing practices expand in use they are broadening the scope of traceability in accommodating this type of detailed product information (Moe 1998). The second benefit identified by Westgren (1999) is traceability’s capacity to improve a firm’s management of product liability. Many harvested seafood species already present potential threats to the public. Consumption of certain species, if improperly handled, can cause sickness and even death in humans. This list includes albacore tuna, which are susceptible to histamine formation when temperature abused; a situation that can lead to scrombroid poisoning. Maintaining traceability throughout the entire food chain provides an effective and means of preserving quality and rapidly recalling defective products. An effective recall, which means
97 having the capability to quickly locate and remove all defective products from the marketplace, will minimize any associated costs and can help save a company’s reputation. Businesses with accredited and verifiable traceability systems, because of their effectiveness in locating products and identifying activities, have the potential of reducing overhead by commanding favorable liability premiums from insurance companies (Gledhill 2002). Information collected by a traceability system can also improve supply­side and inventory management, helping a firm rapidly respond to internal challenges and external market opportunities (Peterson 2002). Another beneficial aspect of traceability is its capacity to provide, fishermen and scientists, with a wealth of new information on many aspects of species biology and ecology. Spatial information collected with this traceability system, through the use of GIS applications, can be transposed onto satellite generated maps of sea surface temperature, chlorophyll concentrations or other oceanographic data. This information, when used in conjunction with temporal and length at age data, can be utilized to track migration patterns and schooling behaviors in relation to current oceanographic conditions. Over time, this technology can provide the industry and fishery managers with an incredible amount of new information that can be used to ensure that the resource is sustainable into the future, both economically and biologically. This research into electronic seafood traceability clearly demonstrates that collecting capture and handling data onboard can be accomplished in an efficient and reliable manner using today’s computer technologies. The system used in these trials performed remarkably well at sea and will supply a solid foundation for continued research into chain traceability; research which is needed in order to provide the U.S. seafood industry with a means to deliver traceable products to their customers. This is important for two reasons: to enable the industry to meet any new or existing legal requirements for traceability and to provide an innovative tool for marketing U.S. seafood. As the global economy continues to grow and more low­price imported seafood products find their way into the marketplace, many seafood businesses in the U.S., both small and large, are finding it harder to remain competitive. Traceability can offer many benefits including improved supply­side management, product
98 quality, and quality assurances that can help a business, or an entire food chain, quickly respond to and take advantage of new market opportunities. Utilizing all the beneficial aspects of traceability can also help a firm remain competitive and lead to more cooperative business relationships and long term success in the marketplace. In the future, research into onboard traceability and chain traceability needs to be continued in the Eastern Pacific albacore industry and other sectors of the U.S. seafood industry so that the industry will be capable of providing traceable products, if and when required. Currently, this system is receiving minor modifications required to remedy the few problems encountered during these trials and will undergoing further onboard testing. In the future, this system will also be tested in other Pacific Northwest fisheries, including salmon, and as a part of an integrated chain traceability system.
99 References Bledsoe, G.E. and Rasco, B.A. 2002. Addressing the risk of bio­terrorism in food production. Food Technology. 56(2): 43­47. Borresen, T. 2003. Traceability in the fishery chain to increase consumer confidence in fish products ­ application of molecular biology techniques. First Joint Trans­Atlantic Fisheries Technology Conference­ TAFT 2003. 11­14 June 2003, Reykjavik, Iceland. K4. Caswell, J.A. and Mojduszka, E.M. 1996. Using informational labeling to influence the market for quality in food products. American Journal of Agricultural Economics. 78:1248­1253. Dickinson, D.L. and Bailey, D.V. 2002. Meat traceability: Are U.S. consumers willing to pay for it? Journal of Agricultural and Resource Economics. 27(2): 348­364. Food Standards Agency. 2002. Traceability in the food chain: A preliminary study [online]. Food Chain Strategy Division. Accessed on 15 November 2003. URL: http://www.foodstandards.gov.uk/multimedia/pdfs/traceabilityinthefoodchain. pdf Frederiksen, M., Osterberg, C., Silberg, S., Larsen, E. and Bremner, A. 2002. Info­ Fisk. Development and validation of an internet based traceability system in a Danish domestic fish chain. Journal of Aquatic Food Products and Technology. 11(2): 13­34. Gledhill, J. 2002. Tracing the line. Special to Food Processing, March 6. 2002. Accessed 3 November 2003. URL: http://www.foodprocessing.com/Web_First/FP.nsf/ArticleID/LKIE­57YN2X Golan, E., Krissoff, B., Kuchler, F., Nelson, K., Price, G. and Calvin, L. 2003. Traceability in the US food supply: Dead end or superhighway? Choices. Second Quarter 2003: 17­20. Hernandez, M.R.P. 2001. Study of the quality management system and product traceability in a fish processing company. United Nations University, Fisheries Training Programme, Final Project 2001. Accessed 11 September 2003. URL: http://www.unuftp.is/proj01/MariaRitaPRF.pdf Kim, H.M., Fox, M.S. and Gruninger, M. 1995. An ontology of quality for enterprise modeling. In Proceedings of WET­ICE, Los Albamitos, CA, USA. Pp. 105­ 116
100 Loureiro M.L. and McCluskey, J.J. 2000. Assessing consumer response to protected geographical identification labeling. Agribusiness. 16(3): 309­320. Moe, T. 1998. Perspectives on traceability in food manufacture. Trends in Food Science and Technology. 9 (1998): 211­214. Peterson, H.C. 2002. The “learning” supply chain: Pipeline or dream? American Journal of Agricultural Economics. 84(5): 1329­1336. Roosen, J., Lusk, J.L. and Fox, J.A. 2003. Consumer demand for and attitudes toward alternative beef labeling strategies in France, Germany and the UK. Agribusiness. 19(1): 77­90. Sporleder, T.L. and Moss, L.E. 2002. Knowledge management in the global food system: Network embeddedness and social capital. American Journal of Agricultural Economics. 84(5): 1345­1352. Thompson, M., Sylvia, G., and Morrissey, M.T. 2003. Seafood traceability in the U.S. Presented at the annual meeting of the Institute of Food Technologists, July 12­16, Chicago, IL. Uchiyama, J.H. and Kazama, T.K. 2003. Updated weight­on­length relationships for pelagic fishes caught in the Central North Pacific Ocean and bottomfishes from the Northwestern Hawaiian Islands. Pacific Islands Fisheries Science Center Administrative Report H­03­01. August 2003. Pp. 44. Umberger, W.J., Feuz, D.M., Calkins, C.R. and Sitz, B.M. 2003. Country­of­Origin labeling of beef products: U.S. consumers’ perceptions. Presented at 2003 FAMPS Conference, Washington, D.C. 20­21 March 2003. Accessed 16 July 2003. URL: http://www.competitivemarkets.com/whats_new/2003/4­7.pdf Unnevehr, L.J., Miller, G.Y. and Gomez, M.I. 1999. Ensuring food safety and quality in farm­level production: Emerging lessons from the pork industry. American Journal of Agricultural Economics. 81(5): 1096­1101. Wessells, C.R., Johnston, R.J. and Donath, H. 1999. Assessing consumer preference for ecolabeled seafood: The influence of species, certifier, and household attributes. American Journal of Agricultural Economics. 81(5): 1084­1089.
101 Conclusions This research project provides a comprehensive investigation into seafood traceability, quality and how today’s computer technologies can be used, by integrating traceability data collection with onboard quality handling practices, as an instrument for marketing. The first manuscript focuses on traceability and how it may impact the U.S. seafood industry, while the second examines different onboard bleeding and handling techniques and their affect on residual blood content. The third manuscript links these two, seemingly disparate subjects, together with the development of an onboard computerized traceability system designed to collect capture data and document onboard handling practices. In order for the industry to create new market opportunities for albacore products it is essential that they: identify what attributes are desirable to consumers, understand how onboard handling practices affect quality, and have the ability to provide adequate documentation to support marketing claims, quality assurance programs and meet traceability requirements. In this section, the key findings of this project will be evaluated in the context of providing the Eastern Pacific albacore fishery, and other U.S. fisheries, with an uncomplicated and cost effective onboard traceability system that will provide a beneficial new marketing tool for U.S. seafood products. Traceability is a relatively new food safety related topic within the U.S. food industry and many in the industry, especially smaller seafood businesses, are not aware of all it may imply. Research into traceability reveals the importance of documenting, not only the physical location of a product, but also any activities taking place on that product that may affect quality. New legislation in the E.U., which makes traceability mandatory on all seafood products, has recently come into affect that will undoubtedly have an impact on the U.S. seafood industry. Whether this impact is positive or negative will largely depend on a firm’s ability to capitalize on the marketing benefits of providing traceable seafood products. Although traceability is not yet mandatory in the U.S., persistent outbreaks of mad cow disease and other food safety issues will only reinforce the perception within government that traceability should be used to improve food safety and consumer confidence. Given
102 these circumstances, the industry must realize that the possibility of mandatory traceability exists and that it may not always be confined to export markets. A pro­ active stance should be taken by the industry to establish minimum traceability standards and develop integrated traceability systems to protect U.S fishermen from potential trade barriers and the loss of vital markets, both here and abroad. The results of the onboard handling experiments indicate that several onboard bleeding and handling methods do have an affect on the amount of residual blood in the muscle tissues. Residual blood can affect product quality because as it deteriorates it can lead to the production of detrimental by­products and stimulate bacterial growth, which, in turn, can lead to rancidity. With the loss of their traditional market for cannery fish the Eastern Pacific albacore fishery needs to introduce guidelines to improve and standardize quality. Recognizing what onboard handling techniques are the most effective at capturing and preserving those quality attributes desired by consumers is the first step in developing new markets for albacore products. By incorporating the bleeding and handling methods identified in this research with proper chilling techniques a comprehensive onboard quality program can be implemented which will enable fishermen to produce high quality albacore products for all segments of the marketplace. The final aspect of this project, which integrates the research from the first two manuscripts, was the design and development of an onboard computerized traceability system. This system was developed, after many hours of at sea research, to record vital product data that will allow products to be tracked through production and, if necessary, traced back to the producer. Onboard trials of this system, which were conducted under normal fishing conditions, proved to be very successful. Although some minor modifications are currently being made to the FDI program the system concluded these trials without a single hardware or software malfunction. It was effective in accurately and efficiently recording traceability data on individual fish, both with and without barcode tags. The system is able to collect temporal and spatial data, for locating a product, as well as information on the vessel and what onboard handling practices it is currently using. Traceability data collected using this system can then be exported to other businesses in the food chain using ODBC
103 applications. This will allow for the implementation of chain traceability without requiring vertically integrated software solutions. The albacore tuna industry in the Eastern Pacific is currently experiencing some difficulties in opening up new markets for troll­caught fish. This has a lot to do with the traditional perception in the marketplace of albacore products being lower in quality than other alternatives. This fishery, along with many others, is also facing other serious issues including the unpredictable nature of fishing, management concerns, and the possibility of new food safety requirements. To facilitate the creation of new markets, improve quality, and maximize the economic benefits of this fishery the industry should consider the use of traceability systems. Traceability can not only help capture and preserve important product attributes, especially those relating to quality and food safety, it can also provide numerous other benefits including improved quality assurance, supply­side management, and, most importantly, it will make available a new marketing tool for seafood businesses in the U.S.
104 Bibliography Abe, H., Dobson, G.P., Hoeger, U. and Parkhouse, W.S. 1985. Role of histidine­ related compounds to intracellular buffering in fish skeletal muscle. American Journal of Physiology. 249:R449­R454. Amano, K., Bito, M., and Kawabata, T. 1953. Handling effect upon biochemical change in the fish muscle immediately after catch­ I. Difference of glycolysis in the frigate mackerel killed by various methods. Bulletin of the Japanese Society of Scientific Fisheries. 19(4):487­498. American Fishermen’s Research Foundation. 2000. Albacore bleeding and spiking trials at Starkist Caribe, August 16­18, 2000. American Fishermen’s Research Foundation Final Report, August 28, 2000. Anderson, J.G. and Anderson, J.L. 1991. Seafood quality: issues for consumer researchers. The Journal of Consumer Affairs. 25(1):144­163. Ando, M., Banno, A., Haitani, M., Hirai, H., Nakagawa, T., and Makinodan, Y. 1996. Influence on post­mortem rigor of fish body and muscular ATP consumption by the destruction of spinal cord in several fishes. Fisheries Science. 62(5):796­799. Ando, M., Nishiyabu, A., Tsukamasa, Y., and Makinodan, Y. 1999. Post­mortem softening of fish muscle during chilled storage as affected by bleeding. Journal of Food Science. 64(3):423­428. Arthur, P.G., West, T.G., Brill, R.W., Schulte, P.M. and Hochachka, P.W. 1992. Recovery metabolism of skipjack tuna (Katsuwonus pelamis) white muscle: rapid and parallel changes of lactate and phosphocreatine after exercise. Canadian Journal of Zoology. 70:1230­1239. Bailey, D.V., Jones, E. and Dickinson, D.L. 2002. Knowledge management and comparative international strategies on vertical information flow in the global food system. American Journal of Agricultural Economics. 84(5):1337­1344. Ben­Gigirey, B., Vieites Baptista de Sousa, J.M., Villa, T.G., and Barros­Velázquez, J. 1999. Histamine and cadaverine production by bacteria isolated from fresh and frozen albacore (Thunnus alalunga). Journal of Food Protection. 62(8):933­939. Bledsoe, G.E. and Rasco, B.A. 2002. Addressing the risk of bio­terrorism in food production. Food Technology. 56(2): 43­47.
105 Borresen, T. 2003. Traceability in the fishery chain to increase consumer confidence in fish products ­ application of molecular biology techniques. First Joint Trans­Atlantic Fisheries Technology Conference­ TAFT 2003. 11­14 June 2003, Reykjavik, Iceland. K4. Brill, R.W. 1996. Selective advantages conferred by the high performance physiology of tunas, billfishes, and dolphin fish. Comparative Biochemistry and Physiology. 113A(1):3­15. Brill, R.W., Cousins, K.L., Jones, D.R., Bushnell, P.G., and Steffensen, J.F. 1998. Blood volume, plasma volume and circulation time in a high­energy­demand teleost, the yellowfin tuna (Thunnus albacares). The Journal of Experimental Biology. 201:647­654. Bushnell, P.G. and Jones, D.R. 1994. Cardiovascular and respiratory physiology of tuna: adaptations for support of exceptionally high metabolic rates. Environmental Biology of Fishes. 40:303­318. Caporale, V., Giovannini, A., Franscesco, D. and Calistri, P. 2001. Importance of the traceability of animals and animal products in epidemiology. [online] Accessed on 3 November 2003. URL: http://www.ole.int/eng/publicat/rt/2002/caporale.pdf Caswell, J.A. 1998. How labeling of safety and process attributes affect markets for food. Agricultural and Resource Economics Review. October 1998: 151­158. CEN Workshop Agreement. 2002. Traceability of fishery products­ specifications of the information to be recorded in captured fish distribution chains. CEN, 2002. Accessed 5 May 2003. URL: http://193.156.107.66/ff/po/EUTrace/WGCaptured/WGC_StandardFinal.doc CCFICS (Codex Committee on Food Import and Export Inspection and Certification Systems). 2003. Discussion paper on traceability / product tracing in the context of food import and export inspection and certification systems. Accessed on 17 November 2003. URL: http://www.bag.admin.ch/verbrau/Codex/CCFICS%20WG%20­ %20Discussion%20Paper.pdf Clapp, S. 2002. A brief history of traceability. Food Traceability Report: Response for Institute of Food Technologists Annual Meeting, Anaheim, California. 2002. Accessed 12 July 2003. URL: http://ift.confex.com/ift/responses/2002/16.doc Clemens, R. and Babcock, B.A. 2002. Meat traceability: Its effect on trade. Iowa Ag Review. 8(1): 8­9.
106 Committee on Resources and the Committee on Commerce, Science and Transportation. 1997. Proposed implementation of a fishing vessel registration and fisheries information system [online]. Accessed on 21 May 2003. URL: http://www.nmfs.noaa.gov/sfa/401.pdf Connel, J.J. 1975. Control of fish quality. Whitefriars Press Ltd., London. pp. 272. Craven, C., Kolbe, E., Morrissey, M., and Sylvia, G. 1997. Onboard factors affecting chilling and freezing rates, and quality of albacore tuna. A Report to the American Fishermen’s Research Foundation. Oregon State University, November 1997. Department of Agriculture, Fisheries and Forrestry, Australia. 2000. From plate to paddock: Turning the tables. Consumer driven demands on global food chains and the implication for Australia [online]. Accessed on 11 September 2003. URL: http://www.affa.gov.au/content/publications.cfm?ObjectID=E1975734­ DD8C­4277­9F389FD4E9E26BB3 Dewar, H., Brill, R.W., and Olson, K.R. 1994. Secondary circulation of the vascular heat exchangers in skipjack tuna, Katsuwonus pelamis. The Journal of Experimental Zoology. 269:566­570. Dickinson, D.L. and Bailey, D.V. 2002. Meat traceability: Are U.S. consumers willing to pay for it? Journal of Agricultural and Resource Economics. 27(2): 348­364. Dickson, K.A. 1995. Unique adaptations of the metabolic biochemistry of tunas and billfishes for life in the pelagic environment. Environmental Biology of Fishes. 42:65­97. Dickson, K.A. 1996. Locomotor muscle of high­performance fishes: what do comparisons of tunas with ectothermic sister taxa reveal? Comparative Biochemistry and Physiology. 113A(1):39­49. Duff, D.W., Fitzgerald, D., Kullman, D., Lipke, D.W., Ward, J., and Olson, K.R. 1987. Blood volume and red cell space in tissues of the rainbow trout, Salmo gairdneri. Comparative Biochemistry and Physiology. 87A(2):393­398. EAN International. 2003. EPCglobal [online]. Accessed on 15 March 2004. URL: http://www.ean­int.org/index800.html Economic Research Service, U.S. Department of Agriculture. 2002. Consumer food safety behavior: Consumer concerns [online]. Accessed 3 November 2003. URL: http://www.ers.usda.gov/Briefing/ConsumerFoodSafety/
107 Economic Research Service, U.S. Department of Agriculture. 2003. Economics of foodborne disease [online]. Accessed 3 November 2003. URL: http://www.ers.usda.gov/Briefing/FoodborneDisease/ Food Standards Agency, Food Chain Strategy Division. 2002. Traceability in the food chain: A preliminary study [online]. Accessed on 15 November 2003. URL: http://www.foodstandards.gov.uk/multimedia/pdfs/traceabilityinthefoodchain. pdf Frederiksen, M., Osterberg, C., Silberg, S., Larsen, E. and Bremner, A. 2002. Info­ Fisk. Development and validation of an internet based traceability system in a Danish domestic fish chain. Journal of Aquatic Food Product Technology. 11(2): 13­34. Gingerich, W.H. and Pityer, R.A. 1989. Comparison of whole body and tissue blood volumes in rainbow trout (Salmo gairdneri) with 125I bovine serum albumin and 51Cr­erythrocyte tracers. Fish Physiology and Biochemistry. 6(1):39­47. Gingerich, W.H., Pityer, R.A., and Rach, J.J. 1990. Whole body and tissue blood volumes of two strains of rainbow trout (Oncorhynchus mykiss). Comparative Biochemistry and Physiology. 97A(4):615­620. Gledhill, J. 2002. Tracing the line. Special to Food Processing, March 6. 2002. Accessed 3 November 2003. URL: http://www.foodprocessing.com/Web_First/FP.nsf/ArticleID/LKIE­57YN2X Golan, E., Krissoff, B. and Kuchler, F. 2002. Traceability for food marketing and food safety: What’s the next step? Agricultural Outlook. Jan­Feb 2002: 21­25. Golan, E., Krissoff, B., Kuchler, F., Nelson, K., Price, G. and Calvin, L. 2003. Traceability in the US food supply: Dead end or superhighway? Choices. Second Quarter 2003: 17­20. Harold, R.W. 2001. An introduction to appearance analysis. [online] Secondsight. A reprint from GAFTWorld. 84:1­7. http://www.color.org/ss84.pdf#search='Richard%20%20W.%20Harold%20an %20introduction%20to%20appearance%20analysis' Hashimoto, T., Tanaka, K. and Niwa, H. 2003. Trial of farmed fish traceability in Japan. First Joint Trans­Atlantic Fisheries technology Conference­ TAFT 2003. 11­14 June 2003, Reykjavik, Iceland. Hernandez, M.R.P. 2001. Study of the quality management system and product traceability in a fish processing company. United Nations University, Fisheries Training Programme, Final Project 2001. Accessed 11 September 2003. URL: http://www.unuftp.is/proj01/MariaRitaPRF.pdf
108 Hooker, N.H., Nayga, R.M. Jr. and Siebert, J.W. 1999. Preserving and communicating food safety gains. American Journal of Agricultural Economics. 81(5): 1102­1106. ISO 8402. 1994. International Standards. Quality management and quality assurance: vocabulary. URL: http://home.netvigator.com/~timwwh01/ISO­ 8402.HTM. Jensen, F.B. 2001. Comparative analysis of autoxidation of haemoglobin. Journal of Experimental Biology. 204: 2029­2033. Jones, D.R., Brill, R.W., and Mense, D.C. 1986. The influence of blood gas properties on gas tensions and pH of ventral and dorsal aortic blood in free­ swimming tuna, Euthynnus affinis. Journal of Experimental Biology. 120:201­ 213. Jones, D.R., Brill, R.W., and Bushnell, P.G. 1993. Ventricular and arterial dynamics of anaesthetized and swimming tuna. Journal of Experimental Biology. 182:97­112. Kim, H.M., Fox, M.S. and Gruninger, M. 1995. An ontology of quality for enterprise modeling. In Proceedings of WET­ICE, Los Albamitos, CA, USA. Pp. 105­ 116, IEEE Kim, S.H., An, H., and Price, R.J. 1999. Histamine formation and bacterial spoilage of albacore harvested off the U.S. Northwest coast. Journal of Food Science. 64(2):340­ 343. Kim, S.H., Ben­Gigirey, B., Barros­Velázquez, J., Price, R.J., and An, H. 2000. Histamine and biogenic amine production by Morganella morganii isolated from temperature­abused albacore. Journal of Food Protection. 63(2):244­251. Korsmeyer, K.E., Dewar, H., Lai, N.C. and Graham, J.B. 1996. The aerobic capacity of tunas: adaptations for multiple metabolic demands. Comparative Biochemistry and Physiology. 113(A):17­24. Krissoff, B., Kuchler, F., Nelson, K., Perry, J. and Somwaru, A. 2004. Country­of­ Origin labeling: Theory and observations. Outlook, Economic Research Service, United States Department of Agriculture. WRS­04­02, January 2004. Accessed 21 June 2004. URL: http:\\www.ers.usda.gov/publications/WRS04/jan04/wrs0402/ Kristinsson, H.G. 2003. Personnal commincation . Monday, November 17 th , 2003.
109 Laurs, R.M., Ulevitch, R., and Morrison, D.C. 1978. Estimates of blood volume in the albacore tuna. In The Physiological Ecology of Tunas. Edited by G.D. Sharp and A.E. Dizon. Academic Press, New York. pp. 135­139. Loureiro M.L. and McCluskey, J.J. 2000. Assessing consumer response to protected geographical identification labeling. Agribusiness. 16(3): 309­320. Luzuriaga, D.A., Balaban, M.O., and Yeralan, S. 1997. Analysis of quality attributes of white shrimp by machine vision. Journal of Food Science. 62:113­118. MacDubhghaill, U. 2000. The outlook for the UK fishing industry [online]. Accessed 8 February 2003. URL: http://www.intrafish.com/intrafish­ analysis/UK_2000_48_eng/ Manning, J.M., Dumoulin, A., Li, X. and Manning, L.R. 1998. Normal and abnormal protein subunit interactions in hemoglobins. Journal of Biological Chemistry. 273:19359­19362. Mathieu­Costello, O., Agey, P..J., Lodgemann, R.B., Brill, R.W. and Hochachka, P.W. 1992. Capillary­fiber geometrical relationships in tuna red muscle. Canadian Journal of Zoology. 70:1218­1229. Mathieu­Costello, O., Brill, R.W. and Hochachka, P.W. 1996. Structural basis for oxygen delivery: muscle capillaries and manifolds in tuna red muscle. Comparative Biochemistry and Physiology. 113A(1):25.31. Mermelstein, N.H. 2002. A look into the future of food science and technology. Food Technology. 56(1): 48­55. Mochizuki, S. and Sato, A. 1996. Effects of various killings procedures on post­ mortem changes in the muscle of chub mackerel and round scad. Nippon Suisan Gakkaishi. 62(3):453­457. Mochizuki, S., Norita, Y., and Maeno, K. 1998. Effects of bleeding on post­mortem changes in the muscle of horse mackerel. Nippon Suisan Gakkaishi. 64(2):276­279. Moe, T. 1998. Perspectives on traceability in food manufacture. Trends in Food Science and Technology. 9 (1998): 211­214. Morrissey, M.T., Rasmussen, R. and Okada, T. 2004. Mercury content in Pacific troll­caught albacore tuna (Thunnus alalunga). Journal of Aquatic Food Product Technology. 13(4):41­52. Nakayama, T., Toyoda, T., and Ooi, A. 1996. Delay in rigor mortis of red sea­bream by spinal cord destruction. Fisheries Science. 62(3):478­482.
110 Oka, H., Ohno, K., and Ninomiya, J. 1990. Changes in texture during cold storage of cultured yellowtail meat prepared by different killing methods. Nippon Suisan Gakkaishi. 56(10):1673­1678. Ollinger, M. and Ballenger, N. 2003. Weighing incentives for food safety in meat and poultry [online]. Amber Waves, April 2003 1(2):34­41. Accessed 11 November 2003. URL: http://www.ers.usda.gov/AmberWaves/April03/Features/WeighingIncentives. htm Palsson, P.G., Storoy, J., Frederiksen, M. and Olsen, P. 2000. Traceability and electronic transmission of qualitative data for fish products. Icelandic Fisheries Laboratory, 2001. Accessed 11 Sept. 2003. URL: http://www.norden.org/fisk/sk/sluttrapport_Traceability.pdf Peterson, H.C. 2002. The “learning” supply chain: Pipeline or dream? American Journal of Agricultural Economics. 84(5): 1329­1336. Price, R.J., Melvin, E.F., and Bell, J.W. 1991. Postmortem changes in chilled round, bled and dressed albacore. Journal of Food Science. 56(2):318­321. Price, R.J. and Melvin, E.F. 1994. Recommendations for on board handling of albacore tuna. University of California Cooperative Extension, Sea Grant Extension Program publication, UCSGEP 94­1. July 1994. Ramsey F.L. and Shafer, D.W. 2002. The statistical sleuth: a course in methods of data analysis. Duxbury, Pacific Grove, California. p.742. Rasmussen, RS, Morrissey, MT, and Carroll, S. [In Press]. Effect of Seasonality, Location, and Size on Lipid Content in North Pacific Troll­Caught Albacore Tuna (Thunnus alalunga). Journal of Aquatic Food Product Technology. Richards, M.P. and Hultin, H.O. 2001. Rancidity development in a fish model system as affected by phospholipids. Journal of Food Lipids. 8:215­230. Richards, M.P. and Hultin, H.O. 2002. Contributions of blood and blood components to lipid oxidation in fish muscle. Journal of Agricultural and Food Chemistry. 50:555­564. Roosen, J., Lusk, J.L. and Fox, J.A. 2003. Consumer demand for and attitudes toward alternative beef labeling strategies in France, Germany and the UK. Agribusiness. 19(1): 77­90.
111 Sporleder, T.L. and Moss, L.E. 2002. Knowledge management in the global food system: Network embeddedness and social capital. American Journal of Agricultural Economics. 84(5): 1345­1352. Swift, D.J. 1983. Blood component value changes in the Atlantic mackerel (Scomber scombrus L.) subjected to capture, handling and confinement. Comparative Biochemistry and Physiology. 76A(4):795­802. Sylvia, G. and Peters, G. 1992. Product characteristics and market demand for Pacific whiting. In the Proceedings for the Pacific Whiting Harvesting, Processing and Quality Assurance Workshop, March 30­31, Newport, Oregon. Edited by Gil Sylvia and Michael Morrissey. pp 82­86. Tall, A. 2001. Traceability procedures based on FDA and CFIA regulations­ an understanding. INFOFISH International, 5: 49­51. Terayama, M. and Yamanaka, H. 2000. Effects of bleeding on the quality of skipjack. Nippon Suisan Gakkaishi. 66(5):852­858. Thompson, M., Sylvia, G., and Morrissey, M.T. 2003. Seafood traceability in the U.S. Presented at the annual meeting of the Institute of Food Technologists, July 12­16, Chicago, IL. Tibbits, G.F. 1996. Towards a molecular explanation of the high performance of the tuna heart. Comparative Biochemistry and Physiology. 113A(1):77­82. Tort, L., González­Arch, F., Torres, P., and Hidalgo, J. 1991. On the blood volume of the Mediterranean dogfish, Scyliorhinus canicula. Fish Physiology and Biochemistry. 9(2):173­177. Tretsven, W.I. and Patten, B.G. 1981. Effect of arterial incisions on the amount of bleeding and flesh quality of rainbow trout. Marine Fisheries Review. 43(4): 16­18. Uchiyama, J.H. and Kazama, T.K. 2003. Updated weight­on­length relationships for pelagic fishes caught in the Central North Pacific Ocean and bottomfishes from the Northwestern Hawaiian Islands. Pacific Islands Fisheries Science Center Administrative Report H­03­01. August 2003. Pp. 44. Umberger, W.J., Feuz, D.M., Calkins, C.R. and Sitz, B.M. 2003. Country­of­Origin labeling of beef products: U.S. consumers’ perceptions. Presented at 2003 FAMPS Conference, Washington, D.C. 20­21 March 2003. Accessed 16 July 2003. URL: http://www.competitivemarkets.com/whats_new/2003/4­7.pdf
112 Unnevehr, L.J., Miller, G.Y. and Gomez, M.I. 1999. Ensuring food safety and quality in farm­level production: Emerging lessons from the pork industry. American Journal of Agricultural Economics. 81(5): 1096­1101. Voigt, M.N. and Botta, J.R. 1990. Advances in fisheries technology and biotechnology for increased profitability. Technomic Publishing Company Inc., Lancaster, Pennsylvania. pp 481­485. Wagner, G.L. and Glassheim, E. 2002. Traceability of agricultural products [online]. Accessed on 11 September 2003. URL: http:// www.ngplains.org/ArticleView~article_id~50~s_article_category_id~2~.asp Wallat, G.K., Luzuriaga, D.A., Balaban, M.O., and Chapman, F.A. 2001. Analysis of skin color development in live goldfish using a color machine vision system. North American Journal of Aquaculture. 64(1):79­84. Weber, J. and Haman, F. 1996. Pathways for metabolic fuels and oxygen in high performance fish. Comparative Biochemistry and Physiology. 113A(1):33­38. Wessells, C.R., Johnston, R.J. and Donath, H. 1999. Assessing consumer preference for ecolabeled seafood: The influence of species, certifier, and household attributes. American Journal of Agricultural Economics. 81(5): 1084­1089. Westgren, R.E. 1999. Delivering food safety, food quality, and sustainable production practices: The label rouge poultry system in France. American Journal of Agricultural Economics. 81(5): 1107­1111. Wheeler, S. and Morrissey, M.T. 2003. Quantification and distribution of lipid, moisture and fatty acids in West Coast albacore tuna ( Thunnus alalunga ). Journal of Aquatic Food Product Technology. 12(2):3­16. Wood, C.M., McHanon, B.R., and McDonald, D.G. 1977. An analysis of changes in blood pH following exhausting activity in the starry flounder, Platichthys stellatus. Journal of Experimental Biology. 69:173­185.