‘\.4 [[fl --‘\\\1\ t ' ‘ “"1".” OVERDUE FINES: 25¢ per dey per ite- RETURNIM LIBRARY MATERIALS: Piece in book return to remove charge from circulation records SELECTED TREATMENTS AND PROPERTIES OF MECHANICALLY DEBONED CARP (Cxprinus carpio) FLESH By Thomas Edward Rippen A THESIS Submitted to Michigan State University in partial fu1fi11ment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1981 l/ / L9H} 4.12 ABSTRACT SELECTED TREATMENTS AND PROPERTIES OF MECHANICALLY DEBONED CARP (Cyprinus carpio) FLESH By Thomas Edward Rippen Lake Huron carp were mechanically deboned and subjected to 3 topics of study: (l) efficiency of deboning operation and character- istics of minced flesh, (2) effect of 2 antioxidants on lipid stability during frozen storage and, (3) effect of 6 treatments on color, flavor and other characteristics of the flesh. Deboning yielded 42% mince containing l3-25% fat. Tenox 2 was more effective than Freez-gard as an antioxidant. Washing minced flesh with water reduced yield, increased shear resistance, reduced color intensity, heme pigment con- tent and flavor intensity, and appeared to lower TBA numbers during storage. Addition of NaHCO3 to the wash water was more effective than plain water or a NaCl wash at decreasing red color and storage TBA numbers. The NaCl wash or the addition of 5% vegetable fat to water washed mince improved overall acceptability. Hexane extraction or H202 bleaching was not successful. ACKNOWLEDGMENTS The author wishes to express his appreciation for the guidance and assistance given by his major professor, Dr. J. F. Price, and for the continuous support of guidance committee members, Dr. L. E. Dawson, Dr. M. A. Uebersax and Dr. N. R. Kevern. A special thank you is extended to technician Rose Gartner for the many hours she dedicated to laboratory analyses. The direction that the author's degree program and employment future has taken is largely credited to the efforts of Dr. A. E. Reynolds. He was willing to take a chance on an unseasoned student for which the student is extremely grateful. Appreciation is also extended to the Michigan Sea Grant program and Doctors Dice and Booren for making a graduate assistantship possible and meaningful. The pursuit of an advanced degree became feasible and enjoy- able by the support of the author's parents, sisters and closest friend, Rosie. Their encouragement and understanding will not be forgotten. ii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . iv LIST OF FIGURES . . . LIST OF APPENDICES . . . . . . . . . . . . . . vii Chapter I. INTRODUCTION . . . . . . . . . . . . . . l 11. REVIEW OF LITERATURE . . . . . . . . . . . 3 Underutilized Fish and Rationale for Mechanical Deboning . . . . . . . . . . . 3 Yield . . . . . . . . . . . 5 Composition . . . . . . . . . 6 Color . . . . . . . . . . . . . . 7 Flavor . . . . . . . . . . . . . . 12 Lipid Stability . . . . . . . . . . . . . l4 Texture . . . . . . . . . . . . . . . . 20 Microbiology . . . . . . . . . . . . . . 2l III. EXPERIMENTAL . . . . . . . . . . . . . . 23 Materials . . . . . . . . . . . . . . . 23 Methods . . . . . . . . . . . . . . . . 32 IV. RESULTS AND DISCUSSION . . . . . . . . . . . 46 Part I: Deboner Efficiency and Characteristics of Mechanically Deboned Carp . . . . . . . . 46 Part II: Antioxidants . . . . . . . . . . . 55 Part III: Corrective Treatments . . . . . . . 63 V. SUMMARY . . . . . . . . . . . . . . . . 96 APPENDICES . . . . . . . . . . . . . . . . . lOO BIBLIOGRAPHY . . . . . . . . . . . . . . . . 116 iii 10. 11. 12. 13. LIST OF TABLES Proximate composition of fish . . . . . . . . Effect of processing minced carp by one or two passes through mechanical deboner, n = l replication Yields of carp processed through mechanical deboner, n = 8 replications . . . . . . . . . Characteristics of mechanically deboned carp harvested during winter and spring . . . . . Selected correlation coefficients for parameters of freshly minced carp . . . . . . . . . Effect of antioxidants on TBA number (mg malgn- aldehyde/10009) during frozen storage at -26 C, 3 replications per treatment . . . . . . . Total heme pigment and pH of treated minced carp, 2 replications per treatment . . . . . . . . Residual sodium chloride in sodium chloride washed minced carp, 3 replications per treatment . . . Effect of treatments on parameters of minced carp, expressed as mean percent of corresponding control . Effect of treatments on properties of minced carp Solubilized protein (by Biuret) in wash water of 3 treatments, 3 replications per treatment . . . Bone and scale content as affected by water wash treatment, n = 3 replicates . . . . . . . . Effect of treatments on lipid stability during storage at -26°C, recorded as TBA difference values (TBA number of control minus TBA number of treatment) iv Page 47 49 50 54 56 66 66 69 7O 72 77 Table Page l4. .Effect of cooking on treated minced carp, previously frozen and stored lO weeks at -260C, n = 3 replicates . . . . . . . . . . . 86 l5. Sensory rankings of treated minced carp, 49 judgments . . . . . . . . . . . . . . . 9l LIST OF FIGURES Figure Page 1. Chemical forms of myoglobin as they affect color of meat . . . . . . . . . . . . . . 9 . Preparation of antioxidant treatments . . . . . . 25 Preparation of four washed minced carp treatments . . 27 Preparation of hydrogen peroxide treatment . . . . 29 Preparation of hexane extracted treatment . . . . . 3l 0‘ U" h (A) N O O C . Performance of prospective judges during screening trial, .3% to accept . . . . . . . . . . . . 42 7. Effect of two antioxidants on TBA values of minced carp during frozen storage at -26 C . . . . . . . 57 8. TBA values during frozen storage at -26 C of minced carp treated by water washing or water washing plus 5% added hydrogenated vegetable fat . . . . . . . 79 9. Effect of washing minced carp in 0.5% w/v solutions of NaCl or NaHCO3 on TBA values during frozen storage at -26 C . . . . . . . . . . . . . . . . 80 lO. Effect of hydrogen peroxide treatment on TBA values of minced carp during frozen storage at -26 C . . . 8l ll. Effect of hexane extraction plus 5% added hydrogenated vegetable fat on TBA values of minced carp during frozen storage at -26 C . . . . . . . . . . . 82 vi LIST OF APPENDICES Appendix Page A.l. Analysis of variance of the effect of Tenox 2, Freez-gard and untreated control on TBA-measured lipid oxidation during frozen storage . . . . . . l02 A.2. Analysis of variance of difference values (control minus treatment) for the effect of Tenox 2, Freez-gard and untreated control on TBA-measured lipid oxidation during frozen storage . . . . . . l03 A.3. Analysis of variance for proximate composition of "corrective" treatments . . . . . . . . . 104 A.4. Analysis of variance of the effect of "corrective" treatments on TBA-measured lipid oxidation during frozen storage . . . . . . . . . . . . . l05 A.5. Analysis of variance of effect of cooking on Hunter L value . . . . . . . . . . . . . 106 A.6. Analysis of variance of effect of cooking on Hunter a value . . . . . . . . . . . . . 107 A.7. Analysis of variance of effect of cooking on Hunter b value . . . . . . . . . . . . . 108 A.8. Analysis of variance of effect of cooking on shear force . . . . . . . . . . . . . . l09 B.l. Score sheet used to screen panelists . . . . . . lll B.2. Score sheet used to sensory evaluate color of "corrective" treatments . . . . . . . . . ll2 8.3. Score sheet used to sensory evaluate flavor of "corrective" treatments . . . . . . . . . . l13 B.4. Score sheet used to sensory evaluate texture of "corrective" treatments . . . . . . . . . . ll4 8.5. Score sheet used to sensory evaluate over-all preferences of "corrective" treatments . . . . . ll5 vii CHAPTER I INTRODUCTION During the past 100 years the commercial fishing industry in the Great Lakes has been characterized by shifts in the fish species targeted and has experienced significant fluctuations in landed tonnage. Stocks have been affected by the extent of exploitation, alteration of habitat, the introduction of parasitic and competitive species, the development of modern fishing methods and probably the release of toxic compounds. For most of its history the Michigan industry was supported primarily by catches of lake herring, lake trout and lake whitefish. Total production reached a peak of 47.5 million pounds per year during l905-l909 and has generally declined to the current level of approximately l2 million pounds per year, now largely com- posed of alewife and lake whitefish (Tainter and White, 1977). The economic condition of the Great Lakes commercial fishery would be strengthened by expanding the market and value of under- harvested species, including sucker, carp, buffalo, burbot, sheeps- head, smelt and menominee. Carp harvested in l977 totaled just 4,800,000 pounds nationwide and l,l4l,000 pounds from Michigan waters. However, these figures could probably be increased by expanding fish- ing effort in shallow water areas of Lake Michigan, Lake Huron and lower Great Lakes (Mattingly and Kevern. l979). Carp is considered a valuable food fish in eastern Europe, Asia and Japan but demand is generally low in U.S. markets. The advent of mechanical deboning equipment and recent advances in seafood tech- nology should help to overcome problems with fine bones, earthy flavors or storage stability. Since very little information is available on the suitability of carp for use in the deboned, minced form, a pre- liminary research project was developed to characterize minced Lake Huron carp flesh and to control anticipated deficiencies as a basis for subsequent product development studies. The investigation described in the following report consists of three study topics: (1) the efficiency of meat-bone separation, and the composition and quality of freshly minced carp; (2) the effect of frozen storage on lipid oxidation with and without commer- cial antioxidants; and (3) the effectiveness of six treatments designed to lighten color, reduce flavor and/or extend frozen storage life. CHAPTER II REVIEW OF LITERATURE Underutilized Fish and Rationale for MechanicaT Deboning In recent years the United States has been faced with an increased demand for fishery products accompanied by a general decline in the availability of traditional, high value species of finfish and shellfish. In addition to an increase in population, per capita consumption of fish rose from approximately 11 pounds as recently as 1970 (Miyauchi and Steinberg, 1971) to 13.7 pounds in 1979 (USDA 1980). For economic reasons much of the potential U.S. catch is not harvested (Steinberg, 1974) and most of the production of herring-like fishes is processed to make animal feed and indus- trial products rather than used for human food (Lee, 1963; Okada et a1., 1973). Baker (1978) estimated that 70 percent of marine fishes and 80 percent of Great Lakes species are underused. He also noted that a substantial portion of those fish captured incidental to target species are discarded due to the absence of established markets. Even for traditionally harvested species the amount of edible flesh which is recovered may be only 50 percent of what is technologically attainable (Steinberg, 1974; Nobel, 1973). Many authors have proposed the adoption of deboning-mincing equipment and related technology to more completely utilize fishery resources (Keay, 1979; King and Carver, 1971; Lanier and Thomas, 3 1978). A method for mechanical deboning of fish was developed in Japan circa 1940 and has been an important process there since the early 19505 (Anon. 1970; King and Carver, 1971). Fish are usually headed, gutted and split then passed through a machine which forces muscle particles through a perforated drum, either by squeezing the fish between the drum and a pressure belt or by constricting the fish between the drum and an auger that applies increasing pressure as it turns. This results in separation of minced nearly boneless flesh on one side of the drum from bones, scales and skin on the other side. The design and operation of these machines is summarized by Lanier and Thomas (1978). Certain species of fish may be underutilized due to the presence of intermuscular bones which are difficult to remove by hand or by mechanical filleting (Dawson et a1., 1978; Noble, 1974). Odd shaped species, such as those with unusually large heads or spines, are also difficult to process and are characterized by low fillet yields (Bremner, 1977). Keay (1979) and Teeny and Miyauchi (1972) indicated that small species of fish may not be economically processed by conventional methods. Unpleasant species appearance, name, flesh color, flavor or social connotation are other frequently suggested explanations for poor demand and underharvest of certain species (Baker, 1978; King, ca 1978; Teeny and Miyauchi, 1972; Patashnik, 1974). Mechanical deboning is generally effective at bone removal even from notably boney species. Zapata (1978) recovered minced sucker containing 0.13 percent bone,and Apolinario (1975) determined 0.16 percent bone in mechanically separated tilapia flesh. These levels of bone content are less than what might be expected in fillets (Baker, 1978). The minced form allows for incorporation of various ingredients and additives, and permits extrusion or molding into assorted shapes. Such flexibility may serve to reduce problems of acceptability associated with species identity or stability (Bligh and Regier, 1976), flavor, texture (Hewson and Kemmish, 1976) or color (Jaurequi, 1978; Moledina et a1., 1977). Yield Yield of mechanically deboned flesh depends on fish species and form, and machine operation. Yields ranging from 33 to 58 percent of round weight are commonly reported for various species (Reseck and Waters, 1979; Kudo,eta1.,1973). Okada et al. (1973) reported a low of 28 percent yield for cod and a high of 66 percent for herring. According to Miyauchi and Steinberg (1971) the yield of minced cod may be partially affected by the high percentage of bone and skin in that species (16.7% compared to a normal of approximately 11%). However, Crawford et al.(1972b)achieved a yield of 40.4 percent of round weight as minced cod. The reason for this discrepancy in reported yield is unclear. A small, boney African species, bongo fish, may yield 70 percent minced flesh (Noble, 1974). Removal of bones and other waste by hand filleting results in approximately 25-30 percent fillet yield for most species (Miyauchi and Steinberg, 1971). Crawford et al.(1972b)reported an increased yield of 38.5 percent for mechanically deboned English sole compared to fillet weight and a 79.6 percent increase for Dover sole. Filleting wastes have been processed in a mechanical deboner to yield 31.2 percent flesh from ocean perch frames and 72.2 percent flesh from small pollock frames (King and Carver, 1971). Moledina et a1. (1977) estimated 60 percent as a typical yield from frames. Lobster bodies, which are traditionally discarded during processing, yield up to 55 percent edible tissue when passed through a mechanical deboner (Noble, 1973). Adjusting machine belt tension or drum perforation size affects the efficiency of meat-bone separation. Yield of minced flesh from headed and gutted black rockfish varied from 33 percent under "light" belt pressure to 47 percent under "medium belt pressure (Miyauchi et a1., 1975). Bremner (1977) found that drum perforation sizes of 5mm and 10mm resulted in respective yields of 55.6 percent and 60 percent from a headed and gutted Australian fish species. Composition Average proximate composition of the edible portion of some fish species are listed in Table 1. For a given species composition may vary due to nutritional status (Cutting, 1969), season (Finne et a1., 1980) or location (Aitken, 1976). Certain species taken from deep water may have abnormally high moisturezprotein ratios during extended spawning periods (Aitken, 1976; U.S.0.C., 1980). Sablefish caught at 320 fathoms contained 71.0 percent water and 11.5 percent protein while those caught at 450 fathoms contained 81.8 percent water and 8.7 percent protein (U.S.D.C., 1980). Such fish are soft in texture and are sometimes known as "slinks" (Waterman, 1980). TABLE 1.—-Proximate composition of fish. Species Moisture,% Protein,% Fat,% Ash,% Author (many) 70-80 15-20 0.5-20 -- Merritt, (1974) sucker 80.7 16.6 2.0 -- Dawood, (1979) cod 81.8 16.9 0.2 1.0 Bello and Pigott, (1979) herring 72.3 15.6 11.5 1.3 Bello and Pigott, (1979) salmon 67.0 17.0 14.0 -- Cutting, (1969) The composition of normally fatty marine fish, such as herring and mackerel, may change dramatically with season. Aitken (1976) noted that herring sometimes range from less than 1 percent fat and 80 per- cent water to 30 percent fat and 54 percent water. Composition of mechanically deboned freshwater sucker, a lean species, varied but was not significantly related to season or location (Dawson et a1., 1978). Fat content of fish is generally highest in tissue from belly, lateral line and back midline regions (Duttweiler, 1978, Rippen, 1980). The process of mechanical flesh separation may alter composi- tion. Crawford et al.(19723)reported significantly lower moisture contents after deboning in four of six examined marine species when compared to intact fillets. For five of the six species fat content was significantly higher after deboning. The authors attribute the difference to machine inclusion of subcutaneous fat, which is normally removed during hand filleting, and to moisture lost during machine separation. Similar results were reported by Webb et a1. (1976). I Wong et a1. (1978) recovered minced fish through four drum orifice sizes and recorded a small inverse relationship to moisture content. The authors suggest that small (2mm) perforation diameter has no appreciable effect on passage of water but inhibits passage of solids. 99193_ Color of fish flesh is primarily related to the content of the heme pigments, myoglobin and hemoglobin, and in some species to carotenoid content. Less important intrinsic pigments in fish flesh include cobalamine (a cobalt-containing porphyrin), flavins (yellow coenzymes) and cytochromes (a class of heme pigments). The relative content of heme pigments reported in mullet fillets was 63 percent myoglobin and 37 percent hemoglobin (Silberstein and Lillard, 1978) compared to approximately 90 percent of the pigment attributed to myglobin in well bled red meat (Lawrie 1966, Price 1981). Redness caused by accumulation of carotenes is considered desirable in products made from salmon (Schmidt and Cuthbert, 1971) and shrimp (Simon, 1971). However, King (1974, ca 1978) noted that heme color is undesirable in fish sticks and portions which are traditionally produced from white fish. Federal standards for minced fish blocks downgrade color in these products (Fed. Register, 1979). Searobin fish is considered too dark for use in Newburg or chowder (Baker and Darfler, 1979). Combining diagrams by Giffee et a1. (1960) and Rust and Olson (1973) results in the scheme for common color states of myoglobin in red meat shown in Figure 1. Dehydration and pH affect color intensity, and chemical form varies with oxygen partial pressure, presence of Myoglobin (purplish red) ++ Oxygenation (02) 01%?13310263) Fe ++ Fe Sulfide + oxidation Reductants + oxidatiOn Free and oxidized porphyrins (brown, yellow, green, colorless Figure 1. Chemical forms of myoglobin as they affect color of meat. 10 reducing agents, oxidizing agents or heavy metals, bacterial con- tamination and temperature (Giffee et a1., 1960; Rust and Olson, 1973). Heme content variability among fish is usually related to extent of physical exertion. Migratory or pelagic species possess a greater proportion of dark muscle and higher concentrations of myoglobin and hemoglobin than do more sedentary species (Love, 1975; Mai and Kinsella, 1979; Simidu, 1961). A similar relationship exists for red meat animals (e.g. Rickansrud, 1967). Color problems may be significant in mechanically deboned fish, in part due to a major artery and vein system closely associated with the backbone (Lagler et a1., 1962). Fish frames are known to release blood when pressed, especially from under the backbone (King, 1974). This observation may explain the findings of Silberstein and Lillard (1978) who reported an increase in hemoglobin content following deboning. Lee and Toledo (1977) suggested that contact with iron surfaces on deboning machines may catalyze oxidation of myoglobin to the brown metmyoglobin form. Discoloration may also result from reaction of metals with amine, sulfhydryl and phenolic groups (Moledina et a1., 1977). Metal and light catalyzed browning was observed in minced whiting by Anderson and Mendelsohn (1971). Maillard (nonenzymatic) browning is known to occur in fish by reaction of amino acids with either glucose, sugars arising from nucleotide degradation or carbonyls produced during lipid oxidation (Markakis, l979; Pedraja, 1970; Ramamurthy et a1., 1976). 11 Mechanical deboning occasionally incorporates pieces of dark skin or peritoneum in the mince (Lanier and Thomas, 1978). Jauregui and Baker (1980) determined that gray discoloration in fresh minced fish is probably due to release of skin melanins. The largest con- centrations of melanins were measured at high machine pressure settings and at the far end of an auger type deboner. Frozen minced flesh of three marine species became less red and more yellow during six months of storage and was somewhat darker when cooked (Nakayama and Yamamoto, 1977). Such yellowing and browning has been associated with lipid oxidation during frozen storage (Hansen, 1972). Fading of salmon carotenoid pigments during frozen storage also may be related to lipid oxidation (Botta et a1., 1973). A variety of approaches to correct color problems has been proposed. By adding phosphate to minced freshwater mullet, Baker et a1. (1977) achieved less yellowing and improved color accepta- bility scores after frozen storage. Moledina et a1. (1977) found that metal catalyzed reactions and perhaps other oxidative color changes in deboned flounder were minimized during frozen storage by the addition of 0.1% citric acid, 0.2% Kena (a phosphate mixture), 0.25% NaZEDTA and either 0.1% ascorbic acid or 0.1% erythorbate. Acidification to pH 5.3 with citric acid and ascorbic acid lightened color without producing an objectionable chalky, granular appearance noticed at lower pH levels. Diluting or masking color with starch or another white vegetable product (King, 1974), fat (Lanier and Thomas, 1978), titanium dioxide (Jauregui, 1978) or smoke (Bello and Pigott, 1979) 12 has been partially successful for certain products. Since much of the dark muscle is often close to the skin, several researchers suggested operating deboning machines at reduced belt pressure to allow pigment-rich tissue to pass with the wastes (Miyauchi et a1., 1975; Patashnik et a1., 1973; Steinberg, 1975). White V-cuts or other light colored fillet trimmings can be combined with frames to dilute color (King, 1974). Keay (l979) recommended removal of the backbone dorsal to the visceral cavity or mixing light colored fish species with more pigmented species. Heme pigments are water soluble and can be washed from minced fish with cold water - a common practice in Japan (Okada et a1., 1973; Kudo et a1., 1973; Miyauchi 1972). Similarly, Beuchat (1973) observed that pigments were leached from skinned catfish when stored in ice. Wash- ing also removes other proteins, principally sarcoplasmic (Setty et a1., 1974). Bello and Pigott (1979) noted that washing is an expen- sive process and "should be avoided." Prior to deboning Apolinario (1975) decolorized the skin of catfish by dipping them into a 5 percent sodium hydroxide solution for 30 seconds at 155 F. Jauregui (1978) removed fish skin with sodium hydroxide and partially oxidized the melanin in others by bleaching with sodium hypochlorite. Patashnik et a1. (1973) lightened apparent flesh color by forming an emulsion for use as a base in spreads. Small oil globules refract and reflect light from the product. 13 Liam: Characteristic fresh fish flavors are caused by a wide variety of volatile compounds. Howgate (1976) reported improved flavor acceptability for cod when allowed to stand in ice for l to 4 days. Raja and Moorjani (1971) attributed such improvement primarily to the formation of inosine 5'monophosphate (IMP) resulting from the enzymatic deamination and dephosphorylation of adenosine triphosphate. The authors claimed that subsequent enzymatic degradation of IMP is responsible for "the loss of sweet flavour." Flavor improvement in shrimp during initial stages of iced storage may result from the release of free amino acids by native enzymes (Pedraja, 1970). Off flavor development during storage of fresh fish is primarily a result of autolytic and microbial degradation which may be encouraged by mincing (Miller et a1., 1972; Pedraja, 1970). Miller et al. (1972) identified up to 21 volatile ketone, aldehydes, alcohols and amines in ground canary rockfish muscle stored on ice. Production of dimethylamine in gadoid fish is primarily associated with muscles containing high concentrations of heme pigments (Castell et a1., 1971). Dimethylamine formation in silver hake is enhanced by mincing (Lall et a1., 1975). Intrinsic earthy or grassy flavors are often due to organic compounds absorbed by fish from their environ- ment (Rippen, 1980). Mechanical deboning may directly influence flavor. Crawford et al.(1972b)reported an approximately 40 percent reduction in flavor panel scores for minced English sole and Dover sole when compared to intact fillets. The authors attributed the results to the mechanical 14 inclusion of volatile compounds associated with the skin in these species. No significant flavor changes as a result of mechanical deboning were observed with three other species. Inclusion of kidney material in minced fish may contribute off flavors (Dingle and Hines, 1975). Similarly, blood may carry a metallic flavor to minced fish products (Keay, l979). Howgate (1976) suggested that cell rupture and subsequent release of intracellular materials may accelerate enzymatically induced flavor changes in mechanically deboned fish. According to Field (1974) deboner heat may alter flavor of such raw materials as lamb breasts and broiler necks, especially when machines are adjusted for maximum yields. Flavor of minced fish has been improved by mixing 1:1 with minced shrimp (Babbitt et a1., 1974) or blending off flavored finfish with more acceptable finfish (Steinberg 1975). Patashnik (1974) masked intrinsic flavor of minced carp with onion or smoke flavorings. Many researchers proposed a post-deboning ice water wash to reduce odor and flavor of minced fish (Kudo et a1., 1973; Miyauchi et a1., 1975; Okada et a1., 1973; Patashnik et a1., 1973). In addition to removing undesirable compounds elevated moisture levels in the washed mince may dilute flavor (King, 1974). Cooking tends to drive off flavors and odors especially if fish are heated uncovered so that vapors escape (Beuchat, 1973). Phosphates may improve the flavor and aroma of minced fish (Baker et a1., 1977). However, Manohar et a1. (1973) found no consistent flavor preference for walleye or whitefish due to presence or absence of sodium tripolyphosphate. 15 Lipid Stability Shelf life of frozen fishery products is often limited by the development of undesirable flavors and odors resulting from the oxi- dative deterioration of fats and oils. Lipid oxidation is a complex series of reactions initiated between unsaturated oils and oxygen and involves the formation of hydroperoxide intermediates and such terminal products as acids, aldehydes, ketones, alcohols and hydrocar- bons. A review of this topic may be found in Labuza (1971). Even small concentrations of the lower molecular weight compounds may contribute characteristic off flavors to fish. McGill et a1. (1977) determined that hept-c4-enal and, to a lesser extent, two other aldehydes are the primary compounds responsible for cold storage flavor in cod. The authors reported an average flavor recog- nition threshold for hept-c4-enal of only 4x10'5 3 ppm in water which compares to an olfactory threshold of 1.5x10' ppm in oil (Labuza, 1971). The type and concentration of end products created depends on fat content and other compositional characteristics of the fish. Lean fish (e.g. cod) generally oxidize to produce cardboard-like flavors, and fatty fish (e.g. mackerel) become typically rancid (Atkinson and Wessels, 1975; Cole and Keay, 1976). Saltwater mullet is highly susceptable to oxidative rancidity (Deng et a1., 1977; Fischer and Deng, 1977). When analyzed by Finne et a1. (1980) mullet lipids were composed of 34.5 percent polyunsaturated fatty acids, including 23.1 percent that contained 5 or 6 double bonds. A high level of fatty acid unsaturation and instability is common to many species of fish 16 (Stansby, 1973; Lindsay, 1975), especially marine species (Labuza, 1971). As might be expected, tissues that are most susceptable to oxidation are usually highest in fat content such as belly flap, lateral line, dark (red) muscle and skin tissues (Stansby, 1973; Ke et a1., 1977; Lee and Toledo, 1977). Phospholipids were recognized as being relatively more important than triglycerides in the oxidation of cooked pork (Youna- than and Watts, 1960). The frequently reported loss of phospholipids from fish during storage may reflect oxidative degradation directly or enzymatic hydrolysis (Mai and Kinsella, 1979 Wood et a1., 1969). b’ Due to the higher fat content of dark muscle compared to white muscle in sucker (6.2% versus 1.4%), dark muscle was found to contain 41.5% more phosphatidylcholine than white muscle (Mai and Kinsella, 19793). This was true even though white muscle contained more phosphatidyl- choline when expressed on a percent of lipid basis. The readily oxidizable nature of red muscle in fish is probably also related to heme pigment and iron content. Silberstein and Lillard (1978) demonstrated proportionately faster rates of oleic acid oxidation with increased heme protein concentration in mullet extracts. In a kinetic model system they also determined that metmyoglobin was a stronger catalyst than methemoglobin. This is important since mechanical deboning of mullet resulted in an increase of total heme pigment from 4.41 mg/g prior to deboning to 5.77 mg/g following debonding while the hemoglobinzmyoglobin ratio rose from 0.57 to 1.20. These results suggest that although heme-catalyzed 17 oxidation may increase after mechanical deboning, the rate of increase due to the added hemoglobin should be less than would be expected if myoglobin was solely responsible for the increased pig- ment content. Deng et a1. (1978) demonstrated the reactive potential of ground dark muscle from mullet. Ascorbic acid oxidized at a much faster rate in red or mixed flesh than in all light colored flesh. This appeared to be a function of heme and iron content since red muscle and white muscle contained 15.97 ppm heme iron (40.75 ppm total iron) and 0.76 ppm heme iron (2.60 ppm total iron), respectively. Lee et a1. (1975) reported that hemeproteins are the pre- dominant catalysts of lipid oxidation in mechanically deboned chicken. They found antioxidant as well as prooxidant characteristics of hemeproteins depending on relative concentration to substrate where a polyunsaturated fat:heme molar ratio of 500 maximized oxidation. Similar findings were observed by Hirano and Olcott (1971) in linoleate solutions. They suggested that at high concentrations, heme and hemeproteins may form appreciable quantities of oxidized porphyrin derivatives capable of acting as free radical scavengers. At low concentrations, the prooxidant effect was possibly due to their ability to decompose peroxides with the subsequent generation of free radicals for chain initiation. 0f several antioxidants investigated, Fischer and Deng (l977) achieved the most inhibition of lipid oxidation with cyanide which is known to bind strongly to heme proteins and stabilize them. This prompted their conclusion that heme iron is the major catalyst of 18 lipid oxidation in mullet flesh. The researchers also acknowledged the importance of nonheme iron arising from certain enzymes and nonenzyme sources such as ferritin. They reported evidence that 56 to 75 percent of total iron in mullet is nonheme iron. Silberstein and Lillard (1978) found that 2 to 14 percent of total iron was nonheme iron in mullet phosphate buffer extracts but suspected that their procedure was unable to account for all nonheme iron. In a cooked red meat model system nonheme iron possessed prooxidant acti- vi ty but metmyoglobin demonstrated no activity (Love and Pearson , 1974) . Labuza's (1971) discussion of trace metal and heme catalyzed lipid oxidation theory can be summarized as follows: metals such as iron, copper, cobalt, nickel and manganese lower the activation energy required for initiation. This is possible by interaction with and subsequent decomposition of peroxides or by direct radical initiation with the substrate. The catalytic activity of a metal depends on its changing oxidation strate properties. Metals of greatest activity pass through a change of +2 to +3 where the +3 valence is most effi- cient at hydroperoxide decomposition. A slow regeneration of the +2 state favors the +3 form yet releases the metal from the peroxy com- plex (the metal is reduced) for reuse. The peroxyl radical (R00.) produced becomes available to enter propagation reactions. In heme pigments, ferrous or ferric iron is bound to the 4 nitrogens of a porphyrin ring as a planar chelate. The fifth coordination position is bound to the nitrogen of a histidine in the globin (globular protein) and the sixth position is available for complexing with oxygen or other moiety. During oxidation, a 19 hydroperoxide radical can bind to the sixth position despite possible steric hinderance from the globin. Because the electron orbital structure of iron is thermodynamically satisfied when complexed in this manner oxidation might be expected to terminate. However, Labuza (1971) suggested that due to the close proximity of the two electrostatically negative oxygens cleavage is likely if repulsed by the hemeprotein. The oxy-radical (R0.) formed is highly unstable and enters the propagation phase. Unless closely controlled mechanical deboning promotes condi- tions conducive to lipid oxidation. Oxidation was accelerated when fish flesh was allowed to contact iron parts on a meat-bone separator (Lee and Toledo, 1977). Mai and Kinsella (1979a) postulated that "... the deboning process and freezing may accentuate the release and mixing of prooxidants from dark muscle." Many researchers have noted the potential for increased lipid oxidation during and following mechanical deboning due to greater flesh surface area and incorporation of oxygen (Cole and Keay, 1976; Teeny and Miyauchi, 1972; Bremner, 1977; Patashnik et a1., 1973). Conversely, Steinberg (1975) pointed out that the minced form offers the opportunity for exposing more surface to antioxidants. Dawson et a1. (1978) reported that mechanically deboned sucker was less vulnerable to lipid oxidation than ground eviscerated sucker (bones and skin included) or belly flaps, although more oxidizable than loin muscle. Washing minced fish in water has been suggested to improve lipid stability as well as to improve color and flavor. Jauregui 20 (1978) and Miyauchi (1972) reported a reduction in lipid content following washing and, as previously discussed, hematins and other water soluble constituents are at least partially removed during the washing process which probably contributes to slower oxidation rate (Steinberg, 1975). Washed minced fish is commonly dewatered to a water content 3 to 10 percent higher than prewash flesh (Miyauchi et a1., 1975; Patashnik, 1974; Patashnik et a1., 1976) which may minimize contact of reactants with substrate (Labuza, 1971). Texture The act of mechanical deboning results in a loss of muscle integrity. Webb et al. (1976) believed that texture loss was asso- ciated with the shearing of myofibrillar proteins and their subse- quent denaturation. However, the view is not substantiated by research. In a study by Wong and Yamamoto (1978) a direct relation- ship existed between fish texture and machine perforation size. This was attributed to an elevated moisture content resulting from increased compression observed in flesh processed through small holes. Texture may range from "mushy" at a high water content to "rubbery" at a lower water content (Lanier et a1., 1980). Patashnik (1974) reported improved texture for products derived from carp when deboned with a machine equipped with 7mm perforations compared to 1.4mm perforations. Many factors affect the texture of minced fish products including species, maturity, season, location and freshness (Miyauchi 21 et a1., 1973), and temperature, pH and ionic strength (Lanier et a1., 1980; Moledina et a1., 1977; Okada et a1., 1973). Release of proteo- lytic enzymes during the mincing operation has been implicated in the softening of deboned fish. Cathepsins, originating from lysosomes, are most active in an acid environment at moderate temperatures, while alkaline proteases of cytoplasmic origin are functional at alkaline pH and at -20°C to 60°C (Lin et a1. 1980; Lin and Lanier, 1980). The enzymes are 200 to 1500 times more active in kidney tissue than in muscle and represent the major "contaminant" in minced croaker (Lin et a1., 1980). Protein denaturation is manifested as a loss of salt extracta- bility and water holding capacity with concomitant toughening. The toughening commonly associated with frozen fish is probably due more to hydrogen bonding and hydrophobic interactions between polypeptides than to disulfide linkages (Iwata and Okada, 1971). The protein conformational changes required to encourage such bonding may be due to interaction with free fatty acids (Anderson and Favest, 1969), products of lipid oxidation such as malonaldehyde (Anderson, 1970; Jarenback and Liljemark, 1975) and dimethylamine or formaldehyde in gadoid species (Dingle et a1., 1977; Babbitt et a1., 1972). Washing minced fish in water is known to improve binding properties and water holding capacity (Miyauchi et a1. 1973; Kudo et a1., 1973 Miyauchi et a1., 1975). This effect is usually attri- buted to a relative increase in myofibrillar proteins due to loss of water soluble proteins (Okada et a1., 1973; Patashnik et a1. 1976) and to removal of proteases (Okada et a1. 1973; Lanier et a1., 1980). 22 The latter study also suggested that water extraction of prooxidants may aid the retention of protein functionality. Microbiology Freshly caught whole fish possess bacteria in gills, slime and digestive tract and are essentially free of organisms in the tissues (Reay and Shewan, 1949; Hunter, 1933). These bacteria vary in composition depending on location. According to Shewan (1962) fish found in tropical waters carry significant quantities of mesophilic varieties while fish from northern waters are associated with over 95 percent psychrotrophs. Researchers have related spoilage of iced or refrigerated fish primarily to psychrotrophic varieties, particularly species of Pseudomonas and Achromobacter (Reay and Shewan, 1949; Castell and Anderson, 1948). Freshwater fish are spoiled by a somewhat different microflora than marine species and reportedly maintain quality longer than saltwater fish (Nair et a1., 1971, 1974). The penetration of surface bacteria into the flesh is greatly assisted by mechanical deboning. Raccach and Baker (1978) found a ten-fold increase in bacteria count following deboning but acknowl- edged a possible deficiency in the sampling method employed. They suggested that mincing may release nutrients, such as amino acids and vitamins, that improve the conditions for microbial growth. Pedraja (1970) pointed out that bacteria are incapable of attacking intact proteins and must be assisted by native enzymes or mechanical rupture. CHAPTER III EXPERIMENTAL Materials Procurement of Fish All carp used in this study were caught commercially by trap net or gill net from Saginaw Bay, Lake Huron. December and January batches were purchased from Beardsley Fish Co., Standish, Michigan and all other batches from Bay Port Fish Co., Bay Port, Michigan. Carp were boxed in ice 0-5 days at time of purchase, were reiced and transported without additional refrigeration to the Michigan State University Meat Laboratory. When required the fish were topped with additional ice but always stored at 2°C and processed within 2 days. Most fish weighed in the range of 2.3-6.7 kg. Fish Preparation and Mechanical Deboner Operation Carp were weighed as a group when yield data were desired and manually headed, gutted and split dorsoventrally parallel to the backbone. The split carp were washed under cool running tap water, using a hand brush to facilitate removal of kidney material. They were reweighed when needed and immediately layered with crushed ice until deboned. The mechanical deboner was a Bibun model SDX13 (Bibun Co., Fukuyama Kiroshima, Japan) belt type machine equipped with a 3mm 23 24 perforation size drum. The dressed carp were passed through the machine flesh side to the drum and the minced flesh was recovered with a plastic lug. On one occasion flesh temperatures were recorded prior to deboning, at the drum mouth and upon completion of the deboning operation. The minced fish was covered with plastic film and stored at 2°C until analyzed the same day or packaged and frozen for later analysis depending on the parameter to be tested. A pre- liminary trial was conducted to determine the need for passing the minced flesh through the machine a second time to remove remaining bone and scale residue. In this instance first and second pass samples were collected for yield, proximate composition, bone and scale, Hunter color difference, shear press shear force, TBA and total plate count data. Treatment Preparation Antioxidants A procedural flow diagram for preparation of control and antioxidant treatments is presented in Figure 2. Freshly minced carp flesh was mixed with a stainless steel paddle and combined with 0.02% w/v Tenox 2 (based on antioxidant content, not carrier), 0.18% w/v Freezgard (formula FP-88E) or no antioxidant in a Hobart paddle type mixer (Hobart Corp., Troy, Ohio) set at low speed for 3 minutes. Tenox 2 (Eastman Chemical Products, Inc., Kingsport, Tenn.) is a mixture of 20% butylated hydroxyanisole, 6% propyl gallate, 4% citric acid and 70% propylene glycol. Freezgard (Stauffer Chemical Co., Westport, Conn.) is a blend of sodium hexametaphosphate, NaCl and sodium erythorbate in undisclosed proportions. 25 MECHANICALLY DEBONED CARP TENOX 2 FREEZGARD CONTROL 0.02% 0.18% - MIX 3 MINUTES PACKAGE 759 PER WHIRLPAC BAG, FREEZE AND STORE AT -26°C TBA ANALYSIS AT 0, 1, 2, 3, 6, 9, 12 MO., -26°C Figure 2. Preparation of antioxidant treatments. 26 The control and two treatments were packaged in whirlpac bags (Nasco), approximately 759 per bag with free air removed in so far as possible, then identified, frozen and stored at ~260C. Samples were subsequently analyzed for TBA value at months 0,1,2,3,6,9 and 12, where month 0 was frozen for 1-3 days prior to analysis. The proce- dures were replicated 3 times using a different batch of fish for each replication. Corrective Treatments and Parameters Investigated Four water wash treatments were developed based on methods described by Patashnik et a1. (1976) and Kudo et a1. (1973) for plain water washing, defatting and dewatering of minced fish flesh. See Figure 3 for a flow diagram of the procedure. Freshly deboned carp was washed in ice water at a 4:1 weight ratio of ice water to fish. The mixture was manually agitated with a stainless steel paddle for 3 minutes, allowed to settle for 1 minute and the coagulated fat was skimmed off the surface. The supernatant was slowly poured off through a fiberglass window screen that was tied over a plastic tub for collecting wash water. The procedure was repeated with a 3:1 ice water to fish ratio. Following the second fat skimming step the entire slurry was poured onto the screen, then covered and dewatered by gravity 2-4 hours at 2°C. The dewatering process was encouraged by gentle agitation at 20 minute intervals. The washed mince was considered to be sufficiently dewatered by a subjective evaluation corresponding to a yield ranging from 78 to 82% of original whole mince weight. 27 CONTROL MECHANICALLY DEBONED CARP 2% NaCl BY WT. 0F MINCE ,x’ (0.5% w/v SOLUTION) WASH 4:1, ICE WATERzMINCE \ WASH 3:1, ICE WATER:MINCE - \ 2% NaHCO BY WT. 0F MINCE (0.5%3w/v SOLUTION) AGITATE 3 MINUTES SETTLE 1 MINUTE SKIM FAT 9 DISCARD mince supernatant ' POUR OFF SUPERNATANT THROUGH SCREEN water SOLUBLE. washed SCREEN-DEWATER 2-4 HRS., 2°C li-uid PROTEIN mince MIXING EVERY 20 MIN. DETERMINATION mix 5% ADDED HYDROGENATED mince 3 min. VEGETABLE FAT mix 3 min. Vr OBJECTIVE ANALYSES OF RAW AND COOKED MINCE SENSORY EVALUATION OF COOKED MINCE TBA OF RAW MINCE AT 0, 1, 2, 3, 6, 9, 12 MO., -26°C Figure 3. Preparation of four washed minced carp treatments. 28 Sodium chloride and sodium bicarbonate treatments were prepared as above except that either 2% NaCl or NaHCO3 by weight of mince was predissolved and added to the primary wash (0.5% w/v solution). The second wash was a plain ice water wash only followed by dewatering to 72-78% of prewash weight. The fourth treatment consisted of adding 5% w/v hydrogenated vegetable fat to plain water washed mince. The fat was composed of hydrogenated soybean, palm and cottonseed oils which did not contain antioxidants (Holsum Foods, Waukesha, Wisc.). The fish was added slowly to the vegetable fat while mixing at slow speed with a Hobart paddle type mixer, mixing a total of 3 minutes. A corresponding control was similarly mixed 3 minutes. All other wash treatments corresponded to an unmixed control. Water wash and control samples were packaged in polyethelene freezer bags and stored at -26°C for later comparison of bone and scale content (4 replicates); NaCl and control samples for NaCl determination (3 replicates). The wash waters were paddle agitated, then sampled and frozen in identi- fied capped vials for subsequent analysis for Biuret protein. A hydrogen peroxide bleaching procedure was prepared following the method of James and McCrudden (1976), Figure 4. Freshly deboned carp was mixed with 1.0% sodium tripolyphosphate and 0.85% H202 (by weight of mince) while mixing at medium speed in a Hobart paddle type mixer. The pH was adjusted to 10.5 with 5N NaOH as measured by a model 10 Corning pH meter (Corning Scientific Instruments, Medfield, Mass.) fitted with a Sargent glass pH electrode (S-30072-15, Sargent- Welch Scientific Co., Detroit, Mich.). After 3 minutes of mixing the mince was allowed to stand 12 minutes at 16°C, then adjusted to pH 29 mix CONTROL 5 min MECHANICALLY DEBONED CARP ADD 1.0% SODIUM TRIPOLYPHOSHATE AND 0.85% H202 ADJUST TO pH 10.5 WITH 5N NaOH MIXING 3 MIN. HOLD 12 MIN. AT 16°C ADJUST TO pH 6 0 WITH 3N CITRIC ACID MIXING 2 MIN. DECOMPOSE H202 WITH EXCESS CATALASE OBJECTIVE ANALYSIS OF RAW AND COOKED MINCE SENSORY EVALUATION OF COOKED MINCE TBA 0F RAW MINCE AT 0, 1, 2, 3, 6, 9, 12 MO., -26°C Figure 4. Preparation of hydrogen peroxide treatment. 30 6.0 with 3N citric acid while mixing an additional 2 minutes. Pre— dissolved bovine liver catalase (Sigma Chemical Co., St. Louis, Mo.) was added in excess near the end of mixing to decompose residual H202. A control was similarly mixed 5 minutes. An hexane extracted treatment was prepared as diagrammed in Figure 5. Cold hexane was added 2:1 (Hexane: mince, by weight) and manually stirred with a glass rod for 2 minutes. The mixture was filtered for approximately 5 minutes through four layers of cheese- cloth with a Buchner funnel attached to a vacuum flask and water aspirator. The mince was then spread in a thin layer (about 1.5cm) on cheesecloth suspended over a glass baking dish to assist in removal of hexane. It was transferred to a vacuum chamber (model 5831, National Appliance Co., Portland, Oregon) and evacuated to 30 in.hg vacuum with a vacuum pump (model 1405, Sargent-Welch Scientific Co.) for 3 hours at 16°C. The mince was mixed at 45 minute intervals during the vacuum step. Due to equipment capacity limitations a maximum of 6009 of mince (pretreatment weight) could be treated at one time and usually two batches were prepared per replication. Hydrogenated vegetable fat (previously described) was added to the extracted mince at the rate of 5% of the pretreatment mince weight by adding fish to the fat while mixing one minute at Slow Speed in a Hobart paddle type mixer. A control was similarly mixed for 1 minute. All "corrective" treatments were packaged, frozen, stored at -26°C and subjected to TBA analyses according to the procedure pre- viously described for antioxidant treatments. Samples were also CONTROL 1 me MECHANICALLY DEBONED CARP m1”. HEXANE : MINCE, 2:1 STIRRING 2 MIN. solvent FILTER —4> DISCARD mince REMOVE HEXANE, 30 IN. Hg VAC. 16°C, 3 HRS. 5% ADDED HYDROGENATED VEGETABLE FAT MIX 1 MIN. OBJECTIVE ANALYSES OF RAW MINCE 0 TBA AT 0, 1, 2, 3, 6, 9, 12 MO., -26 C Figure 5. Preparation of hexane extracted treatment. 32 Whirlpac or vacuum packaged and stored at -26°C up to 3 days for TBA analysis or longer when destined for proximate composition. Fresh samples, held covered up to 4 hours at 20C were analyzed for total plate count, TBA, Hunter color difference (L,a,b), shear force resist- ance, and pH. Other samples were vacuum packaged in barrier bags, frozen and stored at -26°C for 3 weeks prior to sensory evaluation or for 10 weeks prior to determination of heme pigment and cooking effects. Methods Microbial Analysis Minced carp flesh was analyzed for total plate count (aerobic and facultative microorganisms) by the procedure of Frazier et a1. (1968). Samples were removed from appropriate lots of mince and stored on stainless steel plates covered with plastic film in a 2°C cooler until plated (10 min. to 3 hrs.). The order that samples were plated was varied so that treatments were exposed to approxi- mately the same average holding time. Approximately 11 grams of sample was accurately weighed to two decimal places, blended 2 minutes in a single speed Waring blender (Waring Products Corp., Winsted, Conn.) with 99ml of 0.0003M sterile phosphate buffer. Appropriate serial dilutions were prepared by pipetting 1.0m1 of aliquot from previous mixed dilutions into 99ml sterile phosphate buffer blanks. Samples were plated in triplicate with 2 plates per dilution. Sterile plate count agar was held at 45°C prior to pouring onto the plates which were then swirled in a figure 8 pattern and allowed to cool. Plates were inverted, covered 33 loosely with aluminum foil and incubated 72 hours at room temperature (approximately 25°C). Plates containing 30-300 colonies were counted and total plate count recorded as microorganisms/g. All equipment and materials used in this procedure were autoclave sterilized except petri plates which were sterilized by the manufacturer (Miles Laboratories, Inc., Naperville, 111.). Bone and Scale Bone and scale content was determined by a method modified from one proposed by Wong and Yamamoto (1974). Triplicate 1009 samples of minced carp were stirred at room temperature (25°C) over- night in 2000ml of 3M urea and 0.02 N NaOH on a Multi-Magnestir (Lab-line Instruments Inc., Melrose Park, 111.). The contents were filtered through a 425 micron sieve and the residue resuspended in 50ml of the urea-NaOH solution for 4 hours with a magnetic stirrer. Bone and scale fragments were allowed to settle for 30 minutes. The liquid was carefully poured off through a 425 micron sieve and the remaining residue was washed several times with distilled water, then dried in a 105°C oven for several hours. Due to the presence of fatty material the residue was stirred with petroleum ether and filtered through dried and tared Watman No. 1 filter papter. After an initial period of air drying the samples were oven dried at 105°C for 12 hours, cooled in a desiccator and weighed. The cleaned residue weight (g) was recorded as percent bone and scale. 34 Moisture The moisture content of minced carp was determined by the A.0.A.C. (1965) procedure (page 3, paragraph 23.003). Each of four replicate subsamples of about 5 grams were weighed accurately into a tared and dried aluminum dish and then dried overnight (approximately 18 hours) in a convection oven at 105°C. The dried sample was trans- ferred to a desiccator until cool and accurately weighed. Weight loss was recorded as moisture and expressed as percentage of original weight. [at_ The extractable lipid fraction was determined by the Goldfisch extraction method described by the A.0.A.C. (1965) procedure (page 3, paragraph 23.005). The aluminum dishes containing the samples pre- viously analyzed for moisture were folded, placed into a porous porcelain thimble and extracted into a previously dried and tared beaker with anhydrous ether for 4 hours on the Goldfisch apparatus (Laboratory Construction Co., St. Louis, Mo.). The extract was dried for a minimum of 4 hours in a 105°C convection oven, then cooled in a desiccator and accurately weighed. Fat content was recorded as a percentage of fresh fish weight. Protein Protein content was calculated from the micro kjeldahl nitro- gen determination of the A.0.A.C. (1965, page 3, paragraph 23.009). Approximately 0.59 of minced carp were accurately weighed in triplicate into micro kjeldahl digestion flasks. To this was added lml 10% 35 CuSO4, lg anhydrous NaZSO4, 7ml H2504 and a few glass beads. The flasks were heated on a rotary kjeldahl digestion apparatus until the solution became transparent green, then digested an additional 30 minutes. The flasks were allowed to cool prior to adding 15 ml deionized water. Ten ml 2% boric acid and 3 drops Bromcresol green were added to erlenmeyer flasks which were mounted on distillation units such that the outlet tubes were positioned below the surface of the boric acid. The digestion flasks were attached to form a closed system and an excess of 44% NaOH solution was added to the sample flasks. Steam was immediately let into the system and allowed to distill for 7 minutes. The erlenmeyer flasks were then lowered and distillation was continued for another 3 minutes. The boric acid containing recovered ammonia was titrated to the Bromcresol endpoint with standard H2504. Percent protein was calculated by the formula: % protein =(m1 H2504) (normality of H2504) (14) (6.25) (100) weight of samfile, mg 4 Where, 1 = 6.25 molecular weight of nitrogen = 100/16% nitrogen in protein (assumed). Thibarbituric Acid Test (TBA) TBA numbers were determined by the method of Tarladgis et a1. (1960). Triplicate 109 samples of minced carp were homogenized for one minute with 50ml distilled water in a Virtis homogenizer (Model 6-105-AF, Virtis Co., Gardiner, N.Y.) set at medium speed. The slurry for each replicate was transferred to a 500ml distillation flask with 47.5ml distilled water and the pH was adjusted to 1.5 with 36 2.5m1 4 N HCl. Several glass beads and a few drops of Antifoam A (Dow Corning Corp., Midland, MI) were also added to the flask. The flask was connected to a 30.5cm distilling tube followed by a conden- ser unit. The sample mixture was heated and allowed to distill until the first 50 ml of distillate was collected. Two 5ml aliquots of the distillate were pipetted to test tubes to which were added 5ml of 0.02M thiobarbituric acid in 90% redistilled glacial acetic acid. The tubes were capped, agitated, and heated in a boiling water bath for 35 minutes and then cooled in cold water for 10 minutes. Color complex development was measured by absorbance at 538nm against a TBA reagent blank containing 5m1 distilled water instead of distillate. TBA value was calculated by multiplying mean absorbance by the distillation constant 7.8, and was recorded as m9 malonaldehyde/10009 sample. Shear Force Resistance of raw or cooked minced carp to Shear was deter- mined with a Kramer shear press (Model TR 3 Texurecorder, Food Technology Corp., Rockville, Md.) characterized by a ram descention speed of 0.52cm/sec. and fitted with a CS-l multiblade shear- compression cell and a 30001b compression ring. Triplicate 1009 (approximately) raw samples were weighed, spread across the bottom of the cell and sheared. Triplicate cooked samples were cut from por- tions with a template to fit the cell, then weighed (approximately 759) and sheared. Total force (compression plus shear) was recorded as lb shear force/g sample and was calculated from chart peak height by the formula: 37 1 (peak height 1b/9 force = (30001b ring) x (range) x 100' sample wt., 9 Where, range = chart recorder sensitivity setting. Color Color was characterized by a model 0 25-2 Hunterlab color difference meter (Hunter Associates Laboratory, Inc., Fairfax, Va.) which was standardized against a white standard. Two replicate 1009 samples were each measured twice for "L", "a" and "b" values and each replicate was rotated 45° between determinations. These values represent reflectance ranges from black to white (0 to 100, L value), green to red (-a to +a) and blue to yellow (-b to +b). The samples were usually taken from the Kramer shear press after shearing and were uniformly pressed into a transparent dish to minimize textural differences. Heme‘pigments The concentration of total heme pigments was determined by the combined methods of Rickansrud and Hendrickson (1967) and Fleming et a1. (1960). Duplicate 509 samples of minced carp were blended 3 minutes with 100ml of 0.01N cold (4°C) acetate buffer adjusted to pH 4.5. They were centrifuged at 2000x g for 15 minutes, the super- natant then filtered through 4 layers of cheese cloth with the aid of a Buchner funnel, vacuum flask and water aspirator. The funnel was cleared with approximately 75ml of the cold acetate buffer and the supernatant contrifuged again at 2000x9 for 15 minutes. This final supernatant was filtered through Watman No 1 filter paper and brought to volume with distilled water in a 250ml volumetric flask. To 20ml 38 of aliquot was added 2m1 K3Fe(CN)6 reagent (0.9m Mole/100ml) and 3ml of KCN reagent (8.0m Mole/100ml) to stabilize pigments. Absorb- ance was read at 540nm and converted to heme pigment content by: . . _ absorbance 17,000 x 0.3 x 1000 M9 total heme pigment/g flSh - 711,3007 sample wt., 9 Where,11,300 = molar extinction coefficent 17,000 = estimated equivalent weight of pigments 0.3 = volume of extract in liters Sodium Chloride Chloride as NaCl was determined in control and NaCl washed, minced carp by the Volhard titration procedure (Stine, 1978). Triplicate 109 samples were accurately weighed and transferred to erlenmeyer flasks. Twenty m1 of 0.1 N AgNO3 (standardized) followed by 20ml of nitric acid were added to each flask and then boiled under a hood until dissolved. Digestion of organic matter was assisted by adding concentrated KMnO4 as required. The solution was boiled until nearly colorless, then approximately 25ml distilled water was added and boiling continued for another 5 minutes. The flasks were cooled and distilled water was added making to about 100ml. A small quantity of acetone was poured down the inside wall of each flask to prevent resolution of precipitated AgCl. The flasks received approximately lml Fe(NH4)(SO4)2 indicator solution acidified with nitric acid. They were then titrated to a permanent light brick red endpoint with standardized 0.1N NH4SCN and calcu- lated for NaCl content with the following formula: 39 Percent NaCl==(Blank-NH&SCN, ml) (normality of NEASCN) (0.0584)(100) 'Samfile wt., 9 Where, 0.0584 = milliequivalent wt. of NaCl Blank = theoretical ml of NH4SCN needed to react with all AgNO3 (= 20ml if used 20ml 0.1N AgN03). Soluble Protein Solublized protein removed in wash treatment waters was deter- mined by the Biuret method of Goa (1953). Duplicate 0.1 or 0.2ml samples of wash water were brought to 2ml volume. Two ml of 6% NaOH solution was added, then the samples were mixed prior to receiv- ing 0.2ml Biuret reagent. After mixing again and allowed to stand for 15 minutes, absorbance was measured at 540nm. A standard curve was prepared with bovine serum albumin made to a concentration of 125mg BSA/25ml. Aliquots of the standard containing a range of O-Smg protein were measured by the Biuret procedure and the constant, mg protein/absorbance, was determined. Protein content of the wash waters was then calculated from the formula: absorbance of sample - absorbance of blank mgiprotein/EbSOFbance sample, ml mg protein/ml = ( Buiret reagent is made by the following procedure: 1. Dissolve 1739 Na citrate and 1009 Na2C03 in 500ml distilled water. 2. Dissolve 17.39 CuSO4°5H20 in 100ml distilled water. 3. Add solutions together and make to lOOOml - discard if reddish precipitate forms. 40 Cooking of Fish A control and 3 replicates of "corrective" treatments were prepared from one batch of fish and then cooked for 40 minutes at 175°C in covered aluminum baking dishes (351g/dish) with a convection oven (model CV912, Crimsco, Inc.). Due to residual solvent the hexane extracted treatment was not cooked. After removing from the oven the dishes were set on one end for 5 minutes and then on the opposite end for another 5 minutes to allow cooked out fluids to escape and the fish to cool. Triplicate samples were cut, weighed and sheared, then analyzed for Hunter L, a and b values. Yield of cooked flesh (percent of precook weight) was determinedonly once. SensoryAEvaluation Minced carp was mixed as one large uniform batch and prepared into the 6 "corrective" treatments plus a control. The hexane extracted mince was not utilized for taste panel study due to resi- dual solvent. The minces were vacuum packaged, frozen and stored at -26°C for 10 weeks prior to sensory evaluation. Judge Selection Twenty-three prospective taste panelists were screened for their ability to detect differences in flavor intensity of minced carp. Since two research projects had similar panel requirements, a system was employed to develop one screened panel for both sets of evaluations. A standard fish ball formulation was used to make two batches - one contained water washed minced carp and one contained unwashed minced carp. The fat content was adjusted to yield products 41 with similar proximate compositions. The balls were deep fat fried at 165°C to a uniform brown color, however some distinction between treatments was possible based on outside and inside color. Eight successive triangle tests were given each judge - 4 tests with 2 samples containing water washed mince against one sample with unwashed mince and 4 vice versa (see Appendix 8.1 for score sheet). The order of presentation of the 2 arrangements was determined randomly from a table of arrangements. The number of incorrect decisions (inability to detect the odd sample) was totaled for each panelist. Prior to evaluation it was tentatively decided that 4 incorrect judgments was acceptable but that 5 or more would be grounds for rejection. This was based on the test Characteristic, p = 1/3 by chance only. The criterion proved workable since after plotting the results, a bimodal distribution occurred among the judges with the second peak beginning at 5 or more errors (Figure 6). Five panelists were rejected on this basis, leaving 18 screened judges who were invited to attend 3 subsequent panel sessions. Attendance was reasonably good with 15 judges participating in the first session and 17 judges attending each of the last 2 sessions. Preparation and Evaluation of Treatments Fish treatments were evaluated without adding seasonings or other ingredients. Approximately 159 samples were portioned into individual foil pouches without added seasonings or other Number of judges 42 accept 18 *--+-9 reject 5 l O Figure 6. 2 4 6 8 Number of incorrect judgments Performance of prospective judges during screening trial, p29, to accept. 43 ingredients and were baked in a convection oven at 175°C for 15 minutes. They were then held at 93°C until just prior to serving. Each treatment was assigned a list of five 3-digit code numbers between 100 and 1000 from a random number table before every tasting session. The order that samples were placed around the paper serving plates was randomized differently for each panelist. The samples were served in individual aluminum weighing dishes which were coded with numbered tape tabs. Evaluation sessions were held at 10:00 a.m. or 3:00 p.m. in the taste panel facilities of the Food Science Building, Michigan State University, where 8 fluorescent lit partitioned booths were used adjacent to the preparation kitchen. Judges were supplied with water, plastic forks, napkins, spittle cups, pencils, a set of 4 score sheets and a plate of samples. Judges were informed to rank the six samples in the following order of stated categories: color intensity (dark to light), color preference (most preferred to least preferred), flavor intensity (strongest to weakest), flavor preference (most preferred to least preferred), texture (firmest to softest), texture preference (most preferred to least preferred) and overall preference (like best to like least). The score Sheets explained that the 6 samples could be evaluated as often as needed and in any order. Ranking as a means of sensory evaluation and its statistical analysis are discussed by Kahan et a1. (1973). 44 Statistical Analysis Analysis of variance tables were generated by submitting raw coded data into the genstat packaged program interfaced with the main computer system at Michigan State University. Analysis of variance for 2 way classifications containing unequal replications involved computer selection of best fit data by iterative least squares analysis which resulted in the loss of one or more degrees of freedom. This condition occurred in the lipid stability storage studies due to failure to determine month 0 TBA values for the first of 3 replica- tions in the 2 antioxidant treatments and to storage of 3 replications of control, plain water wash, NaCl wash and NaHCO3 wash treatments compared to 2 replications of wash plus fat, H202 and hexane extracted treatments. Where indicated by significant variance ratios, means were compared by Tukey's procedure (Steel and Torrie, 1960). Following the antioxidant storage study, the mean treatment TBA values were compared to the control by Dunnett's procedure (Steel and Torrie, 1960) as well as being compared to the control and each other with Tukey's method. Possible correlations among proximate composition, bone and scale, color, shear force, TBA and total plate count data following the deboning operation were investigated by applying simple corre- lation coefficients and consulting a table of significant r values (Snedecor and Cochran, 1967). Sensory determined mean ranks were analyzed by the method described by Kahan et a1. (1973) for ranked data. The procedure 45 is based on F statistics which when significant permit comparisons by least significant difference. Since the L.S.D. method is relatively liberal at detecting differences only means different at the 1% level were reported as significant. Statistical analyses were performed on actual measurements when possible or practical. However, since not all treatments were processed on the same day and mixed (agitated) treatments corresponded with mixed controls, not all "corrective" treatments could be directly compared. Consequently, values were expressed as percent of corre- sponding control or simple difference from corresponding control depending on the data analyzed. Comparing arithmatic differences also helped to reduce replication variance due to intrinsic charac- teristics unique to each batch of fish. CHAPTER IV RESULTS AND DISCUSSION Part I: Deboner Efficiency and Characteristics ofTMechanically DeboneOFCarp Based on research by Zapata (1978) a preliminary deboning trial was conducted to indicate the need for passing minced carp through the mechanical deboner a second time. Although an additional 30% reduction in bone and scale residue was realized after a second deboning pass, the first pass determination of bone and scale content was low, 0.034% (Table 2). This compares to 0.15% bone and scale for sucker following two passes through a deboner equipped with 5mm holes (Zapata, 1980). Table 2 also indicates that a loss of textural integrity (reduced shear resistance) resulted from the second pass, as well as a possible increase in lipid oxidation (elevated TBA value). Due to these preliminary findings a second deboning pass was not consi- dered valuable for carp and no further investigation was attempted. During a later deboning trial, temperatures were recorded as follows: dressed carp prior to deboning, 1.l°C; minced carp at machine drum exit, 4.4°C; maximum temperature of mince before returning to cooler, 8.9°C. It is reasonable that elevated temperatures encountered during a two pass process may accelerate proteolysis and lipid oxidation since reaction rates for both are temperature dependent. 46 47 .psmwmz cmwm open: mo acmogmem .psmwoz mmma umcme Go “smegmam .usmwoz commmcu mo acmugoap copxm.P pn.m Pup.o w.- m.m m.~e cmo.o m¢.mp pw.¢— mp.on Fe mm a.m— on eopxm.P m~.m mom.o m.PF ~.o~ m.o¢ emo.o N¢.mp m¢.mp m¢.wo w.o~ Am\m5mwcmmsov .aczoo mumpm Longs: Amxvvmmmmgo Amlvmpocz umgpu .copumoppamg Pu .meoamu Poopcmgums sauces“ mommea oz“ go one An memo nmoc_e mcpmmmuoga eo uummmm--. N m4m

“oz PA.m.nHNvupmw> Amxvmocwz A.8.awxvvpmw> Amxvvmmmmgo Amxvmposz .mcomumuwpamg m u.c..gmconwu Peuecwsums.;m=oggu commmuoga acme Go mupmw>u-.m m4mgo= agou uuconou appau—cagume mo muvumegmuuegezu-u.e u4m.umgmp. an umamewumo we: use mcwmmws we: mapm> x3 .Amo.vav u:memmw.n a.u=eu.w.:m.m mew mpg..um.mq=m ucmgmwe.n An umuecm.mmu mzo. cwguwz mw=.m>muune m~.. .eee.eae. x ee.. .m.o .eee.eee. mm.o wemh mcems eo mmocmgmw$.u mo msoggm useuceum tNP.m u~_.m a.m.w nmem.— mmm.P amps.— nomn.. cam: om.P me.N me.~ m... mm.~ m... om.o F¢—.p N xocmh mm.~ x3¢m.m mm.m om.~ m¢.~ om.P om.m Fw¢.~ unmmINmmLu o~.m zpw.n pm.¢ mm.m eo.m mm.p m~.N en.~ pogucou ewe: N. m e m N . o . eeee.eee. .oe .mE.» .pcmEummgp Lou mco.ueo..gmg m .9 mm- pm mmegoum cmNoge m:.e:u Am coo.\mv>;mu.eco.es as. .6353: 7.0) are generally characterized by a soft or "Sloppy" texture (Love, 1975; Kelly et a1., 1966). The soft condition reported in these cooked fillets is often accompanied by increased hydration. an increase in bound water (not necessarily total water) might serve to strengthen raw texture by swelling and yet inhibit formation of crosslinks between proteins when cooked. Evidence will be presented in a later section, "Effect of Cooking on Treatments" to support this explanation for NaHCO3 treated minced carp. No apparent tex- ture change resulted from H202 or hexane treatment (Table 9). 75 Effect on Microbial Status and Lipid Oxidation None of the treatments were significantly different from each other or their corresponding control for total plate count or TBA value (Table 9). Lee and Toledo (1977) also found no significant reduction in TBA value by washing minced fish if it was then stored fresh. However, the plain water, plus fat and NaHCO3 washes appeared to reduce TBA number slightly (Table 10), possibly due to loss of polyunsaturated lipids and removal of heme or metallic catalysts. By contrast, a slight prooxidant effect was introduced by NaCl treatment. Increased TBA numbers associated with H202 and hexane treatments may reflect the elevated temperatures encountered during processing or other factors that are discussed in a later section. Bacteria counts dropped after exposure to H202 which is sometimes used medicinally as a topical antiseptic. James and McCrudden (1976) believed antimicrobial activity to be an important benefit of high pH H202 bleaching of minced cod frames. Whether the 84% reduction in bacterial load observed in minced carp trans- lates to extended shelf life is not clear without further study (Table 9). Long exposure to warm conditions was probably responsible for an apparent increase of organisms following hexane treatment. Lower plate counts after washing was probably due to removal of bacteria with the wash water. 76 Bone and Scale The possibility that removal of oil and water soluble components during the wash procedure might concentrate bone and scale residue was investigated. Although an increase was noted, it was not significantly different than the control (Table 12). Lipid StabilitygDuring Frozen Storage The effect of all treatments on production of TBA reactive compounds during storage is summarized by difference from control in Table 13 and illustrated by actual TBA numbers in Figures 8-11. Analysis of variance for difference values produced F statistics that were significant for effect of time and treatment on TBA numbers (Appendix A.4). TBA values for months 9 and 12 were sig- nificantly higher than previously determined times (p<.05) and month 12 had higher TBA numbers than month 6 (p<.05, Table 13). Figures 8-11 suggest that the controls and hexane treatment were primarily responsible for the significant lipid oxidation after 6 months. Data for the 4 wash treatments are plotted on two figures (8 and 9) for clarity, however the control is the same and they are directly comparable. Analysis of treatment mean differences by Tukey's or Student- Neuman-Keul's procedure (Steel and Torrie, 1960) yielded no signi- ficant distinctions among treatments. However, when treatments were compared only to their control by Dunnett's procedure (Steel and Torrie, 1960) the NaHCO3 mean TBA number proved significantly lower 77 TABLE 12.--Bone and scale content as affected by water wash treatment, n=3 replicates. Treatment . Bone and Scale (%) Control 0.043a1 Water wash 0.0513 aMeans designated by different superscripts are significantly differ- ent (p<.05). 78 .882.» 8 88888..8888 .888.8 8 88.88..888. ..mo.va. p:88888.8 8.8888.8.=u.8 888 888888888888 888888888 88 888888.888 8:28.88 8.58.: 8:882x3 ..mo.va. ucmgmwm.n 8.8:8u.8.8888 888 888.88888888 8888888.8 .8 888888.888 8:88 88:88: 88882888 888.8 - 88888888. x 88.. 888.8 - 88.. 888.8 - 88888888. 8:885 .8 88888888888 88 8888.8,8888888m 888.8 8888.. 888..8 88..8 888.8 888.8- 88..8- e888 3x88.8 88.8 .8.. 88.8 88.8- 88.8- 88.8- 88.8- 8888888xe wc8xm: 3x88.8 88.8 88.. 8..8- 88.8- 8...- 88.8- 88.8- 8888: 3x88.. 88.8 88.. 8... 88.8 .8.8 88.8 8..8 8.8. 88.8 :88: .8883 388.. 88.8 88.8 88.. 88.8 88.8 88.8 88.8 .888; 888882 3x.8.8 88.8 88.. 88.8 .8.8 88.8 88.8 88.8- .8883 .882 .8.88.8 88.8 88.. 8... .8.8 88.8 88.8 8..8 .8883 8888: x88.8 88.8 88.8 88.8 88.8 88.8 88.8 88.8 ..888888 888: 8. 8 8 8 8 . 8 .82 .85.. .AacmEp8mgu mo 8885:: mocm -88...8 <8. 88 88888888 .8888- 88 8888888 88.888 88...8888 8.8.. 88 8888888888 .8 8888.8--.8. 8.88. .888 88888888> 888888888888 88888 88 88.8 888888: 88883 88 888888: 88883 88 8888888 8888 888885 88 8888- 88 8888888 888888 888888 88:.8> <88 A88880EV mth N— 8 o m N .8 888888 88 88 88 88 o .8 BID 8: 838 888883 88.8: 0 «8'8. 888883 88:: 0'0 ._o~:.zou 88 Cfizfll (60001/apfiuapleu019w 6w) V81 80 0 own an mmmgoum cmNogw mcwgau mmapm> \3 am.o cw qgmu umucms mcmcmmz $0 pumwmm .m mgsmwu Amzpcosv mth up a m m N P o mu \U'DJ- all-IIIIIIIIIIIIIIIImu hv xv fl I I ‘0 ... . .4. a. .., c 0 AV DID 8:23 mouzmz ”Val-lll”? om1m<3 Fonz Cl 0 ..omhzou (60001/apfiuapleu019w 5w) V81 81 ~_. .uomm- pm momgopm cm~og$ mcvgau agmu umucms mo mszm> mcmxm; we gummmm umamcmmoguxc umuum am mapa covpumgpxm Amcucoev wsz NF m m DID 85653 H.233: OIIcIO ..oEzoo [:10 DO 0\ .__mL=mwu D (60001/apfiuepleu019w 5w) V81 83 than the control during storage (Table l3). Treatment x time inter- action was not significant (Appendix A.4). The possible inhibition of lipid oxidation by washing can be attributed to reduced lipid content and smaller concentrations of heme and perhaps metallic prooxidants, however the additional pro- tection afforded by NaHCO3 washing is difficult to explain. Perhaps more efficient removal of hemeproteins discussed previously is sufficient to describe the improved stability. Also, pH is known to be related to activation and inhibition of heme and nonheme iron catalyzed oxidation. Hills (1965) determined that nonheme iron is only catalytic at acidic pH with an optimal pH of 5.5. Nonheme iron may be of enzymatic origin as in zanthine oxidase or of non- enzymatic origin as in ferritin. Consequently, the alkaline nature of the NaHCO3 treatment may eliminate nonheme iron as a prooxidant. The explanation may not be that simple and is complicated by evi- dence that hemeprotein catalysis is accelerated at alkaline pH (Fischer and Deng, l977). Use of NaCl or hydrogenated vegetable fat had no appreciable effect on TBA number compared to the plain water fish (Table 13). All of the wash treatments appeared to stabilize lipids and results indicate that washed mince carp resists TBA measured lipid oxida- tion for at least one year when properly packaged and stored. TBA values during the first 6 months of storage of the H202 treated mince were higher than for the washes compared to the controls (i.e. negative differences) (Table 13). A trend similar to the washes was observed in that little change occurred with 84 time which indicates inhibition of exponential propagation reactions. Several factors may have been involved. Yu and Sinnhuber (1964) observed a reduction in TBA value when malonaldehyde preparations were exposed to H202 and suggested that malonaldehyde might be oxidized to malonic acid. Hydrogen peroxide treatment was demonstrated to reduce the rate of lipid oxidation in chicken probably by oxidative destruction of heme- proteins (Lee et al., l975). If hemeproteins are major catalysts of lipid oxidation in minced carp as suspected, less change in TBA with time should be observed. Simultaneously, non-heme iron cata- lyzed oxidation may have been activated by the low pH (pH 6.0) pre- scribed by the H202 procedure (Fischer and Deng, l977). Other reasons why the H202 treatment might be expected to encourage lipid oxidation include the time of exposure to high temperature (l6°C) conditions, use of catalase (a hematin enzyme) to cleave residual H202, direct oxidation of lipids by H202 and the production of oxygen by the reaction: catalase Apparently none of these factors or their cumulative effect was sufficient to promote free radical propagation as measured by TBA analysis. Of all treatments, hexane extraction was least different from the control during 9 months of storage (Figure ll). Hydrogen- ated vegetable fat has been shown to be more stable than carp oil and the solvent was removed under vacuum which might minimize 85 initiation of lipid oxidation reactions. However, the samples were held approximately 3 hours at 16°C and some residual hexane remained after treatment. According to Labuza (l97l) unsaturated lipids oxidize at an increasing rate in solvents of decreasing polarity due to less competition between substrate and solvent for catalysts. Effect of Cooking Treatments were compared for relative effect of cooking on color and shear resistance. Because of residual hexane no attempt was made to cook hexane treated mince. Analysis of variance for Hunter color measurements produced significant variance ratios (F) for form (raw versus cooked), treatment and form x treatment inter- action except for Hunter "b" data which showed no form x treatment interaction (Appendices A.5-A.7). Heat denaturation of hemeproteins and permanent oxidation of heme iron account for the loss of color and graying noticed when meat is cooked (Bodwell and McClain, l97l). Consequently, the large mean increase of Hunter L number measured in the control upon cooking is almost certainly a result of high initial concen- tration of hemeproteins (Table l4). This effect of cooking was responsible for significantly more whitening in the control than in the treatments (p<.05). An implication is that products made from treatments designed to lighten color may not realize as much improvement when cooked as expected based on raw color. However, cooked color was still appreciably darker for the control than for the wash treatments and its "L" value for both forms (combined) was significantly smaller (darker) than the treatments (p<.05). 86 633.32 2... v .232 .338. 3 as 8.38 3..n ..ma.vav acouoye.v >_ucau.u.:o.n as. saga «no; =.zu.3 mua_cuucunau acacoop.u an ou.~:a.«ov cog: .vo4oou 33:.3 to; so; u:_a> coon o» acoamusgou moueogo...u oc.xa_—o~ nonvcunsoasmu ..mc.va. “encode—v xpucou,e_=o.m on. Aug» «nos c—ga_x «an.su»goaam “canoe..u an you-co.moo cog: .vo‘oou na—n tag ‘0 pouOu ca ucoamougou «cans Loans-sun o:.xo——o~ nun.suwuoaam— Au. -- 0.3 92 92 3.: 0.8 9: .32» 38 - .3..- on}: on"..- . anmd- gang? .26. .23: -- .sn.o uNP.m .oo.o aoo.o .sm.o okn.~ «o._ .om.n .m°._ p~n.n ..o.o gho._ .~n.o coco. Laden -- ...N- no..- . .~.p- . . . ..._- . a.._- go..- -- .a.o o.... .N... a~.~_ ..._ an N. _ _. an.~_ o.o. no.~_ .~... .o.o_ .o.o a - ..T ..n. no.7 of? o... a.m. -. .1 u .3. .3.- .. .3- .5 a so..- .3 do..- a... . do..- .2 o a: ...2 . -- ~.~. o.- ~.o- N... a.m. .°_- - A... a no.3 4.3 o $3 5.3 s $.33 .53 3 3.3 6.3 a 1.3 3.3 on 1.3 .2. 4 ....o oo‘ooo .a.. ...,e uuaooo .:.c ....u uo.ooo .:.c ....u oo.ooo .xoc ..e.o uo.ooo .xoc ..e.u uoeooo .xa. ~....u ipuogooo .3.. Loans...“ nuouooeu‘o agony: ~o~z cw.) nouzoz gm.) .uoz use . gag: couoz- In. gnu: Lug.) .oguceu acumwmogh - .moaou._aug nu: .uoom- an «awn: o. vocouw was co~och x—mso.>ocn .acou vouc_s voyage» co a:.aoOu ea uuuymm-.cp man

epm umoxmmz u m .go>mpe pmmmcogum u —m emesmmmga ammo? u m .umgemmmga umos u pm “mooeme_ u e .omoxeau u FF .Apo.vmv acmgmmmwv xpaceummwcmmm wee muawcomcmasm acmgmmmwc cum; m:E=_ou segue: mxcmmmuunm 91 omm.m ammo.e aoo.m omm.m oeo._ oem.m eoo.~ New: upm.m namo.e , amo.m apm.m gem.e aae.~ oo~.m mowwmn nag .1: £33 a; 386 i: am: is, :2 #3 3:. aem.~ oma.~ nme.~ nooo.~ eaom.¢ onmm.m new.e emwmn ae¢.m ama.~ nme.~ a_o.m a_m.e oNo.m amo.m gummy ame.m ae~.e ae¢.¢ ao~.m aem.m ame.m a8.. Poeocoo m.mmga ppmem>o «.mmea mgauxm» emespxm» N.emga eo>mpu mgo>wpu ~.mmea gopou Axcwg cmev ovumwgmpomgugo FLO you unmEHMmLh .mpcmsmvsn me .ggmu umucms uwummgp mo mmcwxcmg xgomcmm-.mp m4m

eo mogzom menacemwcmwm com: $0 saw eo mmmeumo . . _ .. _ “...mmmeoum :mNogw mcwcau cowumuwxo uwaPF naeammwsr eo mwmzpec<-.P.< xHazmao< 103 NaN.oNN om Peace ucmgw mNm.N wp_.~Np Nm peace mmN.~ Pmo.Ne em szuwmmm OF. pmm.P mNe.N nmo.mN NF pcmuwxopucm x meek Po. mem.o FNN.m mNm.ae m we?» po. mom.~_ Nmm.ep Nmm.nm o ~muoe Fo. aux.“ Nmu.m nee-mm a paaummmm mN.A mue.N Nmp.eN mNm.me N ucmuwxopuc< mo. 8: . ~83 8w. : N 858:8”. e eo e menacm .mmcmzam scammed cowumwcm> eo moesom mocmquwcmwm cam: eo 53m mo momemmo .mmmeopm canoe» mavezu cowumvwxo ovum, emesmmmsi mucmgmmmwu mo mucmmem> we mwmxpmc<-.N.< xmozmao< 104 N~.mens up pouch usage mp.Nme oc.Nw~o m— peach ne._a mm.NmN m poauwmmz —o. Nan.Np oN.moo— «p.mvoo m acmsuemgh mN. ome.n mm.owN «N.Fom N copuauvpnmz :Pouoca ea.—ocm— up ~ouo» vzoew Nm.m—m mm.essmp mp —euoh mF.NN os.oem m puauymmz —c. mom.on Nm.Nm—N mp.mN_mp u nauseous» —c. oNN._p mm.m—m me.NNm— N cowuuuVFgmx a Nu.—oom up ~uuoh venue um.nmp me.momN mp .uuoh Nm.me oe.N_¢ m —u=u*mo¢ —o. Nmo.m m_.o—v mo.noeN m acmsuuogh ma. omm.~ e—.oen aN.Nao N covuouppaoa assumvo: u an m ogeaom mogoscm scammed co*uov¢po> mo mugaom mucouvaco_m cam: ea 23m we awesome .mucusuoug» =o>vauucgou= mo covu.moasou «paevxocn so» encapsu> we mvmapuc eo mugsom mucmuwwwcmmm new: we sum mo mmmcmma .wmmgopm :mwoce mcwgan compmvwxo vwawp cmgammmE-mpumcgou= mo pummew esp mo mocmegm> to mmmmpmc<-.e.< x~azmaa< 106 Nee.Noe mm peach eeege eum.mN oNN.mPN om peach NeN.o eeN.m ON Feeewmem Fe. mpm.N~_ Noe.mN emo.me_ m pceEpeeeu x seem Fe. eme.Fme mw_.mpp Nee.mem m ucesueegp Po. eNo.NmN mem.oe oem.NmP m peach mN.A mmp.o mmo.o Pno.o N Feeewmem poo. NNm.NN_m mu¢.Nmp mu¢.Nm_ p Axe; .m> eexeeev seem mN.A mpN.F mpm.o mme.o N :e_ueu__eem e ee a eeeeem meeeeem seeeeem eewuepee> we eeceem eeeeewewcewm see: we saw ee meegmeo .eepe> 4 Levee: ee mewxeee we Humvee ee eeeewee> we mwmxpec eexeeev seem mN.A ON¢.o mNo.o Nmo.o N :ewueewpeem u ee u ecesem megeeem Eeeeeem :ewuewee> ee eeceem eeeeewwmemwm :eez ee Eem we meegmeo .eepgeeeueex ee meieee we flute ee 85.22, o6 memaPa=<-.e.< xeezmaa< 108 wmm.me mm Pepew eeeeu wom._ meo.mm om Page» mww.e eee.m 0N weeewmem mN.A com.o ome.o mwN.o m acesueeeu x age; we. eeN.mm ONo.w wmo.mm m peeeueeew _o. ome.m¢ mmo.m ewF.eN m .euew mN.A mmm.o mew.o mNm.o N weeewmex we. Nme.ee_ Nee.mN Nee.MN F Asa; .m> eoxooov seed mN.x eme.e eee.e wep.e N coweaow.aom m we w egeeem meeeeem Eeeeeew eewaewee> we eueeem eeeeewwwcmwm see: we Eem we meeemeo .eepe> e Lepeez ee mcwxeee we peewwe we eeeewee> we mwmxpee<-.w.< xHozmee< 109 wmm.mw mm Pepew eceew wme._ eem.me Om weeew OF0.0 FON.O ON peeewmem PO. ON0.00e ONO.¢ NPF.ON m pceeueecp x Ewe; —O. www.NOe mee.¢ ONN.mN m weeseeeew POO. New.mOOw ON0.0_ mmO.Nm m weeew ON.A eNO.w O_0.0 PN0.0 N weeewmem .ee. m_e.m__m m_e.Nm m_e.Nm _ fixed .3) eoxooov seem ON.A NN0.0 ON0.0 ON0.0 N :eweeewpeem ..— wo “— mwwscm mewzcm Eoummw... cozmwwm> we muwzom eeceewwwemwm new: we Eem we meegOeO .eeeew geese ee Oewxeee we ueewwe we eecewce> we mwmxpae<-.e.< xeezeaa< APPENDIX 8 SCORE SHEETS USED FOR SENSORY EVALUATION 110 APPENDIX 8 SCORE SHEETS USED FOR SENSORY EVALUATION Fish Evaluation Name Booth # Date Taste the three samples as often as needed to determine Flavor differ- ences. Two of these samples are the same, but one is different. Write the code of the samples which is different in the space below. Which of the samples do you Prefer for Flavor? Same ‘ , Different , No preference , Check One. Comments: APPENDIX B.l.--Score sheet used to screen panelists. 111 112 FISH COLOR EVALUATION NAME DATE BOOTH INSPECT THE SAMPLES AND ARRANGE AS NEEDED To DETERMINE COLOR DIFFERENCES. DO NOT TASTE THE SAMPLES- RANK THE SAMPLES IN ORDER OF COLQB_1NIENSLII. AND HRITE THE SAMPLE CODES BELOW- CODE DARKEST mmanP-a LIGHTEST RANK THE SAMPLES IN ORDER OF CQLQB_EB£E£BEHQE- WRITE THE SAMPLE CODES BELOW- CODE 1 MOST PREFERED COLOR 2 3 u S 6 LEAST PREFERED COLOR COMMENTS APPENDIX B.2.--Score sheet used to sensory evaluate color of “corrective" treatments. 113 FISH FLAVOR EVALUATION NAME DATE BOOTH TASTE THE SAMPLES AS OPTEN As NEEDED AND IN ANY ORDER YOU NISH. RANK THE SAMPLES IN ORDER OF ELAyOR_1N1ENSIIx, AND wRITE THE SAMPLE CODES BELOW- CODE STRONGEST FLAVOR aim-=WNl-l WEAKEST FLAVOR RANK THE SAMPLES IN ORDER OF ELA!QB.EBEEERENCE- WRITE THE SAMPLE CODES BELOW- CODE MOST PREFERED FLAVOR O'IU'I-DWNH LEAST PREFERED FLAVOR COMMENTS APPENDIX B.3.--Score sheet used to sensory evaluate flavor of "correc- tive" treatments. 114 FISH TEXTURE EVALUATION NAME BOOTH DATE TASTE THE SAMPLES As OFTEN As NEEDED AND IN ANY ORDER you HISH. RANK THE SAMPLES IN ORDER OF IEXIDB£_EIBMNESS- AND HRITE THE SAMPLE CODES DELow. CODE 1 FIRMEST 2 3 q S 6 SOFTEST RANK THE SAMPLES IN ORDER OF TEXTURE PREFERENCE- WRITE THE SAMPLE CODES BELON- CODE MOST PREFERED TEXTURE O'TU'lJ:\NNlI—| LEAST PREFERED TEXTURE COMMENTS APPENDIX B.4.--Score sheet used to sensory evaluate texture of "correc- tive" treatments. 115 FISH ACCEPTABILITY NAME BOOTH DATE EVALUATE THE SAMPLES AS OFTEN AS NEEDED AND IN ANY ORDER You NISH. RANK THE SAMPLES IN ORDER OF OVER-ALL EBEEEBENCE ACCORDING TO YOUR owN STANDARDS- WRITE THE SAMPLE CODES DELOH. CODE LIKE BEST CDU'l-DWND—I LIKE LEAST COMMENTS APPENDIX B.5.--Score sheet used to sensory evaluate over-all preferences of "corrective" treatments. BIBLIOGRAPHY 116 BIBLIOGRAPHY Aitken, A. l976. Changes in water content of fish during processing. Chem. and Industry 18: 1048. Anderson, M. L. l970. Technical notes for industry: new recommenda- tions for preservation of fish by freezing. Commercial Fish. Rev. 32(10): l5. Anderson, M. L. and Favest, B. 1969. Reaction of free fatty acids with protein in cod muscle frozen and stored at -29°C after aging in ice. J. Fish. Res. Board Can. 26: 2727. Anderson, M. L. and Mendelsohn, J. M. l97l. A study to develop new products from whiting or other underutilized species. Techni- cal Assistance Project. U.S. Dept. of Commerce. Angel, 5. and Baker, R. C. l977. A study of the composition of three popular varieties of fish in Israel, with a View towards further processing. J. Food Technol. l2: 27. Anon. l970.) Getting more meat from fish. Commercial Fish. Rev. 32 ll : 23. AOAC. 1965. "Official Methods of Analysis," lOth ed. Association of Official Agricultural Chemists, Washington, D. C. Apolinario, K. M. l975. Recovery and utilization of boneless flesh mechanically separated from tilapia, buffalo, and channel catfish. M.S. thesis, Auburn University, Auburn, Alabama. Asghar, A. and Pearson, A. M. l980. Influence of ante- and post- mortem treatments upon muscle composition and meat quality. Advances in Food Res. 26: 53. Atkinson, A. and Wessels, J.P.H. l975. Flavour and texture changes during the storage of frozen blocks of shredded hake. Fish. Ind. Res. Inst. Annual Report 29. Babbitt, J. K., Crawford, 0. L. and Law, 0. K. 1972. Decomposition of trimethylamine oxide and changes in protein extractability during frozen storage of minced and intact hake (Merluccius productus) muscle. J. Agr. Food Chem. 20: 1052. Baker, R. C. 1978. Fish, a wasted resource. New York's Food and Life Sci. ll(4): 12. 117 118 Baker, R. C. and Darfler, J. M. 1979. Tasty dishes from minced fish: development of products from minced fish. Booklet 5, New York Sea Grant Inst., State University of NY, Albany, NY. Baker, R. C., Regenstein, J. M., Raccach, M. and Darfler, J. M. 1977. Frozen minced fish; development of products from minced fish. Booklet 3. New York Sea Grant Inst., State University of NY, Albany, NY. Bello, R. A. and Pigott, G. M. 1979. A new approach to utilizing minced fish flesh in dried products. J. Food Sci. 44: 355. Beuchat, L. R. 1973. Hypoxanthine measurement in assessing freshness of chilled channel catfish (Ictalurus punctatus). J. Agr. Food Chem. 21: 453. Bligh, E. G. and Regier, L. W. 1976. The potential and limitations of minced fish. In "The Production and Utilisation of Mechanically Recovered Fish Flesh (Minced Fish)," p. 73. Conference pro- ceedings, Ministry of Agric., Fish. and Food, Torry Research Station, Aberdeen, Scotland. Bodwell, C. E. and McClain, P. E. 1971. Chemistry of animal tissues. In "The Science of Meat and Meat Products," 2nd ed-SILNF- Price and B. S. Schweigert eds. W. H. Freeman and Co. , San FranCIsco. Botta, J. R. and Richards, J. F. 1973. Thiobarbituric acid value, total long-chain free fatty acids, and flavor of Pacific halibut (Hippoglossus stenolepis) and Chinook salmon (Oncorhyncus tshawytscha) frozen at sea. J. Fish. Res. Board Can. 30: 63. Botta, J. R., Richards, J. F. and Tomlinson, N. 1973. Flesh pH, color, thaw drip, and mineral concentration of Pacific halibut (Hippoglossus stenolepis) and Chinook salmon (Oncorhyncus tshawytscha) frozen at sea. J. Fish. Res. Board Can. 30: 71. Botta, J. R. and Shaw, 0. H. 1975. Chemical and sensory analysis of roughhead grenadier (Macrourus berglax) stored in ice. J. Food Sci. 40: 1249. Bremner, H. A. 1977. Storage trials on the mechanically separated flesh of three Australian mid-water fish species: 1. analytical tests. Food Technol. Australia 29(3): 87. Briskey, E. J. and Kauffman, R. G. 1971. Quality characteristics of muscle as a food. In "The Science of Meat & Meat Products," 2nd ed., J. F. Price and B. S. Schweigert eds. W. H. Freeman and Co., San Francisco. Brotsky, E. and Everson, C. W. 1973. Polyphosphate use in meat and other muscle foods. In "Proceedings Of the Meat Industry Research Conference," p. 107. American Meat Science Associa- tion and American Meat Institute Foundation, Chicago, Ill. 119 Buttkus, H. 1967. The reaction of myosin with malonaldehyde. J. Food Sci. 32: 432. Cann, J. R. and Phelps, E. A. 1955. Binding of salt ions by bovine Y - pseudoglobulin. J. Am. Chem. Soc. 77: 4266. Castell, C. H., and Anderson, G. W. 1948. Bacteria associated with spoilage of cod fillets. J. Fish. Res. Board Can. 7: 370. Castell, C. H., Smith, B. and Neal, W. 1971. Production of dimethyl- amine in muscle of several species of gadoid fish during fish frozen storage, especially in relation to presence of dark muscle. J. Fish. Res. Board Can. 28: 1. Cole, B. J. and Keay, J. N. 1976. The development of rancidity in minced herring products during cold storage. In “The Produc- tion and Utilisation of Mechanically Recovered Fish Flesh (Minced Fish)," p. 66. Conference proceedings, Ministry of Agric.,dFish. and Food, Torry Research Station, Aberdeen, Scot an . Crawford, 0. L., Law, 0. K., and Babbitt, J. K. 1972a. Nutritional characteristics of marine food fish carcass waste and machine- separated flesh. J. Agr. Food Chem. 20: 1048. Crawford, 0. L., Law, 0. K. and Babbitt, J. K. 1972b. Yield and acceptability of machine separated minced flesh from some marine food fish. J. Food Sci. 37: 55. Cutting, C. L. 1969. Fish processing. In "Food Industries Manual," 20th ed., p. 213. A. Woolen ed., Leonard Hill Books, U. K. Dawood, A. A. 1979. Factors affecting the functional properties of fish muscle proteins. PhD dissertation, Michigan State Univer— sity, E. Lansing, Michigan. Dawson, L. E., Uebersax, K. L. and Uebersax, M. A. 1978. Stability Of freshwater sucker (Catostomus spp.) flesh during frozen storage. J. Fish. Res. Board Can. 35: 253. Deng, J. C., Matthews, R. F. and Watson, M. 1977. Effect of chemical and physical treatments on rancidity development of frozen mullet (Mugil cephalus) fillets. J. Food Sci. 42: 344. Dingle, J. R. and Hines, J. A. 1975. Protein instability in minced flesh from fillets and frames of several commercial Atlantic fishes during storage at -5°C. J. Fish. Res. Board Can. 32:775. Dingle, J. R., Keith, R. A. and Lall, B. 1977. Protein instability in frozen storage induced in minced muscle of flatfishes by mixture with muscle of red hake. Can. Inst. Food Sci. Technol. J. 10(3):143. 120 Duttweiler, M. W. 1978. Fish contaminants and human health. New York Sea Grant Bulletin, Cornell University, Ithaca, NY. Fairley, J. L. 1980. Personal communication. Michigan State Univer- sity, E. Lansing, Mich. Federal Registeru 19793. U.S. standards for grades of frozen minced fish blocks. 44(110):32388. Federal Registeru 1979b. Notice and solicitation of information: mechanically deboned poultry. 44(127):37965. Field, R. A. 1974. Mechanically deboned meat. In "Proceedings of the Meat Industry Research Conference," p. 35. Finne, G., Nickelson II, R., Quimby, A. and Connally, N. 1980. Minced fish fresh from nontraditional Gulf of Mexico finfish species: yield and composition. J. Food Sci. 45: 1327. Fischer, J. and Deng, J. C. 1977. Catalysis of lipid oxidation: a study of mullet (Mugil cephalus) dark flesh and emulsion model system. J. Food Sci. 42: 610. Fleming, H. P., Blumer, T. N. and Craig, H. B. 1960. Quantitative estimations of myoglobin and hemoglobin in beef extracts. J. Animal Sci. 19: 1164. Frazier, W. C., Marth, E. H. and Deibel, R. H. 1968. "Laboratory Manual for Food Microbiology." 4th ed., Burgess Publishing Co., Minneapolis. Giffee, J. W., Urbin, M. C., Fox, J. B., Landmann, W. A., Siedler, A. J. and Slivinski, R. A. 1960. Chemistry of animal tissues. In "The Science of Meat and Meat Products " 15? Ed-- D. 56- A.M.I.F. ed., W. H. Freeman and Co., San FranCTSco. Goa, J. 1953. A micro biuret method for protein determination: determination of total protein in cerebrospinal fluid. Scan. J. Clin. Lab. Invent. 5: 218. Goodrich, Jr., D. C. and Whitaker, D. B. 1977. Retail market tests of frozen minced fish. Public, A. E. Res. 77-6, Agric. Exp. Stn., Cornell University, Ithaca, NY. Grabouska, J. and Sikorsky, Z. 1973. Technological quality of minced fish preserved by freezing and additives. Acta Alimentaria 2 3 : 319. Hansen, Poul. 1972. Storage life of prepacked wet fish at 0°C: II. Trout and Herring. J. Food. Technol. 7: 21. 121 Hewson, D. and Kemmish, B. 1976. Some interesting uses for minced fish in processes of reforming, upgrading and extending. In "The Production and Utilisation Of Mechanically Recovered Fish Flesh (Minced Fish)," p. 84. Conference proceedings, Ministry of Agric., Fish. anb Food, Torry Research Station, Aberdeen, Scotland. Hirano, Y. and Olcott, H. S. 1971. Effect of heme compounds on lipid oxidation. J. Am. Oil Chem. Soc. 48: 523. Howgate, P. F. 1976. The sensory properties of minced cod and herring. In "The Production and Utilisation of Mechanically Recovered Fish Flesh (Minced Fish)," p. 49. Conference proceedings, Ministry of Agric., Fish. and Food, Torry Research Station, Aberdeen, Scotland. Hunter, A( g. 1933. Bacterial decomposition of salmon. J. Bacteriology 5 4 : 353. Iredale, D. G. and York, R. K. 1977. Effects of chemical additives on extending the shelf life of frozen minced sucker (Catostomidal) flesh. J. Fish. Res. Board Can. 34: 420. Iwata, K. and Okada, M. 1971. Protein denaturation in stored frozen Alaska pollock muscle 1. protein extractability and kamaboko forming ability of frozen surimi. Bull. Japanese Soc. Fish 37(12): 1191. James, A. L. and McCrudden, J. E. 1976. Whitening of fish with hydro- gen peroxide. In "The Production and Utilisation Of Mechanically Recovered Fish Flesh (Minced Fish)," p. 54. Conference pro- ceedings, Ministry of Agric., Fish. and Food, Torry Research Station, Aberdeen, Scotland. Jarenback, L. and Liljemark, A. 1975. Ultrastructural changes during frozen storage Of cod (Gadus morhua L.): 111. effects of linoleic acid and linoleic acid hydroperoxides on myofibrillar proteins. J. Food Technol. 10: 437. Jauregui, C. A. 1978. Discoloration problems in mechanically deboned fish. Summary of New York Sea Grant Thesis, p. 13. Cornell University, Ithaca, NY. Jauregui, C. A. and Baker, R.C. 1980. A research note: decoloration problems in mechanically deboned fish. J. Food Sci. 45: 1068. John, J. 1974. Some marketing considerations with respect to minced fish products. MFR paper 1107, Marine Fisher. Rev. 36(12). Kahan, 6., Cooper, 0., Papavasiliou, A. and Kramer, A. 1973. Signi- ficance of differences for ranked data. Food Technol. 27(5): 59. 122 Ke, P. J., Ackman, R. G., Linke, B. A. and Nash, D. M. 1977. Differen- tial lipid oxidation in various parts of frozen mackerel. J. Food Technol. 12: 37. Kelly, K., Jones, N. R., Love, R. M. and 011ey,J. 1966.Texture and pHin fish muscle related to "cell fragility" measurements. J. Food Technol. 1: 9. King, F. J. 1974. Improving the supply Of minced blocks for the fish stick trade:- a progress report. MFR paper 998, Marine Fish. Rev. 35(8). King, F. J. ca 1978. Past, present and fugure uses of minced fish in the United States. Presented at the Aberdeen (III) conference on the utilization of minced fish, Aberdeen, Scotland. King, F. J. and Carver, J. H. 1971. How to use nearly all the ocean's food. Commercial Fish. Rev. 32(12): 12. Keay, J. N. 1979. Minced fish. Torry Advisory Note No. 29. Torry Research Station, Aberdeen, Scotland. Kropf, D. H. 1980. Personal communication. Kansas State University, Manhattan, Kan. Kudo, G., Okada, M. and Miyauchi, D. 1973. Gel-forming capacity of washed and unwashed flesh of some Pacific coast species of fish. MFR paper 1021, Marine Fish. Rev. 35(12): lO. Kwon, T. W., Menzel, D. B. and Olcott, H. S. 1965. Reactivity of malonaldehyde with food constituents. J. Food Sci. 30: 808. Labuza, T. P. 1971. Kinetics of lipid oxidation in foods. CRC Crit. Rev. Food Technol. 2(3): 355. Lagler, K. F., Bardach, J. E. and Miller, R. R. 1962. "Ichthyology," John Wiley & Sons, Inc., N.Y. Lall, B. S., Manzer, A. R. and Hiltz, D. F. 1975. Preheat treatment for improvement of frozen storage stability at -10°C in fillets and minced flesh of silver hake (Merluccius bilinearis). J. Fish. Res. Board Can. 32(8): 1450. Lanier, T. C., Lin, T. 5., Hamann, D. D. and Thomas, F. B. 1980. Biological and processing factors affecting functionality: Gel formation in comminuted fish systems. Presented at the Third National Technical Seminar on Mechanical Recovery and Utiliza- tion of Fish, Raleigh, N.C. Lanier, T. C. and Thomas, F. B. 1978. Minced fish: its production and use. Public. UNC-SG - 78-08, Sea Grant College Program, University of North Carolina, Raleigh, N. C. 123 Lawrie, R. A. 1966. Chemical and biochemical constitution Of muscle. In "Meat Science," p. 66. Pergamon Press, New York, N.Y. Lee, C. M. and Toledo, R. T. 1979. Processing and ingredient influ- ences on texture of cooked comminuted fish muscle. J. Food Sci. 44: 1615. Lee, C. F. 1963. Processing fish meal and oil. In "Industrial Fishery Technology." C. F. Stansby ed., p. 219. Reinhold Publishing Corp., New York, N.Y. Lee, Y. B., Horgus, G. L., Kirpatrick, J. A., Berner, D. L. and Forsythe, R. H. 1975. Mechanism of lipid oxidation in mechan- ically deboned chicken meat. J. Food Sci. 40: 964. Lee, C. M. and Toledo, R. T. 1976. Factors affecting textural Charac- teristics of cooked comminuted fish muscle. J. Food Sci. 41: 391. Lee, C. M. and Toledo, R. T. 1977. Degradation of fish muscle during mechanical deboning and storage with emphasis on lipid oxida- tion. J. Food Sci. 42: 1646. Lin, 1. S. and Lanier, T. C. 1980. Properties of an alkaline protease from the skeletal muscle of Atlantic croaker. J. Food Biochem. 4: 17. Lin, T. 5., Sue, H. K. and Lanier, T. C. 1980. Characterization of fish muscle proteases using radio-labled protein substrates. J. Food Sci. 45: 1036. Lindsay, R. C. 1975. A summary of the effect of antioxidants and film packaging on oxidative quality of fish during long term frozen storage. A 10 p. manuscript summarizing M.S. thesis by Steven J. Teets, University of Wisconsin, Madison, Wisc. Love, M. R. 1975. Variability in Atlantic cod (Gadus morhua) from the northeast Atlantic: a review of seasonal and environmental influences on various attributes of the flesh. J. Fish. Res. Board Can. 32: 2333. Love, J. D. and Pearson, A. M. 1974. Metmyoglobin and nonheme iron as prooxidants in cooked meat. J. Agr. Food Chem. 22: 1032. Mahon, J. H. 1961. Tripolyphate salt synergism and its effect on cured meat volume. Proceedings 13th Research Conf. American Meat Institute Foundation, Chicago, 111., p. 59. Mai, J. and Kinsella, J. E. 1979 . Lipid composition of dark and white muscle from white sucker TCatostomus commersoni). J. Food Sci. 44: 1101. 124 Mai, J. and Kinsella, J. E. 1979b. Changes in lipid composition of cooked minced carp (Cyprinus carpio) during frozen storage. J. Food Sci. 44: 1619. Manohar, S. V., Rigby, D. L. and Dugal, L. C. 1973. Effect of sodium tripolyphosphate on thaw drip and taste of fillets of some freshwater fish. J. Fish. Rev. Board Can. 30: 685. Markakis, P. 1979. Personal communication. Michigan State Univérsity, E. Lansing, Mich. Mattingly, R. L. and Kevern, N. R. 1979. Life history, abundance and potential harvest of carp (Cyprinus carpio Linnaeus) from Michigan water's of the Great Lakes. Public. Michu -SG-79-205, Michigan Sea Grant Program. ‘ McGill, A. S., Hardy, R. and Gunstone, F. D. 1977. Further analysis of the volatile components of frozen cold stored cod and the influ- ence of these on flavor. J. Sci. Food Agric. 28: 200. Mendelsohn, J. M. 1974. Minced fish in a new form. MFR paper 1078, Marine Fish. Rev. 36(8):34. Mendelsohn, J. M. 1974. A study: expanded processing techniques, production costs and market survey of underutilized fish species. Technical Assistance Project, U.S. Dept. of Commerce. Merritt, J. H. 1974. Biological aspects on the cooling and freezing of - fish. In “Refrigeration applications to Fish, Fruit and Vege- tables in South East Asia,“ p. 142. Food. Agric. Organ. of the U.N.-Internationa1 Institute of Refrigeration. Miyauchi, D. 1972. Procedure for preparing modified minced-fish blocks. Informational report, U.S. Dept. of Commerce, Pacific Fishery Products Center, Seattle, Wa. Miyauchi, 0., Kudo, G. and Patashnik, M. 1973. Surimi - a semi wet fish protein. MFR Paper 1020, Marine Fish. Rev. 35(12). Miyauchi, D., Patashnik, M. and Kudo, G. 1975. Frozen storage keeping quality Of minced block rockfish (Sebastes spp.) improved by cold water washing and use of fish binder. J. Food Sci. 40: 592. Miyauchi, D. and Steinberg, M. 1971. Machine separation of edible flesh from fish. Fish. Industrial Res. 6(4): 165. Miller, III, A. Scanlan, R. A., Lee, J. S. and Libbey, L. M. 1972. Volatile compounds produced in ground muscle tissue of canary rockfish (Sebastes pinniggr) stored on ice. J. Fish. Res. Board Can. 29: 1125. 125 Moledina, K. H., Regenstin, J. M., Baker, R. C. and Steinkraus, K. H. 1977. Effects of antioxidants and chelators on the stability of frozen stored mechanically deboned flounder meat from racks after filleting. J. Food Sci. 42: 759. Morris, D. M. and Dawson, L. E. 1979. Storage stability of mechanically deboned sucker (Catostomidae) flesh. J. Food Sci. 44: 1093. Nair, R. B., Tharamani, P. K. and Lahiry, N. L. 1971. Studies on chilled storage Of fresh water fish: 1. changes occuring during iced storage. J. Food Sci. Technol. 8: 53. Nair, R. B., Tharamani, P. K. and Lahiry, N. L. 1974. Studies on chilled storage of fresh water fish: II. factors affecting quality. J. Food Sci. Technol. 11: 118. Nakayama, T. and Yamamoto, M. 1977. Physical, chemical and sensory evaluations of frozen-stored deboned (minced) fish flesh. J. of Food Sci. 42: 900. Nicholas, R. C. 1979. Personal communication. Michigan State UniVer- sity, E. Lansing, Mich. Nickelson II, R., Finne, 6., Hanna, M. O. and Vanderzant, C. 1980. Minced fish flesh from nontraditional Gulf of Mexico finfish species: bacteriology. J. Food Sci. 45: 1321. - Noble, J. 1973. Retrieve meat wastes to build new foods. Food Engineering 45(10): 76. Noble, J. 1974. New protein sources from the sea. Food Industries of South Africa. 26(15): 7. Okada, M., Miyauchi, D. and Kudo, G. 1973. Kamaboko -the giant among Japanese processed fishery products. MFR paper 1019, Marine Fish. Rev. 35(12). Patashnik, M. 1978. Unpublished data. Pacific Fisheries Utilization Research Center, Seattle, Wash. Patashnik, M. 1974. Utilization of minced carp muscle. In "Second Technical Seminar on Mechanical Recovery and Utilization of Fish Flesh," p. 28. Boston, Mass., June 12-13. Patashnik, M., Kudo, G. and Miyauchi, D. 1973. Smooth white spread from separated fish flesh forms a base for flavored dips, snack items. Food Prod. Devel. 7(6): 82. Patashnik, M., Kudo, G. and Miyauchi, D. 1974. Bone particle content of some minced fish muscle products. J. Food Sci. 39: 588. 126 Patashnik, M., Miyauchi, D. and Kudo, G. 1974. Controlling bone . particle content in minced fish muscle. MFR paper 1079, Marine Fish. Rev. 36(8): 37. Patashnik, M. Miyauchi, D. and Kudo, G. 1976. Objective evaluation of texture of minced black rockfish (Sebastes spp.) during frozen storage. J. of Food Sci. 41: 609. Pedraja, R. R. 1970. Change of composition of shrimp and other marine ' animals during processing. Food Technol. 24: 1355. Price, J. F. 1981. Personal communication. Michigan State University, E. Lansing, Mich. Raccach, M. and Baker, R. C. 1978. Microbial properties of mechanically deboned fish flesh. J. Food Sci. 43: 1675. Raja, K. C. M., Moorjani, M. N. 1971. Nucleotides- a study of their degradation in fish under various conditions and its Impact on flavour. J. Food Sci. and Technol. 8(1): 3. Ramamurthy, M. S., Kumta, U. S. and Lewis, N. F. 1976. Quality assessment of freeze-dried white pomfret (Stromateus Cinereus) during storage. Levensm.-Wiss. II Technol. 9(3): 176. Ravichander, N. and Keay, J. N. 1976. The production and properties ' of minced fish from several commercially important species. In "The Production and Utilisation of Mechanically Recovered Fish Flesh (Minced Fish), p. 18. Conference proceedings, Ministry of Agric., Fish. and Food, Torry Research Station, Aberdeen, Scotland. Reay, G. A. and Shewan, J. M. 1949. The spoilage of fish and its preservation by chilling. In "Advances in Food Research," Vol. 2, p. 343. Torry Research Station, Aberdeen, Scotland. Reseck, J. and Waters, M. 1979. Acceptability, stability of under- utilized fish species suggest commercial potential. Food Prod. Develop. 13(5): 46. Rickansrud, D. A. and Henrickson, R. L. 1967. Total pigments and myoglobin concentrated in four bovine muscles. J. Food Sci. 32: 57. Rippen, T. E. 1980. Understanding Contaminants in Fish. Extension bull. E-1434, Michigan State University. E. Lansing, Mich. Rust, R. E. and Olson, D. G. 1973. "Meat Curing Principles and Modern Practice," Koch Supplies, Inc., Kansas City, MO. 127 Sadowska, M. and Sikorski, 2. E. 1975. Interaction of different animal proteins in the formation of gels. Lebensm-Wiss. U-Technol.!1:20. Schmidt, P. J. and Cathbert, R. M. 1971. Colour sorting of Pacific salmon. In "Fish Inspection and Quality Control," p. 104. R. Kreuzer ed., Fishing News (Books) Limited, London. Setty, T. M. R., Muddanna, V. and Shetty, H. P. C. 1974. A new method for incorporation of fish meat into domestic products and its prospects. Mysore J. Agric. Sci. 8: 466. Shewan, J. M. 1962. The bacteriology of fresh and spoiling fish and some related chemical changes. In "Recent Advances in Food Science," Vol. I, p. 167. J. Hawthorn, J. Muil eds. London: Butterworths. Silberstein, D. A. and Lillard, D. A. 1978. Factors affecting the autoxidation of lipids in mechanically deboned fish. J. Food Sci. 43: 764. Simidu, W. 1961. Nonprotein nitrogenous compounds. In "Fish as Food," Vol. I, p. 353. G. Borgstrom ed., Academic Press, New York, N.Y. Simon, I. B. 1971. Computer-aided prediction of food storage stability: ~oxidative deterioration of a shrimp product. J. Food Sci. 36: 280. Sims, G. G., Cosham, C. E. and Anderson, W. E. 1975. Hydrogen peroxide bleaching of marinated herring. J. Food Technol. 10: 497. Siu, G. M. and Draper, H. H. 1978. A survey of the malonaldehyde content of retail meats and fish. J. Food. Sci. 43: 1147. Snedecor, G. W. and Cochran, W. G. 1967. "Statistical Methods," 6th ed. The State University Press, Ames, Iowa. Stansby, M. E. 1973. Data for selecting species to meet special dietary needs: polyunsaturates and fat in fish flesh. J. Amer. Diet. Assoc. 63(6): 625. Steel, R. G. and Torrie, J. H. 1960. "Principles and Procedures of Statistics," McGraw-Hill, New York, N. Y. Steinberg, M. A. 1974. Comminuted fish flesh. Alaska Seas and Coasts 2(3): 1. Alaska Sea Grand public., U. of Alaska, Fairbanks. Steinberg, M. A. 1975. USDC issues detailed report on uses Of minced fish. Quick Frozen Foods Intern. 16(3): 115. 128 Stine, C. M. 1978. Personal communication. Michigan State Univer- sity, E. Lansing, Mich. Tainter, S. and White, R. J. 1977. Seines to salmon Charters: 150 years of Michigan Great Lakes fisheries. Extension bull. E1000, Michigan State University, E. Lansing, Mich. Tarladgis, B. G., Watts, B. M., Younathan, M. T. and Dugan, L. 1960. A distillation method for the quantitative determination of malonaldehyde in rancid foods. J. Am. Oil Chem. Soc. 37: 44. Teeny, F. M. and Miyauchi, D. 1972. Preparation and utilization of frozen blocks of minced black rockfish muscle. J. Milk and Food Technology 35(7): 414. Ueda, T. Shimizu, Y. and Simidu, W. 1968. Species difference in fish muscles. Bull. Japanese Soc. Sci. Fish. 34(4): 357. US Dept. of Agriculture. 1980. National food review, spring. Washington, D.C. USDA. 1963. "Composition Of Foods: Raw, Processed, Prepared," Agriculture Handbook No. 8, United States Department of Agriculture, Washington, D.C. USDC. 1980. What causes soft sablefish from deep water? Activities report, Northwest and Alaska Fisheries Center, Seattle, Wash. Waterman, J. J. 1980. Fish smoking: a dictionary. Torry Advisory Note 83, Torry Research Station, Aberdeen, Scotland. Watts, B. M. 1950. Polyphosphates as synergistic antioxidants. J. Am. 0i. Chem. Soc. 27: 48. Webb, N. B., Hardy, E. R., Giddings, G. G. and Howell, A. J. 1976. Influence of mechanical separation upon proximate composition, functional properties and textural characteristics of frozen Atlantic croaker muscle tissue. J. Food Sci. 41: 1277. Wills, E. D. 1965. Mechanisms of lipid peroxide formation in tissue: role of metals and hematin proteins in the catalysis of oxida- tion Of unsaturated fatty acids. Biochem. Biophys. Acta 98: 238. Wong, J., Lau, Y. C. and Yamamoto, M. 1978. Mechanical fish deboners: influence of perforation sizes on bone content and texture of minced fish flesh. J. Food Sci. 43: 807. Wong, J. and Yamamoto, M. 1974. Simple chemical method for isola- ting bone fragments in minced fish flesh. J. Food Sci. 39: 1259. 129 Wood, G., Hintz, L. and Salwin, H. 1969. Decomposition and fifth in foods (Chemical indexes). J.A.O.A.C. 52: 904. Wood, G. and Hintz, L. 1971. Decomposition in foods: lipid changes associated with the degradation of fish tissue. J.A.O.A.C. 54: 1019. Younathan, M. T. and Watts, 8. M. 1960. Oxidation of tissue lipids in cooked pork. Food Research 25: 538. Yu, T. C., Landers, M. K. and Sinnhuber, R. O. 1969. Storage life extension of refrozen silver salmon steaks. Food Technol. 23: 100. Yu, T. C. and Sinnhuber, R. 0. 1967. An improved 2-thiobarbituric acid (TBA) procedure for the measurement of autoxidation in fish oil. J. Amer. Oil Chem. Soc. 44: 256. Yu, T. C. and Sinnhuber, R. O. 1964. Further Observations on the 2-thiobarbituric acid method for the measurement of oxida- tive rancidity. J. Amer. Oil Chem. Soc. 41: 540. Zapata, J. F. 1978. The functional properties of the mechanically deboned sucker (Catostomidae family) flesh. M.S. thesis, Michigan State University, E. Lansing, Mich.