£5525 WHHINIWIHHNIWIHIHHIWHIMHIHIIIHIUII llllllllllllllllllllilllllllllllllllllI ‘ 3 1293 10772 4878 LIBRARY IfificifigguaESaagg University észfiimazsw ‘~' :3” This is to certify that the thesis entitled THE FUNCTIONAL PROPERTIES OF THE MECHANICALLY DEBONED SUCKER (CATOSTOMIDAE FAMILY) FLESH presented by Jorge Fuentes Zapata has been accepted towards fulfillment of the requirements for Master of Science degree in Food Science and Human Nutrition Q Major professor Nov. 3, 1978 OVERDUE FINES ARE 25¢ PER DAY PER ITEM Return to book drop to remove this checkout from your record. THE FUNCTIONAL PROPERTIES OF THE MECHANICALLY DEBONED SUCKER (CATOSTOMIDAE FAMILY) FLESH By Jorge Fuentes Zapata A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Food Science and Human Nutrition 1978 ABSTRACT THE FUNCTIONAL PROPERTIES OF THE MECHANICALLY DEBONED SUCKER (CATOSTOMIDAE FAMILY) FIESH By Jorge Fuentes Zapata Sucker fish, an underutilized freshwater species from the Great Lakes area,were mechanically deboned and the cooked minced flesh evaluated for water holding capacity (WHC) and texture by using various techniques. The effect of the handling of the fish prior to freezer storage as well as the freezing conditions throughout a 13 months period were related to the functional properties of the fish. The centrifuge technique for WHC and the Instron uni- versal testing instrument for texture were found to be the most reliable methods to evaluate the cooked fish matrix. Severe water losses during cooking resulted in fairly poor texture characteristics for the flesh. However, WHC was improved by the addition of 2% salt or 2% soy protein isolate to the control binder system used in this study. The best storage procedure for suckers intended for further processing into fabricated fish products seemed to be dressing, washing and storage in blast freezer at -29° C. ACKNOWLEDGMENTS The author wishes to express his sincere appreciation to Dr. J.F. Price, Professor of Food Science for his help in setting up the project and his invaluable assistance during the completion of the study. Gratitude is also extended to Dr. A.E. Reynolds and Dr. L.E. Dawson of the Department of Food Science and Human Nutrition and to Dr. N. Kevern of the Department of Fisheries and Wildlife for serving on the guidance committee and for reviewing the thesis. Acknowledgments are due to C. LeBlanc for performing the determination of TBA values, bone content and bacterial counts in this study and to Dr. M. Mostafavi for his sug- gestions on the statistical analysis of the results. The author feels grateful to the Ford Foundation, Rio de Janeiro, Brazil for the opportunity and financial assistance granted to him. Finally, the author would like to express most sincere appreciation to his parents for their constant encouragement and to his wife and son for their unfailing assistance and understanding throughout his graduate education. ii TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES INTRODUCTION LITERATURE REVIEW Underutilized Freshwater Fish and Mechanical Deboners . . The Importance of the Structural Proteins of Fish Muscle on the Acceptability of Minced Flesh Products The Effects of Freezer Storage on the Keeping Quality of Fish Flesh . Water Holding Capacity and Texture as Indicators of the Gel- -Forming Ability of the Minced Fish Flesh . . . . . . . . . . . EXPERIMENTAL Classification of the Experimental Work Into Two Study Sections . . . . . . . . . . . . . Section One. The study of the mechanical deboning operations, the binders and the cooking systems as related to texture and water- -holding capacity measurements . Section Two. The effect of the handling and freezer storage conditions on the func- tional properties of the sucker flesh Handling of Fish at Laboratory and Processing Operations . . . . . Dressing and Mechanical Deboning . . Packaging of Fish Material for Freezing Operations . Binders and Binding operations Cooking of the Fish Paste iii Page vii ll 15 15 15 17 19 19 19 21 Techniques and Methods of Analysis Chemical Methods . . . Moisture Content Fat Content Protein Content . . 2- Thiobarbituric Acid (TBA) Test . Bone Fragment Content . Microbiological Method Physical Methods . Measurements Related to Water Holding Capacity . . Thawing drip . . . Smokehouse shrinkage Water losses during water bath cooking : Expressible fluids by the filter paper press method . . . Expressible fluids by the. centrifuge technique . . . Measurements Related to Texture . Shear force by the Kramer Shear Press Shear force by the Instron universal testing instrument . . . Softness by the Penetrometer Statistical Analysis Correlation Coefficient Analysis of Variance RESULTS AND DISCUSSION Section One Section Two SUMMARY REFERENCES CITED . iv LIST OF TABLES Table 1. Some of the physical, chemical and microbio- logical characteristics of the sucker flesh obtained by mechanical deboning Mean and standard deviation of the bacterial load on the surface of the suckers, expres- sed as the log of the number of bacteria/cm2 fish surface . . . . Modifications of the mechanical deboned sucker flesh by washing, use of binders or by the mechanical stress applied to the flesh during the process of blending with binders . . Mean and standard deviation of the water- holding capacity of the cooked sucker flesh by four different techniques and expressed as water losses percent Mean and standard deviation of the shear force by the Instron and by the Kramer shear press, and of the softness by the Penetrometer as a measure of texture of the smokehouse-cooked sucker flesh Correlation coefficients (r) obtained by com- paring the techniques used to assess the functional properties of the mechanically deboned sucker flesh . . Yields of dressing and mechanical deboning operations (Z) as affected by a 13 month freezer storage period . . . . . . . . Drip values of the mechanically deboned sucker flesh through 13 months of frozen storage Drip of the mechanically deboned sucker flesh obtained from fish handled and freezer stored in seven different ways Page 34 36 38 39 44 47 50 53 54 Table 10. 11. 12. 13. Mean values of the bacterial load in the minced sucker flesh obtained from fish stored in freezer for 13 months Means of the TBA values of the mechanically deboned sucker flesh as affected by han- dling and a 13 month period of frozen storage Mean values and standard deviations of the functional properties of the mechanically deboned sucker flesh as affected by han- dling and freezer storage conditions . Mean values and standard deviations of the functional properties of the mechanically deboned sucker flesh as affected by the kind of binder incorporated vi Page 56 57 59 61 LIST OF FIGURES Figure Page 1. Results of the interaction between pre-freezing handling and binder effects on the texture of the smokehouse-cooked sucker flesh . . . . . . 62 vii INTRODUCTION In the recent few years a significant decrease in the availability of many commercially exploited fish species was noted in several traditional Atlantic fishing grounds. In order to supply the increasing market demands for fish and shellfish attempts are made to utilize several abundant, less valuable species, previously not fished on an industrial scale. Fish that are underutilized, whether they be of marine or freshwater origin, usually have certain defects in qual- ity which preclude their gaining consumer acceptance when marketed in the conventional forms, such as dressed fish, fillets, breaded sticks, or portions. The defects may be related to the different morphological characteristics of these fish and/or their less appealing sensory properties. The fish may have poor icing and cold storage character- istics, which make marketing difficult. They may be too bony, or their small size or odd shape may make processing technically or economically infeasible. The advent of deboned (minced) fish flesh and related fishery products on the North American scene in recent years has attracted a considerable amount of attention on the part of many major processors, within government circles and in research-oriented laboratories (Martin, 1972, 1974). 1 Nevertheless, in spite of the fact that deboned fish tech- nology appeared to offer great processing possibilities and to provide a variety of marketing options which were not readily available for unconventional fish species, many problems have hindered the realization of its full utili- zation (Nakayama and Yamamotq 1977). Before underutilized fish can be evaluated for potential use in processed food products, certain basic information on each species is needed to help determine whether the fishery resource is acceptable as human food and can justify the processing equipments and facilities (Miyauchi,1975). The suckers (Catostomidae family), from the Great Lakes area, have been selected as the target freshwater underutilized species for this study since although it probably has the greatest pro- duction potential it is one of the most difficult to market when processed by conventional methods. It was the purpose of this study to evaluate some of the variables affecting the functional properties of the sucker flesh obtained by mechanical deboning. An attempt was made to assess techniques or methods for measuring tex- ture and water-holding capacity as indicators of the func- tional behavior of the minced fish flesh. Finally, the influence of the handling of the fish and the freezing stor— age conditions on the functional parameters of the mechan- ically deboned sucker flesh was also studied. LITERATURE REVIEW Underutilized Freshwater Fish and Mechanical Deboners In the early 1970's only Whitefish and chubs remained as species worth fishing for the Michigan commercial fishery. Economic impact analysis revealed that the commercial fish- 4ery had a dollar multiplier of about four making the 1973 catch worth about 16 million dollars to the state of Michi- gan (Kevern,1975). For these reasons it was decided to eval- uate the potential of the fisheries of underutilized species in the Great Lakes area and an invited symposium on Under- utilized Freshwater Fish Species was held at Michigan State University on June 25-26, 1975. Underutilized species that may have potential for commercial development in the upper .Great Lakes are the suckers, burbot, rainbow smelt, and alewife (Magnuson,l975). The suckers are bottom.dwelling freshwater fish. The white sucker (Catostomus commersoni) is the most common sucker found in Michigan. However, the longnose sucker (C. catostomus) can also be found in the Grand Traverse Bay watershed. Suckers are bottom feeders, the characteristic sucking mouth being adapted for picking up and extracting food items from bottom sediments. Suckers provide a sport fishery in some areas during the spring spawning runs. 3 These fish are excellent eating after a visit to the smoke- house, but because of the many bones in the flesh they are not a popular food fish. Suckers have also been taken com- mercially in the past, referred to by commercial fishermen as mullet (Price and Kelly, 1976) . Mechanized methods of separating the flesh of fish from skin and bones have been in use for more than 20 years. The Japanese have made extensive use of this technology, especially in the field of fish utilization (Miyauchi g£'§1., 1973, and Okada g5 g1., 1973) but only in recent years has the potential of mechanical flesh separation been recognized by the North American food industry (Miyauchi and Steinberg. 1970). This technique permits greater recoveries of edible flesh than obtained by conventional filleting methods. At the present time, there are two basic types of ‘machines used to separate raw fish flesh from skin and bones. In one type of machine, the raw material is squeezed between a belt and.a perforated metal drum moving at the same or dif- ferent speeds, and this results in the softer parts such as the flesh being extruded through the holes of the drum, while the harder or tougher parts, like bones and skin, are passed on to discard. In the other type, called a strainer, the material is forced by a worm-feed into a perforated drum, so that the flesh is forced through the holes and unwanted .material is discharged through an opening at the end of the drum. The separation of flesh is not perfect in either machine, and some small pieces of bone and skin will be found in the product, the size and amount depending of the size of the holes in the drum, as well as other operating conditions and on the nature of the raw material. Large bone particles and scales left in the product constitute major quality defects influencing the consumer acceptability of cooked pro- ducts made from the minced flesh (Dingle g; gl., 1974). In species of low muscle lipids, the dark flesh located just under the skin is considerable higher in lipids than the white flesh that forms the bulk of the muscle. It is frequently desirable to collect white flesh separately. The location of the dark flesh permits some control by regulating the belt pressure exerted against the drum. A single pass at low pressure results in a product with virtually all light meat. A second pass of the resultant waste at maximum pres- sure recovers the balance of the light meat plus a high per- centage of the dark meat that was attached to the skin (Patashnik g2 g1., 1973). Potential source material for deboning may be obtained from.l) racks from a filleting operation, 2) fish which are considered too difficult to fillet efficiently and 3) fish which are diffuclut to market in conventional forms due to their low consumer appeal (Iredale g; al., 1974). There are two basic potential uses for deboned fish flesh. The first is the direct substitution of deboned fish blocks for frozen fish fillet blocks to be made into breaded or battered fish sticks or portions (King, 1973; Teeny and Miyauchi, 1972; ‘Miyauchi gt 31., 1975). This could be an outlet for the flesh recovered from the racks of well accepted species such as cod and haddock. The second use is in formulated products such as gifelte fish, fish sausages and frankfurters and various types of fish portions where the basic character- istics of the fish have been modified by the addition of other components. This second use is particularly applicable to those species which have low consumer appeal when proces- sed by conventional methods (Iredale 2; $1., 1974). The Importance of the Structural Proteins of Fish Muscle on the Acceptability of Minced Flesh Products Based on the differences in their physico-chemical properties fish proteins are broadly categorized as sarco- plasmic and myofibrillar proteins. The sarcoplasmic proteins, forming approximately 15-20% of the total proteins,depending on the fish species, are generally soluble in water or buf- fers of low ionic strength. The myofibrillar proteins con- sisting 60-80% of the total proteins are soluble in salt solutions of high ionic strength. These "structural" or "textural" proteins are particularly important as their denaturation, caused by a variety of factors, induces drastic changes in the quality attributes of the product (Warrier 35 31., 1975). The drip is defined as the exudate of tissue fluids that flow free from fish muscle during holding and storage or during thawing of frozen fish or muscle. Drip leaches along with it soluble proteins, vitamins, minerals, and confers an undesirable appearance on thawed fish. Drip has been regarded as occurring as a result of cell damage caused by freezing. The appearance of DNA in the expressible fluid is an indicator of rupture of cell membrane. However, it has been reported by Seagran, (1958) that cell damage alone cannot account for the release of drip, but it is also related to the capacity of muscle proteins to imbibe free liquid. It is well recongized that the textural qualities associated with muscle such as fibrousness, plasticity, and gel-forming ability are controlled by myofibrillar proteins. Fish muscle contains large proportions of actomyosin, a pro- tein composed of actin and myosin. Investigations carried out by Ueda 3; 31., (1968) on the behavior of the purified actomyosin revealed that thereimne no species differences in intrinsic viscosity value, electrophoretic mobility, salting- in and salting-out range, etc., while there existed a dif- ference in the temperature of denaturation from species to species. Myosin, actin, and tropomyosin are the other com- ponents of the myofibrillar proteins present in fish muscle. Water molecules combine with polar groups of the con- tractile proteins. Fish muscle has a minimum ability to combine water molecules around pH 5. On either side of the isoelectric point, the muscle exhibits higher ability to hold water. Post-mortem changes such as Ca++ liberation, influence the amount of bound water. The presence of sodium chloride increases the water holding capacity principally by the effect of Cl' ions which bring about greater repulsion between the peptide chains (Warrier et 31., 1975). Several salts of weak acids such as sodium pyrophos- phate and sodium tripolyphosphate have been recognized by many workers in increasing the water holding capacity of flesh foods. The influence of these phosphates has been related to their effect in promoting the extraction of pro- tein from the fibrils, principally in the presence of high salt concentration and divalent cations. The concept of muscle hydration has now been extensively applied to improve the quality of frozen fish in terms of reduced drip and cooking loss and textural improvement. The most important biochemical change associated with proteins is denaturation. This structural modification may bring about definite changes in chemical, physical or biolog- ical properties of the protein (Warrier gt al., 1975). Denaturation mechanism is considered to be a two-step process involving the opening of the peptide followed by the split- ting or combination of the molecules. The amount of extract— able protein is often taken as a criterion of denaturation. During frozen storage the extractability of the myofibrillar group of proteins is unaffected (Childs,l973). It is pre- sumed that the proteins of fish can also be damaged as a consequence of continued exposure to concentrated solutes in the frozen-stored muscle of fish with added salt. The inter- action of protein with fatty acids, as well as the tendency ’of myosin to aggregate have also been reported as situations leading to protein insolubilization (Warrier 33 31., 1975). The Effects of Freezing Storage on the Keeping Quality of Fish Flesh A particular problem associated with unconventional fish species is the relatively poor keeping quality during frozen storage of deboned fish flesh obtained from certain species. This is largely due to the characteristic physical disintegration of the flesh resulting from the deboning pro- cess (Miyauchi 33 33., 1975; Patashnik 33 33., 1976; and defitt 33‘3l., 1976). The more drastic changes observed are in the flavor and color of the minced fish flesh as a conse- quence of lipid oxidation during frozen storage. Washing has been advocated as a means for removing soluble constituents believed to be responsible for such deterioration. The theory proposed to justify this practice is that the hemo- proteins, myoglobin and hemoglobin, liberated from either muscle or bone marrow upon deboning catalyze oxidative rancidity, as reported by Lee and Toledo (1977). The same authors demonstrated that iron coming from nonstainless steel parts of the deboning machine accelerated lipid oxi- dation of the minced fish flesh. 10 A test based in the reaction of 2-thiobarbituric acid (TBA) with the oxidation products of unsaturated fatty acids to give a red pigment, has been used to determine the oxida- tive rancidity in minced fish flesh (Sinnhuber and Yu, 1958). The deterioration of mechanically deboned fish flesh during frozen storage is also manifested by protein denatur- ation. In sausage formulations such stale raw material induces an undesirable grainy texture of the final product (Grabowska and Sikorski, 1973). This deterioration has been associated with the production of formaldehyde and dimethyl- amine from.the trimethylamine oxide present, mainly in the muscle of marine fishes (Dingle 33 33., 1977; Hiltz 33 31., 1976). under proper conditions of frozen storage bacterial action is prevented so there is no production of trimethyl- amine (Mills, 1975). The presence of slime is detrimental to the quality of dead fish because of its ability to support vigorous growth of bacteria. Therefore removal of slime by washing has been strongly suggested to increase storage life of fish (Gillespie and Ostovar, 1971). The technique of washing minced fish muscle followed by incorporation of salt, sodium polyphosphates and sugar has been successfully used in the preparation of "surimi", a japanese fish paste product, to save fish muscle protein from denaturation during cold storage (Iwata and Okada, 1971). 11 Water Holding Capacity and Texture as Indicators of the Gel-Forming Ability of the Minced Fish Flesh During the mechanical separation of fish flesh the disruption of tissue integrity allows access of oxygen, the spreading of bacteria and contacts of intracellular and extracellular components. The overall effect is the promo- tion of a variety of biochemical, physical and chemical reactions at rates greater than occur in fillets or in whole fish (Bremner,l977). As a result the minced muscle tends to release water and form drip when heated or when frozen and thawed. The degree of drip formation varies with processing procedures and fish species. The water imbibing power of fish flesh is called either water holding capacity or water binding capacity. These terms are typically used to mean the strength by which water is held in food systems but really are the water content under some given condition. Methods used to measure water- holding capacity (Wierbicki gt 31., 1957; Wierbicki and Deatherage, 1958; Dagbjartsson and Solberg, 1972; Karmas and Turk, 1975; and Bremner, 1977), are quite arbitrary, and they do not give any physical chemical property to express the immobilized portion of the water. Consequently, water holding capacity should be defined in terms of a particular method of measurement, so that the results obtained in dif- ferent laboratories may be comparable (Labuza and Lewicki, 1978). 12 A reason why the leftover parts of fish carcasses, -after the fillets have been removed, are underutilized as a source of fish flesh is the poor water binding quality of such minced flesh. Consumer qualities of minced fish pro- ducts, such as appearance, flavor, as well as drip and skrinkage on cooking depend greatly on the degree of water binding. Therefore, when the water binding of these pro- ducts is improved, its consumer acceptance is generally increased (Karmas and Turk,l976). The moisture content along with the particle size of the mince have a marked influence on the texture of the final cooked fish product. Other important factors deter- mining the texture of cooked minced fish flesh products are the presence of salt and phosphates, the chopping time in -emulsion type fish products, and the cooking temperature and rate (Lee and Toledo,l976). The use of vegetable and milk proteins such as soy protein isolate and sodium.caseinate respectively has been suggested to increase the texture characteristics of fish products beyond the intrinsical ability of the natural proteins present in fish muscle (Karmas and Turk,l976). Textured soy flour has also been successfully used in fish preparations (Daley and Deng,l978; Daley gt 31., 1978). It is obvious that the perception of texture is not solely dependent on the properties of the foodstuff but is strongly influenced by the characteristics of the person 13 examining or consuming the food. In this respect texture is similar to other physical characteristics such as color and flavor. This dual character of food texture has given rise to attempts to construct a texture classification system based on the impression created on the human sensory sys- tem, the texture profile (DeMan,l975). Several types of instruments have been used to evaluate the texture of minced fish flesh products in terms of one or more components of the texture profile system. The Instron Universal Testing Machine has been suggested to evaluate texture profile parameters like shear force, hardness, fracturability, springiness. cohesiviness, gumminess and chewiness from the force-deformation curve obtained by the instrument (Peleg,l976; 800 and Sander,l977a; Lee and Toledo, 1976; Webb gt 31., 1976; and Cross gt 31., 1978). The Kramer Shear Press has been a widely used instrument for food texture measurements. The major use of the shear press is with the universal test cell. It consists of a slotted box through which a set of 10 metal blades is forced down by a hydraulic ram. The action in the cell involves compression, shear and extrusion. With the recording device two types of results can be obtained; maximum force which is indicated by the height of the peak and work which is indicated by the area under the curve (DeMan,l975). The Universal Penetro- meter has been used to measure the consistency (softness) of comminuted fish-matrix agent mixtures 14 ($00 and Sander, 1977b). The Shear Jaw Press designed by Dassow gt 31. (1962), has been successfully used for the determination of shear values on fresh and frozen seafood. There seems to exist a general consensus, among people working in the area of fish processing, that texture charac- teristics of comminuted fish flesh are important factors affecting acceptability by consumers. EXPERIMENTAL Classification of the Experimental Work Into Two Study Sections In the first section, 3 study of the variables modi- fying texture and water-holding capacity of the mechanically deboned sucker flesh, along with the methods used to measure these functional properties, was undertaken. An examination of the conditions of handling and freez- er storage, and their influence on the functional prop- erties of the fish flesh, constituted the work of the second section. Section One: The Study of the Mechanical Deboning Operations, the Binders, and the Cooking Systems as Related to Texture and Water-holding Capacitijeasurements About 600 lb. of fresh sucker from Lake Huron were purchased at AuGres, Michigan and transported on ice to the Meat Laboratory at Michigan State University. The load was composed of some fish caught 24 hours and some caught 72 hours prior to the time of processing. Species composition consisted of about 507; each: Silver Redhorse sucker, beostoma anisurum (Rafinesque), and White sucker, Catostomus commersoni (Lacépede), (Eddy,l974). 15 16 Fish were dressed and passed through a mechanical deboner. The collected minced sucker flesh was then passed one more time through the meat separator to compare the efficiency of the process of separation as well as the chem- ical, physical and microbiological characteristics of the flesh. Samples of minced flesh were collected after the first and second passes and analyzed for moisture, fat, pro- tein, bone residue, TBA number and microbiological load. The sucker flesh was then packaged into Cryovac (poly vinylidene chloride) bags and stored in a -29° C blast fre- ezer. The frozen flesh was thawed as needed, with thawing drip determined at this stage. All further operations up until the cooking step were carried out in a cooler room at 2° C. At this point the fish flesh was blended with differ- ent binders and cooked by two different ways: water bath cooking and smokehouse cooking, the latter after stuffing and linking the fish paste into frankfurter type casings. The water-holding capacity was estimated on the smoke- house-cooked fish paste by either the smokehouse shrinkage, filter paper press method or centrifuge technique, and on the water bath-cooked fish paste by either the blotting paper technique, the filter paper press method or the cen- trifuge technique (see pgs. 28&29). Texture of the smoke- house cooked product was estimated by the shear force obtained from the Kramer Shear Press and from the Instron universal testing instrument, as well as by the softness 17 measured by using the Universal Penetrometer. Section Two: The Effect of the Handling and Freezer Storage Conditions on the Functional Properties of the Sucker Flesh Approximately 350 1b. of fresh sucker from Lake Huron were purchased at Bay Port, Michigan and transported on ice to the Meat Laboratory at Michigan State University. Fish ~had been caught about 24 hours prior to laboratory processing. Species composition was about the same as that of fish used in the first section. Fish was handled and frozen according to five treat- ment groups as follows: A. - Whole fish, fast freezing; B. - Whole fish, slow freezing; C. - Dressed fish, no washing, fast freezing; D. — Dressed fish, washed, fast freezing; E. - Dressed and washed fish, mechanically deboned, fast freezing. A year earlier a similar study had been conducted wherein one treatment involved exposure for 3 minutes of dressed fish to a severe washing and scrubbing action by -placing them in a commercial sand paper-scrubbing type .potato peeler (RE-NU open tank, model SAA, Vacuum Filter Mfg. Co., Chicago, 111.) through which cold water was con- tinuously spraying. Fish treated by this severe washing- scrubbing action were exposed to both fast and slow fre- ezing systems as described in the following. 18 Fish material was vacuum-packed and freezer stored for 13 months. Fast freezing conditions were accomplished in a -29°C blast freezer. Slow freezing was carried out in a -l8° C dead air freezer with an approximate freezing time of 24 hours, at least 16 hours of which occurred in the zone of maximum ice crystal formation. After this period of storage all fish material was thawed at 2°C. Treatment groups A, B, C and D were either dressed and mechanically deboned or mechanically deboned only. The minced fish flesh from these four treatment groups, along with that of treatment group E, was evaluated for func- tional behavior according to the following sequence: fish flesh from each treatment group was blended with regular bin- der into an emulsion type fish paste by chopping and also with regular binder plus 2% salt into a fish paste blended by revolved mixing. The modified flesh was then cooked by water bath and, after being stuffed and linked into frank- furter type casings, in a smokehouse. Water-holding capa— city was determined by the centrifuge method on both the smokehouse cooked and the water bath cooked fish flesh. Tex- ture was determined by measuring the shear force on the smokehouse cooked product using the Instron universal testing instrument. 19 Handling of Fish at Laboratory and Processing Operations Dressing and Mechanical Deboning Fish was dressed by separating the head and removing all viscera (along with the kidney tissue) manually. Fish were then washed in cold water and split lengthwise. The mechanical deboning operation was carried out using a Bibun meat separator (Type SDX 13, Bibun Co. Fukuyama Hiroshima, Japan) equipped with a 5mm hole size drum. The machine was fed the split fish with the muscle side facing the drum to facilitate the process of flesh separation. Yields of both the dressing and mechanical deboning operations were calcu- lated from the weight of the fish material before and after each operation. Packaging of Fish Material for Freezing Storage The minced fish flesh obtained for the study of section one was packaged into polyvinylidene chloride bags (Cryovac, Dewey and Almy Chemical Co., Cambridge, Mass.). Each bag contained about 25 lbs. of flesh. Whole fish, dressed fish and mechanically deboned fish intended for the study in sec- tion two were vacuum packaged into Cryovac bags using a vacuum sealer (Cryovac model FVC-E, Dewey and Almy Chemical Co., Cambridge, Mass.). Each bag contained about 15 lbs. of fish material. 20 Binders and Blending Operations The minced fish flesh was modified by using binder mix- tures and mechanical agitation as described in Table 3. Reg- ular binder (RB) was established, for the purposes of this study, as a mixture of ingredients used along with the minced fish flesh in the following concentrations: Salt ---------------- 1.0 3 Sugar ---------------- 1.0 % Corn 011 -------------- 1.0 % Fish Muscle ------------- 2.5 % Ice ----------------- 5.0 1 Monosodium Glutamate -------- 0.3 % Sodium Tripolyphosphate ------- 0.15 % Sodium Ascorbate ---------- 0.04 % Preparation of the regular binder mix was carried out in a 2000 ml bowl capacity Mixmaster (Mbdel 12C, Sunbeam Appliances Mfg., Chicago, 111.). All ingredients except the corn oil were homogenized for 15 seconds. After the addition of oil the slurry was stirred at maximum speed for 3 minutes. The proteinaceous substances used as binders in this study were soy protein isolate (Cenpro-P, Central Soya, Chicago, Ill.) and sodium caseinate (Milk Proteins Inc., Detroit, Mich.). The blending of fish flesh with binders was achieved in either a silent cutter (Model 84181D, Hobart Mfg. Co., Troy, Ohio) for l, 2 or 4 minutes or in a Kitchen Aid Food Preparer (Model N-50, Hobart Mfg. Co., Troy, Ohio) with a paddle attachment for 15 minutes. 21 Cooking of the Fish Paste Fish flesh was cooked in two different ways: i. - Water bath cooking. Twentyfive grams of fish paste were weighed into open 50 ml plastic centrifuge tubes and cooked in a 70°C water bath for 30 minutes. ii. - Smokehouse cooking. Fish paste was stuffed into size 22 cellulose casings NoJax (Union Carbide, Films and Packaging Div., Chicago, Ill.) and linked as frankfurter type sausages. The product was then cooked in an Elek-Trol laboratory smokehouse (Drying Systems Inc., Chicago, 111.) according to the following schedule: Time (min.) Temperature (°C) Relative Humidity (Z) 10 54.5 25 20 60.0 35 20 65.5 40 20 71.1 65 15 79.4 65 The cocking process was concluded with a 6 minute cold water shower. Techniques and Methods of Analysis Chemical Methods Moisture Content The A.O.A.C. (1965, 23.003) procedure for determining moisture was used. Five grams of fish flesh were accurately weighed into a previously dried and tared aluminum dish (100°C for at 22 least 1 hour). Sample plus dish were then dried overnight for 18-24 hours in an air convection oven at 100°C. Dry sample was cooled in a dessicator and weighed to four decimal places. Loss in weight was reported as moisture for each hundred grams of meat. Three replicates were run for each sample. Fat Content The fat content was determined using the Goldfisch extraction method of the A.O.A.C. (1965, 23.005). Sample used was the same as for moisture analysis, taken directly from the dessicator. The aluminum dish con- taining the dried 5 grams sample was carefully folded into a porous thimble and clipped into the Goldfisch apparatus. Fat was extracted with an anhydrous ether for 4-5 hours into a previously dried and tared beaker. The extract was then dried for 30 minutes at 100°C in an air convection oven, cooled in a dessicator and weighed. The percent fat was cal- culated as grams of fat extracted from each one hundred grams of fish flesh. Three replicates per sample were run. Protein Content Protein was determined following the micro Kjeldahl nitrogen determination method (A.O.A.C. 1965, 23.009). Approximately 0.5g of fish flesh were accurately weighed into a micro Kjeldahl digestion flask followed by 23 addition of 1 ml Cu804 10%, 1 g anhydrous NazSOA, 7 ml H2804 and several glass beads. This mixture was heated on a rotary Kjeldahl digestion unit under a hood until the boiling mix- ture turned clear green. The digestion was allowed to con- tinue for 30 minutes. The flask was then cooled, 15 m1 deionized water were added and the flask cooled again. Ten m1 2% boric acid and 3 drops Bromcresol green indicator solu- tion were placed into a 125 ml erlenmeyer flask. This was secured under the distillation outlet and raised enough to permit the tip to lie below the surface in the flask. Then enough NaOh 44% solution was added to make the solution strongly alkaline. Steam was allowed to enter the system and the sample was distilled for 7 minutes. After this time the erlenmeyer flask was lowered so the tip was out of the liquid and distillation was run for 3 minutes more. The amonnia boric acid solution was titrated to the Bromcresol green end point with standard H2804. The percent protein was calcu- lated using the following formula: 7 Protein = (ml of H2804)x.(Normality of H2804) x (l4)(6.25)(100) Weight of sample in mg. Three replicates were run of each sample. 2 - Thiobarbituric Acid (TBA) Test The TBA values were determined using the method of Tarladgis gt 31., (1960). Four 10 g portions of the minced meat or fish muscle 24 were homogenized with 50 ml of distilled water at medium speed for one minute in a Virtis homogenizer (Model 6-105-AF, Virtis Co., Gardiner, New York). The homogenized mixture was transferred into a 500 ml distilling flask with the aid of 47.5 ml of distilled water. The pH was lowered to 1.5 using 2.5 m1 of 4NHCL. Several drOps of Dow Corning Anti- foam A (Dow Corning Corp., Midland, Michigan) and a few glass beads were added to the mixture. The flask was con- nected to a distilling unit consisting of a 30.5 cm long distilling column connected to the condensor with an elbow joint; a 50 ml graduate cylinder acted as a receiver. After boiling began, the first 50 m1 of distillate were collected. Two 5 ml portions of the distillate were pipetted and transferred to screw top test tubes. Then 5 ml of 0.02 M thiobarbituric acid (Eastman Organic Chemical, Rochester, New York) in 90% redistilled glacial acetic acid were added and the tubes were capped, mixed, and heated in a boiling water bath for 35 minutes. After cooling in cold water for ten minutes the absorbance was determined at 538 nm against a reagent blank in which 5 ml of distilled water were used in place of the distillate. The TBA number was calculated by multiplying the mean absorbance by 7.8, a distillation constant (Tarladgis gt 31., 1960). The TBA value was reported as mg TBA reactive substance per 1000 grams of fish flesh. 25 Bone Fragments Content Bone residue was determined following the method described by Wong and Yamamoto, (1974). A 100 mg sample of minced fish flesh was added to a 4000 ml erlenmeyer flask containing 2000 ml of 3 M urea and 0.02 M NaOH solution. The suspension was stirred gently overnight at room temperature (28° C) on a Multi-Magnestir (Lab-Line Instruments Inc. Melrose Park, 11.). The suspen- sion was then filtered through a 425 microns sieve (Sargent Welch, No. 40). The retained residue of bone fragments, cartilage and scales was transferred to a dried and tared 100 ml beaker and restirred for 4 hours in 50 m1 of 3 M urea and 0.02 M NaOH solution in order to further solubilize small pieces of flesh that still remained with the bone and scale fragments. At the end of this time, fragments were allowed to settle for about half an hour and the urea-NaOH solution again poured off through the sieve. The nearly clean, pro- tein-free residue left in the bottom of the beaker was then washed three or four times with distilled water. The beaker containing the residue was then dried to constant weight in a 105° C oven for 7 hours, cooled in dessicator and reweighed. A separate weight of the scale fraction alone was also recorded. The percentage of bone fragments was calculated using the following formula: 26 (Dryweight of bones) x (100) % Bones = Sample weight Six replicates were run on each sample. Microbiological Method The bacterial load of the fish material was measured on the surface of the fish and on the minced fish flesh according to methods described by Frazier gt 31., (1968). Surface bacterial load was determined by swabbing two fish in each treatment group. Sterile calcium alginate swabs in tubes (CalgitubeC) were run over a small area (approximately 4 square cm) midway down on lateral side of each fish. These were held overnight at 34° F in capped tubes containing 10 ml of a 1% W/v sodium citrate solution. A dilution range of 101 to 105 was prepared and then sam- ples were plated out on Plate Count Agar. Four replicate plates per dilution and per treatment group were incubated at room temperature (29° C) and counted after 72 hours. Results were expressed as number of bacteria per cm2 of fish surface. The bacterial load of the mechanically deboned fish flesh was assessed as follows: Duplicate 11 g samples of the minced fish flesh were weighed into bottles containing 99 m1 of sterile water and several glass beads. A dilution range of 101 to 105 was prepared and the samples were plated, incubated and counted as described for the surface bacterial 27 load determination, above. Results were expressed as number of bacteria per gram of minced fish flesh. Physical Methods Measurements Related to Water-holding Capacity Thawinngrip The thawing drip was determined at room temperature by weighing a 50 g sample of partially frozen fish flesh into a 8.5 cm diameter glass funnel containing a glass wool bed. The liquid was collected into a 10 ml graduated cylinder, and volume measurements were made hourly over a period of 3 - 4 hours. Results were expressed as water losses %. Smokehouse Shrinkage The smokehouse shrinkage was calculated by the difference in weight of the fish paste product before and after the smoke- house cooking operation. The results were expressed as water losses %. Water Losses During Water Bath Cookigg Water losses during water bath cooking of the fish paste were calculated as follows: A 25 g sample of fish paste was cooked in a waterbath as described in Cooking of the Fish Paste, i, above. Following this the centrifuge tubes containing the cooked fish flesh were reversed until 28 all fluids were completely drained from them (about 15 minutes). Next, the meat pellets were removed from the tubes, blotted on paper toweling and weighed. The weight differ- ence with the original sample of raw fish paste was reported as water losses %. Expressible Fluids by the Filter Paper Press Method The water—holding capacity of the cooked fish flesh was determined according to the gravimetric adaptation intro- duced by Karmas and Turk, (1975) of the original Filter Paper Press method described by Wierbicki and Deatherage, (1958). A 300 mg sample of cooked fish flesh was weighed onto a 3/4 inch diameter aluminum foil disc. This was placed on the center of a tared piece of filter paper, 7 cm in diameter (Whatman No. 54), stabilized at the laboratory room temper- ature (29° C) and relative humidity (about 60%) conditions. The sample-liner-filter paper system was then pressed between two plexiglass plates for 1 minute at 500 psi, using a hand operated Carver laboratory press (Fred S. Carver Inc., Hydraulic Equipment, Summit, New Jersey). Then the liner and the meat residue were removed from the filter paper con- taining the expressible fluids and this was immediately weighed. The weight difference with the original filter paper was reported as water losses %. This determination was run on four repetitions per sample of both the 29 smokehouse-cooked and the water-bath-cooked fish paste. Expressible Fluids by the Centrifuge Technique The water-holding capacity of the cooked fish flesh was also determined using the centrifuge technique reported by Bremner, (1977). Ten g.of cooked fish flesh were weighed into 15 ml CorexC>No. 8441 glass centrifuge tubes and centrifugated at 18,000 rpm (39.100 xg) for 1 hour in an automatic refrigerated centrifuge (Sorvall Type RC2B, Rotor S-34, Ivan Sorvall Inc , Norwalk, Conn.). Then the supernatant was drained and tubes were reversed for about 15 minutes on toweling paper to com- pletely drain the remaining fluid. The tubes plus the cooked fish flesh pellets were then weighed. The weight dif- ference of the meat pellet and the original raw paste was expressed as water losses % due to cooking. This technique was applied to both the smokehouse-cooked and the water- bath-cooked products. Three replicates per treatment were run . Measurements Related to Texture Shear Force by the Kramer Shear Press This determination was performed on the smokehouse- cooked fish paste by using a texture test system (TP-2 Texturepress associated with a TR-l Texturecorder, Food Technology Corp., Reston, VA) equipped with a shear 30 compression cell. The fish samples were cut into 8 cm long cylinders, weighed and placed within the test cell. Then they were subjected to shear, compression and extrusion. The resistance force was measured by the deformation of a high strength alloy ring and recorded by the texture recorder. The results were expressed as lb-f per gram of cooked fish paste. Five repetitions per sample were recorded. Shear Force by the Instron, Universal Testing Instrument This determination was performed on the smokehouse cooked fish paste by using an Instron universal testing machine (Model TTBB, Instron Corp., Canton, Mass.) equipped with a Compression Load Cell, range 1 to 50 kg. A meat shear cell was adapted to the upper moving fixture. The instrument was calibrated with a 1 kg weight for full scale displacement of the recorder pointer. The speed of the drive and the speed of the chart were adjusted both at 20 cm/min. Fish flesh frankfurters were cut into about 10 cm long cyl- inders and sheared across the section. The resistance force was expressed in kg-f per cm sectional diameter of the fish cores. At least 10 replicate measurements were recorded for each treatment group. Softness by the Penetrometer This determination was also performed on the smokehouse- cooked fish flesh by using the Universal Penetrometer 31 (Micrometer Adjustment Penetrometer, Arthur H. Thomas Co., Philadelphia), equipped with a 35 g penetration cone. Samples of fish frankfurters (1.8 cm average sectional diam- eter) were cut into 2.6 cm long cylinders and placed verti- cally (section of the meat cylinder facing the penetration cone) just under the penetration device. This was then released for 10 seconds over the fish cores. The penetra- tion, expressed in mm, was read on the instrument scale. The results were expressed in terms of mm penetration per cm sectional diameter of the fish core. At least 10 repe- tition measurements per treatment group were recorded. Statistical Analysis Statistical analysis were achieved in a TI-59 pro- grammable electronic calculator (Texas Instruments Inc., Lubbock, Texas) equipped with a Solid State software statis- tics module (STAT-5859). Correlation Coefficient Correlation coefficients (r) were obtained by entering row data (average of the repetitions) through the Bivariate Data ST-04 program and then applying the special control operation for correlation coefficient. 32 Analysis of Variance (AOV) The one-way AOV was carried out by entering raw data (individual observations) through the one-way AOV ST-06 pro- gram and then analyzing the data through the one-way AOV ST- 15 program. The two-way AOV was carried out by entering raw data (single individual observations) through the two—way AOV ST- 06 program and then analyzing the data through the two-way ST-l6 program. The Tukey's procedure (Steel and Torrie, 1960) was used to compare the mean values of the treatments each time a significant F value was obtained through the AOV. RESULTS AND DISCUSSION Section One Table 1 shows some of the physical, chemical and micro- biological characteristics of the sucker flesh obtained from the mechanical deboner (first-pass flesh) and after passing this flesh through the machine another time (second-pass flesh). The yield of minced sucker flesh from the mechanical :meat-bone Separator was about 50%. Losses of meat and/or juices by a second pass operation seemed not to affect this yield in a drastic way. This yield figure is expressed as a.percent of the initial round weight of fish, however mechanical deboning of dressed suckers yielded about 75% of their dressed weight as minced flesh. The 50% yield seems to be similar to that reported by Morris, (1977) for the same fish species caught through the summer season. Bone and scale residue showed a slight but significant decrease in the second pass flesh. This was probably due to a considerable dimunution in the scale content rather than a lowering of bone content as shown in Table l. The chemical technique for isolating bone fragments by Yamamoto and Wong, (1974) proved to be appropriate not only for the separation and determination of the residue (bone, scale, cartilage 33 34 TABLE No.1 - Some of the physical, chemical and microbiological charac- teristics of the sucker flesh obtained by mechanical deboning (mean values*). Characteristics First pass flesh Second pass flesh Minced flesh yield after mechanical deboning(%) 50.8 48.8 Bone and scale residue(%), N=6 0.19axo.02 0.15bi0.03 ‘Bone residue(%), N=6 0.133:0.01 0.12310.o2 Scale residue(%), N=6 o.ooa:o.01 0.03b10.01 Moisture(%), N=6 81.23aiO.15 81.15310.43 Fat(%), N=6 1.4Zai0.12 1.59a:o.42 Protein(%), N=6 15.983:0.44 15.803i0.85 TBA number as mg TBA/Kg flesh, N=7 0.3018ar0.07 0.2709ai0.03 Bacterial load as log No. bacteria/g flesh, b N=8 4.7laiO.13 4.57 10.09 * Means in same row not followed by the same superscript are significantly different (p<0.05). 35 and skin fragments) but also for the separation of scales from bone fragments as it was performed in this study. The calcium.content of the mechanically deboned sucker flesh has been found to be 0.15 and 0.08% for the first-pass flesh and second-pass flesh respectively (LeBlanc, 1978). These values are well below the proposed 0.75% maximum percent calcium content for mechanically deboned meats (Anonymous, 1976). On the other hand the small size of the bone fragments obtained in this study suggests that the drum hole size and the belt tension used on the mechanical separator permitted good separation of flesh and residue. Moisture, fat, protein and TBA number (an indicator of the rancidity of the fish lipids) were not significantly different when comparing the first-pass flesh to the second- pass flesh. The bacterial load on the minced sucker flesh was slightly lower in the second-pass flesh than in the first- pass flesh. However this difference seems not to be of industrial microbiological importance since the change in the number of bacteria was lower than one log unit. The process of washing the fish after dressing did not affect the bacterial count on the surface of the fish, as can be observed in Table 2. This might be due in part to the poor method of washing used in this study which removed only loose slime from the surface of the fish and in part to the probable diffusion on the skin surface of microorganisms ncéwmaé :.oumw.q ~N.ouwm.q Wm Ahms guano—£5 233 K33 swam ommmmuo swam ommmmun swam oaaom .an .mommusw swam ~80 \ mwumuomn mo panes: may no on ecu mm oommouoxm .muwxosm onu mo oommusm on» no omoH Hmfiuouomn may mo soHumw>mo oumocmum man are: I ~.oz mqm<fi 37 coming from the gut cavity. The ability of slime to support bacterial growth has been shown by Gillespie and Ostovar, (1971). These authors also reported a 20-fold lower count on the slime from commercial Whitefish washed for 15 minutes under cold water jets compared to that on the unwashed fish. Similar reductions have been obtained in this laboratory by using a potato peeler device to scrub and wash dressed suckers (Price,1978). Table 4 shows the results of the water-holding capacity (WHC) studies of the cooked sucker flesh expressed as water losses percent, and measured by four different techniques. The smokehouse shrinkage, the draining and blotting on absorb- ent paper, the filter paper press method, and the centrifuge technique. Fourteen different modifications of the minced sucker flesh were carried out in this study as described in Table 3. The sucker flesh with regular binder (RB), (treatment B-5) was considered as the control binder treatment since the washed sucker flesh with binder or without binder (treat- ments B-3 and B-2 respectively), and the sucker flesh without binder (treatment B-l) showed an extremely soft consistency with very poor textural characteristics. On the other hand 1 and 4 minutes of cutting-mixing (treatments B-4 and B-6 respectively) showed no improvement in the WHC of the sucker flesh when compared to the 2 minute cutting-mixing, the 38 Acowmasao umoev wauuuao mouaaaa N Acowmasao uooEv mcuuuso mouscwa N “page mauvwa cu wcw>ao>mu mouscwa mg AconH=Eo uooav wcfiuuso mmuacfia N Acome=Eo amoev mcfiuuao mouncae N Acowmasao umoEv wcfiuuso mou::NE N Acoamaaao umoav wcfiuuno mwuscwa N Acoamflsao ummev wcfiuuao mouacaa N AconH35o umoev wcauuso mmuacas c AconHSEo umoav wcwuuau mmuacfia N Acowmaaao gooey waauuso sundae N AcONmasao umwsv wcfiuuao mouacfia N Acowmaaam uooav mcwuuso unusuaa N AcoamHDEo ummev wcfiuuao mmuacwa N ooaaqmm mmouum Hmoacwnomz um Nu msfla uHam Ham Nu mafia “Ham uamm uHmm om Hmm NN msHa mafia mafia mafia mafia mafia Aumv oumcaommo Esquom AHmmv condom“ :Nououo mom mm mafia mafia mafia mafia mafia mafia mafia moan mafia mafia mafia mm: hm: mm: mm: hm: hm: mm: mm! hm: hm: Ammv homage umaswou mafia .wcwsmma .mhz wcwnmoa can mm: Amaze gnome gnaw caucus unoEumouu may no coaumfiuumon nonsmoaufiuameH meoo .wumocwp may saga wnfiocwan mo mmooouo mnu wcfiuso amoau onu ou vowaoom mmouum Hmowcmsoma onu he no mucocfin No on: .wcficmma he smoau poxosm coconoo Hmoficonooe mzu No ocowumowwwvoz i m.oz mamuompo cannumm .N macaw :a confluence mm aucoauwouu uovcamN .mocmofluuswfim mo Ho>oa Mm ozu um acououufio mum Loan: smoau ooxooo mono: oxosm way so mmouo momma nouaww onu haw uaooxo .Ago.ovov acououuwo zuuchNuwchm mum unwuuwuooam osmm msu an voonHON no: sch mama ca name:# o.~nm.mN o.~wm.NN N.~ww.aN ~.Nwd.mm I I I I m.~wm.qm muz .cmmfim pm a no on on omxoou ooze; onoam no movie Isoou mwzuwuucou . I . . I . . I . . I . I I I I . I . u a mo 0 5mm mm N Nmm 0m Newmo on m wwm Nm m owe mm owxmoossuww noum3 :0 wood: Inoou wwauwuucwu m.~Hm.Nq m.me.~q m.NHw.mq <.muh.wm I I I I m.~w~.mc vuz no no A m an .ammau ooxooo masocoxOEm osu to among uoomo Houafim o I u a II o o I o o I o o I o o I o o I o o I o o I o a N mm~ me wwomww we owommm we mommu 0m wwommm cc owommm Na 0 ww— Nm 0 ng mm Nomww fin coxooocummnmwwww osu co mmouo uoama nouawm I I I I w. u .MN o.~Ho. H o.quo.NN o. “N.mN N. I .m «z mmq a w n mm mmq N .uoamo ucoauownm so mewuuoHn can wcwcumuc mm.w~ mm.o~ qo.~H~.oN no.0N ~.mwm.o~ ~.Nwo.oN m.cuN.oN w.mue.—N m.mwo.o~ mnz .mwmxcuusm omsozoxoem 0:3 wnu No :oHumE qum mHIm N~Im -Im o~Im oIm wIm mIm mIm Iwumo osu new menu IQUOpa no cocoa: Nmucoeumoub panama .ucouuoa wommoH Coos: ma oomwouaxo ocw mosvficsuou acouwuwuo use“ an cmoau umxusm ooxooo ecu No zuwomomo acmvac; ocuc3 osu mo mcoNumH>mo oumocmum was com: I «.02 m4m<9 40 standard procedure used in this study. Thus, nine differ- ent binder treatments are compared for WHC and texture char- acteristics in Tables 4 and 5 respectively. Void spaces in Table 4 are due to the fact that some binders and WHC tech- niques were tested at the beginning of this experiment and then discarded or replaced by some more advantageous ones. By using any of the reported four techniques the higher the value of water losses percent the poorer was the WHC. The measurement of the smokehouse shrinkage seemed not to be a good indicator of the WHC of the sucker flesh since no statistical differences were found in this study due to the presence of the various binders used (Table 4). The WHC determined by the techniques of draining and blotting the fluids from the water-bath-cooked flesh on absorbent paper was found to be subjective and tedious. However, from the results in Table 4 it can be noted that RB imparted to the flesh better WHC than either of the pro- teinaceous binders, soy protein isolate (SP1) and sodium caseinate (SC) used alone. The 2% SP1 produced a greater increase in the WHC of the fish flesh than 2% SC. This effect could be noted in both situations, the proteinaceous binders used alone and in combination with RB. RB plus 2% SP1 sugfifi- cantly improved the WHC of the sucker flesh when compared to 2% SC as the sole binder used with the fish flesh. The determination of the WHC by the filter paper press technique was carried out on the water-bath-cooked flesh and 41 on the smokehouse-cooked fish flesh, Table 4. The water- bath-cooked flesh had good WHC when 2% SC or 2% SP1 was used in combination with RB. No improvement in WHC was noted when these proteinaceous binders were used alone. In this study the results of WHC by the filter paper press technique cor- related well with those obtained by using the draining and blotting on absorbent paper technique. They also tend to agree with those reported by Karmas and Turk,(1976) as no significant difference between SP1 and SC were detected when used as fish binders alone or in combination with RB. Some improvement in WHC was also detected when RB was used in combination with an additional 2% salt and with 2% added salt plus-either of the two proteinaceous binders used in this study. A significant improvement of the WHC was observed in the flesh which had RB plus 2% salt and was subjected to pad- dle type (Hobart model N-50) mixing for 15 minutes. The incorporation of salt by the paddle-mixing procedure did not reduce the meat particle size and apparently permitted greater retention of water. The gravimetric adaptation of the fil- ter paper press method suggested by Karmas and Turk,(1975) proved to be a good technique for the determination of the WHC of the cooked sucker flesh. In this study, however, it was not possible to obtain good replicates values for each sample due in part to the fact that fish samples were pres- sed in quadruplicate at the same time between three plates 42 of plexiglass and the water from the meat continued to dif- fuse through the filter paper as the liner and meat residue were removed one by one to weigh the filter paper containing the expressed fluids. The centrifuge technique for the determination of the WHC of the fish flesh was based upon the ability of the meat to retain water as a constant centrifugal force (about 40,000 X g) was applied for one hour to the system. This technique appeared to be more reliable than any other used in this study to assess the WHC of the sucker flesh. Results of WHC by the centrifugation method from the water-bath-cooked flesh and from the smokehouse-cooked flesh correlated very well, r = 0.91. Replicate results per sample were fairly easy to reproduce as demonstrated by the low standard devi- ation values reported hnTable 4. An improvement in the WHC of the sucker flesh by this technique was noted by the use of additional 2% salt. The favorable affect of the 15 minute paddledmixing procedure when compared to the 2 minute cutting-mixing procedure could also be observed. Further improvement in WHC was obtained by using the proteinaceous binders in combination with RB and additional 2% salt. Finally, significant improvement of the WHC was observed when RB was used in combination with 2% additional salt and 2% SP1. In this study SPI tended to perform better than SC as binder for sucker flesh. This might be due to the better swelling and gelation properties of SPI when compared to SC 43 (Hermansson,l97l). Table 5 shows the results of various measures of texture of the smokehouse-cooked fish flesh modified by different binder systems. Void spaces in this table are due to the fact that the proteinaceous skin formed on the fish sausage during smokehouse cooking introduced some interference on the texture determination as shear values obtained were due in part to the skin resistance to breaking rather than to the meat resistance to shear; therefore further shear force determinations by both the Instron and Kramer press were carried out on the skinned cores. The higher the shear force value obtained by using these instruments the better the texture or firmness of the product was. By using the Penetrometer to measure the softness of the fish product the higher the penetration length value obtained the poorer the texture or firmness of the product was. It could be noted, from data in Table 5, that the tex- ture of the unpeeled fish cores determined by the Instron was not improved by using either 2% SP1 or 2% SC instead of RB. Fish flesh with RB plus 2% SP1 showed the best value of texture although it was significantly higher only than the fish flesh with RB plus 2% SC. Texture by the Kramer shear press on the same product showed higher shear values for the fish with RB plus 2% SP1. Fish with 2% SP1 alone as binder showed shear values significantly higher than those with 2% SC or with RB plus 2% SC. Texture values of the unpeeled 44 .Q choNuomm Eu ¢.N\coNumuuocoa SE :N mmocumomc .nmon umxoam owNOOu No w\mInN :« mouoN umosmm .u HmcoNuoom Eu o.N\Nwa :N oouoN umosmc .& NocoNuuom Eu ¢.N\Nwa :N mouoN umosmm .m oNomH cN ownNuummo mm mucoEumouu umocNmN .ANo.ovav unouoNNNo hNucmonachm mum uawuomumoam osmm osu an vozoNNoN go: so» memo :N mcmmzN m. I~.a 5. IM.@ m.ono.a e.ono.o m.ona.w ~.on~.a ~.onm.m o.onm.o s.¢-~.oN oNIz Imwmmsmm uwm mm N m co w No No w Luau ooNomo l6: :0 MO u $80..» a Iocom zoommeNNom . I . .OIoo. N.oImN. . INN. N. Inn. N.oHoo.N . “ON. N.ouNm. N.on<. muz .owmmsmm N mmen N m N m wow“ «mm : uwvwn w m :mNu ooaown :0 uoEmux m up money umosm I I I I n.0uNN.N w.mwNw.N n.0wmo.N m.omNN.m m.ome.N oNuz .mwomamm smNu emNmoa In: so muoEmuM >9 mouoN ummzm cl. .1. cl. oI. cl. .lo olo .Io olo "waNm—dmm N o+N N N 0+0 m N w+c N mommN N N mwc m m omN c wowmc N w w+c N N o+m N wSMNN ooNown co caouuch an mouoN unocm I I I I n.1nm.m _.~“N.oN m.on.m s.NHm.oH m.hnm.oN «NIz Immmmamm m a no no m swam ooNooa Ia: co mcouuch hp ouuow unocm munuxou mmommm «NIm mNIm NNIm NNIm oNIm mIm mIm NIm mIm 0» wow: unwasuum IcH can musooooum NmucoEummuH poocNm .smon uoxosm omxooo omsonoxoam ocu No ousuxou No whammos m mm Noumeouuocwm ocu an wmmCNNom osu No new .mmmuo amonm uoEmux ecu we can couuch ecu xn mono“ umwzm ocu No mcoNumN>mo oumocmum can Name: I m.oz mNm oszsn mHma mmem< aNmN> ozNzw<3 ozH3<=H szxe I N.oz MNm aNua I m.oz mNm £06: I O~IOZ MAmn vaONNON No: 30p mEmm mnu cN mammzN wm.0numw.N om.owvmn.c No.0Homm.oN mo.ownmmm.o NN.ouonwo.N No.0wmQN.o ummm cmmoum ummu cmuoum omeONm on 7. uwmw cououm .vozmms .wnNNmmB o: aon schema ummm cououw .chNvam: on Rn noxonm vmonNz uoxonm vmmmoun noxonm vmmmwua poxonm venom poxonm venom uwxunm vmocaz mu ms sou mama osu :« memo: n :no v am can v «N N oe.owmmo.N mm.ohsn~.m~ hm.shss.s~ mw.msp¢.~s Nm.0nvva.m sm.ohso~.s~ m~.~hssm.m~ as.~nsh.sm Nm.ossmo.n Ns.onmo~.m~ -.mhsm.- ms.~hsm.mm an.ossss.N so.Nhsmw.s~ sn.sssss.s~ ma.msso.sm no.0hssos.~ NN.0ummo.vN wc.qumm.nN sm.~hss.sm chNqunoo ownHONm umuwoum vco wchvcmm NsNuzv .u auo.N\qux nu vommouano .NQoNN voxoou manozoxoan on» do ounuxoa NNIzv .N nomoON ON: on vomoouaxu .snomm voxooo mono: onoan on» No unexcqum NsIzs .N summon vonmounxo .zmoNu monocoxoan onu AsIzv .N sashes vommmuaxo .snon sumo poun3 may oNx on voxoou no 033 ON: mm voxoou :0 0:3 mowuummoum Nmnouuonnh .mnowuavcoo mmeONm umnomuu vcm chchm: an vmuumwwm mm :mmNN umxonm vmconov ANNmoNcmsume may no mmNuuoaoua NnGONuucnu mag No mcoHumN>ov van scum cm monNm> can: I .02 NNN//// /////// a“ //////////// / s<