CHAPTER 9
Fish and Shellfish Freezing
Ε. HEEN A N D O . KARSTI
Norwegian Fisheries Research Institute, Bergen, Norway
I. Historical Data 355 II. Fundamental Aspects 357
A. Physical Changes 357 B. Biochemical Aspects 361 C. Protein Denaturation 365 D. Fat Stability; Rancidity 370 E. Microbiological Aspects 374 III. Technological Developments 375
A. Freezing Methods 375 B. Freezing Rate and Time 379 C. Protection, Wrapping, and Glazing 379
IV. Freezing Fish Fillets 383 V. Freezing Fish Blocks and Fish Sticks 384
VI. Storage, Transportation, and Distribution 385
A. Storage Temperatures 385 B. Fluctuating Temperatures 386 C. Cold Storage and Refrigerated Transport 386
VII. Thawing 387 VIII. Freezing at Sea 389
A. Soviet Union 389 B. United Kingdom 391 C. United States 392 D. General 393 IX. Reprocessing 393
A. General 393 B. Refreezing Frozen Fish 394
C. Canning Frozen Fish 395 D. Salting, Drying, and Smoking Frozen Fish 397
X. Freezing Shellfish 397 A. Crustaceans 397 B. Molluscs 403 References 404 I. Historical Data
Notable advances have taken p l a c e in fish freezing during recent decades. O n the other hand, c o m m e r c i a l freezing of fish was practiced m o r e than 1 0 0 years ago (Stevenson, 1 8 9 9 ; see further Borgstrom, 1 9 6 5 ) ,
and m a n h a d long before l e a r n e d t h a t fish could b e preserved b y freez
ing. Particularly in countries w i t h cold climates one resorted to natural 355
freezing in the open air. Different methods of artificial refrigeration later m a d e it possible to b e independent of natural conditions. I c e and salt mixtures w e r e employed for fish freezing early in the nineteenth cen
tury.
Several types o f freezer w e r e devised in m a n y countries (Plank, 1 9 5 2 ) . W h e n the compressor was developed in the latter part of the nineteenth century, it was immediately employed for fish freezing and rapid devel
opment followed. F r e e z i n g fish in low temperature air was first patented in the United K i n g d o m in 1 8 4 2 b y U. Benjamin. I t is not known if his ice-salt mixture was used commercially. E . Piper in C a m d e n ( M a i n e ) obtained a U.S. patent for a similar m e t h o d o f freezing fish in 1 8 6 1 , C. F . Pike, Providence ( R h o d e I s l a n d ) , used this m e t h o d on b o a r d ship as early as 1 8 6 6 - 6 7 , and the procedure was employed b y a M i c h i g a n firm in 1 8 7 5 . Almost simultaneously these n e w methods w e r e introduced in Russia and somewhat later in E u r o p e .
F r e e z i n g fish b y machinery appears to have b e e n first done b y the Russians in 1 8 8 8 at Astrakhan. Compressors w e r e installed on a b a r g e and towed up and down the V o l g a ( B o r o d i n e , 1 8 9 9 ) , and on ships fish
ing on the Caspian. L a t e r a special building was devoted to fish freezing in Astrakhan.
Immersion freezing of fish was introduced b y the F r e n c h engineer, Rouart in 1 8 9 8 . A n u m b e r of freezing houses w e r e equipped with am
monia compressors in 1 8 9 0 along the N e w E n g l a n d coast for the freezing of bait. Siberian salmon and sturgeon w e r e frozen on barges and vessels equipped with compressors from 1 9 0 3 onwards. G e r m a n companies built several fish-freezing establishments in Russia in 1 9 0 8 - 1 9 1 3 .
D u r i n g and after the first world war, pioneering investigations con
cerning the theory of freezing w e r e carried out b y Plank and his collabo
rators ( P l a n k et αι., 1 9 1 6 ) . L a r g e fish-freezing plants operating at — 2 5 ° and — 3 0 ° C . w e r e in business, however, in 1 9 1 1 in Vancouver, British Columbia, and in Seattle, Washington, having recognized the significance of rapid freezing. This period of active experimentation l e d to the development of equipment and freezing apparatus for more rapid freez
ing, rendering products of higher quality.
T w o pioneering accomplishments deserve special mention, first, that of t h e Norwegian engineer, Nekolai D a h l , who developed spray freezing with highly chilled brine solution ( 1 9 1 2 ) . Several American installations employed this procedure. Second, it was further developed and converted to a kind of immersion freezing employing eutectic brine solutions ( — 1 8 . 2 ° C . ) as the freezing medium; this invention was m a d e b y the D a n i s h fish exporter, A. J . A. Ottesen ( 1 9 1 1 - 1 3 ) .
C l a r e n c e Birdseye ( U . S . ) in 1 9 2 9 improved the plate freezer of C o o k e
9. FISH AND SHELLFISH FREEZING 357 ( 1 9 2 5 ) , and M u r p h y put forth t h e idea o f freezing in t h e retail package.
This laid a firm foundation for t h e modern phase of fish freezing. T h e freezing o f fillets was developed b y R . K o l b e ( 1 9 2 6 ) and later b e c a m e the chief m e t h o d for retailing frozen fish. T r o u t and most minor fish are still frozen in the consumer package.
II. Fundamental Aspects
A. PHYSICAL CHANGES
T h e physical changes w h i c h o c c u r during freezing and storage o f frozen products comprise crystallization of i c e with expansion of the volume, and desiccation starting from the surface of the frozen fish.
1. Ice Formation
T h e crystallization of i c e is initiated w h e n the temperature of the fish is lowered to about — 1 ° C . At the s a m e time a concentration occurs of various inorganic salts and organic components present in the fluid of the fish. Consequently the freezing point falls. T h e temperature of the fish must therefore b e lowered further before additional water freezes.
T h e r e is also an increase in the volume of the fish when the water is converted to i c e . A t — 3 ° C . about 7 0 % of the water is frozen. A t — 5 ° C . about 8 5 % is frozen, at — 2 5 ° C . about 9 5 % , and at — 5 0 ° to — 6 0 ° C . almost all the w a t e r in the fish is frozen. T h e larger p a r t of the water consequently freezes b e t w e e n — 1 ° and — 5 ° C , and it is the rate of cooling during this temperature interval w h i c h determines the size of the ice crystals.
2. Freezing Rate
T h e preserving effect of freezing, however, is due to the extent of reduction in rate of c h e m i c a l and biological processes w h e n t h e tem
perature of the fish is lowered. F r e e z i n g also results in partial "dehydra
tion/' w h i c h probably contributes to its preserving effect. On the other hand, the crystallization or freezing of the water has certain unfavorable effects. I t is well known that slow freezing results in formation of large i c e crystals. T h e s e m a y cause t h e tissue of the fish to b e c o m e so porous that perforations of the tissue can often b e seen after the fish is thawed;
it m a y even b e c o m e spongy. R a p i d freezing, on the other hand, results in small i c e crystals, and the quality of quick-frozen fish m a y b e practi
cally equivalent to that of fresh unfrozen fish. T h e rate of freezing there
fore has long b e e n considered of vital importance for obtaining a high quality product. M i c r o s c o p i c examination also shows that slow freezing results in greater destruction of the tissue than quick freezing ( R e u t e r , 1916; Birdseye, 1929; M o r a n and H a l e , 1932; Poole, 1 9 3 5 ) .
O n e of the earliest and still one of the most comprehensive investiga
tions on the effect of the freezing rate on changes in fish tissue was carried out b y R e u t e r ( 1 9 1 6 ) .
Theories of the effect of freezing led to the conclusion that large i c e crystals formed b y slow freezing are able to penetrate the cell walls, re
sulting in larger loss of drip when the fish thaws than w h e n t h e fish is quick-frozen ( T a y l o r , 1927; W e l d , 1 9 2 7 ) . Consequently it was important to freeze the fish as rapidly as possible. As a result of further investiga
tions and n e w techniques and methods, however, opinions with regard to the m e c h a n i c a l d a m a g e to fish tissue w e r e modified ( W o o l r i c h and Barlett, 1 9 4 2 ; L e b e a u x , 1947; B e r g h , 1 9 4 8 ) . M i c r o s c o p i c examinations in these cases were carried out with frozen tissue and not with thawed tissue as in earlier investigations. Certain studies (Plank, 1 9 3 2 ; F i n n , 1933; R e a y , 1 9 3 3 ; Nord, 1 9 3 6 ) appear to show that the changes occurring in freezing and thawing are caused rather b y irreversible changes in the colloidal state of the proteins induced b y the increased concentration of dehydrates in the unfrozen part of the fluid in the fish tissue. T h i s would also explain the drip loss on thawing.
B y microscopic examination of frozen meat, Hiner et al. ( 1 9 4 5 ) fur
ther concluded that rapid freezing caused m o r e d a m a g e to individual cells than slower freezing. W o o l r i c h ( 1 9 4 8 ) , on the other hand, found that cells were not disrupted b y rapid freezing.
Histological investigations (Piskarev, 1 9 5 8 ) on frozen fish revealed that, when freezing fish with considerable post-mortem changes, the ice crystals destroy the sarcolemma m o r e easily and grow to large sizes causing m o r e histological c h a n g e in the tissue.
T h e presence of fluid in t h e interfibrous s p a c e in fish frozen after storage causes the formation of i c e in the interfibrous space, w h i c h can
not b e averted even b y the most rapid freezing.
T h e size of the crystals b e t w e e n the fibers depends on the t i m e of fish storage before freezing, in turn affecting the hydrophilic properties of t h e muscle tissue and its capacity to hold the muscle juice. Piskarev et al. ( 1 9 5 8 ) observed that i c e crystal size is a d e q u a t e for determining the freezing rate and classifying frozen fish as quick- or slow-frozen;
while small ice crystals are certainly a feature of quick freezing, large ice crystals can b e the result not only of slow freezing b u t also of consider
able post-mortem changes in the tissue prior to freezing. ( S e e further Piskarev, 1 9 6 3 . )
A n e w c h e m i c a l t e c h n i q u e used b y L o v e ( 1 9 5 4 , 1 9 5 5 , 1957, 1 9 5 8 a ) has contributed to a b e t t e r understanding of the changes occurring dur
ing freezing of fish. B y this m e t h o d the deoxyribonucleic acid ( D N A ) content in t h e expressible fluid of the fish after thawing was determined,
9. FISH AND SHELLFISH FREEZING 3 5 9 and used to measure the extent to which freezing damages the cells. I t was established that the D N A content in the drip varied with the rate of freezing. L o v e ( 1 9 5 5 ) further found the amount of D N A in expressible fluid with very slow freezing to b e not m u c h above that o f unfrozen fish.
H e found that as freezing b e c o m e s m o r e rapid, little c h a n g e occurs until the time required to cool from 0 ° to — 5 ° F . b e c o m e s less than 1 5 0 min
utes, when a large increase in D N A occurs. According to L o v e ( 1 9 5 5 ) , this seems to b e caused b y the change-over from intercellular to intra
cellular freezing; the i c e masses w h i c h first form within the cells are so large they burst the sarcolemmas, liberating nuclear material among other things into the interstitial spaces and expressible fluid. W i t h m o r e rapid freezing the i c e crystals b e c o m e smaller, and there is a drop in D N A content in the expressible fluid. I n ultrarapid freezing, however, there is a rise in expressible D N A , but probably owing to a different kind of cell damage. T h e s e p h e n o m e n a apply only to fillets frozen from b o t h sides
( L o v e , 1 9 5 5 ) . W i t h freezing from one side, L o v e ( 1 9 5 7 ) found an in
crease in D N A content at freezing times, defined as the n u m b e r of min
utes elapsing until the temperature of the top layer of fish immersed in a box reaches — 5 ° C . ( 2 3 ° F . ) , of 2 5 minutes and 7 0 - 8 5 minutes. T h e dam
a g e was less in ultrarapid freezing under these conditions than in freezing from both sides.
W i t h slow freezing, L o v e ( 1 9 5 8 a ) concluded there was a zone of minimum d a m a g e at a freezing time of about 1 1 5 minutes, w h i c h is the shortest time in which all the i c e seems to b e formed in the intercellular spaces. H e also found a m a x i m u m at 2 0 0 - 5 0 0 minutes, the content of D N A decreasing thereafter. Disruption of cells at a freezing time of 2 5 minutes is, according to L o v e ( 1 9 5 8 b ) , the result of m e c h a n i c a l stress during freezing expansion of i c e crystals of a certain critical size growing within the cell and presumably affecting the cell nuclei.
As far as the pressure during freezing is concerned, investigations ( L o v e a n d Karsti, 1 9 5 8 ) indicate that the pressure on fish during the freezing process has little effect on cell damage, and is less important than the effect of the freezing rate.
3. Desiccation
Desiccation from the surface of the frozen fish takes p l a c e during freezing and storage b e c a u s e there is always transport of w a t e r vapor from the fish to the evaporator or cooling coils. T h e greater the temper
ature differential b e t w e e n fish and evaporator, and the greater t h e veloc
ity of the circulating air, the m o r e moisture will b e transported and the more rime will gather on the evaporator. T h e loss of water from the product and the formation of rime or snow on the evaporator coils are
greater at higher storage temperatures, b e c a u s e the w a r m e r air will take up m o r e moisture. C h a n g e or fluctuation in storage temperature influ
ences the desiccation loss of w e i g h t and quality of t h e fish, and con
tributes to a poor a p p e a r a n c e or results in "freezer burn." I f the desicca
tion is pronounced the fish surface m a y b e c o m e dry a n d fibrous; in some cases the skin m a y c h a n g e color. Several other factors influence the loss of weight, e.g., the kind of wrapping and its sealing and moisture trans
mission characteristics.
Medium-sized w h i t e fish stored in a rather dry room lost 7, 3.5, and 1.5% in weight p e r m o n t h at — 9 , — 2 1 , and — 2 9 ° C , respectively; w h e n the fish w e r e w r a p p e d in p a r c h m e n t and stored in w o o d e n cases the loss was 2 . 5 , 1.0, and 0 . 2 5 % ( R e a y et al., 1 9 5 0 ) . Under certain conditions the loss of w e i g h t m a y b e even greater. Kondrup ( 1 9 4 8 ) found the re
sults given b e l o w for fillets w r a p p e d in different kinds of p a p e r a n d cellophane, stored at — 1 8 ° C . with an air circulation of 1 0 m . / s e c .
% Loss in weight of fillets stored for Wrapping material 3 months 4 months
Parchment substitute (1 layer) 20.5 24.0 Parchment substitute (2 layers) 13.5 16.5
Waxed paper (110 g./m.2) 11.0 14.5
Cellophane MSST 400 (45g./m.2) 0.5 0.8 T h e loss during storage of kippers at — 1 0 , — 2 0 , and — 2 8 ° C . in 2 5
weeks is shown in the tabulation b e l o w ( B a n k s , 1 9 5 2 a ) .
Storage time (in weeks) of kippered herring
Storage temperature 4 8 12 16 20 25
—10°C. 1.6« 3.5 4.6 6.5 7.6 8.9
—20°C. 0.0 0.9 2.2 2.3 3.0 3.2
—28°C. 0.0 0.4 0.8 1.2 1.4 2.1
a Figures show percentage reduction of original weight.
D u r i n g storage of frozen herring t h e desiccation m a y also result in loss of flavor and color and m a y contribute to m o r e rapid development of rancidity (Kuprianoff, 1 9 5 6 b ) .
4. Toughness
F r e e z i n g increases toughness o f fish muscle, w h i c h proceeds pro
gressively during subsequent storage. L u i j p e n ( 1 9 5 7 ) observed that t h e proportion of fish m u s c l e protein soluble in 5 % N a C l solution was not correlated with the development o f toughness during cold storage at or b e l o w — 2 0 ° C . T h e r e f o r e toughness cannot b e measured in terms of per
c e n t a g e of soluble protein nitrogen in the cold-stored fish. Assessment b y
9. FISH AND SHELLFISH FREEZING 3 6 1 a taste panel is time-consuming and furthermore inaccurate. Since fish muscle fibers are extremely short and interspaced with m y o c o m m a t a , instruments such as penetrometers have limited usefulness.
L o v e ( 1 9 5 8 d , 1 9 5 9 ; also L o v e and M a c k e y , 1 9 6 2 ) devised a m e t h o d to measure toughness of uncooked cod muscle. A p i e c e of muscle ( 0 . 2 g.) with fibers of uniform thickness was d e t a c h e d from the m y o t o m e on the "bone" side of the fillet b y means of a "double scalpel." T h e p i e c e of muscle was freed from traces of connective tissues and half of it was added to 2 0 ml. 1 % formaldehyde kept b e l o w 4 ° C . T h e material was then homogenized. I t was found that the unfrozen fish muscle fibers w e r e broken up b y this treatment, reducing the light transmission. F r o z e n fish showed a higher proportion of intact fibers, letting through more light.
Toughness was related rather to the connective tissue, the destruction of which releases the fibrils, than to toughening of the cells as such.
5. Discoloration
F r o z e n tuna and swordfish m a y exhibit green and brown discoloration on cooking, w h i c h lead to rejection b y consumers. I t is known that un
cooked fish m e a t contains three derivatives o f myoglobin, its principal pigment ( B r o w n et al, 1 9 5 8 ; Sano and Hashimoto, 1 9 5 8 ; Sano et al., 1959; Y a m a m o t o , 1 9 6 0 ) . T h e p i g m e n t responsible for the pink color in normal cooked m e a t of tuna is h e m o c h r o m e , derived from the reaction of myoglobin with non-heme constituents.
B r o w n i n g is due to formation of metmyoglobin in the muscle through autoxidation of ferrous myoglobin. G r e e n i n g is due to pigments resulting from the oxidation of h e m o c h r o m e that occurs when the m e a t is unduly exposed to oxidative conditions during and after cooking. Amano ( 1 9 5 0 ) and Amano and T o m i y a ( 1 9 5 5 ) maintain that greening of frozen sword- fish m a y b e related to the hydrogen sulfide produced b y putrefactive bacteria. T h e color develops w h e n hemoglobin or myoglobin is altered by the action of hydrogen sulfide.
Proper evisceration and removal of blood immediately after the catch reduce the risk of discoloration. T a n a k a ( 1 9 6 1 ) suggests that undesirable discoloration in yellowfin tuna m e a t can b e averted if the fish is frozen at full rigor, stored at a temperature of — 2 3 ° to — 2 7 ° C , and defrosted by still air at 1 0 ° C . H i g h e r temperatures are less favorable.
B . BIOCHEMICAL ASPECTS
Although it is impossible to separate the physical changes in fish muscle due to the influence of freezing, storage, and thawing on the mechanism of enzymic processes in the tissue, it is justifiable to discuss the b i o c h e m i c a l changes separately.
1. Glycolysis
T h e glycolytic cycle in animal tissue is now comparatively well under
stood. B a t e - S m i t h ( 1 9 4 4 ) , Sharp ( 1 9 3 4 ) , and Larsen ( 1 9 4 4 ) have con
tributed to our knowledge of the effect of freezing and storage on this system in mammalian and fish muscles. Nord and B i e r ( 1 9 5 2 ) have presented a general theory of the influence of freezing and thawing on colloid systems in their aggregation-disaggregation hypothesis, a n d Kier- meier ( 1 9 4 9 ) has evaluated the enzymic processes during freezing, thawing, and storage.
An over-all picture shows that the rate of glycolysis increases during freezing, and that storage in the frozen state gradually reduces the ac
tivity of one or m o r e of the enzymes involved in the cycle. I t is probable that the latter phenomenon depends on denaturation of one or more protein constituents in the enzymes. Pre-rigor cod m u s c l e exhibited an increased rate of glycolysis after freezing; b u t after 2 months of storage at — 1 7 ° C , glycolysis ceased after thawing ( L a r s e n , 1 9 4 4 ) .
Details of the inhibition of the glycolytic cycle during freezing stor
age are rather scarce. A m o n g the m a n y enzymes involved, aldolase ac
tivity has b e e n studied; no appreciable effect of freezing and storage on this link in the chain was found ( R a m b e c k and Connell, 1 9 5 5 ) .
2. Phosphatase
T h e behavior of muscle phosphatase during freezing and storage is of particular interest, since this enzyme is involved in the splitting of adenosine triphosphate ( A T P ) , w h i c h is known to have a special func
tion in dissociating actomyosin. T h i s particular protein is known to b e denatured during freezing storage, and is at least partly responsible for the undesirable changes in texture of fish muscle during freezing storage.
3. Actomyosin
T h e development of toughness in frozen cod muscle is correlated with the amount of extractable actomyosin (Connell, 1960a; D y e r et al, 1957b; L o v e , 1 9 5 9 ; Luijpen, 1 9 5 7 ) . Similar correlations have b e e n found in other species of fish during frozen storage ( D y e r , 1 9 5 1 ; D y e r et al, 1956; D y e r and Morton, 1 9 5 6 ; Nikkilä, 1957; Nikkilä and L i n k o , 1 9 5 4 ) .
Although actomyosin denaturation has b e e n measured b y various methods (Connell, 1960b; Heen, 1954; Matsumoto, 1 9 5 8 ; Partmann, 1957, 1960; Sawant and Magar, 1 9 6 1 ) , the criterion for denaturation resulting in texture changes of tissue is the loss of extractability. M e a t of terrestrial animals toughens less than that of fish. Connell ( 1 9 6 1 ) observed that myosin extracted from cold-blooded animals aggregates spontaneously at a faster rate than myosin from warm-blooded animals.
9. FISH AND SHELLFISH FREEZING 363 T h e mechanisms b y w h i c h actomyosin ( 1 ) aggregates, ( 2 ) loses m u c h of its b o u n d w a t e r (forming d r i p ) , a n d ( 3 ) b e c o m e s insoluble, are not well understood ( C o n n e l l , 1 9 6 2 ) . Connell ( 1 9 5 8 , 1 9 6 0 c ) observed that only a minor p a r t of the molecular surface of cod actomyosin is responsible for its relative instability, since the other properties of cod and rabbit actomyosin are very similar. T h e r e m a y b e a relationship b e tween lipid hydrolysis and actomyosin denaturation in frozen muscle.
D y e r and F r a z e r ( 1 9 5 9 ) , Olley and L o v e r n ( 1 9 6 0 ; Lovern, 1 9 6 2 ) , and O t a and Nishimoto ( 1 9 6 3 ) reported a correlation b e t w e e n increase in free fatty acid content and decrease in actomyosin extractability in frozen cod muscle. D y e r and F r a z e r ( 1 9 5 9 ) postulated that either the stabiliz
ing effect of intact lipids on actomyosin is destroyed b y lipid hydrolysis, or fatty acids formed from lipid hydrolysis cause actomyosin inextracta- bility. K i n g et al. ( 1 9 6 2 ) support this assumption since the accumulation of free fatty acids, such as linoleic and linolenic, rapidly reduces the solubility of cod actomyosin. O n the other hand, thorough studies of the two p h e n o m e n a in cod, lemon sole, halibut, and dogfish indicate that in this respect there is no simple cause-and-effect relationship ( O l l e y et al, 1 9 6 2 ) .
T h e disappearance of A T P is the sum of the splitting and the re- synthesis; these are both enhanced during freezing, b u t the degradation rate is predominant, so that a total disappearance m a y occur even during the freezing procedure. T h e amount varies not only with variety of fish but also with formation of ice crystals in the tissue ( B i t o and Amano, 1 9 5 9 ) . Resynthesis m a y set in during thawing and rigor m a y occur in the form of a thawing contraction. I f the frozen muscle has b e e n stored for considerable time at moderate temperatures, resynthesis of A T P will not o c c u r and rigor during thawing is not observed.
T h e s e p h e n o m e n a m a y a c c o u n t for s o m e of the discrepancies in opinions on the influence of rigor on frozen fish and on freezing fish b e fore, during, and after rigor mortis.
Splitting of A T P is regulated through a phosphatase system ( A T P a s e ) which is little affected b y freezing and w h o s e activity remains nearly u n c h a n g e d during freezing storage at — 2 0 ° to — 3 0 ° C . B e t w e e n — 2 ° and — 1 0 ° C . a definite loss of A T P takes p l a c e (Partmann, 1 9 5 7 ) . Saito and Arai ( 1 9 5 7 , 1 9 5 8 a , b ) h a v e shown that, at these latter temperatures, inosinic acid is formed in amounts indicating that deaminases are com
paratively active at this temperature level.
4 . Oxidative Changes
Another important group of enzymes is c o n n e c t e d with the oxidative changes in frozen fish. Among these is c y t o c h r o m e oxidase, a powerful
catalyst w h i c h in muscle tissue is activated b y salt and is partly respon
sible for the increased rate of oxidation in fish frozen in brine ( B a n k s , 1 9 5 2 a ) .
T h e "dark" muscle (musculus lateralis superficialis) contains higher amounts of myoglobin ( H a m o i r , 1 9 5 5 ) , and also of vitamins w h i c h are part of the oxidative enzymes ( B r a e k k a n , 1 9 5 6 ) , than the white muscle.
I t is probable that the frequently observed high susceptibility to oxidative changes in this part of the tissue is related to elevated oxidase activity, including the transfer mechanism from molecular oxygen to the final acceptor.
5. Antioxidants
Studies on the effect of antioxidants on oxidative changes in fat and lean fish show a m o r e confused picture. Antioxidants, highly active in the autoxidation of unsaturated fatty acids in the isolated glycerides, may fail when the lipids are in situ in the muscle tissue. Ascorbic acid has b e e n extensively studied in freezing storage of salmon, trout, herring, and tuna ( B a u e r n f e i n d et al, 1 9 5 1 ; Banks, 1 9 5 3 ; H e e n and Karsti, 1 9 5 0 ; Bramsnaes et al, 1 9 5 9 ) with varying results. Olcott ( 1 9 5 7 ) has drawn attention to t h e possibility that different antioxidants m a y b e specific for oils from diverse species o f fish.
T h e oxidation of unsaturated fish lipids is characterized b y an induc
tion period followed b y an accelerated rate of oxygen absorption with simultaneous development of peroxides, rancid odors, and polymerized products. T h e rate of initiation of free radical chains is enhanced b y light, heat, irradiation, and heavy metals. F r e e radicals r e a c t with oxygen to yield peroxy radicals which take hydrogen from substrates, producing hydroperoxides and n e w free radicals. A single chain is thus continued but n e w chains result from the breakdown o f hydroperoxides, producing n e w free radicals. Ultimately a plethora of reaction products results
( O l c o t t , 1962a; see also Olcott, 1 9 6 2 b ) .
T h e attack on fish lipids can b e avoided b y adding antioxidants which a c t as free radical chain breakers or as peroxide decomposers.
Unusual synergistic effects are found with mixtures of antioxidants. T h e cause of these variations is not known, b u t appears to depend on ( 1 ) nature of the substrate, ( 2 ) temperature, and ( 3 ) certain other variables.
T h e nutritional damage from ingestion of oxidized fish lipids is caused b y the toxicity of the peroxides as such or b y further oxidative reactions in vivo. Polymers would also b e damaging as they are n o longer nutrition
ally available. Numerous other end products of oxidation m a y also have deleterious effects. Generally, the adverse effects of fish oils oxidized in vitro or in vivo reflect secondary d a m a g e due to formation of free
9. FISH AND SHELLFISH FREEZING 3 6 5 radicals from the decomposition of peroxides. T h u s the ability of free radicals or peroxides to destroy vitamins A and Ε is reflected in various deficiency symptoms.
Preservation of the nutritive value of fish products depends upon preventing the oxidation of lipids. T h i s can b e achieved b y ( 1 ) remov
ing lipids, ( 2 ) preventing access to air or oxygen, ( 3 ) avoiding contami
nation b y pro-oxidants like heavy metals, ( 4 ) adding antioxidants or combinations of antioxidants and synergists, ( 5 ) avoiding irradiation in
cluding light, and ( 6 ) using low storage temperatures.
M a r c u s e ( 1 9 6 0 , 1961a,b, 1 9 6 2 ) observed that certain amino acids, but not cysteine, have a strong antioxidative effect. T h e antioxidative ca
pacity of different amino acids differs; it is especially pronounced in histidine and tryptophan. M a r c u s e ( 1 9 6 1 a ) found that histidine could reduce the oxidation of herring oil b y 8 5 % . T h e antioxidative effect is enhanced and a pro-oxidative effect can b e suppressed or converted to an antioxidative effect b y the addition of a phosphate or an emulsifier. A strong inhibitory effect is obtained b y the c o m b i n e d addition of a phos
phate and an emulsifier together with the amino acid. T h e action of primary antioxidants, e.g., tocopherol, is greatly e n h a n c e d b y amino acids. T h e antioxidative tendency is stronger in the case of M e linoleate than with linoleic acid or with M e linolenate.
C . PROTEIN DENATURATION 1. Drip
D r i p is defined as the exudate of tissue fluids that flow free from fish m u s c l e during thawing. Its composition m a y vary, b u t its nitrogen content seems to b e derived entirely from the m y o g e n of the muscle cells ( D y e r et al., 1 9 5 6 ; Saito and Arai, 1 9 5 7 ) . D r i p was originally interpreted as a result of cell damage, and this is still valid as far as the exudate re
sulting from freezing and thawing is concerned. F u r t h e r release of free water is taken as a sign of dehydration of protein micelles due to de
naturation of t h e proteins. This phenomenon was confirmed b y Akiba ( 1 9 5 5 ) on Atka m a c k e r e l flesh. T h e free water test has b e e n extensively used as a quality test for frozen fish, b u t has definite limitations ( R e a y and Kuchel, 1 9 3 6 ; Banks, 1 9 5 5 ; Notevarp and H e e n , 1 9 3 9 ; E m p e y and Howard, 1 9 5 1 ) .
2. Protein Solubility
This denaturation, measured as loss of solubility in defined salt solu
tions, decreases with lowering of the storage temperature ( F i n n , 1 9 3 4 ) . T h e denaturation was traced to the globulin fraction of the muscle pro
tein, while the m y o g e n fraction was unaffected ( R e a y , 1 9 3 3 , 1 9 3 4 ) . T h e
great progress in our knowledge of the m e c h a n i s m of muscular contrac
tion ( M o m m a e r t s , 1 9 5 0 ) and of the protein components of fish muscle ( H a m o i r , 1 9 5 5 ) has given a better basis for understanding these phe
nomena.
D y e r ( 1 9 5 4 ) m a d e an extensive study of the denaturation of fish muscle actomyosin and found a general correlation b e t w e e n loss of solu
bility of this protein and organoleptic changes, notably in texture or con
sistency. H e e n ( 1 9 5 4 ) drew attention to the p r o b a b l e limitations in such a correlation, since actomyosin gels show "superprecipitation' in such a way that solubility should not b e expected to reflect texture changes b e yond a certain level. Experiments b y Luijpen ( 1 9 5 7 ) support this view.
Although actomyosin denaturation m a y play an important part in the texture changes during freezing storage, toughness m a y also result from changes in the cell membranes, as suggested b y L o v e and Ironside
( 1 9 5 8 ) .
F r o m his great n u m b e r of experiments, L o v e further assumes that inorganic salts play a decisive role in actomyosin denaturation, and that denaturation is almost arrested w h e n the temperature falls b e l o w the cryohydratic point of N a C l ( — 2 1 . 4 ° C ) . A t storage temperatures at the
— 3 0 ° C . level n o actomyosin denaturation takes place; nonetheless tex
ture changes m a y develop at this temperature.
F r o m storage experiments at very low temperatures and from con
ventional organoleptic and physical tests, H e e n ( 1 9 5 4 ) advanced the hypothesis that the dominating reactions in texture changes are mono- molecular, irreversible, and spontaneous. This would indicate that the reaction rate would r e a c h zero only at 0 ° K . T h e work of L o v e shows that such a relation is not valid as far as actomyosin denaturation is con
cerned, although the experimental figures indicate a first order reaction rate at moderate storage temperatures ( — 1 4 ° C ) .
T h e question whether molecular oxygen plays a part in the denatura
tion of proteins and texture changes is not definitely answered, even though fatty fish are more susceptible to oxidation. Stansby ( 1 9 5 5 ) ob
served a detrimental effect of ready access of air even in lean fish, and Karsti and H a k v a g ( 1 9 6 1 ) observed a profound difference in the texture of shrimps.
T h e possible protective action of lipids on protein denaturation, as suggested b y D y e r et al. ( 1 9 5 7 b ) , is a phenomenon of considerable in
terest.
3. Brining
T h e effect of brining fish fillets prior to freezing, w h i c h gives con
siderable reduction in drip and free water, was originally explained as an osmotic effect. In light of the properties of myosin, its sorption of N a +
9. FISH AND SHELLFISH FREEZING
367
and K + bringing about a shift in isoelectric point and a charging of the protein, the effect of N a C l and K C l m a y simply b e an increased hydra
tion of the higher charged protein micelles. T h e peculiar and somewhat confusing effect of bivalent ions, as M g+ + and C a + + , on the muscle myosin and actomyosin has b e e n little studied in connection with freez
ing denaturation. Experiments with blocking these ions through chelating compounds have b e e n reported ( H e e n , 1 9 5 4 ) .
Glycerol is known to exert a protective action on the preservation b y freezing of b a c t e r i a and spermatozoa. T h e effect seems to b e limited to the freezing period proper and has apparently no influence on the storage denaturation of proteins ( H o l l a n d e r and Nell, 1 9 5 4 ) .
T h e fundamental c h e m i c a l changes on the molecular level in freezing denaturation of proteins are not well understood. Simidu and Hibiki
( 1 9 5 1 ) indicated a similarity to denaturation that is induced b y salts.
It is further established that parallel to denaturation there is a certain liberation o f S H groups ( S e a g r a n , 1 9 5 4 ; Luijpen, 1 9 5 7 ) , a process that seems to b e a general feature of protein denaturation.
X-ray diffraction studies of freeze-denatured fish proteins are very scarce and h a v e revealed little information on the structure of denatured proteins ( H e e n , 1 9 4 2 , unpublished; Seagran, 1 9 5 8 ; Connell, 1 9 5 7 ) .
4. SarcopL·sma
I t has b e e n suggested that disturbance of t h e intracellular hydrated contractile protein equilibrium rather than of the sarcoplasmic fraction is responsible in p a r t for the liberation of drip from t h a w e d fish muscle.
A definite similarity was found in the protein composition of drip and of sarcoplasmic extracts of low ionic strength obtained from fresh fish muscle b y the m e t h o d of D i n g l e et al. ( 1 9 5 5 ) . Patterns obtained with centrifuge drip from frozen and thawed rockfish were compared with those of extracts of low ionic strength from fresh rockfish. No differences were revealed in the n u m b e r of components present. No relative c h a n g e took p l a c e in mobilities of the component proteins of drip as a result of freezing fish muscle. I t is concluded that the sarcoplasmic protein frac
tion of fish muscle is not intimately associated with the origin of the drip, because of ( 1 ) the similarity in n u m b e r and ionic characteristics of the protein constituents of the drip and of the extracts of low ionic strength from fish muscle, and ( 2 ) the a b s e n c e in drip of the contractile protein, actomyosin.
W h i l e in general the rate of denaturation is governed solely b y the subzero storage temperature, b e c o m i n g progressively less as the tem
perature is reduced ( L o v e , 1 9 6 2 ) , one particular condition of freezing can increase the rate of denaturation considerably, even at low temper-
ature ( L o v e , 1 9 5 8 c ) . At the critical freezing rate w h e n the muscle cools from 0 ° to — 5 ° C . in about 6 0 minutes, each cell contains just one i c e crystal which grows up the center. I t was concluded that the minerals and small organic molecules, like sugars ( L o v e , 1 9 5 8 c ) , b e c o m e more concentrated towards one side of each cell under these circumstances than under any other conditions. Denaturation seems to b e related to the presence of concentrated salts, etc., in liquid solution, and therefore would cease altogether at very low temperatures when the last traces of liquid have disappeared.
Akiba ( 1 9 5 5 , 1 9 6 1 ) was the first to point out that if an initial treat
ment prior to freezing is given to keep the water bound, any decline in its amount can b e controlled both in the freezing process and during storage. This is actually the basis of the methods recently heralded as a breakthrough in fish-freezing technology. T h e fillets are submitted to a polyphosphate dip, prior to freezing, w h e r e b y the water-binding prop
erties of the proteins are held at a maximum.
5. Phosphate Dips
Polyphosphates have b e e n employed in a n u m b e r of foods for water- binding purposes. A special symposium on this matter was held in 1957 and the proceedings w e r e published in 1958 (Anonymous, 1 9 5 8 a ) , with special reviews on the mechanism of the polyphosphate effect on pro
teins (Kotter, 1 9 5 8 ) and the use of these compounds in the manufac
turing of cottage cheese ( M a i r - W a l d b u r g , 1 9 5 8 ) and meat sausages ( G r a u , 1 9 5 8 ; see also Bianchi, 1 9 6 0 ) .
No full agreement prevails as to how polyphosphates exert their water-binding effect on actomyosin. Paradoxically enough, there is a dissociation of this molecule into actin and myosin and thus a reduction in viscosity. T h e hydration effect seems to hinge on the presence of so
dium chloride or M g salts, but to b e counteracted b y C a ions.
Polyphosphates are used for the same beneficial hydration effect in fish sausages ( O k a m u r a et al., 1 9 5 8 a , b ) . In this case, they also a c t as a germicide and reduce microbial softening reactions ( s e e V o l u m e I I I , Chapter 1 0 ) .
T h e use of polyphosphates in fish sausages corresponds to the tradi
tional addition of starch as a binding agent, which was frequently the source of contamination or stimulus to microbial growth.
I n latter years, polyphosphates have b e e n employed successfully in frozen fish to reduce the amount of drip ( M e y e r , 1 9 5 6 ; Nikkilä et al., I 9 6 0 ; Mahon, 1962, 1 9 6 3 ; T a n i k a w a et al., 1 9 6 3 ) . T h e mechanism of action appears to b e that indicated above, namely, the effect on the water-bind
ing capacity of the fish proteins.
9. FISH AND SHELLFISH FREEZING 3 6 9 In order to reduce drip b y this process, fish fillets prior to freezing at 0 ° F . are dipped in an aqueous solution containing a single sodium or potassium phosphate or a mixture of two phosphates having a molar ratio of alkali m e t a l oxide to P205 of about 1:1 or 2 : 1 ; the weight o f fluid lost on thawing after storage for up to 8 5 days is decreased as com
pared to control fillets dipped in water or 4 - 5 % N a C l solution (Nikkilä, et al, 1 9 6 0 ; Mahon, 1 9 6 2 , 1 9 6 4 ) . T h e most effective dip solution contains 1 2 . 5 % sodium tripolyphosphate.
No undesirable side effects have b e e n noted ( M a h o n , 1 9 6 2 ) ; these additives seem to aid in reducing yellow discoloration and off-flavors caused b y rancidity during storage, particularly when added to a glazing mixture of alginate or related substance (Nikkilä et al, 1 9 6 0 ) .
6. Rigor Mortis and Freezing
T h e process of rigor mortis, in w h i c h the muscle tissue of animals stiffens at death retarding the post-mortem autolytic and bacterial de
composition of the flesh and its protein, has b e e n the subject of m a n y in
vestigations. ( F o r a review of rigor mortis in fish, see R e a y and Shewan, 1949; Partmann, 1 9 5 4 ; a n d Vol. I, C h a p t e r 8 . )
Rigor mortis in fish starts soon after death and lasts for a varying period. I t lasts longer in fish w h i c h h a v e b e e n less muscularly active prior to death and in fish chilled immediately after the catch. A longer rigor mortis period can b e assured b y careful handling of fish during the c a t c h and on board and is of great economic importance.
T h e chief factor in rigor mortis appears to b e disappearance of A T P . Stiffening o f the m u s c l e reduces the solubility of actomyosin. T h e solu
bility decreases to a low value of about 3 0 % in herring when frozen pre- rigor and does not c h a n g e during 3 0 days of storage (Nikkilä a n d Linko, 1 9 5 6 ) . B u t w h e n herring is frozen during rigor (Nikkilä and Linko, 1 9 5 4 ) , it was found that the rigor gradually resolved during 12 days of storage at — 2 0 ° C , and that the protein solubility returned to normal from the low pre-rigor values. C o d frozen pre-rigor shows less protein denaturation than that frozen post-rigor, and this difference is evident under a w i d e range of conditions. I n the first case the protein is less soluble in salt solutions than in the second instance, provided the muscle had b e e n able to contract freely on defrosting ( L o v e , 1 9 6 2 ) . T h i s con
dition prevails only in small pieces of the order of 1 g. or in thin strips w h i c h are t h a w e d quickly. However, when m u s c l e fragments are t h a w e d slowly, or w h e n whole fillets or whole fish are employed, no contraction takes p l a c e and the protein is not affected. Changes in labile phosphate compounds during rigor freezing have b e e n discussed b y H e e n ( 1 9 5 4 ; see also Tomlinson et al, 1 9 6 2 ) .
T h e timing of the filleting process in relation to the progress of rigor mortis is also important. A fillet taken from a pre-rigor cod is likely to contract permanently with changes in translucency, color, texture, and surface bloom. Steps must b e taken to delay the onset of rigor until the fillet is frozen ( D y e r and Frazer, 1 9 6 1 ) . I f the fillet is then held for long in cold storage, the rigor process will presumably b e fully resolved b e fore thawing, b u t B a n k s ( 1 9 6 2 a ) points to the danger of "thaw rigor."
In trawling operations, with the varying rates of catch, the mending of nets and other interruptions, and the inevitable delays in handling some of the fish w h i c h will b e in varying states of exhaustion and shock if not already dead, it does not seem practicable to freeze fillets pre-rigor
( E d d i e , 1 9 6 2 ) . W h e r e the appearance of thawed fillets is important, it seems essential, as in the context of freezer trawlers, to wait until rigor is resolved before filleting the cod.
7. Nitrogenous Changes
T h e amino nitrogen level rose to approximately 5 times the original in the freezing storage of perch, mackerel, and cod p a c k e d in plastic bags or cellophane. This took p l a c e within 4 months, after which time no further amino nitrogen was formed ( B o t a l l a et al., 1 9 5 5 ) . I t was con
cluded that enzymic proteolysis proceeded in spite of the low tem
perature.
Ammonia showed a trebling to doubling in amount for four months during storage of fillets of perch, mackerel, and cod in various bags (plastics and c e l l o p h a n e ) ( B o t a l l a et al, 1 9 5 5 ) . After this time the level fell, but was maintained above that of the original sample.
A constant build-up of T M A (trimethylamine) was observed in long- term studies on frozen m a c k e r e l fillets. T h e s e fillets w e r e packed in cans to avoid losses through foil transmission. Italian workers found a small initial rise in the T M A level of perch, mackerel, and cod p a c k e d in bags of two plastics and cellophane; after 2 months the level fell off ( B o t a l l a et al., 1 9 5 5 ) . T h e y concluded that no bacterial activity was involved, but rather tissue enzymes w h i c h activate the lactic reduction of T M A O
(trimethylamine o x i d e ) to T M A .
D . F A T STABILITY; RANCIDITY
T h e sensitivity of marine oils and fats to oxidative deterioration is of particular significance in the preservation of fish b y freezing. T h e fundamental reaction is generally a c c e p t e d to b e the process of autoxida- tion of the unsaturated fatty acids, w h i c h are abundant in fish lipids.
No doubt the highly unsaturated fatty acids are very reactive. Conse
quently fatty fish in particular are very susceptible to oxidative changes.
9. FISH AND SHELLFISH FREEZING 3 7 1 Additionally there are e n z y m e systems in the tissue acting as pro-oxidants which m a y play an equally important part in oxidation and the resulting rancidity.
T h e red lateral b a n d is the most sensitive tissue. I t has long been observed that some fatty fish, such as herring, mackerel, salmon, and tuna, present special problems in protecting the "skin side" from rancidity and discoloration. Since molecular oxygen is required for autoxidation of unsaturated fatty acids, protection from the oxygen of the air has b e e n the main r e m e d y in preventing rancidity.
Oxygen is practically insoluble in ice. Glazing the frozen fish gives rather effective protection against development of rancidity, provided the glaze is sufficiently thick and the sublimation of the glaze is counter
a c t e d ( B a n k s , 1 9 5 2 a ) . T h e different practical means of obtaining such a protective glaze and the effect of oxygen removal b y applying a barrier through appropriate p a c k a g i n g will b e dealt with elsewhere.
T h e rate of autoxidation of marine oils is highly dependent on the temperature. Banks ( 1 9 5 2 a ) has given an illustration of the oxidation of herring oil from w h i c h it appears that t h e temperature level of — 3 0 ° to — 3 5 ° C . should protect the fat from oxidation during c o m m e r c i a l stor
age periods. However, w h e n these fats are dispersed in the tissue w e are justified in assuming that the previously mentioned pro-oxidants m a y in
fluence the rate ( a n d possibly the p a t h w a y s ) of oxidation. Practical experiments on a rather large scale show an appreciable development of rancidity even at storage temperatures b e l o w — 3 0 ° C , b u t the rate is considerably slower than at — 2 0 ° C . L o w storage temperature is in fact a very important means of retarding fat oxidation in frozen fish.
T h e accelerating effect of metal catalysts like c o p p e r and iron on oxidative processes is generally recognized ( T a r r , 1 9 4 5 ) , and Castell et al.
( 1 9 6 2 ) have shown that traces of these metals a d d e d to fish muscle cause the development of m a n y off-odors, p r o b a b l y b e c a u s e of changes in fatty materials in fish. T h e early observation of B a n k s ( 1 9 3 5 ) , that even pure salt ( N a C l ) in combination with herring tissue h o m o g e n a t e acted as a catalyst on herring oil, was later interpreted as a physical effect of depressing the cryohydratic point in tissue fluids ( B a n k s , 1 9 5 2 a ) . T a r r ( 1 9 4 7 ) observed that b o t h N a C l and N a N 02 a c c e l e r a t e oxidation of fat in frozen fish, so that light brining intended for freezing causes the fish to b e c o m e rancid m u c h m o r e rapidly.
S o m e antioxidants, of w h i c h w e n o w h a v e a large variety are also active in fish tissue ( T a r r , 1 9 4 7 ) , b u t show extremely varying results in different species as well as in fish of the s a m e species ( B a n k s , 1 9 5 2 b ) . I n northern waters, herring have the highest fat content in summer and the fat composition shows the highest proportion o f unsaturated fatty
acids in this period. Accordingly summer herring is m o r e susceptible to oxidation than winter herring. However, the effect of ascorbic acid as antioxidant, applied as dip and in glazing, was very different in winter and summer herring, showing little if any protection in the latter. T h i s behavior m a y b e explained b y a different distribution of the fat; the summer herring is in a w a y "coated" with oil, forming a barrier to a b sorption of this water-soluble antioxidant. E x t e n s i v e studies on the numer
ous potential antioxidants are n e e d e d before a conclusion can b e drawn as to the merits of these compounds in fish freezing. T h e indication of a certain specificity of the different antioxidative compounds for marine oils from different species is of particular significance ( O l c o t t , 1 9 5 7 ) .
T h e adverse effect of freezing fatty fish in brine is n o w generally recognized. T h e following figures from B a n k s ( 1 9 5 2 a ) illustrate this effect in comparison with air-blast freezing of herring ( T a b l e s I and I I ) .
TABLE la
DEVELOPMENT OF PEROXIDES IN SUPERFICIAL F A T OF LEAN HERRING STORED AT Low TEMPERATURES
Peroxides0 after storage for Storage
temperature
4 8 12 16
(weeks)
20 23 28
Brine-frozen
—20°C. 6.7 10.0 13.3 29.2 25.8 25.8 38.2
—28°C. 1.3 5.0 9.8 7.2 20.2 9.6 17.3
Air-frozen
—20°C. 0.7 1.1 2.0 2.4 3.0 3.5 10.2
—28°C. 0.5 0.2 0.3 0.5 2.0 1.1 2.6
a Banks (1952a).
0 In ml. 0.002 Ν thiosulfate per g.
TABLE II»
DEVELOPMENT OF PEROXIDES IN SUPERFICIAL F A T OF FATTY HERRING STORED AT Low TEMPERATURES
Peroxides0 after storage for Storage
temperature
2 4 8 12 16
(weeks)
20 24 29 Brine-frozen
—20°C. 4.9 5.4 7.4 13.4 24.8 25.2 30.6 34.2
—28°C. 1.8 1.9 3.22 5.2 4.8 14.2 27.0 16.0 Air-frozen
—20°C. 2.6 5.6 9.3 3.2 20.4 24.8 43.7 42.3
—28°C. 2.2 1.7 3.0 2.6 4.5 4.7 7.6 7.8
a Banks (1952a).
& In ml. 0.002 Ν thiosulfate per g.
9. FISH AND SHELLFISH FREEZING 373 In this respect the different behavior of summer and winter herring is very interesting. T h e leaner winter herring is m o r e susceptible to the effect o f salt, p r o b a b l y due to t h e oil's constituting a barrier to t h e penetration o f salt in t h e summer herring.
R e m o v a l of oxygen ( a i r ) either b y evacuation or b y r e p l a c e m e n t with inert gases is a highly efficient means o f reducing oxidative changes in fatty fish ( T a r r , 1 9 4 8 ) ( T a b l e I I I ) . S o m e side effects of C 02 a n d t h e technical complications involved have m a d e this procedure less attractive for practical application. V a c u u m p a c k is p r a c t i c a b l e for some special products and further progress along this line should b e expected.
TABLE III«
DEVELOPMENT OF PEROXIDES IN WHOLE F A T OF HERRING STORED UNDER DIFFERENT CONDITIONS AT — 1 0 ° C . ( 1 4 ° F . )
Peroxides0 after storage for
Conditions of storage 6 weeks 1 4 weeks
In air (unbrined) 7 . 5 1 8 . 4
In air (brined) 9 . 8 7 6 . 3
In C 02 after evacuation (unbrined) 0.1 0.8 In C 02 after evacuation (brined) 1.2 1.5 In C 02 after flushing (unbrined) 1.4 2 . 8 In C 02 after flushing (brined) 3 . 6 6.1 In N2 after evacuation (unbrined) 0.3 0 . 8
In N2 after evacuation (brined) 1.1 1.5
a Tarr ( 1 9 4 8 ) .
0 In ml. 0 . 0 0 2 Ν thiosulfate per g.
In some species, as t h e scombroids with a high content o f myoglobin in the muscle ( d a r k m u s c l e ) , there m a y b e an appreciable though limited source o f oxygen available in addition to supplies from t h e air.
M o s t studies o f oxidative changes in frozen fish have b e e n b a s e d on estimation of the formed hydroperoxides (peroxide value, P V ) , the primary reaction product in fatty acid autoxidation. Although this test m a y indicate "potential" rancidity, it is n o t fully correlated with "or
ganoleptic" rancidity. B o t a l l a et al. ( 1 9 5 5 ) reported a rather rapid in
crease in P V in perch, mackerel, and cod. T h e fate o f t h e peroxides is not fully understood, and t h e secondary reactions p r o b a b l y take quite different paths according to the substrate and conditions. Aldehydes a r e known to b e formed, a n d t h e Kreis test is b a s e d on one o f t h e secondary reactions resulting in epidrin aldehyde. This test also is limited in value and ill-defined. N e w methods that estimate the formation o f carbonyl compounds have b e e n introduced ( H e n i c k et al., 1 9 5 4 ; H o l m et al., 1 9 5 7 ) . The T B A (thiobarbituric a c i d ) m e t h o d should also b e m e n t i o n e d ( s e e further C h a p t e r 2 , this volume; also Andersson a n d Danielsen, 1 9 6 1 ) .
B o y d et al. ( 1 9 5 7 ) observed that a mixture of two or more anti
oxidants has greater protective action against development of rancidity.
T h e y have shown experimentally that peroxide development is markedly inhibited b y a mixture of ascorbic acid ( 0 . 0 4 % ) and ethyl gallate
( 0 . 0 2 % ) in frozen fatty fish stored at + 1 6 ° F .
E . MICROBIOLOGICAL ASPECTS
Investigation of the microbial content of a frozen product will usually show that it contains a certain n u m b e r of bacteria. T h e n u m b e r will vary and is partly dependent on the original load of the raw product prior to freezing and primarily dependent on the treatment and sanitary conditions. T h e n u m b e r of microorganisms in the finally frozen fish also depends on cooling and freezing conditions, temperature during holding, species and physiology of the organisms, and composition of the fish.
Careful washing before freezing is particularly important in the freez
ing of fish. T h e n u m b e r of bacteria may b e considerably reduced b y careful washing before freezing, while inadequate washing and un
satisfactory cleaning of machines and equipment may contaminate the fish and result in high counts. Since cooling prevents or reduces growth of bacteria, if the fish temperature is low enough m a n y bacteria will not grow at all. L o w temperature has a lethal effect on some bacteria and no effect on others. F r e e z i n g therefore does not sterilize the product.
Although most of the bacteria m a y b e destroyed there will always b e some w h i c h are not killed during freezing. As soon as the product is de
frosted, therefore, these microogranisms will p r o c e e d to grow. V e r y often the texture o f the tissue m a y b e better after having b e e n frozen, resulting in more rapid spoilage. M a n y investigations have b e e n carried out and several reviews are available on the effect of freezing on bacteria and the microbiology of frozen products in general ( B e r r y and Magoon, 1934;
W a l l a c e and T a n n e r , 1 9 3 3 ; Haines, 1 9 3 8 ; W e i s e r and Osterud, 1 9 4 5 ; Stille, 1 9 5 0 ; Shewan, 1 9 5 4 ; Borgstrom, 1955, 1 9 6 1 a ) . T h e r e are also books dealing with the bacteriology of frozen fish (Plank, 1 9 5 2 ; Tressler and E v e r s , 1 9 5 7 a ) .
As far as fish are concerned, S h e w a n ( 1 9 5 4 ) summarizes the results as follows, "Freezing causes an initial drop in the numbers of bacteria present, of the order of 6 0 - 9 0 % , and provided the temperature of storage is b e l o w the minimum for growth, there is an exponential drop followed b y a more gradual decline during storage. T h e heavier the initial load the greater the n u m b e r of survivors."
Microorganisms, such as the gram-positive types, can survive freezing while others cannot, and freezing does not seem to c h a n g e the properties of the microorganisms. V e g e t a t i v e cells are more resistant to freezing than
9. FISH AND SHELLFISH FREEZING 3 7 5 spores. It appears t h a t the process of freezing and thawing is more lethal than storage in the frozen state ( J 0 r g e n s e n , 1 9 6 2 ) .
T h e p r e s e n c e of fat, sugar, a n d colloidal materials in t h e frozen m e d i a is in some cases a b l e to protect the bacterial cells, resulting in higher mortality in distilled water than in sea water.
T h e p H also affects the n u m b e r of microorganisms w h i c h survive freezing. F u r t h e r m o r e , there is more rapid microbial killing b e t w e e n
— 1 ° a n d — 1 0 ° C . than at — 2 0 ° C . or lower. At — 1 9 5 ° C . t h e r e appears to b e no freezing death at all.
T h e general conclusion is, therefore, that the freezing proper destroys microorganisms to a certain extent, depending chiefly on the temperature, which in turn determines the freezing rate. F r e e z i n g storage further de
stroys b a c t e r i a due to denaturation of proteins and is more lethal at the temperature interval b e t w e e n 0 ° and — 1 0 ° C . than at lower temperatures.
S o m e microorganisms m a y therefore survive long storage at low tem
peratures.
According to Borgstrom ( 1 9 5 5 , 1 9 6 1 a ) , slow freezing appears to b e more effective in destroying b a c t e r i a than rapid freezing, b u t slow freez
ing is not r e c o m m e n d e d since the b a c t e r i a m a y have time to develop and grow before the temperature b e c o m e s inhibitory, resulting in un
desirable effects on the textural qualities of the fish.
Dussault ( 1 9 5 6 ) found that coliform b a c t e r i a in fish resisted quick freezing and storage, and that five typical cultures of these b a c t e r i a w e r e not eliminated b y rapid freezing and storage at l o w temperature; b u t a reduction can b e obtained if the initial load is not too heavy. Similar results have b e e n reported ( R a j and Liston, 1 9 6 1 ; R a j et al.9 1 9 6 1 ) .
Since t h e original microbial count of the fish prior to freezing and the sanitary conditions are the most important factors determining the b a c terial load o f the frozen fish, the following guidelines s e e m useful:
( 1 ) Maintain cleanliness and good hygiene throughout operations.
( 2 ) R a w fish should b e cooled down as fast as possible and kept near 0 ° C . until freezing.
( 3 ) Filleting, washing, and freezing of fish should b e done shortly after the catch, w h e n the n u m b e r of b a c t e r i a on the fish is still compara
tively low (Anonymous, 1 9 6 1 a ) . III. Technological Developments
A. FREEZING METHODS
A great n u m b e r of methods h a v e b e e n developed. A partial list of quick freezing patents published b y Tressler and E v e r s ( 1 9 5 7 b ) com
prises 2 9 7 patents for the period 1 8 4 2 - 1 9 5 5 in the U n i t e d States alone.
Only a small n u m b e r of methods, however, have b e e n put into practical use in fish freezing, a n d three m a i n principles c a n b e distinguished.
T h e s e are freezing ( 1 ) b y cold air ( i n t u n n e l s ) , ( 2 ) b y brine, and ( 3 ) b y indirect contact.
W h i c h method is preferable in e a c h particular case depends on several factors and to some extent on shape and size of the product. F o r freezing of large fish or freezing in bulk, tunnel or brine freezing is most suitable. W h e n packing in retail sizes, indirect contact freezing is pref
erable. T h e aim is to get the most effective h e a t transfer from t h e fish, and to choose the m e t h o d w h i c h is superior as far as fish quality is con
cerned. W h e n there is a large supply of raw fish in a short season, the demands of high capacity have to b e taken into account.
According to M e e k a n d G r e e n e ( 1 9 4 7 ) , 6 0 % of fish was frozen in the United States b y air blast, 3 0 % b y plate contact, 5 % b y liquid immersion, and less than 5 % b y fog, spray, moving plate, and vacuum. I n Scandina
via, particularly Norway, however, brine freezing is still relatively m o r e important than in other countries, especially for freezing herring and mackerel, etc.
1. Freezing by Cold Air
In air freezing, the fish is frozen b y chilled air circulated b y fans from coils over the fish and then b a c k to t h e coils. A great n u m b e r of different arrangements and constructions of air-blast freezers have b e e n evolved (Helgerud, 1 9 5 4 b ; Anonymous, 1957a; B u t l e r et al, 1 9 5 6 ; Dassow et al, 1956; Kasansky and Khatoutsev, 1 9 4 9 ; Slavin, 1 9 5 8 ; Lorentzen, 1 9 5 8 ) . In some cases the fish are p l a c e d on shelves in the freezing room or on wagons. Large-sized fish are often h u n g from racks on dollies. T h e freez
ing room is often arranged as a tunnel, w h i c h is the most common type used in air freezing of fish. T h e air-blast freezers m a y also b e arranged for more or less continuous freezing b y conveying the fish on belts or dollies through the freezing tunnel. Another arrangement for b a t c h freez
ing is b y employing cabinets. I n this case the product m a y b e p l a c e d on trays.
I n "sharp freezers" the freezing is carried out partly b y cold air cir
culating over the fish and partly b y indirect contact. T h e product is placed directly on the coils or on metal plates chilled inside b y coils. T h e heat transfer is due partly to the contact b e t w e e n the fish and t h e cooling surface, and partly to the air blast w h i c h improves the rate of freezing considerably.
T h e advantage of air freezing over other methods is mainly its flexi
bility. L i t t l e work is required in handling the product, and it is possible to freeze b o t h large and small fish in boxes. O n the other hand, the air
9. FISH AND SHELLFISH FREEZING 377 circulation results in desiccation of the fish surface and formation of ice on the cooling surface, in turn resulting in impaired h e a t transfer, re
tarded freezing, and loss of time b e c a u s e of necessary defrosting of the coils. T o reduce dehydration to a minimum, freezers have therefore b e e n constructed ( F i n n e g a n , 1938a,b; Knowles, 1 9 4 5 ; and van G r e e n e , 1 9 4 4 ) where t h e temperature difference b e t w e e n the coils and the air circulating over the fish is reduced, and the humidity of the air is kept as high as possible. Air freezers are not, however, preferable for freezing of con
sumer packs. T o obtain regular form in the packs it is better to freeze t h e m b e t w e e n plates. Air freezers also require more electrical energy than plate freezers.
2. Brine Freezing
I n freezing with brine, the fish is frozen b y contact with a refrigerated liquid suitable for contact with the fish. Various types of brine are used, but those with sodium chloride are most important and most economical for fish ( H e l g e r u d , 1 9 5 4 b ) . D u r i n g freezing, the temperature of the brine is lowered to about — 1 7 ° C . , and the fish is frozen either b y spray freez
ing or b y immersion in the brine. Details of the m e t h o d vary.
F r e e z i n g b y sprinkling is used for freezing herring and mackerel in Scandinavia, particularly Norway. T h e boxes ( 5 0 k g . ) of fish are placed in stacks. O n top there is a b o x with a perforated base in order to dis
tribute the brine evenly. T h e brine passes from one b o x to the other and d o w n into a tank w h e r e it again passes over the cooling coils and is then p u m p e d up and sprinkled over the boxes of fish until the fish is frozen, i.e., 1-2 hours. T h e arrangement of the freezer m a y vary. T h e boxes are in some cases p l a c e d in a freezing tank or put into the spray on dollies.
Boxes m a y also b e m o v e d forward on a conveyor, w h e r e b y m o r e contin
uous freezing is obtained as the boxes are pushed forward e a c h time a n e w stack enters the freezer.
I n immersion freezing the fish in boxes are pushed forward in a duct under the surface of the brine. T h e r e is less brine foaming than in spray freezing and the amount of labor required is less; a more uniform freez
ing is also obtained. This m e t h o d m a y b e used for b o t h large and small fish.
O n board the tuna clippers the fish are put into large tanks and cooled b y refrigerated sea water to — 1 ° C . W h e n the fish are cooled the sea water is concentrated b y adding salt; the temperature of t h e brine is thereby lowered further, finally producing freezing of the fish. T h e brine is then emptied and the fish held frozen until delivered. B r i n e freezing has also b e e n used at sea on b o a r d the U . S . experimental trawler Delaware. In other methods the brine is sprayed in a fine mist over the
fish, thereby inducing rapid freezing ( T a y l o r , 1921; Zarotschenzeff, 1 9 3 0 ) . Brine penetration. O n e disadvantage of the direct brine freezing is the uptake of salt b y the fish, resulting in reduced keeping properties a n d in a few cases in off-flavors.
Salt penetration in brine-frozen fish has b e e n thoroughly studied at the B o s t o n L a b o r a t o r y of the U . S . F i s h and Wildlife Service. T h e i r find
ings m a y b e summarized as follows. Salt penetration is serious only when the fish are held in the brine ( 2 3 % N a C l ) longer than 4 hours and the brine temperature is held b e l o w 1 0 ° F . F i s h left in the chilled brine after freezing continue to absorb salt. B r i n e held at its freezing point, so that slush ice freezes out, results in some penetration of salt into the fish.
Coating of the fish with a marine plant gel is not r e c o m m e n d e d because, if used on a large scale, it dilutes and contaminates the brine ( B u t l e r , 1 9 5 5 ) .
3. Contact Freezing
T h e most common m e t h o d for contact freezing is with multiplate freezers. T h e fish are p l a c e d b e t w e e n metal plates and are not in contact with the freezing m e d i a or the brine ( T r e s s l e r and E v e r s , 1957b; Hel- gerud, 1 9 5 4 b ; Knowles, 1 9 5 4 ; Plank, 1 9 4 9 ; Slade, 1954a,b; Yule and E d d i e , 1 9 5 3 ) .
T h e freezing m a y b e conducted in molds, where the contact b e t w e e n fish and plates is indirect. T h e fish are p l a c e d in metal containers. This procedure m a y also b e used in spray freezing.
T h e freezer plates have internal channels for circulation of chilled brine or refrigeration medium. This keeps the plates at a low tempera
ture, so that the fish freeze w h e n p l a c e d b e t w e e n the plates. T h e r e are b o t h continuous and discontinuous types, the latter being most common for fish. I t is a very convenient m e t h o d for freezing b o t h retail packages and bulk or large-sized packages. T h e fish fillet is p a c k e d in frames and pressure is kept on the fish during freezing to obtain a more regular sur
face on the packs and blocks. T h e pressure is also important for good contact b e t w e e n fish and plates, giving far better heat transfer.
Plate freezers with horizontal plates ( t h e Birdseye principle) are generally used, but they m a y also b e arranged in a vertical manner ( H e l g e r u d , 1 9 5 4 b ; Plank, 1 9 4 9 ) . T h i s freezer was originally developed for freezing whale meat, and a modified type was later installed on board for bulk freezing of whole fish. Another vertical freezer was developed b y the Torry R e s e a r c h Station, Scotland ( Y u l e and E d d i e , 1 9 5 3 ) ; it was adopted for freezing whole fish on board and is in use on British trawlers.
( S e e also E d d i e , 1 9 6 2 ; Merritt and Templeton, 1 9 6 3 . F o r a review of types of freezers see Anonymous, 1958c, 1962, 1 9 6 3 . )
