CHAPTER 6
Scandinavian Sardines
MARTIN D. S T E N S T R Ö M1
Swedish Institute of Food Preservation Research, Göteborg, Sweden
I. Introduction 2 65
A. Brief Historical Notes 2 65
B. Biological Classification 2 66
C. Fishing 2 70
D. Supply Fluctuation 2 72
II. Technology 2 72
A. Fishing Methods 2 72
B. Processing Methods 2 7^
C. Cover Oils and Sauces 2 83
D. Types of Can; Can Sizes 2 83
E. The Processing Line 2 8^
References 2 88
I. Introduction
A. B R I E F H I S T O R I C A L N O T E S
Clupea pilchardus W a l b a u m constitutes the chief basis of sardine manufacturing in Portugal, Spain, M o r o c c o , and other Mediterranean countries. R e l a t e d clupeoids of similar size lend themselves well to a cor
responding processing, in some cases with slight modifications in the man
ufacturing procedure. B u t the final product should conform to certain specifications as regards appearance, taste, tenderness, etc.
In the seas surrounding the Scandinavian countries there are two small c o m m o n clupeoids, the sprat, Clupea sprattus, and the young her
ring, Clupea harengus. B o t h have b e e n utilized in the manufacture o f sardine packs.
According to S t r 0 m ( 1 9 4 9 ) , t h e packing of Scandinavian sardines started in Norway in 1879. B o t h sprats and herrings were earlier salted or used as fertilizer. T h e output of "sardines" grew rapidly when that produce b e c a m e a c c e p t e d on the export market. A great n u m b e r of fac
tories started operation, first in Norway and later in D e n m a r k and Sweden. Scandinavian sardines are n o w established over the entire world, the largest markets b e i n g the United States, the U n i t e d Kingdom, and Germany, but also South Africa, Australia, etc.
1 Present address: The State. College of Technology, Halmstad, Sweden.
265
B . B I O L O G I C A L C L A S S I F I C A T I O N
1. The Sprat
Sprat (Clupea sprattus L . ) is called "skarpsüT or "vassbuk" in Sweden and "brisling" in D e n m a r k and Norway. Sprat is a rather small herring
like fish. T h e dorsal part is bluish green. T h e sides and ventral part are silvery white. I t measures 17.5 cm. in length. T h e head is short and the lower j a w never exceeds 5 3 % of the h e a d length. T h e sprat is distin
guished from the herring b y its characteristic sharp ventral margin and carries conspicuous spines on the ventral plates. T h e form of the body is identical to that of the herring and shows the same variations. T h e ventral profile is still more curved in comparison with the dorsal. T h e pectoral fins resemble those of the herring, but the ventral fins are in general shorter. T h e i r position constitutes the most important characteristic of the sprat, as the ventral fins are inserted vertically b e l o w the beginning of the dorsal fins or further to the front of the body. T h e geographical range of the sprat is from Lofoten in northern Norway to the Mediter
ranean on the east side of the Atlantic. I t has never b e e n found on the North American coast. I t is common from L o f o t e n to Brittany in F r a n c e but occurs also in the Baltic sea up to latitude 6 4 ° N ( S m i t t , 1 8 9 5 ) .
T h e sprat spawn during April-July in the North S e a and t h e Skagerak and during April-August in the B a l t i c Sea, b u t mainly during M a y - J u n e . T h e y spawn in the open sea, not far from the coast at a depth of 2 0 - 1 0 0 meters.
After hatching the sprat makes its way inside the island b e l t and into the fjords w h e r e it stays during the first year. S o m e of the sprat remain in the fjords. T h e r e are also sprat which spawn and live in the fjords all their life. S o m e sprat on the other hand remain at sea. T h e larvae are about 4 mm. in length; they grow very quickly during the first summer. B y August they m a y reach a size of 4 - 6 cm. and in D e c e m b e r 7 - 8 cm. After a year the length is 9 - 1 1 cm. T h e sprat in the open sea grow faster than the sprat in the fjords. After three summers the size m a y b e still only 1 1 - 1 4 cm. W h e n 5 - 6 years old they m a y have grown insignificantly.
2. The Herring
T h e herring (Clupea harengus L . ) is the most important fish in the Scandinavian countries. I n S w e d e n it is called "sill" and in D e n m a r k and Norway "sild." Young small-sized herring are used as raw material for sardine packs, especially in Norway and D e n m a r k . T h e herring occur over the whole northern Atlantic, on the United States coast from Green
land to C a p e C o d and on the E u r o p e a n coast from the B a y of Biscay to the Arctic coast of Norway and Russia. T h e y are also found in the entire B a l t i c Sea.
6. SCANDINAVIAN SARDINES 267 T h e difference b e t w e e n small herring and sprat is very minute. T h e i r coloration is about the same. T h e h e a d length is 1 9 - 2 2 % o f the body length. I t is three-sided, pyramid-shaped, b u t with a narrow upper surface and the snout is constricted in front of the eyes. T h e most characteristic features are the prominent tip of the lower j a w and the fat cover in the m e m b r a n e concealing a great portion o f t h e eyes and extending over both the sides of the snout and the shoulder girdle. T h e eyes are round, or slightly longer than high, and rather large. T h e insertion of the ventral fin lies always some distance b e h i n d the perpendicular from the begin
ning of t h e dorsal fin. T h e scales of t h e belly are not prominently sharp (Smitt, 1895; Anderson, 1 9 4 2 ) .
T h e herring has m a n y varieties. T h e s e are identified b y minor distinc
tive features such as the n u m b e r of spines, different spawning seasons and sizes. T h e s e properties are, however, not significantly hereditary. S o m e varieties show variations in n u m b e r of spines from one year to the next.
This can b e explained b y the hydrographic conditions of the spawning p l a c e (Anderson, 1 9 4 2 ) .
S o m e herring spawn in the spring and others in the autumn. At spawning time the herring migrate towards the coast or to the grounds in the open sea. T h e spawning occurs at different depths, usually 1 0 - 4 0 meters. N e w l y h a t c h e d herring are 6 - 1 0 m m . After 2 - 3 weeks they start feeding on plant plankton first, later on animal plankton. At 2 - 3 months of age the larvae grow scales and pigments and the appearance changes to that of the adult herring. T h e y then gather in schools and migrate over large distances, preferably along the coasts and into the fjords w h e r e they find plenty of food. T h e schools of young herring often mix with schools of sprat and the catches are usually a mixture of the two. W h e n the herring are about a year of age they leave the coasts and migrate to deeper water. L a r g e numbers of these young herring are caught during the winter season and used for the production of herring m e a l and oil;
a great part is used in Norway and D e n m a r k for the packing of "sild sardines."
3. Feeding; Fat Contents; Seasonal Variations Herring and sprat feed on plankton and plankton-sized larvae of worms and fish. T h e y also eat fish eggs and even their own larvae and small fish. T h e m a i n feed is crayfish species such as copepods and schizo- pods. T h e y sometimes eat small mollusks, w h i c h m a y cause trouble to the canner. T h e s e mollusks are difficult to digest, whereupon t h e intestines develop large amounts of enzyme w h i c h m a y attack the flesh of the fish after the c a t c h and render them unfit for canning. I f the intestines con
tain feed the belly is softened and so w e a k e n e d that it m a y burst either
before or during processing. B o t h appearance and taste are affected, making the product unfit for food.
T h e supply of nourishment to herring and sprat is abundant during summer and fall when the sea water temperature is high, and the fish grow quickly. T h e fat concentrates first in the muscles and then in the belly along the intestines. T h e fat p e r c e n t a g e of herring and sprat changes periodically. As in all fish, fat content is lowest at spawning. Thus the sprat and spring spawning herring are leanest in spring. T h e fat content increases until O c t o b e r and November, when it is at its peak and m a y reach 1 5 - 2 0 % or even higher. T h e n it declines until spawning time, when it is 1 % or less. Before the first spawning the fish usually deposit large amounts of fat. T h e s e fish are preferred as raw material for sardines.
4. Quality Factors
T h e quality of the sardine depends, of course, fundamentally on the quality of the fish and changes with the season. T h e cardinal quality factor is the fat content. Important also are the taste and tenderness of the fish meat. T h e feed must b e removed from the stomach in order to obtain prime quality. S u m m e r caught year-old sprat have the best taste and texture and are p a c k e d as fancy quality. F a t content of fish for can
ning purposes is specified b y regulation; it must b e 7 % or more. Packing of sprat with 6 % fat content is allowed for second-class sardines, but cans must b e marked accordingly to distinguish t h e m from the fancy quality. F a t p e r c e n t a g e of the fish is determined before t h e landing and, if too low, the catch must b e released again. A 5-liter sample must b e taken at random from which 2 0 0 fish are used for the fat determination.
T h e Norwegian data on fat p e r c e n t a g e of summer caught brisling are published e a c h year. I n 1957 the variation was 2 . 5 - 1 5 % (Anonymous, 1 9 5 8 ) . E a c h individual sample, however, m a y have an appreciable indi
vidual variation. J a k o b s e n et al. ( 1 9 4 3 ) analyzed 100 individual fish and found the fat percentage to vary in the range 0 . 7 - 1 9 . 6 % , calculated on the basis of the whole fish. T h e m e a n value was 9.54 ± 1 . 1 with a prob
ability level of 9 8 % . F r o m the same catch, 2 1 samples of 1 0 0 fish each gave results within the range of 8 . 3 - 9 . 8 % . T h e size of the sample was later fixed at 2 0 0 fish where a reliability of 9 8 . 2 % m a y b e expected.
L u n d b o r g ( 1 9 4 4 ) published figures on the fat p e r c e n t a g e of winter caught ( J a n u a r y - F e b r u a r y ) sprats and small herring. Sprats of 10.0-15.1 cm. had a fat p e r c e n t a g e of 4 . 6 - 1 1 . 6 ; herring of 1 1 . 4 - 1 5 . 9 cm. a fat per
centage of 3.5-6.5.
In Polish studies on sprat in the southern Baltic, the p e a k in dry matter was found to occur in O c t o b e r and the lowest values w e r e found
6. SCANDINAVIAN SARDINES 2 6 9 in July. Another important relationship is the inverse ratio b e t w e e n fat and protein ( M i k i c i n s k a , 1 9 5 5 ) , shown in the tabulation.
Fat Protein
Dry matter (% of dry (% of dry
( % ) matter) matter)
20.3-23.2 13.5 75.9
October 31.4-33.2 46.3 47.0
This fluctuation in composition obviously carries nutritional implica
tions. M e y e r ( 1 9 4 3 ) concluded as a result of a very extensive study of the quality factors in herring and sprat of the B a l t i c that the sprat showed a satisfactory fat content from N o v e m b e r to M a r c h ( 1 8 . 7 - 7 . 3 % ) . B y F e b ruary there is a decline which reaches its lowest in July-August. T h e r e are certain persistent differences in the ideal periods in t h e eastern, central, and western parts of the Baltic, the east constantly showing less favorable conditions, smaller-sized fish, and lower fat content ( D i x o n , 1937; M e y e r , 1 9 4 3 ) . T h e most suitable period for catching from the point of view of composition far from coincides with t h e schooling pe
riods which are best for capture.
T a s t e and texture of the fish m e a t c h a n g e with the seasons. T h e summer caught brisling have a delicious taste and are very tender. T h e winter caught brisling and herring h a v e a different taste, b e i n g markedly
"fishy." T h e texture m a y also b e m o r e fibrous since the fish is somewhat older.
T h e taste changes with season are not quite clear. T h e y are almost certainly caused b y variation in content of t h e nonprotein nitrogenous compounds, b u t other factors m a y b e involved. I n the case of brisling and herring, T M A O (trimethylamine o x i d e ) or its decomposition products seem to b e largely responsible for the taste changes. This b a s e has b e e n studied very thoroughly. R o n o l d and J a k o b s e n ( 1 9 4 7 ) found a seasonal variation in T M A O and T M A contents of herring and brisling, with a peak in F e b r u a r y . T M A O is broken down at high temperatures to dimeth- ylamine and formaldehyde ( D u n s t a n and Goulding, 1 8 9 9 ) . As a result, canned brisling ( L u n d e et al, 1 9 3 4 ) and other marine products ( R e a y , 1 9 3 6 ) have higher formaldehyde contents than t h e raw fish.
"Fishiness" flavor appears in various foodstuffs, e.g., milk, butter, cured stored fish, and poultry and hogs that w e r e fed fish m e a l containing excess oil. T h e development of "fishiness" seems to depend on reactions b e t w e e n T M A and fat ( S h e w a n , 1 9 5 1 ) . W h e t h e r similar reactions could take p l a c e in canned Scandinavian sardines is an open question.
Patents on treatment of fish to improve the flavor have b e e n issued.
Mathiesen ( 1 9 4 5 and 1 9 4 6 ) patented the treatment of fish with reducing agents such as N a H2P 03, N a2S203, ( N H4)2S 03, K H S 03, ( C H3)4N H S 03
or organic reducing agents such as ascorbic acid and also gaseous S 02. T h e treatment should b e carried out at p H 6.0. T h e patent claims the
"steam odor" a n d "steam taste" will b e removed and the treatment will h a v e t h e same seasoning effect as about 6 months of storage. I n other words, t h e c h e m i c a l processes w h i c h lead to normal final seasoning are accelerated. Sjöström ( 1 9 4 8 ) patented a treatment of fish with alkali. H e claims t h e palatability of herring is improved b y spraying the fish with a dilute solution of alkali after precooking and before canning. T h e s e treat
ments m a y increase the p H b y 0 . 1 - 0 . 5 unit.
C. F I S H I N G
I n order to establish good quality standards and also to protect the stock from overfishing, b o t h catching and p a c k i n g is regulated b y law in Norway and b y special agreements b e t w e e n canners, fishermen, and state authorities in S w e d e n and D e n m a r k . T h e fishing season for sprat in Norway starts in M a y and ends the last day o f D e c e m b e r . I n S w e d e n the season terminates w h e n the fat content is less than 4 % , w h i c h occurs at the beginning of M a r c h . T h e Danish season is from O c t o b e r till F e b r u a r y .
I n Norway the p a c k i n g of sild sardine, w h i c h is b a s e d on small-sized herring, is permitted until J a n u a r y 3 1 . As was mentioned, the schools of herring and sprat are frequently mixed. T h e Norwegian law therefore de
fines the highest p e r c e n t a g e of herring permitted in sprat catches and the highest p e r c e n t a g e of sprat in herring catches. I f the p e r c e n t a g e exceeds these figures, the catches must b e released. D e n m a r k and Sweden do not have such regulations. Packing stops w h e n the fish quality is not good for canning, the fishing season ends at about the same time. Moreover, the catches n e e d not b e released i f unsuitable for canning; they m a y b e used for fish m e a l production and are delivered to reduction plants.
T h e fishing season for sprats is divided into two distinct periods.
During the first period, spring and summer, the fish contain feed in the stomach. T o get rid of this feed the fish are kept in the purse seines for 3 days. T h e feed is then digested and the intestines are evacuated. D u r i n g fall and winter t h e fish eat less a n d less and there is little or no feed in the stomach. T h e y can then b e landed and brought to the cannery for packing as sardines. T h e permitted periods do not c h a n g e simultaneously in the whole district. At t h e switch, the sprat goes into the manufacture of other products, such as Scandinavian anchovies ( s e e V o l u m e I I I , C h a p t e r 6 ) .
DANISH CATCH Catch of brisling Catch of sild Pack of sardines
OF BRISLING 1952 2,181 3,964 3,371
AND SMALL 1953 2,033 2,126 2,336
Ί HERRING 1954 2,086 3,688 3,467
ABLE I FOR CANNING PURPOSES 1955 1956 2,108 1,961 3,443 4,682 3,347 4,636
AND THE 1957 1,533 3,208 3,651
PACK OF 1958 459 5,646 4,698
SARDINES* 1959 345 6,303 4,355
1960 565 6,144 3,730 a Quantity in 1000 kg. TABLE II NORWEGIAN PRODUCTION OF CANNED BRISLING AND SILD SARDINES0 Sardine type 1953 1954 1955 1956 1957 1958 1959 1960 Brisling 640 420 247 198 534 309 672 489 Sild 490 567 927 635 828 750 618 685 α Compiled by the author. Production in 1000 cases. TABLE III SWEDISH CATCH OF BRISLING AND PRODUCTION OF SARDINES* 195(0 1951 1952 1953 1954 1955 195(5 1957 1958 1959 I960 Catch of brisling 10,115 9,153 11,186 9,150 10,694 8,332 7,115 3,180 2,318 5,935 4,482 Pack of sardines 3,362 3,500 3,391 3,207 3,853 3,328 3,161 2,265 2,189 1,852 2,041 a Quantity in 1000 kg.
CO Ö
s
CO> I to -a ι—»
D . S U P P L Y F L U C T U A T I O N
T h e Danish official statistics give no information on the quantity of fish used for packing o f sardines. T h e following figures are compiled from private sources and from statistical reports (Anonymous, 1 9 5 9 , 1 9 6 1 , 1 9 6 2 ) . T h e figures for catches and total p a c k of sardines give a picture of the supply. In D e n m a r k the catch of brisling and sild for processing and the p a c k of sardines are reported ( J e n s e n , 1 9 5 8 ; Bramsnaes, 1 9 6 4 ) in T a b l e I.
T h e Association of Norwegian Canners reports annually the produc
tion of brisling and sild sardines in Norway. T h e statistics in T a b l e I I are given in numbers of cases containing 1 0 0 cans. T h e shipping weight of a case of D i n g l e y tin plate cans is ( J a k o b s e n , 1 9 5 0 ) 1 9 . 5 - 2 0 . 0 kg.
and that of aluminum cans of equal size 15.36 kg.
Swedish official statistics listed figures up to 1960; all this informa
tion is contained in T a b l e I I I . No sardines are p a c k e d from herring in Sweden.
T h e Swedish figures in T a b l e I I I reveal a decreasing supply of raw material to the sardine packing industry since 1 9 5 5 . T h e Norwegian figures, however, indicate that the lowest point has b e e n passed. T h e Swedish catch in 1957 was the lowest since 1 9 3 6 , and was termed b y both fishermen and canners a "catastrophe." T h e causes of this long- term fluctuation are not clear. I t m a y b e due to overfishing, or to detri
mental hydrographic conditions during the spawning period. Since the main part of the sprat catch is 1-3 years old and never exceeds 6 years, such a fishery is most vulnerable to a sequence of years with poor spawning.
II. Technology
A. F I S H I N G M E T H O D S
1. Development
Since ancient times the sprat and herring fisheries have b e e n an im
portant livelihood to the coastal inhabitants of the Scandinavian coun
tries. T h e only way of preserving fish in the early days was b y salting, smoking, or possibly drying. Parts of the catches often very large were spoiled and h a d to b e used as fertilizer. T h e development of transporta
tion and the emergence of a preserving and canning industry during this century changed the profitability of fishing, and t h e efficiency of fishing operations improved rapidly. Prior to 1900, the only effective fishing gear was the seine, fixed on shore or hauled. Such nets are still used to a great extent. L a r g e r fishing vessels gradually c a m e into use and the first
6. SCANDINAVIAN SARDINES 2 7 3 motor vessels appeared. M u c h larger nets could then b e used, such as purse seines and otter trawls.
2. Haul Seine Fishing
This is the simplest form of seine. I t is used extensively during the summer and fall season in Norway and Sweden. T h e sprat seine is some
what smaller than the herring seine. T h e headline is 1 8 0 - 2 0 0 meters and 2 0 - 2 6 meters in depth. T h e headline is b u o y e d b y cork floats and the groundline weighted to keep the net tight to the bottom. O n e end is re
tained on land, the other is drawn b y a vessel through the water en
circling t h e school of fish and brought b a c k to land, very often the other side of a bay. I f the water is deep and the headline submerges b e l o w the surface, a covering net is laid out over the seine thus preventing escape of fish. T h e seine is then hauled to shore b y means of lines attached to the net. I f the fish contain feed they must b e held alive in the water for 3 days. Other nets—small purse seines—are then used. F i n a l l y the haul seine ends are tightened and a "purse" is accomplished.
I f the c a t c h is large, two or m o r e small purse nets are p l a c e d inside the haul seine and the fish are held in these nets. I f the catch is small, the purse is p l a c e d outside the haul seine and the fish released into it.
T h e headline of the net must b e adequately b u o y e d in order to keep the top of the n e t well above the water surface. After the fish have b e e n held for the prescribed time and the fat content has b e e n established to fulfill specifications, the fish are landed b y means of brails or similar devices.
In the m e a n t i m e the fishing net can b e used for n e w catches.
Haul seine fishing requires six to seven fishermen, one motor vessel, and one or two boats.
3. Purse Seine Fishing
T h e purse seine is a very large, almost rectangular net 3 0 0 - 3 5 0 meters long and 6 0 - 7 0 meters deep. T h e center of the seine ( t h e bunt or landing b a g ) is about 15 meters deeper. I t is designed for fishing in the open sea and is effective in the capture of schooling fish. T h e seine can b e laid out b y the main vessel but is usually laid out b y two seine boats, brought to the fishing ground on b o a r d the main vessel. T h e seine boats are lowered with half the seine in e a c h boat. W h e n the school is found, the bunt is laid out first. T h e n the boats row in opposite semicircular courses around the school, letting out the rest of the net. On the ground rope are rings, through w h i c h a line is drawn. T h e purse is accomplished b y hauling in the slack line. T h e whole purse is then towed to shore and t h e fish are transferred to smaller purse seines. T h e s e are anchored, as was indicated above to keep the fish alive to get rid of feed. I f not,
the fish are brought on board b y hauling the purse seine and sorted ac
cording to size and species. T o handle the purse seine a crew of 1 2 - 2 0 fishermen is required and a motor vessel and two seine boats, depending on the size of the seine. T h e fishing procedure varies in different countries and districts.
4. Trawl Fishing
Most of the sprat and herring during late fall and winter are caught by trawling. W i t h the increase in power of fishing vessels, the otter trawls were developed; they display many variations. I n brief, the otter trawl is a pocket net towed from one or two motor vessels b y ropes ("bridles"), one affixed to e a c h trawl head. T h e s e are attached to another rope ( t h e
"warp"), which runs out from the ship at about 3 times the depth of the fishing ground. T h e wings of the otter trawl net are k e p t apart b y means of two large boards, one attached to each wing. T h e bridles hold them securely to the towing ropes. T h e bridles are adjusted so that the force of water during towing keeps the boards apart. W h e n catching herring, a third board is often attached to t h e headline of the otter trawl, holding this high.
T h e otter trawl is dragged along the bottom or just above it b y means of special devices. An important innovation in 1 9 4 9 was the floating trawl, w h i c h is a square-shaped net, widest at the mouth with the sides tapering into a bag. T h e upper edge is b u o y e d b y floats and the lower edge is kept low b y weights. T h e depth can b e controlled b y the vessel speed.
W h e n the trawl is used at a season when the sprat and herring con
tain no feed, the c a t c h c a n b e brought on board at once. It is then di
rectly sorted for size and species.
B . P R O C E S S I N G M E T H O D S
1. Purpose of the Processing
T h e processing of sardines from clupeoid fish involves the addition of salt, cooking, trimming, filling into cans, adding a cover liquid with an appetizing oil or sauce, sealing and sterilizing the cans. T h e processing should impart the proper color, flavor, texture, and appearance to the fish, thus developing the inherent delicious properties of t h e raw ma
terial.
2. Brining
Proper uptake of salt is very important for the final flavor of the sardines. Scandinavian sardines are salted b y soaking in brine. T h e con
dition and the fat p e r c e n t a g e of the fish determine the degree of salt penetration.
6. SCANDINAVIAN SARDINES 2 7 5 Hansen and J a k o b s e n ( 1 9 4 5 ) have studied this problem. T h e y used fish with different fat percentages and of different sizes. T h e brine con
tained 2 5 , 1 8 , and 1 4 % sodium chloride. T h e y found salt penetration to b e very rapid in strong brine, t h e main part of t h e salt b e i n g absorbed in less than 2 minutes! Salt uptake was slower in fat fish than in lean. This rapid process cannot b e used, however, owing to individual differences of the fish. T h e y r e c o m m e n d a constant process time of 5 minutes in a brine w h i c h imparts a salt p e r c e n t a g e of about 1.5. F a t brisling require 1 8 % brine, m e d i u m fat fish 1 4 % , and lean fish 1 0 % .
This experiment and the recommendations cited refer to b u l k brining.
I f a continuous process is employed the time for treatment can b e reduced and the brine can b e m a d e more concentrated.
3. Precooking
It is difficult to define precooking satisfactorily. I t is m e a n t here in the widest sense. T h e fish must pass a heating process which releases and removes t h e innate w a t e r of the fish meat, prior to the heat treatment in retorts, so that no liquid is exuded in the can during the actual appertiza- tion.
Precooking can b e accomplished in different ways. T h e original F r e n c h m e t h o d was to h e a t the fish in edible oil. Precooking in steam has re
cently r e p l a c e d the old m e t h o d to a large degree in F r a n c e , Portugal, and M o r o c c o . Precooking in steam is also used in t h e Scandinavian countries b u t t h e dominant process is a kind of preliminary smoking, t h e details of which are similar to processes c o m m o n in several other countries.
Precooking of sardines was the subject of a technical symposium held in Paris at the 1 9 5 7 annual meeting of the International Permanent C o m mittee on C a n n e d F o o d s . M e e s e m a e c k e r and Sohier ( 1 9 5 7 ) reported an investigation of the w h o l e range of precooking procedures presently in use in F r a n c e , Portugal, M o r o c c o , and Spain, such as boiling in water or brine, frying in oil, heating in hot air and in steam. T h e results from precooking in hot air and steam are relevant to the smoking process. T h e studies w e r e carried out b o t h on grills and in the can. T h e heating on grills was conducted with standard types. T h e precooking in cans pro
c e e d e d as follows. T h e sardines w e r e b e h e a d e d , eviscerated, and brined;
they w e r e then p l a c e d in the cans. T h e cans w e r e arranged on inclined grills with the open end downward, to allow complete drainage of exuded liquids as t h e y formed. T h e grills w e r e introduced into the oven with hot air or expanded steam a n d subjected to the action of h e a t for t h e required time. O n taking the cans from the oven, oil was a d d e d immediately and the cans w e r e sealed and heat-processed in retorts. T h e weight loss was determined in reference to the original raw fish. T h e amount of liquid
exuded in the can was a measure of the precooking effect. F i s h of dif
ferent fat content were also investigated in this way.
W h e n cooking in hot air three temperature levels were chosen, 1 0 0 ° C , 1 3 0 ° C , and 1 6 0 ° C ; 6 0 minutes at 1 6 0 ° C . was not sufficient. Cooking in the can was a slower operation than heating on grills. T h e total weight loss of the fish increased with temperature and in the most severe cases rose to 3 2 . 4 % on the grills and 2 8 . 7 % when treated in the can. In the latter case the amount of exuded liquid was about 3 times larger.
Precooking in steam was conducted in a steam c h a m b e r at 1 0 0 ° C ; the experiments were carried out simultaneously on grills and in cans.
W e i g h t loss on heating was found to b e more rapid on grills than in cans and exudation of liquids from the fish in cans to occur with some difficulty.
W h e n precooking sardines of different fat content, the required processing time for the fatty fish is m u c h less than for the lean. Thus in order not to exceed 1.5 g. aqueous exudate per can, the cooking must b e continued for 4 0 min. if the sardine is very lean ( 2 % ) and reduced to 2 0 min. if it is very fat.
T h e loss during precooking in steam at 1 0 0 ° C . was calculated from analysis of the raw fresh fish and the cooked fish. T h e authors found that the fish lost 8 % of its fat, 8 % of its nitrogenous matter, and about 1 0 % of its mineral content. T h e y concluded that the fat in the muscles re
mained there and that the drip of fat results from fat in the abdominal cavity and from fusion of the subcutaneous fat.
Precooking in expanded steam is the most rapid procedure.
4. Smoking Process
a. G E N E R A L
T h e smoking process involves drying, cooking, and improvement of palatability. In most kiln types the combustion gases of burning oak wood are active. T h e y supply the required heat, convey the smoke compounds to the fish, and also carry away the evaporated steam from the kiln.
T h e sardines need only a very light smoke treatment in order to affect adequately the color of the product. This process has b e e n studied ex
tensively but several details remain to b e settled. T h e result of the smoking operation still depends to a large degree on the skill of the worker. Progress is generally m a d e through trial and error. Several re
search workers in the United Kingdom and C a n a d a have investigated the behavior of the smoke and its constituents.
Börresen ( 1 9 5 2 ) in Norway gave a comprehensive technological sur
vey of the smoking process, particularly as a water-removing process.
T h e evaporation in hot air, as well as in the combustion gases, can b e analyzed jointly b y calculating the basic characteristics of the gas. T h e
6. SCANDINAVIAN SARDINES 277 h e a t requirements, the vapor capacity, and the h e a t losses to the product, to the equipment, and through radiation are all essential factors.
T h e evaporation should b e so guided, that a large difference is main
tained b e t w e e n the vapor pressure of the product and t h e humidity of the smoke. W h e n the surface of the product is wet, it has a vapor pressure of saturated vapor equal to that of its temperature level. T h e temperature of the fish is then that of t h e w e t b u l b , corresponding to the temperature of the dry b u l b thermometer at the actual humidity state of the gases. W h e n the surface of the fish b e c o m e s dried, the vapor pressure depends m o r e on the hygroscopic properties of the product and on the ability of the vapor to pass from the w e t inside parts of the fish to the surface. T h e vapor pressure m a y decrease b e l o w the pressure of the sat
urated steam at the temperature of the product. T o maintain a safe dif
ferential vapor pressure, either the temperature of the air and thus also of the product must b e increased, or the velocity of the air must b e ac
celerated. I f the vapor pressures are equal, no evaporation will occur, and the water contained in the product m a y boil and the fish burst. As a guide to the drying process the Mollier i-x-diagram is very useful. It is apparent when using it that the temperature of the gases is the only h e a t supply and, therefore, must b e k e p t high. T h e velocity of t h e gas must b e very high at the start of the process, with only a small part of the heat actually being of service. T h e fastest drying would occur if h e a t were added continuously or at short intervals. This h e a t can b e supplied b y the injection of hot air, h e a t exchanging, infrared rays, etc.
According to Norwegian specifications, the sprat or herring should b e reduced in water content to a level of 6 4 % . F a t content of the fish is very important when the process is calculated and controlled. I t is known that while the p e r c e n t a g e of fat-free dry residue is rather constant, the fat content and the w a t e r content vary considerably. T h e i r sum, how
ever, is m o r e or less constant. F l o o d ( 1 9 5 8 ) published a series of determinations for the period 1 9 3 4 - 1 9 5 6 , and found the p e r c e n t a g e of fat-free dry residue in herring to b e 1 7 . 4 - 2 0 . 9 . L u n d b o r g ( 1 9 4 4 ) reports a fat-free dry residue in brisling of 1 7 . 3 - 1 9 % and fat content of 4 . 5 - 1 1 . 6 % . This relationship b e t w e e n fat and w a t e r c a n b e used to calculate the amount of water which must b e removed, at least as an approximate es
timation. A similar inverse relation b e t w e e n fat and w a t e r was reported in Polish studies on sprat in the southern B a l t i c ( M i k i c i n s k a , 1 9 5 5 ) .
Very fat raw fish m a y already have a w a t e r content o f 6 4 % or even less. Current specifications state that water m a y not exceed this figure in the smoked product. This is obviously less relevant in this particular case.
I f the fat-free dry residue is considered as the water-binding substance of the fish, it must carry m u c h more w a t e r w h e n the fat p e r c e n t a g e is
high than i f merely a specification of 6 4 % should b e met. Consequently, it is m o r e difficult to dry fat fish than lean. T h e y exude m o r e liquid in the can when heat-treated, i f the process is performed without special precautions. T h e application of h e a t must b e very careful during the initial period of drying.
T h e capacity of a continuous kiln depends on m a n y factors. T h e c h o i c e o f drying time is an accommodation b e t w e e n h e a t e c o n o m y and plant capacity. I n the ideal process the air discharged from the kiln should b e saturated b y steam at the lowest possible temperature. T o achieve this the air should, w h e n leaving the kiln, pass over a w e t charged prod
uct. I f the product at the discharge end were dry, this in itself would limit the humidity o f the air to a low level b e c a u s e of its hygroscopic status. E v e n in the best kilns with countercurrent air passage the relative humidity of the air seldom exceeds 7 5 - 8 0 % and in the case of parallel current passage 1 5 - 2 0 % .
Börresen ( 1 9 5 2 ) recommends the following steps to increase the dry
ing efficiency:
( 1 ) H i g h e r air temperature.
( 2 ) L o w e r relative humidity.
( 3 ) Adding heat under the process.
( 4 ) Increasing the air velocity.
( 5 ) Uniform distribution of the air current.
( 6 ) M e c h a n i c a l removal of free water before introducing fish into the kiln.
Kverneland ( 1 9 5 3 ) reports experiments on drying fish of classified sizes. H e found that small fish are dried faster than large fish, b u t if the speed of evaporation is related to the surface, the evaporation curves nearly coincide. T h e temperatures, air velocity, and humidity w e r e kept identical in the different experiments; the speed of evaporation was there
fore a function of the surface of the fish.
b . S M O K E M A K I N G
T h e fuel used in the smoking o f Scandinavian sardines is almost al
ways oak wood. A few installations h a v e b e e n constructed w h e r e other fuels w e r e employed for heating. T h e combustion gases o f oak wood will give sufficient heat and high quality of the smoke.
T h e h e a t of combustion of dry w o o d is about 4 5 0 0 kcal, per kg.
( 1 0 , 4 6 0 B T U p e r l b . ) , w h i c h corresponds to normal analytical data of 5 0 % C. 6 % H2, a n d 4 4 % 02. I f the wood contains Wf kg. water p e r kg.
wood, the h e a t of combustion will b e ( 4 5 0 0 — 5 1 0 0 Wf) kcal, per kg.
w o o d or ( 1 0 , 4 6 0 — 1 1 , 8 5 7 Wf) B T U per lb. wood. T h e w e i g h t o f the resulting gases is about 6 times the weight of the dry wood.
6. SCANDINAVIAN SARDINES
279
T h e dew point of the gases, w h i c h indicates the limit of the drying properties, depends on the water content of the wood. I f dry w o o d is burned the d e w point is about 5 0 ° C . ( 1 2 2 ° F . ) , and if the w o o d contains 3 0 % water it is about 5 8 ° C . ( 1 3 6 ° F . ) . I n order to improve the evapora
tion capacity and also to regulate the initial temperature, the combustion gases must b e mixed with considerable amounts of air.
O a k wood is a very expensive source of fuel and its use b y industry has caused a shortage. E x p e r i m e n t s w e r e recently carried out to replace oak wood for heating purposes b y other kinds o f fuel. Coniferous woods cannot b e used; they impart a resinous flavor to the fish, w h i c h is not desirable. B u r n i n g of cokes or mineral oils has b e e n tried, and the com
bustion gases have b e e n injected directly into the kiln or their h e a t was utilized via a heat exchanger. A separate generator c h a r g e d with oak wood gives a small additional amount of flavoring smoke. T h u s oak w o o d is saved, w h i c h makes the process m o r e economical.
T h e c h e m i c a l ingredients of oak wood smoke according to Cutting ( 1 9 5 1 ) are organic compounds such as formaldehyde, acetaldehyde, furfuraldehyde, acetone, diacetyl, methyl a n d ethyl alcohols, formic and acetic and higher fatty acids as well as phenols and tar. T h e proportion of the various substances depends to a large extent on t h e degree o f venti
lation. W i t h no air influx at all, the amounts o f acrid flavored phenols a n d aliphatic acids increase. As ventilation is increased, the amount o f form
aldehyde is r e d u c e d and there is m a r k e d increase in the qantity of resinous compounds.
T h e stream of smoke-laden air must b e controlled during the w h o l e process. T o rely only on t h e natural air suction of the c h i m n e y would result in t h e process b e i n g dependent on atmospheric conditions. F a n s to regulate the air velocity to suit t h e product must b e installed. F a n s of both propeller a n d centrifugal types are employed. D a m p e r s regulate air supply for the burning and also mixing of air a n d combustion gases.
c. E F F E C T ON T H E P R O D U C T
T h e partial drying o f sardines during manufacture has b e e n studied b y the R e s e a r c h L a b o r a t o r y of the Norwegian C a n n i n g Industry at Stavanger.
L u n d e and M a t h i e s e n ( 1 9 3 2 a , b ) analyzed sardines prior to and sub
sequent to the actual drying. T h e i r analyses are given in the tabulation.
Fish % Water % Fat % Protein
Brisling, raw 70.0 13.1 16.4
Brisling, smoked 61.1 15.9 21.4
Herring, raw 75.9 5.5 17.1
Herring, smoked 64.5 6.9 26.0
Neither the type of kiln used nor other details of the process were given.
As was mentioned, analyses may b e misleading b e c a u s e of the large in
dividual variations of the sample. W i t h no information on yield during smoking, these figures do not tell the whole story. I f the raw brisling were dried without loss of fat or protein, the fat content would b e around 1 7 . 0 % and that of protein around 2 1 . 2 % , which shows that some fat is lost but very little protein.
T h e effect of the smoking on b a c t e r i a has b e e n studied b y Aschehoug and Vesterhus ( 1 9 4 0 ) . T h e y found a decrease in the count of aerobes and anaerobes incubated at 3 7 ° C , and a large decrease in the count of aerobes at 2 2 ° C ; but a considerable n u m b e r survived light smoking.
Bacterial spores are found to b e very resistant to smoke ( T a n n e r , 1 9 4 4 ) . O f the spores from a 7-month-old spore culture, 7 1 % survived ex
posure to dense smoke for 7 hours while nonsporing organisms were killed in 1-2 hours. T h e temperature of the smoke is not mentioned.
5. Processing Apparatus
T w o types of kiln are used for smoking sardines. T h e old type is still in use. T h e fire in this type burns directly under the fish. A horizontal continuous kiln, t h e Kvaerner oven, was developed in 1933 and is now widely used. Another continuous but vertical kiln was constructed b y the Norwegian engineer, J . Sterner. In recent years n e w types of oven have b e e n under development: (1) combining steam cooking and smoking, (2) the I M C flash cooker, and ( 3 ) cooking with infrared rays ( t h e L u b e c a c o o k e r ) .
a. O L D T Y P E O F K I L N
This kiln is constructed as a vertical rectangular c h a m b e r with three main sections. T h e lower section has a fire burning directly under the middle section, w h e r e the fish hang on spits arranged in frames. T h e upper section narrows down, with a fan installed to circulate air and gases in the kiln and then to blow the smoke out to the chimney.
T h e size is moderate and each processor has his own construction.
T h e rectangular dimensions are about 0.75 m. X 1.25 m. ( 3 0 " X 5 0 " ) . T h e flames should not reach higher than about 5 0 cm. ( 2 0 i n c h e s ) below the fish. T h e section where the fish hang is about 2.10 m. ( 7 f e e t ) above the floor. T h e total height varies, b u t is usually about 5 - 6 m. ( 1 6 - 2 0 f e e t ) , the chimney not included. T h e fire is controlled b y dampers at the bottom and b y t h e fan. T h e wood is chopped into thin logs and must b e well dried, i.e., to a degree corresponding to drying in air for two years.
T h e fish are charged into the upper section of the kiln for the initial
6 . SCANDINAVIAN SARDINES 2 8 1 drying. T h e frames must b e turned during the process. As the drying ad
vances, they are m o v e d downward. All work is manual. I n the lower sec
tion the fish are exposed to t h e hot gases from the fire and here the smoke is applied. S i n c e the w o o d burns with free flames and the burning wood is red-hot, some effect of infrared heating can b e assumed. T h e process is controlled b y the worker and the result depends entirely on his skill. F u r t h e r m o r e , it is a wood-consuming process. T h e fat drip falls into the fire and cannot b e collected. S o m e canners regard the product from this kiln as superior to those obtained b y other manufacturing proc
esses.
b . K V A E R N E R C O N T I N U O U S K I L N
The K v a e r n e r kiln can b e used for smoke treatment of brisling and herring to sardines, Kieler sprotten, kippered herring, and buckling and can b e regulated to m e e t special specifications. Its main part consists of a horizontal tunnel with three separate heating furnaces outside. T h e frames charged with fish are p l a c e d in conveyor trolleys w h i c h pass the tunnel on rails and are driven b y an endless chain belt. E v e r y car has a wall of steel plates of the same dimensions as the cross section of t h e tunnel.
E a c h car thus constitutes a separate c h a m b e r and the smoke is forced to pass through these up and down consecutively. In t h e rear part of the kiln the final drying and smoking occur. T h e furnace Nos. 1 and 2 provide the heat and smoke required at the b a c k part of the tunnel. Since the hot gases in the loading section have only a heating purpose, other fuels than oak wood can b e used, such as gas, cokes, burning oils, or cheap wood.
A fan gives the proper draught through the tunnel and the gases leave the kiln at the loading end. Part o f the gases can, however, b e recirculated to the mixing c h a m b e r of the No. 1 furnace.
The process time is 4 0 - 5 0 min. and the capacity is about 6 - 7 metric tons per 8 hr. T h e temperature is held at l l ( M 3 0OC . at the end o f the process, and 8 0 - 1 1 0 ° C . at the start.
The Kvaerner kiln has a good heat e c o n o m y and cheap fuels m a y b e used in the furnace.
c. S T E R N E R K I L N
This is designed as a tower, divided into two c h a m b e r s . T h e frames are inserted at t h e b o t t o m o f the left c h a m b e r and elevated b y chains.
During the passage u p w a r d they travel in the same direction as the fuel gases from the bottom. At the top of the left c h a m b e r t h e frames are car
ried over to the right c h a m b e r , w h e r e they m o v e downward, m e e t i n g the smoke from the second furnace ( E g e l a n d , 1 9 5 2 ) .
d. I M C F L A S H T Y P E S A R D I N E C O O K E R F O R S M O K E D PRODUCTS
T h e I M C flash sardine cooker operates with the sardines p a c k e d raw in the can. I t is a complete unit comprising a smoke generator. T h e cans containing the dressed fish pass the steam cooker upside down, thus al
lowing free drainage o f t h e liquid b e i n g formed. T h e cooker is divided into two compartments. I n the first the cans are exposed to steam. Smoke is injected into this c h a m b e r and the smoke substances carried b y the steam are conveyed to the fish, on whose surface the steam condenses.
T h e smoke readily penetrates the fish, yielding a product o f uniform flavor. After cooking the products pass through a lock to the drying chamber, w h e r e a very humid air circulates. S m o k e m i x e d with air is in
j e c t e d here too, in order to give the product the proper smoke taint. E x perimental packs have b e e n described b y L a n g b a l l e ( 1 9 5 6 ) .
T h e flash cooking process is said to b e cheap, since the heat is supplied from a boiler b y steam having a m u c h higher h e a t capacity than hot air.
T h e product is exposed to a very gentle process and the fish temperature never exceeds 9 0 - 9 5 ° C . ( 1 9 4 - 2 0 3 ° F . ) . Loss of fat and nutrients is kept low.
e. I N F R A R E D R A D I A T I O N C O O K E R
Infrared radiation cookers have b e e n successfully tried during the last decade. T h r e e types are available (Lopez Capont, 1 9 5 7 ) : ( 1 ) In M o r o c c o a cooker was designed b y Philips, w h e r e the precooking is done in perforated cylinder, assuring easy drainage of liquids; it is e q u i p p e d with 5 2 parabolic lamps of 2 5 0 watts. ( 2 ) T h e F r e n c h T o q u e r system employs gas for the infrared heating. T h e s e two systems can b o t h b e used for precooking the fish either prior to p a c k i n g in cans or p a c k e d r a w in cans. ( 3 ) T h e H a r t m a n n process is b e i n g tried in the Scandinavian coun
tries, in part experimentally. T h e cooker comprises ( a ) a large c h a m b e r , ( b ) a conveyor w h i c h carries the cans through the cooker 3 times, ( c ) the infrared radiation windings, and ( d ) a fan for blowing hot air into the first part of the cooker. T h e conveyor is equipped with trays on which the cans are placed; during passage through the cooker the trays are twice inverted in order to drain the liquid.
T h e addition of smoke flavor and smoke color can b e accomplished electrostatically or b y means of steam in a c c o r d a n c e with t h e procedures indicated.
Substitution of metal tubes, heated with fuel oil, for electric lamps improved the performance of the infrared cookers (Cheftel, 1 9 5 7 ) .
G o o d results are reported with the L u b e c a infrared cooker ( E r i c h s e n , 1 9 5 8 ) . B o t h fresh and frozen fish were used as raw material. However,
6. SCANDINAVIAN SARDINES 2 8 3 even the slightest trace of rancidity of the fish caused an u n a c c e p t a b l e color and off-flavor. T h e s e changes do not appear during the actual radiation b u t during the final drying in hot air.
C. C O V E R O I L S AND S A U C E S
1. Edible Oils
F a n c y quality sardines are covered with virgin olive oil, i.e., olive oil produced b y m e c h a n i c a l processes (pressing, sedimentation, centrifu
gal separation, filtration, clarification b y c o l d ) or olive oil produced b y blending virgin olive oil and 3 0 % chemically refined virgin olive oil.
F o r standard quality sardines ( w i n t e r p a c k e d or tomato s a u c e ) , her
ring oil of the highest quality m a y also b e used. T h e oil specifications are issued b y the R e s e a r c h L a b o r a t o r y of the Norwegian Canning In
dustry, and are also approved in D e n m a r k a n d Sweden.
T h e amount of oil to b e a d d e d depends on the fat content of the fish, the rule b e i n g that the final fish product should contain not less than 3 0 % for fancy qualities and 2 8 % for other grades.
2. Tomato Sauce
T h e tomato sauce for covering consists of a mixture of at least 5 % olive, herring, or p e a n u t oil and a tomato puree, containing no m o r e than 3 0 - 3 2 % salt-free dry residue. L e a n fish require more oil than fat fish.
T o m a t o puree specifications are issued in Norway b y the R e s e a r c h L a b oratory of the Norwegian Canning Industry, in D e n m a r k b y the Depart
m e n t of Fisheries, and in S w e d e n b y the Swedish Control Service of C a n n e d Products. T h e y conform to some degree to the specifications issued b y Comite International Permanent de la Conserve, w h i c h o b serves U n i t e d States regulations.
D . T Y P E S O F C A N ; C A N S I Z E S
1. Denmark
T h e two sizes used for sardines are shown in the tabulation. B o t h sizes are used for export. T h e D i n g l e y is used for small fish and % 30 club for large fish.
Inside Inside Over-all
width length height Capacity
Denomination (mm.) (mm.) (ml.) (ml.)
14 Dingley 104.4 75.3 22.50 106
14 30 Club 104.4 59.4 30.0 127
2. Norway
List of cans for packing sardines is shown. j4 Dingley predominates.
Inside Inside Over-all
length width height Capacity
Denomination (mm.) (mm.) (mm.) (ml.)
y2 American 116 87 29.3 227
y2 American red 112 83 26.7 183
y2 Rectangular 105 76 40.3 272
y4 Dingley 105 76 21.3 112
y8 105 76 15.5 74
Vie 96 43 17 49
y4 Club 30 105 60 29.7 141
y4 Club 27 105 60 27 128
% Club 25 105 60 25 115
% Club 22 105 60 22.2 95
3 . Sweden
Sardine can sizes are shown. F o r export only 1 1 5 S I S 7 1 0 1 0 3 is used.
T h e others are encountered in the domestic market to a minor degree.
Inside Inside Over-all
length width height Capacity
Denomination (mm.) (mm.) (mm.) (ml.)
115 SIS 710103 105.0 76.0 21.5 115
150 SIS 710103 105.0 60.0 30.5 150
190 SIS 710103 105.0 76.0 30.0 190
Cans are usually m a d e of tin plate. W i n t e r p a c k e d sardines and sardines in tomato are p a c k e d in lacquered cans for protection against corrosion.
Aluminum cans have b e c o m e quite common in Norway.
E . T H E P R O C E S S I N G L I N E
1. Landing
T h e landed fish are stowed in wooden boxes with ice. T h e y are brought as quickly as possible to the canning plant, sometimes via fish auctions. S u m m e r caught brisling are always taken directly to the plant.
All boxes should b e new; boxes m a y b e used a second time only if quite clean and without odor.
2. Temporary Storing
T h e fish should b e quite fresh when processed. T o maintain freshness they may b e stored under refrigeration for a day. F r e e z i n g of raw ma-
6. SCANDINAVIAN SARDINES 2 8 5 terial for later processing has b e e n subject to experimentation as well as commercial trial. I f no precautionary measures are taken to maintain freshness, processing must p r o c e e d immediately b u t no fancy products can b e expected. F r e e z i n g in alginate has b e e n successful ( s e e Chapter 9, Section V ) .
3. Salting
F i s h are salt-treated b y immersion in brines, the strength of which varies with the fat content of the fish. T h e following brine concentration and duration of treatment are r e c o m m e n d e d .
Brining time Brine strength
Fish (min.) (degrees Baume)
Fat 7 15
Medium fat 5 15
Lean 5 10
T h e brining is usually performed in vats of wood or stainless steel.
T h e s e are charged with about 1 0 0 kg. fish. After brining, fish are dis
c h a r g e d from the vats b y means of hoop-nets and m a y b e washed in clean water. T h e salt p e r c e n t a g e of t h e brine is controlled and salt is a d d e d after e a c h lot o f fish to restore the original brine strength. Never
theless brine needs renewal when used repeatedly. W h e n fish are washed after brining, the brine treatment must b e extended a few minutes longer.
Automatic or continuous briners have b e e n devised. Baskets with covers are used in one type. T h e fish in the baskets are carried b y means of chain conveyors through the brining vats. Another design uses an end
less b e l t with paddles. T h e fish are p l a c e d at one end of the vat and the paddles bring them through the brine and discharge t h e m at t h e other end. T h e brine m a y also function as a conveyor system. O n e device con
sists of a spiral channel. T h e fish are p l a c e d in the center and brought through the spiral b y means of the brine. At the end of the spiral, the brine is drained off and recirculated after adding salt. T h e fish are dis
charged to a conveyor belt. T h e length of time for brining is controlled b y the speed of pumping.
I n automatic brining the salt penetrates the fish more rapidly. T h e salt treatment of the fish is also m o r e uniform than b y bulk processing.
T h e salt uptake should never e x c e e d 1.5%.
4. Spitting
During the smoking the fish must h a n g freely, without sticking to
gether. T h e y are h u n g on spits, w h i c h is usually done b y manual labor.
A special spitting table is used. T h e fish are arranged in a row at a b a r
divided into two parts joined along their entire length, the outside part of which is movable. Cavities are bored in the b a r to fit the heads of the fish. Through the whole b a r is a bored hole, fitting the spit and p l a c e d so that the spit will penetrate the eyes of the fish. W h e n the outside part of the b a r is opened the fish are seen hanging on the spit. T h e spits with the fish are arranged on a wooden frame, so that every fish hangs free. T h e frames are then brought to the smoking kiln either directly or p l a c e d in cars, depending on the type of kiln. D u r i n g the spitting, de
fective fish are sorted out. A spitting m a c h i n e has b e e n designed in Norway, but is not yet in common use.
5. Smoking
T h e smoking process and different types of smoking kilns are described above. After smoking, the fish should show a very light brownish color and b e firm and easy to handle. T h e skin should b e smooth and not blistery. I t must not burst.
6. Trimming
T h e frames with the smoked fish are brought to a head-cutting ma
chine. An endless b a n d knife cuts off the h e a d just b e l o w the gills. T h e b e h e a d e d fish fall on trays. T h e s e are passed b y an endless conveyor belt to the can-filling tables. T h e trays are washed, dried, and loaded onto the conveyor for return to the cutting machine.
7. Packing in Cans
T h e can-filling section is usually provided with a long straight con
veying system with working tables arranged perpendicularly to it, on one or both sides. T h e conveying system has three belt conveyors of stainless steel; the first carries trays with smoked fish, the second trays with empty cans, and the third trays with filled cans. T h e s e working tables are spaced for two or four girls.
Prior to loading cans on the conveyors, a small amount of oil or sauce is added b y a machine. T h e cans are then placed on trays of aluminum or stainless steel. T h e fish are sorted b y size and arranged in the cans with the dorsal side down, the cut ends alternately toward one or the other end of the can. Thus only the silver-white part of the fish is visible and the general appearance is most attractive.
T h e n u m b e r of fish in each can depends on size. T h e fancy, summer caught brisling sardines are rather small. I n the rectangular can, the largest sizes are p a c k e d lengthwise and the small ones transversely. In the 34 D i n g l e y can the sardines are p a c k e d in two layers. T h e fish must b e carefully graded b y size and the n u m b e r of fish is standardized as fol
lows:
6. SCANDINAVIAN SARDINES 2 8 7
Can designation Number of sardines
Yi Rectangular 28-34
Y4 Dingley, two layers 16-18, 18-22, 20-24 Y\ Dingley, cross-packed 32-36
Vs
Can 8-10l /1 6 Can 8-12, 10-14
T h e range of sizes of winter caught brisling or sild sardines is larger and the standard n u m b e r of fish e a c h can takes is as follows:
Can designation Count of sardines
Y Rectangular 24-28, 28-32
Τ/4 Dingley, two layers 18-22, 20-24
*4 Dingley, cross-packed 32-38
Vs
Can 8-10, 10-12Club 30, two layers 4-8
14 Club 27, two layers 4-6
% Club 25, two layers 4-7
% Club 22, one layer 4-6
All filled cans are inspected and undesirable cans removed.
8 . Covering with Oil or Sauce
T h e oil or sauce is added b y m a c h i n e in two steps. A small amount is first put into the empty can in order to bring the fish in contact with the covering medium. T h e completing fill is m a d e after the sardines have b e e n p l a c e d in the can. T h e amount o f filling can b e regulated b y the machine.
After the addition of oil or sauce the cans should remain standing for a while in order to allow the escape of air.
9. Closing the Cans
T h e cans are loaded on conveyors, w h i c h bring t h e m to the closing machines. Only double seamers are used and m a n y types of commercial machines are installed in plants for this purpose. At the present time the equipment is quite modern. H i g h speed automatic sealing machines are used to a large extent; t h e y are generally equipped with a coding auxil
iary.
10. Washing the Cans
W h e n cans have b e e n sealed, oil or sauce sticking on the can surface must b e r e m o v e d b y washing in water with detergents. T h e washing is performed b y spraying or dipping.
11. Retorting
T h e cans are p l a c e d in the retort crates. T h e y can b e arranged in tight regular layers or plunged into the carts from conveyors. T h e retorts are usually of the counterpressure type and the processing is carried out in water. T h e pressure outside the cans must b a l a n c e the pressure inside in order to maintain the original shape of the can as m u c h as possible. This is needed particularly with aluminum cans, which are easily damaged unless precautions are taken.
T h e r e c o m m e n d e d heat processing time for the most common sizes of can, y4 Dingley and ys and y1 6 cans, is 6 0 min. at 1 1 2 ° C . ( 2 2 4 ° F . ) . If overpressure is used, a total pressure of 18 p.s.i. is recommended. A process of 7 0 min. at 1 1 0 ° C . should b e used for tomato sardines in order to reduce the risk of browning reactions of the Maillard type.
Final cooling of the cans after heat processing is done b y pumping chilled water into the retort. H o t water is in the m e a n t i m e p u m p e d into a storage tank, from which it is released into the retort to warm up sub
sequent lots of cans.
12. Washing, Labeling, and Packing in Boxes
T h e cans are cleaned in a washing m a c h i n e after retorting. H o t water with detergents removes dirt. Cans should b e heated only enough so that excessive water evaporates.
T h e cans are usually p a c k e d in paperboard crates or boxes before labeling and storing for a minimum of 4 weeks. After labeling they are inspected for defects. Printed labels are common, but lithograph cans, are b e i n g increasingly used. Cans are labeled on the sides, wrapped in parch
ment paper, and sealed with a top label.
Sardines are shipped in wooden or cardboard boxes, each containing 100 cans. Cardboard boxes predominate and are used exclusively in Sweden.
R E F E R E N C E S
Andersson, K. A. (1942). "Fiskar och Fiske i Skandinavien," Vol. I, pp. 227-254.
Natur & Kultur, Stockholm.
Anonymous (1954). The situation. Tidskr. Hermetikind. 40, 2, 67.
Anonymous (1958). Determination of the fat percentage of brisling 1957. Tidskr.
Hermetikind. 44, 245-253.
Anonymous (1959, 1961, 1962). Statistics on canned foods. Konserves 17, 129; 19, 52; 20, 139.
Aschehoug, V., and Vesterhus, R. (1940). Bacteriological investigation of the raw material of kippers and sardines. Tidskr. Hermetikind. 26, 49-53.
B0rresen, R. (1952). Drying problems in the smoking process. Tidskr. Hermetik
ind. 38, 303-326, 353-365.
