7.1. Basics of fresh fish handling and use of ice
7.2. Fish handling in artisanal fisheries
7.3. Improved catch handling in industrial fisheries

7.1 Basics of fresh fish handling and use of ice

Throughout history, man has preferred to consume fresh fish rather than other types of fish products. However, fish spoil very quickly and man has had to develop methods to preserve fish very early in history.

Keeping and transporting live fish

The first obvious way of avoiding spoilage and loss of quality is to keep caught fish alive until consumption. Handling of live fish for trade and consumption has been practised in China with carp probably for more than three thousand years. Today, keeping fish alive for consumption is a common fish-handling practice both in developed and developing countries and at both artisanal and industrial level.

In the case of live fish handling, fish are first conditioned in a container with clean water, while the damaged, sick and dead fish are removed. Fish are put to starve and, if possible, water temperature is reduced in order to reduce metabolic rates and make fish less active. Low metabolic rates decrease the fouling of water with ammonia, nitrite and carbon dioxide that are toxic to fish and impair their ability to extract oxygen from water. Such toxic substances will tend to increase mortality rates. Less active fish allow for an increase in the packing density of fish in the container.

A large number of fish species are usually kept alive in holding basins, floating cages, wells and fish yards. Holding basins, normally associated with fish culture companies, can be equipped with oxygen control, water filtering and circulation and temperature control. However, more simple methods are also used in practice, for instance large palm woven baskets acting as floating cages in rivers (China), or simple fish yards constructed in a backwater of a river or rivulet for large "surubi" (Platystoma spp.), "pacu" (Colossoma spp.) and "pirarucu" (Arapalma gigas) in the Amazonian and Parana basins in South America.

Methods of transporting live fish range from very sophisticated systems installed on trucks that regulate temperature, filter and recycle water and add oxygen (Schoemaker, 1991), to very simple artisanal systems of transporting fish in plastic bags with an oxygen supersaturated atmosphere (Berka, 1986). There are trucks that can transport up to 50 t of live salmon; however, there is also the possibility of transporting a few kilo-grammes of live fish relatively easily in a plastic bag.

By now a large number of species, inter alia, salmon, trout, carp, eel, seabream, flounder, turbot, catfish, Clarias, tilapias, mussels, oysters, cockles, shrimp, crab and lobster are kept alive and transported, very often from one country to another.

There are wide differences in the behaviour and resistance of the various species. Therefore the method of keeping and transporting live fish should be tailored according to the particular species and the length of time it needs to be kept outside its natural habitat before slaughtering. For instance, the lungfish (Protopterus spp.) can be transported and kept alive out of water for long periods, merely by keeping its skin moist.

Some species of fish, noticeably freshwater fish, are more resistant than others to changes in oxygen in solution and the presence of toxic substances. This is probably due to the fact that their biology is adapted to the wide yearly variations in water composition presented by some rivers (cycles of matter in suspension and dissolved oxygen). In these cases, live fish are kept and transported just by changing the water from time to time in the transport containers (See Figures 7.1 (a) and (b)). This method is widely used in the Amazonian, Parana and Orinoco basins in South America; in Asia (particularly in the People's Republic of China, where also more sophisticated methods are used) and in Africa (N'Goma, 1993).

In the case presented in Figure 7.1 (a), aluminium containers with live freshwater fish are stored in the aisles of a public transport vessel. Containers are covered with palm leaves and water hyacinth to prevent the fish from jumping out of the containers and to reduce evaporation. The water in the containers is changed from time to time and an almost continuous visual control is kept on fish. Dead fish are immediately put to smoke-drying (African style) in drum smokers, also transported in the vessels or transporting barges.

In the case presented in Figure 7. 1(b), carp is kept in a metal container drawn by a bicycle. This is a rather common practice in China, and other Asian countries; for instance in Bangkok, live catfish is sold daily by street vendors.


Figure 7.1 (a) Transport of live freshwater fish in Congo (Cuvette Congolaise) (N'Goma, 1993); (b) street vendor of live fish in China today (Suzhou, 1993, photo H. Lupin) 

The most recent development is the keeping and transporting of fish in a state of hibernation. In this method, the body temperature of live fish is reduced drastically in order to reduce fish metabolism and to eliminate fish movement completely. The method greatly reduces death rates and increases package density, but careful temperature control should be exercised to maintain the hibernation temperature. There is an appropriate hibernation temperature for each species. Although the method is already utilized for instance to transport live "kuruma" shrimp (Penaeus japonicus) and lobster in pre-chilled wet sawdust, it should be considered an experimental technique for most of the species.

Although keeping and transporting live fish is becoming more and more important, it is not a viable solution for most of the bulk fish captures in the world.

Chilling fish with ice

Historical evidence proves that the Ancient Chinese utilized natural ice to preserve fish more than three thousand years ago. Natural ice mixed with seaweed was also used by the Ancient Romans to keep fish fresh. However, it was the development of mechanical refrigeration which made ice readily available for use in fish preservation.

In developed countries, particularly in USA and some European countries, the tradition of chilling fish with ice dates back more than a century. The practical advantages of utilizing ice in fresh fish handling are therefore well established. However, it is worthwhile for young generations of fish technologists and newcomers to the field, to review them, paying attention to the main points of this technique.

Ice is utilized in fish preservation for one or more of the following reasons:

(i) Temperature reduction. By reducing temperature to about 0C the growth of spoilage and pathogenic micro-organisms (see section 6) is reduced, thus reducing the spoilage rate and reducing or eliminating some safety risks.

Temperature reduction also reduces the rate of enzymatic reactions, in particular those linked to early post mortem changes extending, if properly applied, the rigor mortis period.

Fish temperature reduction is by far the most important effect of ice utilization. Therefore, the quicker the ice chills the better. Although cold-shock reactions have been reported in a few tropical species when iced, leading to a loss of yield of fillets (Curran et al., 1986), the advantage of quick chilling usually outweighs other considerations. The development of ad hoc fish handling methods is of course not ruled out in the case of species that could present cold-shock behaviour.

(ii) Melting ice keeps fish moist. This action mainly prevents surface dehydration and reduces weight losses. Melting water also increases the heat transport between fish and ice surfaces (water conducts heat better than air): the quickest practical chilling rate is obtained in a slurry of water and ice (e.g., the CSW system).

If, for some reason, ice is not utilized immediately after catching the fish, it is worthwhile keeping the fish moist. Evaporative cooling usually reduces the surface temperature of fish below the optimum growth temperature of common spoilage and pathogenic bacteria; although it does not prevent spoiling.

Ice should also be utilized in relation with chilling rooms to keep fish moist. It is advisable to keep chilling room temperature slightly above 0C (e.g., 3-4C).

However, water has a leaching effect and may drain away colour pigments from fish skin and gills. Ice melting water can also leach micronutrients in the case of fillets and extract relatively large amounts of soluble substances in some species (e.g., squid).

Depending on the species, severity of leaching and market requirements, an ad hoc handling procedure may be justified. In general, it has been found that drainage of ice meltwater is advisable in boxes and containers and that permanence of fish in chilled sea water (CSW) and refrigerated seawater (RSW) should be carefully assessed if leaching and other effects (e.g., uptake of salt from the seawater, whitening of fish eyes and gills) are to be avoided.

During the past there was much discussion about allowing drainage from one fish box to another, and consequent reduction or increase of bacterial load by washing with drainage water. Today, apart from the fact that in many cases box design allows for external drainage of each box in a stack, it is recognized that these aspects have less importance when compared with the need for quick reduction in temperature.

(iii) Advantageous physical properties. Ice has some advantages when compared with other cooling methods, including refrigeration by air. The properties can be listed as follows:

(a) Ice has a large cooling capacity. The latent heat of fusion of ice is about 80 kcal/kg. This means that a comparatively small amount of ice will be needed to cool 1 kg of fish.

For example, for I kg of lean fish at 25C, about 0.25 kg of melted ice will be needed to reduce its temperature to 0C (see Equation 7.c). The reason why more ice is needed in practice is mainly because ice melting should compensate for thermal losses.

The correct understanding of this ice characteristic is the main reason for the introduction of insulated fish containers in fish handling, particularly in tropical climates. The rationale is: ice keeps fish and the insulated container keeps ice. The possibility to handle fish with reduced amounts of ice improves the efficiency and economics of fresh fish handling (more volume available for fish in containers, trucks and cold storage rooms, less weight to transport and handle, reduction in ice consumption, less water consumed and less water drained).

(b) Ice melting is a self-contained temperature control system. Ice melting is a change in the physical state of ice (from solid to liquid), and in current conditions it occurs at a constant temperature (0C).

This is a very fortunate property without which it would be impossible to put fresh fish of uniform quality on the market. Ice that melts around a fish has this property on all contact points. In the case of mechanical refrigeration systems (e.g., air and RSW) a mechanical or electronic control system (properly tuned) is needed; nevertheless, controlled temperature will be always an average temperature.

Depending on the volume, design and control scheme of mechanical refrigeration systems, different temperature gradients may appear in chill storage rooms and RSW holds, with fish slow freezing in one comer and maybe above 4C in another comer. Although the need for proper records and control of temperature of chill storage rooms has been emphasized recently in connection with the application of HACCP (Hazard Analysis Critical Control Point) to fresh fish handling, it is clear that the only system that can assure accurate temperature control at the local level (e.g., in any box within a chill storage room) is ice melting.

Ice made of sea water melts at a lower temperature than fresh water ice, depending on the salt content. Theoretically with 3.5 % of salt content (the average salt content of seawater) seawater ice will melt at about - 2.1C However, as ice made out of seawater is physically unstable (ice will tend to separate from salt), brine will leach out during storage lowering the overall temperature (and this is the reason why sea water ice always seems wet). In these conditions, fish may become partially frozen in storage conditions and there may be some intake of salt by the fish muscle. Therefore, it cannot be said that ice made out of seawater has a proper self-controlled temperature system.

There is a narrow range of temperature below 0C before fish muscle starts to freeze. The freezing point of fish muscle depends on the concentration of different solutes in the tissue fluids: for cod and haddock, it is in the range of -0.8 to - 1 C, for halibut -1 to -1.2C, and for herring about -1.4C (Sikorski, 1990).

The process of keeping fish below 0C and above the freezing point is called superchilling, and it allows achievement of dramatic increases in overall keeping times. In principle it could be obtained using seawater ice or mixtures of seawater and freshwater ice, or ice made out of a 2% brine and/or mechanical refrigeration. However, in large volumes it is very difficult to control temperature so precisely and temperature gradients, partial freezing of fish in some pockets and hence lack of uniformity in quality are unavoidable (see section 6. 1).

(iv) Convenience. Ice has a number of practical properties that makes its use advantageous. They are:

(a) It is a portable cooling method. It can be easily stored, transported and used. Depending on the type of ice, it can be distributed uniformly around fish.

(b) Raw material to produce ice is widely available. Although clean, pure water is becoming increasingly difficult to find, it is still possible to consider it a widely available raw material. When there is no assurance that freshwater to produce ice will be up to the standard of drinking water, it should be properly treated, e.g., chlorination.

Clean seawater can also be utilized to produce ice. Ice from seawater is usually produced where freshwater is expensive or in short supply. However, it should be remembered that harbour waters are hardly suitable for this purpose.

(c) Ice can be a relatively cheap method of preserving fish. This is particularly true if ice is properly produced (avoiding wastage of energy at ice plant level), stored (to avoid losses) and utilized properly (not wasted).

(d) Ice is a safe food-grade substance. If produced properly and utilizing drinking water, ice is a safe food substance and does not entail any harm either to consumers or those handling it. Ice should be handled as food.

(v) Extended shelf life. The overall reason for icing fish is to extend fresh fish shelf life in a relative simple way as compared to storage of un-iced fish at ambient temperatures above 0C (see Chapter 6). However, extension of shelf life is not an end in itself, it is a means for producing safe fresh fish of acceptable quality.

Most landed fish can be considered a commodity, that is, an article of trade. Unlike other food commodities, it is usually highly perishable and it is thus in the interest of the seller and the buyer to ensure fish safety at least until it is consumed or further processed into a less perishable product. Ice and refrigeration in general, by making possible extension of fish shelf life, convert fresh fish into a true trade commodity, both at local and international level.

Ice is used to make fish safe and of better quality to consumers. It is also used because otherwise the current fish trade at local and international level would be impossible. Shelf life is extended because there is a strong economic reason to do so. Fishermen and fish processors who fail to handle fresh fish appropriately ignore the essence of their business. The inability to recognize fresh fish also as a trade commodity is at the root of misunderstandings and difficulties linked to the improvement of fish handling methods and prevention of post-harvest losses.

Types of ice

Ice can be produced in different shapes; the most commonly utilized in fish utilization are flake, plate, tube and block. Block ice is ground before being utilized to chill fish.

Ice from freshwater, of whatever source, is always ice and small differences in salt content or water hardness do not have any practical influence, even if compared with ice made out of distilled water. The physical characteristics of the different types of ice are given in Table 7.1.

Cooling capacity is expressed by weight of ice (80 kcal/kg); therefore it is clear from Table 7.1 that the same volume of two different types of ice will not have the same cooling capacity. Ice volume per unit of weight can be more than twice that of water, and this is important when ice stowage and volume occupied by ice in a box or container are considered. Ice necessary to cool fish to 0C or to compensate for thermal losses is always expressed in kilogrammes.

Under tropical conditions ice starts to melt very quickly. Part of the melted water drains away but part is retained on the ice surface. The larger the ice surface per unit of weight the larger the amount of water retained on the ice surface. Direct calorimetric determinations show that at 27C the water on the surface of flake ice at steady conditions is around 12-16% of the total weight and in crushed ice, 10-14% (Boeri et al., 1985). To avoid this problem, ice may be subcooled; however, under tropical conditions this effect is quickly lost. Therefore a given weight of wet ice will not have the same cooling capacity as the same weight of dry (or subcooled) ice, and this should be taken into account when making estimations of ice consumption.

Table 7.1 Physical characteristics of ice utilized in chilling fish. Adapted from Myers (1981)


Approximate Dimensions (1)

Specific volume (m3/t) (2)

 Specific weight (t/m3)


10/20 - 2/3 mm

2.2 -2.3 



30/50 - 8/15 mm

1.7 - 1.8



50(D)- 10/12 mm

1.6 - 2.0



Variable (3)



Crushed block


1.4 - 1.5

0.71 -0.66


(1) They depend on the type and adjustment of the ice machine.
(2) Indicative values, it is advisable to determine them in practice for each type of ice plant.
(3) Usually in blocks of 25 or 50 kg each.

There is always the question of which is the "best" ice to chill fish. There is no single answer. In general, flake ice will allow for an easier, more uniform and gentle distribution of ice around fish and in the box or container and will produce very little or no mechanical damage to fish and will chill fish rather more quickly than the other types of ice (see Figure 7.2). On the other hand, flake ice will tend to occupy more volume of the box or container for the same cooling capacity and if wet, its cooling capacity will be reduced more than the other types of ice (since it has a higher area per unit of weight).

With crushed ice there is always the risk of large and sharp pieces of ice that can damage fish physically. However, crushed ice usually contains fines that melt quickly on the fish surface and large pieces of ice that tend to last longer and compensate for thermal losses. Block ice requires less stowage volume for transport, melts slowly, and contains less water at the time it is crushed than flake or plate ice. For these reasons, many artisanal fishermen utilize block ice (e.g., in Colombia, Senegal and the Philippines).

Probably tube ice and crushed ice are more suitable for use in CSW systems if ice is wet (as it normally is under tropical conditions), since they will contain less water on their surfaces.

There are also economic and maintenance aspects that may play a role in deciding for one type of ice or another. The fish technologist should be prepared to analyze the different aspects involved.

Cooling rates

Cooling rates depend mainly on the surface per unit of weight of fish exposed to ice or chilled ice/water slurry. The larger the area per unit of weight the quicker the cooling rate and the shorter the time required to reach a temperature around 0C at the thermal centre of the fish. This concept is also expressed as "the thicker the fish the lower the cooling rate".

Small species such as shrimp, sardines, anchovies and jack mackerels cool very quickly if properly handled (e.g., in CSW or CW). Large fish (e.g., tuna, bonito, large sharks) could take considerable time to cool. Fish with fat layers and thick skin will take longer to cool than lean fish and fish with thin skin of the same size.

In the case of large fish, it is advisable to gut them and to put ice into the empty belly as well as around it. In large sharks, gutting alone may not be enough to prevent spoilage during chilling, and therefore it is advisable to gut the shark, to skin it and to cut the flesh into sizeable portions (e.g., 2-3 cm thick) and to chill them as soon as possible. Chilled sea water (CSW) has in this case the advantage of extracting some of the urea present in shark muscle (see section 4.4). However, this is an extreme case, since in current situations fillets kept in ice will last less time than gutted fish or whole fish (because of the unavoidable microbial invasion of the flesh) and will lose soluble substances.

Typical curves for cooling fish in ice, using different types of ice and chilled water (CW) are shown in Figure 7.2.

From Figure 7.2 it is clear that the quickest method to chill fish is with chilled water (CW) or chilled sea water (CSW), although the practical difference with flake ice is not great. There are, however, noticeable differences after the quick initial drop in temperature with crushed block ice and tube ice, due to differences in contact areas between fish and ice and flow of melt-water.

Cooling curves may also be affected by the type of container and external temperature. Since ice will melt to cool fish and simultaneously to compensate for thermal losses, temperature gradients may appear in actual boxes and containers. This type of temperature gradient could affect the cooling rate, particularly in boxes at the top or side of the stacks, and more likely with tube and block crushed ice.

Curves such as those shown in Figure 7.2 are useful to determine the critical limit of chilling rates when applying HACCP to fresh fish handling. For instance, in specifying a critical limit for chilling fish "to be at 4.5C in the thermal centre in no more than 4 hours", in the case of Figure 7.2 it could be achieved only by using flake ice or CW (or CSW).

In most cases the delay in reaching 0C in the thermal centre of the fish may not have much practical influence because the surface temperature of the fish will be at 0C. On the other hand, warming-up of the fish is much riskier because the fish surface temperature (which is actually the riskiest point) will almost immediately be at the external temperature, and therefore ready for spoilage. As large fish will take longer than small fish to warm up and also have less surface area (where spoilage starts) per unit of volume than small fish, they usually take a little longer to spoil than small fish. This circumstance has been widely used (and abused) in practice in the handling of large species (e.g., tuna and Nile perch).


Figure 7.2 Chilling of large yellow croaker (Pseudosciaena crocea) with three different types of ice and chilled water (CW). Ice-to-fish ratio 1: 1; the same type of insulated containers (with drainage) was used in a parallel experiment (data obtained at the FAO/DANIDA National Workshop on Advances in Chilling and Processing Technology of Fish, Shanghai, China, June 1986)

Small species will warm up very quickly and definitely more quickly than large species (warming-up the same reason for which they cool faster). Although warming-up studies of fresh fish have received little attention in the past, they are necessary within an HACCP scheme, to determine critical limits (e.g., maximum time fish can be handled without ice in a fish processing line).

With application of HACCP and HACCP-based systems, thermometers including electronic thermometers, should be a standard tool in fish processing plants. Therefore, it is advisable to perform fish cooling and warming-up trials on actual conditions.

Ice consumption

Ice consumption can be assessed as the sum of two components: the ice necessary to cool fish to 0C and the ice to compensate for thermal losses through the sides of the box or container.

Ice necessary to cool fish to 0C

The amount of ice theoretically necessary to cool down fish from a temperature Tf to 0C using ice can easily be calculated from the following energy balance:

L mi = mf cpf (Tf - 0)                                                              7.a


L = latent heat of fusion of ice (80 kcal/kg)
mi = mass of ice to be melted (kg)
mf = mass of fish to be cooled (kg)
cpf = specific heat capacity of fish (kcal/kg C)

From (7.a) it emerges that:

mi = mf cpf Tf / L                                                                      7.b

The specific heat capacity of lean fish is approximately 0. 8 (kcal/kg C). This means that as a first approximation:

mi = mf Tf / 100                                                                          7.c

This is a very convenient formula, easily remembered, to quickly estimate the quantity of ice needed to cool fish to 0C.

Fatty fish have lower cpf values than lean fish and, in theory, require less ice per kilogramme than lean fish; however, for safety purposes it is advisable to make calculations as if fish were always lean. Refinements in the determination of cpf are possible; however, they do not drastically alter the results.

The theoretical quantity necessary to cool fish to 0C is relatively small and in practice much more ice is used to keep chilled fish. If we relate the proper fish handling principle of surrounding middle and large sized fish with ice, to the approximate dimensions of ice pieces (see Table 7. 1), it is clear that with some types of ice (tube, crushed block and plate) greater quantities are required for physical considerations alone.

However, the main reason for using more ice is losses. There are losses due to wet ice and ice spilt during fish handling, but by far the most important losses are thermal losses.

Ice necessary to compensate for thermal losses

In principle, the energy balance between the energy taken by the melted ice to compensate heat from outside the box or container could be expressed as follows:

  L   (dMi/dt) = U A (Te - Ti)                                                    7. d


Mi = mass of ice melted to compensate for thermal losses (kg)
U = overall heat transfer coefficient (kcal/hour m2 C)
A = surface area of the container (m2)
Te = external temperature
Ti = ice temperature (usually taken as 0 C)
t = time (hours)

Equation (7.d) can be easily integrated (assuming Te = constant) and the result can be expressed as:

Mi = Mio - (U A Te / L) t                                               7. e

It is possible to estimate thermal losses, calculating U and measuring A. However, this type of calculation will seldom give an accurate indication of ice requirements, for a number of practical factors (lack of reliable data on materials and conditions, irregularities in the construction of containers, irregular geometric shape of boxes and containers, influence of lid and drainage, radiation effect, type of stack).

More accurate calculations of ice requirements can be made if meltage tests are used to determine the overall heat transfer coefficient of the box or container, under actual working conditions (Boeri et al., 1985; Lupin, 1986 a).

Ice meltage tests are very easy to conduct and no fish are needed. Containers or boxes should be filled with ice and weighed before commencing the test. At given periods, the melted water is drained (if it has not already drained) and the container is weighed again. The reduction of weight is an indication of the ice lost due to thermal losses. In Figure 7.3 the results of two ice meltage tests obtained under field conditions are presented.

Initially, some ice will be melted to cool down the walls of the box or container; depending on the relative size and weight of the container, wall materials and thickness and entity of the thermal losses this amount may be negligible. If it is not, the container can be cooled down before starting the test, or the ice necessary to cool down the container can be calculated by the difference disregarding the first part of the meltage test. A constant air surrounding temperature would be preferable and it can be achieved during short periods (e.g., the testing of a plastic box in tropical conditions). However, reasonably constant temperatures may be achieved during the intervals between weight loss measurements and an average used in the calculations.

Results as shown by Figure 7.3 can be interpolated empirically by a straight line equation of the form:

Mi = Mio - K t                                                         7.f

Comparing Equations 7.e and 7.f, it is clear that:

K = (Uef Aef Te / L)


Uef = overall effective heat transfer coefficient
Aef = effective surface area








Figure 7.3 Results of ice meltage tests under field conditions. ( ) standard plastic box (not insulated) 40 kg total capacity, (x) insulated plastic fish container (Metabox 70, DK). Both kept in the shade, un-stacked, flake ice, average external temperature (Te) 28C. (Data obtained during the FAO/DANIDA National Workshop on Fish Technology and Quality Control, Bissau, Guinea-Bissau, March 1986)

 From Expression 7.g it follows that:

K = K' - Te                                                                           7. h

and eventually K' could be determined, if experiments can be conducted at different controlled temperatures.

The advantage of meltage tests is that K can be obtained experimentally from the slope of straight lines, as appears in Figure 7.3, either graphically or by numerical regression (now found as sub-routine in common pocket scientific calculators). In the case of the straight lines appearing in Figure 7.3 the correlations found are as follows:

Plastic box:

Mi = 10.29 - 1.13 t ,  r = -0.995                                        7.i
K = 1. 13 kg of ice/hour

Insulated container:

Mi = 9.86 - 0.17 t ,  r = 0.998                                            7.j
K = 0. 17 kg of ice/hour

where r = correlation coefficient.

From 7.i and 7.j it follows that the ice consumption due to thermal losses in these conditions will be 6.6 times greater in the plastic box than in the insulated container. It is clear that under tropical conditions it will be practically impossible to handle fish in ice properly utilizing only non-insulated boxes, and that insulated containers will be needed, even if additional mechanical refrigeration is used.

The total amount of ice needed will be the result of adding mi (see Equations 7.b and 7.c) to Mi (according to expression 7.f) once t (the time fish should be kept chilled in the box or container in the particular case) has been estimated.

Under tropical conditions it may happen that, depending on the estimated t, total available volume in the box or container might not be enough even for ice to compensate for thermal losses, or the remaining volume for fish could be insufficient to make the chilling operation attractive.

In such cases it might be feasible to introduce one or more re-icing steps, or to resort to additional mechanical refrigeration (see Figure 7.5 to observe the effect of storage in a chill room on ice consumption). In practice, an indication of when re-icing is needed would be given to foremen or people in charge.

An analytical approach to this problem in connection with the estimation of the right ice-to-fish ratio in insulated containers can be found in Lupin (1986 b).

Ice consumption in the shade and in the sun

An important consideration, particularly in tropical countries, is the increased ice consumption in boxes and insulated containers when exposed to the sun. Figure 7.4 gives the results of a experimental meltage test conducted with a box in the shade and the same box (same colour) in the sun.

The plastic box in the shade is the same plastic box of Figure 7.3 (see Equation 7.i). The correlation for the plastic box in the sun is:

Mi = 9.62 - 3.126 t                                                           7. k

This means that for this condition and this type of box, the ice consumption in the sun will be 2.75 times that in the shade (3.126/1.13). This considerable difference is due to the radiation effect. Depending on the surface material, type of material, colour of the surface and solar irradiation, it will be a surface radiation temperature, that is higher than dry bulb temperature. Direct measurements on plastic surfaces of boxes and containers on field conditions, in tropical countries, have given values of surface radiation temperature up to 70C.


Figure 7.4 Results of ice meltage tests under field conditions. () plastic box in the shade, (x) plastic box in the sun. Plastic boxes, 40 kg capacity, red colour, unstacked, flake ice, external average temperature (dry bulb) 28C. (Data obtained during the FAO/DANIDA National Workshop on Fish Technology and Quality Control, Bissau, Guinea-Bissau, March 1986) 

It is clear that there is little practical possibility in tropical countries to handle chilled fish in plastic boxes exposed to the sun. An increase in ice consumption, even if less dramatic than in plastic boxes, can be measured in insulated containers exposed to the sun.

The obvious advice in this case is to keep and handle fish boxes and containers in the shade. This measure can be complemented by covering the boxes or containers with a wet tarpaulin. The wet tarpaulin will reduce the temperature of the air in contact with boxes and containers to the wet bulb temperature (some degrees below the dry bulb temperature, depending on the Equilibrium Relative Humidity - ERH - of the air), and will practically stop noticeable radiation effect (since there are always radiation effects between a body and its background).

Ice consumption in stacks of boxes and containers

In a stack of boxes or containers not all of them will lose ice in the same way. Figure 7.5 gives the results of an ice meltage test conducted on a stack of boxes. Boxes or containers at the top will consume more ice than boxes and containers at the bottom, and those in the middle will consume less than either.

Figure 7.5 Results of ice meltage tests during storage in a stack of plastic boxes. Plastic boxes 35 kg in a chill storage room at 5C, flake ice (from Boeri et al. (1985)

Jensen and Hansen (1973) and Hansen (1981) presented a system ("Icibox"), mainly for artisanal fisheries. In this system, a stack of plastic boxes were insulated by placing wooden frames, filled with polystyrene, at the top and at the bottom of the stack, and covering the whole with a case made out of canvas or oil skin. A similar system, composed of stacks of styropor boxes, accommodated in a pallet, and covered by an insulated mat of high reflective (Al) surface, is used in practice for shipment of fresh fish by air (e.g., it is utilized to ship fresh fillets of Nile perch from Lake Victoria to Europe).

Results of Figure 7.5 are also of interest to demonstrate the effect of a chill room on fresh fish handling. The use of chill rooms drastically reduces the ice consumption in plastic boxes, avoiding the need of re-icing. In a fish handling system chilling fish with ice, mechanical refrigeration is used to reduce the ice consumption and not to chill fish.

Although analytical models of ice consumption (e.g., Equations 7.a to 7.h) can be applied directly to estimate the ice consumption in simple and repetitive fish handling operations, their main importance is that they can help in arriving at solutions for the proper handling of chilled fish in rational. way (as seen from Figures 7.3, 7.4 and 7.5).

Ice consumption in the sides of boxes and containers

It is necessary to bear in mind that ice will not melt uniformly in the interior of a box or container, but meltage will follow the pattern of temperature gradients between the interior of the box/container and the ambient. In Figure 7.6, a commercial plastic box with chilled hake shows the lack of ice in the sides due to the temperature gradients at the walls.

Following Figure 7.5, and supposing that a simple box could be devided into five subboxes, it is clear that the bottom and top of boxes and containers should receive more ice to compensate for thermal losses, the top receiving more ice than the bottom. However, in practice more ice should also be put in the sides of boxes and containers.


Figure 7.6 Commercial plastic box with chilled hake (M. hubbsi) showing the effects of lack of ice in the sides (photo H. Lupin)

The box of Figure 7.6 was initially prepared with enough ice, and it can be seen that ice is still abundant on topof the box. However, after a period of storage in a chill room, ice has melted, mainly on the sides, leaving some fish and parts of fish exposed to the air with a consequent rise in temperature and dehydration. In addition, ice and fish have formed a compact mass that can produce physical damage to exposed fish when the box is moved.

In chilled fish onboard fishing vessels or transported by truck, this problem may not exist if there is a continuous gentle movement which allows for ice melt water from the top to move to the sides. However, in chill rooms or storage rooms (insulated containers) it would be advisable to re-ice if this problem is observed. Under tropical conditions this effect is observed, even with insulated containers, in less than 24 hours of storage.

7.2 Fish handling in artisanal fisheries

Artisanal fisheries, existing both in developed and developing countries, encompass a very wide range of fishing boats from pirogues and canoes (large and small) to small outboard and onboard engine vessels, utilizing also a variety of fishing gears. It is difficult to find a common denominator; however, from a fish handling point of view, artisanal vessels handle relatively small amounts of fish (when compared with industrial vessels) and fishing journeys are usually short (usually less than one day and very often only a few hours).

In general, in tropical fisheries the artisanal fleet land a variety of species, although there are examples of the use of selective fishing gear. In temperate and cold climates artisanal fleets can focus more easily on specific species according to the period of the year; nevertheless, they may land a variety of species to respond to the market demand.

Although very often artisanal fisheries are seen as an unsophisticated practice, closer scrutiny will reveal that in many cases they are passing through a process change. There are many reasons for this process but very often the main driving forces are: urbanization, fish exports and competition with the industrial fleet.

This change in the scenario of artisanal fisheries is essential to understanding the fish handling problems faced by the artisanal and small sector of the fish industry, particularly in developing countries.

When the artisanal fleet was serving small villages, the amount of fish handled was very low; the customers usually bought the fish direct from the landing places, fishermen knew customers and their tastes, and fish was consumed within a few hours (e.g., fish caught at 06.00 h, landed and sold at 10.00 h, cooked and consumed by 13.00 h). In this situation, ice was not used, and gutting was unknown; very often fish arrived at landing places in rigor mortis (depending on fish species and fishing gear), and fish handling was at most reduced to covering the fish from the sun, keeping it moist and keeping off the flies. In Figure 7.7 two cases of landing un-iced fish by artisanal fishermen are shown.

Figure 7.7 Landing by artisanal fishermen: (a) un-iced shrimp by artisanal fishermen (El Salvador, September 1987, photo H. Lupin); (b) un-iced fish (Bukova, Tanzania, 1994, photo S.P. Chen)

With urbanization and the request for safer and more quality products (as a result of exports and competition with industrial fish) conditions changed drastically. Large cities also demanded increased fish supplies, and thus middlemen and fish processors had to go to more distant landing places for fish. The amount of fish handled increased, fishing journeys lasted longer and/or passive fishing gears like gillnets were set to fish for longer times, a chain of middlemen and/or official fish markets replaced the direct buyer at the beach and, as a result of growing business (fish for income), in some places the catch effort also increased with a consequent increase in the number of fishing boats and an increase in the efficiency of the fishing gears.

In one way or another, each of the new circumstances added hours to the time which passed between catching the fish and eating or processing it (e.g., freezing). This increase in exposure of un- iced fish to ambient temperature (or water temperature for a dead fish in a gillnet), even though brief (e.g., an additional 6-12 hours), dramatically changed the situation regarding fish spoilage and safety.

In the new situation, fish remained at ambient temperature some 13-19 or more hours. It could be already spoiled, at terminal quality and/or could present public health hazards (e.g., from the development of C. botulinum toxin to histamine formation). In addition to the safety and quality aspects, post-harvest losses, non-existent at subsistence level and very low at the village stage, become important. For instance, it is estimated that the post-harvest losses of Nile perch caught artisanally in Uganda amount to 25-30% of the total catch.

The situation described in previous paragraphs, and cases like those shown in Figure 7.7, moved extension services in developing countries and international technical assistance to focus on the problem of introducing improved fish handling methods at the artisanal level. The basic technical solution is the introduction of ice, proper fish handling methods and insulated containers, which is the approach utilized by most of the artisanal fleet in developed countries.

There are several examples where this approach was adopted by fishermen in developing countries and has become a self-sustained technology. Two very interesting cases to analyze are the introduction of insulated containers onboard of "navas", the traditional fishing vessels of Kakinada in Andhra Pradesh, India (Clucas, 1991) and the introduction of insulated fish containers in the pirogue fleet of Senegal (Coackley and Karnicki, 1984). The sketch of an insulated fish container for Senegalese pirogues is shown in Figure 7.8.

The insulated container of Figure 7.8 was designed to fit existing pirogues, according to the type of catch and needs expressed by fishermen. The materials and tools needed to construct the insulated container are available to fishermen in Senegal, even though some of them are imported (e.g., foam sheets and resin).

The example of Senegalese fishermen is now spreading steadily to similar fisheries in Gambia, Guinea-Bissau and Guinea which are adopting the use of insulated containers similar to those of Senegal. However, the process of diffusion and adoption of a technology, even if relatively simple, is not as straightforward as could be supposed. A pirogue with two insulated containers onboard is shown in Figure 7.9.

Once artisanal fishermen become aware of the rationale of insulated containers, they tend to favour large insulated fish containers rather than small ones. The reason is clear from Equations 7.e and 7.g, as for the same volume of fish and ice, large containers will present less external area than the area presented by several small containers. For example, a large cubic insulated fish container can be envisaged of a side measuring x m, and eight cubic insulated containers of sides equal to x/2 m presenting the same total volume as the large one. The eight containers will have an external area twice that of the big container, thus increasing the ice consumption by two, and decreasing the amount of fish that can be transported.

Other reasons are that small containers will cost more than a large one of the same total volume (simply because they need more material); small containers are not always easy to secure safely onboard small boats, and large containers allow for transport of large ice bars that can be crushed at sea (reducing stowage rate). However, large containers are difficult to handle and sometimes canoes and pirogues are very small or narrow and they cannot accommodate large insulated fish containers. This is the case for relatively small insulated fish containers. An example is shown in Figure 7. 10.


Figure 7.8 Sketch diagram of a two-hatch insulated container for Senegalese pirogues (after Coackley and Karnicki, 1985) 

Figure 7.9 A Senegalese pirogue at the beach, carrying two insulated containers (photo B. Diakit, 1992)

Figure 7.10 Small insulated container installed onboard an artisanal fish catamaran (The Philippines, 1982, photo H. Lupin) 

A serious constraint in many artisanal fisheries is the relatively high cost of industrial containers and the difficulty in finding appropriate industrial materials to construct them. For this reason, efforts have been made to develop artisanal containers made from locally available materials (Villadsen et al, 1979; Govindan, 1985; Clucas and Whitehead, 1987; Makene, Mgawe and Mlay, 1989; Wood and Cole, 1989; Johnson and Clucas, 1990; Lupin, 1994).

In some cases, the correct approach could be to add insulation to local fish containers; in other cases it could be necessary to develop a new container. In general, artisanal fish could be cheaper than industrial fish containers, but they will not last as long. An artisanal insulated container developed at Mbegani (Tanzania), based on the local basket container ("tenga") is shown in Figure 7.11.

A key factor in the construction of artisanal insulated containers is the selection of insulation material. There are a number of materials available: inter alia, sawdust, coconut fibre, straw, rice husks, dried grass, old tires and rejected cotton.

However, the use of such materials presents problems: the materials become wet very quickly (with the exception of old tires), losing their insulating capacity and increasing the weight of the container. When wet, most of them tend to rot very quickly. The solution is to put them inside a plastic bag (waterproof); however, in this case they tend to settle, leaving part of the walls without insulation.

With a view to overcoming these problems, the concept of "insulated pillows" was developed in various FAO/DANIDA fish technology workshops. This concept is very simple: the insulating material (e.g., coconut fibres) is placed inside one plastic tube of the type usually found to produce ordinary small polyethylene bags (10 cm in diameter); the insulating material is pressed before sealing the tube; the tube is sealed by heat at both ends (e.g., every 20 cm), and with some practice it is possible to produce a strip of "pillows". It is advisable to utilize a second tube to reduce the incidence of punctures due to fish spines and bones.

Figure 7.11 (a) Sketch of an artisanal insulated container (the "Mbegani fish container") developed and utilized in Tanzania; (b) The "Mbegani fish container" on a bicycle to distribute fresh fish. This container was initially developed at the FAO/DANIDA National Workshop on Fish Technology and Quality Control, held at Mbegani, Tanzania, May-June 1984

The strip of "insulated pillows" can then be placed between the internal and the external walls of the container. Once the container is finished with an insulated lid and handles, fish and ice can be put in a large resistant plastic bag, as shown in Figure 7.11 (a). The use of the plastic bag extends the lifespan of the container and improves fish quality.

This example indicates the type of practical problems found when developing an artisanal insulated fish container, and the possible solutions.

Why is ice not always used to chill fish when necessary

Despite the knowledge on the advantages of fish chilling, ice it is not as widely used as it should be, particularly at artisanal level in developing countries. Which are the main  problems found in practice? Some of the problems that can be found are as follows:

(i) Ice should be produced mechanically

This obvious statement implies, inter alia, that it is not possible to produce ice artisanally for practical purposes (machines and energy are required). To produce ice under tropical conditions, from 55 to 85 kWh/l ton of ice (depending on the type of ice) are necessary whereas, in cold and temperate countries from 40 to 60 kWh are required for the same purpose. This may be a large power requirement for many locations in developing countries, particularly in islands and places relatively far from large cities or electricity networks. Ice plants require maintenance and hence trained people and spare parts (in many cases this requires access to hard currency).

A cold chain will also require chill rooms (onboard and on land), insulated containers, insulated trucks and other auxiliary equipment (e.g., water treatment units, electric generators). Besides increasing the cost, all this equipment will increase the technological difficulty associated with the fish cold chain.

(ii) Ice is produced and used within an economic context

In developed countries ice is very cheap and costs only a fraction of the price of fresh fish. In developing countries ice is very often expensive when compared with fresh fish prices.

A survey conducted in 1986 by the FAO/DANIDA Project on Training on Fish Technology and Quality Control on current fish and ice prices in fourteen African countries demonstrated that in all cases and for all the fish species, I kg of ice increased the fish price at least twice the rate recorded in developed countries. The cheaper the fish the worse the situation. For instance, in the case of small pelagics, the percentage of increase in the fish cost per kilogramme of ice added, was 40% for the "yaboy" of Senegal, 16-25% for the sardinella of Congo, and 66 % for the sardinella of Mauritania and the anchovy of Togo. The market price for fish, in this case, acts as a deterrent for the use of ice.

According to the relative cost of ice to fish, ice may or may not be used. For instance, in Accra, Ghana in 1992, it was found that using ice to chill small pelagics (Ghanian herring) in a proportion of 2 kg ice: 1 kg fish would increase the cost of fish by 32-40%. However, in the case of snapper, for the same ratio of ice to fish the cost increase would be in the range of 4.5-5.7%. The result is that ice chilling of snapper is relatively common in Accra, whereas ice is not utilized to chill small pelagics.

Very often fish compete with other sources of demand (soft drinks, beer), even if the ice machine was initially installed to supply ice for chilling fish. This and energy losses at the ice plants contribute to increase the market price of ice.

In addition to producing and utilizing ice on a sustainable basis, economic aspects must be considered (e.g., depreciation, reserves, investment). Moreover, in the case of ice manufacture there is a strong influence of the scale of production. Low ice prices in developed countries are also the result of large ice plants located at the fishing harbours that supply a large number of companies and fishing boats.

(iii) Practical constraints

Introduction of ice into fish handling systems that are not accustomed to using it can create practical problems. For instance, from Table 7.1 it is clear that the introduction of ice will increase the volume required for storage and distribution, and will reduce the effective fish hold in vessels. The use of ice will also increase the weight to be handled. This will have a number of implications such as an increased workload for the fishermen, fish processors and fishmongers, and an increase in costs and investment.

From Figures 7.3 and 7.4 it is clear that the total amount of ice needed per 1 kg of fish, in the complete cycle from the sea to the consumer will be much higher in tropical countries than in cold and temperate regions. As an indication, the average consumption of ice in the Cuban fishery industry was estimated at around 5 kg of ice per 1 kg of fish handled (including ice losses), although higher values (up to 8-10 kg of ice per 1 kg of fish) have been recorded in single industries in tropical countries; this necessitates large storage and transport capacities.

Freshwater or seawater utilized for producing ice should comply with standards (microbiological and chemical) for potable water and should be readily available in the volumes required. This is not always possible particularly in countries with energy problems (blackouts) and without (or with erratic) public tap- water distribution. If water has to be treated, this implies additional costs and additional equipment to operate and maintain.

Properly trained personnel are required to operate the ice plant and auxiliary equipment efficiently, and to handle ice and fish properly. Although many developing countries have made efforts to train people, in many cases there is a lack of technical personnel ranging from well trained fish technologists to refrigeration mechanics or electricians, or simply plant foremen.

Moreover, in many developing countries it is increasingly difficult to keep technical and professional schools operating in this field, thus jeopardizing the possibility of self-sustained training, and hence fishery industry developments.

(iv) Ice is not an additive

Knowledgeable people (e.g., fishmongers) are quickly aware of the fact that ice is not an additive. Therefore, when there is a delay in icing, ice is not usually utilized (even if available) because it will not improve fish quality. Consumers could also be intuitively aware of this fact, and they prefer to be presented with the fish as it is (e.g., at the terminal state of its quality) rather than in ice, because in this case ice will increase the price of fish but not enhance its quality. Due to the above and to the problems associated with the transition between artisanal and industrial or semi-industrial fisheries, already discussed, consumers in some countries (e.g., in Saint Lucia and Libya) tend to believe that iced fish is not fresh fish.

A need for chilled fish can develop if a market for iced fish (not just a market for "fresh fish") is developed, and to develop a market for iced fish where it does not already exist may be a very difficult and expensive endeavour as is the introduction of any other food product.

(v) Need for appropriate fish handling technologies

To chill and keep fish with ice is a very simple technique. A more complicated picture emerges when actual fish handling systems are analysed, including the economic aspect.

From a comparative study on the same fish handling operation, utilizing ice and insulated containers, carried out in both a developed and a developing country, it was seen that in developed countries, the more "appropriate" technology would aim at reducing wage costs (e.g., chutes to handle ice and fish, special tables to handle containers and boxes and conveyors to move them, machines that mix ice and fish automatically); in developing countries the main concern would be to reduce ice consumption, and to increase the fish : ice ratio in the containers (Lupin, 1986 b).

The same study found that a twentyfold difference in wage costs between developing countries and developed countries cannot offset a tenfold difference in the cost of ice. There is no "comparative advantage" in low wages in developing countries with regard to fresh fish handling. Advanced technology on fish handling from developed countries could make work easier for people in developing countries, but might not improve the economics of the operation as a whole.

There is obviously no single solution to the problems discussed above. However, it is clear that it is the problem to be solved in the coming decade in the field of fresh fish handling. With total catches having reached a plateau, losses due to the lack of ice utilization could be ill-afforded, and developing countries and artisanal fishermen in particular should not be deprived of potential market opportunities.

7.3 Improved catch handling in industrial fisheries

The aims of modern catch handling are the following:

To meet these aims, equipment and handling procedures that will eliminate heavy lifting, unsuitable working positions and rough handling of fish must be introduced. By doing so, the catch handling time is accelerated and the chilling process initiated much earlier than was previously the case (Olsen, 1992). The typical unit operations in catch handling are shown in Figure 7.12.

Figure 7.12 Typical unit operations in catch handling of pelagic and demersal fish

Important general aspects in modern catch handling are:

Catch handling can be done in several ways ranging from manual methods to fully automated operations. How many operations will be used in practice and the order in which they are done depends on the fish species, the fishing gear used, vessel size, duration of the voyage and the market which has to be supplied.

Transferring catch from gear to vessel

Midwater trawlers and purse seiners fishing pelagic fish use tackling in lifts of up to 4 t, pumping or brailing for bringing the catch onboard. When lifting huge hauls (100 t or more) onboard by these methods, the danger of losing fish and gear always exists if the fish start to sink after having been brought to the surface. The speed of which the fish may sink depends on the species, catching depth and weather conditions during hauling.

Pumping the catch onboard using submersible pumps without bruising the fish can be difficult, as it is not easy to control the fish-to-water ratio during pumping.

In recent years, the so-called P/V pump (P/V - pressure/vacuum) has found increasing use. The P/V-pump principle is that an accumulation tank of 500-1500 1 size is alternately put under vacuum and pressure by a water-ring vacuum-pump (Figure 7.13). The fish, together with some water, are sucked through a hose and a valve into the tank of the system. When the tank is full, it is pressurized by changing the vacuum and pressure side connections from the tank to the pump and the fish/water mix flows through a valve and a hose into a strainer. The P/V-pump is claimed to handle the fish more gently than other fish pump types, but the capacity is generally lower, mostly because of the alternating way of operations. This problem can be solved by having two P/V-tanks running in phase opposition using only one vacuum-pump.

Figure 7.13 Working principle of a P/V pump

Small gillnetters (10-15 m) haul the nets with the net hauler, and very often store their catch in the net until landing. Here the net is drawn through a net shaker by two men in order to free the fish from the gear. It has been shown that the violent way in which the shaker works can be harmful to the men's hands, arms and shoulders. Ergonomic precautions have therefore been suggested to overcome this problem.

Trawlers and seiners (Danish and Scottish) tackle the catch into pounds. Commonly used pounds are those with a raised bottom which can be hoisted hydraulically. The purpose of these designs is to provide good working conditions for the crew (Figure 7.14). Also gillnetters may use a work-saving pound system, which is often connected with a conveyor to bring fish to the gutting-table.

Figure 7.14 Deck lay-out for trawler using machine gutting of demersal fish 
1. Tackle pound, 2. Hoisting pound, 3. Gutting table, 4. Bleeding/washing machine, 5. Gutting machine, 6. Chair.

Holding of catch before handling

When large catches are to be handled, or if for other reasons catch handling cannot start immediately, it is convenient and necessary to prechill the catch during holding in deck-pounds using ice or in tanks using Refrigerated Sea Water (RSW) or a mixture of ice and sea water (Chilled Sea Water, CSW).

Prechilling holding systems are mostly used on pelagic trawlers which grade their catches in size before storing in boxes or in portable CSW-containers. It is also essential to prechill when the pelagic fish are soft and feeding and therefore very prone to bellyburst. Prechilling tanks are unloaded by elevator or P/V-pumps. If no sorting is done onboard, the fish is conveyed directly for chilled storage in the hold. A system for holding demersal fish in tanks is shown in Figure 7.15.

Figure 7.15 System comprising CSW raw material holding tanks before manual or machine gutting of fish 


Pelagic fish are sometimes sorted or graded onboard according to size. The equipment used operates on the basis of thickness of fish using principles such as:

Grading by thickness can meet the demand for the high capacity needed in pelagic fish handling, but it is generally accepted that the correlations between thickness and length or weight are not too good (Hewitt, 1980). The most important point, often forgotten, for making a grader function at its optimum is even feeding. This could be done with an elevator delivering to a (vibrating) water sprayed chute leading to the inlet guide chute of the grading machine.

Sometimes it is necessary to install a manual sorting conveyor before the grading machine for removal of larger fish and debris, e.g., in the fishery for argentine with by catch of grenadier.

Sorting and grading of demersal fish by species and by size is normally done by hand. However, some automatic systems of sorting according to width are in use. Static or dynamic weighing by marine weighing systems are also in use with good results. Research is under way using a computerized vision system for species and size grading.


In order to obtain optimal quality in a white fillet, many white-fleshed demersal fish (but not all) need to be bled and gutted immediately after capture. The best procedures from an economic, biological and practical point of view are still under discussion (see section 3.2 on bleeding and section 6.4 on gutting).

The vast majority of fishermen are handling the fish in the easiest and also the fastest way, which means the fish are bled and gutted in one single operation. This may be done manually, but gutting machines have been introduced to obtain even more speed. The fish are transported to and from the fisherman by suitable conveyor systems. Using machines, round fish can be gutted with a speed of approximately 55 fish/minute for fish length up to 52 cm and 35 fish/minute for fish length up to 75 cm. Gutting by machine is 6-7 times faster than hand-gutting.

Existing gutting machines for round fish of the type using a circular saw blade for cutting and removing the guts destroy the valuable roe and liver. A new type of gutting machine which copies the manual gutting procedure is now available on the market. Gutting speed of this machine is 35-40 fish/minute and the roe and liver can be saved (Olsen, 1991). Flatfish can also be gutted by a recently developed machine. The speed of this machine is about 30 fish/minute.

After gutting, the fish are conveyed to the washing or bleeding operation. This may be done in pounds, often with raised bottom or in special bleeding tanks, frequently with a hydraulically-operated tilting system and rotating washing drums are also used (Figure 7.15); and special equipment such as the Norwegian and British fish washer may be used.

After catch handling (sorting, grading, gutting, etc.), the fish may be passed to an intermediate storage silo or batch holding system for the different sizes or grades before being dropped by chute to the hold, or the chutes may lead directly from the grading machines to the hold (Figure 7.16).


Figure 7.16 "Polar"-system. Mechanized sorting and boxing of herring 1. Herring sorting machine, 2,3,4. Conveyors, 5. Flexible dosing tube.

Chilling/chilled storage

Demersal fish have traditionally been stored on shelves or in boxes. Boxing has a big advantage over shelf storage as it reduces the static pressure on the fish and also facilitates unloading.

Shelf storage is done by alternating layers of ice and fish from one layer of ice and fish (single shelving 25 cm between shelves) up to ice/fish layers 100 cm deep. In practice, shelving often allows better temperature control than boxing and therefore also a longer storage life. Because excessive handling during unloading and excessive pressure on the fish have- a negative effect on quality, e.g., appearance, boxing is preferable to shelving, given proper icing.

In pelagic fisheries, boxed fish will be untouched until processed, but in demersal fisheries the catch is often only sorted by species onboard and not graded by size and weighed. These operations are carried out after landing before auction whereby some of the handling and quality advantages of boxing are lost.

In the near future when integrated quality assurance systems have been introduced, these unit operations will be carried out onboard and an informative label on each box will give details of factors of importance for first-hand sale (including freshness).

In general, two types of plastic fish boxes are used: stack-only and nest/stack boxes (Figures 17 a and 17 b).

To overcome some of the space problems in using stack-only boxes, the nest/stack type has been developed. These occupy only approximately a third of the space needed when stored empty compared to when the boxes are loaded with fish and ice.

Fig. 7.17 a Stack-only boxes                      Figure 7.17 b Nest/stack boxes

This type of box is widely used in France, the Netherlands and Germany and also in some Danish ports. When a system tailor-made for a certain type of plastic box is designed, the quality advantages of using boxes can be fully utilized onboard. The key points to consider are:

  1. The handling rate necessary to prevent quality loss because of delayed icing. Prechilling can be of advantage to compensate lack in handling rate.
  2. Handling methods which make it possible to guarantee that the icing procedure is sufficient to chill the fish to 0C and maintain this temperature until landing.
  3. The hold construction must be constructed such that safe and easy stacking of the boxes can take place.
  4. Hold insulation of a relatively high quality should be considered. A small mechanical refrigeration plant can be of advantage. Air temperature in the hold should be + 1-3C

RSW-storage (Refrigerated Sea Water is a well established practice which has been refined both theoretically and practically since its introduction in the 1960s in Canada where it was developed for salmon and herring storage (Roach et al, 1967). At the beginning, most RSW vessels were salmon- packers and because of some failures in design which were attributed either to insufficient refrigeration or circulation systems, a standard for control of RSW-systems was established. Since vessels are different, the RSW-installation has to be studied carefully in every fishery to determine its real capability. Therefore, methods for rating each individual system and vessel and providing general specifications and guidelines for the proper installation have been suggested by the Canadian technicians (Gibbard and Roach, 1976).

In order to obtain maximum shelf life from RSW-systems, temperature homogeneity in the region of -1C is very important. The factors affecting temperature homogeneity were recently studied in Denmark (Kraus, 1992). The most important conclusions were that the inflow of the chilled seawater in the bottom of the tank must take place over the whole tank bottom area, and that filling capacity for securing water circulation and temperature homogeneity is dependent on fish species. The necessary chilling rate was suggested to be: fish temperature must be below 3C within four hours and below 0C after 16 hours, and the temperature should be kept between -1.5C and 0C until unloading.

The CSW system has also been developed in Canada as a much cheaper means an investment point of view - to obtain rapid uniform chilling of fish. The most popular method used is the so-called "Champagne" method where rapid heat transfer between fish and ice is obtained by agitation with compressed air introduced at the bottom of the tanks, instead of using circulation pumps as in RSW and some earlier CSW designs (Figure 7.18) (Kelmann, 1977; Lee, 1985). An indication of the chilling rate for herring could be: reduction of fish temperature from 15C to 0C within two hours. The concept of a CSW system is to load well insulated tanks at the harbour with the amount of ice necessary to chill the catch to between 0' and - 1C and maintain this temperature until unloading.

Figure 7.18 Chilled seawater system: piping layout

The Canadian west-coast fishermen are achieving this in practice by using a minimum of seawater when they start loading the tank and by forcing air through the ice-sea-water-fish-mixture only during loading, and stop forcing air immediately when the tankis full. Thereafter they will force the air only for 5-10 minutes with 3-4 hours' interval. The air agitation therefore only serves as a method to overcome local temperature differences in the tank. The objective is to obtain a uniform mixture of fish and ice in order to secure temperature homogeneity.

A proven rule-of-thumb for estimating the amount of ice necessary is simply to observe the amount of ice left in the tank at unloading, and compare it with temperature readings, which should be in the -1C range measured in the landed fish. The starting situation should be conservative, which at sea-temperature around 12-14C, for a trip lasting 7 days and with 10 cm polyurethane insulation, is 25% ice by weight of the tank capacity. The amount of ice is adjusted according to the observations on the following trips.

An analytical approach to estimate necessary ice quantities in a CSW tank system has been developed. The quantity of ice required takes into consideration tank size, catch volume, time at sea, water temperature, hold insulation and hold flooding strategy (Kolbe et al., 1985).

CSW "Champagne" systems can also be used in small coastal vessels, e.g., in a fishery for small pelagic fish with vessels of 10-14 m length with a fish carrying capacity from 3 to 10 t fish (Roach, 1980).

Another way of loading a CSW tank, which is in practical use in Denmark, is to add the necessary amount of ice to the fish during loading by mixing a controlled stream of fish with a controlled stream of ice. The greatest amount of ice is added to the fish during loading. When the tank is full the voids are filled with seawater from a hose and the tank is left undisturbed, except for watercirculation by pumping or compressed air blowing for 5-10 minutes of 4-hour intervals. The ice is bulk-stored in the forward hold and the ice is shovelled into a conveyor flush with the floor. The conveyor then leads the ice to the mixing point at the deck.

The use of portable CSW containers for pelagic fish handling was tested in the early 1970s (Eddie and Hopper, 1974). The approximately 2 m3 heat insulated containers were loaded with the necessary amount of ice from the harbour and agitated with compressed air in a similar way as for CSW-tanks. The main advantages with this method are that the fish will be undisturbed until processed and easily unloaded. The disadvantages are: marketing problems and reduced pay-load on existing vessels (Eddie, 1980). Portable 1.1m3 CSW containers are used to a limited extent in combination with the earlier mentioned conveyor system originally laid-out for boxing without the above-mentioned reduced pay-load compared to boxing (Anon., 1986). Also, small coastal vessels can use insulated portable CSW containers (Figure 7.19).

Figure 7.19 Some of the 10 pieces of 200 l CSW containers placed on deck on a 15 GRT cod gillnet wooden boat 


Shelfed fish are unloaded, using baskets or boxes which are filled as the shelves are removed. The fish are tackled from the hold and emptied on a conveyor leading to the manual grading and weighing process.

Boxed fish iced in 20 or 40 kg boxes at sea will normally be unloaded in pallet loads of, for instance, twelve 40 kg boxes per pallet. Swedish boats use hydraulic deck-mounted cranes and a special pallet fork during unloading. An unloading rate of approximately 30 t/h is possible by this method.

Danish coastal vessels, landing their pelagic catches daily, use quay mounted P/V-pumps for unloading their catches, which often are iced in pens in layers up to approximatly 1 m height. It is necessary only to add small quantities of water to make the pump function properly. The fish is delivered to a strainer from where a conveyer leads the fish to a size grader. The strained water is recirculated to the hold. Grading machines with up to 30 t/h are often installed.

In Scandinavia the 30-50 in RSW/CSW vessels still use brailing to a limited extent when unloading their catches at a rate of 30 to 50 t/h. The main disadvantage of this method is that very big hatches are needed to obtain reasonable unloading rates.

P/V-pumps have recently been introduced for unloading herring and mackerel. Thus vessels with small tanks, e.g., 30 in , and small hatches can also be unloaded at a rate similar to or higher than the above-mentioned brailing rate. P/V-pumping rates will typically be around 40-50 t/h. The fish can be transported directly in a tube system into the factory where representative samples are taken for quality assessment.