THE KINDS AND LEVELS OF POST-HARVEST LOSSES IN
AFRICAN INLAND FISHERIES

by

Geoffrey R. Ames
Natural Resources Institute, Chatham, UK

1. INTRODUCTION

At first sight, post-harvest losses appear a straightforward matter. An amount of fish is caught but a smaller quantity reaches the consumer. The difference is what concerns us now. However, the real situation is far from straightforward. Losses are not simply a matter of quantities of material. We must also consider losses of value, what the fish is worth in monetary terms through the handling, processing, distribution and marketing cycles. There are losses of quality, when stale or mouldy fish becomes less attractive to consumers. However, these can be regarded as either losses in material or more usually, losses in value; as the quality drops there is nearly always a decrease in value. Also there are losses in nutritional value, when the fish contribute less towards the diet of consumer than it did, or might have done. It is convenient to start by considering these different types of loss.

2. LOSSES OF MATERIAL

One of the most obvious cases of loss of material is insect infestation of cured fish: blowflies attack fish during drying and beetles attack dried or smoked fish after it has been processed. The flesh of the fish is eaten by the insects before reaching the human consumer. There is also spoilage, when fish becomes putrid and uneatable due to biological decomposition. If the marketing or processing systems break down (for example, if flooded roads prevent distribution or if fish cannot be preserved because of a lack of fuel for smoking or because of adverse weather) then the fish may be completely unusable. Also smoked or dried fish may break into pieces during packing or distribution and the fragments may be lost. Other factors in both spoilage and fragmentation need to be considered further, in a different context.

3. LOSSES IN VALUE

3.1 The losses of material will inevitably involve a loss in value, as the fisherman, processor, or distributor has less weight of material to sell. Further, the material may command a lower relative price. Dried fish which has been attacked and partly eaten by insects will be less attractive to consumers than undamaged fish, and its price per kilogramme will usually be lower. Not only is there less to sell, but what can be sold is worth relatively less. Here we have a material loss and a bigger financial loss, as someone in the chain has lost the value of the weight of fish eaten by insects and a drop in value of the remainder.

3.2 Spoilage of wet fish is accompanied by a loss in value. Consumers will pay more for fish which is in good condition. The relative value of good/fair/poor quality fish is a complicated matter, and varies between countries and places. Sometimes people will pay more for fresh or iced fish, and sometimes not. However, no one will pay as much for fish which is starting to become putrid as they would for fish in better condition. As fish spoils, its value drops even though this can happen in different ways. If fish can be landed and sold to the final consumer within a few hours of catching, the fish is fresh enough for this effect to be negligible. If ice is used, preferably on board or at least after landing (or after catching for land-based fishing), there is not likely to be any change in the value for many days; but if the fish cannot be iced there will normally be a quick drop in value.

3.3 Losses in material tend to be absolute, as the fish can either be sold or eaten or it cannot. Losses in value are progressive and gradual and much more subjective. Different groups of people may have a different view of what something is “worth”.

3.4 In most parts of the world fresh (wet) fish commands a higher price than traditional cured fish¹. Consequently the need to sun-dry or smoke fish in order to preserve it during the marketing cycle is itself a financial loss. The fishermen and/or those involved in the initial marketing would have received more money had it been possible to preserve the fish in ice. This type of loss is not usually included in discussions of “post-harvest losses” but it clearly is a loss to someone, and it provides a first example of how difficult it is to define “losses”. However, in the author's view it is misleading to regard all cured fish production as “losses”.

¹ Using the term “cured fish” to denote products which have a longer shelf life than the unprocessed material. Flavour modified products such as smoked salmon, which are still highly perishable, are not included here.

4. NUTRITIONAL LOSSES

4.1 These are a factor when fish is preserved by smoking, as many procedures involve heating the fish to fairly high temperatures such as 150°C or more. Under these conditions, the nutritional value of the fish protein can be damaged. In particular there are serious losses in the availability of lysine and other essential amino-acids. Recent work at NRI has indicated that similar damage could occur at much lower temperatures such as 60°C, so perhaps losses of lysine are also involved in sun-drying.

4.2 Nevertheless these losses in nutritional value will not necessarily be significant if potential consumers receive an adequate diet from other sources. Many people in various parts of the world eat only relatively small quantities of cured fish yet may obtain enough lysine from other foods. However, many others are heavily dependent on cured fish for their protein needs. The author is not aware of any studies of the nutritional status of the diets of those specific groups in developing countries which are heavily dependent on cured fish. The latter are often lower income groups who may not be able to afford other protein foods richer in lysine, so this would appear to be a problem of some significance.

4.3 As fish spoils there will be a loss in nutritional value. The bacteria which cause the spoilage themselves eat the protein which is intended to be the main human food. This is not generally a major factor, as bacterial action generates nitrogenous substances with objectionable smells. The fish will become highly unattractive before there is too much nutritional damage. It could be more significant in freshwater fish, when less violent ammoniacal odours are produced than in marine fish, where trimethylamine and its breakdown products rapidly impart strong odours.

5. IS IT ACTUALLY LOST?

5.1 After a discussion of what losses there are, further factors need to be considered. When fish undergoes bacteriological spoilage, the loss in quality is accompanied by a reduction in commercial value. This may bring the product within the reach of low-income groups who could not afford better quality. The people involved in fish marketing (and perhaps the fishermen) may have suffered a loss of potential income by not selling at the best possible price, but someone else may have gained by having access to a still nutritious food. Even when fish has deteriorated so much that it cannot be sold at all, and it is thrown away, there are regrettably often people who are so impoverished that they would be glad to take the least spoiled fish. This can be regarded as a loss in value for the fisherman or trader, but a social gain for very low-income groups.

5.2 In a different context, fish which is lost in one sense may return as a different food. When cured fish undergoes fragmentation one can regard the smaller fragments as lost. These fragments are often fed to chickens; and as people eat the chickens, are the fragments really lost? If one is specifically referring to losses of fish as direct human food, fragments fed to chickens must be regarded as lost.

5.3 Another form of loss is the excessive consumption of fuel by some smoking kilns. In many African, countries fuel wood supplies are under heavy pressure, leading to deforestation and soil erosion. If smoking kilns use more fuel than really necessary, the country itself suffers a loss in terms of energy resources and environment. This is not generally regarded as a post-harvest loss of fish, and it is not considered further here.

6. SUMMARY DEFINITION OF TYPES OF LOSS

The type of action needed in each case is set out in Table 1.

Table 1

Action areas for losses

7. MAGNITUDE OF LOSSES

7.1 In 1988 the present author and his colleagues presented a review of post-harvest losses of cured fish, in which they summarized all the published data which could be traced, together with much unpublished material. This covered material losses of the main types of cured fish, and it was specified whether the data were obtained by measurements or if they were estimates. The summary is reproduced here, as Table 2, and it will be seen that there is a very wide variation between the results and that a high proportion are estimates rather than measurements.

7.2 There does not seem to have been any similar compilation of estimates of measurements of material losses of wet fish. Few studies of the losses of value have been carried out. Again, the reports that have been published are mostly estimates rather than measurements.

8. PROBLEMS OF MEASUREMENTS AND QUANTIFICATION

8.1 It will be noted from the Table that many of the published figures for post-harvest losses are “Estimates”. To estimate losses of cured fish due to insect infestation the principle is fairly straightforward: how much fish have the insects eaten? The issue is not really so simple as not all the fish in a batch will have been attacked to the same extent. Some fish may be severely damaged, with a very substantial loss in weight, while others may have little visible damage. The former may be thrown away while the latter may be sold at the normal price for undamaged fish. If, say, one fish in ten is discarded then the trader will have lost 10% of his fish, but this does not correspond to an average loss in weight of 10% over the whole batch.

8.2 Many of the estimates are purely visual. Someone has looked at a lot of fish and guessed what proportion of the weight has been lost. However good the intentions the observer's assessment can only be a guess. Often the estimates are based on the opinion expressed by processors or traders as to the quantity they lost. Few such people are likely to have kept detailed records, and the quantities they suggest may be inflated for effect (like the fisherman's legendary “one that got away”) or to arouse sympathy.

8.3 Another factor is that estimates are usually based on observations at a very few places on very few occasions. Quite often someone has visited a few landing sites or markets during one or two weeks in the dry season and repeated the observations in a similar way during the wet season. The observer may conclude that losses were, say, negligible in the dry season, 20% in the wet season, and therefore averaged 10% over the year. Such an approach is unscientific, even in the rare event of wet and dry seasons both lasting for six months, with uniform landings of fish throughout the year. The observer will be aware of these uncertainties, and will qualify his estimates by saying that losses “may be as much as…” or “could be about…”. Unfortunately, such phrases are likely to be omitted when other people refer to the report or publication. When the original estimate has been referred to in a number of other publications it may become “it is generally accepted that losses are…”.

8.4 Measuring losses is not an easy task. For example, in measuring blowfly losses during drying, one starts with wet fish, that is mostly water, and ends with a very different product with much less water. Even if the weight of fish consumed by the insect is known, is this given as a percentage on a wet or dry basis? In considering beetle infestation of cured fish in store, allowance must be made for spontaneous charges in moisture content. The product may dry out, losing weight through dehydration as well as beetle attack. Alternatively, if the ambient humidity increases, the product may gain more weight by absorption of water than it loses through insect infestation.

8.5 Wood (1985) has drawn up a detailed scheme for assessing losses of cured fish which allows for changes in moisture, bone and salt content but it is, unavoidably, somewhat complicated. Table 3 shows the steps involved. As for the estimates, the measurements of cured fish losses have usually been carried out in a very limited number of places in the stated country. Even if wet and dry seasons have been considered separately, it is usually only within a single year. There do not appear to have been any studies of cured fish losses comparing losses in different parts of a country or on a year-by-year basis.

8.6 Spoilage losses of wet fish are also difficult to assess. In any marketing system for perishable produce, those concerned are prepared to sustain a certain proportion of losses due to spoilage. Someone who has bought a quantity of fish is likely to have a fair idea of how much he has had to throw away before selling the remainder, although comments to an outside observer may not necessarily be accurate. In a regular marketing channel, such as the distribution of wet fish in ice from the landing site to market in a distant city, the marketing system will be finely adjusted to minimize losses. Traders are not likely to sustain losses of say 50% on a regular basis, unless the price of the remaining 50% is high enough to compensate for the loss. Usually if losses would be substantial those concerned will improve the handling procedures, or opt for a nearer, easier market.

8.7 Spoilage of wet fish is often due to adverse weather conditions or other emergencies. If smoking is the only available means of preservation, and the kilns or barbeques cannot be used in heavy rain, there may be no alternative to letting the fish spoil. Again, floods may wash away roads, preventing fish being taken to market, or preventing the wood needed for smoking to be brought to the processing sites. All these situations are likely to occur irregularly, and affect specific localities, making it even harder to judge what the average level of losses might be.

9. A COMPREHENSIVE PROCEDURE FOR ASSESSING LOSSES

9.1 If an accurate assessment of losses is to be made, information must be obtained or collated on:

1. The fish resources: the main species landed, the quantities involved, and seasonal variations

2. For cured fish, the types of process in use and the amount of fish processed

3. For wet or cured fish, the nature of the distribution system: where the fish is sent, and by what transport systems; who is involved in distribution and marketing; how much fish is involved

4. Packing systems

5. Storage facilities

6. Acceptability of the wet or cured fish at various product levels; at what level of quality does the fish become unacceptable to particular consumer groups.

9.2 This is likely to involve a preliminary study to collect information on these aspects. Often it will be necessary to determine what the distribution system is and what markets are supplied.

9.3 Then the study plan needs to be decided. What products are to be considered: wet fish and/or which types of loss: material and/or value? cured fish, and which types of the latter. How many geographical locations are to be involved: if the number is small then the results may not be typical of the region or country as a whole, but increasing the number of sites adds greatly to the work involved.

9.4 There are also variations with time. There will almost certainly be seasonal variations, as handling and processing will usually be more difficult in wet seasons. One year's observations are not sufficient as there can be great differences between one year and the next. To understand what losses are, the studies should be continued for at least three years.

9.5 Only then can losses be measured. As explained by Wood (1985) this involves assessing quality and measuring quantities at each stage in the processing or handling. An illustration of the system is shown in Table 3. If losses in value are being considered, a careful evaluation must also be made of the prices and costs at each stage in the system.

10. DO WE NEED TO KNOW?

10.1 Having shown how much is involved by accurately assessing losses, the above question must be asked. In a sense we do not need to know what losses are. The important thing, surely, is to reduce or avoid them, so as to increase the income of those involved in fish production and to provide more food for human consumption. A massive three-year study of losses over a wide geographical area will not in itself provide anyone with more food. The pressure to establish what losses are arises for two reasons.

10.2 The first is the simple, procedural need to convince outsiders that there is a serious problem which needs to be tackled. This applies equally to scientists in national institutes, seeking funds from their own governments, and those who work for aid agencies. It is difficult to obtain funds if we can only assert that there are serious losses: potential donors will wish to know the extent of these losses. We can say that losses are X million tonnes or Y million dollars, even if these numbers are only guesses. At the same time, it is seldom easy to acquire funds for a project which aims only at establishing what is happening; we are usually expected to do something about it. For these purposes we have to depend on optimistic generalizations made from “estimates” or a few measurements.

10.3 Further, we need to know what losses occur, and when, if we are to understand what processing systems need to be improved, and in what respect. Careful studies of losses should indicate where improvements are most needed and what should be changed. To assess the effectiveness of any changes base-line data are required. This clarifies and reduces the extent to which losses need to be studied. It is not necessary to know what they are over a whole country or region but only in the specific places where new methods or equipment are being tested. First, are there serious losses which call for attention? Fishermen, processors, traders can tell us although we ought to look at the systems to check these reports. Having identified a system and area where losses are clearly significant, a detailed study to obtain base-line data should be undertaken. One year's observations should suffice, as we can normally check the situation in subsequent years while we are trying to develop improvements. Hopefully these limited base-line studies will provide us a good enough case to obtain funds and approval for action programmes to reduce losses and provide more food for those who most need it.

11. PESTICIDES

11.1 Two more general issues need to be considered in the context of post-harvest losses. One is pesticides, in view of the importance of cured fish in so much of inland Africa, and in view of the widespread incidence or danger of insect infestation.

11.2 Pesticides have a bad image these days. They are poisons: that is why we use them, to poison pests, and are always likely to affect something other than the target as well. It is obviously better to use another means of controlling insect pests, if there is an effective and acceptable alternative. As far as blowflies are concerned, nothing promising has yet been found. Screens will keep blowflies off drying fish, provided the fish can be put behind the screens without blowflies also going there. Even then, blowflies have been observed to deposit their eggs on the screens from which they fall onto the fish beneath.

11.3 Unless an effective alternative is developed, pesticides appear to be a necessity and at that point it is a matter of balancing risks. Everything possible must be done to reduce the risk, by ensuring that if pesticides are used they should be the safer ones rather than the more dangerous. Some governments have tried to ban the use of all pesticides on food, but this is not easy to enforce. Fish processors are not likely to allow blowflies to eat their fish and their income if they can use pesticides when no one is looking. It seems better to encourage them to use only the safer pesticides. For fish and for other foods there is an effective pesticide which has low toxicity to man and leaves no harmful residues when used properly: pirimiphos methyl. The use of this substance on fish has been approved by the FAO/WHO Codex Alimentarius Committee, and it is very effective against blowflies and beetles (research is continuing on the optimum procedures and levels for use on beetles). All efforts must be made to discourage the use of unapproved pesticides which are probably far more dangerous.

11.4 Pesticides present another danger, apart from their deliberate use to control insect infestation. That is when pesticides are used on land, say on agricultural crops, and the pesticides are washed into the rivers, and penetrate the food chain and the fish. Fish technologists can do little to prevent this, but need to ensure that colleagues concerned with food production on land take every possible care to avoid this problem.

12. GENERAL ECONOMIC DEVELOPMENT

The other general issue is what effect is general economic development likely to have on post-harvest losses. In the author's view it is likely to reduce them. Most consumers want fresh fish (or frozen, in which the essential fresh form is retained). Cured fish is an unavoidable but inferior alternative when distribution and marketing systems are such that wet fish cannot reach the consumer. As roads are improved and more ice plants are set up, more fish will be sent to more distant markets. As living standards and incomes rise, more people will be able to pay the extra cost of ice and transportation. This is already happening throughout the world. Iced fish is now being sent long distances in many African countries, such as from lake Victoria to Mombasa. In India only 10 or 15 years ago there was little trade in iced fish; now large quantities of fish of all types in ice are being sent long distances by road and rail.

If this change in emphasis from cured fish to wet fish does happen more attention will have to be given to the losses involved. We are likely to be less concerned about losses of material in cured fish, and more concerned with losses of quality and value in the distribution and marketing of wet fish, less concerned with insect infestation and more concerned with handling systems and procedures, containers, packaging materials, etc. The change should bring more income to the fishing communities. Hopefully, this will happen without depriving low income groups of the cured fish on which they depend for so much of their protein food.

This change may be quite gradual. Cured fish seems likely to remain very important in Africa for a long time yet, especially in inland areas, which have many small landing sites, often with poor communications. Over the last ten years, in Africa as a whole there has been no significant change in the proportion of fish catches which are processed by curing (FAO, 1989). Thus the inherent problems of cured fish will remain: unsatisfactory processing equipment, charring, fragmentation, insect infestation, and the losses in value which they represent.

13. REFERENCES

Ahmed, M. et al. 1978. Radiation disinfestation studies on sun dried fish. Proc. IPEC 18 (3) : 310–21.

Aref, A., A. Timbely and J. Daget. 1965. Fish and fish processing in the Republic of Mali. 3. On the destruction of dried fish by dermestid insects. Alexandria J.Agric.Res., 12 (2) : 95–108.

Bostock, T.W. 1987. Marine fisheries of Gujarat: post-harvest losses and possibilities for development. Report L 75 Tropical Development and Research Institute, London, UK.

Bouaré, F. 1986. Prévention des pertes de poissons fumés (test d'utilisation du tetrachlorvinphos) au Niger. In: Fish Processing in Africa. Proceedings of the FAO Expert Consultation on Fish Technology in Africa, Lusaka. FAO Fish.Rep., (329) Supplement.

Diouf, N. 1980. Trial on protection of dried fish by pyrethrum and solar energy. Paper presented at FAO Expert Consultation on Fish Technology in Africa, Dar-es-Salaam, Tanzania, 11–15 February 1980.

Doe, P.E. et al. 1977. A polythene tent drier for improved sun-drying of fish. Food Technol.Aust., 29(11):437–41.

Duguent, J.S. C. Brou Bleou, Tamboura and Djenepo. 1985. Evaluation of the efficacy of deltamethrin compared with pirimiphos-methyl and tetrachlorvinphos against D. maculatus and N. rufipes on Sarotherodon niloticus at Mopti (Mali). Int.Pest Control, 27 (4) 92–93, 96, 99.

FAO 1989. FAO Yearbook: Fishery Statistics Vol. 65, 1987.

FAO/PNUD 1970. Rapport au Gouvernement du Mali sur les améliorations possibles de l'utilisation des produits de la pêche. Basé sur les travaux de A. Szabo, Expert FAO/AT du traitement du poisson. Rapp. FAO/PNUD (AT), 72 p.

Golop, P., J.R. Cox and K. Kilminster. 1987. Evaluation of insecticide dips as protectants of stored dried fish from dermestid beetle infestation. J.Stored Prod.Res., 23 (1) 47–56.

Guggenheim, H. 1980. Introduction of better processing and packaging methods for dried fish. GCP/RAF/157/DEN Field Report No 2 (Unpubl. Proj. rep.).

Mastaller, M. 1981. Feasibility and evaluation of methods for drying reef fishes. In: Proceedings of Fourth Int.Coral Reef Symposium, Manila 1981. 1, 105–109 ed. Gómez, E.D. Birkeland, C.E. Buddemeises, R.W. Univ. Phil. Manila. ISBN 971-03-00156.

Meynell, P.J. 1978. Reducing blowfly spoilage during sun-drying of fish in Malawi using pyrethrum. Proc. IPEC. 18 (3): 9–12.

Mills, A. 1979. Handling and processing fish on Lake Chad, 55 p. (Unpubl. Proj. rep).

Poulter, R.G., G.R. Ames and N.J. Evans (1989). Post-harvest Losses in Traditionally Processed Fish Products in Less Developed Countries. In: Post-harvest Fishery Losses, ed M.T. Morissey, p 133–145 (University of Rhode Is.).

Rollings, M.J. and L.A.W. Hayward. 1962. Dried fish handling in Nigeria. Food Manuf., 37:498–501

Taylor, R.W.D. and N.J. Evans. 1982. Laboratory evaluation of four insecticides for controlling Dermestes maculatus on smoke-dried fish. Int.Pest Control, 24(2):42–45, 49.

Toury, J., A. Wane, R. Giorgi and J. Cros. 1970. Le poisson dans la ration alimentaire au Sénégal, aspects quantitatifs et qualitatifs, modes de conservation. Food and Nutrition in Africa 8, 7–14 July 1970.

Walker, D.J. and L. Donegan. 1988. Protection of fish from blowfly infestation in wet season in Malawi by dipping in dilute aqueous insecticide preparation before sun drying. Int.Pest Control, 30 (4), 92–94, 98.

Walker, D.J. and N.J. Evans. 1984. Report on two visits to the Gambia to evaluate the use of insecticides in reducing losses in traditionally cured fish. Unpublished report R 1245 (S) of Tropical Development and Research Institute, London, UK.

Watanabe, K. 1971. Improved method of preparing dried sardine from Lake Tanganyika. Rep. Min. Lands Nat. Res., Dept. Fisheries Res. Inst. Zambia.

Wood, C.D. 1983. Losses of traditionally cured fish: a case study. In: Proceedings of the Workshop on the Production and Storage of Dried Fish. University of Agriculture, Serdang, Malaysia, 2–5 November 1982. FAO Fish.Rep., (279) Suppl: 265 p. 172–184.

Wood, C.D. 1985. A methodology for the assessment of losses of cured fish and evaluation of countermeasures. In: Proceedings of the FAO Expert Consultation on Fish Technology in Africa, Lusaka, p. 360–368 (FAO Fish.Rep., (329) Supplement).

Wood, C.D., N.J. Evans and D.J. Walker. 1987. The effect of salt on the susceptibility of dried whiting to attack by dermestid beetles. Trop.Sci., (27):223–228.

Wood, C.D. and D.J. Walker. 1986. Losses of traditionally cured fish in Africa. Paper presented at the FAO Expert Consultation on Fish Technology in Africa, Lusaka. (FAO Fish.Rep., (329) Supplement).

Table 2

Physical loss of cured fish in the tropics

M = Measured;
E = Estimate
* = Laboratory trials at 27°C and 70% relative humidity

Table 3

Loss measurement over two consecutive processing stages