Although any Senegalese seems familiar with fish balls ("boulettes") and stuffed fish ("farci")[4], neither the origin of the practice of fillet mincing, nor the expansion of the operations in markets or landing sites could be accurately dated and explained. However, it was known that in Saint-Louis region, mullet poached with minced fish fillets were being used in a local dish ("farci") for some generations. It was also acknowledged that fish mincing in public places is not common in rural areas.
Where it is known, the process of mincing involves mainly Sardinella maderensis, S. aurita and Ethmalosa fimbriata, for their relative abundance and their affordability. While low-income groups in urban zones purchase their fish then mince it in the site of purchase, medium income groups usually mince at home themselves. With regard to high-income strata of the population it was said that their minced semi-product is made up of high value (demersal) species, which are also minced at home.
The 3-week study was conducted during the Sardinella glut fishing season according to a planned methodology based on the available information:
The major landing sites of small pelagics which provide 87 percent of sardinella and bonga shad were visited. These are Thiès (Mbour, Joal, Kayar) and Dakar (Dakar, Pikine) regions.
The main fresh fish market in Dakar ("Marché central"), with a daily throughput of 60 to 100 tonnes of fish (50 to 60 percent of sardinella and bonga shad) originating from Thiès region and Hann landing site. It is a place where many women provide fish mincing services. The fish unloaded there is mostly sold on a wholesale basis and dispatched to Dakar secondary markets and processing sites (Pentium Senegal being the largest one) or to hinterland.
Some important secondary markets with fish mincing operations.
Data were collected from minced fish operators, consumers, fisheries inspectors and markets administrators. This was done in collaboration with human and logistical support from DOPM while the Institute of Food Technology (ITA) collaborated in the samples analysis. Accordingly, the study was divided into two parts:
A questionnaire was designed to collect the information regarding the social role of fish mincing business, its profitability, the extent of the activity, the safety and hygienic conditions under which fish handling and mincing operations were taking place, the marketing channels, the use of the semi-product, and the potentials for its expansion.
Table 2 is a breakdown of the several people met. It takes into account some practical constraints and field realities, which to some extent restricted the number of interviewees. However, measures were taken to interview a sample representing at least 10 percent of people operating in a given market during the survey.
Table 2: Outlook of the sites visited.
Site |
Number of |
Number of |
Number of |
Number of |
Interview of |
DAKAR CITY |
|||||
Central market |
36 |
2 |
3 |
59 |
Yes |
Hann artisanal landing site |
11 |
2 |
2 |
57 |
Yes |
Cambérène market |
43* |
3 |
2 |
45 |
Yes |
Tilène market |
27 |
3 |
5 |
27 |
No |
Pikine market[5] |
32 |
3 |
2 |
28 |
Yes |
Castor market |
24 |
3 |
2 |
21 |
No |
TOTAL |
173 |
17 |
16 |
237 |
|
MBOUR CITY |
|||||
Landing site |
71 |
3 |
3 |
NF |
Yes |
Market |
4 |
0 |
00 |
NF |
|
TOTAL |
75 |
3 |
3 |
- |
|
JOAL CITY |
|||||
Landing site |
00 |
01** |
00 |
8 |
No |
Big market |
4 |
2 |
00 |
4 |
|
Small market |
5 |
1 |
00 |
5 |
|
TOTAL |
9 |
4 |
00 |
17 |
|
KAYAR CITY |
|||||
Landing site |
00 |
00 |
00 |
16 |
Yes |
Market |
00 |
00 |
00 |
00 |
|
TOTAL |
00 |
00 |
00 |
16 |
|
OVERALL TOTAL |
257 |
24 |
19 |
303 |
|
* 22 among them were combining the fish cleaning and mincing activity.
** A former fish mincing operator met incidentally at this site informed the survey team that they were 3 women there some 8 to 10 months ago, but their mincers were no longer operational.
NF: Not found on that day.
When a response from direct interview seemed not realistic, in order to avoid biased answers, a second visit was paid to cross-check the relevant information. During this validation phase, which was practically feasible in Dakar, other operators in the same site were taken into account. This was found useful for data linked to the daily quantity of raw material minced and for the income earned. The team noted that the best way to collect reliable data on these figures was to stay by the operators stall for at least an hour, to count the customers, assess the weight of fish brought, the quantity of waste dumped, and estimate the money collected during that period of time. The quantity of fish was determined based of the teams field experience. Depending on the size, a heap of 3 pieces of sardinella or bonga shad generally weights between 650 g to 1 kg.
The team seized the opportunity of this exercise to evaluate the sanitary conditions within the stalls and the working practices.
The sampling was conducted after the first round of interviews, or during the validation of some data (in the case of Dakar region).
Uniform samples of sardinella or bonga shad purchased by the survey team were minced at different sites under the prevailing preparation conditions, and then analysed for their composition, their freshness, and their microbiological status. Owing to weaknesses in the diverse schemes of fish handling and mincing recorded, which involved possible risk of cross-contamination, the trials consisted in sampling minced fish after a thorough cleaning of the mincer, then a second sample was taken after the turn of at least six customers. Some stability tests were conducted to assess the shelf-life at ambient and low temperatures. Some samples of plastic bags used as packaging material were also taken to assess the safety (essentially microbiological evaluation). A sample processed at ITA served as reference.
As per the data in Table 2, no fish mincing operation was reported in Kayar, and it was a very limited practice in Joal, despite the importance of small pelagic landings at these places. These are relatively smaller cities compared to Dakar and Mbour, the latter being the city where this business attracts the highest number of women in one landing site. The field investigations confirmed the indication of the fisheries officers that minced fish fillets production is confined to urban areas. However, any Senegalese knows fish mincing. Fish flesh is traditionally pounded to obtain a paste that has the same intended use as the actual minced fillets. Therefore the fact that no fish mincing operators were found in rural markets and landing sites does not mean that the practice is unknown.
Contrary to information provided by market administrators and fisheries officers the information collected in the field indicated that: (i) fish mincing is a much wider activity than initially presented to the survey team, (ii) its importance has little linkage with the volume of landings but rather on the size of the site and its comparative socio-economic status.
Table 3: Estimated quantity of raw material ("fresh" fish) minced in the sites[6] visited.
Site |
Quantity |
Dakar Central market |
324-2 880 |
Hann artisanal landing site |
22-572 |
Camberene market |
645-2 785 |
Tilene market |
810-1 080 |
Pikine market |
768-960 |
Castor market |
240-1 248 |
Average for Dakar |
468-1 588 |
Mbour landing site |
2 250-4 350 |
Joal bigger market |
12-100 |
Joal smaller market |
55-265 |
Average for Joal |
34-183 |
On average 1.5 tonnes of fresh Sardinella or Ethmalosa is minced daily per site visited (Table 3). Lower figures reported were for the lean fishing season. The survey team noted however that during the period of the study, supposed to be a sardinella season, some sites were facing a critical problem of supply in raw material while others were oversupplied, with obvious implications for prices and physical losses.
The data in Dakar Central Market indicate that the 36 operators daily mince an average of 1.7 tonnes of fresh sardinella or bonga shad. For ease of reference, this figure was compared with that of Pentium Senegal, the largest fish processing centre in Dakar. This site produces 80 percent of fish processed in Dakar. It supplies the citys secondary markets as well as hinterland and neighbouring countries. About 3 to 4 tonnes of raw material is used daily by 220 artisanal fish operators (173 women processors and 43 male fish cleaners) working at the place. This gives an idea of the actual extent and dynamics of the fish mincing operations.
Photo 1: A family-owned fish mincing unit at work.
Except from Tilène market where only men, and in Hann landing site where few men operate, fish fillets mincing can be described as an exclusive women activity. They represented 89 percent of the minced fish operators counted.
At least half of the minced fish operators interviewed were either former fresh fish sellers, or have been combining the two activities together with fish cleaning. Fish mincing is a paid service for the owner of the fish: the mincers do not produce minced fish for sale. It is a non risky and profitable business compared to the sale of fish. They were therefore not ready to go into the sale of readily minced fish fillets, unless some cooling facilities are provided to them to preserve the minced product (reducing thus the risk factor) and they have access to the raw material at a wholesale price.
The report of the presence of men among the operators, and the fact that no operator sells minced fish, but provided mincing service against payment were additional information brought to the attention of fisheries officers and market administrators.
Table 4: Characteristics of the minced fish producers.
Place |
Number of operators |
Number of women |
Age |
House-hold size |
Working experience (month) |
Working hours per day |
Working days per week |
Dakar |
173 |
143 |
35 |
9.8 |
39 |
8 |
6.4 |
Mbour |
75 |
75 |
28 |
7.3 |
23 |
6 |
6.5 |
Joal |
9 |
9 |
30 |
9 |
30 |
6 |
6.7 |
Average/total |
257 |
227 |
31 |
8.7 |
31 |
6.6 |
6.5 |
Table 4 shows that fish mincing is a full-time activity for the operators involved. They work at least 6 days a week, for an average of 7 hours per day (from 6 to 12 a.m.). They appreciated their job and deemed it has fewer constraints, compared to house helps who work longer and do not have time to take care of their family. This is a huge advantage when one knows that in addition to their involvement in income generating activity, these women are responsible for household tasks, including feeding, clothing the kids and supporting husband and relatives. Apart from the youngest among them who lived with their mother, most of the operators had to prepare the meals and take care of the kids after closing from work.
It was reported that fish mincing is being done in markets and landing sites for more than 10 years. However its expansion is a recent phenomenon: about 60 percent of the interviewees had been involved in the business three years or less.
The operators were between 28 and 52 year old (average 31) and represent typical lower income families in urban areas of up to 15 persons. The household size of disadvantaged strata in Dakar is between 9.5 and 11.9. These data give an indication of the number of people directly "fed" by the fish mincing business.
Photo 2: An operator gutting fish.
Diagram 1: Common processing schemes of minced sardinella fillets. |
Diagram 2: Processing schemes of minced sardinella fillets in a particular place. |
Customers of minced sardinella or bonga shad usually purchase themselves the necessary quantity of fish then pay for the service of an operator to mince it on the same site. The process can be described into two main schemes (diagrams 1 and 2) depending on the relative safety of the fish preparation method. There are some common points to the two diagrams and some peculiarities that indicating potential sources of hazards:
- As seen in pictures 2 to 4 representing some production steps, wearing jewels during the operation is a common practice. So is handling fish without washing hands after shaking peoples hands or after holding a child or vegetables and spices.
- It was noted that fish collected from the customer was seldom washed before cleaning started. Yet washing the fish is the best way to get rid of 90 percent of the bacteria at the surface of the fish and to decrease the bacterial load during the subsequent handling steps. Washing fish before cleaning was applied only by few women out of the 13 at Castor market. These women used water at least three times (steps 1, 4, 8 and 9 in diagram 2), and this made the difference in the flow and nature (social status) of the customers observed during the survey, though the service was much more expensive than the women operating differently.
The existence of a second group of women fish mincers was a contrast in the hygienic status of fish handling practices and working conditions in that particular market of Dakar. Based on the fisheries officers briefing and the teams field observation on the predominance of demersal fish, Castor market is located in a less populous district. Groups of women fish cleaners, sellers and mincers have put their stalls just outside the entrance of the market fish on a very dense situation. They were operating under the worst conditions seen during the whole study. It was said that these women were on a temporary site, they paid daily taxes but did not have access to any basic service.
Photo 3: Removal of a skin on a butterfly fillet.
Photo 4: Removal of a backbone from a skin-on butterfly fillet
They operated all according to diagram 1 and charged less than the group established within the premises of the market (CFAF25 against CFAF50 to 75 per heap of 3 fish). The longest queue of customers in this market was with three women who applied comparatively better hygienic working practices, though their service was more expensive:
- The repeated use of the same water to wash the fish from several consumers was a common point to all the operators counted. Water was changed depending on the operators visual appreciation of its dirtiness, and this can vary from 4 to 5 customers packages (average estimate of 6 to 8 kg of fish) in the best case (Diagram 2), to 5 to 8 customers (8 to 15 kg of fish) in the case of Diagram 1.
- Apart from the repeated use of water for many customers, it was noted that some operators did not use potable water at all. Some market sites were not supplied with tap water. This led to the use of lower quality water from wells.
- The mincer was used several times without being washed. Less than 2 percent of the workers counted washed the mincer between customers. These ones were in the group of operators within Castor market. The fragility of the apparatus requires that it be dismantled from time to time to get rid of the scraps of fish that often block it. It was noted during the survey that the apparatus stopped functioning after an average of 8 to 10 kg of fish. While in most cases the scraps of fish were simply removed with a stick or with fingers, some of the women in Diagram 2 seized the opportunity of dismantling to wash their mincer. Cleaning the mincer between customers would not be a critical matter if fish were thoroughly washed with potable water when taken from the hands of the customer.
- The sanitary conditions of the fish handling and preparation environment are poor: the floor is often sandy, dirty, or of damaged cement or tiles; lack of drainage facilities for waste water; irregular and inadequate removal of solid wastes, absence of toilet facilities and potable water at some places. The operators were aware of these conditions and at least two thirds among them unanimously stated that they would have attracted much more customers and boosted their business if hygiene within the places was improved.
Photo 5: Last washing of skinless single fillets (Diagram 2 operator).
Photo 6: Skin-on butterfly fillets ready to be minced.
It is clear that improvements to be brought in the fish mincing operations require that basic facilities and services be provided. The responsibility of the market or landing sites administrators is involved, as they collect regular taxes and several levies from the operators. It is fundamental that appropriate action be taken accordingly:
- The packaging material is a plastic bag either brought by the customer or given to him/her for free. Plastic bags are commonly sold in markets and landing sites and are meant for multiple purposes, including packing vegetables, other foodstuffs, clothes, shoes, etc. As it was not particularly designed for packaging prepared food, its sterility may not be guaranteed for packaging minced fish, known as relatively sensitive product compared to whole fish.
- The results from ITA confirmed the inadequate working practices and environmental conditions. The load before cleaning the mincer illustrates that customers in the earlier periods of the operation have safer product than the following ones (contamination increases with the number of uses of the mincer.).
The presence of Clostridium was noted in a sample in one location. This observation was made in empty plastic bags (without minced fillets) from the same location, which had seven counts of Clostridium. The bacteria in that sample could therefore come from the first clients products or from the plastic bag. In fact, the analytical findings on these bags were that before being used, they already contained substantial bacteria. With exception of the absence of Salmonella, they had some bacterial counts lower than samples in diagram 1 and diagram 2, but higher than the control sample.
The control sample from ITA served to establish storage time. Based on the microbial status, a freshly prepared minced fillet of good quality (from sardinella fish grade 6/7 in Torry quotation) and handled hygienically must be cooked within 12 hours if kept at ambient temperature, but within 36 hours in the fridge (4 ºC). The data are of interest for housewives as there is usually a delay between mincing the product and cooking. This can be particularly long in households where only one meal is taken a day, which is prepared in the afternoon, while fish is minced generally early in the morning.
These analyses, which were the first on this type of products, constitute a good starting point as guide for customers, operators and extension officers. They would however need to be replicated and combined with histamine analyses, including testing samples of various degrees of freshness.
Table 5: Microbiological assessment of samples of minced sardinella fillets.
Bacteria (UFC/g) |
Diagram 1 |
Diagram 2 |
Control |
||
Before |
After |
Before |
After |
||
Total bacteria count |
3.4x 105 |
7.2 × 104 |
1.2 × 104 |
104 |
1.1 × 103 |
Total coliforms |
2.6 × 104 |
8 × 102 |
2 × 104 |
4 × 101 |
3.3 × 102 |
Faecal coliforms |
1.3 × 104 |
3.6 × 102 |
< 10 |
101 |
1.2 × 102 |
Pathogenic staphylococcus |
< 100 |
< 100 |
< 100 |
< 100 |
<100 |
Clostridium |
0 |
0 |
3 |
0 |
0 |
Salmonella (UFC/25g) |
Absence |
Absence |
Absence |
Absence |
Absence |
Table 6: Freshness and chemical assessment of samples of minced sardinella fillets.
Sample |
Freshness |
Chemical parameters |
|||
Sensorial |
N-TVB |
Histamine |
Fat % |
Protein % |
|
Control |
9 |
25 |
8.35 |
6.64 |
22.34 |
Diagram 1 |
6 |
25 |
16.54 |
16.37 |
20.22 |
Diagram 2 |
7 |
20 |
<Detection limit |
7.84 |
20.34 |
Photo 7: Mincing operation.
Photo 8: Minced sardinella fillets being packaged.
Photo 9: Mincer being dismantled for cleaning and washing.
Photo 10: dismantled parts of a mincer ready for cleaning.
The basic equipment used for fish mincing is represented in Table 7. The variation in the price of dustbins and knives lies in their size or quality (strength).
Table 7: Inventory of the equipment used for fish mincing.
Type |
Cost per unit |
Number per |
Useful life |
Mincer |
7 500-9 000 |
1 |
5-6 |
Wooden table |
7 000 |
1 |
12-18 |
Wooden plank |
Salvage equipment |
0-1 |
2 |
Wooden bench |
1 200 |
1 |
12 |
Wooden stool |
400-500 |
1-2 |
10 |
Plastic pan (40 l) |
1 000 |
1-2 |
5-6 |
Plastic bucket (10 l) |
500 |
2 |
5 |
Plastic bin |
700-1 000 |
1 |
2-3 |
Plastic strainer |
400 |
1 |
3 |
Other container (1-2 l) |
400 |
0-3 |
3 |
Knife |
300-500 |
1-2 |
2-3 |
Depending on the site or the working habit of the operator, a table or stool is used to support the mincer, as shown in pictures 1, 7 and 9. When a table or a concrete cement block is used as a display, a bench served as a seat. A chopping plank is usually placed on top of the cement block to immobilize the mincer. Using two stools is a common practice in Mbour landing seat and Dakar central market, one serving to immobilize the mincer and another one as a sit. On the other hand a table or a concrete block, then a bench as a seat, were the main tools used in all markets premises and in Hann landing site.
The buckets and basins were generally used to store water, and to clean and wash the fish. It was noted that the more tools the worker has, the longer the lifespan of each, and that women who had a processing scheme similar to diagram 2 had generally more containers.
The mincer is an imported machine with a bowl of approximately 300 g-capacity. All the operators had one unit for fish mincing; some may own a second one to mince vegetables, leaves and spices.
The cost of a mincer varies depending on the place of purchase or the bargaining skill of the operator. The interviews reported an average cost of CFAF8 000. An overwhelming majority of the respondents mentioned that the mincers lifespan varied from 5 to 6 months (up to one year). This short lifespan was the critical concern of the persons met.
Among issues raised by the operators the mincers weaknesses were the first preoccupation (Table 8).
Table 8: Problems faced by the fish mincing operators.
Problem |
Number of respondents |
|||
Dakar |
Mbour |
Joal |
Total |
|
Fragility of the mincer |
9 |
00 |
2 |
11 |
Capacity of the mincer |
6 |
00 |
3 |
9 |
Wound when operating the mincer |
3 |
00 |
1 |
4 |
Wound when removing fishs spine |
7 |
00 |
2 |
9 |
Working environment (space, table, hygiene) |
5 |
3 |
00 |
8 |
Supply in/cost of potable water |
2 |
00 |
02 |
4 |
High levies and taxes |
1 |
00 |
00 |
1 |
Customers bargaining the cost of service |
1 |
00 |
00 |
1 |
Total |
34 |
3 |
10 |
47 |
It can be noted from Table 8 that 51 percent of the problems are closely linked to the mincer. The short lifespan of the apparatus, which is the direct consequence of its fragility, combined with the lack of spare parts on the local market were very disconcerting facts for the operators. The justification of these shortcomings could be the intensive use of the machine, which is not actually designed for "semi-industrial" purpose but rather for household use.
This calls for an improved design, which would be stronger, more durable, safer, and have appropriate capacity. The actual capacity needed would be 2 to 3 times more as the current bowl cannot hold more than two single fillets at once.
The tediousness of the process of deboning with the fingers, in particular for good quality fish (firm vertebral column), was the second concern raised by the operators. There is hence a need to use a manual backbone remover to prevent any wound that represents a potential source of contamination of the product.
A local manufacturer was met to discuss the feasibility of addressing these two types of weaknesses. The problem raised by the manager was that unless he was assured of the funding of the pilot assessment and potential markets, he would not provide to the survey team broad figures of the retail cost of improved equipment. He thought however that it could not be time consuming to design this prototype and test it, and is ready to concur if any institution would meet his conditions.
Other relevant operational deficiencies raised by the operators were the bad working environment and the lack of potable water.
As previously mentioned, fish fillets are minced on request from customers who bring their fish to the operator. Four types of customers were identified, based on difference in the quantity of fish purchased and the intended use of the semi-product:
- Housewives: the family size of the customers interviewed is 11 (between 5 to 20), this is the size of a typical low-income family. A quantity of 1 to 2 kg of fresh fish is minced per purchaser on a daily basis or at least three times a week. sardinella is the predominant species, but some women also use bonga shad (Ethmalosa fimbriata), horse mackerel (Trachurus sp.), Spanish mackerel (Scomberomorus sp.) or skipjack tuna (Katsuwonus pelamis). Women who supplied food to groups of purse seine fishermen are also in this category of purchasers.
- Although they may have a fish mincer at home, a number of women belonging to the medium income stratum minced their fish in markets and landing sites in order to save time once at home to prepare the meals. They were seen in most instances at the stalls of operators working according to diagram 2. Their frequency of purchase is less than the low-income groups (once to twice a week).
- Restaurants and chop bar owners: they preferred sardinella species. They manage low standard restaurants with an estimated 15 to 20 daily consumers of fish balls. Depending on the consumers affluence in the restaurant they purchased and minced 4 to 8 kg of raw material 4 to 5 times a week.
- Traders of ready-to-eat minced fish products at the entrance of schools and colleges. They have the same purchase frequency and quantity as the owners of restaurants and chop bars, but always purchase whole Spanish mackerel in addition to minced sardinella fillets.
- Traders of packaged minced fillets. Usually on a daily basis, they purchase 5 to 10 kg of sardinella that they mince, package and cover with ice. It is then sold in apartment districts ("quartiers HLM"). They have been reported only in Dakar central market.
The majority of the customers interviewed insisted that fish should be minced in their presence, but in fact many left the fish with the operator to complete ones shopping in the market. The purchasers thus ignored the conditions under which their product had been prepared, and whether the packaged product was actually made from the fish they left. This was one of the contradictions reported during the survey.
The cost of mincing varied generally from CFAF10 per fish to CFAF25 per heap of 2-3 pieces, depending on the size of the fish. The exception was the pricing in the stalls inside Castor market where it varied from CFA F50 to 75 per heap of 3 pieces. Some cases of bargaining about the charges for the service have been noted.
Diagram 3: Utilization schemes and final products of minced sardinella. |
Six types of final products were identified from the minced fish fillets. The most frequent are fish balls in tomato sauce, served with a local dish made up of fried rice (Thiebou Dienn). Diagram 3 presents the utilization scheme and the several forms of end products.
Other end products reported by medium income women, but not represented on the diagram were fish balls as attractive presentations for salads, and two types of entrée one made up of stuffed eggs and another one with spiced fried minced fillets wrapping egg.
Though none of the customers interviewed had mentioned the existence of sandwich in schools and colleges, the possibility of the sale of this ready-to-eat product should not be underestimated. In effect the sandwich prepared in family served for breakfast, and also as snacks to be taken away by school children.
All these products are probably interesting market niches to be explored.
[4] According to people met,
"farci" is a bigger size fish balls. [5] Zinc 1 Market. [6] Quantity based on the adjusted information of the interviewees then multiplied by the number of operators reported on the site. [7] According to Torry scoring table. |