Protocol for choosing species to be analysed

The products to be analysed as a matter of priority were defined by project partners according to a range of criteria:

• food consumption data (SECODIP, INCA, GIRA and SIC data), 
• production, import and export data,
• requirements drawn up by professionals in the sector,
• the harvesting of existing data from the databases.

Several data sources have been studied by project partners so that the analyses were performed on products which are representative of French consumption:

• the SECODIP panel, which allows us to evaluate the percentage of households which have bought the product at least once during the year under consideration;
• the INCA survey, which provides the average consumption of each product in grams per day according to age category (children or adults)
• GIRA SIC data on non-domestic catering

The SECODIP panel consists of a quantitative record of household purchases produced by IT scanning, over a minimum period of 36 weeks per year. Since purchases are recorded per household, consumption per individual is difficult to estimate using SECODIP. However, it is possible to divide the total quantity bought per household by the number of people which make up the household. But we are then assuming that consumption is evenly distributed between household members (no distinction made between adults and children, for instance). Furthermore, as the SECODIP data are purchase data, they do not take into account non-domestic catering, self-sufficient consumption (products fished by the consumer himself) or items which are not consumed (non-edible parts, edible leftovers left on the plate).

While the SECODIP data do not make it possible to produce a reliable estimation of individual consumption, they do enable us to evaluate precisely the percentage of buyers of a product in a household. This is defined for each product as the percentage of households which have bought the product at least once during the year under consideration. The OFIMER Observatoire Économique (the National Inter-Professional Office for Seafood and Aquaculture’s Economic Monitoring Body) provided this percentage of buyers after analysing the SECODIP data.

 The INCA 99 survey was performed by CREDOC in 1998-99. It is a national individual survey into food consumption. The data presented in the project were communicated by the AFSSA’s (the French Food Safety Agency) Food Consumption Monitoring Body (Observatoire des Consommations Alimentaires). Data on food consumption were obtained from consumption notebooks completed by the consumers themselves over a period of seven consecutive days. All foodstuffs ingested are accounted for, whether eaten at home or outside the home. The quantities entered in the notebooks are the quantities actually consumed. The survey involved 3,003 individuals - both adults and children. National representativeness was ensured by stratification (age, sex, individual socio-professional category and household size). The sample of adults contains 1,985 individuals aged 15 and over. In this survey, children are defined as individuals aged 3 to 14. The INCA survey therefore enables us to distinguish between consumer levels according to age category (children or adult), and above all provides the average consumption of each product in grams per day for each of these two age categories. However, the INCA data do not make it possible to accurately identify all aquatic products consumed: over half of these products consumed are listed as fish without further clarification. Furthermore, when an aquatic product is mentioned, only its common name is used and no mention is made of the scientific name of the species consumed. We thus know about tuna consumption levels, without knowing how these are divided between Thunnus alalunga, Thunnus albacares, Thunnus thynnus, Thunnus obesus and Katsuwonis pelamis. According to the INCA survey, the most consumed fish among adults are, in decreasing order: salmon, tuna, Alaskan Pollock, trout, sole, cod, sardines, hake, mackerel and dab. Among children, the classification of species is similar to that for adults, with salted cod being listed among the ten most eaten fish, however. In terms of shellfish, crustaceans and molluscs, the most eaten foodstuffs among adults are, in decreasing order: prawns, mussels, oysters, squid and crab. Among children, the same products are mentioned but in a different order. Consumer levels obtained by the INCA survey are significantly lower than the percentages of buying households when we compare values species by species. This difference may be due to the fact that a household may be considered as ‘buying’ however many individuals in the household consume the product, while the number of consumers is an individual item of information.  

The 2001 GIRA SIC  data on non-domestic catering (RHF) provided by OFIMER's Economic Monitoring Body are significant sources of information since RHF accounts for ¼ of purchases of aquatic products and since it is possible that the aquatic products involved in RHF are different from those bought by households. The data on non-domestic catering class salmon as the most consumed fish (if we exempt an undifferentiated category: "all white fish"), like the INCA data. On the other hand, they identify fish which are apparently more consumed outside the home. In decreasing order (in tons consumed outside the home), these fish are: trout, sea bass, sole, ray, monkfish and tuna. Sea bass and ray were not among the ten most eaten fish in the INCA survey.

In order to assess the importance of a species to professionals in the sector, we took into account the production, import and export data provided to use by OFIMER's Economic Monitoring Body. Next, the catch data recorded at auctions were taken into consideration in order to evaluate the proportion of producers’ sales accounted for by a species. To determine the most important species, the working group took into account the tonnages imported and the tonnages produced.

On the basis of these production, import and export data, OFIMER’s Economic Monitoring Body used the supply balance method to determine the apparent consumption (in whole equivalent kg) per inhabitant (species coefficients were applied).

Finally, professionals’ requests were studied. They were collected by professional organisations via their members or subscribers. They highlighted the species which the sector would like to see increasing in prestige because of their nutritional characteristics. In general these are species of major commercial importance or species which are difficult to market, and with nutritional composition data which seem to offer opportunities to add value.

The project included a phase in which data was collected from partners and from scientific literature. Indeed, before recommending that one product or another be analysed, it was essential to ensure that the data available for the product had been collected in order to recommend for analysis only those products for which composition data were lacking. The data collected were entered into the CIQUAL database to supplement and improve the data already present. CIQUAL then evaluated the quality of these data on the basis of a set of specifications before providing the project partners with a table of compiled data on the products which interest the working group.

Based on the data from consumption, production, the database and those collected, the species were categorised according to the need to obtain reliable nutritional composition data for them.

Three categories of priority were defined:

1 : high priority

2 : medium priority

3 : low priority