Improve Scientific and Technical Advices for Fisheries Management

Shared Ressources
Center for informations
Center for data
Internet Links
Photo Galery
Activity : Collection and management of some specific information

In data collection, one should not simply think that « what can easily be conceived can be easily carried out and obtained ». Thus, obvious information – considering what they mean – as well as essential information – for future processing – such as species to which an individual fish belongs, its size or length or weight can pose important problems as soon as they need to be observed and recorded for high quantities or total number of fishes and in technical conditions that are not really controlled. We must remind us that neither the scientist nor the investigator knows with accuracy the set of landed fishes, but the fisherman or the salesman. 

Two levels of issues seem to emerge around data collection in the monitoring of fisheries.

§         The first level is more general: it is about the designing or modelling of data collection, which is generally neglected, sometimes because of a lack of awareness about its interest but also a lack of representation tools adapted to this use.

§         The second level is broader and corresponds to problems encountered on the types of information, particularly sought, but which acquisition in routine is difficult for some reasons.  

These two levels will be addressed in this paper note.

1. Designing the data collection process: give precision on the observation model to eliminate the inconsistencies and improve the feasibility of surveys.

Data acquisition refers to a protocol of work and a representation of objects of the real world (fishery), that jointly constitute the observation model, which materializes in fine in the form of a survey questionnaire. This model is well worth getting to be explicitly described and justified. Not doing this could lead to ambiguities and difficulties in achieving data collection, processing and interpretation: for instance, when a survey form for landings wants to ask « the selling price of fishes», what are we exactly talking about: the price of fish species contained in the canoe and will be sold after the landing (but how can we know that as these fishes are not sold yet)? Or the price at which we think we can sell them (but, in this case, aren’t we referring to the cost of the transaction for this species as a characteristic of the commercial activity on the beach that day, i.e. another subject of study, with another most appropriate person to talk to). Moreover, if we prefer the first case, it is not realistic to define the price as a descriptor of the species but rather as a descriptor of the transaction (sale) of a set of fishes.

This common example of ambiguities of forms demonstrates how much the lack of designing stage or a modelling stage of data collection can be prejudicial to the feasibility of surveys and the clarity of what collected data means.

2. Particular difficulties encountered on such specific information types.

The wish to collect key information for use in the analysis and interpretation process, especially in the framework of the objectives of dynamic modelling of the resource, is obviously justified. However, this wish must be confronted with costs and feasibility dimensions on the field, and we have to be aware of real obstacles that may arise when one wants to collect these information. In artisanal fisheries, three types of data of major interest raise particular difficulties, according to which we must seek innovative solutions.

2.1. The « species » or « taxon » data. This data consists of designating the specific identity of a fish or a set of fishes (which is landed and measured/weighted). Two difficulties appear then:  

§         The firs one is that in certain cases all the investigators do not have the capacity to recognise the whole set of fishes caught at the level of the species. Some fishes are then identified at high taxonomic levels (genus, family) or under vernacular or commercial categories name.  

§         The other difficulty comes from the fact that in several cases, the biomass of fish at landing and the speed at which the fishes are landed (in containers – 20 or 50 kg boxes – that are not necessarily open on sight), do not enable the identification of all the individuals at the level of the species, whatever competencies the investigator may have.  

Attempts of providing responses to these difficulties have been tested in some countries of the Istam group.

2.2. Data: « size» of individuals, i.e. generally the total length or sometimes the standard length, in cm.

The difficulty comes from the type of statistics (indicators) that scientists want to compute from the collected data:  the structure (distribution) of lengths and the mean of lengths. The production of such statistics assumes that a high number of measures of lengths on a non-biased individual sample is done. The first condition is problematic because only the handling of the fish is allowed, and the handling of a lot of individuals can be rejected by the fisherman or the purchaser of the fish. The second condition (which assumes that individuals are selected on a random basis) is almost impossible to achieve in the real conditions of a survey which is carried out on a landing site. 

We can think about new indicators of length, coarser but more feasible and more robust regarding the risks of sampling bias.

We can also wonder whether it is better to give up the collection of length data on landing sites and only measure fishes in factories and other processing sites. 

2.3. The data « fishing place or area ».

For it is artisanal fisheries, it is difficult to get information on the « area of fishing », i.e.  to know the zone(s) where the fish has effectively been caught – be it by a passive gear or a fishing action. Indeed, canoes do not have recordable GPS and the canoe captains are not compelled (nor a majority of them probably do not wish) to accurately take note of the fishing areas.

We can think about several solutions to alleviate difficulties in collecting spatial information relating to catches: working from a pre-established list of names of zones, collaborate with a group of more broad minded fishermen promising to deteriorate the precision of the information that they will provide, ask for less precise information (distance to the coastline, orientation and journey time from the last fishing action).

3. Suggestion of work within the Istam group.

To concretely deal with the data collection theme,we think that the appropriate way to proceed consists of working from surveys forms and protocols of the countries or case studies, and to study how the items 1, 2.1, 2.2 and 2.3 are addressed in it.

Shared ressources

Persons involved in the case studies
Pierre Labrosse, Biology Aboubacar Sidibe, Biology
Youssouf Hawa Camara, Biology Jerome Guitton, Database Administrator
Ely Beibou, Informatique Cheik Abdallahi Inejih , Biology
Oumar Hamet Wagne, Mohamed Soumah, Informatique
Pierre Morand, Biostatistics Emilie Leblond, Fisheries scientist
Pierre Chavance, Biology

Group email