The phenomenon commonly called the selectivity of the fishing gears occurs when a fishing gear such as trawl catches only fish of a certain sizes. Such an explanation is given in the study entitled "The stock dynamics" (Andrzej Kompowski , Jan Horbowy ; MIR 1990).

Fish, depending on age or length are susceptible or not susceptible to certain fishing gear.

Prof. Joseph Świniarski in his publication entitled "Technology of the catch of marine organisms (Wydawnictwo Morskie , Gdańsk 1993) reports that the selectivity properties of fishing trawls relate mainly to the end of the belly and a codend. He also stresses the importance of correct determination of the mesh size in these parts of the trawl. He further reports that the selectivity properties depend on the interaction of technical and biological factors. As a biological factors he lists, among others,
ecological conditions of fish habitat and, associated with these conditions, the reactions of fish to the gear used. The technical factors mentioned by prof. Świniarski are mainly physical and mechanical properties of materials, angle of attack of the trawl, mesh shape and its tension.

Selectivity is often presented with a curve, which usually takes the form of Sigma. The size of fish when they become caught is marked with the symbol Lo. The symbol L100 indicates that we are dealing with the full susceptibility of fish to the gear. But the most important point of the selectivity curve is the value of L50. This is a value that shows us a fish size at which half of all the fish of this size run away from the tool. The slope of the curve refers to the size range of escaping fish and, in the context of discards, generally the goal is to obtain that range as small as possible. Recently, there is much discussion about the selectivity of fishing gear and its
improvement. Usually this term is associated with the size of the mesh and is is reflected by a schematic thinking that the larger mesh is equal to the greater selectivity. A typical example of this approach, we see even in the activities of politicians, who in 2009 made another decision to increase the minimum mesh size to 120 mm ("inside mesh”) in the codends used, inter alia, in cod fishery. But it was rather a decision calculated on the "media" effect, having no justification in the results of scientific research.

Much less attention is given to changes in the construction of fishing gear as a potentially better way to achieve improved selectivity in all its aspects. It is the design of the gear, mainly codends, that provides more opportunities to achieve this goal. The BACOMA and T90 codends used in the Baltic fisheries can serve as an example.
Research and current practice shows that T90 codends, at least in the initial period of their exploitation, are characterized by greater selectivity than BACOMA codends of the same mesh

Unfortunately, these constructions are also not perfect, as evidenced by research and resulting from research the selectivity curves of these tools.
One reason for limiting the escape of undersized fish is the construction of the end part of both T90 and BACOMA codends, where we have a situation when constantnumber of meshes making the codend circumference lead to one place ("codline"), thereby causing the mesh in this place are almost or completely closed. Coefficient of mesh opening (Ux) decreases with decreasing diameter of the bag.

The following picture (picture 1) shows a different shape that take the mesh, depending on where they are located.
Our proposal relates to the T90 codend and the codend with the BACOMA window and shows a new design of its tip in which meshes are completely open and take the shape of a square mesh (T45), such that we have in the BACOMA windows. Generally, the change is relatively simple and, above all, universal, that could also be applied in other codends used in fisheries.

It is worth emphasizing the important information for the fishermen that the proposed change does not require making of new codends and can be done in existing tools.
The change in the construction consists of cutting two wedges across the width of the so-called. "working meshes" at the end of each panel. Their depth depends obviously on the width of the codend panel.

Re-sewing of the above mentioned wedges, which we called the an “envelope” tip, provides an area where all meshes will be turned by 45 ° in relation to the selvage.
Such an arrangement results in getting the so-called square meshes and thus they will be open during the whole process of trawling.

The undersized fish tired of trying to escape, which usually stumble at the end of codend in a closed meshes, due to these changes will have now greater opportunity to escape. Therefore, they will not block the mesh at the end of codend (at least to alesser extent than previously), which in turn creates a greater chance to escape for next undersized individuals.

The new construction will also reduce the so-called effect of "buckets", thus pushing the water in the opposite direction. This will change the hydrodynamic flow in the gear and, therefore, the stumbling of fish in the meshes is expected to be smaller The expected improvement of selectivity of a new gear does not lead to increase of the size of fish escaping, but to obtaining a higher number of escapes of undersized fish and to reduce the range of the selectivity of the fishing gear. The reduction of the selectivity range will help to reduce discards, which at is one of the fundamental aim of CFP.

The change of the construction currently in force for the new one involves changing the way of binding the so-called "codline". The proposal we presented requires further consultation with the fishermen and the use of their practical knowledge of the fishing process.

We hope that it will interest not only fishermen or persons responsible for making and implementing fisheries management instruments, but also among nongovernmental organizations that work with fishermen in the area of the conservation of the seas and their resources.

Krzysztof Stanuch, MSc Eng.
Andrzej Boczkowski, Eng.
Baltic Net Ltd