The aim of these tests was: to demonstrate the rigging, operation and performance of several BRDS; to demonstrate catch data collection methods, and to observe the reaction of fish and shrimp to BRDS. All aims were achieved with a high level of success.

As planned the tests/demonstration carried out have demonstrated the feasibility to reduce by-catch by using an appropriate device.

In general, great interest was shown by the observers from the Arab countries, as well as by the participants and the fishery authorities of Iran; expressing the importance they give to by-catch related problems.

The participants agreed that further research is required in the various fisheries to select appropriate by-catch reduction devices and to adjust these according to local situations.

As a matter of fact, in order to rigorously test the effectiveness of these BRDs, further testing and refinement of the testing technique is required. Suggested improvements include standardising shot duration, direction and location, and to account for any difference or bias in catching efficiency between sides, each BRD should be tested on both port and starboard trawls and, obviously increasing the number of shots for each BRD.

The results indicate that all BRDs except the fisheye (because of too large loss of shrimp) were effective in reducing fish bycatch. Experience in Australia and other countries, however, has indicated that a BRD suitable in one fishing location may be unsuitable in the next. Reasons for this include different bycatch species between locations, differences in the relative proportions of each bycatch species, and different seasonal and diurnal behavioural characteristics. Additional testing in all provinces is therefore recommended.

The diameter of the hoop used in the SNM + Cone combination should not be too small compared to the diameter of the base of the cone to limit the space between the two circles leading to the codend, to stimulate more fish to swim backward and, therefore, to increase the effectiveness of the cone. However, care should be taken if large catches are expected/possible to ensure that fouling of the catch does not occur.

For further investigation regarding BRDs’ efficiency and most appropriate models the catch categories should be more detailed. To allow assessment of juvenile catch rates separating commercial fish species into adult/large and juvenile (or retained and discarded) categories is recommended. The use of large and small bycatch categories is useful as the success of many BRDs is based on swimming performance, which is often size related. The use of the above-mentioned four categories: shrimp, commercial fish adult, commercial fish juvenile, non-commercial fish, will eliminate much of the current confusion and allow for a more detailed assessment of BRD performance.

Random subsamples from both codends should be taken for each shot and all species identified, measured and weighed. This information may yield valuable information which can be useful to determine the most successful BRD for a particular species or to give an insight into performance variations between fishing grounds.

The minimum authorized codend mesh size in this fishery is 45 mm, however, it is common knowledge that many trawlers are using mesh sizes as small as 30 mm. It is generally accepted that the exclusion of fish from a trawl occurs mainly in the codend and the use of undersized meshes significantly reduces small and juvenile fish exclusion rates. Fishermen must therefore be prevented from continuing this practice and informed of the benefits of using larger meshes. Rigorous comparative testing of 30 mm and 45 mm is recommended and the results promulgated to industry. The use of fine mesh codend covers also impacts on the exclusion of small animals and juveniles, making their escape almost impossible. The use of covers should either be discouraged or their design modified to improve escape rates.

In order to encourage fishermen to use BRDs an extension program promoting the performance and relative merits of using BRDs is recommended. This could include providing BRDs for testing by fishermen to gain first hand experience in the use and merits of these devices on their own fishing vessel. A similar extension programme is currently being successfully used in Australia and many fishermen are now developing their own BRDs to optimise bycatch exclusion rates.