WORKSHOP AGENDA ITEM 5: METHODS TO REDUCE BYCATCH

MOZAMBIQUE (1)

Bycatch reduction in Mozambican prawn trawl fisheries
Domingos Gove, Centro Desenvolvimento Sustentavel, Maputo

Shrimp are one of the main sources of foreign income to Mozambique, and most of the catches are made on the Sofala Bank (Anon., 1979; GTA, 1990). In Mozambique, bycatch comprises about 70 percent of total catch, and, although several species of turtles occur in Mozambique (Hughes 1973; Gove and Magane, 1996), there is nothing known about the impacts of turtles caught in the bycatch of prawn trawling (Magane, Sousa and Pacule, 1998). Studies in the United States in the 1980s, however, showed that prawn trawling was a major source of turtle mortality, and the National Marine Fisheries Service (NMFS) began investigations of turtle excluder devices (TEDs). TEDs were shown to reduce catches of turtles, as well as reducing bycatch by up to 40 percent, without reducing prawn catches. Catch sorting times were also reduced, and hence prawn quality and value were improved. Trawler fuel consumption was also reduced.

TEDs have not been used in Africa, despite the fact that a regional training course on their design and use was held in Malindi in 1997 (Wamukoya and Salm, 1998). There is a tendency for them to be used in the western world, but not elsewhere - possibly because of a lack of information and few demonstrations of their application. An experiment testing the use of TEDs in Mozambique was undertaken in 2001 (Gove et al., 2001). The objectives were to determine the numbers of turtles caught by trawlers, and to test the effect of TEDs on: bycatch (including turtles), prawn catches, catch sorting time and prawn quality. Trawler captains and fishing authorities were interviewed, and onboard testing was conducted in August and November on the southern Sofala Bank. The TED was obtained from the NMFS, but had to be modified to fit in the local gear.

Based on interviews with trawler captains and scaling up by total fleet effort, it was estimated that 1 932-5 436 turtles were caught by trawlers (semi-industrial and indsutrial) per year, mostly in summer. This was considered a substantial source of mortality for turtles in Mozambique. The onboard experiments showed that the TEDs did not reduce prawn catches or bycatch, although catches of large elasmobranchs and rocks were reduced. The catch sorting time was not affected by using a TED, and prawn quality and value was improved.

It was recommended that the report describing the experiment and its results should be distributed to the stakeholders, combined with seminars on the experiment. It was also recommended that legislation be introduced that requires trawlers to use TEDs, and that catches by trawlers with TEDs be certified. Similar studies and follow-up experiments need to be undertaken throughout the region.

MOZAMBIQUE (2)

Bycatch reduction in Mozambican prawn trawl fisheries
Barbara Palha de Sousa, Instituto de Investigacao Pesqueira, Maputo

In 1993, over a 15 day period, experiments with a 0.9 m x 1.5 m aluminium grid in the aft and top section of a shallow prawn trawl were undertaken on the fishing grounds for shallow water shrimps along the coast of Mozambique. For the most abundant species (white and brown prawn), the grid gave a better size selection (more small shrimps were excluded from the catches) compared to an experimental codend with 60 mm mesh size. The best results were obtained with a grid with 14 mm bar distance and when the grid had an angle of attack close to 33 degrees.

The fish bycatch was normally between two and four times the weight of shrimps caught, and the catches contained mainly species with a maximum length of 25-30 cm. As an overall result the exclusion of fish by weight through the 60 mm codend was much higher (83 percent) than for any grid version tested (36.4 percent to 76 percent). The size selection on fish was sharper with grids compared to the 60 mm codend. The results were preliminary due to the short test period and the small number of hauls using 54 mm and the experimental 60 mm codends (Isaksen and Larsen, 1993).

In 1995, following the recommendation from the 1993 grid experiments, further experiments were undertaken with the “top-grid“ system with a 14 mm bar spacing to further understand the effect of the grids. The results showed that 60 mm codends gave very clean catches retaining the biggest tigertooth croaker (Otolithes ruber) and bearded croaker (Johnius dussumieri). With the 54 mm codend, smaller sizes and greater numbers of the same species were retained. As for the 1993 experiments, the grid had a very good separation of Indian pellona (Pellona ditchela) and orangemouth thryssa (Thryssa vitrirostris).

As with the results for the TED experiment descibed by Gove in this report, further testing is required before use of the grids is implemented.

SOUTH AFRICA

Bycatch reduction in South African prawn trawl fisheries
Sean Fennessy, Oceanographic Research Institute, Durban

The discarded bycatch of the prawn trawl fishery mostly consists of small, low value fish species (Fennessy, 1993). Increased utilization of the discards is unlikely because the fishery is economically marginal, and increasing the amount of retained bycatch that is packed would increase packaging costs without adding much value to the landed catch. It is also not known whether increased harvesting of the bycatch would be sustainable and whether there would be ecological or fisheries impacts. Consequently, I do not believe that increased use of the discarded prawn trawl bycatch is likely in South Africa. Instead, I suggest that the current discarded catch should be reduced by means of bycatch reduction devices (BRDs).

During September 2000, I investigated the effects of a square mesh panel on the bycatch of a twin net boom trawler operating on the Tugela Bank shallow water prawn grounds. The panel was inserted in the codend of one trawlnet while the other net served as a control. The objectives of the study were to reduce the bycatch, but without reducing the prawn catch. Catches of prawns, retained bycatch and discards in panel and non-panel trawls were compared. The panel was successful in reducing overall discarded catch quantities, without reducing prawn catches (Table 1). In fact, on 10 occasions, prawn catches in the panel trawl were greater than in the non-panel trawl. Catches of retained fish (mostly O. ruber) were reduced in the panel trawl. Other retained catch categories (crabs, cuttlefish) did not appear to be affected by the insertion of a panel. Differences in catch rates of discarded fishes appeared to be species specific, with catches of J. amblycephalus, O. ruber and P. olivaceum being lower in the panel trawls. For A. thorpei and P. sextarius), differences in catches between panel and non-panel trawls were not apparent.

Table 1: Numbers of Tugela Bank trawls incorporating a square mesh panel, with catch quantities either greater than, less than or equal to a non-panel (standard) trawl.

Catch rates of discards and retained fishes were significantly higher in trawls without a square-mesh panel, while prawn catches by weight were not affected (Figure 1). Mean catch rates of the main discarded fish species were all lower in trawls with a panel and, in the case of three species, significantly so.

It was anticipated that the size distributions of fishes caught in panel and non-panel trawls would be different, because of the potentially greater ability of smaller fish to escape through the panel relative to larger fish. However, this does not appear to be the case, as there was not much difference in size. The panel therefore appeared to permit the escape of all sizes of the most common fishes encountered. This contention is supported by the much reduced catch rates of retained fishes in panel trawls on the Tugela Bank. Most retained fish were O. ruber that, despite being much larger than discarded O. ruber, were still able to escape through the square mesh, largely accounting for reduced catches of the retained fish component in panel trawls.

Figure 1: Comparison of mean catch rates (plus one standard deviation) by panel and non-panel prawn trawls on the Tugela Bank. P = Panel, N = Non-panel; an asterisk^* indicates a significant difference (Wilcoxon paired signed-rank test: P < 0.05).

The results of this work show that the square-mesh panel holds considerable potential for the reduction of fish bycatch in the South African prawn trawl fishery. However, several aspects need further investigation. Firstly, the panel needs to be tested during the main prawn season, i.e. when prawn catches are greatest (April to June), in order to determine whether prawn catches will be compromized beyond the viability of the fishery. Secondly, variances associated with the mean catch rates are high, so greater numbers of replicates are required. Thirdly, the square-mesh panel that was used in this investigation does not allow the release of larger fishes such as P. commersonnii, sharks or rays. The importance of individual species’ behaviour is indicated by the differential catch rates of some fishes relative to others.

Although implementation of legislation requiring South African prawn trawlers to utilize square-mesh panels is feasible, based on these results, it is apparent that additional experimental work is required to optimize the application of bycatch reduction devices. However, such research is also dependent on the cooperation of the industry and the availability of suitable vessels.