Broadhurst et al Effects of square-mesh panels in codends and of haulback delay on bycatch reduction 



413 



One suite of modifications to trawls that has been 

 successful throughout the world in reducing 

 bycatches while retaining the target catch involves 

 the use of square-mesh panels in codends ( Robertson 

 and Stewart, 1988; Carr, 1989; Briggs, 1992; 

 Broadhurst and Kennelly, 1994; Robertson 2 ; Isaksen 

 and Valdemarsen 3 ; Suuronen 4 ). Quantifying the ef- 

 fectiveness of square-mesh panels in codends has 

 been approached by a variety of experimental meth- 

 ods. At the simplest level, catches and bycatches from 

 control and modified codends have been compared 

 by using alternate tows (Robertson and Stewart, 

 1988; Broadhurst and Kennelly, 1994), by towing both 

 codends as twin gear (Briggs, 1992), by using trou- 

 ser trawls (Walsh et al., 1992; Suuronen 4 ), or by tow- 

 ing either gear from two adjacent vessels (Thor- 

 steinsson, 1992). Alternatively, covers of small-mesh 

 netting over modified codends have provided esti- 

 mates of the quantities offish escaping (Robertson 

 and Stewart, 1988). 



Although these kinds of experiments have deter- 

 mined the effectiveness of square-mesh panels in 

 codends, without direct observations by divers or 

 cameras there is no information available on the 

 patterns of behavior of escaping fish or on the period 

 when escapement occurs during tows. The latter 

 point raises an important issue with respect to the 

 survival rates of escaped catch. Because fish are ex- 

 hausted to varying degrees owing to the stress asso- 

 ciated with capture (see Wardle, 1983; Main and 

 Sangster 5 ), the ultimate efficiency of square-mesh 

 panels in reducing bycatch mortality is determined 

 by the rate at which nontarget individuals are ex- 

 cluded. The survival of these individuals is probably 

 inversely proportional to the time spent in the trawl, 

 because studies have shown that fish that remain in 

 the codend are subjected to high levels of stress and 

 fatigue (Main and Sangster 5 ; Main and Sangster 6 ). 



Observations from divers, from video cameras, and 

 from towed submersibles, have provided conflicting 

 information on the behavior of escaping fish (Wardle, 

 1983, 1989; Watson, 1989; Briggs, 1992; Watson et 

 al., 1993). For example, in the Irish Sea, Briggs ( 1992) 

 observed that the escape of European whiting, 

 Merlangius merlangus, occurred continually through- 

 out the duration of the tow. In contrast, Watson et 

 al's (1993) studies in the Gulf of Mexico showed that 

 red snapper, Lutjanus campechanus, maintained 

 their position within the codend throughout the tow 

 and escaped only when haulback was initiated. 



In the NSW oceanic prawn-trawl fishery there is 

 commonly a delay of up to 15 seconds (s) between 

 slowing the vessel and engaging the winch to haul 

 in the trawl (termed haulback delay). Because of this 

 delay, possible effects, such as those observed by 

 Watson et al. ( 1993), need to be investigated prior to 

 interpreting the effectiveness of square-mesh pan- 

 els in codends. Direct observations of such effects, 

 however, always assume that the divers or cameras 

 have no artificial effects on the behavior of the fish 

 or on the normal commercial operation of the gear. 

 While it may be reasonable to assume minimal ef- 

 fects in conditions with adequate light, Wardle ( 1989) 

 found that the use of cameras with artificial light at 

 night or in turbid conditions (below 10" 3 lux) disori- 

 entated fish and disturbed their behavior. The pos- 

 sibility of such effects makes this technique unsuit- 

 able for assessing the effects of haulback delay on 

 square-mesh panels in the NSW oceanic prawn-trawl 

 fishery which operates at night. 



Our specific goals in this experiment were 1) to 

 investigate the effectiveness of square-mesh panels 

 in reducing the bycatch from the NSW oceanic prawn- 

 trawl fishery and 2) via a manipulative experiment, 

 to provide an accurate and inexpensive means to de- 

 termine the effects of haulback delay on the perfor- 

 mance of these designs. 



2 Robertson. J. H. B. 1983. Square mesh cod-end selectivity 

 experiments on whiting tMerlangius merlangus (L.l) and had- 

 dock ( Melanogrammus aeglefin us (L.l). Int. Counc. Explor. Sea, 

 council meeting 1983/B:25, 13 p. [Mimeo.] 



3 Isaksen, B., and J. W. Valdemarsen. 1986. Selectivity experi- 

 ments with square mesh codends in bottom trawl. Int. Coun. 

 Explor. Sea, council meeting 1986/B:28, 18 p. [Mimeo] 



4 Suuronen, P. 1990. Preliminary trials with a square mesh 

 codend in herring trawls. Int. Counc. Explor. Sea, council meet- 

 ing 1990/B:28, 14 p. [Mimeo.] 



5 Main. J., and G. I. Sangster. 1988. Scale damage and sur- 

 vival of young gadoid fish escaping from the cod-end of a dem- 

 ersal trawl. In J. DeAlteris (ed.), Proceedings of selectivity and 

 survivability workshop, p. 17-34. Univ. Rhode Island Sea 

 Grant Advisory Service, Narragansett, RI. 



6 Main, J., and G. I. Sangster. 1991. A study of haddock 

 iMerlanogrammus aeglefinus (L.)l behavior in diamond and 

 square-mesh codends. Scot. Fish. Work. Paper 19/91, 25 p. 



Materials and methods 



Two experiments were conducted on commercial 

 prawn-trawl grounds east of Yamba, New South 

 Wales (29°26'S, 153°22'E), between August and Oc- 

 tober 1994 by using a commercial prawn trawler ( 13.8 

 m ). Three florida flyers ( mesh size 42 mm ), each with 

 a headline length of 12.8 m, were rigged in a stan- 

 dard triple gear configuration (see Andrew et al., 

 1991, for details) and towed at 2.5 knots. Each of the 

 identical outside nets were rigged with zippers (no. 

 10 nylon open-ended auto lock plastic slides) to fa- 

 cilitate changing of the codends (see Broadhurst et 

 al., in press). Because the middle net was not rigged 



