540 



Fishery Bulletin 104(4) 



designed from an understanding of the behavior of 

 fish and squid were effective in allowing fish to escape 

 and did not significantly reduce the catches of squid. 

 The escape of fish through these BRDs was a direct 

 consequence of 1) their behavior and species-specific 

 swimming abilities in the trawl (Wardle, 1983; Watson, 

 1989) and 2) the location of the BRDs with respect to 

 the influences of water-flow (Broadhurst et al., 2002). 

 More specifically, when fish enter the codend they are 

 often herded together and herding invariably initiates 

 movement towards the top and sides. The density of 

 the school and any species-specific behavioral responses 

 (Watson, 1989) determine the extent of such movement. 

 Panels of square-shaped mesh, if strategically posi- 

 tioned, have been demonstrated to allow actively swim- 

 ming fish to escape the trawl through the large open- 

 ings in the square-shaped mesh (Broadhurst, 2000). An 

 important contributing factor is the effect of anteriorly 

 displaced water in front of the catch, which reduces 

 relative flow and assists small fish to maintain their 

 position in the codend. Broadhurst et al. (2002) dem- 

 onstrated that these effects diminish with increasing 

 distance forward and that the optimal location for BRDs 

 designed on fish behavior is up to 1.2 m anterior to the 

 last row of meshes in the posterior section of the codend 

 (i.e., the position of all BRDs examined in our study). 



The behavior of squid in trawls, codends, or near 

 BRDs has not been examined in detail. Glass et al. 

 (1999), however, observed that the reaction of schooling 

 L. pealeii during initial detection of the gear was simi- 

 lar to fish. Individuals orientated away and then, as a 

 result of compensatory movements in response to shifts 

 in their visual field (termed the optomotor response), 

 attempted to maintain position in the mouth and body 

 of the net. After a short period, all squid were observed 

 to stop swimming, turn, rise in the net, and fall back 

 toward the posterior section of the trawl and toward 

 the codend. Their behavior in this region of the net 

 was not documented, but, given the lack of significant 

 reductions in the catches of squid (or other cephalopods) 

 by prawn trawls containing behaviour-based BRDs (e.g., 

 Broadhurst et al. 2002) and the results observed in our 

 study, active escape is probably limited. 



There was a nonsignificant reduction in the weight of 

 squid from the codends containing BRDs in experiment 

 1, and especially the diamond BRD, but this reduction 

 probably occurred during retrieval of the codend. Typi- 

 cally, once the trawl is winched to the surface, a delay 

 of up to three minutes can occur while the vessel is 

 stopped, the retrieval line is connected, and the codend 

 is brought onboard. During this period, any squid in 

 the anterior codend or extension section may wash for- 

 ward and out of the escape opening in the BRD. Watson 

 (1989) observed similar effects during the retrieval of 

 prawn trawls in the Gulf of Mexico and, more recently. 

 Brewer et al. (1998) highlighted this escape route as be- 

 ing a major cause of loss of prawns from several BRDs 

 in trawls used in the Gulf of Carpentaria, Australia. A 

 simple way of minimizing the potential for such losses 

 when using the diamond BRD would be to keep the 



vessel moving, thus maintaining drag on the codend, 

 during retrieval. A simpler and more practical solu- 

 tion would be to substitute the square-mesh BRD with 

 panels made from 75-mm mesh hung on the bar, since 

 the 75-CSMP and 75-panel codends tested in Experi- 

 ment 3 showed no evidence of loss of squids. Assuming 

 that fish escaping from the CSMP survive the process 

 (Broadhurst et al., 1997), the results presented in this 

 study support the adoption of the CSMP codend in Bro- 

 ken Bay squid trawls. As a consequence of this study, 

 and the co-operative relationship established between 

 the NSW Department of Primary Industries and com- 

 mercial fishermen, a square-mesh panel BRD has been 

 voluntarily adopted throughout the Hawkesbury River 

 squid-trawl fishery. 



Acknowledgments 



This study was funded by the Australian Research 

 Council Linkage Scheme (Project C00106975) and the 

 NSW Department of Primary Industries. The work 

 would not have been possible without the technical 

 expertise and support of Broken Bay squid fishermen 

 Mark Peterson, Carl Blacklidge, Graeme Hillyard, Chris 

 Stapleton, and Rolf Norington. We thank Bob Hunt, 

 Katie O'Donnell, Simon Gartenstein, Andres Grigali- 

 unas, Craig Myers, Sophie Diller, and David Blockley for 

 their technical assistance. Steve Kennelly developed the 

 original proposal for this study and suggested valuable 

 improvements to the manuscript. Constructive com- 

 ments from two anonymous reviewers were also greatly 

 appreciated. 



Literature cited 



Brewer, D., N. Rawlinson, S. Eayrs, and C. Burridge. 



1998. An assessment of bycatch reduction devices in a 

 tropical Australian prawn trawl fishery. Fish. Res. 

 36:195-215. 



Broadhurst. M. K. 



2000. Modifications to reduce bycatch in prawn trawls: 



a review and framework for development. Rev. Fish 



Biol. Fish. 10:27-60. 

 Broadhurst, M. K., and S. J. Kennelly. 



1996. Effects of the circumference of codends and a new 

 design of square-mesh panel in reducing unwanted by- 

 catch in the New South Wales oceanic prawn-trawl fish- 

 ery, Australia. Fish. Res. 27:203-214. 



Broadhurst. M. K., S. J. Kennelly, and D. T. Barker. 



1997. Simulated escape of juvenile sand whiting (S(7- 

 lago ciliata, Cuvier) through square-meshes: effects on 

 scale-loss and survival. Fish. Res. 32:51-60. 



Broadhurst, M. K., S. J. Kennelly, and C. A. Gray. 



2002. Optimal positioning and design of behavioural-type 



bycatch reduction devices (BRDs) involving square-mesh 



panels in penaeid prawn-trawl codends. Mar. Freshw. 



Res. 53:813-823. 



Broadhurst, M. K., R. B. Larsen, S. J. Kennelly, and P. E. 



McShane. 



1999. Use and success of composite square-mesh codends 



