534 



Fishery Bullelm 104(4) 



In New South Wales (NSW) Australia, arrow squid 

 {Nototodarus gouldi), mitre squid (Photololigo sp.) and 

 southern calamari (Sepioteuthis australis) are impor- 

 tant in the legally retained bycatch from fish and prawn 

 trawlers working in oceanic waters. In addition, two 

 species of squid are specifically targeted by up to 20 

 trawlers with modified, single-rigged prawn trawls in 

 Broken Bay (Fig. 1). Broad squid (P. etherldgei) (typi- 

 cally 30-290 mm mantle length [ML] when harvested) 

 represent the majority of the total catch (approximately 

 25-50 t per year since 1997), although the smaller 

 bottle squid (Loliolus noctilitca) (30-90 mm ML when 

 harvested) are also retained and sold, primarily as 

 bait for recreational anglers. Additional details on the 

 local biology of P. etheridgei can be found in O'Donnell 

 (2004). 



The designs of trawls used to catch these squid vary 

 among operators, but all are restricted by lengths of 

 the headline (<11 m) and sweeps (<5 m) and minimal 

 and maximal mesh openings of 40-60 mm in the body 

 and 40-45 mm in the codend. To maximize catches of 

 the smaller bottle squid, all operators use codends of 

 the minimal legal mesh size of 40 mm, with anterior 

 sections of 100 meshes in circumference attached to 

 posterior sections of at least 150 meshes in circumfer- 

 ence (made with thick twine), which are designed to 

 reduce lateral-mesh openings (Broadhurst and Ken- 

 nelly, 1996). 



The configuration of conventional codends in the Bro- 

 ken Bay squid-trawl fishery means that these trawls 

 are poorly selective and therefore, in addition to the 

 targeted squid, they also retain large quantities of by- 

 catch. Preliminary data from a 3-year observer-based 

 study of catches in the early 1990s (DPI-) indicated a 

 bycatch-to-squid ratio of up to 4:1 (by weight). Although 

 a subset of this bycatch includes some species that the 

 fishermen are legally permitted to retain, most of the 

 bycatch is small, unwanted fish, including juveniles of 

 several commercially and recreationally important spe- 

 cies. Concerns over the mortality of these individuals 

 and the negative impacts on stocks led us to test the 

 utility of modifications to fish- and prawn-trawl codends 

 for improving selection of species in squid trawls. Such 

 experimental work has not previously been completed 

 in squid-trawling fisheries. 



A series of explicit hypotheses were tested in the 

 fishery. It was proposed that devices designed to reduce 

 bycatch would 1) cause differential changes in catches, 

 by reducing bycatch but causing no important reduction 

 in catches of squid; 2A) not cause disproportionate by- 

 catch of small or large fish (i.e., would not influence the 

 distributions of sizes offish caught); 2B) not cause dif- 

 ferential effects, so that subsequent incidental catches 

 had similar mixtures of types and relative abundances 

 offish species as before; and, 2C) not cause relatively 



^ DPI (New South Wales Department of Primary 

 Industries). Unpubl. data. 1992. [Data are on file at 

 New South Wales Department of Primary Industries, PO 

 Box 21, Cronulla, 2230, NSW, Australia.] 



Figure 1 



Location of Broken Bay and the two areas trawled iPatonga, 

 and Flint and Steel). Broken Bay is 35 km north of Sydney, 

 New South Wales, Australia. 



larger catches of particular fish species, nor of juveniles 

 of commercial species. 



Materials and methods 



This research was completed on two commercial squid- 

 trawl grounds in Broken Bay NSW (Fig. 1) between 

 February and September 2002 by using a chartered 

 squid trawler (9.2 m in length). The vessel was rigged 

 with a single, conventional 4-seam otter trawl made 

 from 60-mm (all mesh sizes refer to stretched opening) 

 polyethylene (PE) netting (headline of 11 m; Fig. 2) 

 towed between 0.8 and 1.8 m/s over a sandy substratum 

 at depths ranging from 5 to 15 m. A zipper (Buraschi 

 S146R, Milan, Italy, 1.5 m in length) was attached at 

 the posterior body of the trawl to facilitate changing 

 codends (Fig. 2). 



Codends and BRDs examined 



Five codends of different design were made from dark 

 PE netting with total lengths and anterior fishing cir- 

 cumferences (i.e., the expected circumference of the 

 codend during towing — [Broadhurst et al., 1999]) of 

 2.4 and 1.5 m, respectively (Fig. 3, A-E). All codends 

 had zippers (Buraschi S146R, 1.5 m in length) attached 

 at their anterior ends. Four of the codends were made 

 within existing regulations and had anterior sections 

 made from 40-mm knotted netting (2-mm diameter-o 

 twisted twine) with a length and circumference of 35 



