Fisheries Research 72 (2005) 81–95 A bycatch assessment of the gillnet monkfish Lophius gastrophysus fishery off southern Brazil Jose´ Angel Alvarez Perez∗, Roberto Wahrlich Centro de Ciˆencias Tecnol´ogicas da Terra e do Mar-CTTMar, Universidade do Vale do Itaja´ı-UNIVALI, Cx. Postal 360, Itaja´ı SC CEP 88.302-202, Brazil Received 23 June 2003; received in revised form 7 September 2004; accepted 2 October 2004 Abstract Bycatch production in the recently developed monkfish (Lophius gastrophysus) fishery off southern Brazil was assessed from 14 observer trips conducted by a chartered gillnet fleet during 2001. Catch composition and discard ratios were estimated in 523 selected sets conducted between latitudes 22◦44S and 34◦21S and longitudes 33◦37W and 52◦23W and between the 132 and 607 m isobaths. The monkfish constituted on average 40.7% of all organisms entangled in each observed set. Among non-targeted species geryonid crabs (mostly the royal crab, Chaceon ramosae) and spider crabs (Majidae) were the main components representing 22.6% and 8.5% of total entangled organisms, respectively. The remaining 23% included mostly teleosts and elasmobranchs, such as beard fish (Polimixia lowei), silver john dory (Zenopsis conchiffer), gulf hake (Urophycis cirrata), argentine hake (Merluccius hubbsi), angel shark (Squatina argentina) and various skates (Rajidae). Geryonid crabs and angel sharks were the most retained and processed bycatch components (88.8, 50.0% of numerical catches, respectively). Discard ratios were higher than 75% in all other species entangled by gillnets, including abundant or valuable species, such as the wreckfish (Polyprion americanus). Catch rates of large-size bycatch components, such as dolphins, birds, sharks, rays and the wreckfish were poorly correlated with fishing effort and seem to have been mostly affected by the overlap between these species concentrations or higher diversity areas and the southernmost monkfish fishing grounds. Geryonid and majid crabs were correlated with monkfish catches and fishing effort, being most frequently caught north of 29◦S. This pattern was related not only to a target/not-target species distribution area overlap, but also to an elevated vulnerability of these crustaceans to gillnet operations. Proposed measures for bycatch reduction in the monkfish fishery in future, include area closures within monkfish southern fishing grounds and effort reduction measures. Whereas the former would minimize mortality on wreckfish, angel shark and also royal crab populations, effort reduction would benefit particularly the royal crab, which constitutes a conflict aspect of deep-water trap and gillnet fishery development off southern Brazil. © 2004 Elsevier B.V. All rights reserved. Keywords: Southern Brazil; Monkfish; Deep-sea gillnet fishery; Bycatch ∗ Corresponding author. Tel.: +55 47 341 7714; fax: +55 47 341 7715. E-mail address: [email protected] (J.A.A. Perez). 0165-7836/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.fishres.2004.10.011 82 J.A.A. Perez, R. Wahrlich / Fisheries Research 72 (2005) 81–95 1. Introduction (eastern Canada and Grand Banks) and L. piscatorius and L. budegassa (the NW Iberian Peninsula) (Bruno Deep-water demersal fishing operations off south- et al., 2001; Kulka and Miri, 2001; Pineiro˜ et al., 2001). ◦ ◦ ern Brazil (20 S–34 S) have increased in recent years The new fishing and processing technology motivated as a response to (a) decreasing catch rates of mostly a great interest of the government and the fishing indus- overexploited shelf resources (Valentini et al., 1991; try, since it not only seemed to meet international mar- Haimovici, 1998) and (b) the expansion of interna- ket demands more adequately, but also caught mostly tional markets for traditionally discarded or poorly large and mature fish as compared to traditional trawl- known species. Although gillnet and longline deep- ing practices (Perez et al., 2002, 2003). water operations directed towards rockfish and sharks While the positive aspects of this new fishery were have existed, since mid 1980s (Barcellos et al., 1991; evident, its impact on the deep marine ecosystem, par- Haimovici et al., 1997; Peres and Haimovici, 1998), ticular due to bycatch of other commercial and non- deep-water fisheries development greatly acceler- commercial species, raised concern and required fur- ated after 1999, as a large part of the shrimp and ther investigation to provide ecological support for up- groundfish trawlers expanded their fishing grounds coming management actions. towards outer shelf and slope areas, aiming for During 2001, these operations were observed and re- emerging valuable resources, such as flatfishes (Par- ported in detail providing extensive technological and alichthys spp.), pink cusk-eel (Genipterus brasilien- biological information, which has been the basis for a sis), monkfish (Lophius gastrophysus), skates (fam- complete assessment of the monkfish fishery off south- ily Rajidae), argentinean squid (IIlex argentinus) and ern Brazil (Perez et al., 2002). As part of this monitor- deep-water lobster (Metanephrops rubellus)(Perez ing, the present study aims at producing an analysis on et al.,2001). bycatch of the Brazilian monkfish gillnet fishery as a In addition to this process, the Ministry of Agricul- preliminary assessment of its potential impact on the ture (Brazilian Government) launched in 1998 a deep- Southwest Atlantic deep-water ecosystem. water fishing development policy in an attempt to in- duce the economical exploitation of the outer bounds of Brazilian EEZ. This policy was based on the char- 2. Material and methods tering of foreign vessels by national companies and an intense technological and scientific monitoring of 2.1. Data collection and analysis these deep-water fishing operations. Together, such ac- tions would allow (a) the improvement of the knowl- Monkfish gillnet fishery was conducted off south- edge on potential resources; (b) the evaluation of large- ern Brazil in 2001 by nine authorized vessels originally scale deep-water fishing operations profitability and (c) from Spain. These vessels were 26.7–39.6 m long with the absorption of adequate technologies for deep-water gross tonnage ranging between 72 and 362 t and frozen fishing, handling and processing. Since 2000, chartered stocking capacity ranging between 45 and 110 t. Fish- vessels using pots, bottom longlines, bottom gillnets ing trips included 24–86 sets, each of them consisting and otter trawls started to operate in 200–900 m deep of 60–400 stringed nets, which were kept immersed for areas off southern Brazil being fully monitored by ob- 110.6 h on average (±1.5 S.E.M.). The nets were 50 m servers and a VMS (Vessel Monitoring System) (Perez long and 13.5 meshes high and were anchored to the et al., 2003). sea floor by 60–80 kg weights. Mesh size used by all The monkfish has been one of the main targets iden- vessels was 280 mm (stretched mesh). tified in this fishing expansion process. In 2001, 8830 t Bycatch data were obtained during 14 gillnet fish- were landed, mostly by national trawlers (62%) and ing trips conducted by eight observed vessels between chartered gillneters (36%), the latter introducing in the January and December 2001 (Table 1). These data were area both the monkfish directed bottom gillnet fishing generated as part of a scientific cooperation program method and on board handling and processing prac- established between the Ministry of Agriculture and tices (Perez et al., 2002). This method is conducted in University of “Vale do Itaja´ı” (Santa Catarina State, other fishing areas of the Atlantic for L. americanus Brazil) (Perez et al., 2003). J.A.A. Perez, R. Wahrlich / Fisheries Research 72 (2005) 81–95 83 Table 1 Summary of the gillnet chartered vessels commercial trips off southern Brazil between January and December 2001, analyzed for bycatch assessment Vessel Period Latitude range Longitude range Depth Conducted tows Immersed Nets Monkfish range (m) (analyzed) (analyzed) catch kg Antoxo 24 Jun–06 Aug −29.139 to −32.256 −47.796 to −50.153 283–574 24 (20) 7719 (1835) 39386 Belen 12 Jul–04 Sep −26.010 to −28.130 −46.010 to −49.500 289–443 49 (22) 15966 (2339) 103223 Eder sands 12 Jun–13 Aug −30.553 to −32.041 −48.440 to −50.038 318–508 50 (36) 16595 (3705) 231500 Juno 12 Jan–02 Mar −23.668 to −26.837 −41.736 to −45.845 300–425 46 (43) 17470 (5060) 123762 19 Jun–17 Jul −30.491 to −31.828 −48.150 to −49.987 338–450 47 (47) 17560 (7420) 125642 19 Sep–10 Nov −28.810 to −33.658 −33.620 to −51.050 244–535 41 (19) 12860 (2469) 28648 Slebech 04 Nov–20 Dec −22.730 to −28.993 −45.779 to −47.832 293–421 41 (39) 14418 (4578) 69386 South coast 24 Feb–30 Apr −23.682 to −24.749 −41.067 to −44.529 252–561 61 (27) 20358 (2689) 60149 18 Oct–17 Dec −23.070 to −23.820 −41.330 to −42.410 132–607 67 (58) 22857 (7410) 100914 Suffolk 04 Feb–06 Apr −24.007 to −25.681 −43.075 to −45.350 222–543 56 (36) 17346 (3487) 131557 chieftain 09 Apr–25 Jun −24.828 to −29.402 −44.573 to −48.201 245–420 47 (27) 18150 (3293) 53130 27 Jun–03 Sep −24.881 to −27.000 −44.661 to −52.383 314–483 61 (54) 23923 (7301) 131296 Titan 17 Jul–24 Sep −28.673 to −34.349 −47.430 to −51.702 243–520 63 (48) 21770 (14490) 95925 l6 Oct–11 Dec −28.830 to −33.940 −47.690 to −51.310 200–485 61 (47) 17750 (14500) 39256 Total 12 Jan–20 Dec −22.730 to −34.349 −33.620 to −52.383 132–607 714 (523) 244739 (80576) 1315674 Observers assessed the catch composition of 80,576 nets immersed during 523 selected fishing sets con- ducted between latitudes 22◦44S and 34◦21S and lon- gitudes 33◦37W and 52◦23W, and between the 132 and 607 m isobaths (Table 1; Fig.