Estimated drift gillnet selectivity for 

 albacore Thunnus alalunga 



Norman Bartoo 

 David Holts 



La Jolla Laboratory, Southwest Fisheries Science Center 

 National Marine Fisheries Service. NO/V\ 

 P.O. Box 271, La Jolla, California 92038-0271 



Drift gillnet (DGN) fisheries, some 

 directed toward albacore Thunnus 

 alalunga and some toward other 

 species, have operated in the north 

 and south Pacific Ocean. These 

 fisheries, some of which had their 

 beginnings decades ago, expanded 

 in the early 1980s and exerted a 

 substantial, but as yet unknown, 

 mortality on Pacific albacore popu- 

 lations. Knowledge of the selectiv- 

 ity of DGNs is useful for applying 

 complex population-dynamics tech- 

 niques to the albacore population. 



Several DGN experimental sur- 

 veys conducted in the north and 

 south Pacific Ocean have produced 

 data on albacore distribution and 

 gear selectivity. In 1984-85 the 

 Southwest Fisheries Science Cen- 

 ter of the National Marine Fisher- 

 ies Service (NMFS) undertook an 

 experiment to quantify DGN mesh- 

 size selectivity for albacore. The ex- 

 periment produced size-selectivity 

 data for a single mesh size used in 

 the U.S. DGN commercial fishery 

 for albacore ( 184 mm; all mesh sizes 

 referred to are stretch mesh mea- 

 surements), in the eastern North 

 Pacific Ocean. Other experiments 

 designed to sample albacore were 

 conducted by the Japan Marine 

 Fishery Resource Center in 1980 

 (JAMARC 1983), 1982 (JAMARC 

 1985), and 1983 (JAMARC 1986). 

 These experiments used other mesh 

 sizes which produced complimen- 

 tary data used in this study. The 

 objective of this analysis was to 

 quantify drift gillnet mesh-size 

 selectivity for albacore, using exist- 



ing data collected from various ex- 

 periments and surveys in the Pa- 

 cific Ocean. 



Methods 



Various experiments providing data 

 have been conducted independently 

 at different times and in different 

 areas (Fig. 1). Consequently, some 

 assumptions had to be made con- 

 cerning comparability of nets used 

 and availability and vulnerability of 

 fish sizes to be sampled. However, 

 there was enough information to 

 provide insights into net selectivity 

 for albacore. The nets used in all 

 the experiments were similar to 

 typical Japanese commercial drift 

 nets. All nets had similar mesh- 

 hanging ratios, twine/mesh-size 

 ratios, and were of multifilament 

 construction (Table 1). These are the 

 most important factors affecting the 

 selectivity curve for a given mesh 

 size(Hamley 1975). 



Detailed descriptions for the 

 JAMARC experiments, including 

 set locations, timing, mesh size 

 placement, etc, are contained in the 

 JAMARC references. These experi- 

 ments were designed to obtain in- 

 formation on resource distribution 

 and catch rates for possible future 

 commercial fisheries. In brief, the 

 JAMARC North Pacific experiment 

 was a lyr survey beginning in the 

 spring of 1980 (Fig. 1). It targeted 

 albacore using nets with mesh sizes 

 of 130-250mm. The JAMARC 1982 

 and 1983 experiments in the South 



Pacific were conducted in the area 

 shown in Fig. 1 and targeted all 

 tuna-like species, using 180 mm 

 mesh nets in 1982, and 118 mm and 

 160-180 mm mesh nets in 1983. 



The NMFS experiment targeted 

 albacore. A total of 27 night sets 

 was made: 12 sets in waters south 

 of Point Conception and within 

 300 mi of shore; 7 sets in the vicin- 

 ity of the Guide Seamount south of 

 the Farallon Islands; and 8 sets 

 approximately 1000 mi west of Or- 

 egon near 142° W long, and 45° N 

 lat. (Fig. 1). All sets used 184mm 

 mesh nets. 



In all the experiments using mesh 

 sizes of >130mm, the amount of 

 gear fished during each set ranged 

 from 2760 m to 10,800 m, and the 

 number of sets ranged from 11 to 

 176 (Table 1). Nets in each experi- 

 ment were deployed and retrieved 

 as in typical commercial operations. 

 Nets were deployed at dusk, and 

 retrieval began at first light often 

 extending into midmorning. Soak 

 times were -10-16 h/set. 



Data collected 



Data published by JAMARC are 

 limited to fork length (FL) frequen- 

 cies by mesh size by set. The NMFS 

 data include fork length-frequencies, 

 maximum or largest body girth, and 

 opercular girths for each fish, and 

 notation on the mode of capture for 

 each fish (tangled, wedged, or 

 gilled). Fish were described as 

 wedged or gilled when inserted into 

 a mesh so that the entire perimeter 

 of the mesh held the fish. This oc- 

 curred at locations on the body from 

 the snout to the point of maximum 

 body girth. Fish were described as 

 tangled when snared or tangled by 

 fins or tail. Length-frequencies for 

 U.S. troll-caught fish from the same 

 time and areas are available for 



Manuscript accepted 9 February 1993. 

 Fishery Bulletin, U.S. 92:371-378 ( 1993). 



37) 



