Abstract. -To resolve the uncer- 

 tainty in estimating capture depths 

 of fish on pelagic longline gear, elec- 

 tronic microchip hook timers were 

 attached to branch lines to record 

 when bites occurred, and time-depth 

 recorders (TDRs) were attached to 

 longline gear, off Hawaii in January 

 1989 and January-February 1990. 

 Hook timers indicated that 32% of 

 the striped marlin Tetrapturus au- 

 dax, 21% of the spearfish T. angus- 

 tirostris, and 12% of the bigeye tuna 

 Thunrvus ohesus were caught on sink- 

 ing or rising hooks, demonstrating 

 that capture time data are needed to 

 correctly estimate capture depth. 

 Recorded and predicted longline 

 depths differed greatly, indicating 

 that TDRs are essential for describ- 

 ing depth distributions of fish from 

 longline catches. Most (> 60%) of the 

 spearfish and striped marlin were 

 caught on settled hooks (not sinking 

 or rising) at depths of < 120 m, where- 

 as most bigeye tuna were caught at 

 depths of > 200 m. This suggests that 

 eliminating shallow hooks could 

 substantially reduce the bycatch of 

 spearfish, striped marlin, and other 

 recreationally important billfish 

 without reducing fishing efficiency 

 for bigeye tuna. Bigeye tuna and 

 striped marlin survived up to 6-9 

 hours after capture, and over 50% of 

 12 frequently-caught taxa were alive 

 when retrieved, suggesting that the 

 release of live fish can be an effec- 

 tive management option. 



Depth, capture time, and hooked 

 longevity of longline-caught 

 pelagic fish: Timing bites 

 of fish \N\th chips 



Christofer H. Boggs 



Honolulu Laboratory, Southwest Fisheries Science Center 



National Marine Fisheries Service, NOAA 



2570 Dole Street, Honolulu, Hawaii 96822-2396 



Manuscript accepted 5 May 1992. 

 Fishery Bulletin, U.S. 90:642-658 (1992). 



Targeting specific depths can improve 

 longline catches of desired species, 

 such as bigeye tuna Thunnus obesus 

 (Saito 1975, Hanamoto 1976, Suzuki 

 et al. 1977, Suzuki and Kume 1982), 

 and reduce bycatch of other species, 

 such as billfish (Suzuki 1989). How- 

 ever, considerable uncertainty exists 

 in estimating the fishing depth of 

 longline gear. Predicted longline depth 

 based on catenary geometry, line 

 length, and distance between floats 

 (Yoshihara 1954) differs from true 

 depth (Saito 1973, Hanamoto 1974, 

 Nishi 1990) because of currents and 

 other factors, yet depth is often in- 

 ferred rather than measured (Suzuki 

 et al. 1977, Suzuki and Kume 1982, 

 Hanamoto 1987, Grudinin 1989). Fur- 

 thermore, fish may be caught while 

 the hooks are sinking, during deploy- 

 ment of the gear, or rising during its 

 retrieval (Saito 1973), making cap- 

 ture depths impossible to estimate ac- 

 curately without known capttu-e times. 

 Accurate estimates of fishing depth 

 can be made if time-depth recorders 

 (TDRs) are attached to longline gear. 

 Longline studies using TDRs (Saito 

 et al. 1970, Saito 1973, Yamaguchi 

 1989, Nishi 1990) have also inter- 

 preted TDR depth fluctuations as 

 records of times of capture, but few 

 such measurements exist. Instead, 

 capture has been assumed to occur 

 when the gear is settled, so capture 

 depth has been estimated as settled 

 hook depth (Hanamoto 1976, Suzuki 

 and Kume 1982). Hook timers, de- 



signed to indicate when each hook is 

 struck (Somerton et al. 1988), pro- 

 vide a way to measure capture times 

 and survival times of hooked fish. 

 Capture times, together with TDR 

 records, can be used to estimate cap- 

 ture depths accurately. 



Billfish catch rates in recreational 

 fisheries may be negatively affected 

 by nearby longline fisheries (Squire 

 and Au 1990), and interest in finding 

 ways to reduce the longline take of 

 billfish without reducing fishing effi- 

 ciency for target species is increas- 

 ing (Rockefeller 1989). Information 

 on capture depth, capture time, and 

 hooked longevity can be used to 

 design fishing methods that reduce 

 billfish mortality. Data on the selec- 

 tivity and efficiency of longline gear 

 at various depths are also critical for 

 stock assessments (Suzuki 1989). 



The present study improves meth- 

 ods for estimating capture depths of 

 fish on longline gear using electronic 

 timing devices, and describes the 

 depth distributions and capture times 

 of tunas, billfishes, sharks, and other 

 pelagic fishes in Hawaiian waters in 

 winter. Water temperature and dis- 

 solved oxygen (DO) were measured 

 to describe the physical habitat in the 

 study area, since these variables ap- 

 pear to cause geographic variation 

 in depth distributions of fish (Hana- 

 moto 1975, 1987). Relative fishing 

 efficiency and the bycatch of bill- 

 fish were predicted for several gear 

 configurations. 



642 



