654 



Fishery Bulletin 90(4|, 1992 



example, since deep gear is more efficient than regular 

 gear for bigeye tuna, this species must occupy a rela- 

 tively deep habitat (Suzuki et al. 1977). More specific 

 information on habitat depth is provided by catches 

 and CPUEh by hook position (Hanamoto 1979 and 

 1987, Hanamoto et al. 1982, Suzuki and Kume 1982), 

 especially when TDRs are used to record gear depth 

 (Saito 1973 and 1975, Hanamoto 1974, Nishi 1990). 

 Capture depth estimates without TDR records ignore 

 major variations in actual gear depth (Fig. 2; Nishi 

 1990), and those without hook timers are biased by the 

 inclusion of inappropriate hook depths. 



A possible source of bias in the present study is the 

 inclusion of some falsely confirmed depths due to fish 

 being caught with timers already activated. The pro- 

 portion of false estimates should be similar to the fre- 

 quency of timers that were without fish and were trig- 

 gered while settled, which was only 3.9% in 1989 and 

 2.6% in 1990 (Table 3). Thus it is unlikely that >4% 

 of confirmed capture depths in this study are incorrect 

 because of false timer readings. 



Many pelagic longline studies (Saito 1975, Hanamoto 

 1976, Yang and Gong 1988) assume that fish are caught 

 while hooks are at settled depths. Supporting this 

 assumption, Saito (1973) has shown that albacore Thun- 

 nus alalunga are caught almost exclusively by settled 

 hooks, based on capture times indicated by fluctuations 

 in TDR records. Using hook timers, the present study 

 adds new information: Almost 90% of bigeye and 

 yellowfin tuna also are caught while hooks are at 

 settled depths (Table 4). However, hook timers indicate 

 this generalization does not extend to striped marlin, 

 spearfish, mahimahi (Table 4), and most of the com- 

 mercially unimportant species (Fig. 5). Although most 

 of these fish are also caught on settled hooks, a substan- 

 tial fraction are not, and this must be considered when 

 quantifying their depth ranges (Fig. 4). 



Besides the present study, little information exists 

 on longline capture depths for mahimahi, spearfish, and 

 striped marlin. In the study area, CPUEn values for 

 these species (Fig. 7) indicate maximum abundance at 

 depths in the mixed layer for mahimahi (<100m, 

 24°-25°C; Fig. 10), extending into the top of the ther- 

 mocline for striped marlin (120 m, 20°C) and into the 

 middle of the thermocline for spearfish. Striped marlin 

 are reported to be caught most frequently on longline 

 hooks closest to the surface (60-90 m) in the eastern 

 tropical Pacific and Indian Oceans, but they may be 

 more abundant above this depth (Hanamoto 1979, 

 Hanamoto et al. 1982). Mahimahi and spearfish may 

 also be more abundant above the uppermost stratum 

 (40-80 m) in the present study, since their catch rates 

 appear to increase towards the surface (Fig. 7). 



Striped marlin are also reported caught on deep 

 longline hooks (~200m; Hanamoto et al. 1982) and at 



the deep end of vertical longline gear (336 m; Saito 

 1973); but in the present study, their deepest confirmed 

 capture depth is 210m. Tracking data on striped marlin 

 off California indicate a shallow (< 60 m) depth distribu- 

 tion with most of the daytime spent within 10 m of the 

 surface (Holts and Bedford 1989). 



The depth distribution (200-400 m) of bigeye tuna in 

 the present study is deeper than in many previous 

 reports (Hanamoto 1974, 50-160 m; Saito 1975, 207- 

 245 m; Suzuki and Kume 1982, 1 70-300 m; Yang and 

 Gong 1988, 260-300m; Nishi 1990, 140-180m), al- 

 though these studies have found bigeye tuna are most 

 abundant on the deepest hooks fished. Hanamoto 

 (1987) hypothesizes a habitat depth of 250-400 m for 

 the central Pacific Ocean at latitude 25°N, based on 

 the observed maximum longline CPUE at tempera- 

 tures of 10°-15°C. The highest CPUEd values in the 

 present study are at the cold, deep end of this range 

 (Fig. 7), deeper than most hooks used in commercial 

 fishing gear. However, the CPUEq value at 280-400 

 m is not significantly different from that at 200-400 m 

 (Fig. 8). Although these results may be specific to 

 January and February, perhaps commercial CPUE 

 could be improved by increasing fishing depth, at least 

 during winter months. 



Seasonal and geographic variation in temperature 

 and DO profiles may affect the depth preferences of 

 pelagic fish. Hanamoto (1975, 1987) has hypothesized 

 that the deep end of bigeye tuna habitat is limited by 

 DO concentrations below ImL/L (1.4mg/L) and by 

 temperatures below 10°C. Results of the present study 

 suggest that bigeye tuna are seldom caught in waters 

 with a DO concentration of ~<2mg/L (Fig. 10). Oxy- 

 gen concentrations of ~2-3mg/L cause significant 

 reductions in bigeye tuna cardiac output (1.9-2. 6 mg/L) 

 and heart rate (2.7-3.5 mg/L), suggesting that bigeye 

 tuna cannot maintain a full range of activity at lower 

 DO concentrations (Bushnell et al. 1990). 



Longline data to support the hypothesis of a 10°C 

 temperature limit independent of the DO limit are 

 sparse. Few hooks have been deployed in waters colder 

 than 9°-10°C with DO concentrations of >lmL/L 

 (Hanamoto 1975, 1987). In the present study, the only 

 area with DO values > 2 mg/L and temperatures <8°C 

 was at lat. 10°-20°N (Fig. 10). Currents prevented 

 hooks from reaching cold (6°-8°C) water in this area. 



Sonic tracking of bigeye tuna around Hawaii indi- 

 cates a depth distribution slightly shallower than that 

 in longline studies (Hanamoto 1987, 250-400 m; pres- 

 ent study, 200-400m). Holland et al. (1990) have 

 reported that tracked bigeye tuna spend most of the 

 daytime at 200-240m in 14°-17°C water. This may be 

 due to the association of the tracked bigeye tuna with 

 fish aggregating devices or due to a size-related differ- 

 ence. The 72- to 74 cm bigeye tuna studied by Holland 



