Boggs: Estimating capture depths of longline-caught pelagic fish 



649 



while the gear was settled rather than whOe it was sink- 

 ing or rising (Fig. 5), probably because the gear spent 

 much more time in the settled position. However, 

 substantial numbers of mahimahi, billfish, and other 

 species were caught while the gear was rising (Fig. 5), 

 which explains how fish were caught on deep-positioned 

 hooks (unconfirmed depths) when their confirmed 

 depth distribution was shallow (confirmed depths; 

 Fig. 4). 



For many species, catch-per-unit-time (CPUT) may 

 have been highest while the gear was rising. The CPUT 

 values at <3 and ^3h before recovery were not directly 

 comparable because short sets resulted in lower effort 

 (number of hooks with timers) >3h before recovery. 

 The catch in the 1-2 h and 2-3 h periods (Fig. 5) must 

 be divided by 2 for comparison with the CPUT in the 

 0.5h and 0.5-<1.0h periods. For periods of 0.5-1.0, 

 1-2, and 2-3h before recovery, CPUT values were less 

 than during the rising period for ribbonfish, pomfrets, 

 mahimahi, lancetfish, striped marlin, spearfish, brown 

 ray, and whitetip shark (Fig. 5). In contrast, the values 

 for yellowfin and skipjack tunas were not much dif- 

 ferent between settled and moving gear and were 

 highest for bigeye tuna 1-2 h and 3-4 h before recovery. 

 Blue shark CPUT values were highest l-2h before 

 recovery. 



Relatively large numbers of fish were categorized as 

 caught at recovery (Fig. 5). However, estimates of the 

 delay between recovery and reading each timer were 

 not precise (± 1 minute). Thus some fish caught "at 

 recovery" actually had timers triggered after recovery. 

 Hook timers that did not catch fish were most often 

 triggered "at recovery" (Table 3), suggesting that 

 handling activated the timers. A similar lack of preci- 

 sion affected capture times "at deployment." 



Hooks with timers triggered by small fish or without 

 catching fish may have resulted in false capture times 

 if larger fish were caught later on those hooks. For- 

 tunately, small (<10kg) fish, particularly lancetfish 

 (Table 2), were most frequently caught without trig- 

 gering the timers (Fig. 5). It was unusual for larger 

 (~10-90kg) fish, such as tunas, billfishes, or sharks 

 (Table 2, Fig. 5), to be caught without triggering the 

 timers. The increase in breaking strength of the trig- 

 gers in 1990 (Fig. 1) decreased the relative number of 

 small fish that triggered timers, and reduced the pro- 

 portion of timers triggered without catching fish from 

 18.5% in 1989 to 9.7% in 1990 (Table 3). 



Survival and release 



Over 56% of the fish other than wahoo and skipjack 

 tuna were alive when recovered, and for most species, 

 survival was higher than 70% (Table 2). Based on fish 

 with hook-timer data, over half of the bigeye tuna 

 recovered up to 9 h after capture were alive. None of 

 the 11 bigeye tuna recovered 1-2 h after capture were 

 dead, and the shortest period between capture and 

 recovery of dead bigeye tuna was 2-3 h (Fig. 6). Striped 

 marlin were less hardy, with over half recovered dead 

 > 3 h after capture; nevertheless, many were recovered 

 alive up to 6-8 h after capture (Fig. 6). Spearfish were 

 the least hardy: The longest survival time was 5-6 h, 

 and dead fish were recovered at < 1-2 h after capture 

 (Fig. 6). 



Of the 29 bigeye tuna, 35 striped marlin, and 11 

 spearfish tagged during the study, 2 bigeye tuna and 

 1 striped marlin were recaptured 3-10 months later. 

 These three fish were tagged after having been on 

 branch lines for 3-6 h. The marlin had been injected 



Table 3 



Frequencies of activated hook timers on branch Mnes without fish (as percentage of total timers) categorized by elapsed time since 

 the timers were triggered (range of values from individual sets in parentheses). 



Year 



Elapsed time 



Before retrieval 



After deployment 



At retrieval 

 (•Slmin) 



Rising 

 (>l-30min) 



Settled 

 (>30min) 



Sinking 

 (•eSOmin) 



At deployment 

 (<2min) 



Activated 



before 

 deployment 



No. 

 timers 



Branch lines 



with timers 



(%) 



1989 

 1990 

 Combined 



6.4 



(-)* 



3.8 



(1.1-6.2) 



4.7* 



1.2 

 (0-3.2) 



1.5 

 (0-4.8) 



1.4 



3.9 



(0.8-7.1) 



2.6 

 (0.8-6.5) 



3.1 



1.0 

 (0-5,7) 



0.4 



(0-1.1) 



0.6 



4.8 

 (1.6-7.1) 



1.0 

 (0-2.6) 



2.4 



1.0 

 (0-3.3) 



0.4 

 (0-2.4) 



0.6 



3744 

 (126-356) 



6492 

 (167-418) 



10236 



86.0 

 (61-100) 



64.5 

 (34-99) 



71.0 



'Number recorded only during the first set in 1989. To calculate the combined frequency (4.7%), frequency was assumed to be 6.4% 

 throughout 1989. 



