Boggs Estimating capture depths of longline-caught pelagic fish 



645 



sinking (<0.5h after gear deployment), (5) at deploy- 

 ment (< 2 min after deployment), and (6) before deploy- 

 ment (timer triggered before setting commenced). 

 Timers activated but without fish were categorized 

 similarly except all settled categories (>0.5-9.0h) were 

 combined. Untriggered hook timers with fish also were 

 tallied, and hooks with timers that were damaged, 

 broken loose, or tangled too badly to be triggered were 

 counted as hooks without timers. 



The numbers of fish caught while the gear was sink- 

 ing, settled, and rising were summarized. The uncon- 

 firmed depth of capture of each fish was defined as the 

 settled depth of the hook. Capture depths were con- 

 sidered confirmed only if hook timers indicated the cap- 

 ture occurred within the period in which the gear was 

 settled. 



Catch and effort 



Live fish that were not needed as specimens were 

 tagged and released. Steel head "H" type dart tags 

 (Squire 1987) were applied using a 3m tagging pole 

 while the fish remained in the water. Billfish were also 

 injected with 5-20 mg oxytetracycline/kg of fish using 

 pole-mounted syringes (Foreman 1987) to mark hard 

 parts for validation of growth increments. Fish were 

 released by cutting the branch lines close to the hooks 

 with a tree-trimming pole. The condition (alive or dead) 

 of the retained fish was noted, and it was weighed to 

 the nearest 0.5kg or measured to the nearest 0.1cm. 

 For the five most-frequently-caught species of commer- 

 cial importance, catch, number of hooks, and number 

 of hooks with timers were stratified into 40 m strata 

 (40- < 80 m, 80- < 120 m, and so on) based on settled 

 hook depths. The catch-per-unit-effort (CPUE) in each 

 depth stratum was examined in two ways: (1) by con- 

 firmed capture depth (CPUEd in number of fish/1000 

 hooks with timers) representing the depth distribution 

 of fish; and (2) by settled hook depth (CPUEh in num- 

 ber of fish/ 1000 hooks), representing the total effec- 

 tiveness of hooks while sinking, settled, or rising. 



The CPUEh for each depth was used to predict 

 catch rates of "standardized" types of gear to illustrate 

 the use of catch by hook position in estimating rela- 

 tive gear efficiency for different gear configurations. 

 Total CPUE for each standardized gear configuration 

 was estimated by calculating the weighted average 

 CPUEh , with weights corresponding to a given num- 

 ber of hooks per depth stratum for each configuration. 

 Total CPUE was calculated from 1989 and 1990 data 

 separately and averaged. Gear efficiency was calcu- 

 lated as the ratio of the predicted CPUE for each con- 

 figuration to that of the regular configuration. 



Standardized regular and deep longline gear config- 

 urations were assumed to have 6 and 13 hooks/basket, 



respectively. A shortening rate of 0.6 and the dimen- 

 sions in Suzuki et al. (1977; without adjustment for cur- 

 rents) indicated hook depths of about 95, 140, and 

 170 m (for regular gear) and 100, 145, 190, 230, 265, 

 290, and 300 m (for deep gear). These depths corre- 

 spond roughly to the midpoints of hook depth strata 

 in the present study (100, 140, 180, 220, 260, and 

 300 m). 



In addition to regular and deep gear, CPUE values 

 for two hypothetical gear types were predicted: (1) 

 shallow gear for which hooks are limited to the first 

 three depth strata of this study; and (2) a proposed new 

 gear for which no hooks would be deployed in the first 

 three depth strata and the distribution of deeper hooks 

 would match that of deep gear. The shallow gear con- 

 figuration may be representative of that achieved by 

 Hawaii's longline fishermen in 1989 and 1990 when 

 they first began using monofilament longline and had 

 difficulty achieving the depths formerly fished with 

 traditional rope gear. With the rope gear, slack was 

 obtained by manually throwing the baskets with the 

 main line partially coiled. The [predicted] CPUE for the 

 new gear type was estimated to indicate the reduction 

 in bycatch of some species by the elimination of shallow 

 hooks. 



To show CPUE as it would appear in a study of gear 

 configurations without hook position, capture depth, 

 or capture time information, CPUEs values were cal- 

 culated from catch and effort by set type. Sets were 

 categorized on the basis of depth (TDR depth plus 

 branch line length) into three groups: 60- < 200 m (reg- 

 ular), 200- < 330 m (deep), and 330-530 m (very deep). 

 The first two groups contained depth ranges roughly 

 comparable to those expected for regular and deep 

 longline gear types, assuming a variety of shortening 

 rates and variation due to ocean currents (Suzuki et 

 al. 1977). 



Oceanography 



Vertical temperature structure in the area of each set 

 was measured by expendable bathythermographs 

 (XBTs; 400 m depth) and conductivity-temperature- 

 depth casts (CTDs; 500- 1000 m depth, usually 500 m) 

 before or after each set. Water samples were taken 

 with Niskin bottles to measure DO and to calibrate DO 

 measurements made by CTDs. 



Many of the TDRs were equipped with a second chan- 

 nel to record temperature. The TDRs were attached 

 to the CTD probe to calibrate depth and temperature 

 measurements. The TDR temperature data were used 

 to estimate set depths exceeding 400 m (the lower limit 

 for accurate range depth measurement from the 

 TDRs). 



