496 



Fishery Bulletin 90(3), 1992 



chosen for the work. This nearly circular rise (Fig. 1) 

 with a crest at 421m (230fm), has an area of 1187ha 

 (horizontal planar area <640m or 350 fm) and is 

 isolated from the Islands of Kauai and Niihau by depths 

 >730m (400 fm; Fig. 2). The site lies in the required 

 depth range for H. laevigatus and has relatively high 

 densities of the target species. 



The intensive fishing experiment was conducted 

 13-24 May 1986. During each of the 12 days of the 

 experiment, 6-14 pyramidal shrimp traps were set be- 

 tween depths of 421 and 695 m (230 and 380 fm). 

 Following the Leslie method (Ricker 1975, Seber 1982), 

 catchability was estimated directly from the slope of 

 the linear regression of CPUE on corrected cumulative 

 catch. That is, 



ca 



57 



56 







55 



to 



0) 



■+-> 



54 



53 I I I I I I I I I I I I I I I I I I I I I I I I I 

 59 58 57 56 55 54 



Longitude 

 (minutes W of 159°) 



Figure 1 



Contour map of the Kaulakahi Channel study site. Depths are in fathoms, 

 and circles represent trap-set locations. 



22°30' 



CPUEi = q • Bj 



= q(Bo - K,) 

 = qBo - qKi, 



where CPUE; is the catch-per-unit-effort 

 on day i (kg/ trap-night), q is the catchabil- 

 ity coefficient/trap-night of the pyramid 

 traps, Bj is the average biomass (kg) pres- 

 ent on day i, Bq is the biomass (kg) of 

 shrimp present at the start of the experi- 

 ment, and Kj is the corrected cumulative 

 removals for day i, defined as 



i-l 



K, 



2 n = l 



0) 



^ 22 00' 



CO 



-J 



160° 30' 



WO'OO' 



159'30' 



Longitude (W) 



Figure 2 



Contour map of the Kauai-Niihau area depicting the 100, 200, 300, 400, 500, 600, and 

 lOOOfm isobaths. Note the depletion study site, i.e., the small rise located mid-channel. 



where Cn(n = l,2 i) is the catch (kg) 



taken on each day of the experiment. Note 

 that the estimate of catchability (q) pertains 

 strictly to the interaction between the traps 

 we used and the stock resident in the study 

 area, which is normalized to unit 

 area after multiplying by 1187ha 

 (i.e., the area of the study site). 



Determination of 

 habitat areas 



The distribution of H. laevigatus 

 is strongly dependent on bottom 

 depth. Little or no shrimp occur 

 in waters outside the 366-915 m 

 (200-500 fm) range (Struhsaker 

 and Aasted 1974, Gooding 1984, 

 Dailey and Ralston 1986). In the 

 main portion of the Hawaiian 

 Archipelago (Hawaii, Maui, La- 

 nai, Kahoolawe, Molokai, Oahu, 

 Kauai, and Niihau) a number of 

 islands share a common 915m 

 (500 fm) depth contour (e.g, Kau- 

 ai and Niihau; Fig. 2). Even so, 

 in our study each island was 

 treated as a separate stock for 



