Ra Iston and Tagami: Exploitable biomass of Heterocarpus laevigatas in Hawaiian Is , Part 1 



495 



these findings could form a basis for estimating the 

 potential yield of the shrimp resource. 



To determine the exploitable biomass of shrimp at 

 spatially discrete locations in Hawaii, the formula 

 (Ricker 1975) was used: 



CPUE = q 



where CPUE is the catch-per-unit-effort, q is the catch- 

 ability coefficient of the fishing gear, B is exploitable 

 biomass, and A is the area occupied by the population. 

 This relationship is based on the explicit assumption 

 that catch rate is strictly proportional to shrimp den- 

 sity (B/A), and that q is the proportionality constant 

 equating these quantities. By rearrangement we have, 



B = 



CPUE • A 



Thus, to estimate the exploitable shrimp biomass at a 

 locality we need (1) an unbiased estimate of catch rate, 

 (2) a measure of the habitat area over which the catch 

 rate prevails, and (3) knowledge of the sampling gear's 

 efficiency (i.e., an estimate of q). 



To accomplish these three objectives, the study was 

 divided into two phases. First, a depletion experiment 

 was conducted to estimate the catchability coefficient 

 (q). This was followed by a depth-stratified sampling 

 program for H. laevigatus around each of the main 

 islands of the Hawaiian archipelago (i.e., Hawaii, Maui, 

 Kahoolawe, Lanai, Molokai, Oahu, Kauai, and Niihau). 



Methods 



face, each trap was emptied and the contents were 

 sorted to species, counted, and weighed to the nearest 

 0.01kg. Random subsamples of ~200 H. laevigatus 

 were routinely collected, from which carapace length 

 (CL) was measured to the nearest 0.1mm using dial 

 calipers. For all measured shrimp, sex was determined 

 by examining the endopodite of the first pleopod 

 (spatulate in males, pinnate in females; see King and 

 Moffitt 1984). In addition, the ovigerous condition of 

 females was recorded. 



Shrimp trapping was conducted during a series of nine 

 cruises of the NOAA ship Townsend Cromwell (Table 

 1). During each cruise, standard fishing gear was util- 

 ized to gather CPUE statistics at specific geographical 

 locations. The gear employed was a top-loading pyra- 

 midal shrimp trap, identical in construction to those 

 used commercially in Hawaii from 1983 to 1984. Each 

 trap was made of welded steel reinforcement bars, had 

 a 1.83 m- base, an overall volume of 1.84m^, and was 

 covered by 1.27x2.54 cm mesh hardware cloth. A full 

 description and illustration of the gear is given in 

 Tagami and Barrows (1988). 



Typically, 6-10 solitary traps were set daily and 

 allowed to soak overnight. Traps were generally de- 

 ployed in the afternoon and hauled the following morn- 

 ing, being in the water for a period of 16-20 hours. All 

 traps were baited with approximately 3 kg of chopped 

 mackerel Scomber japonicus. After hauling to the sur- 



Depletion experiment 



To estimate q, an intensive fishing experiment was con- 

 ducted (see also Ralston 1986). Depletion experiments, 

 including the Leslie method used here (Ricker 1975), 

 have three restrictive assumptions. First, all individuals 

 in the exploitable portion of the population are equally 

 likely to be captured with the fishing gear. Second, the 

 fished population is closed, or else additions exactly 

 balance removals other than those due to fishing. Third, 

 fishing removals account for all changes in stock bio- 

 mass, such that natural mortality, growth, and recruit- 

 ment have negligible effects during the period of 

 fishing. Thus, the best site for a depletion experiment 

 is a naturally isolated, small area so that removals can 

 be carried out over as short a time-interval as possible. 

 A small rise midway in the Kaulakahi Channel (21° 

 54.5'N, 159°56.5'W) separating Kauai and Niihau was 



