MOFFITT ET AL.: DISTRIBUTION AND YIELD OF DEEPWATER SHRIMP 



pelago were conducted on seven cruises of the 

 NOAA ship Townsend Cromwell (TC) between April 

 1982 and Augnst 1984 and one charter cruise of the 

 University of Guam vessel Pesquedot (PQ) in August 

 1984. The standard gear used consisted of strings 

 of five canvas-covered, half-round shrimp traps set 

 about 40 m apart. The traps were constructed of a 

 reinforcing bar frame (about 90 cm long, 65 cm wide, 

 and 45 cm high) wrapped with 2.5 x 1.3 cm mesh, 

 18-gauge welded wire. An entry cone with an open- 

 ing of approximately 10 cm was located at each end 

 of the trap. A figure of this general trap design is 

 shown in Gooding (1984). The ground lines attach- 

 ing the traps together and the main lines attaching 

 the ground line to the surface buoys were of 13 mm 

 polypropylene line. Strings were usually set in the 

 afternoon and retrieved the following morning. 

 Normal soaking times ranged from 15 to 24 hours. 

 Traps were baited with Pacific mackerel, Scomber 

 japonicus. 



Ordinarily, subsamples of 100 specimens of each 

 of the three major species of Heterocarpus {H. laevi- 

 gatus, H. ensifer, and H. longirostris) were saved 

 each day from each depth sampled. Two sampling 

 sites, Esmeralda Bank and Pagan Island, were 

 visited on each of six Cromwell cruises and large 

 subsamples of 400 H. laevigatus were saved on each 

 visit. All specimens were returned to the laboratory 

 where the carapace length, sex, and reproductive 

 condition were recorded. The areas of suitable 

 habitat for H. laevigatus at each island and bank 

 location was estimated from charts using a computer 

 aided planimeter. 



In this study, the yield assessment approach 

 described by Polovina and Ralston (1986) was em- 

 ployed. A systematic survey of 22 islands and banks 

 gave information on the depth range of the primary 

 species and their relative abundance by area. An in- 

 tensive fishing experiment produced an estimate of 

 catchability using the Leslie method (Ralston 1986). 

 This information combined with an estimate of the 

 area of suitable habitat for each island or bank was 

 used to estimate available biomass by location. Esti- 

 mates of growth were obtained by application of 

 Elefan I (Pauly 1982) to a site specific time series 

 of length-frequency data. The ratio of mortality to 

 growth and asymptotic length was estimated from 

 a large length-frequency sample (Wetherall et al. in 

 press). Equilibrium yield as a function of fishing 

 mortality was determined from the Beverton and 

 Holt (1956) yield-per-recruit equation as the product 

 of yield per unexploited trappable biomass and the 

 trappable recruited biomass estimate obtained from 

 the systematic sampling and intensive fishing (Polo- 



vina and Ralston 1986). Estimates of recommended 

 yield from the equilibrium yield equation were ob- 

 tained based on marginal yield and minimum spawn- 

 ing stock biomass considerations. 



RESULTS^ 



Throughout the course of this survey a total effort 

 of 2,508 trap-nights was expended at 527 shrimp 

 trapping stations. The total catch of pandalid shrimp 

 was 5,188 kg for an overall catch rate of 2.07 

 kg/trap-night. Over 99% of this catch was composed 

 of//, ensifer, H. laevigatus, and//, longirostris. A 

 complete list of shrimp species taken during this 

 study is given in Table 1. 



DEPTH AND SIZE DISTRIBUTION 



A plot of catch per unit effort (CPUE) versus 

 depth for the three major species (Fig. 2) shows that 

 they inhabit different depth strata. Heterocarpus en- 

 sifer is the shallowest dwelling of these species. The 

 maximum catch rate for this species was 0.17 kg/ 

 trap-night at a depth of 366 m (200 fathoms). Un- 

 fortunately, 366 m was the shallowest depth tar- 

 geted throughout most of the survey. A small 

 amount of effort was expended in the 137-274 m 

 (75-150 fathoms) depth range on cruises TC-84-02 

 and PQ-84-01. Heterocarpus ensifer catches in this 

 depth range were negligible (only five shrimp in 37 

 trap-nights). Heterocarpus laevigatus, the most 

 abundant shrimp taken in our survey, was caught 



^Portions of this section are also presented in Polovina et a!. 

 1985. 



341 



