FISHERY BULLETIN: VOL. 84, NO. 4 



completed in the Marianas (including Guam) and Fiji 

 (Wilder 1977; King 1983; King and Butler 1985; 

 Moffitt and Polovina 5 ). 



Evidence also exists to show that this species is 

 highly susceptible to trapping (Ralston 1986) and, 

 according to commercial fishermen, depletion of the 

 resource has occurred over certain fishing grounds 

 in Hawaii (S. Barrows 6 ). Because estimates of the 

 shrimp's productive capacity which are currently 

 available are preliminary at best and a fishery has 

 developed rapidly, this study set out to examine 

 aspects of the life history of the Hawaiian stock of 

 H. laevigatus to obtain information useful in devel- 

 oping a basis for management of the fishery. 



METHODS 



All sampling was conducted by commercial fishing 

 vessels owned by the Hawaiian Shrimp Company 

 (Easy Rider, Mokihana, and the Easy Rider Too) 

 over the 16-mo period from August 1983 to Novem- 

 ber 1984. During this time, six 35-60 d cruises were 

 completed and samples were obtained during 9 of 

 the 12 calendar months (Table 1). Fishing was con- 

 ducted throughout much of the Hawaiian Archi- 

 pelago, from Gardner Pinnacles south to the Island 

 of Hawaii (Fig. 1). Samples were collected at all of 

 the seven main islands (Hawaii, Kauai, Lanai, Maui, 

 Molokai, Niihau, and Oahu) and from Necker, 

 French Frigate Shoals, and Gardner Pinnacles in 

 the Northwestern Hawaiian Islands. 



All shrimp were caught during overnight sets of 

 baited pyramidal traps, which measured 1.5 x 1.8 

 m with a funnel opening at the top center. Fishing 

 was targeted between depths of 500 and 700 m, 



5 Moffitt, R. B., and J. J. Polovina. The distribution and yield 

 assessment of the deepwater shrimp resource in the Marianas. 

 Manuscr. in prep. Southwest Fisheries Center Honolulu Labora- 

 tory, National Marine Fisheries Service, NOAA, 2570 Dole Street, 

 Honolulu, HI 96822-2396. 



6 S. Barrows, Hawaiian Shrimp Company, 737 Bishop Street, 

 Suite 2910, Honolulu, HI 96813, pers. commun. 1985. 



Table 1 .—Temporal and geographic distribution of Heterocarpus 

 laevigatus samples (FFS = French Frigate Shoals). 



although some catches were made in both shallower 

 and deeper water because of the trap drift. The best 

 catch rates were found in areas of hard rough bot- 

 tom; otherwise, all sampling sites were to all ap- 

 pearances similar. 



Systematic subsamples of the catch were taken 

 from every other trap on every second fishing day 

 by randomly scooping approximately 0.9 kg (2 lb) 

 of shrimp from traps prior to emptying. Samples 

 were placed in double bags with tags recording date, 

 location, depth, and condition, and were then frozen 

 and packed for transfer to the laboratory. There all 

 shrimp were identified to species; sexed; examined 

 for embryos on the pleopods; measured to the near- 

 est 0.1 mm for carapace length (CL), carapace width 

 (CW), and total length (TL); and weighed to the 

 nearest 0.1 g on a top loading scale. The data were 

 then keypunched and stored for analysis. 



Size-frequency distributions of H. laevigatus were 

 analyzed by the regression method of Wetherall et 

 al. (in press) to estimate maximum size (L m of the 

 von Bertalanffy growth equation) and the ratio of 

 total instantaneous mortality rate (Z) to von Berta- 

 lanffy growth coefficient (K). Additionally, the 

 growth coefficient of H. laevigatus was estimated 

 by following the progression of size modes evident 

 in three large samples taken: 1) 24 October to 6 

 November 1983, 2) 24 April to 11 May 1984, and 

 3) 3 September to 6 November 1984. Sample sizes 

 of N = 2,021, 1,991, and 1,438 were obtained in 

 these respective samples, accounting for 74% of all 

 shrimp measured in the study. Modal progression 

 of size distributions was determined by the 

 ELEFAN I computer program of Pauly (1982). 



RESULTS 



A total of 7,368 H. laevigatus were measured and 

 examined for CL, sex, and the presence of eggs 

 (Table 1). Of these 3,956 were females (32.6% of 

 which were ovigerous) and 3,412 were males. This 

 corresponds to an overall male to female sex ratio 

 of 1:1.16, departing significantly from equality (P 

 < 0.0001). Measurements of TL, CW, and weight 

 were obtained from 5,920 of the shrimp sampled. 



Due to an imbalance in sampling, the effects of 

 location and time on the distribution of H. laevigatus 

 could not be completely separated. We therefore 

 assume that all samples were drawn from statis- 

 tically homogeneous locations in order to isolate and 

 examine temporal and depth effects. The strength 

 of this assumption is based largely on our personal 

 observations and those of fishermen that seasonal 

 change seems to account for most major population 



916 



