Abstract. - Five submersible 

 dives were conducted to evaluate the 

 behavior of deepwater shrimp and 

 the relationship of their density to 

 bottom type and trap yield. Differ- 

 ences in behavior of two species of 

 Heterocarpus were observed: H. en- 

 sifer tended to group around large 

 anemones and other benthic relief 

 over otherwise flat, sandy bottom 

 and were very active in the presence 

 of a baited container; whereas H. lae- 

 vigatus were solitary and showed 

 little activity around a baited con- 

 tainer. Greater densities of H. laevi- 

 gatus were observed on volcanic than 

 on coralline substrate, indicating a 

 possible association wdth this bottom 

 type. Trap catches were regressed 

 against observed H. laeingatus den- 

 sities yielding an estimate of the 

 catchability coefficient. This coeffi- 

 cient differed from that obtained 

 from a previously conducted Leslie 

 model depletion study. Factors con- 

 tributing to this difference may in- 

 clude comparing estimates of catch- 

 ability based on data from different 

 areas, bias in the estimate of catch- 

 ability based on observed density, 

 and bias in the estimate of catchabil- 

 ity from the depletion study. A com- 

 bined fishing and visual census study 

 is suggested as the best assessment 

 technique. 



An assessment of the exploitable 

 biomass of Heterocarpus laevigatus 

 in the main Hawaiian Islands. 

 Part 2: Observations from 

 a submersible 



Robert B. Moffitt 

 Frank A. Parrlsh 



Honolulu Laboratory, Southwest Fisheries Science Center 



National Marine Fisheries Service. NOAA 



2570 Dole Street, Honolulu, Hawaii 96822-2396 



Manuscript accepted 26 May 1992. 

 Fishery Bulletin, U.S. 90:476-482(1992). 



Tropical deepwater pandalid shrimp 

 have potential for commercial har- 

 vesting in many areas of the Pacific 

 (Struhsaker and Aasted 1974, Wilder 

 1977, Moffitt 1983, King 1984, Taga- 

 mi and Barrows 1988). These shrimp 

 are readily trapped but not easily 

 trawled (Struhsaker and Aasted 

 1974). The largest and most commer- 

 cially desirable species in Hawaii is 

 Heterocarpus laevigatus (Tagami and 

 Barrows 1988); the smaller H. ensifer 

 has less commercial appeal but is also 

 abundant (Struhsaker and Aasted 

 1974). In the early 1980s, several 

 boats initiated a trap fishery target- 

 ing H. laevigatus, and landings rose 

 to a high of 159 metric tons (t) in 1984 

 (HDLNR 1986). By 1985, most ves- 

 sels left the fishery, and the annual 

 landings dropped to <6t/yr (West. 

 Pac. Fish. Inf. Network, NMFS 

 Honolulu Lab., unpubl. data). 



Early predictions of maximum sus- 

 tainable yield for Hawaiian shrimp, 

 based on little or no direct data, were 

 as much as 1000-2000t/yr (Struh- 

 saker and Aasted 1974, HDLNR 

 1979). Recent research on population 

 dynamics combined with systematic 

 trapping surveys has resulted in 

 more refined estimates of exploitable 

 biomass and maximum sustainable 

 yield for H. la£vigatus in various 

 island locations (Dailey and Ralston 

 1986, Ralston 1986, Moffitt and Polo- 



vina 1987, Ralston and Tagami 1992). 

 The most recent estimate of exploit- 

 able biomass for the main Hawaiian 

 Islands, 271-1050 1, is based on an 

 estimate of the catchability coeffi- 

 cient (q) obtained through a Leslie 

 model depletion study, coupled with 

 catch-per-unit-effort (CPUE) values 

 and habitat area estimates obtained 

 through systematic trapping (Ralston 

 and Tagami 1992). 



The relationship of observed tar- 

 get-species density to fishing-gear 

 CPUE has been used to estimate 

 stock biomass and catchability (Rals- 

 ton et al. 1986, Kulbicki 1988). Esti- 

 mates of abimdance obtained through 

 visual census techniques are general- 

 ly higher than those based on catches 

 of fishing gear, and the relative reli- 

 ability of the various assessment 

 methods must be analyzed on a case- 

 by-case basis (Uzmann et al. 1977, 

 Powles and Barans 1980, Kulbicki 

 and Wantiez 1990, Matlock et al. 

 1991). 



In the present study, we conducted 

 submersible dives at several sites in 

 the Hawaiian Islands to observe 

 shrimp behavior both away from and 

 in the vicinity of a baited container 

 and to record density and substrate 

 associations of //. laevigatus. Obser- 

 vations of shrimp behavior and sub- 

 strate associations have applications 

 to commercial fishermen in terms of 



476 



