Observations of Deepwater , ensifer, From a Submersible off the Island of Hawaii

REGINALD M. GOODING, JEFFREY J. POLOVINA, and MURRAY D. DAILEY

Introduction ensifer which has a shallower preferred data acquired by other means and in de­ depth range of 300-600 m (Clarke, 1972; veloping more efficient gear and fishing In Hawaii and other Pacific island Struhsaker and Aasted, 1974; Gooding, techniques. groups there has been considerable inter­ 1984; Dailey and Ralston, 1986), have In February 1984, we had the opportu­ est in the commercial potential of deep­ been targets of trap fisheries. However, nity to use the Hawaii Undersea Research water caridean of the Het­ H. laevigatus is generally considered to Laboratory (HURL) submersible, erocarpus. Exploratory trapping surveys have the greater commercial potential be­ Malwlii (Fig. I), to make observations of have shown that deepwater shrimps are cause it is larger. H. ensifer off the northeast coast of the widely distributed throughout the central At present there is no Fishery Manage­ Island of Hawaii. The project was con­ and western Pacific (Clarke, 1972; Struh­ ment Plan (FMP) for deepwater shrimps. fined to H. ensifer grounds because the saker and Aasted, 1974; Wilder, 1977; The Western Pacific Regional Fishery maximum operational depth of the Intes, 1978; Brown and King, 1979; Management Council (Council) has iden­ Malwlii is less than the depth at which H. King, 1980, 1981a, 1981b, 1982, 1984, tified the Heterocarpus resource as one in laevigatus occurs. The objectives of the 1986; Moffitt, 1983; Gooding, 1984; need of further basic research for effec­ dives were to: I) Determine factors that Dailey and Ralston, 1986; Ralston, 1986; tive future management!. The Honolulu might cause variation in H. ensifer catch Moffitt and Polovina, 1987). Laboratory is presently conducting rates within a string of five traps, 2) ob­ The Southwest Fisheries Center's "intensive fishing experiments" on iso­ serve the behavior of shrimp in the vicin­ (SWFC) Honolulu Laboratory of the Na­ lated H. laevigatus grounds to acquire es­ ity of two different trap designs, tional Marine Fisheries Service, has con­ timates of absolute and relative abun­ 3) document observed distribution and ducted surveys of deepwater shrimp re­ dance of exploitable stocks (Ralston, abundance of shrimp relative to type of sources in Hawaii (Gooding, 1984) and 1986). For management purposes there is substrate, 4) set bait on the bottom as an the Mariana Islands (Ralston, 1986; Mof­ a need for research into population biol­ attractant to acquire further insight into fitt and Polovina, 1987). ogy and stock assessment, growth rates shrimp density relative to type of sub­ Despite the recent economic failure of and natural mortality, and migration pat­ strate, and 5) collect shrimp specimens, a large commercial endeavor that was en­ terns. Commercial catch and effort data bottom sediment, and water samples for a gaged in a trap fishery for Heterocarpus and research survey data from surface laboratory assay of incident bacteria. A laevigatus in Hawaii, (Schlais, 1982, vessels provide much of this information. problem for the shrimp industry that has 1983) the deepwater shrimp resource Visual observations from a research sub­ limited the marketability of Heterocar­ continues to offer promise for commer­ mersible can give a valuable insight into pus spp. has been a rapid deterioration in cial exploitation in Hawaii and indeed areas such as general behavior and mi­ the meat, so-called "mushy tail." This is throughout the southern and western crodistribution relative to types of fishing believed to be caused by bacteria of the Pacific islands. In the Hawaiian Islands gear and the nature of the bottom (Ral­ genus Vibrio. Four of the genus H. laevigatus, which occurs most abun­ ston et aI., 1986). Such direct observa­ had been identified by one of the authors dantly in depths of 450-700 m and H. tions may be very useful in interpreting (Dailey) in cultures of shrimp landed from commercial traps. We hoped to de­ termine if these bacteria are either always lWestern Pacific Regional Fishery Management on the shrimp and cause no problem for Reginald M. Gooding and Jeffrey J. Polovina are Council. 1984. Draft assessment of resources, the living , or are subsequently with the Honolulu Laboratory, Southwest Fish­ existing and potential fisheries, and management eries Center, National Marine Fisheries Service, needs for selected species in the picked up in the upper water column, or NOAA, 2570 Dole Street, Honolulu, HI 96822­ Western Pacific region. Rep. on file at Western even after the shrimp has been landed on 2396. Murray D. Dailey is with the Department Pacific Regional Fishery Management Council, of Biology, California State University at Long 1164 Bishop Street, Suite 1405, Honolulu, HI the vessel. Beach, Long Beach, CA 90840. 96813. Because the Malwlii became available

32 Marine Fisheries Review for the project on very short notice, we were able to fabricate and transport only six shrimp traps to Hawaii from Oahu. Four traps were lost early in the project so we were unable to acquire any infonna­ tion on the catch of similar traps on a string in relation to substrate variations. Methods The Makalii is a two-man, battery­ powered, one-atmosphere submersible owned and operated by HURL, Univer­ sity of Hawaii, and funded by the Na­ tional Oceanic and Atmospheric Admin­ istration (NOAA), National Undersea Research Program (NURP). It is 4.8 m long and has a spherical capsule 1.5 m in diameter. With a pilot and one observer and up to 95 kg payload, it has an opera­ tional depth capability of about 366 m and a dive duration of 4-5 hours with Figure I.-The NOAA submersible, Makalii. emergency life support for 72 hours. Nonnal operating speeds range from 1 to 3 knots. During this study, equipment that was used included: Hydraulic manipulator, two-color video cameras with monitors, recorders, and video lights, externally mounted 35 mm still camera and strobe, current and temperature meters, dicta­ phone tape recorder, directional antenna, and sonic pingers for site relocation. A Motorola Mini-Ranger Falcon2 naviga­ tion system provided very precise (within 15 m) position fixes for the support ves­ sel which tracks the submersible with an Edo Western submersible tracking sys­ tem. The Makalii is also equipped with a Neil Brown environmental monitoring package for continuous recording of tem­ perature, salinity, conductivity, pH, oxy­ gen, and depth. While on the bottom, observations were continuously video- and voice-recorded and still photographs were frequently made with the 35 mm camera and strobe. Observations were made on two types of traps. The first was a half-rounded de­ sign with an entrance at each end (Fig. Figure 2.-Half-round shrimp trap. 2). This design has been used during past shrimp surveys by the SWFC Honolulu Laboratory (Gooding, 1984). The second trap was a larger pyramid-shaped design (Schlais, 1983). Traps were baited with with a single entrance at the apex (Fig. Pacific mackerel, Scomber japonicus, 2Mention of trade names or commercial fllTl1s does not imply endorsement by the authors or the 3). This was the preferred trap used in the and were set from the submersible sup­ National Marine Fisheries Service, NOAA. fonner Hawaiian commercial fishery port vessel. Attached sonic pingers en­

50(1). 1988 33 Results and Discussion Figure 4 shows the location of the dive t sites. Depths ranged from 345 to 380 m, E and bottom temperature was 7°-8°C. (')" cO... Within these narrow ranges, temperature t and depth were not correlated with either the apparent abundance of H. ensifer as observed from the submersible or the size of trap catches. Bottom currents were 0.2-0.3 knot (10-17 cm/second). The dives showed that for the most part the bottom was a flat basalt plain largely con­ sisting of a shallow layer of brownish silty sand overlying volcanic lava flows. 1~ " Occasionally the nearly flat relief was in­ (J;) terspersed with small areas of low lava outcrops covered with limestone cement rising to about 20 cm above the sand. 1 Although there was considerable amount of sediment and apparent turbidity in the 2.54 x 2.54cm upper and midwater column, water clar­ WIRE MESH SPOT-WELDED TO ity at the bottom was very good except STEEL 'RE-BAR" when the fine silt was disturbed by the submersible. Trap Observations Five dives (Makalii dive No. 216­ 220), totaling 16.75 hours of bottom ob­ servation time, were made during 5-13 Figure 3.-Pyramid shrimp trap. February 1984. Table I summarizes the schedule of dive and trapping operations. Prior to the arrival of the traps from Hon­ olulu, on dive No. 216 we conducted a preliminary survey to determine a suit­ abled the submersible to locate them and the plastic bag tom away with the able area for subsequent trap sets and ob­ when it reached the bottom. Submersible manipulator. servational dives. The area surveyed on observations were made on trap sets For the bacteriological tests, individ­ dive No. 216 consisted almost entirely of shortly after setting (1.5-5 hours) or ual H. ensifer were collected from the silty sand overlying a hard flat substrate. about 24 hours after setting. The traps bottom with the submersible suction col­ There was a high incidence of H. ensifer were retrieved after either 24 or 48 hours lector. A bottom sediment sample was widely distributed throughout the region on the bottom. After removal from the collected from the area in which shrimp traversed by the Makalii. Shrimp densi­ traps the catches were held in crushed ice were located and a sample of water close ties were estimated at up to I per m3 in until later processing at dockside. Indi­ to the bottom was collected in a modified some areas. Subsequent operations were vidual shrimps were not weighed. Mean Van Dom bottle. In the laboratory 0.1 rnl conducted in the same general vicinity. individual weights were derived by di­ of raw inoculant from each of the sam­ During the following four dives (217­ viding total weight of catch by number of ples was plated on TCBS to determine 220) we made observations on H. ensifer shrimp. total Vibrio and on marine agar to deter­ in relation to varying combinations of The density of shrimp in an area was mine total bacteria. trap configuration, trap soak time, and estimated by monitoring the number of The submersible observations were bottom type. shrimp aggregating to a can of tuna with conducted on the Mauna Kea ledge about I) Dive 217. A string of five half­ numerous small punctures on all its sur­ 12 n.mi. northeast of Hilo harbor. This round traps, spaced 16 m apart. All of the faces during 10-minute intervals. The general area was recommended by Edith traps in the string landed upright in a cans were sealed in plastic bags and car­ Chave of HURL who had seen numerous depth of 347 m. They were clustered ried in the submersible's sample storage H. ensifer there on an earlier Makalii closely together on flat rubbly substrate basket. They were placed on the sea floor dive. interspersed with low outcrops. During

34 Marine Fisheries Review the observation period the traps had been 20·~=~"J7~======:i;i:=~s:==3~:::::======~'0!!UC====1 on the bottom for from 1.5 to 4 hours. 350 "'0 vdj' 110 '435 rln Only a few scattered H. ensifer were seen 1250 in the vicinity of the traps and the largest '" ,'00 1354 ". "7 1110 catch in anyone trap appeared to be about ...... ~-- IB' 910 , .. four shrimp. A survey in the vicinity of ,06..... "0 ". "0 ...... ·~,.~.,~013 the site showed the traps had landed on .3 ,,'" u' lB' 51. 2 .40 contiguously rocky rubbly ground appar­ 77 "5 ". , ••0 .40 ently devoid of silty sand substrate in the .0 ..... '35 "0 900 ...... 500 625 .00 "lOla vicinity. were not lB' I "0 610 920 ./ seen on the rocky ground but unidentified \12' "" .90 • 500 //" striped shrimp were visible in small caves I '95 "5 20. 525//'" lOBO "' in the substrate. These probably were 110"" "0 .00 '0'0 i Plesionika longirostris which were sub­ I 190 "0, 180 180 3'0 975 : 1365 sequently identified on dive 219. During .; ~"-'5 "lll"S60 ....~ '50 17 0 J"o 190 630 w:" /' ,'00 ,,30 retrieval of the traps the following day, '; ",":: 5' " 'J~I ••.(~, "0, ''', 311 ",,.­ ~t. s·!O~ the groundline broke and four of the traps :19 24 2' u ,.. ". "3 .. pS 1220 1240 • '18 2Q 0 (A SIt ~ J. ,,. 180 520 ./ 1160 in the string were lost. The recovered trap .... -,1, 20~\V 2,~ 33 '" ':' ~r<· ...:,o, '67 ~o.o . d 69 H ;r. ( . h d) 14 ~"-IO'" d,h.,-, 1.1(11)51)47 11 5~··- •. '-$ 140 ."·l~ 151! contalOe . ensljer not welg e . '," ·,lui- 35 \ 1480 1450 2) Dive 218. A single pyramid trap" 51 (400 which landed upright on sandy bottom at 51 \... 1480 '410 1200 a depth of 349 m. No rocky areas were " \ \ 1420 seen in the vicinity. The first observa­ ~\:: 62°"0 "~"\ '05. tions were made about 24 hours after the Hawaii 1420 trap had been set. The trap contained a 19·40··+------+-...:a-,.....--+-...... :·.:..··..l,·\-7:'2.,.,.5-.....,~---'''''1 large number of H. ensifer with many Scale-t250.000 shrimp climbing up the sides, and in the funnel entrance. Many others were on the Figure 4.-Dive sites off the Island of Hawaii (soundings are shown in fathoms). bottom surrounding the trap. The average size of the shrimp seemed so large that at first the observers thought that despite the relatively shallow depth, some of them Table 1.-Trapping observatlona 01 Heterourpua enalfer during MlJIuIIII dives 011 HawaII. were H. laevigatus. The trap was hauled Dive number the following morning after a soak-time 216 217 218 219 of about 48 hours. It contained 914 H. Observation 220 ensifer weighing 16.8 kg with a mean Depth of traps (m) 375·385' 347 349 347 347 individual weight of 18.4 g. No H. laevi­Bottom type Silty sand over Lava rubble and Silty sand over Silty sand ove' Silty sand over gatus were found in the catch. hard substrate low outcroppings hard substrate hard substrate hard substrate 3) Dive 219. A pyramid trap and a Trap soak time when No traps set 1.5·4 24-27 24·27 1.5·4 observed (h) half-round trap spaced 15 m apart were Trap soak time when 26 48 48 21 deployed together. Both traps landed up­ hauled (h) right on very fine silty sand overlying a Abundance of H. ensiler High Low High High High hard substrate. They were about 10 m observed in the area apart at a depth of 347 m. The traps had Abundance of H. ensiler High High High been set the previous morning so they observed in pyramid had been on the bottom for about 24 trap hours when observations began. Both Abundance of H ensiler Very few High High traps were filled with shrimp. Heterocar­observed in half·round trap pus ensifer were walking over the half­ No. of H. ensiler landed 914 1,497 1,676 round trap and on the sides of the pyra­ in pyramid trap mid trap (Fig. 5). The surrounding sandy No. of H. ensifer landed 69 1,016 1.110 bottom was covered with H. ensifer. As in half·round trap

had been observed during the previous 1Range of depths observed during the dive. trap set, the average size of the H. ensifer was uniformly large with very few small individuals in or around the traps. Shrimp

50(1), 1988 35 weighing 21 kg. The mean individual weights of H. ensifer were 19. 1 g in the pyramid trap and 18.9 g in the half-round trap. These observations showed that H. en­ sifer are attracted very rapidly to baited traps, and that a large percentage of the within olfactory range (depend­ ing on current conditions) may enter a trap within 2-3 hours of setting. Both traps with top entrances and those with entrances close to the substrate are effi­ cient in allowing H. ensifer to enter. However, if the shrimps are satiated or the bait depleted, they apparently can more easily exit a trap with a horizontally oriented entrance near the bottom than one with a vertical top entrance. Thus small traps with bottom entrances may be more suitable for relatively short soak pe­ riods. For instance, an early morning set, Figure 5.-Half-round trap on the bottom in a depth of 347 m; dive 219. and haul before noon. For longer soaks (overnight) a taller trap with a top en­ trance may hold the catch more effi­ ciently. were frequently seen both entering and half-round trap spaced 15 m apart. Both King (1981) cited data which suggests exiting the half-round trap. When they traps landed on typical silty sand ground that H. ensifer and H. laevigatus only entered the conical entrance, which on about 10 m apart in a depth of 347 m. The enter traps during night. In surveys by the this type of trap was at ground level, they half-round trap was upright but the pyra­ SWFC Honolulu Laboratory, good usually climbed in. However, they also mid trap was on its side. The Makalii catches of H. laevigatus were made dur­ 3 were observed swimming in and out. Al­ reached the traps about 1.5 hours after ing daylight sets . From deepwater photo though shrimp sometimes swam to they had been set earlier that morning. sequences of baited traps made off Palau, heights approaching 2 m above the bot­ The surrounding area was already cov­ Saunders (1984) found that the upper tom, none was observed swimming into ered with H. ensifer and shrimp were range of H. ensifer appeared to be the apical entrance of the pyramid trap. crawling over both the traps. Despite a strongly influenced by daily photic fluc­ They were observed in various stages of soak time of less than 2 hours they both tuations. They were abundant in night se­ climbing up the sides of the trap towards appeared to contain more shrimp than we quences at 150-250 m depths, but were the entrance, and into the entrance. No had thus far seen. As on the previous not recorded in daylight sequences shal­ shrimp were observed exiting the pyra­ dive, shrimp were both entering and exit­ lower than 274 m. However, at greater mid trap. A survey of the bottom 10 the ing the half-round trap. Unfortunately, depths there appeared to be no differ­ vicinity around the traps showed it con­ because of the danger of becoming fouled ences in either activity or numbers of in­ sisted mostly of silty sand. The exception in the groundline, we were unable to get dividuals in day versus night photo se­ to the otherwise flat sandy topography into a position to observe the entrance of quence. Our trap sets were all well below was a small area of lava outcrops about the pyramid trap which in its overturned 274 m. It is also likely that photic pene­ 30 m from the trap site. There solitary position lay close to the bottom. tration is relatively limited off the east striped shrimp, P. longirostris, were in During the 2.5-hour period between coast of Hawaii, which is subjected to holes and cracks in the hard substrate. the first and last observations of the half­ considerable natural runoff as well as The traps were hauled the following round trap there did not seem to be a sugar factory effluent compared with morning after being in the water for about noticeable increase in H. ensifer around Palauan waters. This study showed con­ 48 hours. The pyramid trap caught 1,497 or inside the trap. When it was hauled 21 clusively that, during peak daylight hours H. ensifer weighing 28 kg. The half­ hours later, the catch, although substan­ at the depths observed, H. ensifer are ac­ round trap caught 1,016 H. ensifer, tial, did not appear to be much different tive feeders and will readily enter traps. weighing 19 kg, and 7 P. longirostris. from what we had observed from the sub­ The mean individual weight of H. ensifer mersible. The pyramid trap caught 1,676 3Honolulu Laboratory, Southwest Fisheries Cen­ taken in each trap was 18.7 g. H. ensifer weighing 32 kg and the half­ ter, 2570 Dole St., Honolulu, HI 96822-2396. 4) Dive 220. A pyramid trap and a round trap caught I ,110 H. ensifer Unpubl. data.

36 Marine Fisheries Review However, we are not aware of any inves­ ,tigation which has rigorously compared the relative merits of day and night trap­ ping for Heterocarpus spp, The shrimp catches in both trap types were H ensifer with the exception of the seven P, lon­ girostris caught in the half-round trap ob­ served on dive 219, No H. laevigatus were found in any of the catches, We were impressed by the uniformly large size of H ensifer viewed from the submersible and caught in the traps, The mean individual weights of H, ensifer . from the two types of traps were virtually identical, at about 18,8 g, This was quite large when compared with a mean weight of 12 g for H, ensifer taken during trap­ ping surveys in the NWHI (Gooding, 1984), and a maximum weight for H, en­ sifer of about 16 g reported by Struhsaker and Aasted (1974), Some studies have Figure 6.-Heterocarpus ensifer aggregated around tuna can in a depth of indicated that large H ensifer occur 347 m; dive 220. within the depth range of maximum abundance, with smaller animals occur­ ring both shallower and deeper (Clarke,

1972; Struhsaker and Aasted, 1974; Table 2.-Rate of Heterocarpus ens/fer aggregation to a punctured tuna can as the number of animals King, 198Ia), However, Moffitt and observed over a 1o-mlnute period.

Polovina (1987) found no significant Rocky outcropped bottom Sandy rubble Silty sand bottom change in the size of H, ensifer with Elapsed bottom time Dive 219 Dive 219 Dive 220 Dive 219 Dive 219 Dive 220 Dive 220 depth, Our observations were made at the (min.) Test 1 Test 2 Test 1 Test 3 Test 4 Test 2 Test 3 shallow end of the range of maximum 1 0 0 1 2 2 7 abundance for H ensifer found by Good­ 2 0 0 1 5 8 14 3 0 0 1 9 13 20+ ing (1984), The submersible surveys re­ 4 0 0 1 12 20 30+ 5 0 0 2 20+ 30+ vealed virtually no small individuals in 6 1 0 2 30+ the vicinity of the traps or the surround­ 7 1 0 2 8 1 0 2 ing area, Thus we speculate that smaller 9 1 0 2 size classes were occupying other habi­ 10 1 0 2 tats, The shrimp were not methodically sexed but inspections of the catches showed the presence of both sexes with a high incidence of berried females, On H. laevigatus surveys in the Mari­ Tuna Can Observations them left their holes to approach the can. ana Islands, Ralston (1986) found that On the sandy ground there was initially a pyramid traps outperformed half-round Punctured tuna cans (Fig. 6) were de­ rapid increase in the aggregation rate of traps by a ratio of nearly 6: I. In this ployed on three substrate types: Silty H. ensifer with time which appeared to study, for the two sets (dive 219 and sand, rocky outcroppings, and sandy rub­ start leveling off after about 7-8 minutes. 220), when pyramid and half-round traps ble. Table 2 lists the number of H. ensifer We were unable to make counts beyond lay adjacent to one another on the same which aggregated to the bait in each test about 30 animals. These crude tests type of ground, the pyramid traps caught during 10 minutes of observation. The H. clearly indicate a much higher density of only 1.5 times (60 kg) the catch of the ensifer were apparently following the H. ensifer on the sandy silt substrate than half-round traps (40 kg). odor gradient and invariably approached on the two others. On dive 220 the pyramid trap that was the can from the down-current direction. lying on its side fished equally well rela­ On one rocky ground test the can was Other Behavior tive to the attached half-round trap as did placed about 2 m directly up-current of Heterocarpus ensifer apparently were the upright pyramid observed on dive an outcropping in which we could see neither attracted, repelled, nor affected in 219. several P. longirostris; however, none of any way by the lights from the sub­

50(1), 1988 37 mersible. When, after a period of total countered on Mauna Kea ledge, i.e., low The indication of a 4+ growth on the H. darkness, the flood lights were turned on profile outcropping and silty sand overly­ ensifer compared with a 2+ growth in the to large aggregations of H. ensifer gath­ ing flat substrate, the preferred type of surrounding bottom material implies that ered around traps, we were unable to dis­ ground for H. ensifer was clearly the lat­ the shrimp are likely a primary substrate cern any changes in their behavior. On ter. Of the six traps we recovered, the for the Vibrio. A 4+ on the marine agar dive 220 we had the opportunity for just two half-round traps that landed on flat indicates other bacteria, primarily Pseu­ a few seconds to observe H. ensifer in sand had a mean catch rate of 20 kg per domonas spp. are ubiquitous on the bot­ something approaching an undisturbed trap whereas the half-round trap observed tom as well as in the water column. natural state. As the Makalii, following on rocky ground caught only 69 H. en­ Vibrio alginolyticus, V. para­ its descent from the surface, approached sifer (probably about 1.2 kg). The three haemolyticus, and Vibrio sp. have also to within about 3 m of the flat sandy bot­ pyramid traps, all of which were on sand, been found commonly on marine mam­ tom, we were able to see H. ensifer all had a mean catch rate of 25.6 kg per trap. mals inhabiting the upper water column over the sandy surface out to the limits of Conversely, with the exception of a sin­ (Dailey, 1985). This would tend to verify visibility. They were solitarily dis­ gle P. longirostris observed on the bot­ that animals rather than the water column tributed, about I per m2. As the sub­ tom in front of the half-round trap on dive or bottom serve as the primary substrate mersible settled on the bottom, some of 219, we did not see any other of this spe­ for these microorganisms. the shrimp which were within about 1-2 cies on sandy ground. Seven P. longiros­ m from the Makalii swam vertically to as tris were caught in the trap which was Conclusions high as about 2 m from the bottom. More lying about 30 m from an area of exposed Because of various logistic factors the distant shrimps, although suddenly sub­ hard substrate on which P. longirostris observations and tests conducted during jected to greatly increased light, re­ were seen under ledges and in small these dives were somewhat limited in mained undisturbed. Our impression was holes. No P. longirostris were caught in scope and sophistication. Nevertheless, that the disturbed H. ensifer were react­ the pyramid trap on the same string. the results show that valuable insights ing to the sudden physical presence of the However, the half-round trap on dive 217 into trap siting and trap design relative to submersible in their midst rather than to that lay on rocky ground where we had deepwater shrimp ecology and behavior the light. seen P. longirostris, did not catch any P. and also more general aspects of the biol­ We did not see any burrowing in the longirostris. Unfortunately, the other ogy of shrimp can be acquired using a sand by H. ensifer on this occasion or four traps on the same string were lost. research submersible. This type of infor­ during other observations. This supports Bacterial Tests mation can complement and supplement aquaria observations which indicated that data collected by more conventional Heterocarpus sp. did not burrow in the The results of the bacteriological cul­ methods. substrate (King, 1986). ture assays (Table 3) indicate that Vibrio The HURL has recently acquired the It is well known from the condition of alginolyticus and another Vibrio sp. Pisces V, a three-man, one-atmosphere shrimp in trap catches that Heterocarpus found commonly in and on both H. en­ submersible that can reach depths of are cannibalistic. We did not usually ob­ sifer and H. laevigatus are apparently not 2,000 m. Planned SWFC Honolulu Lab­ serve any overt aggressive behavior acquired in the trap during hauling or oratory studies on H. laevigatus from the amongst shrimp under the various condi­ after the shrimp are landed on the vessel. Pisces V will provide information about tions of our study. However, in two in­ this potentially more valuable species stances it was shown vividly that an H. which, we hope, will contribute to the ensifer that finds itself at a disadvantage development and management of deep­

to its fellows quickly becomes fair game. Table 3.-Reaults of bacteriological tests to determine water shrimp fisheries in Hawaii and When the tails of animals within a trap the distribution of Vibrio sp. other Pacific island areas. protruded through the mesh, shrimp on Item TeSS Marine agar Acknowledgment the outside would start feeding on the Shrimp 4+ 4+ tail, and on one occasion during observa­ Vibrio alginolyticus and We would like to thank the members one other Vibrio sp. (not tions of shrimp attracted to a tuna can, an identified) of the Hawaii Undersea Research Labo­ H. ensifer which was injured by the sub­ Bottom ratory program for their excellent sup­ mersible's manipulating arm was imme­ sediment 2+ 4+ port. One Vibrio sp. found. The diately attacked by other shrimp in the same as the unidentified area. species from the shrimp Literature Cited

Water Brown, I. W., and M. G. King. 1979. Bottom Habitat column 0 4+ Deepwater shrimp trapping project. Report on No Vibrio sp. recovered. Phase I. Minist. Agric. Fish., Fish. Div. Our observations from the Makalii, Tech. Rep. I, Fiji, 20 p. the subsequent trap catches, and the tuna 1+ ~ 100 colonies/ml. Clarke, T. A. 1972. Exploration for deep benthic 2+ ~ 100-500 colonies/ml. fish and crustacean resources in Hawaii. Univ. can aggregation tests showed that of the 3+ ~ 500-2.000 colonies/ml. ~ Hawaii, Hawaii Inst. Mar. BioI., Honolulu, two distinct types of substrate we en- 4+ 2.000 colonies/ml. Tech. Rep. 29, 18 p.

38 Marine Fisheries Review Dailey, M. D. 1985. Diseases of marine in Vanuatu: A preliminary survey off Port ment for the caridean shrimp Heterocarpus mammalia: Cetacea. In O. Kinne (editor), Vila. Mar. Fish. Rev. 43(12):10-17. laevigatus at Alamagan Island in the Mariana Diseases of marine animals, p. 805-846. BioI. ____. 1981b. The deepwater shrimps of Archipelago. Fish. Bull. (U.S.) 84:927-934. Anstalt Helgol. Hamburg. 14(2), Chapter 7. Tonga: A preliminary survey near ____, R. M. Gooding, and G. M. Ludwig. ___,..---- and S. Ralston. 1986. Aspects of the Nuku'alofa. Rep. Ins!. Mar. Resour., Univ. 1986. An ecological survey and comparison of reproductive biology, spatial distribution, South Pac., Suva, Fiji. 28 p. bottom fish resource assessments (sub­ growth, and mortality of the deepwater ____. 1982. Report on the South Pacific mersible versus handline fishing) at Johnston caridean shrimp, Heterocarpus laevigatus, in Commission deepwater shrimp assessment Atoll. Fish. Bull. (U.S.) 84:141-154. Hawaii. Fish. Bull. (U.S.) 84:915-925. consultancy in Papua New Guinea. South Pac. Saunders, W. B. 1984. The role and status of Gooding, R. M. 1984. Trapping surveys for the Comm., Noumea, New Caledonia, 24 p. Nautilus in its natural habitat: Evidence from deepwater caridean shrimps, Heterocarpus ____. 1984. The species and depth deep-water remote camera photosequences. laevigatus and H. ensifer, in the Northwestern distribution of deepwater caridean shrimps Paleobiology 10:469-486. Hawaiian Islands. Mar. Fish Rev. (, ) near some southwest Schlais, J. F. 1982. Deepwater shrimps, a new 46(2): 18-26. Pacific islands. Crustaceana 47(2): 174-191. Pacific fishery. Sea Frontiers, Nov.-Dec. Intes, A. 1978. Deep trap fishing in New ____. 1986. Fishery resources of Pacific ____. 1983. Fledgling Hawaiian shrimp Caledonia and adjacent islands. Preliminary island countries. Part I. Deepwater shrimps. fishery could someday rival the Gulf catch. experiments (Peche profonde aux casiers en FAO Fish. Tech. Pap. 272.1, 45 p. Natl. Fisherman 64(3):8. Nouvelle Caledonie et iles adjacentes. Essais Moffitt, R. B. 1983. Heterocarpus longirostris Struhsaker, P., and D. C. Aasted. 1974. Deep­ preliminaires.) O.R.S.T.O.M. (Noumea) MacGilchrist from the Northern Mariana Is­ water shrimp trapping in the Hawaiian Is­ Rapp. Sci. Tech. 2, 10 p. lands. Fish. Bull. (U.S.) 83:434-436. lands. Mar. Fish. Rev. 36(10):24-30. King, M. G. 1980. A trapping survey for ____ and J. J. Polovina. 1987. Distribu­ Wilder, M. J. 1977. Biological aspects of the deepwater shrimp (Decapoda: natantia) in tion and yield of the deepwater shrimp Hetero­ fisheries potential of two deep water shrimp, Western Samoa. Rep. Ins!. Mar. Resour., carpus resource in the Marianas. Fish. Bull. Heterocarpus ensifer and H. laevigatus in Univ. South Pac., Suva, Fiji, 26 p. (U.S.) 85:339-349. waters surrounding Guam. M.S. Thesis, ____. 1981 a. Deepwater shrimp resources Ralston,S. 1986. An intensive fishery experi­ Univ. Guam, 79 p.

50(1), 1988 39