Deepwater Shrimp Resources in Vanuatu: A Preliminary Survey off Port Vila

MICHAEL G. KING

Introduction Carids are caught commercially off taken in April 1980 to examine deep­ Alaska (Barr, 1970), California (Dahl­ water shrimp resources near Vanuatu Natantian decapod custaceans repre­ strom, 1970), and Chile (Hancock and (previously known as the New Hebrides). sent one of the world's most commer­ Henriquez, 1968). In several other areas Vanuatu, including the Banks and cially valuable group of marine species. experimental trapping for deepwater Torres group of islands, consists of about This group consists of two large sub­ shrimps has been carried out: Hawaii 40 mountainous islands, besides numer­ groups: 1) Penaeid shrimps which form (Struhsaker and Aasted, 1974), Guam ous islets and rocks. They lie approxi­ the basis of valuable trawl fisheries in (Wilder, 1977), New Caledonia (Intes, mately between lat. 13° and 20V20S and many tropical and subtropical countries, 1978), Fiji (Brown and King, 1979), and long. 1661ho and 1700 E(Fig.1). Vanuatu and 2) carid shrimps, which are more Western Samoa (King, 1980). appears to be situated on a bank with widely distributed but less commonly The distribution of deepwater carids depths of about 650-750 m and which exploited. Interest in the exploitation of in the above countries suggests that simi­ runs approximately in a northwest to deepwater carid shrimps has recently lar resources may exist in other PacifIc southeast direction (PacifIc Islands Pi­ developed, however. Islands with suitable offshore bathyme­ lot, 1969). Carid biology is not well known, al­ try. This preliminary survey was under- Vanuatu is of volcanic origin and cor­ though a few species, which form the al formations are generally restricted to basis of commercial fIsheries, have been isolated and fringing reefs rather than studied. Many carids are deepwater, Michael G. King is Lecturer in Fisheries Biolo­ barrier reefs. In many islands, depths gy, Institute of Marine Resources, The Univer­ benthic species living on the sea floor sity of the South Pacific. P.O. Box 1168, Suva, drop to over 200 m within less than 1 and may make vertical diel migrations Fiji. n.mi. from the coast. For logistic reasons, up into the mid-water masses. The fe­ males carry the eggs externally (in con­ trast with penaeid shrimps) under the abdomen on the pleopods, which are especially adapted for the purpose. The life cycle of carid shrimps appears to be more extended than that of most penaeids.

ABSTRACT-Baited traps were used to survey deepwater shrimps in Vcmuatu in the Southwestern Pacific. Two-entrance box traps (some covered with hessian sacking; the others uncovered) were set in depthsfrom 229 m to 650 m out from Port Vila. Shrimp catch weight did not differ significantly be­ tween covered and uncovered traps but dif­ fered markedly with depth. Mean catch rates variedfrom 0.04 kg pertrap in the 200-300 m depth range to 2.83 kg pertrap in the 500-600 m depth range. Seven species ofcarid shrimp were provisionally identified from the catch. In terms ofabundance, Plesionika longiros­ tris, Heterocarpus ensifer, and H. laevigatus had the greatest commercial potential. Ovigerous specimens of Plesionika longiros/ris.

10 Mari'le Fisheries Review lSOOE . Efote Guam

D·

Port Vila

~\ ~ ua . '4ao '" = I. OFiji New~ Caledonia

30'

o .650 I 3n.mi.

Figure 1.-Vanuatu and other island countries mentioned in Figure 2. - The survey area near Port Vila. The position of the the text. depth prolile presented in the results (Fig. 5) is indicated by the broken line. Trapping locations are marked with black dots and the depth in meters. (Source: Chart 4787, Service Hydro­ graphique de la Marine, Paris.)

the preliminary survey was confined to waters adjacent to Port Vila the capital of Vanuatu on the Island of Efate (Fig. 2). Efate is a mountainous island, espe­ T cially in the northwest, with Mount 110 40 Macdonald reaching 672 m, and sloping 1 down to Efate Bay on the southern part of the island. Port Vila is a well sheltered harbor on the inner (northeast) part of Mele Bay.

Fishing Gear and Methods Traps The basic trap design was that of a box 80X70 cm and 40 cm deep (Fig. 3). The trap framework was welded from 8 mm diameter steel rod, reinforced by two diagonal cross bars on the base and one longitudinal rod, to which the ac­ Figure 3. - The basic trap design used in the survey. A and Brepresent the top plan cess gate was hinged on the top. The and the side view of the trap, respectively. entry cones at each end were tapered in from the edges and extended about 24 cm into the interior of the trap. The inner opening of the cone was 10 cm in The trap frames and cones were cov­ oval mesh size of 10 mm maximum di­ diameter. ered with molded plastic screen with an ameter. On about one- third of the traps,

December 1981, 43(12) /I Figure 4.-The ar­ via a short length of rope to prevent the rangement of the traps from being lifted by the pressure of trap fishing rig used in the survey. The wind and current against the floats and length of drop line rope. used (i.e.. between buoys and bottom Fishing Vessel and Equipment weight) was equiva­ The boat used was a 8.5 m wood and lent to the depth of the water plus an ex­ plywood catamaran powered by a 25 cess of 25 percent h.p. (18.7 kw) outboard engine. The (minimum). winch concisted of a 3 h.p. Briggs and Strattonl 4-stroke engine which powered a geared trap hauler via a belt drive. The trap hauler was of the type used in the Australian rock lobster trapping fishery (i.e., in which the trap line is ftrmly gripped by a pulley, and stripped off by a Polypropylene rope metal peeler). (10 mm dial A skipper model 603 echo sounder fttted with a narrow- beam ferrite trans­ ducer was used to obtain depth proftles and measure the depth at each trap drop. The echo sounder was rigged as a port­ able unit with a separate 12-volt lead­ :-- 15 m -,I acid battery and the transducer was mounted on the end of a length of pipe. A hand compass was used to ftx the position from shore marks at each drop location. ,.::,:. Bait Chain I intended to use the same bait (10 kg) throughoutthesurvey,preferablyskip­ jack tuna, Katsuwonus pelamis, as this was the most successful of several baits tested in Fij{ However tuna was not always available in Vanuatu and, in the initial trap sets, a variety of baits was the sides, bottoms, and tops (i.e., exclud­ ond marker buoy. The second marker used including boneftsh, Alhula vulpes; ing the cones) were covered with hessian buoy was lashed to a 3 m wood or bam­ Indian mackerel, Rastrelliger kanagurta; sacking. It has been claimed that "cov­ boo pole fitted with a brightly colored and sardines, Sardinella spp. Sardines ered" traps (i.e., covered with hessian or flag and a 4 kg scrap chain weight (Fig. were the most easily obtained local bait burlap) are more efficient in catching 4). The buoy was attached approximate­ and this species was used throughout shrimp than "uncovered" traps (Struh­ ly 1.8 m from the top of the pole. A the remainder of the survey. About 500 saker and Aasted, 1974). Reasons for "man-overboard" flashing light was at­ g of bait was placed in plastic netting this could include the hessian cover, tached to the pole above the buoy. containers (about 10 mm mesh size) and either concentrating substances from The drop line was of 10 mm diameter suspended in the center of each trap the bait at the entrances of the traps, or polypropylene and was made from coils (Fig. 3). providing additional shelter for the (about 210 m in length) which were shrimps (i.e., acting as "habitat" traps). joined together as necessary. Atone end The Fishing Operation of the bottom line, 4 loops were spliced Except during the initial trap con­ The Fishing Rig into the rope at 15 m intervals, the last struction stage and during certain peri- The trap ftshing line consisted of two loop being an eyesplice at the end of the parts: A drop line from the marker buoys rope. This spacing was incorporated to and a bottom line from which the traps decrease competitive effects between 'Mention of trade names or commercial firms were strung (Fig. 4). The top of the drop the traps. The last three loops were used does not imply endorsement by the National Marine Fishenes Service, NOAA. line was attached to a 30 cm diameter to attach the trap lines and, from the 'King, M. G. University of the South Pacific, P. buoy which was itself attached to a sec- fourth loop, a 10 kg weight was attached O. Box 1168, Suva, Fiji. Unpub!. data.

/2 Marine Fisheries Review ods of maintenance, two strings of three traps each were used. Each string of traps was left in the sea overnight and hauled in, emptied, and rebaited before being reset in a new location on the following day. At each location the traps and weight were lowered over the windward side of the boat. With the motor off, the boat would drift off to leeward with the trap acting as a sea anchor during the drop. Pressure was always kept on the drop line as the ropes were let out to avoid tangling the traps. On the rare, windless days, the boat's motor was used at the lowest possible speed to keep the pres­ sure on the drop line. During the drop, additional coils of rope or half coils were successively added to the drop line to provide sufficient rope for the depth of water (equivalent to the depth of water The FAG-designed catamaran used in the Vanuatu deepwater shrimp surwv. plus a minimum of 25 percent). At the end of the coil, the float buoy and mark­ er buoy were attached and cast off from the boat. On the following morning, the re­ corded landmarks were used to locate the marker buoy. Rope was hauled in 0------and flayed down in bins or directly into the catamaran floats. Once on deck, the catch from each trap was placed in sep­ arate plastic bags and labelled before i:~- storing on ice. 600 ~i==,;==:::;r==~,==::;:,==:;:, ==~, =~=;:=:....-....., Survey Area and Method o 6 Initial depth soundings were made in Distance (n,mi.) several areas and a full depth profile was recorded along a transect running out at Figure 5.-A depth protile measured hy the echo-sounder trace made a1pni! the 0 238 true bearing from Point Malapoa transect indicated in Figure 2. in MeIe Bay (Fig. 1). The survey was conducted by setting traps at various depths along the tran­ sect. Two strings of three traps were intended to be set over each night in predetermined depths (at 50 m depth was placed in a labelled plastic bag and Results intervals) from 200 to 700 m, but the stored in a portable ice box. Once ashore, OtIshore Bathymetry program had to be modified due to gear shrimp were separated according to spe­ and weather problems. cies and into ovigerous (egg- bearing) The echosounder was used to record At each set location the depth was and nonovigerous groups, before being the depths and sea bottom profiles in recorded both at the time of setting and counted and weighed. A large sample, several areas offshore from Mele Bay. hauling; the mean of these was used to usually the catch from one trap in each The bottom profile along the selected define the "estimated fishing depth." string, was set aside for measuring. transect shown in Figure 2 is presented Also, at each set location, a hand com­ Shrimp carapace lengths (distance be­ in Figure 5, which is drawn from the pass was used to record the bearings of tween the postorbital eye socket and the echosounder recording paper. shore marks. These and other details posterior median edge of the cephalo­ Temperature were recorded at sea on a daily log sheet. thorax) were recorded to the nearest The catch from each individual trap millimeter. Temperatures were recorded on a

December 1981. 4](12) , , Table 1.-Provisional list of scientific and common names of deepwater shrimps found in Vanuatu waters. One additional species is, so far. unidentified. SCIentific name Common name - Parapandalus serratilrons Pyjama shrimp Plesionika enS/S Striped gladiator shrimp' Plesionika longlfos/ris Stars and stripes shrimp He/erocarpus ensifer (sibogae?) Armed nylon shrimp Heterocarpus gibbosus Humpback nylon shrimp' He/erocarpus laevigatus Smooth nylon shrimp' 'Standard FAO name (Halthuis. 19801.

and second abdominal pleura in the lat­ ter species; however, this character ap­ pears to be highly variable. On the basis of this preliminary sur­ vey it appears that three of the species "4fl'1'.'I''''''IIIPI'flfl/llJr'Ijl~T'IIIP'T'Ill\II'I'QQlIll'llllIIT'tlllI'rlll1llT'I'IPI'I''liIIIl'I'1'11I11:r'I~III'I'~IIPI'['IIII"'l'''~ found may be present in sufficient quan­ ~ :8 0, f " 12 13 14 f!5 18 17 t8 19 21 22 23 2A Ie tity to be of commercial interest. These include the two abundant Heterocarpus Deepwater shrimps. clockwise from lOp. are Hell?rocQ/1JIIS 11I1!1·i~all". PI""01Iika species, H. ensifer and H. laevigatus. as longirosl,.is. H. ellsi/i?,. and H. ~ihh0.\·II\. well as the smaller Plesionika longiros­ tris. The length frequencies (from all 0 40 17.S C at the shallowest depth surveyed depths combined) of these three species (229 m) to 7.S oC at the deepest depth are shown in Figure 6. 30 (650 mi. Catch Analyses 20 As Heterocarpus ensifer occurred 10 Depth Minimum (m) temperature (Oc) over the entire depth range surveyed, it 229 17.5 384 110 was decided to analyze the catches of 454 9.5 this species by bothdepth and trap type. 650 7.5 (;'100 All catch data were grouped into three c Q) depth ranges: Shallow «350m); middle i} 80 (350-450m); deep (>450m). After cull­ J! Species Caught 60 ing suspect data from damaged traps, During the survey, seven species of the mean catch weights were tabulated 40 carid shrimps were provisionally identi­ in a 3X2 matrix of 3 depth ranges and 2 20 fied from the catch. All belonged to the trap types. family Pandalidae and at least three gen­ An analysisofthevariancein thedata era: Heterocarpus, Plesionika, and presented in Table2 is shown in Table3. Parapandalus. In common with other There was no significant difference in deepwater shrimps, these are character­ the mean catch rates of H. enslfer be­ istically colored pink to deep red and tween covered and uncovered traps o 10 20 30 40 50 Carapace length (mm) the ovigerious females often carry blue (F=O.O 1, P>O.OS) but catch rates eggs. varied significantly with depth (F= Figure 6.- Length- frequency dia­ The species found are listed in Table 22.S0,P<0.OS). grams of the three most abundant 1 in approximate order of increasing species of deepwater shrimp (samples size. Scientific names are provisional and Depth Distribution from all depths combined). common names are from Holthuis (1980) where possible or from King Analyses presented in the previous (1980). One species found has not yet section indicated that the catch weight been identified and the name of another of H. ensifer varied significantly with species, Heterocarpus ensifer, is provi­ depth. Using all catch data from the modilied maximum/minimum ther­ sional. Heterocarpus ensifer is distin­ survey, the contribution of various mometer attached to one of the traps guished from H. sibogae only by the shrimp species to the total catch at each used at each depth. These ranged from presence of dorsal carina on the lirst depth was calculated. These contribu-

14 Marine Fisheries Review Table 2.- Mean catch (kg) per trap of Heterocarpus Table 3.-A two-way analysis of variance in catches Table 4. - Species composition by weight expressed as ensifer and variance for each trap type at each of 3 (kg/trap) of H. ensifer by trap type and depth range. a percentage of the total shrimp catch in each depth. depth ranges. Source df 5.5. F Depth P. long;- H. H. Lae- Other Depth range Covered Uncovered Row means (m) rostris ensiler vigatus species Columns (traps) 1 2.67 0.D1 (P>0.05) Shallow X =0.06 004 0.05 229 15 54 31 «350m) 5' =0.05 0.04 Rows (depths) 2 189 22.50IP<0.05) 262 51 48 1 324 47 49 4 Middle X =095 1.23 109 Error 2 0.08 384 27 73 0 (350-450 ml 5' =0.89 026 421 <-1 99 <1 436 <1 99 <1 Deep X =150 120 135 ~ 2000 (>450m) 5' =0.67 0.16 461 <1 99 <1 560 57 43 0 Column rlleans 084 0.82 650 68 32 0

Figure 7.-The depth distribution (mean catch per trap) of the four most abundant deepwater shrimp caught during the "3 1 trapping survey. c a ~ 0 200 300 400 500 600 700 Depth range (m)

Figure 8. - Mean catch per trap (kg) per 100 m depth.

tions expressed as percentages are given the mean catch weight per trap as an ent survey, however, further research is in Table 4. index of relative abundance. Indices for necessary to discover how shrimp abun­ Species such as Parapandalus serrati-­ the four major species are shown in dance alters in depths >600 m. frons and Plesionilw ensis which oc­ Figure 7. These data are from a limited Size by Depth curred in low trap abundance are geographical area and collected over a grouped together as "other species." Ple­ short time scale; the fIgure may there­ Length-frequency data were exam­ sionilw longirostris accounted for up to fore not represent the general distribu­ ined to determine whether the size of about half the shallow water catch. Het­ tion throughout Vanuatu or over a com­ shrimp changed with depth. Heterocar­ erocarpus enslfer occurred throughout plete seasonal cycle. pus enslferwas the only species caught the whole sampled depth range and, in all of the depth samples; length­ Catch by Depth from about 400 to 500 m, the catches frequencydatafromselecteddepthsare consisted almost exclusively of this spe­ The catch datawere also examined to shown in Figure 9. This shows that cies. In the deepest depths sampled the determine how the total biomass of shrimp size differed markedly with proportion of H laevigatus in the catch shrimp caught relates to depth. These depth,withagreaterproportionoflarge increased with a corresponding decrease data were grouped in 100 m depth shrimp in the middle part (421-461 m) in the proportion of H ensifer. ranges and are presented as a histogram of the depth range sampled. Other species of shrimp usually ac­ in Figure 8. Although the total length- frequency countedfor <5 percentofthe total catch The histogram indicates that total diagram (samplesfrom all depthscom­ in any depth. An exception was in the mean catches were low in shallower bined) presented in Figure 6 show pos­ shallowest depth sampled (229 m) where water but increased with depth to a max­ sibly 3 major size classes (i.e., at modal the small shrimp Parapandalus serrati­ imum in 500-600 m. At this depth range lengths of 14,23, and 33 mm) for H frons accounted for 31 percent of the the mean catch rate was 2.83 kg per ensifer, Figure 9 shows that the propor­ small catch. trap. Beyond 600 m, trap abundance of tion of the shrimp in each size class The distribution of each species of shrimp appears to decrease. Taking into differs with depth. Samples from shal­ shrimp by depth was estimated by using account the limited nature of the pres- lower water (e.g., 324 m) consisted al-

December 1981, 43(12) /5 most entirely of small shrimp from the species. Figure 10 combines data from first and second size class. In middle all depths (n= 1,846) and shows the 80 range depths (e.g.,461 m) shrimp in the number of ovigerous individuals, ex­

70 smaller size classes were not abundant pressed as a percentage of the total but appeared again in samples from number, in each size class. At about 30 60 deeper water. mm carapace length, half the popula­ Surveys carried out in Hawaii (Struh­ tion consists of ovigerous females; at 50 saker and Aasted, 1974) also indicated >35 mm, thepopulationconsistsentire­

40 that smaller H. ensiferare found above ly of ovigerous females. Both H. ens/fer and below the depth ofmaximum abun­ and H. laevigatus may be protandrous 30 dance. hermaphrodites, existing initially as small males before changing sex to be­ 20 Sexuality come females. 10 As H. ens/fer was obtained over a 324m Discussion large size range (8-40 mm carapace o+---r''-,.---'~~-----':''''''303'540 length) it was possible to determine the At least six of the seven species of 20 proportion of ovigerous (egg bearing) pandalid shrimps found in the Vanuatu survey have been obtained in other Pa­ ~ 10 individuals in each size class for this ., 384m cificcountriesincludingFiji(Brownand I 0+-~--L,-~~---1'--~L;>__~ King, 1979), New Caledonia (lntes, 1978), Guam (Wilder, 1977), Hawaii 20 (Struhsaker and Aasted, 1974), and Western Samoa(King, 1980). Although 10 much of this literature makes no refer­

100, ence to the smaller shrimp, there is 90 someinformationonthelargerHetero­ 80 carpus species. The depth distribution 70 of H. ensifer and H. laevigatus in these ~ .,e 60 countries is shown in Figure 11. Hetero­ 0) .~ 50 carpus gibbosus, which accounts for a c 40 .., substantial proportion ofthecatches in .,~ .' a.. 30 Fiji (at about the 400- 500 m depth) was 20 only found in very small numbers in the 10 Vanuatu survey. .' Takingintoaccountthelimitationsot o 10 15 20 25 30 35 40 0 10 20 30 40 50 this preliminary survey, the catches of Carapace length (mm) Carapace length (mm) shrimp obtained were sufficient to en­ courage a further examination of the Figure 9.-Length frequency of sam­ Figure 10.-The percentage of oviger­ fisheries potential. In decidingon a fish­ ples of H. ensi/er from selected ous females in each size class (data depths. The depth is indicated at the from samples taken at all depths com­ ing strategy, it is not only necessary to right of each graph. bined: n=1.846. consider catch weight but also the size of individual shrimps. In H. ens/fer, for example, the highest proportionoflarg­ er individuals (from the third size class, Figure II.-The depth dis­ Fig. 9) was found in depths of about tribution (light line) and 450- 550 m. Although catch weight was depth of greatest abun­ dance (heavy line) of H. ell­ still high in depths greater than this, a ens~fe=r~~~~~§~~~.~ ~E=~F siler and H. laevigailis. large proportion of the deeper catch Heterocorpus~ .. "·D BC References are: Al Hawaii consisted of smaller shrimp (first and A (Struhsaker and Aasted. second sizeclasses). Trappingindepths 1974); B) Guam (Wilder, 1977,: C) New Caledonia of about 550 m appears ideal in that --~------F Heterocorpus loevigotus ~'''E (Intes. 1978); D) Fiji reasonableweightsoflargerindividuals -:::0 ",::;;:::::==:]B::A~---- (Brown and King. 1979): ofH. ensifer(>1.5 kg pertrap)aswellas C E) this paper; and F) West­ H. laevigatus (>1.0 kg per trap) were ern Samoa (King. 1980l. obtained. The greatest weights of a 100 200 300 400 500 600 700 800 900 1.000 Dotted lines indicate the Depth (m) limits of the respective shrimp (all speciescombined) obtained surveys. during the survey was in the 500-600 m

16 Marine Fisheries Review depth range where the mean catch rate winch. It would be possible to use shore Vanuatu Fisheries Division for the proj­ was 2.83 kg per trap. bearings in place of the echo sounder to ect. The help of Tony Jarvis, Alison Plesionika longirostris, although determine fishing locations and depths Baxter, Joseph Kalo, and Paul and Linda smallerthanHeterocarpusspp.,isworth but this method would depend on the Mackin was much appreciated. someconsideration from acommercial availability of charts with accurate At the University of the South Pacific, point of view. This species has a higher soundings; the method would be less the support of the Director of the Insti­ "recovery rate" (the weight of high­ reliable where the sea bottom profile is tute of Marine Resources, U. Raj, and value tail meat expressed as a percent­ as undulating as that found during this the assistance of Nand Ram was appre­ age of total weight) than the other spe­ survey (Fig. 5). ciated. M. Capra kindly offered advice cies; it is also obtained in shallower Anyexaminationofthecostsinvolved on the manuscript. water (about 350-400 m) where fishing should include special reference to the Financial support for the project was costs are obviously less than in deeper increase of costs with fIshing depth and provided by the Hanns Seidel Founda­ water. During this survey, however, the likely incidence of trap loss due to tion of the Federal Republic of Ger­ catches of P longirostris were always heavy weather, shark attack, etc. Dur­ many. Specimens of all shrimps caught <0.5 kg per trap. Further work is need­ ing the present survey two traps were were sent to B. Kensely at the Smithson­ ed todetermine the trapabundanceand damaged,presumablybysharks. Oneset ian Institution. depth distribution of this species in of traps was lost during the survey, pos­ Literature Cited Vanuatu. sibly due to an incorrectly tied knot at The economics of any proposed the float buoy. Thecostofbait may also Barr, L. 1970. Alaska's fishery resources: The shrimp trapping fishery should be care­ beconsiderable,al thoughsardines, Sar­ shrimps. Fish. Res. Board. Can., Fish. Leaf!. fully examined. It would bedesirable to 631,10 p. dinella spp.,werecaught by cast- netand Brown, I. w., and M. G. King. 1979. Deep­ usesmalliocallymadefishingcraft.The successfully used as bait in this survey. water shrimp trapping project: Report on 8.5 mcatamaran used in the preliminary Phase I. Fish. Div. Fiji, Tych. Rep. 1,30 p. From experiences in Fiji it has been Chart No. 4787 (1893). lie Efate, Archipel des survey was suitable for setting traps be­ Nouvelles - Hebrides. Service Hydrograph­ cause of its deck space and high initial found that the handling of deepwater ique de la Marine, Paris. stability. However, such a boat may not shrimp requires considerable care­ Dahlstrom, W. A. 1970. Synopsis of biological more so, for example, than finfish. Im­ data on the ocean shrimp Panda/us jordani be able to carry the total number of Rathbun, 1902./n M. N. Mistakidis (editor), traps needed for a commercial opera­ mediately after being caught, shrimp Proc. World Sci. Conf. BioI. Culture of tion. Due to the close proximity of the must be placed in crushed ice or chilled shrimps and prawns: Mexico City. FAO Fish brine, and sodium metabisulphite may Rep. 57(4): 1377-1416. fIshing grounds, it may be possible to Hancock, D. A., and G. Henriquez. 1968. Stock relay smaller numbers of traps (e.g., be used as a dipping solution to extend assessment in the shrimp (Helerocarpus strings of 15 traps at a time) out to the thecold storagelifeoftheshrimp.Cook­ reedt) fishery of Chile. In M. N. Mistakidis ing time for deepwater shrimp must be (editor). Proc. World Sci. Conf. BioI. Culture droplocations.Thedevelopmentofcol­ of shrimps and prawns: Mexico City. FAO lapsible traps would also help in this kept short; approximately 2 minutes in Fish Rep. 57(2):443-465. respect. boiling water, usually suffices. Holthuis, L. B. 1980. FAO species catalogue. Vol. 1. Shrimps and prawns of the world. An In Fiji, when traps were sometimes In conclusion, ithas been established annotated catalogue of species of interest to leftsetlongerthan forthe usual18- hour that deepwater carid shrimps are pres­ fisheries. FAO Fish. Synop. (1251 Vol. 1,261 p. (overnight) period there was consider­ entinVanuatuwaters.Catchratesmade Intes, A. 1978. Peche profonde aux casiers en Nouvelle-Caledonie et iles adjacentes: Essais able cannibalism. Results from a series during the preliminary survey off Port preliminaires O.R.S.T.O.M. (Noumea) Rap­ of 24- hour stations suggest that H. en­ Vila were high enough to consider car­ ports Scient. et Tech. 2. 10 p. H. King, M. G. 1980. A trapping survey for deep­ slfer and laevigatlls, in spite of their rying out commercial trials. Further water shrimp (Decapoda: natantia) in Western depth of distribution, only enter traps work would be necessary to elucidate Samoa. Rep. Inst. Mar. Resour., Univ. South during the night (footnote 2). Cannibal­ several points regarding the biological, PacifJc, Fiji, 26 p. ism was not evident in traps set over­ economic, handling, and marketing as­ Pacific Islands Pilot, Volume II. 1969. Ninth edi­ tion. Hydrographic Department. Royal Navy, night in my Vanuatu study. pects of any proposed fishery. United Kingdom. Other than the fIshing boat itself. Struhsaker, P., and D. C. Aasted. 1974. Deep­ Acknowledgments water shrimp trapping in the Hawaiian Is­ there are several other expensive items lands. Mar. Fish. Rev. 36(10):24-30. offIshingequipmentrequired including I wish to thank the Minister for Natu­ Wilder, M. 1. 1977. Biological aspects and !ish­ rope, an echo-sounder capable of de­ ral Resources, Thomas Reuben, and eries potential of two deepwater shrimps. Heterocarpus ensller and H. faelligalUs. in tecting the depths involved, and a me­ FAa fisheries adviser M. Louis Devam­ waters surrounding Guam. Master's Thesis, chanical or hydraulic pot- hauling bez for providing the facilities of the Univ. of Guam, 79 p.

December 1981. 4J(/2) /7