Tridacnid Stocks on Helen Reef, Palau, Western CaroUnels Sli~s

WENDY HIRSCHBERGER

Introduction lying at approximately lat. 3°N and Figure I.-Helen Island at Helen Reef long. 131°E (Fig. 1, 2), occupy about atoll, in Palau's southwest islands. Overexp]oitation of tridacnid clam 216 km 2 with a small island (Helen populations appears to be a current Island) located at the northern end. problem in many areas of the Indo­ Depths inside the lagoon exceed 60 m; remote area is uninhabited and receives Pacific. One area where some of the outside, the bottom slopes steeply. This only occasional visits from U.S. Trust effects of harvesting have been studied Territory outer island support ships and is Helen Reef, a small atoll in the south foreign fishing vessels. Palau District, Western Caroline Is­ Wendy Hirschberger is with the Northwest and In May 1971, the NOAA research Alaska Fisheries Center, National Marine lands, Trust Territory of the Pacific Is­ Fisheries Service, NOAA, 2725 Montlake vessel Townsend Cromwell, operated lands. The submerged reef and lagoon, Boulevard East, Seattle, WA 98112. by the National Marine Fisheries Ser-

ABSTRACT-A survey of Palau Distrids ratios of live individuals to empty shells showed less than 5 percent dead. These Helen Reef was conducted in May 1976 to continued to remain very low. Only Hip­ smaller species may possibly be used as continue monitoring changes in tridacnid popus hippopus showed a substantial de­ natural population indicators at Helen clam abundances. The densities of the four crease in the percentage dead. This is in Reef. is still rare, prob­ largest tridacnid species were slightly contrast to Tridacna maxima which is sub­ ably owing to poor environmental condi­ higher than observed in 1975; however, ject tn relatively no fishing mortality and tions for this species.

8 Marine Fisheries Review vice, NOAA, conducted a survey of the referred to as the"," as old Tridacna squamosa usually occurs on Trust Territory's marine resources. individuals may attain lengths of up to surfaces in depths less than 15 Only a stop was made at Helen Reef at 135 cm (about 4V2 feet). They usually m, most often in protected environ­ this time, and the ship returned in occur on sandy areas of the reef or in ments such as reef canyons and March 1972 to survey the tridacnid areas ofcoral rubble and reef degenera­ fissures, sheltered lagoons, and marine clam populations. In their report, Hes­ tion, within intertidal regions to ap­ lakes. These tridacnids are often refer­ ter and Jones (1974) concluded that the proximate depths of20 m. Although T. red to as "fluted " because of tridacnid populations at Helen Reef gigas exhibits a small byssus orifice, their characteristic broad, plate-like were large and that Tridacna gigas and the adults remain unattached to any projections. In size, they can range up T. derasa could possibly withstand a substrate. The dorsal margin and ribs to 45 cm (over 17 inches). Although moderate, controlled fishery. exhibit deep undulation. Tridacna de­ they are never imbedded, T. squamosa A resurvey effort was undertaken in rasa, also a species capable ofreaching are often found nestled among coral and April 1975 by a team of biologists sup­ large sizes, reportedly attain lengths of anchored with a weak but abundant ported by the Palau District Marine Re­ 50 cm (20 inches). These clams often byssus. Hippopus hippopus grow to 40 sources Office, a division of the Trust live towards the outer edge of coral cm (close to 16 inches) in size and usu­ Territory Department of Marine Re­ reefs and are usually unattached, al­ ally occur in sandy areas of the reef sources, in response to increasing re­ though very young individuals «10 never more than 6 m in depth. You ng ports of foreign fishing vessels in the cm length) develop weak byssal an­ H. hippopus attach by a weak, sparse Helen Reef area. They found that chors. appear to prefer byssus which disappears with age. tridacnid clam populations had appar­ the more shallow reef areas (approxi­ These clams have heavy triangular ently been reduced since 1972 (Bryan mate depth range, 4-10 m). shells and an undulating dorsal margin and McConnell, 1976). There are two slightly smaller tridac­ with very sharp, triangular, interdigitat­ In 1976 the Palau Marine Resources nids of commercial importance. mg processes. Office requested another resurvey in re­ sponse to continued reports of unau­ thorized fishing vessels in the Helen Reef area. There was concern that con­ Figure 2.-Palau District's southwest islands. tinued fishing in this area could lead to severe stock depletions, with small chances ofrecovery. This report gives a description of the 1976 resurvey effort, monitoring changes in the abundance of tridacnid clams on Helen Reef, and provides population densities of the area with comparison to earlier sur­ veys. Natural History Because of its isolation, Helen Reef Yap I.. at one time supported large tridacnid clam populations. In one instance,

Motoda (1938) reported observing ap­ J Palau Is. proximately 38-46 T. gigas within 100 2 CAROLINE ISLANDS m . Although populations have been greatly reduced over the years, all six . Sansaral I. species of the Tridacnidae family still Pula Anna I.. 'Merir I occur on Helen Reef: T. gigas (Linne), T. derasa (Roding), T. squamosa Tabi 1,'1 ... HELEN REEF I Lamarck, T. maxima (Roding), T. crocea Lamarck, and Hippopus hip­ popus (Linne). These beautiful and " sometimes colossal bivalves are found ~:"9> ~ .. 4.::""7'~'1_ only in the Indo-Pacific faunal region ... . (Fig. 3). ~""'~T The two largest Tridacna species are ,'. .... T. gigas (Fig. 4a) and T. derasa. C'll". ~..... Tridacna gigas is the species usually

February J980 9 60° 80° 100° 160° 180° 160° 140°

. _. .. A~~II __• H"'~A",,,,,,,,

______0_ '-_._ 4. ~_.

20°t----~I,_-+____::,.L____"t-h4_h_-+,;_____----'-+__'o__;;M=...... ".

''''-1".1'''' 0'i" '/."'" (~1-' 001-- ·--:;---t-----f--~t~~:_l~~.-f~~~~""''"":::_If__-~;_-'--+---,--'1o.::.;.~,,-,-.. f--I\--'~----~-=.~C:.:.'~''~-~I---+---I ',,: ~ ~«",: . . ,.\'lOINI" •• I("t"; '. ,-rOKII..a.U IS.

20° COOK ". Tridacna gigas

60° 80° 100° 120° 140° 160° 180° 160°

20° 1------.::~~--I---_,..L~~-_l--tL_f:.,;:.--7"=t:;==:..",_.;;:_--_+-....:,....,M=••;;:-:I.... H.. la...... ' .~:: ~.t,,'"'''''''11''1.1.. ~. '.-~

. ~OI.NUl .. ...TOKIU.U II.

200 .....::.-----j------t--·----:?¥- ..·~o~~... C.OOK ------,..

Tridacna derasa

20° 60° 80° 100° 160° 180° 160°

' .. ~A~";I__•

Ii... "':'.... 'A,,.. ------_..... -. 20°t---H~~ ---~'f"''<;:__----=.~_+---.-',_J

\. O°I----..-A~_l'ii_-----1

Tridacna sauamosa Figure 3.-Geographical ranges of tridacnid clams (from Rosewater, J965).

10 Marine Fisheries Review 80° 160° 180° 160° 140°

' .. ~A~"I...... __• 1 :"" """, ------_~_~_~/~ ------20° I---...::.J\--+--,,...L---'_.L-+----.;h~--_I_ ,-----'-I:--+.M;;....'.A..o.:-+---c;;;""--+-----~-~··;;;·~"'b_----+_-__I . .. s..,...... -,.. ."~ M

tI"''''' r ,.' ....'~ .~t4~ ',.'.:~ O·l..lqa.·

.._ .....1"...... '., "'T.~~.

200~,---+----+--_-=::;;;j;#~====~~

Hippopus hippopus

100° 160° 180° 160°

",

'T'

200t--/f-+Hf--±-=::::==f===:::::=-t---7-t:::::==::t:-~--:-_'_jr.::.'~·:.,'T:\·:::;:-~'-"':·:"'':.t_--:-;;,~,:;,'::-.-..;:,~=.tr·"-"-·...... _·_·-t-.-__-...-.-__-__-.-'__Ij.-.-..-__-__-.-..-.--1.. ------.;..._.~"-I.;;;.::;:.- -

Tridacna maxima

80° 100° 160° 180° 140°

~.': ,.",·,./1 _ "f"~""'I""", ------'.... -.-'''.------200r---'p-t~,...L-__'l--::t_-i7'lr-~;----'-t_~;;,;;-;:;;:-+---,=--f__-----;:.~·~::r_--_+____1

·",'I'".. ~.M.Alt'".I.LI" . '~I ...... ~ - ...... :.. ,.,../"- ~.'iltvt ,~f. '~J'- 0° l----~._-f_---_\~~~'=--e:~~f-,~~~-.",,""'- ~--:_:--'--\;:+---,----.:'!,"'.':c.--:c.-._+....:.._---'- ..,._i_..:~~,~••~..~-=-,--+--1 ~~':.'. " . ","OINI" " '...... :; f((·.•• '<"<"~ ','TOI(Il"Il.\lIS ..oc.o.· ...... "" .. '. "'Ton. .~~ , ..."l&.

2001-:. -----l------+------:;;..,(;-----t~~----'-t-~~---'w

Tridacna crocea Figure 3. - Continued.

February 1980 IJ Figure 4a.-Tridacna gigas, about 3 feet in length, on the Figure 4b.-Tridacna maxima partially embedded in coral. reef at low tide.

The larger Tridacna species are the reef flats and reef edge, remaining for counting two or three assigned clam clams which are most often utilized by firmly anchored by a byssus. species. Palauans and foreign fishermen. The Total population sizes (Nj) Survey Methods for Helen mantle and the adductor muscle tissues Reef atoll were estimated by the equa­ are highly favored food items, the Attempts were made during the May tion (Cochran, 1963): larger clams producing a higher yield 1976 survey to duplicate, as closely as per expended fishing effort. Except for possible, sampling locations and field i=1~ A a small trade as a tourist item, shells are methods used during the 1975 survey N= --YilJ - J [ n W discarded. (Bryan and McConnell, 1976). Tran­ ~ Ii The two smallest species, T. maxima sect lengths were determined from the i=1 (Fig. 4b) and T. crocea, are not usually U.S. Navy Hydrographic Office Chart where Yij is the number of individuals harvested by foreign fishermen and are of Helen Reef, No. 6072. The same of tridacnid clam species j observed minimally used for food by Palauans. transects as reported by Bryan and along transect i, Ii is the length of trans­ Tridacna maxima can grow to 35 cm McConnell (1976) were resurveyed. ect i, w is the transect path width ( = 2 (close to 14 inches). It can often be The bottom depths along transects var­ m), and A is the total reef area 7 2 confused with T. crocea, but T. max­ ied from approximately 6 m on the (= 5.34 X 10 m ; Hester and Jones, ima is distinguished by its more triangu­ ends, to a minimum of I m along the 1974), for n total transects. lar shape; its byssus keeps it firmly an­ midsection. Two observers were Survey Results and Discussion chored halfway embedded in coral and pulled slowly behind a boat while all coral heads. is the occurrences of live or dead tridacnid During the resurvey, we were able to smallest species, approximately 15 cm clams within a 2 m wide path below the repeat eight transects from the 1975 (6 inches maximum) and is found com­ boat were recorded on plastic writing survey (Fig. 5). Summary data col­ pletely embedded in coral heads on the tablets. Each biologist was responsible lected from the transects are presented

12 Marine Fisheries Review Table 1.-Summary of results from the 1976 survey of Halen Reef, Western Caroline Islands. Length Transects surveyed Location and of Area sur- No. of live Tran- direction transect veyed individuals 1976 Density of '1975 Density of -1975 and 1976 sect of transect (m) (m') counted clamslloo m' clamsll00 m' ---1975 only To lat. 2"54'42"N 970 1.940 3 Hippapus hippopus 0.20 0.10 long. 131'45'14"E 61 Tridacna maxima 3.10 3.10 2700 true - T. derasa 0.10

T. lat. 2'56'20"N '970 1.940 2 H. hippopus 0.10 long. 131'46'53"E 1 T. gigas 0.05 270' true 2 T. derasa 0.10 0.10 34 T. maxima 1.80 2.00 1 T. squamosa 0.10 0.10

T,o lal. 2'58'00"N 1.130 2.260 5 H. hippopus 0.20 long. 131'48'16"E 47 T. maxima 2.10 2.00 270' true

T" lat. 2"59'05"N 800 1,600 7 H. hippopus 0.40 0.20 long. 131'48'30"E 1 T. gigas 0.10 270' true 1 T. squamosa 0.10 11 T. maxima 0.70 2.7

T12 lat. 3'OO'30"N 645 1,290 26 H. hippopus 2.00 long. 131'48'55"E 4 T. maxima 0.30 0.20 0' true

T" lat.2"58'06"N 1,610 3,220 18 H. hippopus 060 0.30 long. 131'49'24"E 2 T. gigas 010 90' true 4 T. derasa 0.10 0.03 1 T. squamosa 0.Q3 106 T. maxima 3.30 4.30

Figure 5.-Locations of sampling T14 lat. 2'56'30"N 1.130 2.260 1 H. hippopus 0.04 0.30 transects made on Helen Reef atoll, long. 131'49'51"E 1 T. derasa 0.04 90° true 11 T. maxima 0.50 0.20 May 1976. T" lat. 2'58'00"N 480 960 1 H. hippopus 0.10 long. 131'49'00"E 38 T. maxime 4.00 5.70 1800 true ,Data from Bryan and McConnell (1976). in Table I; also included for compari­ son are the 1975 estimates for tridacnid population densities. In many in­ stances, the 1975 and 1976 es timates Table 2.-Stock estimates for trldacnlds on Helen Reef, Western CarOline Islands. were quite close. For each survey, the 1972 Est. of 1975 Est. of 1975 Est. of 1976 mean observed densities were ex­ standing standing slanding stock Est. of 3 panded to provide estimates of the total Species stock' stock' fremTa, .. ,T15 standi ng stock' Tridacna gigas 49.8xl()3 8.6xl()3 None observed 13.8xl()3 (4.7 x l()3- 22.9xl()3) populations on Helen Reef (Table 2). Tridacna derasa 32.8xl()3 12.9xl()3 14.1 xl0' 24.2xl()3 (8.1 xl()3- 40.3xl()3) Because of the reduced survey size in Tridacna squamosa 1.2xl()3 4.3xl0' 3.5xl0' 10.4x 1()3 (3.8xl()3- 17.0xl()3) Hippopus hippopus 44.6x 10' 47.4xl0' 70.5xl0' 217.5xl0' (89.1 xl()3-345.9xl()3) 1976, estimates of standing stocks for Tridacna maxima 1.7xl0' 1.4xlO' 1.3xlO' 1.1 xlO' (0.7xlO'- l.4xlO') 1975 were recalculated using only the Tridacna cracea 3.7xlO' Ubiquitous Ubiquitous Ubiquitous eight northern transects that were re­ 'Data from Hester and Jones (1974). 'Data from Bryan and McConnell (1976). computed from all transects. surveyed in 1976. This was done to 'Data from Bryan and McConnell (1976). computed from only the eight transects resurveyed in 1976. allow a more direct comparison be­ 480 percent confidence limits in parentheses. tween the two survey results. The 1972 estimates for the larger clams, T. gigas and T. derasa, were high compared with 1975 and 1976 es­ and the absence ofobservation in 1975. for T. derasa. All empty shells of T. timates. These species showed a large Recru itment can also be a consideration gigas and T. derasa observed were reduction in apparent population sizes when observing population fluctuations heavily encrusted with marine or­ between 1972 and 1975. The apparent but the extent that apparent increases ganisms, displaying little evidence of differences in the most recent stock es­ were caused by recruitment remains recent harvesting. timates for T. derasa were relatively unknown. Table 3 presents a compari­ Tridacna squamosa has been re­ small and should be attributed to survey son of the ratios of live and dead indi­ ported at low densities during all previ­ variation. Some transect variation is in­ viduals for the two most recent surveys. ous surveys. In 1972, Hester and Jones evitable despite attempts to duplicate In 1976, both species still showed large (1974) observed only one T. squamosa; 1975 methods, which may help explain numbers of dead individuals, although in 1975, Bryan and McConnell (1976) the observance ofT. gigas in the north­ the proportions of dead individuals counted only two individuals. These ern transects during the 1976 survey were slightly less than 1975, especially earlier investigators indicated that this

February 1980 13 tridacnid species is either very rare or Table 3.-Comparlson ofthe number of live and dead trldacnld clams observed during the 1975 and 1976 surveys, Helen Reef, Western Caroline Islands. not easily distinguished from T. max­ 1975' 1976 ima in this particular environment. Dur­ Ratio of live: Percent Ratio of live: Percent ing the 1976 survey, however, the two Species dead individuals dead dead individuals dead species were readily distinguished. Al­ Tridacna gigas 4:206 98.1 4:47 91.5 Tridacna derasa 6:168 96.6 7:24 70.8 though three individuals of T. Tridacna squamosa 2:1 33.3 3:0 0 squamosa were identified, their densi­ Hippopus hippopus 22:458 95:4 63:39 38.1 Tridacna maxima 629:23 35 312:5 1.6 ties were still low compared with all Tridacna cracea Ubiquitous Ubiquitous other species. Tridacna squamosa ap­ 'Data from Bryan and McConnell (1976). pears to prefer relatively protected areas, so wave exposure may be a limit­ ing factor for T. squamosa at Helen Reef. Although their relatively large size would make them desirable for Table 4.-Comparlson of the number of live trldacnlds counted In 1975 along transects In north (transects 8to 15) and south (transects harvesting, since only one dead indi­ 1 to 7) regions of Helan Reef, Western Caroline Islands (from Bryan vidual was observed in 1975, their and McConnell, 1976). population was apparently not strongly Live Live Percent individuals individuals located affected by this activity. counted in counted in in north reef south reef north reef The 1976 survey showed a substan­ Species transects transects transects tial increase in the apparent abundance Tridacna gigas 0 4 0 of H. hippopus, compared with the Tridacna derasa 4 2 66.7 Tridacna squamosa 1 1 50.0 1972 and 1975 surveys, but the ratio of Hippopus hippopus 20 2 90.9 dead individuals to empty shells was Tridacna maxima 367 262 58.3 Tridacna cracea UbiqUitous Ubiquitous still high. This species is characteristi­ cally very angular in appearance with dusky, sandy mantle tones. Their cryp­ tic coloration makes them one of the more difficul t tridacnids to count in the numbers was only noted during the survey. This apparent preference for field, perhaps contributing to an under­ 1975 and 1976 surveys. Lack of fishing the north reef region by H. hippopus estimation of abundance by earlier in­ pressure may allow these smallest would contribute to a population over­ vestigators. tridacnids to maintain a more constant estimation in the 1976 survey. Other The most curious condition observed population size. species appeared to be relatively evenly during the 1976 survey was the absence In Table 4 a comparison was made represented in both north and south re­ of large numbers of empty shells re­ between the numbers of live tridacnids gions. ported by Bryan and McConnell found in north and south areas of the Excluding T. maxima and T. crocea (1976). Foreign fishing boats had been reef in 1975 to examine the influences because of their lack of value to foreign observed in the area between the 1975 of large-scale distribution on popula­ fishermen, less than half (43.6 percent) and 1976 surveys, possibly taking (\cm estimates. The reefwas sampled by the tridacnid clams observed on Helen shells for the tourist market. Alterna­ transects 8-15 in the north and by tran­ Reef in 1975 were alive. In 1976, the tively, heavy seas and rough weather sects 1-7 in the south. Transects 8-15 in conditions were much the same with could have broken up old shells and the north correspond to the areas that only 40.1 percent of the clams surveyed swept them off the reefs. were resurveyed in 1976. The percen­ alive. In natural reef environments, it is The two smallest tridacnid species tage of clams located in the northern unusual to see so many empty shells have shown little variation in apparent eight transects in 1975 was calculated unless fishing pressures have been population densities over the years. The for each species. There appeared to be high. Large percentages ofempty shells small T. maxima have shown low var­ differences among some of the tridac­ have been observed following concen­ iance in stock size, with very little vari­ nids in location preference. In 1975, T. trated fishing activities (Motoda, 1938; ation in apparent densities from 1972 gigas was observed only in the south, Bryan and McConnell, 1976). Because through 1976. The ubiquitous T. while almost 91 percent ofH. hippopus of the lack of fishing pressure on the crocea was not surveyed in 1975 and were found in the north. Since T. gigas smaller species, perhaps T. maxima, 1976 because of the small size of indi­ is usually quite visible, the apparent and maybe T. crocea, indicate the nor­ viduals and high population densities. skewed distribution may only be a mal ratios of live to dead individuals Large coral heads can be embedded product of sampling variability since representative of tridacnid populations with many T. crocea, and it would take only small numbers of this species were subject only to natural sources of mor­ much time and a more detailed sampl­ observed. It is possible that a prefer­ tality. ing effort to survey this species reli­ ence for the south reef area may exist In May 1976, the effects of recent ably. Therefore, its presence in large but it is impossible to verify from this harvests were evident at Helen Reef

/4 Marine Fisheries Review with tridacnid populations reduced at self-fertilization. If this is the case, Office of Marine Resources, who from much higher levels of abundance then there may be some minimum funded the survey. (Motoda, 1938; Hester and Jones, threshold density from which reduced 1974). The conditions of the clam re­ populations would be unable to re­ Literature Cited sources observed in 1976 emphasized cover. Bryan, P. G., and D. B. McConnell. 1976. Status the need to prevent their further uncon­ Harvests of tridacnid clams from of giant clam stocks (Tridacnidae) on Helen trolled harvesting, especially since Helen Reef should continue to be re­ Reef, Palau, Western Caroline Islands, April most of the clams observed in 1976 1975. Mar. Fish. Rev. 38(4):15-18. stricted until increases in stock abun­ Cochran, W. G. 1963. Sampling techniques. 2d appeared smaller than the mean sizes dances have been observed and ed. J. Wiley & Sons, Inc., N.Y., 413 p. reported in 1972 (Hester and Jones, documented. Additional research is Hester, F. J., and E. C. Jones. 1974. A survey of giant clams, Tridacnidae, on Helen Reef, a 1974). Because the recruitment of needed to determine allowable clam western Pacific atoll. Mar. Fish. Rev. tridacnid clams appears to be slow and harvest levels, threshold densities, and 36(7): 17 -22. irregular, large population depletions stock recovery rates. Motoda, S. 1938. On the ecology, shell form, etc. of the Tridacnidae of the South Seas. J. could have serious long-term effects. Acknowledgments Sapporo Soc. Agric. Forest. 29: 375-40 I. Wada (1952, 1954) suggested that Rosewater, 1. 1965. The family Tridacnidae in because of the nature of spawning My thanks for survey assistance to the Indo-Pacific. Indo-Pac. 1:347­ 396. behavior in tridacnid clams, Daniel McConnell, Michael Wada, S. K. 1952. Protandric functional her­ self-fertilization rarely occurs. Other McGrenra, Steven Patris, and Jeffrey maphroditism in the tridacnid clams. Oceanogr. Mag. (Tokyo) 4:23-30. observations have also shown no nor­ June. My deep appreciation for support 1954. Spawning in the tridacnid mal larval development after attempts from Toshiro Paulis, Palau District clams. Jpn. J. Zool. II :273-285.

February 1980 /5