This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Larvae of Nearshore Fishes in Oceanic Waters of the Central EquatorialPacific1
THOMAS A. CLARKE2
ABSTRACT: Larvae of 72 taxa of nearshore fishes were identified from midwater trawl samples taken in oceanic waters between Hawai'i and Tahiti.Catches of nearshore fish larvae and number of taxa caught declined with distance from the closest island. Most of the taxa were taken only within 300 kmof the nearest island; only eight taxa were taken more than twice at greaterdistances. Highest catches were at stations close to major island groups, theHawaiian or Society islands. Among stations closest to small isolated islands,densities were higher relative to distance from shore within the North Equatorial Countercurrent and the Equatorial Undercurrent; these strong eastward-flowing currents routinely transport larvae > 1000 km from likely sourcesupstream. Even in the zones of higher abundance, densities of nearshore larvaewere much lower than in coastal waters, and adequate sampling in oceanicwaters requires larger, faster nets than those typically used for ichthyoplanktonstudies.
THE RELATIVE ABILITIES of different taxa ofnearshore fishes to disperse across oceanicbarriers are for the most part inferred fromknowledge of larval life span and adult distribution (e.g., Brothers and Thresher 1985,Victor 1986). There are few data on occurrence of larvae or juveniles of nearshore species at distances well removed from the nearest shallow-water sources. Richards's (1984)data indicated that some nearshore fish larvae occurred several hundred kilometersfrom land in the Caribbean, and Hare andCowen (1991) documented transport of labrid larvae over distances of 600-700 km alongthe coast of the eastern United States via theGulf Stream and warm-core rings. Data fromtruly oceanic areas, well removed from continents or large islands, are much fewer. Leis(1983, 1984) and Victor (1987) reported offshore captures of nearshore larvae (mostlylabrids) in the eastern Pacific and discussed
their occurrence relative to larval life spansand estimated velocities of prevailing currents.
In this paper I present data on nearshorefish larvae taken in a series of oceanic midwater trawl collections in the central equatorial Pacific between Hawai'i and Tahiti.Though the sampling program was originallydesigned to investigate distribution of smalloceanic fishes, the sample locations rangedbetween 24 and almost 1200 km from theclosest nearshore habitat. The samples weretaken with a large net, and the volume filtered at each station was considerably largerthan that of plankton tows taken by mostprevious studies. The primary purpose of thisreport is to document the types and frequency of nearshore larvae as a function ofdistance from shore and to consider the roleof prevailing currents and other physicalprocesses in larval dispersal.
Samples examined were taken on a seriesof cruises between Hawai'i and Tahiti during
134
1 Sample collection and rough sorting were supportedby NSF OCE77-09202 and the remainder of the analysesby the University of Hawai'i~awai'nnstituteOnVla= MATERI1\:J:;S-AND-ME-TIrine Biology. Manuscript accepted 15 May 1994.
2 Hawai'i Institute of Marine Biology, P.O. Box 1346,Kiine'ohe, Hawai'i 96744.
So . d200 I I I Claty I'slan S I1750~0-----'-----716:::-:0"",,,0-----'------::'1~50==0;----"""':"'-----:1~400W
FIGURE L Map of the central Pacific between Hawai'i and Tahiti showing locations of 71 nighttime midwatertrawl samples (triangles) and of relevant islands. Small islands are indicated by solid circles. For the Society andTuamotu islands, only the individual islands closest to the stations are shown. Because of their distance from thesamples, several islands in the southwest comer of the map are omitted. Broad.J>.had_ed_lines.show_the..appJ:Oximateboundaries between the North Equatorial Current (NEC), North Equatorial Countercurrent (ECC), and SouthEquatorial Current (SEC). Arrows indicate the direction of these currents and of the Equatorial Undercurrent (EUC).
136
1977-1980. The sampling program and procedures were given in Clarke (1987) and arebriefly summarized here. Oblique tows with a3-m Isaacs-Kidd midwater trawl lined withca. 6-mm (1j4-inch) mesh and with a I-mdiameter cod end of 0.333-mm mesh weretaken at ca. 10 intervals of latitude between20 0 Nand 40 S along 153 0 Wand 158 0 Wand between 200 Nand 18 0 S along 1500 W(Figure 1). The zone between 5 and 80 N,1500 W was sampled on five different cruises,and four other stations were sampled twiceduring the program. Seventy-one sampleswere taken at night between the surface and300-400 m; the volume sampled in the upper100 m, where most nearshore larvae probably occur (Leis 1991), was 20,000-40,000 m 3
per tow. Twenty-nine other samples, mostlytaken during the day, fished to greater depthsand were lowered and raised rapidly throughthe upper 300 m thus sampling smaller volumes in the upper 100 m « 10,000 m3 pertow). Unless explicitly noted, the results anddiscussion refer only to the 71 night tows.
Fish larvae and juveniles from the sampleswere sorted using a 3 x magnifier; previousexperience had shown that ca. 90% of thelarvae under 3-5 mm and all larger larvaewere recovered by this procedure. Larvae ofoceanic taxa were separated, and all larvae ofnearshore or potentially nearshore taxa wereidentified to at least family and to lowertaxonomic levels if possible. Notochord orstandard length was measured to the nearest0.1 mm. Because I usually was not certainhow many species were included in identifications to genus, I have dispensed with thecustomary "sp." or "spp." after genericnames. Distinct forms of certain families(mainly labrids) that could not be definitelyidentified to genus are identified as "LabridT," etc. These designations are identical withthose in Clarke (1991), where tentative generic identifications are given. The distancesalong a great circle arc between each sampleand the nearest island were calculated from
_the-latitudeLand_Jungitudes of the stationsand the islands. Locations and directions ofprevailing currents in the study area werebased on Wyrtki and Kilonsky (1992).
PACIFIC SCIENCE, Volume 49, April 1995
RESULTS AND DISCUSSION
A total of 392 larvae or small juveniles ofnearshore or potentially nearshore taxa wasfound in the samples (Table 1); all but 12were from night tows. Those of the ostraciidLactoria diaphana (Bloch & Schneider) (n = 5)were excluded from further analyses. All weretaken > 500 km from the nearest island, andL. diaphana is known to mature and spawnpelagically (Leis and Moyer 1985). Larvae ofcertain other taxa could also have originatedfrom pelagic spawning. The Echeneidae andsome of the Tetraodontiformes may haveincluded species that spawn pelagically. Zenion and the carangid genus Decapturus arecoastal pelagic as adults, but could occasionally spawn at considerable distances fromshore. The congrid leptocephali could include deep benthic species that originatedfrom seamounts or the ocean floor. None ofthese taxa were, however, taken frequently,and their inclusion had little effect on theoverall results. The frequently captured Ectreposebastes, which may also spawn offshore, is considered below.
There was a general decline in frequencyof capture of nearshore larvae, mean numbers per tow, and number of taxa with distance from the nearest island (Table 2). Thegreatest decline occurred between 200 and300 km distance; consequently, captureswithin 300 km were considered "inshore"and those at > 300 km, "offshore." Otherstudies have suggested that nearshore fishlarvae decline rapidly in abundance and diversity a few hundred kilometers from thenearest shore. Leis (1984) noted that nearshore fish larvae were very rarely taken > 500km offshore of the Central American coast,and Fedoryako (1989) reported that "coastal"larvae in the "upper epipelagial" declinedmost rapidly at distances of 100-300 kmfrom shore in the Pacific and 300-600 kmfrom shore in the Indian Ocean.
Catches were, however, obviously influenced by factors other than distance fromshore. The three highest catches were fromstations close to major groups of high islands. One (170 S, 1500 W; 34 larvae) was
tiiR 5' ·'b'Mllf4 P" "¥i'M' II iii !it .. ip"."1411.t1fWJ1i$Z!lIJb!EA£&iiiJ&:!&!M!
TABLE 1
LIST OF LARVAL FISH TAXA COLLEcrED, NUMBER OF SPECIMENS AND THEIR SIZE RANGE, AND RANGE OF DISTANCESBE1WEEN COLLECTION SITES AND NEAREST NEARSHORE HABITAT FOR 71 NIGHTTIME AND 29 OTHER TRAWL SAMPLES
BothidaeBothus 6 (13-34) 117-555Engyprosopon 16 (5-18) 140-272Pelecanichthys I (138) 522
Soleidae I (28) 140Cynoglossidae 1 (17) 44
only 44 km away from the small island of mesoscale eddies are frequently generatedTetiaroa and almost as close to the larger is- and eventually spun off downstream (patzertlands of Moorea and Tahiti in the Society 1969). Lobel and Robinson (1988) consideredIslands. Stations at 19.9° N (96 larvae) and the role of these eddies in entraining larvae17.8° N (25 larvae) along 158° W were 141 and returning them to nearshore habitats.and 272 km, respectively, away from the The high catches reported here could meannearest point in the Hawaiian Islands-and that considerable numbers of entrained near> 1200 km from Johnston Atoll, the next shore larvae are also carried well offshore bynearest island. These catches are comparable eddies that move downstream from the areawith those reported by ClarKe1f9-9T)forwliere tliey were generatea.similar samples taken much closer to shore in Inspection of the data from the remainingHawai'i. The two stations were almost di- 68 night tows, all of which were closest torectly west of the island of Hawai'i, where small isolated islands, indicated that stations
k liM! -@HAM fI§I§"YMW4MP·PEM
Larvae of Nearshore Fishes from Oceanic Waters-CLARKE
TABLE 2
139
SUMMARY OF CATCH DATA FOR NEARSHORE FISH LARVAE FROM 71 NIGHTI1ME MIDWATER TRAWL Tows IN THE
CENTRAL EQUATORIAL PACIFIC
DISTANCE NO. TOWS MEAN LARVAE PER TOW
(km) ( + TOWS) (RANGE, +TOWS) MEAN NO. TAXA PER TOW
NOTE: Given for each lOO-kIn increment of distance from nearest nearshore habitat are nwnber of tows, nwnher of positive tows(+tows), mean nwnber of larvae per tow, range for positive tows only, and mean nwnher oflarval taxa (see Table I) per tow.
at 5-8 0 Nand 20 S-2° N tended to havehigher catches, relative to distance from thenearest island, than those from other latitudes (Figure 2). This tendency was in partbecause 5-8 0 N, 1500 W was sampled repeatedly during the survey, and thus a disproportionate share of the stations was fromthat zone. Nevertheless, densities of larvaewere slightly higher in both the latitudinalzones mentioned above. Assuming that ca.30,000 m 3 was filtered in the upper 100 m pertow, densities for stations> 300 km from thenearest island were ca. 0.5/104 m3 for 5-8 0 Nand 0.7/104 m3 for 20 S-2° N as comparedwith ca. 0.3/104 m3 for offshore stations atother latitudes. The highest catch of the daytows, four larvae, was at 5.9 0 N, 1500 W, andthree of the four specimens of "Thalassoma"larvae (apparently the same as my Labrid T)reported by Leis (1983) were from 5-8 0 N,151 0 W.
Both of these zones of higher abundancelie within strong eastward-flowing currents(Figure I). The North Equatorial Countercurrent (ECC) is centered at ca. 6-7 0 N, andthe Equatorial Undercurrent (EUC) is centered on the equator. Though most of thestations 'at 5-8 0 N were closest to ChristmasIsland, the latter lies south of the ECC in thewestward-flowing South Equatorial Current(SEC) and is downstream of most of the sta-
tions. Transport to 5-80 N from Christmaswould require northward transport in an areawhere currents are predominately east-westand subsequent entrainment in the ECC. Themore likely sources for larvae at 5-8 0 N werePalmyra Island and Kingman Reef, whichwere farther away than Christmas but directly upstream in the ECC. Similarly, thestations at 20 S-2° N were closest to eitherChristmas or Malden islands. The former isat the northern edge of the EUC and the latter well south of it, and flow in the upper 100m is either mostly or exclusively westward inthe SEC. The more likely source for larvaenear the equator was probably Jarvis Island,which is located upstream nearly in the center of the EUC. Though the surface waters atJarvis also flow west in the SEC, the strongerEUC is centered at ca. 100 m and extendsvery close to the surface. Leis (1984) suggested that the EUC might be too cool forlarvae ofshallow-water fishes, but in fact temperatures in the upper 100 m of the EUC areonly 1-2°C lower than in the surroundingwaters of the SEC.
Most of the 72 taxa in Table I were onlytaken close to islands; 21 were taken only atone or two of the three stations close to largeisland groups: the two stations west ofHawai'i and the one near Tahiti. Seven othertaxa were taken only within 200 km of the
140
20
o
o
o
•
PACIFIC SCIENCE, Volume 49, April 1995
w~c:r:::sd 10Z
o
o
•
• •0
• •0 0 • •
0 • 0
CD 0 0 •• 0 • <DOO • • • •
• • o • 0 • 0 00 • • • •
500 1000DISTANCE (km)
FIGURE 2. Number of nearshore fish larvae captured versus distance from the nearest island for 68 nighttime Inidwater trawl samples taken in the central equatorial Pacific. Samples taken at latitudes of strong eastward currents (58° Nand 2° S-2° N) are indicated by solid circles, others by open circles. Data are from only those stations closest tosmall isolated islands; data from three other samples taken close to either the Hawaiian or Society islands have beenoInitted (see text).
nearest island, and 10 more only within 300Ian. Included in these were Parapercis andLabrid S, the two most abundant taxa inClarke's (1991) samples from ca. 13 Ian offHawai'i, and families such as Gobiidae andSynodontidae, which were important in thelatter samples. Larvae of Parapercis, Synodontidae, and Gobiidae were also found primarily very close to shore at Johnston Atollby Boehlert et al. (1992). These results implythat larvae of most of the 38 "inshore" taxararely occur at extreme distances from thenearest adult populations.
Twenty-five taxa were taken at only oneor two stations > 300 Ian offshore. Nine ofthese were also taken more than twice in-
shore: Congrinae, Scorpaeninae, Apogoninae, Carangidae, Pomacanthidae, Synagrops,Acanthurus, Ctenochaetus, and Bothus. Theoffshore captures of the scorpaenine, oneSynagrops, both Acanthurus, and both Ctenochaetus were downstream of Palmyra/Kingman (5-80 N, 1500W), and those of oneSynagrops and one Bothus were near theequator and downstream of Jarvis. The offshore captures of the congrine (Rhechias) andthe pomacanthid were near 150 N, 1580 W,conceivably within the range of transport byeddies from Hawai'i. Most of these taxa wereregularly taken close to shore by Clarke(1991). (Eel and flatfish larvae were not covered in the latter report, but my unpublished
1.....~.:&iJ£i • e ?E -H1t!li31tiWJ.iIli8dUHtS%-..'MMiWhWhi iii WM"?W
Larvae of Nearshore Fishes from Oceanic Waters-CLARKE 141
data indicate that larvae of congrines and downstream of Palmyra/Kingman. MurBothus were frequently present in the samples aenids were taken in about equal numbersconsidered.) inshore and offshore, and Luzonichthys,
Sixteen taxa taken only once or twice off- Anthiinae (most were probably Anthias),shore were not taken frequently inshore either. Labrid T, Chromis, and Naso more frequentlyNine (Muraenosocidae, Ophidiidae, Leth- inshore than offshore. For the last six taxa,rinidae, Epinephelinae, Echeneidae, Balisti- 22 of the 28 offshore captures were downdae, Plagiotremus, Diodontidae, and Peleca- stream of either Palmyra/Kingman or Jarvis.nichthys) were not taken near shore, and In none of these six taxa were there any obseven (Holocentrinae, Decapturus, Chaeto- vious differences in size composition betweendontidae, Cirrhitidae, Labrid N, Ammodyti- inshore and offshore catches.dae, and Callionymidae) were represented by Overall, the results indicate that larvae ofonly one or two inshore captures. Seven of most taxa of nearshore fishes usually occurthe 18 offshore larvae in these groups were within 300 km of the nearest source and thattaken at 5-80 N and four at 20 S-2° N. Ex- most, but not all, captures at > 300 km arecept for Decapturus, which is a coastal pe- associated with meso- or larger scale physicallagic species, and Callionymidae, larvae of transport mechanisms (i.e., strong currents ornone of these taxa appear to be particularly eddies). Though densities of nearshore fishcommon even in waters close to shore (e.g., larvae at distances > 300 km from shore areClarke 1991, BoeWert et al. 1992). ca. 100 times lower than densities of most
Larvae of the setarchine scorpaenid Ectre- taxa close to shore (Clarke 1991), the offshoreposebastes (presumably E. imus Garman) densities of nearshore larvae in the ECC andwere one of the most frequently taken taxa, EVC are still equivalent to ca. 5000 larvaeand most (13/20) were taken > 300 km off- per square kilometer in terms of surface area.shore. Though adults occur in fairly deep Given the horizontal extent, depth, and prewater, they appear to be island-associated vailing current velocities of the ECC and(Struhsaker 1973). Moser et al. (1977), how- EVC above 100 m depth (Wyrtki and Kilonever, reported larvae from 100 to 1300 W sky 1982), rougWy 105-106 nearshore laralong the equator-far from either South vae are transported eastward across 1500 WAmerica or the central Pacific islands. Only every day. That eastward "leakage" from theseven of the 13 individuals taken offshore in Line Islands postulated by Leis (1983) is thisthe study reported here were from either near high is not surprising because Leis also calthe equator or 5-80 N (i.e., directly down- culated that passive transport from the Linestream of likely nearshore sources), and one Islands to ca. 1500 W would take only ca. 20specimen, only 5.5 mm long, was from 18.40 days-less than the larval life span of theN, 1500 W- > 500 km east and upstream great majority of nearshore fishes.of the island of Hawai'i. Thus at least some One of the main reasons that so few nearof the offshore captures probably resulted shore larvae have been reported from offfrom offshore reproduction. Inclusion of shore areas is that most ichthyoplanktonEctreposebastes larvae, however, had little surveys have sampled only a few hundred toeffect on the overall trends considered above. a thousand cubic meters per tow with rela-
Eight other taxa were represented by more tively small, slow nets. I previously pointedthan two offshore captures. Of these, Mulli- out (Clarke 1991) that such tows are indae were taken only offshore. One was from sufficient for sampling many taxa even in7.90 N, 1500 Wand downstream of King- waters close to adult populations. Because ofman/Palmyra, and the others were from 10.3 the much lower densities estimated for offand lIS N-north of the ECC and too far shore areas, not only is the probability ofsoutl} to -invoKe transport oy eamesTrom -- encountenng even a smgle nearshore larvaHawai'i. Pseudogrammidae were taken only very low, but, as Leis (1984) pointed out,once inshore. The inshore capture and three most larvae dispersed this far would be largeof the offshore captures were from down- enough to avoid most plankton nets. Furtherstream of Jarvis; one offshore capture was sampling with large, fast nets is needed to
142
confirm trends in the data reported here andcould provide direct evidence of dispersal oflarvae to even greater distances than shownhere.
ACKNOWLEDGMENTS
I thank B. Mundy for assistance in identifying certain larvae.
LITERATURE CITED
BOEHLERT, G. W., W. WATSON, and L. C.SUN. 1992. Horizontal and vertical distributions of larval fishes around an isolated oceanic island in the tropical Pacific.Deep-Sea Res. 39: 439-466.
BROTHERS, E. B., and R. E. THRESHER. 1985.Pelagic duration, dispersal and the distribution of Indo-Pacific coral-reef fishes.Pages 53-69 in M. L. Reaka, ed. Theecology of coral reefs. NOAA SymposiumSeries Undersea Research, Vol. 3. NOAAUndersea Research Program, Rockville,Maryland.
CLARKE, T. A. 1987. The distribution ofvertically migrating fishes across the central equatorial Pacific. BioI. Oceanogr. 4:47-81.
---. 1991. Larvae of nearshore fishes inoceanic waters near Oahu, Hawaii. NOAA(Natl. Oceanic Atmos. Adm.) Tech. Rep.NMFS (Natl. Mar. Fish. Serv.) 101.
FEDORYAKO, B. I. 1989. The influence of thevicinity of islands and continental shelf onthe ichthyofauna of the upper epipelagialin the equatorial zones of the IndianOcean and the western Pacific. J. Ichthyol.29: 30-35.
HARE, J. A., and R. K. COWEN. 1991. Expatriation of Xyrichthys novacula (Pisces:Labridae) larvae: Evidence of rapid crossslope exchange. J. Mar. Res. 49:801-823.
LEIS, J. M. 1983. Coral reef fish larvae(Labridae) in the East Pacific Barrier.
--TopeiaT98J:-82o=828~-------------
---. 1984. Larval fish dispersal and theEast Pacific Barrier. Oceanogr. Trop. 19:181-192.
PACIFIC SCIENCE, Volume 49, April 1995
---. 1991. The pelagic stage of reef fishes:The larval biology of coral reef fishes.Pages 183-230 in P. F. Sale, ed. The ecology of fishes on coral reefs. AcademicPress, San Diego.
LEIS, J. M., and J. T. MOYER. 1985. Development of eggs, larvae, and pelagic juveniles of three Indo-Pacific ostraciid fishes(Tetraodontiformes): Ostracion meleagris,Lactoria jornasini, and L. diaphana. Jpn.J. Ichthyol. 32: 189-202.
LOBEL, P. S., and A. R. ROBINSON. 1988.Larval fishes and zooplankton in a cyclonic eddy in Hawaiian waters. J. Plankton Res. 10: 1209-1223.
MOSER, H. G., E. H. AHLSTROM, and E. M.SANDKNOP. 1977. Guide to the identification of scorpionfish larvae (Family Scorpaenidae) in the eastern Pacific with comparative notes on species of Sebastes andHelicolenus from other oceans. NOAA(Natl. Oceanic Atmos. Adm.) Tech. Rep.NMFS (Natl. Mar. Fish. Serv.) Circ. 402.
PATZERT, W. C. 1969. Eddies in Hawaiianwaters. Report 69-8, Hawai'i Institute ofGeophysics, University of Hawai'i, Honolulu.
RICHARDS, W. J. 1984. Kinds and abundances of fish larvae in the Caribbean Seaand adjacent areas. NOAA (Natl. OceanicAtmos. Adm.) Tech. Rep. NMFS (Natl.Mar. Fish. Serv.) SSRF (Spec. Sci. Rep.Fish.) 776.
STRUHSAKER, P. J. 1973. A contribution tothe systematics and ecology of Hawaiianbathyal fishes. Ph.D. dissertation, University of Hawai'i, Honolulu.
VICTOR, B. C. 1986. Duration of the planktonic larval stage of one hundred speciesof Pacific and Atlantic wrasses (familyLabridae). Mar. BioI. (Berl.) 90:317-326.
---. 1987. Growth, dispersal, and identification of planktonic labrid and pomacentrid reef-fish larvae in the eastern Pacific Ocean. Mar. BioI. (Berl.) 95: 145-152.
WYRTKI, K., and B. KILONSKY. 1982. Transequatorial water structure during the
-HawaiilOTaliifisnuttle experimenLRe-port HIG-82-5, Hawai'i Institute of Geophysics, University of Hawai'i, Honolulu.