Early life historiesof antarctic fishes

RICHARD L. RADTKE and ADOLF K. KELLERMANN

Division of Oceanic BiologyHawaii Institute of Geophysics

University of HawaiiHonolulu, Hawaii 96822

DAVID J. SHAFER

Department of ZoologyUniversity of Hawaii

Honolulu, Hawaii 96822

JAMES J . RUZICKA

Department of OceanographyUniversity of Hawaii

Honolulu, Hawaii 96822

Recruitment processes of marine fishes and the identificationof stages in the early life history (which are most vulnerableto mortality) are important topics of study in fish biology.Despite the ecological and commercial importance of antarcticfishes, little is known about their larval and juvenile devel-opmental stages. Studies have focused on the spatial and tem-poral occurrence of larvae, as related to hydrographical regimes,and on their feeding ecology and the interannual variability ofabundance levels (e.g., Kellermann 1986; 1987; Kellermannand Kock 1988). To understand relationships between earlylife history processes and biotic and physical environment, itis necessary to determine the time scale of major events indevelopment and the growth rates of individual life stages.Unfortunately, critical assessment of these parameters in thefield requires a logistically difficult sampling program. There-fore, we are obtaining such information from the otoliths oflarval antarctic fishes.

Fish otoliths are calcified structural components of the innerear's equilibrium and auditory sensory system and may con-tain an historical record of biological and ecological informationencountered during a fish's lifespan. Otoliths consist of a pro-tein matrix into which calcium zones are deposited with adaily, anticyclic periodicity (Mugiya 1987) moderated by dielmetabolic cycles as synchronized to ecological and environ-mental parameters. The daily nature of microincrement for-mation has been demonstrated for the antarctic fishesNototheniops nudifrons and Trematomus newnesi (Hourigan andRadtke 1989; Radtke et al. 1989). During periods of slow growth,more calcium and less protein is deposited. Although it isknown that increment formation is initiated the day followinghatch in a variety of species, such investigations are lackingfor the notothenioid antarctic fishes. Critical events in earlylife history—including larval hatch, first feeding and yolk ab-sorption—may be documented in otolith microstructure bydistinct discontinuities, providing information on the relativetime scales of larval ontogeny.

To validate the periodicity of otolith deposition zones, larvaeof the notothenioid fish Notothenia neglecta were reared fromfield-caught eggs (figure 1) on board the icebreaking research

Figure 1. Field-caught egg with embryo of the antarctic fish No-tothenia neglecta, 4.42 millimeters in diameter. Note the advancedstage of the embryo and the yolk-sac.

vessel RIV Polarstern. Eggs were collected near Elephant Islandand in the northern Weddell Sea in October 1988. They con-tained embryos at an advanced stage of development, showingspontaneous heartbeats and body movements. Eggs were rearedin 60-liter inert plastic tanks at densities of 2-4 eggs per tank.After hatching, larvae were removed and placed in separatetanks for continued rearing under controlled temperature andlight regimes. Water temperature varied between –1.0° and1.8°C, and the light cycle was standardized to 16 hours light!8 hours dark. During the week following hatch, larvae weremarked with tetracycline, which deposits a fluorescent markon the increment at the day of formation (Hourigan and Radtke1989). Larvae were either starved or fed on a diet of wildzooplankton. Following completion of the cruise leg, eggs andlarvae were flown to the Alfred-Wegener-Institut for Polar andMarine Research, Bremerhaven, where rearing experimentswere continued. Survival of larvae ranged from 15 to 68 days.These experiments will document the initiation of otolith mi-croincrement formation and its periodicity in larval antarcticfish. Investigation of microincrement symmetry will provideinformation on the effect of starvation on growth-incrementformation.

To investigate hatching periodicity and early life history dy-namics from field-caught larvae of antarctic fishes, otolith mi-crostructure was analyzed in larvae of six species of icefishes,Channichthyidae, caught during the expeditions of RIV Polar-stern in Bransfield Strait and adjacent waters in spring 1987and 1988. Large sample sizes were available for Chionodracorastrospinos us, Chaenodraco wilson i, and Pagetopsis rnacropterus.Larvae of C. rastrospinosus hatch in early spring, and yolk-saclarvae initiate feeding on overwintered late furcilia larvae ofthe antarctic krill Euphausia superba and on larval nototheniidfishes (Kellermann 1986). Hatching in C. wilsoni and P. ma-cropterus is likely to occur during winter (Kellermann 1989),when the area is poorly accessible by research vessels. In these

194 ANTARCTIC JOURNAL

/j ;,,j i-,

4A

%

species, investigations on otolith microstructure are the mostreliable means to assess early life history events. Otoliths froma 34.3 millimeters (total length) C. wilsoni larva displayed adistinct core and a second concentric disruption which arelikely to represent hatching and yolk absorption marks (figure2). Sixty-two microincrements were counted in this specimenbetween the two events suggesting a considerably long yolk-absorption period. Such long yolk-absorption periods havealso been observed in other antarctic fishes through rearingexperiments of larval Harpagifer antarcticus (Daniels 1978) andNototheniops nudifrons (Hourigan and Radtke 1989).

Additional larval samples of N. gibberifrons, C. rastrospinosus,N. larseni, and K. anderssoni were obtained in limited quantityfor study during a January 1989 cruise aboard the NationalOceanic and Atmospheric Administration ship Surveyor. Lar-vae were collected from Bransfield and Gerlache straits andSouth Georgia in deep (60-90 meters and 140-180 meters) obliquebongo tows, shallow "yo-yo" 30-meter bongo tows, and IssacsKidd Midwater Trawls. Smaller collections of C. antarcticus, A.skottsbergi, G. opisthopterus, N. nudifrons, and C. aceratus werealso obtained.

Because spawning in most notothenioid fishes appears tobe fixed and concluded within a few weeks, hatching occursindependent of the pack-ice retreat and subsequent onset ofthe production cycle in spring. The large larval yolk reservesseem to be sufficient to compensate for periods of food short-age, which may occur if pack-ice retreat is delayed in the springand fish larvae hatch before the onset of egg production ofcopepods following the spring bloom (Kellermann 1989). Oto-lith microstructure may reveal in situ information on the tem-

poral and spatial variability of yolk absorption and growthrates. If it is possible to relate such data to the biotic andphysical environment of yolk-sac larvae, the factors contrib-uting to the survival and performance of larvae may be iden-tified and used to construct models of spawning and life historytactics.

Thanks are due to the crews of the WV Polarstern and Na-tional Oceanic and Atmospheric Administration ship Surveyorfor help in collecting and rearing. Hordes of laboratory workerstoo numerous to mention helped with sample preparation anddata analyses. This research was supported by National Sci-ence Foundation grants DPP 85-21017 and DPP 88-16521.

References

Daniels, R.A. 1978. Nesting behavior of Harpagifer bispinis in ArthurHarbor, Antarctic Peninsula. Journal of Fish Biology, 12, 465-474.

Hourigan, T.F., and R.L. Radtke. 1989. Reproduction of the Antarcticfish Nototheniops nudifrons. Marine Biology, 100, 277-283.

Kellermann, A. 1986. On the biology of early life stages of notothenioidfishes (Pisces) off the Antarctic Peninsula. Bcrichte zur Polarforschung,31, 149.

Kellermann, A. 1987. Food and feeding ecology of postlarval and ju-venile Pleuragramnia antarcticum (Pisces; Notothenioidei) in the sea-sonal pack ice zone off the Antarctic Peninsula. Polar Biology, 7, 307-315.

Kellermann, A. 1989. The larval fish community in the zone of seasonalice cover and its seasonal and interannual variability. Archiv fürFischereiwissenschaft, 39 (Beih. 1) 89-109.

Kellermann, A., and K.-H. Kock. 1988. Patterns of spatial and temporaldistribution and their variation in early life stages of Antarctic fish

Figure 2. Otolith section of a 34.3-millimeter length Chaenodraco wilsoni larva showing microincrements and the marks produced at hatchand yolk absorption, respectively.

1989 REVIEW 195

in the Antarctic Peninsula region. In D. Sehrhage (Ed.), AntarcticOcean Resources Variability. Berlin, Heidelberg, New York: Springer.

Mugiya, Y. 1987. Phase difference between calcification and organicmatrix formation in the diurnal growth of otoliths in the Rainbow

trout, Salino gairdneri. Fisheries Bulletin of the U.S., 85(3), 395-401.Radtke, R.L., T.E. Targett, A. Kellermann, J.L. Bell, and K. T. Hill.

1989. Antarctic fish growth: Profile of Trenatomus Ilewnesi. MarineEcology Progress Series, 57, 103-117.

Treadmilling of antarcticfish microtubules

at low temperatures

H. WILLIAM DETIIcH, III

Department of BiologyNortheastern University

Boston, Massachusetts 02115

RIcHAIW H. HIMEs

Department of BiochemistryUniversity of Kansas

Lawrence, Kanses 66045

The cytoplasmic microtubules of eukaryotic cells performessential functions in many cellular processes, including mi-tosis, organelle transport, and nerve growth and regeneration.The polymerization of microtubules from their subunit pro-teins, tubulin alpha-beta dimers and microtubule-associatedproteins (MAPs), is an entropically driven reaction mediatedlargely by the release of structured water at regions of inter-subunit contact (Correia and Williams 1983). Because subunitassociation is driven by an increase in entropy, microtubuleformation is favored by high temperatures. The microtubulesof warm-blooded animals, for example, assemble from theirsubunits at physiological body temperatures (30-37°C) and de-polymerize at lower temperatures (0-4°C). By contrast, thecytoplasmic microtubules of the antarctic fishes, a group ofpoikilotherms adapted to temperatures in the range -1.8 to+ 2°C, must assemble in an unfavorable thermal environment.The long-term goal of our project is to determine the biochem-ical adaptations that enable the microtubules of antarctic fishesto assemble and function efficiently at low temperatures. Aspart of this effort, we have initiated studies of the dynamicsof antarctic fish microtubules polymerized to a steady state invitro.

Cytoplasmic microtubules in vitro are dynamic polymers thatadd and lose tubulin dimers by several end-dependent mech-anisms. "Tread milling" involves the net and balanced additionand loss of tubulin dimers at opposite microtubule ends (Mar-golis and Wilson 1978) at polymer steady state. Consequently,subunits that enter a microtubule at one end eventually dis-sociate from the other. For MAP-rich and MAP-free mam-malian microtubules at 30-37°C, the rates of this subunit "flux"are approximately 1 and 52 micrometers per hour, respectively(Margolis and Wilson 1978; Farrell etal. 1987; Hotani and Horio1988). If treadmilling is important physiologically, then themicrotubules of the cold-adapted antarctic fishes should ex-hibit similar behavior at low temperatures.

During the past year, we completed studies of the tread-milling of the microtubules of antarctic fishes at near-physio-logical and supraphysiologica! temperatures (Himes and Detnch1989). Samples of pure, MAP-free brain tubulin from Not othieniagibberifrons, purified as described previously (Detrich andOverton 1986), were polymerized at 5, 10, or 20°C for intervalssufficient to achieve steady states in polymer mass and stablelength distributions. Tubulin incorporation into and loss fromthe steady-state microtubules were measured by a modificationof the radio-labeled guanosine 5'-triphosphate (GTP) markerprocedure (Margolis and Wilson 1978; Himes and Detrich 1989).Microtubule length distributions were determined by negative-stain electron microscopy. Figure 1 shows the initial rates of

1.0

0.8

C

a0.6-o-1-i0Q)

0.4Z-o

CD0.2

MCI0 10 20 30 40 50 60 70 80

Minutes

Figure 1. Temporal-dependence of the incorporation of [3H]GTP-labeled tubulin into microtubules at three temperatures. Samplesof a preparation of N. gibberifrons tubulin were polymerized attemperatures of 5, 10, or 20°C for intervals (300 minutes at 5°C and10°C, 220 minutes at 20°C) sufficient to attain a steady state inmicrotubule mass and to achieve stable length distributions. Themicrotubules were then exposed to pulses of [3H]GTP for the timesindicated, collected by centrifugation, and analyzed for radiolabelincorporation (Himes and Detrich 1989). The uptake of radiolabelednucleotide by the microtubules (Mts) is plotted as a function ofpulse duration at 5°C (open circles), at 10°C (closed circles), andat 20°C (open triangles). Tubulin concentrations: 5°C, 1.1 milligramsper milliliter (mg/ml); 10°C, 0.86 milligrams per milliliter, and 20°C,0.59 milligrams per milliliter. Microtubule number concentrations:5°C, 1.5 x 10- 10 molar (M); 10°C, 3.4 x 10 10 molar; 20°C, 3.0 x10 10 molar. The microtubule length distributions correspondingto these experiments are shown in figure 2. Reprinted from Himesand Detrich (1989) with permission. (Copyright 1989 AmericanChemical Society.) (jiM denotes micromolar.)

196 AN [ARCTIC JOURNAL