strongly not to be abundant where adult E. superba were abun-dant, and likewise with Thysanoessa. The larval vs. adult separa-tions occurred in both horizontal and vertical directions. All thismakes good ecological sense, but we do not know how suchdistributions are generated or maintained.

As yet, our analyses are incomplete, and the samples weobtained are simply inadequate to answer most of the intriguingquestions: (1) How is such a huge swarm generated and main-tained? (2) Where does the food come from to support such amass of euphausiids? (Or, indeed, are individuals in the swarmactively feeding? Euphausiids can live for long periods withouteating.) (3) What is/are the function(s) of such a swarm? (Proba-bly not breeding, for our samples contained not one egg-carry-ing female euphausiid nor one loose euphausiid egg.) (4) Whatpercentage of the total antarctic krill population is representedin these huge aggregations? (5) How many such swarms arethere scattered throughout the Antarctic? And finally, given allthis uncertainty about superswarms, (6) precisely what will bethe consequence to krill (and to the whales, sea birds, squid,penguins, and seals which depend upon them) of intensivefishing pressures that seem inevitable given man's need forprotein and his past failure to exploit intelligently marine livingresources?

Many of these questions will be investigated by us during ournext cruise (funded by the National Science Foundation forFebruary-March 1984 on the R/v Melville), which will be dedi-cated to an intensive investigation of the Elephant Islandsuperswarm.

Cruise Vulcan was funded by National Science Foundationgrant DPP 79-21295 to Scripps Institution of Oceanography.Analyses of krill samples were funded by National ScienceFoundation grant DPP 81-20492 to the San Diego Natural HistoryMuseum. I thank M. Macaulay for permission to use some of hisacoustic information, particularly figures 2 and 3.

References

Macaulay, M. 1983. Antarctic krill (Euphausia ;uperha) swarms fromElephant Island. Antarctic Journal of the U.S., 18(5).

Shulenberger, E., J. H. Wormuth, and V. J . Loeb. In press. Super-swarms of Euphausia superba Dana. I: Overview of structure andcomposition. Journal of Crustacean Biology.

Wormuth, J. H. 1983. Zooplankton associated with superswarms ofantarctic krill. Antarctic Journal of the U.S., 18(5).

Antarctic krill (Euphausia superba)swarms from Elephant Island

MICHAEL MACAULAY

Northwest and Alaska Fisheries CenterSeattle, Washington 98115

The National Oceanic and Atmospheric Administrationfunded a project to participate with a collaborative project sup-ported by the National Science Foundation. The purpose of theexperiment was to investigate open-sea patches of krill, as wellas to gather other biological and physical data. The observationspresented in this paper were taken on the iIv Melville during thesecond survey of the Scotia Sea (cruise Vulcan, leg 7) 26 Febru-ary to 27 March 1981.

Primary acoustic observations were made by simultaneoussoundings with 50- and 120-kilohertz sounder, systems. A side-looking 105-kilohertz system was used to detect the presence ofkrill in the upper 5 to 10 meters which would not be detected bythe downward-directed systems. Frequency modulated analogrecordings of the envelope-detected signal from each of thethree frequencies were made in the field.

Analyses were done using the NWAFC Acoustic Research Con-tainer computer system by echo integration (SCAR 1981). Thismethod estimates the abundance (weight per unit volume) oftargets present in the water column for any depth intervalwhere target strength is known or can be estimated by length-frequency or length-weight relationships. The estimates of

average abundance over the total water column selected wascalculated to 250 meters or to bottom, whichever occurred first.

Transects were subdivided to give a similar number of obser-vations in each group and combined into blocks of area forcalculation of biomass. These blocks were selected to give arange of sizes for statistical comparison. The variance associatedwith these biomass estimates was computed by two indepen-dent methods. The first used serial correlation; the second usedthe cluster sampling estimator method of Williamson (1982) asan independent method for comparison.

During the survey, numerous patches and large con-centrations of krill were found. The average size patch was 100to 500 meters in horizontal extent by 40 to 60 meters in verticalextent with mean abundance of 50 to 200 grams per cubic meter(approximately 75 to 300 individuals per cubic meter). The verylarge swarm found northwest of Elephant Island containedabundances up to 600 to 800 grams per cubic meter, but typ-ically abundance was 300 to 500 grams per cubic meter. Thelarge biomass was due to distributions that extended as deep as250 meters and the large area covered (150 square kilometers).In the shallow area northwest from Elephant Island, con-centrations of krill were found in layers associated with the 100-to 150-meter depth contour.

Two large swarms were observed in the vicinity of ElephantIsland. The first occurred from 1 to 3 March and the secondappeared on 7 March and was being fished by more than 40Soviet trawlers. The Soviet fleet had also been fishing in thevicinity of the earlier swarm. After a survey period in BransfieldStrait, the Melville returned to Elephant Island. This secondoccupation was characterized by a lower incidence of patcheswith some associated with shallow regions near the island. TheSoviet fleet had also dispersed and only a few vessels remainedin the area.

1983 REVIEW 197

60 52 Cotoi,, I',tprvals1) 520

- - - --- --- 10 0 kq inC 200

54-00

56 o

58 Aoo

61 ' 0

N0©©Oi

2 I 0 a

\\

/6 &... 1115535 550 5430

Figure 1. Contour plot of data from 5-8 March. (Blocks A, B, and C.)

Biomass as well as 95 percent confidence limits around thebiomass estimates were computed using the two variance mod-els (table). The variance calculations by area indicate that thesimple variance underestimates the variance associated withsampling serial data of this type. The intent of the comparisonof variance calculation methods was to demonstrate the need for

Biomass by block and variance estimators (95 percent confidencelimits) for estimated biomasse

SimpleSerialClusterBlockBiomassvariancecorrelationanalysis

(N) (MMT)(MMT)(MMT)(MMT)

A (396)0.12210.00290.09090.0620B (316)0.20670.01190.19600.2164C (976)2.12800.02592.08901.22900 (348)0.09880.00170.06300.0343E (1288)0.20590.00200.24530.0861F (40)0.13170.01210.08270.1191

a "MMT" denotes million metric tons; N" denotes number of observationsused in calculations. 95 percent confidence interval is 1.96 \/'i?

some type of estimator which includes the effect of serialcorrelation.

The potential impact of fishing on such a population (if thereare aspects of reproductive behavior associated with its forma-tion) make better understanding of the causes and frequency ofoccurrence of this phenomenon imperative. This large con-centration of biomass also demands special treatment duringacoustic surveys. Because of the relatively rare occurrence andassociation with adjacent geographic features, such as islandsand ice edges, large swarms must be treated separately from theusual transect data.

The work was supported by special allocation of funds fromthe National Oceanic and Atmospheric Administration.

References

Scientific Committee on Antarctic Research (SCAR) 1981. Calculation ofthe constants needed to scale the output of an echo integrator. BIO-MASS handbook (Series 7)

Williamson, N. J. 1982. Cluster sampling estimation of the variance ofabundance estimates derived from quantitative echo sounder sur-veys. Canadian Journal of Fisheries and Aquatic Sciences, 39(1), 229-231.

198 ANTARCTIC JOURNAL

1 E

60 52 Co tit our ê tit ervn1c--

(15

?fl\ -50

am

58

61 0

--

M. / l-

-----

6 L5535 5 5 0

54 30

Figure 2. Contour plot of data from 20-21 March. (Block E.)

Zooplankton associated withsuperswarms of antarctic krill

JOHN H. WORMUTH

Department of OceanographyTexas A&M University

College Station, Texas 77843

A field program involving simultaneous acoustic and netsampling was undertaken in February-March 1981 (see Mac-aulay, Antarctic Journal, this issue, for acoustics and Shulen-berger, this issue, for net sampling details). Both horizontal andstratified oblique tows were taken in and around a large krillswarm in the vicinity of Elephant Island. Most tows were withinthe upper 80 meters because a large portion of the krill swarmwas in this portion of the water column.

We used a multiple opening-closing net and environmentalsensing system (MocNEss) which collected eight discrete sam-

ples per tow. We positioned the net according to the depth of thekrill "seen" by the acoustics. We took 25 tows and have com-pleted the taxonomic analysis of all 200 samples (see Shulen-berger, Wormuth, and Loeb in press, for additional details). Thesamples were sorted for krill (Shulenberger, Antarctic Journal,this issue), and all other specimens were identified. With fewexceptions, copepods dominated the samples numerically. Thecopepods consisted mainly of five species (Calanus propinquus,Calanoides acutus, Metridia gerlachei, Rhincalanus gigas, and Eu-

chaeta antarctica). The rank order of numerical abundance ofthese five species was very consistent regardless of depth, timeof day, or horizontal distance. These five species also show asignificant tendency to covary in numerical abundance in thesame tows. This suggests that these copepod species are re-sponding to the same environmental stimuli.

Within horizontal tows our spatial scale sampled per net waskept as constant as possible within each tow. This resolutionranged from about 90 meters to 300 meters. It was not unusualto see 1-1.5 orders of magnitude variation in the numericalabundance of copepod species between successive nets (fig-ure). This variation does not appear to be related to variations inthe numerical abundances of the euphausiid species, except in afew extreme cases.

1983 REVIEW 199