Kottmeier, S.T., and C.W. Sullivan. 1988. Sea ice microbial commu-nities (SIMCO). 9. Effects of temperature and salinity on rates ofmetabolism and growth of autotrophs and heterotrophs. Polar Bi-ology, 8, 293-304.

Palmisano, A.C., and C.W. Sullivan. 1983. Sea ice microbial com-munities (SIMCO). I. Distribution, abundance, and primary pro-duction of ice microalgae in McMurdo Sound, Antarctica in 1980.Polar Biology, 2, 171-177.

V'

Stoecker, D.K., A. E. Michaels and L.H. Davis. 1987. Large proportionof marine planktonic ciliates found to contain functional chioro-plasts. Nature, 326 (6,115), 790-792.

Thomsen, HA., K.R. Buck, P.A. Bolt, and D.L. Garrison. In prepa-ration. Fine structure and biology of Cryothecomonas gen. nov. (Pro-tista incerta sedis) from the ice biota. Canadian Journal of Zoology.

Seasonal changesin cell size and abundance

of bacterioplanktonduring the Phaeocystis sp. bloom

in McMurdo Sound

MARY Purr

Oceanography DepartmentOld Dominion UniversityNorfolk, Virginia 23529

K. YNGVE BØRSHEIM

University of TrondheimInstitute of Technology

Laboratory of BiotechnologyTrondheim, Norway

GERALDINE MIcELI and DIANE K. STOECKER

Biology DepartmentWoods Hole Oceanographic Institution

Woods Hole, Massachusetts 02543

Cell size must be known to determine the biomass of naturalbacterioplankton (Psenner 1990). In addition, cell size containsinformation about both bacterial growth rates and protozoangrazing pressure (Andersson, Larsson, and Hagstrom 1986).Bacterial cell size and growth rate are positively correlated(Ammerman et al. 1984). The average cell size of natural bac-terioplankton, however, is generally smaller than in experi-mental manipulations where grazing pressure is reduced. Anumber of protozoans selectively graze large bacteria (Gon-zalez, Sherr, and Sherr 1990). Thus, in temperate waters, pro-tozoan grazing apparently maintains even rapidly growingbacterioplankton below their maximum obtainable size (Am-merman et al. 1984). Here, we examine season changes in cellsize to obtain qualitative information about how grazing influ-ences the bacterioplankton community in McMurdo Sound.

Seawater samples were collected at six depths between 0and 100 meters from station IE at the edge of the sea ice ofeastern McMurdo Sound or through an established hole in theannual sea-ice at station H about 30-60 kilometers south ofstation IE. Direct counts of bacterioplankton abundance were

made for subsamples filtered directly onto 0.2-micrometer fil-ters (TBAC) and in subsamples passed through a 2-micrometerfilter prior to filtration onto the 0.2-micrometer filter (BAC).The difference between TBAC and BAC was used to estimatethe number of bacteria in the >2-micrometer fraction (>2BAC).The >2BAC fraction includes the largest free-living bacterio-plankton as well as bacteria attached to particles. Bacterio-plankton size was estimated from linear dimensions of cells inphotomicrographs of TBAC samples pooled from 15 and 50meters.

Areal concentrations of chlorophyll integrated over the up-per 100 meters remained low (<10 milligrams per square me-ter) until late December when a bloom of large (>200micrometers) Phaeocystis sp. colonies caused a 10-30-fold in-crease in chlorophyll (figure, block A).

Bacterioplankton abundance and cell volume (determinedfrom photomicrographs) both increased by factors of about 2-3 between late November and late December (figure, blocks Band C). Transmission electron microscopy confirmed the in-crease in cell volume in bacterioplankton at or above 50 metersat station IE (table 1). The increase in average cell size in lateDecember was due largely to an increase in the number ofvery large rods and cocci.

Maximum abundance of >2BAC coincided with the onsetof the Phaeocystis sp. bloom and the period of maximum cellsize (figure, block D). In separate experiments in early January,about 33 percent of the bacterioplankton did not pass a 202-micrometer mesh (data not shown). Thus, increased abun-dance of >2BAC was due, at least in part, to changes in thenumber of bacteria associated with large particles, probablyPhaeocystis sp. colonies rather than increased bacterioplanktoncell size. Microscopy confirmed that Phaeocystis sp. colonieswere often colonized by dense accumulations of bacteria.

By mid-January, the Phaeocystis sp. bloom was in decline andchlorophyll concentrations had decreased relative to the earlyJanuary maximum. Bacterioplankton abundance remainedconstant but cell size decreased to pre-bloom levels at bothstations.

Seasonal changes in bacterioplankton size in McMurdo Soundgreatly affect the estimation of bacterioplankton standing stock.For example, at the ice edge at 15 and 50 meters, bacterio-plankton biovolume (abundance X size), and presumably bio-mass, increased by a factor of 7 between 27 November and 27December whereas the change in bacterioplankton abundancealone was only about 2.

During the Phaeocystis sp. bloom, bacterioplankton growthrates reportedly increase by over an order of magnitude relativeto pre-bloom rates, presumably because of increased substrateavailability (Kottmeier et al. 1987; Guillard and Hellebust 1971).Abundances and estimated clearance rates of heterotrophicflagellates, the major grazers of bacterioplankton in temperate

1990 REVIEW 199

0

a)o0C

><C

in

0C\J

-0 co 0

CU

C

0

a)C)

L)ecemberJanuary

DecemberJanuary

120

100

80

60

40

20

0

40

30

20

10

0

Depth

015255075

100

015255075

100

Date

11 Nov 1989

Mean27 Dec 1989

Mean

a in cubic micrometers.

200

Cell sizea

0.046±004 (104)0.046±010 (21)0.044±013 (17)0.064±.012 (24)0.038±008 (19)0.083±018 (45)0.053:t .0060.18 ±041 (45)0.12 ±006(101)

ND0.10 ±.006 (81)

ND0.028 ±.003 (60)0.107±.031

0.12

0.1

0.08(I)

a)0 0.06

0.04

0.02

ft.

T7\AJanuary

C)

S.

I:\

DecemberJanuary

400

A)

C'1 300

enE I'5. 200-c0.00-co 100

ii

December0 14

McMurdo Sound, 1989-1990 field season. Seasonal patterns at the edge of the annual sea-ice (station IE, triangles) and beneath the sea-ice (station H, squares) of: A. Areal chlorophyll. B. Areal bacterioplankton abundance (TBAC). C. Bacterioplankton cell size measured fromepifluorescence photomicrographs of pooled TBAC samples from 15 and 50 meters. D. Bacterioplankton not passing a 2-micrometer filter(>213AC). Arrow denotes data when colonial forms of Phaeocystis sp. first observed. Note different scale in blocks B and D. All values(except block C) represent values integrated over the upper 100 meters of the water column. (tm 3 denotes cubic micrometers. mg . rn2denotes milligrams per square meter.)

Table 1. McMurdo Sound, 1989-1990 field season.Bacterloplankton cell size in samples collected at station IE

estimated from transmission electron microscopy (Borsheim,Bratbak, and Heldal 1990). Means ±SEM (standard error of the

mean). Number of cells sized in parentheses. ND Is no data.

systems, were very low prior to the bloom in late November(table 2). The increase in bacterioplankton abundance and cellsize in December likely reflects, at least in part, a temporal lagbetween increased growth rates of bacterioplankton during thePhaeocystis sp. bloom and development of the heterotrophicflagellate populations.

In addition, Phaeocystis sp. colonies are apparently poorlygrazed relative to other algae (Verity and Smayda 1989). Bac-terioplankton attached to the colonial matrix of Phaeocystis sp.may grow rapidly in the presence of enhanced substrate andmaintain larger average sizes than their free-living counter-parts because of reduced grazing pressure. Phaeocystis sp. col-onies may provide a refuge from grazing for a portion of thebacterioplankton, hence increasing the average size of the totalbacterioplankton population.

By late December, the heterotrophic flagellate communitywas better developed (table 2). The decrease in bacterioplank-ton abundance and size during January suggests that proto-zoan grazing more strongly influences bacterioplankton duringthe declining phase of the Phaeocystis sp. bloom.

Understanding bacterivory and its impact on bacterioplank -ton requires better estimates of protozoan grazing than arecurrently available for polar systems (table 2). We are analyzing

ANTARCTIC JOURNAL

Table 2. McMurdo Sound, 1989-1990 field season. Seasonalchanges in abundance and estimated clearance rates of

heterotrophic flagellates at 15 meters. A clearance rate of 5nanoliters per heterotrophic flagellate per day determined on 19January 1990 from uptake of fluorescently labelled bacteria (Puttand Miceli unpublished data) was used to calculate a preliminary

estimate of community clearance rate.

Date StationAbundance Clearance rate

26 Nov 1989H 0.1 127 Nov 1989IE 0.2 226 Dec 1989H 1.5 1827 Dec 1989IE 0.9 11

a Includes heterotrophic dinoflagellates in the nanoplankton (1-20 mi-crometer) size category. Shown in cells per liter x 106.b Percent at 15 meters per day.

data collected during the 1989-1990 field season to estimatemore accurately bacterivory at both the single species and com-munity level in McMurdo Sound.

L. Davis, E. Zettler, S. Kottmeier, and the staff of the VXE-6 and USCG helicopter operations assisted greatly with thisproject. This work was supported in part by National ScienceFoundation grant DPP 89-16840 awarded to M. Putt and D.K.Stoecker.

References

Ammerman, J.W., J.A. Fuhrman, A. Hagstrom, and F. Azam. 1984.Bacterioplankton growth in seawater. I. Growth kinetics and cellularcharacteristics in seawater cultures. Marine Ecology Progress Series,18, 31-39.

Andersson, A., U. Larsson, and A. Hagstrom. 1986. Size selectivegrazing by a microflagellate on pelagic bacteria. Marine Ecology Prog-ress Series, 33, 51-57.

Borsheim, KY., G. Bratbak, and M. Heldal. 1990. Enumeration andbiomass estimation of planktonic bacteria and viruses by transmis-sion electron microscopy. Applied and Environmental Microbiology, 56,352-356.

Gonzalez, J.M., E.B. Sherr, and B.F. Sherr. 1990. Size-selective grazingon bacteria by natural assemblages of estuarine flagellates and cil-iates. Applied and Environmental Microbiology, 56, 583-589.

Guillard, R.R.L., and J. A. Hellebust. 1971. Growth and the productionof extracellular substance by two strains of Phaeocystis pouchetii. Jour-nal of Phycology, 7, 330-338.

Kottmeier, ST., S. McGrath Grossi, and C.W. Sullivan. 1987. Sea icemicrobial communities. VIII. Bacterial production in annual sea iceof McMurdo Sound, Antarctica. Marine Ecology Progress Series, 35,175-188.

Psenner, R. 1990. From image analysis to chemical analysis of bacteria:A long term study? Limnnology and Oceanography, 35, 234-237.

Verity, PG., and T.J. Smayda. 1989. Nutritional value of Phaeocystispouchetii (Prymnesiophyceae) and other phytoplankton for Acartiaspp. (Copepoda): Ingestion, egg production, and growth of nauplii.Marine Biology, 100, 161-171.

Quantitative analysisof the vertical distributionof overstory macroalgae

near Anvers Island, Antarctica

CHARLES D. AMSLER,* DAVID R. LAuI, LANGDON B. QUETIN,

ROBERT J . ROWLEY, ROBIN M. Ross, and MICHAEL NEUSHUL

Marine Science InstituteUniversity of California

Santa Barbara, California 93106

In response to the grounding of the Ba/iia Paraiso, the Na-tional Science Foundation sent our research team to PalmerStation to assess the impact of the accident on the environ-ment. We had the opportunity to describe quantitatively theepilithic macroalgal and invertebrate communities near AnversIsland. Although the first scuba-diving observations of theantarctic benthos (in early 1958) were primarily of macroalgae(Neushul 1959, 1961, 1965), only DeLaca and Lipps (1976) andRichardson (1979) have quantitatively described antarctic ma-croalgal vertical distribution. DeLaca and Lipps (1976) de-scribed two transect sites, also near Palmer Station, apparentlybased on single 1-square-meter quadrats per depth.

* Present address: Department of Microbiology and immunology, Univer-sity of Illinois at Chicago, Chicago, Illinois 60680.

In the present investigation, five study sites were locatednear Anvers Island in March or early April 1989: one on thesoutheast side of DeLaca Island, one each on the north andsouth sides of Hermit Island, and two on the northeastern sideof Dream Island (figure 1). All sites were located adjacent tosteep, rocky shores. The rock substrate extended to at least a20-meter depth subtidally except at the north side of HermitIsland site. Permanently marked transects were established ateach site. Transect locations and underwater maps are avail-able from the authors. Nine-meter-long transects were locatedalong depth contours at 2, 5, 10, 15, and 20 meters, but notall depths were sampled at each site (table). The cover of ma-croalgal overstory was recorded over 1-by-9 meter paths alongthe length of each transect using an underwater 8 millimetervideo camera. Six 0.78-square meter quadrats (78 by 100 cen-timeters) with 25 regularly spaced dots per quadrat were su-perimposed over the video images at regular intervals alongeach transect. The macroalga under each dot was identifiedand percent cover for each taxon determined (Littler and Littler1985). Voucher collections of macroalgae for species identifi-cation were made from areas immediately adjacent to eachtransect. In the summary analysis of data from all sites, equalweight was given to each depth even though not all depthswere sampled at each site (table). The overstory macroalgalcover was quantitatively removed from three 0.125-square-me-ter quadrats (35.4 by 35.4 centimeters) along each transect andthe substrate video-recorded. The percentage cover of encrust-ing algae was determined from these recordings using 25 reg-ularly spaced dots superimposed over the video image. Allinvertebrates and remaining non-encrusting algae from thesequadrats were subsequently removed for later quantification.

1990 REVIEW 201