J.I. Saiz-SalinasDpto de Biologia Animal y Genetica, Fac. de Ciencias,Universidad del Pais Vasco, E-48080 Bilbao, Vizcaya, Spain

A. Ramos ( )Instituto Espan8 ol de Oceanografıa, Puerto Pesquero s/n,E-29640 Fuengirola, Malaga, Spainfax: 34-5-2463808; e-mail: bentart@ccuma.sci.uma.es

F.J. GarcıaLaboratorio de Biologia Marina, Fac. de Biologia,Avda Reina Mercedes 6, Universidad de Sevilla,E-41080 Sevilla, Spain

J.S. TroncosoDpto de Recursos Naturales y Medio Ambiente, Fac. de Ciencias,Universidad de Vigo, E-36200 Vigo, Pontevedra, Spain

G. San MartinDpto de Biologia, Fac. de Ciencias, Universidad Autonomade Madrid, E-28049 Madrid, Spain

C. SanzDpto de Didactica, Ciencias Experimentales y Matematicas,Universidad de Barcelona, E-08014 Barcelona, Spain

C. PalacinDpto de Biologia Animal, Fac. Biologicas, Universidad deBarcelona, Avda Diagonal, 645, E-08028 Barcelona, Spain

Polar Biol (1997) 17: 393—400 ( Springer-Verlag 1997

ORIGINAL PAPER

J.I. Sa{ iz-Salinas · A. Ramos · F.J. Garci{aJ.S. Troncoso · G. San Martin · C. Sanz · C. Palacin

Quantitative analysis of macrobenthic soft-bottom assemblagesin South Shetland waters (Antarctica)

Received: 7 March 1996/Accepted: 26 July 1996

Abstract Macrobenthic assemblages were investigatedat 26 stations located around Livingston Island, De-ception Island and the Bransfield Strait at depthsranging from 42 to 671 m. Representatives of 30 majortaxa were found. The maximal density was 5,260 speci-mens · m~2 at Livingston Island; the mean abundanceper station ranged from 160 to 4,380 specimens ·m~2.The total biomass of the macrozoobenthos declinedwith depth, with mean values of 3,201 g ·m~2 at shal-lower depths ((100 m) and 210 g · m~2 further down('100 m). After multivariate analysis (cluster analysis,MDS) based on Bray-Curtis dissimilarities, most sta-tions could be assigned to one of three groups onthe basis of distinct biomass differences between sites.The first cluster with a rich Ascidiacea biomass is

common on shallower bottoms. The second, withOphiuroidea as a characteristic group, is common ondeeper bottoms. The absence of an ‘indicator’ taxon ischaracteristic of the remaining cluster of those stationswith the lowest biomass values. No significant correla-tions were detected between macrobenthic biomass andany sediment parameters measured, probably becausepart of the benthos (i.e. the epifauna) could be betterexplained by the coupling with a highly productivewater column. The role of the epi-infauna sensu Gal-lardo as the main factor structuring benthic assem-blages in the investigated area is discussed.

Introduction

Research over the past few decades has established thatthe Antarctic benthos, especially on the shelf, is charac-terized by high biomass levels (Knox 1994). However,according to Gutt (1991), biomass exhibits extremevariation and thus wide variation rather than highbiomass is to be regarded as characteristic of the Ant-arctic benthos. Suspension-feeding epifauna in Antarc-tic waters appear to have a biomass one or two ordersof magnitude larger than that of deposit-feeding in-fauna (Gallardo 1987). A third benthic component,‘epi-infauna’, was proposed by Gallardo (1987). Thisterm was coined for the Antarctic epifaunal sessilesuspension feeders occurring on soft substrata. Thissoft-bottom epi-infauna shares many attributes withthe epifauna living on hard substrates, such as largebiomass and a similar feeding mode.

The South Shetland Islands are one of the principalAntarctic investigation areas (Arntz et al. 1994), wheretwo Sites of Special Scientific Interest have been desig-nated. Studies were mainly carried out in more or lessclosed bays of Greenwich Island (Gallardo and Castillo1969; Gallardo et al. 1977), Deception Island (Retamal etal. 1982), King George Island (Jazdzewski et al. 1986) andAnvers Island (Lowry 1975; Richardson and Hedgpeth

Fig. 1 Position of the samplingstations during the BENTART-95 cruise

1977). Gallardo (1992) has summarized all benthic fieldstudies carried out to date at the South Shetland Is-lands. The results of Muhlenhardt-Siegel (1988, 1989)were based on dredge hauls around the South Shetland,Elephant and South Orkney Islands in both summerand winter.

Spain as a member state of the SCAR, has manageddifferent Antarctic scientific programmes close to itsAntarctic station (Base Juan Carlos I) on LivingstonIsland. Two projects, BENTART-94 (Olaso 1994) andBENTART-95 (Ramos 1995), were devoted to studyingthe fauna of the seabed around Livingston and Decep-tion Islands and the Bransfield Strait. The aims of thepresent paper are (1) to distinguish faunal assemblages

and (2) to identify environmental variables probablyresponsible for the distribution of the benthic assem-blages.

Materials and methods

Field sampling

The research programme of the BENTART-95 cruise aboard RVHesperides was carried out from 16 January to 4 February 1995,during the austral night. Twenty-six stations located around Livin-gston Island, Deception Island and in the Bransfield Strait in water42—671 m deep (Fig. 1, Table 1) were sampled using a quantitativevan Veen (VV) grab of 0.1 m2 sampling area and a boxcorer (BC)

394

Tab

le1

Sta

tion

list

with

loca

tion

and

envi

ronm

enta

lpa

ram

eter

s.A

llse

dim

ent

par

amet

ers

refe

rto

the

surfac

eof

the

seab

ed.

(DD

ecep

tion

Isla

nd,¸

Liv

ings

ton

Isla

nd,

BB

ransfi

eld

Stra

it,n

dnot

dete

rmin

ed)

Stat

ion

D1

D2

L3

L4

L5

L6

L7

L8

L9

L10

D11

D12

L13

Lat

itud

eS

62°5

5@62

°56@

62°3

7@62

°38@

62°4

1@62

°43@

62°4

4@62

°44@

62°3

9@62

°40@

62°5

7@62

°57@

62°3

8@L

ongi

tude

W60

°36@

60°3

8@60

°23@

60°2

5@60

°31@

60°2

7@60

°28@

60°3

0@60

°39@

60°3

8@60

°39@

60°3

6@60

°41@

Dep

th(m

)46

.115

3.4

68.5

182.

526

3.8

121.

476

.211

417

0.4

216.

316

2.8

162.

845

.1pH

(atse

abed

surfac

e)7.

296.

987.

067.

287.

177.

17.

277.

017.

147.

26.

346.

057.

15R

edox

pote

ntial

atse

abed

surfac

e(m

V)

248

!40

!24

!19

!16

!10

0!

75nd

!83

72!

9721

962

.8C

arbo

nate

s(%

)3.

125

3.12

53.

125

3.90

639

.06

3.12

53.

906

4.68

839

.06

39.0

64.

688

3.12

539

.06

Sedi

men

torg

anic

mat

ter

(%)

4.66

94.

212

7.45

89.

861

9.17

78.

198

7.18

45.

597.

398

9.73

5.75

49.

062

8.66

7M

edia

npar

ticl

esize

(lm

)45

.77

22.9

413

.85

9.39

59.

898

7.27

89.

867

6.87

227

.44

13.7

110

.18

12.2

924

.22

Sort

ing

coeffi

cien

t2.

592.

772.

792.

272.

262.

217

2.63

2.04

3.22

2.73

2.41

2.25

3.14

Stat

ion

L14

L15

L16

D17

D18

L19

L20

L21

D22

B23

B24

B25

L29

Lat

itud

eS

62°3

8@62

°45@

62°4

5@62

°59@

62°5

7@62

°44@

62°4

5@62

°50@

63°0

3@63

°57@

63°5

8@63

°56@

62°0

5@L

ongi

tude

W60

°40@

60°3

5@60

°34@

60°3

5@60

°39@

60°3

1@60

°31@

60°2

4@60

°39@

60°5

7@60

°52@

60°4

5@60

°25@

Dep

th(m

)72

.435

2.2

416

107.

816

2.4

240

393.

964

9.3

178.

810

328

644

023

9.5

pH(a

tse

abed

surfac

e)7.

127.

37.

176.

415.

727.

637.

337.

567.

587.

44nd

7.12

7.11

Red

oxpo

tential

atse

abed

surfac

e(m

V)

!54

182

127

87.9

169

5nd

273

nd

112.

7nd

nd10

1C

arbo

nate

s(%

)54

.69

3.90

639

.06

3.12

53.

906

3.12

5nd

3.90

63.

906

18.7

5nd

4.68

83.

125

Sedi

men

torg

anic

mat

ter

(%)

9.46

98.

106

8.07

74.

325

5.22

29.

188

nd

2.82

81.

976

11.3

1nd

5.14

33.

585

Med

ian

par

ticl

esize

(lm

)18

.99

12.9

13.6

824

.52

10.5

412

.28

nd

539.

618

0.7

10.0

2nd

38.2

638

.89

Sort

ing

coeffi

cien

t2.

892.

492.

662.

382.

272.

6nd

1.31

1.3

2.57

nd

4.1

3.69

395

Fig. 3 Dendrogram using complete linkage. Three main clusters aredelineated, characterized by Ascidiacea, Ophiuroidea and low bio-mass values

Fig. 2 Mean biomass (solid bars, kg WW ·m~2), densities (light bars,specimens ·m~2]1,000) and number of taxa (line) at the samplingstations

with a maximum 60-cm breakthrough and an effective samplingarea of 30]20 cm. Four VV grabs plus 1 BC were accomplished ateach station. The first VV grab was used immediately after samplingfor measures of temperature and pH at the surface with a HannaHI 9025c pH meter, and later subsampled for sediment grain sizedistributions, organic content and carbonates. The BC was used toobtain vertical profiles of redox (Eh) values, measured with an OrionORP 9678 electrode coupled to a Hanna model HI 9025c mV meter,and for subsequent meiobenthic studies. Readings of Eh were takenafter an arbitrary 60-s period (Pearson and Stanley 1979). A stan-dard redox solution (Hanna HI 7020) was used as reference. Thethree remaining replicate VV grabs were used to obtain means andstandard deviations of the quantified biotic variables. The content ofeach replicate was sieved using three mesh sizes (5, 1 and 0.5 mm)and later separated into major taxonomic groups (Table 2). Onlyspecimens retained on a 1.0-mm screen were counted on board theship and their wet weight (WW) biomass was determined accordingto the recommendations of Jazdzewski et al. (1986) and Muhlen-hardt-Siegel (1988). Finally, the biological material was preservedand distributed to taxonomists in Spain for further identification.Analyses of sediment particle size distribution, organic content (loss-on-ignition method) and carbonates were performed followingBuchanan (1984).

Data analyses

Data analysis followed the scheme proposed by Field et al. (1982).‘Root-root’ transformed biomass values were used to constructa Bray-Curtis dissimilarity matrix. Classification was performedusing the complete linkage clustering technique, whereas ordination(multidimensional scaling, MDS) was used to evaluate the groupseparation derived by cluster analysis. ‘Indicator’ taxa separatingeach station group were identified by the magnitude of the Kruskal-Wallis test. H values were calculated for each taxon and listed inorder of the magnitude of its contribution to the different stationgroups. Spearman’s rank correlation coefficients were used to assessthe strength and direction of the relationship between macrobenthicbiomass and environmental variables (pH, Eh, organic matter, car-bonates, median particle size, sorting coefficient of sediments andwater depth).

Results

Number of taxa, abundance and biomass

Figure 2 presents the number of taxa as well as themean abundance and biomass at each sampling sta-tion. The stations at Deception Island, all taken insidePort Foster, were characterized by relatively low valuesfor all three parameters (6—10 taxa, 160—876 speci-mens · m~2, 50—220 g ·m~2), except for the relativelyshallow station D1 (45 m depth) where the highestbiomass was found (6,673 g ·m~2). The stations ofLivingston Island exhibited large variations in thenumber of taxa, abundances and biomass. The highestbiomass value was recorded at L13 (5,205 g · m~2),the lowest (23.3 g · m~2) at the deep offshore stationL21 (653 m). Abundances varied between 4,380 speci-mens · m~2 (L19) and 240 specimens ·m~2 (L21), thenumber of taxa between 32 (L7) and 9 (L10).

Overall, the average macrobenthic biomass was1,169 g ·m~2. Ascidians accounted for almost half of it

(46.7%), followed by sponges (19.9%) and polychaetes(13.4%); the rest (20%) was distributed among severalgroups such as bryozoans, bivalves and echinoderms.In terms of abundance, polychaetes were the mostimportant group (about 50%), followed by bivalves.Crustaceans, echinoderms, ascidians, sponges andother groups had much lower densities.

Benthic assemblages

Three main groups could be discriminated both in thecluster dendrogram (Fig. 3) and the MDS plot (Fig. 4).The first cluster, ‘Ascidiacea dominance’ (A), consistedof shallower stations (mostly(100 m) strongly domin-ated in biomass by epifaunal Ascidiacea (Fig. 5,Table 2). Together with other filter-feeding taxa,such as sponges and bryozoans, they accounted forover 85% of the mean WW biomass (3,201 g · m~2).

396

Fig. 4 MDS plot. Stress is 0.16. Station groups delineated in thedendrogram in Fig. 3 are also present here

Fig. 5 Biomass proportions of major taxonomic components in thestation clusters distinguished in the study area. Pies area is propor-tional to the mean biomass (g wet weight · m~2). A Ascidiaceadominance, O Ophiuroidea abundance, ¸B low biomass (no ‘indi-cator’)

The second cluster ‘Ophiuroidea abundance’ (O) andthe third cluster ‘low biomass’ (LB) were formed bydeeper stations (mostly'100 m), characterized by thepresence of epibenthic Ophiuroidea and the absence ofa distinctive epifauna, respectively. The biomass in thetwo last clusters was dominated by polychaetes (Fig. 5).The third cluster (LB) included stations with the lowestmean biomass (102 g ·m~2), while that of cluster O wasslightly higher (271 g ·m~2).

Ordination based on MDS was used to evaluate thegroup separation derived by the cluster analysis. Thetwo-dimensional plot (Fig. 4) corroborated the results

of the classification. Two major groups (A and LB) aresegregated along dimension 1, which can be identifiedas an ‘increasing biomass gradient’. Group O withintermediate biomass values and populated distinc-tively by an Ophiuroidea epifauna remains in thecentre of the plot.

Non-parametric Kruskal-Wallis tests of biomass dif-ferences among the clusters, performed for each taxon,confirmed the importance of epifaunal Ascidiacea(H"21.4; P(0.0001) and Ophiuroidea (H"11.1;P"0.0038) as the core taxa defining cluster groupsA and O, respectively. However, no ‘indicator’ taxonwas found for the remaining group LB. Using majortaxonomic categories, a particular infaunal taxon couldnot be found to explain the existence of this thirdgroup.

Relationship between biotic and environmentalvariables

The sediments at the stations investigated consistedmainly of poorly sorted silty clays, with relatively highcontents of organic matter (Table 1). Carbonate con-tents were low, probably due to strong terrestrial in-fluences; higher values were found close to Walker Bay.The superficial sediments appeared to be oxidized,as shown by Eh values greater than !100 mV. ThepH values in surface sediments were below 7.0 at sta-tions located inside Port Foster (Deception Island). Itremains to be demonstrated these values indicate theprevalence of acidic conditions induced by episodesof volcanic activity within this sea-breached caldera.Spearman’s rank correlations between biomass dataand abiotic environmental variables (Table 1) showedthe absence of significant values, with the exception ofdepth (r

4"!0.57).

Discussion

WW biomass values from the Antarctic peninsula andadjacent islands range from 51 to 46,000 g ·m~2,with a grand mean of about 1,000 g ·m~2 (Gutt 1991).According to Knox (1994), epifaunal biomass is gen-erally much greater than infaunal biomass (up to4,200 g ·m~2 and up to 700 g · m~2, respectively). Theaverage biomass in our study (1,169 g ·m~2) is in theranges of those previously compiled in the literature byJazdzewski et al. (1986), Muhlenhardt-Siegel (1988),Gerdes et al. (1992) and Knox (1994) from differentAntarctic locations.

The absence of significant correlations between mac-robenthic biomass and any of the sediment parametersexamined suggests that water depth (or depth-relatedfactors) plays the most important role in structuring themacrobenthic biomass along a vertical gradient of

397

Tab

le2

Mac

rozo

oben

thic

taxa

inth

ew

ater

soft

he

inve

stig

ated

area

.Mea

ns($

stan

dar

ddev

iations)

ofab

undan

ce(spec

imen

s·m

~2 )

and

bio

mas

s(g

wet

wei

ght·

m~2 )

are

give

nfo

ral

lthest

atio

nsco

mbi

ned

asw

ell

asfo

rea

chof

the

thre

eben

thic

asse

mbl

ages

distingu

ished

Tota

lA

ssem

bla

ge‘‘A

’’A

ssem

bla

ge‘‘L

B’’

Ass

embl

age

‘‘O’’

Bio

mas

sA

bundan

ceBio

mas

sA

bundan

ceBio

mas

sA

bundan

ceBio

mas

sA

bundan

ce

Porife

ra19

9.05

8($

818.

236)

1.59

6($

4.38

4)76

2.5

($1,

461.

829)

4.45

0($

7.37

2)0

($0)

0($

0)0.

278

($0.

622)

0.83

3($

1.98

8)A

ctin

aria

0.14

5($

0.47

1)0.

435

($1.

495)

0.27

8($

0.62

2)1.

117

($2.

497)

0($

0)0

($0)

0.13

9($

0.46

2)0.

275

($0.

912)

Hyd

roid

ea0.

218

($0.

562)

0.57

8($

1.60

1)0.

278

($0.

622)

1.11

7($

2.49

7)0

($0)

0($

0)0.

278

($0.

622)

0.55

0($

1.23

0)G

orgo

nar

ia0.

507

($2.

052)

0.28

7($

0.93

0)1.

945

($3.

654)

1.10

0($

1.55

6)0

($0)

0($

0)0

($0)

0($

0)B

ryozo

a53

.317

($16

4.65

8)5.

217

($12

.819

)19

9.94

5($

273.

499)

18.9

00($

19.3

23)

0.66

6($

1.33

2)0.

660

($1.

320)

1.94

2($

4.39

4)0.

275

($0.

912)

Nem

ertini

3.98

5($

13.7

96)

5.50

4($

11.6

51)

11.9

38($

25.2

35)

12.2

17($

19.7

71)

0.33

4($

0.66

8)1.

340

($2.

680)

1.94

2($

1.72

9)3.

883

($5.

419

Nem

atod

a0.

507

($1.

035)

2.02

6($

5.18

0)1.

110

($1.

570)

6.11

7($

8.70

3)0.

668

($0.

818)

1.32

0($

1.61

7)0.

139

($0.

462)

0.27

5($

0.91

2)Ech

iura

6.21

7($

29.1

62)

0.14

3($

0.67

3)0

($0)

0($

0)0

($0)

0($

0)11

.917

($39

.523

)0.

275

($0.

912)

Priap

ulid

a0.

870

($1.

379)

2.31

3($

3.85

8)0.

557

($0.

787)

1.10

0($

1.55

6)1.

332

($1.

631)

4.66

0($

5.80

2)0.

834

($1.

444)

1.94

2($

3.18

1)Sip

uncu

la1.

740

($2.

113)

14.4

91($

30.3

36)

4.16

7($

2.09

8)25

.550

($18

.609

)1.

002

($0.

818)

6.02

0($

6.12

3)0.

834

($1.

444)

12.4

92($

38.4

71)

Poly

chae

ta13

3.67

0($

110.

112)

794.

900

($83

7.02

1)19

0.33

3($

142.

729)

575.

500

($31

7.62

8)80

.080

($59

.026

)45

8.66

0($

241.

340)

127.

667

($93

.613

)1,

044.

700

($1,

065.

140)

Olig

och

aeta

0.14

5($

0.47

1)5.

070

($21

.117

)0

($0)

0($

0)0.

334

($0.

668)

20.6

60($

41.3

20)

0.13

9($

0.46

2)1.

108

($3.

676)

Hirudin

ea0.

073

($0.

341)

0.14

3($

0.67

3)0

($0)

0($

0)0

($0)

0($

0)0.

139

($0.

462)

0.27

5($

0.91

2)Sole

noga

stra

0.43

5($

0.88

2)1.

304

($3.

650)

0.83

5($

0.83

5)3.

333

($6.

098)

0.33

4($

0.66

8)0.

660

($1.

320)

0.27

8($

0.92

0)0.

558

($1.

852)

Gas

tropoda

4.05

7($

7.68

4)7.

678

($11

.400

)6.

667

($6.

597)

21.6

67($

13.2

97)

0.33

4($

0.66

8)0.

660

($1.

320)

4.30

3($

9.05

5)3.

608

($5.

002)

Sca

phopoda

1.37

7($

2.12

3)6.

059

($10

.127

)0

($0)

0($

0)1.

000

($1.

333)

6.00

0($

9.03

1)2.

223

($2.

485)

8.60

8($

11.5

00)

Biv

alvi

a14

.057

($14

.305

)16

3.18

7($

212.

671)

13.0

55($

7.28

7)25

5.00

0($

322.

692)

9.98

0($

4.06

6)67

.340

($42

.340

)16

.256

($18

.615

)15

7.21

7($

160.

619)

Pyc

nogo

nid

a1.

305

($1.

901)

4.92

2($

8.83

2)2.

223

($2.

078)

11.6

67($

12.4

16)

0.33

4($

0.66

8)0.

660

($1.

320)

1.25

0($

1.93

9)3.

325

($6.

232)

Cum

acea

1.66

7($

2.82

3)6.

083

($10

.340

)4.

443

($4.

158)

16.6

67($

14.6

58)

0($

0)0

($0)

0.97

3($

1.06

7)3.

325

($4.

299)

Isopoda

2.02

8($

3.86

2)2.

752

($3.

361)

4.71

8($

6.47

3)3.

883

($3.

563)

0.66

6($

1.33

2)1.

340

($2.

680)

1.25

0($

1.38

1)2.

775

($3.

292)

Am

phip

oda

6.01

7($

5.26

8)53

.917

($98

.656

)11

.400

($6.

696)

141.

117

($16

0.44

5)2.

334

($1.

669)

7.32

0($

4.89

6)4.

861

($2.

843)

29.7

33($

21.9

63)

Ost

raco

da

0.21

8($

0.56

2)0.

430

($1.

111)

0.27

8($

0.62

2)0.

550

($1.

230)

0.33

4($

0.66

8)0.

660

($1.

320)

0.13

9($

0.46

2)0.

275

($0.

912)

Tan

aidac

ea0.

290

($0.

800)

0.86

5($

2.80

9)1.

112

($1.

242)

3.31

7($

4.70

2)0

($0)

0($

0)0

($0)

0($

0)A

ster

oid

ea7.

448

($23

.041

)2.

609

($4.

173)

26.6

07($

39.1

75)

5.00

0($

5.70

9)0

($0)

0($

0)0.

973

($1.

265)

2.5

($3.

373)

Ophiu

roid

ea31

.578

($33

.502

)56

.904

($60

.640

)31

.943

($29

.226

)10

8.31

7($

82.6

94)

1.66

8($

1.05

3)17

.980

($5.

571)

43.8

58($

34.6

42)

41.6

67($

36.8

78)

Ech

inoid

ea21

.939

($55

.171

)3.

913

($5.

797)

21.6

62($

28.7

32)

7.78

3($

7.13

2)0.

666

($1.

332)

1.34

0($

2.68

0)30

.942

($71

.770

)3.

050

($4.

999)

Hol

othuro

idea

30.6

96($

71.5

72)

4.78

7($

9.21

5)84

.112

($10

7.86

8)12

.783

($13

.664

)0

($0)

0($

0)16

.778

($44

.581

)2.

783

($4.

887)

Ente

ropneu

sta

0.29

0($

0.80

0)1.

161

($3.

050)

0.55

5($

1.24

1)1.

667

($3.

727)

0($

0)0

($0)

0.27

8($

0.62

2)1.

392

($3.

184)

Asc

idia

cea

466.

841

($1,

293.

461)

61.0

13($

169.

588)

1,78

8.16

7($

2,01

2.76

3)23

1.66

7($

266.

150)

0.33

4($

0.66

8)0.

660

($1.

320)

0.55

6($

1.03

9)0.

833

($1.

988)

Ost

eich

thye

s7.

826

($36

.708

)0.

291

($1.

366)

30.0

00($

67.0

82)

1.11

7($

2.49

7)0

($0)

0($

0)0

($0)

0($

0)

398

increasing depth. Since benthic assemblages dependheavily on the supply of food from the water column,the vertical pattern of biomass in the investigated areasuggests a possible response to the degree of plank-ton/benthos interaction.

As has been recently noted (Dayton et al. 1994), theAntarctic seas show two contrasting situations: ahighly productive plankton system capable of sup-porting large populations of tetrapod vertebrates andcephalopods, but benthic assemblages characterized bylow growth rates and secondary production. Since theepifaunal filter-feeders obtain their food from sedi-menting phytoplankton and resuspension from thesediments (Knox 1994), they are most prevalent atshallower depths (Barnes and Hughes 1988) where theycan achieve large biomasses provided that sufficientconcentrations of suspended particulate organic matterare available in the water column. However, deposit-feeding infauna will be succesful on sediments regard-less of their bathymetric position, although largequantities of sediment may have to be sorted and/orconsumed (Barnes and Hughes 1988) to concentratea comparable quantity of food. It is always morerewarding to wipe off particles retained on passivefilters placed in the water column than to swallow largequantities of soft organic-rich sediments with theirlarge percentage of inorganic ash. In fact, Gallardo(1987) suggested that the relatively small biomasses ofinfaunal organisms on Antarctic soft bottoms werecaused by filter-feeders reducing the food materialreaching the sediments. The group of shallow-stations(A), delimited in the classification and ordinationmethods, is defined by the dominance of suspension-feeders (Ascidiacea, Porifera and Bryozoa) anddelineates the most eutrophic environments in the in-vestigated area. Three stations (L6, L7 and L8) are closeto a cape (Miers Bluff ), suggesting the existence ofan area of relatively rapid water flow off LivingstonIsland, which promotes a high biomass of suspension-feeders.

Both deeper station clusters O and LB showed a de-posit-feeding polychaete biomass dominance. The maindifference between them lies in the presence in clusterO of a rich epifauna of ophiuroids, some of which areimportant predators in the Antarctic (Sieg and Wagele1990; Dayton 1990; Arntz et al. 1994). In fact, aprey-predator dependence can be suggested, sinceophiuroids are absent at those stations with the lowestbiomass values (corresponding to station cluster LB).Stomach content analysis of aggregated ophiuroidson deep bottoms of the Scotia and Weddell Seas(Sokolova 1993) demonstrated that they used a con-tinuous ‘rain of planktonic corpses’ as a food source.Thus, large aggregations of ophiuroids may indicatethat aphotic conditions on deeper bottoms ('100 m)are strongly affected by the large productivity of theeuphotic water masses above. On the other hand, lowendobenthic biomass (as shown by the LB cluster

group) can be used to detect oligotrophic areas withlow levels of pelagic production (Ahn and Kang1991).

The differences in faunal composition between sta-tions are probably due to different trophic dynamics.The relationship between pelagic productivity andbenthic biomass was studied by Dayton and Oliver(1977). They related the existence of a high benthicbiomass to eutrophic water columns, whereas lowerbenthic biomasses resulted from oligotrophic, nutrient-poor water flowing above the benthos. Shallow watersclose to the coast of the investigated area probablydevelop high productivity in the euphotic zone, promo-ting the dominance of a suspension-feeding epifauna tosuch an extent that infaunal assemblages appearsmothered by it (Gallardo 1987). In fact, the high con-centrations of particles measured in the Weddell Seawere limited to a depth of about 80 m (Rabitti et al.1990). On the other hand, deep aphotic water massescannot sustain a rich epifauna, promoting the domi-nance of a deposit-feeding infauna unable to attain thelarge biomasses of the epifauna, since the food qualityof soft organic-rich sediments is always lower due totheir inherently higher percentage of inorganic ash(Valiela 1984).

In conclusion, our results show (1) a trophic zona-tion of the Antarctic benthos based on water depth and(2) the displacement of the deposit-feeding infauna todeeper bottoms when competing for food and spacewith the suspension-feeding epifauna in the shallowerwaters of nutrient-enriched environments, as suggestedpreviously by Gallardo (1987).

Acknowledgements The BENTART cruises were carried out underthe auspices of two Spanish Council of Scientific and TechnicalResearch (CICYT) Antarctic Programmes (no. ANT 93—0996 andno. ANT 94—1161-E). The completion of the work in Spain wasachieved under contract no. ANT 95—1011 of the Spanish CICYT.We would like to express our thanks to Antonio Flores for reviewingthe English version of the manuscript, Monica Jimenez for drawingthe figures and four external reviewers for their constructive com-ments and suggestions. The officers and crew of RV Hesperidesplayed a vital part in the materialization of this project, and weexpress our gratitude to all of them.

References

Ahn IY, Kang YC (1991) Preliminary study on the macrobenthiccommunity of Maxwell Bay, South Shetland Islands, Antarctica.Korean J Polar Res 2: 61—71

Arntz WE, Brey T, Gallardo VA (1994) Antarctic zoobenthos. In:Ansell AD, Gibson RN, Barnes M (eds) Oceanography andmarine biology: an annual review, vol. 32. UCL Press, London,pp 241—304

Barnes RSK, Hughes RN (1988) An introduction to marine ecology,2nd edn. Blackwell Scientific, Oxford

Buchanan JB (1984) Sediment analysis. In: Holme NA, McIntyreAD (eds) Methods for the study of marine benthos. BlackwellScientific, Oxford, pp 41—65

399

Dayton PK (1990) Polar benthos. In: Smith WO (ed) Polar oceano-graphy. Part B: chemistry, biology and geology. Academic Press,London, pp 631—685

Dayton PK, Oliver JS (1977) Antarctic soft-bottom benthos inoligotrophic and eutrophic environments. Science 197: 55—58

Dayton PK, Mordida BJ, Bacon F (1994) Polar marine communi-ties. Am Zool 34: 90—99

Field JG, Clarke KR, Warwick RM (1982) A practical strategy foranalysing multispecies distribution patterns. Mar Ecol Prog Ser46: 7—12

Gallardo VA (1987) The sublittoral macrofaunal benthos of theAntarctic shelf. Environ Int 13: 71—87

Gallardo VA (1992) Estudios bentonicos en bahias someras antar-ticas de Archipielago de las Islas Shetland del Sur. In: GallardoVA, Ferretti O, Moyano HI (eds) Oceanografıa en Antartica.ENEA — Proyecto Antartica, pp 383—393

Gallardo VA, Castillo JC (1969) Quantitative benthic survey of theinfauna of Chile Bay (Greenwich Island, South Shetland Islands).Gayana Zool 16: 3—17

Gallardo VA, Castillo JG, Retamal MA, Yan8 ez A, Moyano HI,Hermosilla JG (1977) Quantitative studies on the soft-bottommacrobenthic animal communities of shallow Antarctic Bays. In:Llano GA (ed) Adaptations within Antarctic ecosystems. Proc3rd SCAR Symp Antarct Biol. Smithsonian Institution, Wash-ington DC, pp 361—387

Gerdes D, Klages M, Arntz WE, Herman RL, Galeron J, Hain S(1992) Quantitative investigations on macrobenthos communi-ties of the southeastern Weddell Sea shelf based on multiboxcorer samples. Polar Biol 12: 291—301

Gutt J (1991) Are Weddell Sea holothurians typical representativesof the Antarctic benthos? A comparative study with new results.Meeresforschung 33: 312—329

Jazdzewski K, Jurasz W, Kittel W, Presler P, Sicinski J (1986)Abundance and biomass estimates of the benthic fauna in Ad-miralty Bay, King George Island. Polar Biol 6: 5—16

Knox GA (1994) Benthic communities. In: The biology of the South-ern Ocean. Cambridge University Press, New York, pp 193—220

Lowry JK (1975) Soft-bottom macrobenthic community of ArthurHarbor, Antarctica. Antarct Res 23: 1—19

Muhlenhardt-Siegel U (1988) Some results on quantitative invest-igations of macrozoobenthos in the Scotia Arc (Antarctica).Polar Biol 8: 241—248

Muhlenhardt-Siegel U (1989) Quantitative investigations of Antarc-tic zoobenthos communities in winter (May/June) 1986 withspecial reference to the sediment structure. Arch Fischereiwiss 39:123—141

Olaso I (1994) Campan8 a Bentart 94 en el B.I.O. Hesperides enfebrero de 1994. Instituto Espan8 ol de Oceanografıa, Santander(Final Report for the Spanish CICYT)

Pearson TH, Stanley SO (1979) Comparative measurement of theredox potential of marine sediments as a rapid means of assessingthe effect of organic pollution. Mar Biol 53: 371—379

Rabitti S, Gouleau D, Boldrin A (1990) Suspended matter, phyto-plankton and nutrients. In: Arntz W, Ernst W, Hempel I (eds)The Expedition Antarktis VII/4 (Epos leg 3) and VII/5 of RV‘‘Polarstern’’ in 1989. Ber Polarforsch 68: 50—66

Ramos A (1995) Informe de la Campan8 a Bentart-95. Instituto Es-pan8 ol de Oceanografıa, Fuengirola (Malaga) (Final Report forthe Spanish CICYT)

Retamal MA, Quintana R, Neira F (1982) Analisis cuali y cuan-titativo de las comunidades bentonicas en Bahia Foster (IslaDecepcion) (XXXV Expedicion Antartica Chilena, enero 1981).Publ Inst Antart Chil Ser Cient 29: 5—15

Richardson MD, Hedgpeth JW (1977) Antarctic soft-bottom macro-benthic community adaptations to a cold, stable, highly, pro-ductive, glacially affected environment. In: Llano GA (ed)Adaptations within Antarctic ecosystems. Gulf Publishing,Houston, pp 181—196

Sieg J, Wagele JW (eds) (1990) Fauna der Antarktis. Parey, Berlin.Sokolova MN (1993) New data on the trophic structure of macro-

benthos in the abyss of the Atlantic Ocean. In: Vinogradova NG(ed) The deep-sea bottom fauna in the southern part of theAtlantic Ocean. Trans P.P. Shirshov Inst Oceanol, vol 127.Russian Academy of Sciences, ‘‘Nauka’’, Moscow, pp 50—64(in Russian with English summary)

Valiela I (1984) Marine ecological processes. Springer, Berlin Heidel-berg, New York

.

400