DISTRIBUTION OF HEAVY METALS IN SURFACE SEDIMENTS FROMAN ANTARCTIC MARINE ECOSYSTEM

SANTIAGO ANDRADE1, ARMANDO POBLET2, MARCELO SCAGLIOLA3,CRISTIAN VODOPIVEZ4, ANTONIO CURTOSI4, ADÁN PUCCI1 and JORGE

MARCOVECCHIO1,2∗1 Instituto Argentino de Oceanografía, Bahía Blanca, Argentina;2 Universidad Nacional de Mar

del Plata, Mar del Plata, Argentina;3 OSSE, Mar del Plata, Argentina;4 Instituto AntárticoArgentino (IAA), Buenos Aires, Argentina

(∗ author for correspondence, e-mail: jorgemar@criba.edu.ar)

(Received 16 June 1998; accepted 28 October 1999)

Abstract. The concentrations of lead, cadmium, copper, chromium, iron, manganese and zinc insurface sediments collected from Potter Cove, in the25 de Mayo Island(King George Island),Antarctica, and its drainage basin, were measured by atomic absorption spectroscopy. The obtainedresults were use to determine the areal and vertical distribution of the metals of in the Cove andpotential sources of these metals to this environment. The geochemical data suggest that most ofthe metals found in Potter Cove constitute a redistribution of autochthonous materials within theecosystem. Therefore, the metal concentrations can be considered to be present at natural backgroundlevels in surface sediments.

Keywords: Antarctica, marine ecosystem, surface sediments

1. Introduction

Potter Cove is an Antartic marine ecosystem located within the 25 de Mayo Islandin The Southern Shetland Islands archipelago. This ecosystem is stable and has alarge percentage of indigenous species. Thus, its structure is simple in comparisonwith other ecosystems. The Antarctic area has been recognized as a comparativelyclosed environment (Hondaet al., 1987), basically sustained by the very slowatmospheric mixing between northern and southern zones of the Antartic Con-vergence. These natural barriers may prevent the atmospheric deposition of globalpollutants which affect many areas of the planet. However, when toxic compoundsare releaved in the area, e.g., from Research Stations such as Jubany Station, in Pot-ter Cove, the contaminants cannot be easily dispersed, and the pollutants becomelocalized, concentrated and accumulated in the sediments.

The geographical distribution of heavy metals in surface marine sediments isregulated not only by their concentrations, but also by sediment physico-chemicalcharacteristics, mineralogic composition, grain size distribution, organic mattercontent, etc. as well as several environmental conditions (marine currents, wind,continental runoff, etc.) (Salomons and Förstner, 1984).

Environmental Monitoring and Assessment66: 147–158, 2001.© 2001Kluwer Academic Publishers. Printed in the Netherlands.

148 S. ANDRADE ET AL.

Figure 1.Location of sampling stations.

Moreover, within this abiotic compartment, the transport and deposition of vari-ous natural and man-made substances are dependent on the nature of the particleswith which they are associated. In addition, and looking for an integral under-standing of trace metal processes occurring in surface sediments, several conditionsmight be included as has been opportunely pointed out by different authors: i.e.,physical and chemical characteristics of surface sediments (Salomons and Förstner,1984), metals bio-geochemical cycles involved within the considered environment(Barcellos and Lacerda, 1996), as well as the role of sediments as a dynamic sitefor biological and chemical reactions and as a food source for marine organisms(Muir et al., 1999).

Even though numerous papers have been written on the occurrence, concen-tration and distribution of heavy metals within marine sediments (Katz and Ka-plan, 1981, Ryan and Windom, 1988), data portaining to antarctic environments isscarce, and in the specific case of the Potter Cove, only two papers are available(Alam and Sadiq, 1993; Scagliolaet al., 1994).

This paper includes preliminary results on the concentration and spatial distri-bution of lead, cadmium, manganese, copper, iron and zinc of surface sediments ofPotter Cove, 25 de Mayo Island, Antarctica.

DISTRIBUTION OF HEAVY METALS IN SURFACE SEDIMENTS 149

2. Materials and Methods

This study was conducted in Potter Cove, which is a small bay located between62◦14′ and 62◦15′ S. and 58◦39′ and 58◦43′ W., in the ‘25 de Mayo Island’ (KingGeorge Island), on the Southern Shetland Archipelago, Antarctica (Figure 1). TheCove is a small fiord which exhibits an outer and internal zone separated by a risein the bottom topography to a depth of 30 m. The internal area has a maximumdepth of 50 m, and its bottom sediment is predominantly composed of mud (<

63µm). Its northern and eastern coasts are limited by several glaciers, while thesouthern boundary is formed by a sandy beach. Observations of surface waterssuggest a wind-controlled circulation pattern exists within the cove which gener-ates a cyclonic gyre promoting the input of shelf water to the cove, and delayingthe corresponding output (Kloser, 1994).

Field work and sampling were carried out during an Argentine Antarctic Sum-mer Cruise in 1994–95. Three marine transects were designed, including depthsfrom 5m, 10m, 20m and 30m. Samples of surface sediments were collected atfourteen stations, using plastic sledges following the removal of an upper layer ofgravel. Surface sediment samples were also collected from the ‘The Lagoon’ andfrom two streams which flow between the lagoon and Potter Cove (Figure 1).

All sediment samples were placed in plastic bags, and stored in a freezer (at –20 ◦C) until their analytical treatment at the laboratory. These samples were dried,at 40± 5◦C for 48 h until developing constant weight. Subsequently, they werecarefully sifted through different iron steel meshes to determine their grain sizedistribution.

Samples of seawater and freshwater were collected at the same stations, usingpolypropilene sampling bottles. These samples were vacuum-filtered through a0.45µ pore size cellulose acetate filter. The retained material on the filter (sus-pended particulate matter) was kept in a freezer, at –20◦C, until its analysis at thelaboratory.

The analytical determination of heavy metals (Pb, Zn, Cu, Fe, Mn and Cd) in thesediments and suspended particulate matter (SPM) follow the method previouslyreported by Marcovecchioet al. (1988). Subsamples of 500± 50 mg of sedimentwere removed, and digested with a perchloric and nitric acids mixture (1:3) ina thermostatic bath (at 90± 10 ◦C), up to minimum volume (less than 1 ml).Solutions were made up to 10 ml with 0.7% nitric acid, and the absorbances ofeach metal were measured by atomic absorption spectrophotometry. Membranefilters with SPM samples were digested using the same procedure.

A Shimadzu AA-640-13 atomic absorption spectrophotometer was utilized forall the analyses, working with air/acetylene flame and deuterium background cor-rection, and the data were processed with a personal computer using appropiatesoftware. Reagents of analytical grade were utilized for the blanks and calibrationcurves, and the analytical quality (AQ) was checked against areference mater-ial (Pond Sediment, R.M. #2), provided by The National Institute for Environ-

150 S. ANDRADE ET AL.

TABLE I

Percentages of recovery in the analysisof reference materials to assess analyt-ical quality

Metal Percentage of Recovery

analyzed (range)

Cd 91.4 – 99.3%

Cu 93.1 – 99.5%

Cr 92.8 – 99.2%

Fe 95.6 – 101.7%

Mn 91.2 – 97.9%

Pb 94.7 – 98.8%

Zn 96.5 – 102.3%

mental Studies (NIES) from Tsukuba, Japan (Table I). The obtained results werecopmpared through one-way analysis of variance (ANOVA).

3. Results and Discussion

3.1. TRACE METALS: MARINE SURFACE SEDIMENTS

The concentration (mean value± standard deviation for each studied transect)of selected trace metals bulk sediment samples of the surface sediments fromPotter Cove are presented in Table II and Figure 2. Most of the evaluated heavymetals exhibit a similar distribution trend along the studied area, even though theirconcentration ranges are completely different.

The spatial distribution of lead, chromium, copper, zinc and manganese con-centrations suggests that they are controlled by similar processes and that thesemetals exhibit a similar behaviour in the evaluated environment; this fact was veri-fied through an analysis of co-variance (Table III). Unlike the above mentioneddescription, iron possesses a specific spatial pattern of distribution, consideringthat its concentration was higher (in a different order of magnitude) than the othermetals along the whole studied area (Table II; Figure 2). This fact is related withiron condition of ‘abundant element’ (Salomons and Förstner, 1984), and providesa good start-point to interpret the natural background of metals in surface sediments(Schroppet al., 1990; Amínet al, 1996).

The concentrations of cadmium as determined in the sediments of the PotterCove were – in every case – below the detection limit of the utilized analyticalmethod (0.25µg g−1). These results did not agree with those previously reported

DIS

TR

IBU

TIO

NO

FH

EAV

YM

ETA

LSIN

SU

RFA

CE

SE

DIM

EN

TS

151

TABLE II

Mean, standard deviation and ranges of heavy metals concentrations detected in surface marine sediments from Potter Cove. (Expressedin µg g−1, dry weight)

Zn Cu Cr Pb Mn Fe

(µg g−1) (µg g−1) (µg g−1) (µg g−1) (mg g−1) (µg g−1)

Transect #1 n = 4 53.69± 6.09 140.97± 9.95 6.73± 0.63 3.92± 0.97 1.05± 0.05 15.51± 3.79

{45.98–63.02} {128.5–156.3} {6.01–7.72} {3.06–5.52} {0.99–1.13} {12.01–21.39}

Transect #2 n= 5 51.49± 5.79 115.30± 20.43 6.42± 1.16 3.68± 0.77 0.94± 0.02 13.26± 4.83

{45.53–59.12} {83.64–137.81} {5.09–8.11} {2.94–4.84} {0.92–0.98} {6.55–18.63}

Transect #3 n= 5 52.23± 4.78 94.11± 16.29 5.66± 1.22 3.40± 0.95 0.88± 0.05 10.86± 3.71

{44.96–56.66} {73.37–110.61} {4.11–7.29} {2.29–4.64} {0.79–0.95} {5.15–14.25}

152 S. ANDRADE ET AL.

Figure 2.Heavy metals concentrations in surface marine sediments from Potter Cove.

by Alam and Sadiq (1993) for sediments collected close to Jubany Station, in whichcadmium residues (19.2± 2.8µg g−1 dry wt.) were evaluated in surface sediments.

Higher contents of zinc, copper, lead and manganese were recorded in the Sta-tion beneath 20 m of water along Transect #1 (T1 20 m) (Figure 1). In this areaof Potter Cove seawater circulation is slower than in any other area resulting inan increase in the deposition of the suspended particulate matter. Moreover, insummer, particulate matter deposition is enhanced by the influx of material fromPotter Stream (B). The highest heavy metal levels within the other transects have

DISTRIBUTION OF HEAVY METALS IN SURFACE SEDIMENTS 153

TABLE III

Correlation matrix showing the coefficients of correlation between dif-ferent pairs of variables measured in bottom sediment. S + C: silt + clay(< 63µm)

Cu Cr Fe Mn Pb Zn S + C

Cu 1.00 –

Cr 0.67a 1.00 –

Fe 0.53b 0.57b 1.00 –

Mn 0.89a 0.65b 0.62b 1.00 –

Pb 0.64b 0.63b 0.29 0.53b 1.00 –

Zn 0.62b 0.69a 0.35 0.56b 0.76a 1.00 –

S + C 0.89a 0.79a 0.47 0.80a 0.70a 0.62b 1.00

a Significant at the 95% level (P< 0.05)b Significant at the 99% level (P< 0.01)

been recorded with the Station at 10 m depth along Transect #2 (T2 10 m) and at20 m depth along Transect #3 (T3 20 m).

Zinc, iron, lead and chromium contents in surface sediments have shown nonsignificant differences when a test for comparison of mean values (ANOVA) – ina 25% significance level – were performed between considered transects. Unlikethis, copper and manganese have exhibited high significant differences (p< 0.01)in the same test.

The heavy metal concentration reported in this paper are similar to those re-ported as natural background values for different regions all over the world byseveral authors (Table IV). However, copper levels measured during this study aresignificantly higher than those reported for other environments. These high copperconcentrations could be explained keeping in mind the nature of the studied sedi-ments which were produced by glacier erosion of volcanic rocks (mainly basalticand andesitic) composed primarily of olivine and pyroxene and by plagioclase andpyroxene, respectively (Fourcade, 1960). Salomons and Förstner (1984) have re-ported that during magmatic differentiation, copper is incorporated – among others– into olivine, pyroxene and plagioclase with mean concentrations of 115 ppm, 120ppm and 62 ppm, respectively.

Even though, in order to understand the geochemical distribution of the studiedmetals in the surface sediments, a study of geochemical partitioning of these ele-ments (meaning the separation of the metal geochemical fractions which could bepresent in sediments: (i) exchangeable adsorbed metals; (ii) oxidisable metal com-plexes; (iii) metals in carbonates; (iv) reducible compounds; (v) residual metals) isnecessary, which will be developed in the near future.

Among other authors, Förstner and Wittmann (1983) have opportunely reportedthe occurrence of an increase of heavy metal concentrations linked with decreasing

154S

.AN

DR

AD

EE

TA

L.

TABLE IV

Natural levels of heavy metals detected in surface sediments from different systems worldwide compared with those measuredin The Potter Cove

Zn Cu Cr Pb Mn Fe Cd

(µg g−1) (µg g−1) (µg g−1) (µg g−1) (mg g−1) (mg g−1) (µg g−1)

Jubany 76.7 48.6 1.1 73.1 0.59 34.47 19.2 Alam and Sadiq,

Station 1993

King 46.5 52.1 2.6 120.9 0.28 23.73 10.1 Alam and Sadiq,

George Is 1993

Beagle 38.68 9.01 – 7.87 – 4.35 ND Aminet al.

Channel 1996

Mediterranean 20 15 – 10–93 0.05–2.5 – – Dassenakiset al.

background 1996

(UNEP, 1993)

Severn 241.6 39.2 – 74 – – – French,

Estuary (UK) 1993

Southern 44.4 9.02 25.5 10.5 – – 0.43 Katz and Kaplan,

California 1981

Potter Cove 52.47 116.79 6.27 3.66 0.96 13.20 ND This study

DISTRIBUTION OF HEAVY METALS IN SURFACE SEDIMENTS 155

Figure 3.Percentage of grain size< 63µm in surface marine sediments.

grain size of the sediment. Heavy metal concentrations as presented in this studywere determined in total sediment. Even though, the percentages of sediment smal-ler than 63µm in size within samples collected from the sampling stations havebeen compared in Figure 3. In a first approach, it could be observed that sedimentsin 5 m of water allong Transect #3 (T3 – 5 m), which have the lowest percentageof sediment size<63µm in size also have the lowest concentrations of chromium,lead, copper, zinc and manganese (Table II; Figure 3). In addition, sediments from20 m of water on Transect #1 (T1 – 20 m), which have showed to contain thehighest percentage of grain size<63 µ have the highest contents of the studiedheavy metals (Table II; Figure 3). Nevertheless, a detailed study considering metalcontents in each grain size sediment fraction must be developed in the near future,which would allow to fully understand not only the metal spatial distribution trendsbut also the relationships between metals in this environment.

When new data on geochemical partitioning of metals and metal concentrationsin grain size sediment fractions were available, will be probably possible to identifyif the occurrence of these elements in Potter Cove is the natural background of thesystem (own of rocks and sediments in the area), or, by contrary, if any relationwith human activities at the Scientific Bases exists.

156 S. ANDRADE ET AL.

TABLE V

Heavy metals concentrations in surface sediments and suspended particulate matter from thesampled streams (expressed inµg g−1 dry weight). SPM: suspended particulate matter

Zn Cu Cr Pb Mn Fe

(µg g−1) (µg g−1) (µg g−1) (µg g−1) (mg g−1) (mg g−1)

Sediments Lagoon 63.39 82.20 2.95 5.79 1.43 30.51

Stream A 78.82 58.66 3.64 6.18 1.29 30.50

Outlet A 58.65 42.72 3.42 2.57 1.04 28.41

Stream B 67.94 59.33 5.02 3.80 1.31 27.77

Outlet B 52.55 48.47 2.60 4.87 0.93 23.46

SPM Stream A 42.76 139.69 2.91 5.73 1.55 69.12

Stream B 35.97 81.10 2.80 2.37 1.72 32.81

3.2. TRACE METALS IN FRESHWATER SEDIMENTS AND SUSPENDED

PARTICULATE MATTER

During the summer months, two streams flow into the Potter Cove: A and B (Fig-ure 1). The second one (Stream B) heads in the glacier, and follows downstreemthrough a narrow canyon developed in volcanic and glacial sediments. Fartherdownstream, it traverses sandy beach sediments before terminating in the coveclose to our first sampling transect (T1) (Figure 1).

The other considered freshwater system (Stream A) is underlain by glacierdrift, and is feed by a lake. It terminates within Potter Cove between the samplingtransections #2 and #3 (T2 and T3) (Figure 1). Stream B exhibits a suspendedparticulate matter concentration ranging between 570 and 900 mg L−1. In contraststream A has showed an average suspended particulate matter content of 38 mgL−1.

Heavy metal concentrations have been determined within samples collectedfrom the sediments of the lake, the channel bed of both streams, and their cor-responding outlets. The studied metals (Zn, Cu, Cr, Pb, Mn and Fe) exhibit con-centrations with different range of values (Table V).

Heavy metal contents in the suspended particulate matter from both streamswere also measured (Table V). Neither sediments nor suspended particulate mattersamples have presented detectable cadmium concentrations, which is consistentwith the results obtained for the surface sediments of the Cove. The obtained resultshave pointed out that the highest copper, zinc, lead and iron concentrations in thesuspended particulate matter were recorded in the Stream A (Figure 1 and Table V).Considering the annual runoff volume, the streams flows can vary as a function ofglacial melting. However, the flow within Stream B is always the highest (Varela,

DISTRIBUTION OF HEAVY METALS IN SURFACE SEDIMENTS 157

1994). Similarly, suspended particulate matter concentrations were always higherin Stream B than in Stream A. Thus, while the metal contents in suspended partic-ulate matter are highest in Stream A, the net heavy metals budget as contributed toPotter Cove through S.P.M. is higher for Stream B than for Stream A.

This fact has also been reflected by the corresponding heavy metal values asdetermined in the first sampling transection (T1) (Table II), which is located nearthe mouth of Stream B.

4. Concluding Comments

The results obtained in this study provides an overview of selected heavy metalconcentrations within surface sediments from the Potter Cove, at ‘25 de Mayo’(King George) Island, Southern Shetland Archipelago, in Antarctica. These are ofgreat importance considering the limited data on metal concentrations within thisarea.

Neither the surface sediments nor suspended particulate matter exhibit detect-able levels of cadmium, presumably because of the low background levels withinthe underlying rocks. This is important given that cadmium is extremely toxic,even at very low concentrations (Friberget al., 1986). In contrast, high copperconcentrations have been recorded in the studied sediments; although this highcopper levels, its origin would presumably be natural, considering both the inputof the whole related freshwater system but also the natural high Cu background ofthe mineral components of Potter Cove sediments. In the particular case of iron itwas observed that concentrations in Potter Cove sediments were lower than thosecorresponding to the freshwater system (Table II; Table V). This fact seems toindicates the occurrence of a complex geochemical process which disable iron tobe incorporated to the corresponding marine system; in this sense, a more detailedstudied will be required to solve this uncertainty.

The quantification of the continental transference of heavy metals to the PotterCove – which only occurs during summer season – has showed that the PotterStream (Stream B) is the most important carrier for the considered system.

Finally, it is important to recognize that the concentrations of manganese, chro-mium, iron, lead and zinc as recorded in the studied surface sediments of the PotterCove constitute the natural background levels of the system.

Acknowledgements

The authors are greatly indebted to Lic. Silvia G. De Marco (Univ. Nac. Mar delPlata), and to Lic. Laura Ferrer (CONICET), who kindly read the manuscript, andoffered many helpful suggestions.

This research was partially financed by Instituto Antártico Argentino-DirecciónNacional del Antártico (IAA-DNA) of Argentina.

158 S. ANDRADE ET AL.

References

Alam I. A. and Sadiq, M.: 1993, Metal concentrations in antarctic sediment samples collected duringthe Trans-Antarctica 1990 Expedition,Mar. Pollut. Bull.26 (9), 523.

Amín, O., Ferrer L. and Marcovecchio, J.: 1996, Heavy metal concentrations in littoral sedimentsfrom the Beagle Channel, Tierra del Fuego, Argentina,Environ.Monit.Assess.41, 219–231.

Barcellos, C. and de Lacerda, L. D.: 1996, Metal and sediment dynamics in the Sepetiba Bay, Rio deJaneiro, in G. Prieto and L. Lesmes (eds),Environmental Geochemistry in Tropical Countries,Ch. A: Geochemistry of Aqueous Environments, pp. 10–16.

Dassenakis, M., Kloukiniotou, M. and Pavlidou, A.: 1996, The influence of long existing pollutionon trace metal levels in a small tidal Mediterranean Bay,Mar. Pollut. Bull.32 (3), 275–282.

Förstner, U. and Wittmann, G. T. W. (eds): 1983,Metal pollution in the aquatic environment,Springer-Verlag, Heidelberg, 486 pp.

Fourcade, N. H.: 1960, Estudio geológico y petrográfico de Caleta Potter, isla 25 de Mayo, IslasShetland del Sur.Publ. N◦ 8, Instituto Antártico Argentino: 115 pp.

French, P. W.: 1993, Post-Industrial pollutant levels in contemporary Severn estuary intertidalsediments, compared to Pre-Industrial levels,Mar. Pollut. Bull.26 (1), 30–35.

Friberg L., Kjellström, T. and Nordberg, G. F.: 1986, in L. Friberg, G. F. Nordberg and V. B. Vouk(eds),Handbook of the Toxicology of Metals, Elsevier Sci. Publ., Amsterdam (The Netherlands),Vol. II, pp. 130–183.

Honda K., Yamamoto, Y. and Tatsukawa, R.: 1987, Distribution of heavy metals in Antarctic marineecosistem,Proc. NIPR Symp. Polar Biol.1, 184–197.

Katz, A. and Kaplan, I. R.: 1981, Heavy metals behavior in coastal sediments of Southern California:a critical review and synthesis,Mar. Chem.10, 261–299.

Klöser, H.: 1994, Descripción básica de Caleta Potter y costas abiertas adyacentes, in Estructura ydinámica de un ecosistema costero antártico,Contrib. N◦ 419, Instituto Antártico Argentino, 5–8.

Marcovecchio, J. E., Moreno, V. J. and Pérez, A.: 1988, Determination of heavy metal concentrationsin biota of Bahía Blanca, Argentina,Sci. Tot. Environ.75, 181–190.

Muir, D., Braune, B., DeMarch, B., Norstrom, R., Wagemann, R., Lockhart, L., Hargrave, B., Bright,D., Addison, R., Payne, J. and Reimer, K.: 1999, Spatial and temporal trends and effects ofcontaminants in the Canadian Artic marine ecosystem: a review,Sci. Tot. Environ.230, 83–144.

Ryan, J. D. and Windom, H. L.: 1988, A geochemical and statistical approach for assessing metalpollution in coastal sediments, in U. Seelinger, L. D. Lacerda and S. R. Patchineelam (eds),Metals in coastal environments of Latin America, Springer-Verlag, Heidelberg, pp. 47–58.

Salomons, W. and Förstner, U. (eds): 1984,Metals in the hidrocycle, Springer-Verlag, Heidelberg,350 pp.

Scagliola M., Poblet, A., Vodopivez, C., Curtosi, A. and Marcovecchio, J.: 1995, Distribuciónde metales pesados en sedimentos marinos superficiales de la Caleta Potter, Isla 25 de Mayo(Shetland del Sur), Antártida.III Jorn. Comunic. Invest. Antárticas, DNA/ IAA, 401–408.

Schropp, S., Lewis, F., Windom, H., Ryan, J., Calder, F. and Burney, L.: 1990, Interpretation ofmetal concentrations in estuarine sediments of Florida using aluminium as a reference element,Estuaries13, 227–235.

UNEP: 1993, Preliminary assessment of the state of pollution of the Mediterranean Sea by zinc,copper and their compounds and proposed measures. UNEP(OCA)/ MED/WG.66/ Inf. 3.

Varela, L.: 1994, Estudio sobre el escurrimiento fluvial de arroyos de deshielo,Contrib. N◦ 419,Instituto Antártico Argentino, 31–35.