Relative sea level curves for the South Shetland Islands and Marguerite Bay, Antarctic Peninsula

ARTICLE IN PRESS

Quaternary Science Reviews 24 (2005) 1203–1216

Relative sea level curves for the South Shetland Islands and Marguerite Bay, Antarctic Peninsula M.J. Bentleya,, D.A. Hodgsonb, J.A. Smitha,b, N.J. Coxa a

Department of Geography, University of Durham, South Road, Durham, DH1 3LE, UK British Antarctic Survey, Natural Environment Research Council, High Cross, Madingley Road, Cambridge CB3 0ET, UK

b

Received 10 October 2003; accepted 17 October 2004

Abstract This paper presents preliminary relative sea level curves for the Marguerite Bay region and for the South Shetland Islands. The Marguerite Bay curve is constrained by both new and previously published 14C dates on penguin remains and shells, and on two isolation basins dating back to 6500 14C yr BP. Extrapolation back to the marine limit yields a minimum deglaciation date for Marguerite Bay of ca 9000 14C yr BP. Analysis of beach clasts suggests that there was a period of increased wave activity, perhaps related to a reduction in summer sea-ice extent, between ca 3500 and ca 2400 14C yr BP. The curve for the South Shetland Islands is derived entirely from published 14C dates from isolation basins and on whalebone, penguin bone and seal bone. The curve shows an initial relative sea level fall, which was interrupted by a period in the mid-Holocene when relative sea level rose to a highstand of between 14.5 and 16 m above mean sea level (amsl), before falling again. r 2004 Elsevier Ltd. All rights reserved.

1. Introduction Relative sea level (RSL) variation records the interplay between eustatic sea level change and local isostatic changes. Global eustatic sea level change is reasonably well known (e.g. Fleming et al., 1998) and so RSL variation is an important source of information on former maximum ice thicknesses, the timing of deglaciation, and subsequent changes in ice cover. Where there are good field data on former RSL, combined studies of RSL curves and modelled glacial-isostatic adjustment (GIA) provide a powerful tool for determining past ice sheet history. In Antarctica there is poor coverage of RSL data, due largely to the lack of coastal ice-free areas where coastal deposits can accumulate, and to a paucity of organic material for radiocarbon dating. A small number of areas where robust RSL curves have been determined include the Scott Coast (Hall and Corresponding author. Tel.: +44 191 334 1859/1800; fax: +44 191 334 1801. E-mail address: [email protected] (M.J. Bentley).

0277-3791/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.quascirev.2004.10.004

Denton, 1999) and the Vestfold and Larsemann Hills, East Antarctica (Zwartz et al., 1998; Verleyen et al., 2004). This paper provides preliminary RSL curves for the northern Antarctic Peninsula (South Shetland Islands) and south-central Antarctic Peninsula (Marguerite Bay) (Fig. 1), and thus helps us to infer the timing of deglaciation, former ice volumes and to provide field constraints for glacial-isostatic modelling of Antarctic deglaciation.

2. Derivation of RSL curves There are two main sources of the elevation and age data necessary for RSL curves. The first, and more common, source is morphological evidence such as raised shorelines, deltas, etc., that have been dated using embedded organic material such as shells. The second source is to use isolation basins to determine past sea level change (Sundelin, 1919). These are former marine basins or inlets that have been isostatically raised above

ARTICLE IN PRESS M.J. Bentley et al. / Quaternary Science Reviews 24 (2005) 1203–1216

e

tla

nd

Isl

an

ds

So

ut

h

S

h

1204

P e n i n s u la

James Ross Island

Weddell

Anta

ic rct

Marguerite Bay

Bellingshausen

McDaniel, 2002), shells or seal hair. In the South Shetlands dates come from whalebone, penguin bone, seal bone, wood and seaweed. Penguin bones provide only minimum limiting dates as penguin rookeries may be located tens of metres above sea level, depending on species and local geography. Conventional 14C dates are reported throughout this paper and are presented as in the original publications. However, instead of using the original authors’ reservoir corrections we have applied systematic reservoir corrections for each type of material, namely penguin bone (11307134 yr, Berkman and Forman, 1996), whalebone and sealbone (14247200 yr, Berkman and Forman, 1996) and marine sediment (13007100 yr, Berkman et al., 1998). Thus, any changes in the reservoir corrections would affect equally all those dates on a particular type of material.

Sea

Sea

3. Site descriptions and beach surveys 3.1. Marguerite Bay

Amundsen Sea

1:20,000,000 0

250

500

750

1000

kilometres

Fig. 1. Location map of Antarctic Peninsula. The two study areas are boxed. Ice shelves are shown by dark shading.

contemporary sea level and become freshwater basins. The timing of the marine–freshwater transition (derived from 14C dating) can be used to determine when the inlet sill or threshold of the basin was raised above sea level. This approach can yield more precise age and elevation data than morphological evidence and has been used to good effect to produce RSL curves for northwest Europe and parts of the Arctic. It is also less subject to errors such as reworking of the organic material in beaches, and because the freshwater sediment (or macrofossils within it) is usually dated directly above the marine–freshwater transition marine reservoir corrections are not usually a problem. The curves presented in this paper have been determined by a combined approach. Dates from isolation basins have been used wherever possible, but where lakes do not exist close to the coast a variety of types of morphological data have been used. In the case of the Marguerite Bay curve the majority of dates are on material from abandoned penguin colonies (Emslie and

The RSL curve for Marguerite Bay uses data from seven sites, spread over an area of approximately 60 km
30 km in the northern part of the bay (Fig. 2a, Table 1). The sites are Horseshoe Island, Pourquoi Pas Island, and Calmette Bay in the northeast of the bay, and Ginger Islands, Lagoon Island, Anchorage Island and Rothera Point in the north and west of the bay. We have combined the data from these sites because they are sufficiently close together (o100 km) that there are unlikely to be major local differences in RSL history. The sites were used to determine the elevation of the marine limit (using Autoset Level), minimum and maximum dates for raised beach formation, the timing of isolation, and variations in the roundness of beach material through time. 3.1.1. Marine limit The highest known occurrences of wave-cut platforms and raised beaches in the region, reported here for the first time, are on Pourquoi Pas Island and Calmette Bay, where we surveyed the marine limit at 41 m above mean sea level (amsl). In Gaul Cove, an inlet on the northeast side of Horseshoe Island, the highest beach level is 22 m amsl. A knoll rising to 31 m close to the main UK base at Rothera Point has rare occurrences of rounded beach material wedged in crevices near its summit, implying that this is a minimum value for the marine limit. There is also a report of a marine limit at 50–55 m elsewhere on Horseshoe Island (C. Hjort, pers. comm. 2003), measured with an altimeter. 3.1.2. Timing of isolation The sediment of ‘‘Skua Lake’’ (671480 4000 S, 0671180 5000 W, 3.5 m amsl) was cored by Wassell and

ARTICLE IN PRESS M.J. Bentley et al. / Quaternary Science Reviews 24 (2005) 1203–1216

1205

59°0'W Kilometers

Ginger Is

4

Mondsee

62°10'S

L

A

N

S

U

62°10'S

E

N

I

Tiefersee

Kitezh Lake

Lagoon I. Anchorage I.

A n t a r c t i c

Horseshoe I. 68°0'S

Calmette Bay

Marguerite

Bay 0

3

2

P

Pourquoi Pas Island

1

25

F I L D E S

67°0'S

P e n i n s u l a

Ad el ai de Is la nd

0

58°55'W

50

Long Lake

Jurasee

Yanou Lake

Kilometers 68°0'W

50°0'W

58°55'W

(c) (a)

0

58°0'W 58˚0'W

60˚0'W 60°0'W

62°0'W 62˚0'W

50

100

Kilometers

S

62°0'S

e

King George Island

a

Fildes Peninsula

S

c

o

t

i a Livingston Island

Nelson Island

Byers Peninsula

S

Smith Island

63°0'S

Deception Island

B

ra

n

ai tr

i tt

d el i sf

(b) Fig. 2. Location of dated RSL sites: (a) raised beach and isolation basin sites in Marguerite Bay, (b) South Shetland Islands, (c) Isolation basin sites on Fildes Peninsula, King George Island, (d) raised beach and isolation basin sites on Byers Peninsula, Livingston Island. See Tables 1 and 2 for site details.

ARTICLE IN PRESS M.J. Bentley et al. / Quaternary Science Reviews 24 (2005) 1203–1216

1206

(d) Fig. 2. (Continued)

Ha˚kansson (1992), who dated both the marine sediment below the transition and the freshwater sediment above it. This transition shows that the glacio-isostatic uplift of the site had progressed such that the marine inlet became isolated from the sea, and thus provides a fix on RSL. According to these dates (Table 1) isolation of ‘‘Skua Lake’’ occurred between 1860 and 1320 14C yr BP (corrected ages, see Table 1). However, no sill depth or lake depth is given for the lake and so we assume here that isolation relates to the 3.5 m lake level rather than a higher (sill) altitude. We cored ‘‘Narrows Lake’’ (671 36.0600 S, 067112.6780 W, 19.4 m amsl) on Pourquoi Pas Island and dated the freshwater sediment above a marine–freshwater transition to 6420750 14C yr BP. 3.1.3. Minimum dates for raised beach formation Dates on penguin remains from various sites have been used to determine the minimum ages for formation of the raised beaches that the colonies occupy (Table 1). Emslie and McDaniel (2002) demonstrated that on the Ginger Islands dates from a pit excavated in an existing colony at 5 m amsl show that the colony has been occupied since at least 2030 14C yr BP. On Lagoon Island, dates from pits excavated at 17 m amsl in abandoned colonies suggest that the island was colonised by 4980 14C yr BP, thus giving a minimum age for deglaciation in the region. A shell date of 4310740 (Beta-178159) and a penguin bone date of 3870740 14 C yr BP (Beta-178160) from an inter-ridge silt and sand deposit on Anchorage Island provide minimum ages for the beach. Taken together with the dates reported by Emslie and McDaniel (2002) there is a

reasonable spread of minimum dates up to ca 17 m altitude (Fig. 3). Above this, however, there are no constraints on the RSL curve. We attempted to rectify this by sampling fur seal hair and a penguin feather buried in beaches at Horseshoe Island and Calmette Bay, but these yielded late Holocene dates that are clearly not close minima for the age of the beaches. So far, we have found no other dateable beach material above 17 m altitude in the area. 3.1.4. Maximum dates for raised beach formation Two penguin bones from Anchorage Island (Table 1) do not come from former colony sites but rather from trenches excavated in coarse (410 cm) beach material and are assumed to have been reworked into the beach matrix by wave action during formation of the beach (e.g. Hall and Denton, 1999). For this reason these two samples are interpreted as maximum constraints on the age of the beaches in which they are buried. Our overall curve would not change markedly if this interpretation is incorrect. 3.1.5. Variations in the roundness of beach material We measured clast roundness for several of the beach ridges making up the flights of raised beaches on Anchorage Island and Rothera Point East Beach. Clasts were sampled from the crests of beach ridges, and a minimum of 150 clasts were measured per ridge. Clast roundness was assessed on a standard ordered scale from very angular to well rounded (Powers, 1953). Results are shown in Fig. 4. Cumulative curves higher on the diagrams in Fig. 4b represent less rounding. Thus

Table 1 Radiocarbon dates on organic material from beaches and isolation basins, Marguerite Bay Altitude amsl (m)c Material dated

Conventional age (yr BP)

Ginger Islands

Lev. 3 (7–12 cm)

Beta-141904

1090740

Lev. 4 (12–17 cm)

Beta-141905

1580750

4507140

5

Lev. 5 (17–22 cm)

Beta-141906

1770740

6407140

5

Lev. 7 (27–32 cm)

Beta-144171

3130740

20007140

5

Lev. 2 (7–15 cm)

Beta-141907

1910740

7807140

9.16

Lev. 3 (15–20 cm)

Beta-141908

4440750

33107140

9.11

Lev. 4 (20–25 cm)

Beta-141909

4390750

32607140

9.06

Lev. 5 (25–30 cm)

Beta-141910

4550750

34207140

9.16

Lev. 6 (30–35 cm)

Beta-141911

4240750

31107140

8.96

Lev. 2 (2–15 cm)

Beta-141912

2980750

18507140

17.27

Lev. 3 (15–25 cm)

Beta-141913

3510770

23807150

17.17

Lev. 4 (25–35 cm)

Beta-141914

5610780

44807160

17.07

Lev. 5 (35–40 cm)

Beta-141915

5920750

47907140

17.02

Lev. 5 (35–40 cm)

Beta-141916

6080750

49507140

17.02

Lev. 2 (8–15 cm)

Beta-141917

2490750

13607140

16.06

Lev. 4 (25–32 cm)

Beta-144170

5550740

44207140

15.89

Lagoon Island (NE) site 3

Lagoon Island (NE) site 4

modern

5

Reference

Comment

Penguin bone (collagen) Penguin bone (collagen) Penguin bone (collagen) Penguin bone (collagen)

Emslie and McDaniel (2002) Emslie and McDaniel (2002) Emslie and McDaniel (2002) Emslie and McDaniel (2002)

Minimum age for beach

Penguin bone (collagen) Penguin bone (collagen) Penguin bone (collagen) Penguin bone (collagen) Penguin bone (collagen) Penguin bone (collagen) Penguin bone (collagen) Penguin bone (collagen) Penguin bone (collagen) Penguin bone (collagen) Penguin bone (collagen) Eggshell carbonate

Emslie and McDaniel (2002) Emslie and McDaniel (2002) Emslie and McDaniel (2002) Emslie and McDaniel (2002) Emslie and McDaniel (2002) Emslie and McDaniel (2002) Emslie and McDaniel (2002) Emslie and McDaniel (2002) Emslie and McDaniel (2002) Emslie and McDaniel (2002) Emslie and McDaniel (2002) Emslie and McDaniel (2002)

Minimum age for beach

Minimum age for beach Minimum age for beach Minimum age for beach

Minimum age for beach Minimum age for beach Minimum age for beach Minimum age for beach Minimum age for beach Minimum age for beach Minimum age for beach Minimum age for beach Minimum age for beach Minimum age for beach Minimum age for beach

ARTICLE IN PRESS

Laboratory no.

Lagoon Island (NE) site 2

C

Corrected 14C age (yr BP)b

Stratigraphic positiona

M.J. Bentley et al. / Quaternary Science Reviews 24 (2005) 1203–1216

14

Site

1207

1208

Table 1 (continued ) Corrected 14C age (yr BP)b

Altitude amsl (m)c Material dated

Reference

Comment

2310740

11807140

21.36

This study

Minimum age for beach

Lu-3143

1440750

1320770d

22–27 cm depth

Lu-3099

3160760

Pourquoi Pas Island (P4)

34 cm depth

Beta-178164

Anchorage Island (West beach)

ca 15 cm depth

Site

Stratigraphic positiona

Laboratory no.

Conventional age (yr BP)

Horseshoe Island (H8)

35 cm depth

Beta-178162

Horseshoe Island (‘‘Skua Lake’’)

9–13 cm depth

C

Wassell and Minimum date of isolation Ha˚kansson (1992)

18607120

Penguin feather quill Lake height 3.5 m, Freshwater sill height not sediment given Marine sediment

2970740

16707110

35.47

Fur seal hair

This study

Minimum age for beach

AA-44096

4474750

33407140

4.6

This study

ca 15 cm depth

CAMS-86753

4480740

33507140

6.1

15 cm depth

Beta-178159

4310740

30107110

8.25

This study

Bone in inter-ridge swale, maximum age Bone in inter-ridge swale, maximum age Minimum age for beach

20 cm depth

Beta-178160

3870740

27407140

8.25

Penguin bone (collagen) Penguin bone (collagen) Shell (Nacella concina) Penguin bone (collagen)

This study

Minimum age for beach

60 cm depth

Beta-180801

6420750

6420750

Freshwater sediment

This study

Minimum date of isolation

This study

We denote unofficial place names by quotation marks and give their co-ordinates following the first occurrence in the text. All official names appear in the SCAR Composite Gazetteer of Antarctica (http://www3.pnra.it/SCAR_GAZE). a ‘Lev’. refers to excavation levels (depth) within pits dug in penguin colonies. b Radiocarbon dates corrected for marine reservoir effect (see text for discussion). All dates and errors are rounded to the nearest 10 years. c Surveyed altitude corrected to stratigraphic depth of sample. Heights given an error of 0.2 m to account for surveying inaccuracies, and errors in calculating tidal corrections. d Corrected for date at lake bed surface of 120745 (Lu-3100) (Wassell and Ha˚kansson, 1992).

ARTICLE IN PRESS

19.4

Wassell and Maximum date of isolation Ha˚kansson (1992) M.J. Bentley et al. / Quaternary Science Reviews 24 (2005) 1203–1216

Pourquoi Pas Island (‘‘Narrows lake’’)

14

ARTICLE IN PRESS M.J. Bentley et al. / Quaternary Science Reviews 24 (2005) 1203–1216

1209

Fig. 3. RSL curve for Marguerite Bay. Dates are plotted with 1-sigma error bars for 14C ages and 70.2 m for altitudes. The thick dashed line shows an approximate fit to the data. The y-intercept of the curve is at 2 m amsl, the altitude of the modern beach in this area. Grey lines show the zone of increased clast roundness. (Symbols: isolation contacts ¼ shaded squares; penguin bones ¼ shaded diamonds; shell ¼ solid triangle; fur seal hair ¼ open circle; penguin feather ¼ solid circle.)

at Anchorage Island, west beach, sites 1 and 4 are less rounded and sites 2 and 3 are more rounded, with site 5 in between. Similarly at Rothera Point, sites 1 and 2 are least rounded and roundness increases through sites 5, 3 and 4. The data in Fig. 4 show that there are changes in the roundness of beach material with altitude. Between 4.5 and 8 m amsl clasts are substantially more rounded than the material in the beaches above and below this. This change appears to be independent of beach aspect because the beaches at Rothera Point (E-facing) and Anchorage Island (W-facing) both yield the same result. 3.2. South Shetland Islands The RSL curve for the South Shetland Islands (Figs. 2 and 5) has been developed entirely from previously published radiocarbon dates. A number of workers have published analyses of cores from lakes on the South Shetland Islands in order to infer palaeo-environmental change, but this paper is the first systematic attempt to use the published dates on these lakes as isolation basins to yield a relative sea level curve. The isolation data have been combined with morphological data, primarily whalebone and penguin bone dates. The RSL data have been compiled from three islands: King George, Livingston and Nelson Islands (Fig. 2b). The most important sites are a series of lakes on Fildes Peninsula (Fig. 2c), the largest ice-free area of King

George Island. The lakes range in altitude from ca 14.5 to 450 m and have been repeatedly cored as part of attempts to determine the past climate history of this part of the Antarctic Peninsula (Ma¨usbacher et al., 1989; Martinez-Macchiavello et al., 1996; Yang and Harwood, 1997; del Valle et al., 2002) and because they are close to several research bases. Byers Peninsula (Fig. 2d) on Livingston Island also has a number of beaches and lakes that have been previously studied. We use the data from the different sites to compile a single RSL curve but acknowledge that there may be some significant local differences in RSL history between locally glaciated islands lying close to the margin of a formerly expanded ice sheet. In most of the lake basins the analysis of cores shows continuous freshwater sedimentation from the basal parts of the core to recent (e.g. Midge Lake). However, in cores from three lakes on the Fildes Peninsula there are contacts between marine and freshwater sediments. The first of these is Lake Kitezh (termed ‘‘Kiteschsee’’ by Ma¨usbacher et al., 1989) (621110 3000 S, 581580 0000 W, 16 m amsl) where a marine–freshwater contact occurs at 185 cm depth (Ma¨usbacher et al., 1989). A date on freshwater moss fragments from 184 to 181 cm yields a date of 61807150 14C yr BP (HD 11163-10998). A date on the marine sediments below at ca 189 cm depth yields a date of 69507195 14C yr BP (HD 11162-10997). The moss date shows that the site emerged above contemporary sea level at ca 6180 14C yr BP.

ARTICLE IN PRESS M.J. Bentley et al. / Quaternary Science Reviews 24 (2005) 1203–1216

1210

Anchorage Island (W)

Site 5

Site 3

10.00 9.00 Site 2

Height (amsl)

8.00 7.00 6.00

Site 4

5.00

Site 1

4.00 3.00 2.00 1.00 0.00 0.00

10.00

20.00

30.00 40.00 50.00 Horizontal distance (m) Rothera East Beach

60.00

70.00

80.00

Site 5

Site 4 12.00 Site 3

Height (amsl)

10.00 8.00

Site 1

Site 2

6.00 4.00 2.00 0.00 0

(a)

20

40

60

80

100

120

140

160

Horizontal distance (m)

(b)

Fig. 4. Changes in clast roundness with altitude. (a) Beach profiles for Anchorage Island, west beach and Rothera Point East beach. (b) Cumulative frequency plots of clast roundness. Cumulative curves higher on the diagrams represent less rounding. Cumulative percentages are calculated to the midpoint of each class to reduce the problem of convergence as all cumulative curves approach 100%. The inverse normal scale helps to straighten the curves and thus to show differences between sites more clearly.

The second core is from Long Lake (termed ‘‘Langer See’’ by Ma¨usbacher et al., 1989) (621120 2000 S, 581580 0000 W, 16 m amsl) where a marine–freshwater transition occurs between 140 and 82 cm depth (Martinez-Macchiavello et al., 1996). Freshwater sediment from 68 to 78 cm yields a date of 24607260 14C yr BP, implying that RSL fell below the 16 m altitude sometime prior to 2460 14C yr BP. The third site is Yanou Lake (621130 1800 S, 581570 1200 W, 14.5 m amsl) where a study of the diatom stratigraphy in a 7.7-m-long core shows a complex sequence of freshwater–marine and marine–freshwater transitions (Yang and Harwood, 1997). The lower part of the core comprises freshwater sediments that are overlain by marine sediments at 6.16–6.07 m depth. These are in turn overlain by a thin (4 cm) zone of freshwater sediments, which then gives way to 85 cm of marine sediments. The upper parts of the core are exclusively dominated by freshwater diatoms. Thus the site records a complex history of RSL change with two episodes of marine submergence (freshwater– marine contacts) superimposed on the overall sequence of emergence (marine–freshwater contacts).

None of these contacts were dated by Yang and Harwood (1997) but the core nevertheless provides a useful constraint on the RSL curve (Fig. 5) as the sequence of transitions at 14.5 m altitude must be satisfied by the curve. All the other lake basins studied here are above 16 m altitude and show continuous freshwater sedimentation from the oldest dated horizons through to the present. Thus, they can all be used as minimum constraints on former sea level. Most biological and chemical assemblages in the freshwater sequences show minor climate or catchment-related fluctuations but one mid-Holocene change in a freshwater core from a lake on the Fildes Peninsula may reflect changing sea level. In a core from ‘‘Tiefersee’’ (621110 1000 S, 581540 3000 W, 17.5 m amsl) Schmidt et al. (1990) reported a change in sedimentary characteristics and an increase in halophilic (salt-loving) diatoms between 4700 and 3200 14C yr BP. One of their explanations for this change was an increased influence of sea spray during this interval. Curl (1980), Sugden and John (1973), and Hansom (1979) all collected whalebone or seal bone from gravel beaches at various elevations around the South Shetland

ARTICLE IN PRESS M.J. Bentley et al. / Quaternary Science Reviews 24 (2005) 1203–1216

1211

Fig. 5. RSL curve for South Shetland Islands. (a) Data plotted with 1-sigma error bars for 14C ages and 70.2 m for altitudes. Dotted line shows altitude of Yanou Lake (b) Data with overlaid interpretation of marine, freshwater and salt-spray-affected zones of sediment cores. The thick dashed line shows an approximate RSL curve fit to the data. The y-intercept of the RSL curve is at 2 m amsl, the altitude of the modern beach in this area. (Symbols: isolation contacts ¼ shaded squares; minima for isolation (freshwater sediment) ¼ shaded circles; penguin bones ¼ shaded diamonds; whalebone (maxima for beach) ¼ shaded triangles.)

Islands. The interpretation of the whalebone ages depends on the reservoir correction used: Curl (1980) suggested the correction was likely to be in the range 500–600 years, but Berkman and Forman (1996) suggest 14247200 from a larger data set, and we adopt that value here. This means that, once corrected, most whalebone samples yield ‘‘modern’’ ages and provide little constraint on the RSL curve. However, if they were corrected by the smaller values of Curl (1980) then some of the whalebones would provide constraints on the RSL curve during the

last ca 1000 yr. Most of the whalebone samples were buried to some depth (several tens of centimetres) in gravel beaches. The bones are large so it seems likely that wave activity would be required to bury the bones in the gravel. For this reason the bones are likely to be older than the beaches in which they are embedded and so their ages are interpreted here to be maximum ages. On Fildes Peninsula, King George Island, Barsch and Ma¨usbacher (1986) dated penguin bones (54307140; 55207160 14C yr BP) from 18 m altitude. del Valle et al.

ARTICLE IN PRESS 1212

M.J. Bentley et al. / Quaternary Science Reviews 24 (2005) 1203–1216

(2002) recently reported dates on penguin bones (5750740; 5840740 14C yr BP) from open sections at 14.7–16.7 m altitude on the Potter Peninsula, King George Island. As with the dates reported from Marguerite Bay by Emslie and McDaniel (2002) all of these penguin bone dates are interpreted as minima for RSL. 4. Discussion 4.1. Marguerite Bay The RSL curve for Marguerite Bay in Fig. 3 has been determined from the radiocarbon dates in Table 1 and is based on an assumption of exponential decay from the 41 m marine limit on Pourquoi Pas Island and Calmette Bay. The curve is constrained by minimum dates, two maxima, and the dates for isolation of ‘‘Skua Lake’’ and ‘‘Narrows Lake’’, back to ca 6500 14C yr BP. However, neither the maximum limit of the curve nor its shape older than 6500 14C yr BP is well constrained. Here, we have assumed an exponential shape prior to 6500 14C yr BP because of evidence that the former ice margin stood well offshore from the Peninsula (Bentley and Anderson, 1998). As the maximum age (i.e. lower limit) of the RSL curve is not tightly constrained, the curve generally gives a maximum estimate of the rates of uplift for this region. Estimated rates of uplift are 9.5 mm/yr (shortly after deglaciation) and ca 1 mm/yr at the present. The latter estimate is highly sensitive to the shape of the curve but could be tested with GPS or satellite measurements of present-day crustal uplift. However, it may lie below the easily measurable range of uplift rates. Modelled values of uplift from a series of hypothesized glacial (un)loading histories (Ivins et al., 2000) range from 7.5 to 12.5 mm/yr (progressive unloading from 14 to 5.5 ka) to 2.5 to 3.0 mm/yr (oscillatory loading in Holocene) although the precise values are sensitive to parameters used in the Earth model such as mantle viscosity and lithospheric thickness. Extrapolation of the curve from 6500 14C yr BP back to the regional marine limit at 41 m amsl (assuming exponential decay of uplift) yields an estimate of ca 9000 14 C yr BP as a minimum for deglaciation of inner Marguerite Bay. Using Hjort’s (pers. comm. 2003) marine limit of 50–55 m would extend the date for deglaciation back to ca 10,000 14C yr BP. In comparison, Pudsey et al. (1994) suggested that the middle and outer continental shelf north of Marguerite Bay were deglaciated prior to ca 11,000 14C yr BP, but with the inner shelf perhaps deglaciating later. Sugden and Clapperton (1980) suggested that George VI Sound, at the southern end of Marguerite Bay, was free of ice by ca 6000 14C yr BP, with subsequent formation of an ice shelf. Beaches on Rothera Point and Anchorage Island show variations in roundness of beach material with the

youngest and oldest beaches made up of relatively angular material whilst intermediate beaches (ca 4.5–8 m altitude) are made of more rounded clasts (Fig. 4). This suggests that there was a period of greater wave activity during the formation of the intermediate beaches. The most likely reason for this is reduced summer sea-ice cover. The timing of this change can be estimated from the RSL curve as between ca 3500 and ca 2400 14C yr BP. This correlates with other evidence for a warm period at approximately 4000–2000 14C yr BP in the mid- to late Holocene (Bjo¨rck et al., 1996; see Jones et al., 2000 for a review). 4.2. South Shetlands The RSL curve for the South Shetlands (Fig. 5b) is more complex than the simple exponential uplift curve for Marguerite Bay. There are four main phases shown by the curve: (i) early RSL rise; (ii) early to midHolocene RSL fall, interrupted by (iii) mid-Holocene rise and highstand in relative sea level; and finally (iv) RSL fall during the late Holocene. The first phase of early RSL rise is very poorly constrained and is entirely undated. The only evidence for this rise is the freshwater–marine contact near the base of the core from Yanou Lake at 14.5 m altitude. As it is undated it may substantially pre-date the Holocene, and we cannot exclude the possibility that it may even date to a previous interglacial period. The second phase of early to mid-Holocene RSL fall is constrained by the marine–freshwater isolation contact at 6180 14C yr BP in ‘‘Kiteschsee’’ (16 m) and the undated transition in Yanou Lake, and is partly constrained by the freshwater sequences in nearby Jurasee (47 m), and ‘‘Mondsee’’ (621100 4000 S, 581550 5000 W, ca 50 m) and Midge Lake (ca 70 m) on Livingston Island (Table 2). The data provide some key constraints on both the timing and the altitude of the third phase of RSL change (ca 5800–ca 3000 14C yr BP) during the mid-Holocene highstand. The changes in marine and freshwater sedimentation in Yanou Lake show that after RSL fall through 16 m (‘‘Kiteschsee’’ isolation) and 14.5 m (Yanou Lake) RSL rose back above 14.5 m; however, the continuous freshwater sedimentation in the upper parts of the ‘‘Kiteschsee’’ (16 m) core and in the ‘‘Tiefersee’’ (17.5 m) core imply that the RSL rise did not exceed 16 m (Fig. 5b). Therefore, there is a narrow constraint for the maximum RSL reached during the mid-Holocene highstand of 14.5–16 m amsl. This RSL rise took place sometime after 6180 14C yr BP and the date from Long Lake and the morphological data suggest that RSL was falling again by ca 2500 14C yr BP. Interestingly, Schmidt et al. (1990) reported evidence of a period of enhanced sea spray between 4700 and 3200 14 C yr BP in the ‘‘Tiefersee’’ core. It seems likely that the highstand resulted in an increased proximity of the coast

Island

Stratigraphic position

Laboratory code

Byers Peninsula buried in gravel Birm-50a

Corrected 14C age (yr BP)b

10567130

modern

Alt. amsl Material dated (m) 3.00

Byers Peninsula buried 30 cm

SRR-1086a

2823740

14007200

10

Byers Peninsula buried 40 cm

SRR-1087a

3121735

17007200

10

South Beaches South Beaches

— Modern beach

I-7869 DIC-372

1025780 840775

modern modern

4.50 1.80

South Beaches

—

I-7870

2530785

11107220

8 2

Start Point

King George Island

Conventional C age (yr BP)

14

Reference

Comment

Whalebone (collagen) Whalebone (collagen) Whalebone (collagen) Whalebone Whalebone (collagen) Whalebone (collagen) Whalebone

Sugden and John (1973) Hansom (1979)

Maximum age for beach

Hansom (1979)

Maximum age for beach

Curl (1980) Curl (1980)

Maximum age for beach Maximum age for beach

Curl (1980)

Maximum age for beach

Curl (1980)

Maximum age for beach

Curl (1980) Barsch and Ma¨usbacher (1986) Barsch and Ma¨usbacher (1986) Sugden and John (1973) Sugden and John (1973) Curl (1980)

Maximum age for beach Minimum age for beach

49057100

34807220

970750 6560755

modern 54307140

1.80 18

Sealbone Penguin bone

Fildes Peninsula —

HD9425-9100a

6650790

55207160

18

Penguin bone

Barton Peninsula Not reported

—

Birm-224a

13907140

modern

6.50

Modern beach

Birm-496a

674766

modern

1.00

Keller Peninsula Modern beach

DIC-367a

1000745

modern

1.80

Maxwell Bay

—

DIC-368a

12007110

modern

2.75

Maxwell Bay

—

DIC-369a

1210755

modern

2.75

Maxwell Bay

—

DIC-371a

13607165

modern

6.00

Maxwell Bay

—

DIC-373a

1440755

207210

6.00

Whalebone (collagen) Whalebone (collagen) Whalebone (collagen) Whalebone (collagen) Whalebone (collagen) Whalebone (collagen)

Minimum age for beach Maximum age for beach Maximum age for beach Maximum age for beach

Curl (1980)

Maximum age for beach

Curl (1980)

Maximum age for beach

Curl (1980)

Maximum age for beach

Curl (1980)

Maximum age for beach

1213

Exposed on rear I-7872 of storm beach Barnard Point Modern beach DIC-370 Fildes Peninsula — HD8426-9106a

Maximum age for beach

ARTICLE IN PRESS

Livingston Island

Site

M.J. Bentley et al. / Quaternary Science Reviews 24 (2005) 1203–1216

Table 2 Radiocarbon dates on organic material from beaches and isolation basins, South Shetland Islands. (Locations of sites are shown in Figs. 2c and d.)

1214

Table 2 (continued ) Island

Nelson Island

Site

Stratigraphic position

Alt. amsl Material dated (m)

1223781

1223780

5.50

Whalebone (collagen) Seaweed

14307470

14307470

5.50

Seaweed

802743

802743

5.50

Tree trunk (Nothofagus antarctica) Freshwater sediment Gyttja rich in moss

Buried in gravel Birm-17a

—

—

Birm-14a

1.81–1.84 m

HD11163-10998 61807150

61807150

16

0.68–0.72 m

Gd 4641

24607260

24607260

16

Kitezh Lake

1.55–1.62 m

Lu 3679

34607140

34607140

16

Not given

Tiefersee

1.55 m

53807165

53807165

18

TieferSee Mondsee Jurasee

1.65 m 3.20 m 3.60 m

HD 1116111420 Not given Not given Not given

53807165 72007250 87007300

53807165 72007250 87007300

18 50 47

Freshwater sediment Not given Not given Freshwater sediment

—

—

—

15

Sugden and John (1973) Sugden and John (1973) Sugden and John (1973)

Maximum age for beach

‘‘Pingfo’’

open section

Not givena

5750740

46207140

16

Penguin bone

Ma¨usbacher et al. (1989) MartinezMacchiavello et al. (1996) MartinezMacchiavello et al. (1996) Ma¨usbacher et al. (1989) Schmidt et al. (1990) Schmidt et al. (1990) Ma¨usbacher et al. (1989) Yang and Harwood (1997) del Valle et al. (2002)

‘‘Pingfo’’ Midge Lake

open section 2m

Not givena Ua-1220

5840740 37357250

47107140 37357250

15 70

Penguin bone Moss

del Valle et al. (2002) Bjo¨rck et al. (1991)

Precise location of dates, which we are unable to determine. Radiocarbon dates corrected for marine reservoir effect (see text for discussion). All dates and errors are rounded to the nearest 10 years.

b

Comment

Maximum age for beach Maximum age for beach

Date of isolation Minimum age of isolation

Minimum age of isolation

Minimum age of isolation

Minimum age of isolation Complex F–M–F–M–F sequence Minimum age for beach Minimum age for beach Minimum age of isolation

ARTICLE IN PRESS

Marian Cove

Reference

M.J. Bentley et al. / Quaternary Science Reviews 24 (2005) 1203–1216

a

Corrected 14C age (yr BP)b

—

Yanou Lake

Livingston Island— Byers Peninsula

Birm-16a

Conventional C age (yr BP)

14

Marian Cove

King George I. (Fildes Kiteschsee Peninsula) Long Lake

King George I.— Potter Peninsula

Laboratory code

ARTICLE IN PRESS M.J. Bentley et al. / Quaternary Science Reviews 24 (2005) 1203–1216

to ‘‘Tiefersee’’ and thus greater sea-spray deposition in the lake basin. No comparable change was reported from the (lower altitude) ‘‘Kiteschsee’’, but this site is much further from the coast, so it may have been located too far from the mid-Holocene coastline to have shown any significant effect of the rise. The penguin bones reported in open section on Potter Peninsula, King George Island (del Valle et al., 2002) were deposited ca 5000 14C yr BP at a similar altitude to the maximum height of the transgression and so may reflect onshore reworking (roll-over) of beach material during the rise in RSL. A whalebone dated to 4255 14C yr BP occurs at a low altitude, suggesting that it has been reworked to a lower altitude beach during subsequent sea level fall. This was also suggested by Curl (1980) on the grounds that the bones were scattered and thus appeared to have been reworked. The fourth phase showing RSL fall is constrained by the present sea level and by dates on bones in raised beaches. Whalebones, penguin bone and a seal bone imply that during the late Holocene sea level fell below 10 m at approximately 2500 14C yr BP (Fig. 5b). The curve cannot be used to provide a confident estimate of the timing of deglaciation because of the poor constraints on RSL behaviour in the earliest parts of the curve (Fig. 5). No dated isolation contacts older than ca 6000 14C yr BP have been reported, although continuous freshwater sedimentation in Jurasee (47 m amsl) is reported since 8700 14C yr BP by Ma¨usbacher et al. (1989). This remains the best estimate of onshore glacier retreat in the South Shetlands. However, the distribution of lakes relative to the ice cap margin on King George Island means that there may be a complex pattern of dates for deglaciation. The curve suggests that modern rates of uplift in the South Shetland Islands are ca 3 mm/yr. For comparison, Ivins et al. (2000) gave modelled values for the South Shetland Islands ranging from ca 1 to 2 mm/yr (progressive unloading from 14 to 5.5 ka) to 0 to 2.5 mm/yr (oscillatory loading in Holocene), although the precise values are sensitive to parameters such as mantle viscosity and lithospheric thickness used in the earth model. The role of tectonics cannot be ignored on King George Island and Livingston Island because they are close to active volcanism (e.g. Deception Island) and there are reports of neotectonic features (Palla`s et al., 1995). However, Palla`s et al. (1997) suggested that only a small proportion of uplift in the South Shetland Islands has a tectonic origin and they suggested a maximum tectonic uplift rate of 0.4 m/ka. Apart from this relatively small effect they argued that the majority of RSL change in the islands was due to glacio-hydroisostatic processes (Palla`s et al., 1997). The RSL curve in Fig. 5b is not corrected for the tectonic effect, but the effect of such an uplift, if constant, would be to lower

1215

the curve by 0.4 m/ka, such that, for example, instead of being at ca 14.5 m at 5000 14C yr BP the curve would pass through 12.5 m. The shape of the curve would not change. Palla`s et al. (1995) also demonstrated that raised beaches on Hurd Peninsula, Livingston Island (Fig. 2) are not displaced along known fault outcrops and so differential fault movement is not thought to have been a problem in this area. For this reason we believe that the shape of the RSL curve is a reliable indicator of the pattern of RSL change in the area, although regional uplift may mean that the curve in Fig. 5b provides an overestimate of the total amount of post-glacial isostatic uplift.

5. Conclusions and implications In this paper, we have presented two preliminary RSL curves for the Antarctic Peninsula. The first RSL curve for the Marguerite Bay region is reasonably well constrained by dates on penguin remains and two isolation basins from ca 6500 14C yr BP to the present. The curve indicates exponential RSL fall since deglaciation. Extrapolation of the curve to the regional marine limit suggests that deglaciation of the inner part of Marguerite Bay occurred sometime prior to ca 9000 14 C yr BP. Estimated rates of uplift are 9.5 mm/yr (shortly after deglaciation), and 1 mm/yr (present). Marguerite Bay beaches exhibit clast roundness variations with altitude that suggest there was a period of enhanced wave activity, perhaps related to a reduction in summer sea-ice extent, between ca 3500 and ca 2400 14 C yr BP. The RSL curve for the South Shetland Islands shows a more complex behaviour with a prominent midHolocene highstand. The curve shows that initial RSL fall was interrupted by RSL rise sometime after 6180 14 C yr BP, and reached between 14.5 and 16 m amsl. The subsequent fall is not well constrained but was probably well underway by ca 2400 14C yr BP. The date of deglaciation is not well constrained by the curve and a period of early RSL rise remains undated. We are now using coupled glaciological and glacio-isostatic models to investigate the ice loading history required to produce both the Marguerite Bay and South Shetlands RSL curves. Although some modelling exercises have been attempted previously (Ivins et al., 2000; Palla`s et al., 1997), these were hampered by the lack of field RSL data with which to constrain the models. In particular, the use of more than one curve, as provided here, can provide powerful constraints for models.

Acknowledgements Fieldwork was undertaken during projects funded by the British Antarctic Survey (BAS) SAGES and

ARTICLE IN PRESS 1216

M.J. Bentley et al. / Quaternary Science Reviews 24 (2005) 1203–1216

NERC-AFI programmes. Thanks are due to BAS for logistic support, and Steve Emslie, O´lafur Ingo´lfsson, Jenny McDaniel, Emmanuel Le Meur, Richard Burt, and Chandrika Nath for assistance in the various field campaigns. Glenn Milne provided a helpful discussion on the South Shetland Islands RSL curve. Peter Fretwell drafted Figs. 1 and 2. Kurt Lambeck and Christian Hjort provided helpful reviews. References Barsch, D., Ma¨usbacher, R., 1986. Beitra¨ge zur Vergletscherungsgeschichte und zur Relieftwicklung der Su¨dshetland Inseln. Zeitschrift fu¨r Geomorphologie 61, 25–37. Bentley, M.J., Anderson, J.B., 1998. Glacial and marine geological evidence for the ice sheet configuration in the Weddell SeaAntarctic Peninsula region during the last glacial maximum. Antarctic Science 10, 307–323. Berkman, P.A., Forman, S.L., 1996. Pre-bomb radiocarbon and the reservoir correction for calcareous marine species in the Southern Ocean. Geophysical Research Letters 23, 363–366. Berkman, P.A., Andrews, J.T., Bjo¨rck, S., Colhoun, E.A., Emslie, S.D., Goodwin, I.D., Hall, B.L., Hart, C.P., Hirakawa, K., Igarashi, A., Ingo´lfsson, O., Lo´pez-Martı´ nez, J., Lyons, W.B., Mabin, M.C.G., Quilty, P.G., Taviani, M., Yoshida, Y., 1998. Circum-Antarctic coastal environmental shifts during the late Quaternary reflected by emerged marine deposits. Antarctic Science 10, 345–362. Bjo¨rck, S., Ha˚kansson, H., Zale, R., Karle´n, W., Jo¨nsson, B., 1991. A late Holocene lake sediment sequence from Livingston Island, South Shetland Islands. Antarctic Science 3, 61–72. Bjo¨rck, S., Olsson, S., Ellis-Evans, J.C., Hakansson, H., Humlum, O., de Lirio, J., 1996. Late Holocene palaeoclimatic records from lake sediments on James Ross Island. Palaeogeography, Palaeoclimatology, Palaeoecology 121, 195–220. Curl, J., 1980. A glacial history of the South Shetland Islands, Antarctica. Institute of Polar Studies Report 63, Ohio State University, Columbus, OH. del Valle, R.A., Montalti, D., Inbar, M., 2002. Mid-Holocene macrofossil-bearing raised marine beaches at Potter Peninsula, King George Island, South Shetland Islands. Antarctic Science 14, 263–269. Emslie, S.D., McDaniel, J.D., 2002. Ade´lie penguin diet and climate change during the middle to late Holocene in northern Marguerite Bay, Antarctic Peninsula. Polar Biology 25, 222–229. Fleming, K., Johnston, P., Zwartz, D., Yokoyama, Y., Lambeck, K., Chappell, J., 1998. Refining the eustatic sea-level curve since the last glacial maximum using far- and intermediate-field sites. Earth and Planetary Science Letters 163, 327–342. Hall, B.L., Denton, G.H., 1999. New relative sea-level curves for the southern Scott Coast, Antarctica: evidence for Holocene deglaciation of the western Ross Sea. Journal of Quaternary Science 14, 641–650. Hansom, J., 1979. Radiocarbon dating of a raised beach at 10 m in the South Shetlands. British Antarctic Survey Bulletin 49, 287–288.

Ivins, E.R., Raymond, C.A., James, T.S., 2000. The influence of 5000 year-old and younger glacial mass variability on present-day crustal rebound in the Antarctic Peninsula. Earth Planets and Space 52, 1023–1029. Jones, V.J., Hodgson, D.A., Chepstow-Lusty, A., 2000. Palaeolimnological evidence for marked Holocene environmental changes on Signy Island, Antarctica. The Holocene 10, 43–60. Martinez-Macchiavello, J.C., Tatur, A., Servant-Vildary, S., del Valle, R., 1996. Holocene environmental change in a marine–estuarine–lacustrine sediment sequence, King George Island, South Shetland Islands. Antarctic Science 8, 313–322. Ma¨usbacher, R., Mu¨ller, J., Schmidt, R., 1989. Evolution of postglacial sedimentation in Antarctic lakes (King George Island). Zeitschrift fu¨r Geomorphologie 33, 219–234. Palla`s, R., Vilaplana, J.M., Sa`bat, F., 1995. Geomorphological and neotectonic features of Hurd Peninsula, Livingston Island, South Shetland Islands. Antarctic Science 7, 395–406. Palla`s, R., James, T.S., Sa`bat, F., Vilaplana, J.M., Grant, D.R., 1997. Holocene uplift in the South Shetland Islands: evaluation of tectonics and glacio-isostasy. In: The Antarctic Region: Geological Evolution and Processes. Terra Antarctica Publication, Siena, pp. 861–868. Powers, M., 1953. A new roundness scale for sedimentary particles. Journal of Sedimentary Petrology 23, 117–119. Pudsey, C.J., Barker, P.F., Larter, R.D., 1994. Ice-sheet retreat from the Antarctic Peninsula shelf. Continental Shelf Research 14, 1647–1675. Schmidt, R., Ma¨usbacher, R., Muller, J., 1990. Holocene diatom flora and stratigraphy from sediment cores of two Antarctic Lakes (King George Island). Journal of Paleolimnology 3, 55–90. Sugden, D.E., John, B., 1973. The ages of glacier fluctuations in the South Shetland Islands, Antarctica. In: Palaeoecology of Africa, the Surrounding Islands and Antarctica. Balkema, Cape Town, pp. 141–159. Sugden, D.E., Clapperton, C.M., 1980. West Antarctic ice sheet fluctuations in the Antarctic Peninsula area. Nature 286, 378–381. Sundelin, U., 1919. Uber die spa¨tquarta¨re Geschichte der kustengegenden O¨stergo¨tlands und Sma˚lands. Bulletin Geological Institut Upsala 16, 195–242. Verleyen, E., Hodgson, D.A., Sabbe, K., Vanhoutte, K., Vyverman, W., 2004. Coastal oceanographic conditions in the Prydz Bay region (East Antarctica) during the Holocene recorded in an isolation basin. The Holocene 14, 246–257. Wassell, A., Ha˚kansson, H., 1992. Diatom stratigraphy in a lake on Horseshoe Island, Antarctica. Diatom Research 7, 157–194. Yang, S., Harwood, D.M., 1997. Late Quaternary environmental fluctuations based on diatoms from Yanou Lake, King George Island, Fildes Peninsula, Antarctica. In: The Antarctic Region: Geological Evolution and Processes. Terra Antarctica Publication, Siena, pp. 853–859. Zwartz, D., Bird, M., Stone, J., Lambeck, K., 1998. Holocene sea-level change and ice-sheet history in the Vestfold Hills, East Antarctica. Earth and Planetary Science Letters 155, 131–145.