sign in sign up
<<
Home , East Antarctica, Geology of Antarctica

and Planetary Science Letters 290 (2010) 281–288

Contents lists available at ScienceDirect

Earth and Planetary Science Letters

journal homepage: www.elsevier.com/locate/epsl

History of ice sheet elevation in East : Paleoclimatic implications

Xiaohan Liu a,b,⁎, Feixin Huang a,c, Ping Kong b, Aimin Fang b, Xiaoli Li b, Yitai Ju c a Institute of Tibetan Research, Chinese Academy of Sciences, Beijing, 100085, China b Institute of and Geophysics, Chinese Academy of Sciences, Beijing, 100029, China c Mineral Resource Institute of China Metallurgical Geology Bureau, Beijing, 100025, China article info abstract

Article history: The multi-disciplinary study of past ice surface elevations in the Grove Mountains of interior Received 17 July 2007 provides direct land-based data on the behavior of the East Ice Sheet since the Pliocene. The glacial Received in revised form 3 December 2009 geology, the ages of cold soils, the depositional environment of younger moraine sedimentary Accepted 6 December 2009 boulders and their spore–pollen assemblages combine to imply a possible significant shrinkage of the Ice Available online 13 January 2010 Sheet before the Middle Pliocene Epoch, with the Ice Sheet margin retreating south of the Grove Mountains Editor: Y. Ricard (∼450 km south from its present coastal position). Exposure age measurements of bedrock indicate that the elevation of the ice surface in the Grove Mountains subsequently rose at about mid-Pliocene to at Keywords: least 200 m higher than today's levels. The ice surface then progressively lowered, with some minor Grove Mountains fluctuations. Middle to Late Pleistocene exposure ages found on the lowest samples, at the ice/bedrock fluctuation of ice surface contact line, indicate a long period with ice surface elevations kept at the current level or complex – spore pollen assemblage fluctuation history during the Quaternary Epoch. cosmogenic nuclide exposure age © 2009 Elsevier B.V. All rights reserved. Pliocene

1. Introduction constrain ice sheet models during known glacial maxima and minima in the post-14 Ma time-frame, when the earth's geography and its climate Interest in the past behavior of the East (EAIS) has system have roughly resembled their present configurations. been piqued by the growing awareness of an active subglacial Land-based data concerning the past ice surface elevation in the hydrological system (Wingham et al., 2006). This and other findings interior of the EAIS are scarce because of extremely difficult access. question whether the EAIS is as stable as had hitherto been assumed. As Outside the region, only a few paleo ice levels the world's largest glacial system, the EAIS first formed around 34 Ma ago, have been reported (Ingolfsson et al., 1998). In the Lambert Graben, associated in part with the thermal isolation of Antarctica by the opening northern Prince Charles Mountains, glacial sedimentary strata have of the in conjunction with declining global CO2 levels recorded fluctuations in Neogene glacial ice extent (Mabin, 1992; (DeConto and Pollard, 2003). The ice sheet there oscillated on Adamson et al., 1997; Hambrey and McKelvey, 2000a,b), and observa- Milankovitch frequencies (Naish et al., 2001).TheEAISextendedbeyond tions of glacial topography reveal expansion and retreat of the EAIS its present continental margin during glacial times and retreated during margin in the region of Dronning Maud Land () (Jonsson, interglacial times to expose a coastal terrain that supported a low 1988; Holmlund and Näslund, 1994; Lintinen and Nenonen, 1997). This woodland forest (Barrett, 2007). Around 14 Ma ago the EAIS entered a paper provides new data on past ice sheet elevations from the Grove more stable state, during which many believe ice persisted in central East Mountains, 400 km inland from the coast at Prydz Bay, where the Antarctica with the ice volume and ice extent varying modestly, perhaps nunataks protrude hundreds of meters above the contemporary ice by less than 1/3 of today's 66 m of sea level equivalent (Kennett and surface (e.g., Mt. Harding, at 2338 m). Field studies were undertaken in Hodell, 1993; Barrett, 1996). A proposal in the 1980's that the EAIS might this region by the Chinese National Antarctic Research Expedition have shrunk by as much as 2/3 to allow the ocean to flood central East (CHINARE) from 1998 to 2005. The results of this endeavor include Antarctica (Burckle and Potter, 1996)hasbeencontested(Hicock et al., interpretations of glacial geology, recognition and analyses of cold desert 1996; Stroeven and Prentice, 1997; Harwood and Webb, 1998), but soils, lithologic analyses of sedimentary boulders of moraines, along with continues to have some credence (Haywood et al., 2002). Regardless of analyses of spore–pollen assemblages. Samples of bedrock from two that issue, a key constraint on our understanding of the EAIS behavior nunataks provide in situ cosmogenic nuclide 10Be and 26Al exposure ages. remains—the lack of direct evidence of ice sheet surface levels to 2. Geographic setting

⁎ Corresponding author. Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing, 100029, China. Fax: +86 10 62375165. The Grove Mountains emerge as a group of 64 isolated nunataks 2 E-mail address: [email protected] (X. Liu). scattered over an area of ∼3200 km within the EAIS (72°20′Sto73°10′

0012-821X/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.epsl.2009.12.008 282 X. Liu et al. / Earth and Planetary Science Letters 290 (2010) 281–288

S, and 73°50′Eto75°40′E). The nunataks can be divided into 5 ranges are usually jagged, with summit areas containing well-developed forming a ridge–valley topography with an NNE trend. The regional ice ventifacts facing the dominant wind from the SE (Fig. A1). The wind has flows north-westward, perpendicular to the ridges and away from the cut meters into hard rock, suggesting that these higher areas have been ice central part of the EAIS (i.e., Dome Argus), where the ice surface free for at least several hundred thousand years. However the lower elevation exceeds 4 050 m above sea level (Fig. 1). slopes, within ∼100 m from recent ice surface, contain more recent glacial The Grove Mountains nunataks extend along the local seasonal features, such as fresh trimlines and erratics. Lower nunataks have been equilibrium line separating the more coastal zone of net ablation from eroded into “roches moutonnée” forms with oriented surface striae, the inland zone of net accumulation. Because they form obstacles to presumably from the flow of overriding ice (Fig. A2). We consider this ice flow, these nunataks are responsible for the existence of boundary between the effects of wind and glacial erosion to indicate the neighboring blue ice , where wind-induced ablation exceeds level of a former persistent ice surface, perhaps during the late Pliocene or the local snow accumulation. The ice surface elevation in this area early Quaternary, that has not been subsequently overtopped. averages ∼2000 m above sea level. Mount Harding in the central part of the Grove Mountains appears These nunataks consist mainly of upper to crescent shaped, open to the northwest. Both the northern and southern facies metamorphic rocks, syn-orogenic to late orogenic granite, and ends form steep crests, protruding ∼200 m above the current ice post tectonic granodioritic aplite and pegmatite. The time of surface. The central segment of the ridge-line between two summits deformation (∼550 Ma) indicates that the Grove Mountains, like descends progressively until it reaches the ice surface at a central col, similar rocks in Prydz Bay, form part of the Pan-African orogenic belt with a relic ice tongue hanging on the lee side. A stagnant field of blue (Liu et al., 2003). The absence of active structures and earthquakes, ice, tens of square kilometers wide, lies inside the crescent. An arc- and the lack of Cenozoic suggest that this region, along shaped ice-cored moraine ∼100 m wide and over 5 km long extends with Prydz Bay, has been geologically stable since at least the Late along the western edge of this blue ice field (Fig. A3). An old ice tongue Era (Tingey, 1991). once overrode the central col of Mount Harding, leaving this terminal moraine as the ice surface lowered. The gravel cover of this moraine has 3. Glacial geology protected the blue ice beneath from ablation since the last ice retreat, leaving it ∼25 m above the surface of the surrounding ice. The slopes of nunataks facing due the upstream of ice flow show smoothly abraded and striated bedrock, with occasional patches of 4. Sedimentary boulders and spore–pollen assemblages diamicton. The slopes facing the downstream of ice flow typically form bluffs that have been steepened by glacial abrasion. The nunataks leave The ice-cored moraine consists mainly of clasts of local metamorphic trails of superglacial debris tens of km in length on the blue ice surface, rock derived from Mount Harding, sub-angular to angular in shape and marking the present ice flow trace. The upper parts of the higher nunataks ranging in size from centimeters to meters. It also contains exotic clasts

Fig. 1. The sketch map showing locality, landscape and ice flow lines in central part of Grove Mountains. X. Liu et al. / Earth and Planetary Science Letters 290 (2010) 281–288 283 of similar size range taken by the ice flow from some distance to the assemblages representing a continental flora. The compositions of south-east, beneath the interior of the EAIS. The exotic clasts include these assemblages resemble the Neogene Weddellian biogeocenose, such as high-pressure granulite, ophiolitic rocks, including the major pollen types such as Penaceae, Podocarpus, meteorites, and some sedimentary rocks with subrounded shapes and Araucardiaceae, Chenopodiaceae and Artemisia (Fang et al., 2005b). varnished, striated, and pitted surfaces. Lithologies of the sedimentary The herbaceous angiosperms are important for distinguishing and boulders range from weakly consolidated diamicton to well consolidat- subdividing the Neogene . The first herbage occurs in the early ed and gravel-bearing argillaceous sandstone with clay or , with increasing species and abundance into the middle calcareous cement (Fig. A4). Most of these sedimentary boulders should Miocene. These reached their peak in the Pliocene, when major have ages older than the formation of the continental scale stable EAIS assemblages occurred in some special environments, such as the (e.g. 14 Ma), but some of them, the glaciofluvial genetic, weakly assemblage of Chenopodiaceae, Atemisia and Polygonaceae. Although consolidated diamicton, exhibit strictly younger features than those of the abundance of herbaceous pollen is not high in our samples, some well consolidated sedimentary boulders. Atemisiaepollenites, Chenopodiaceae and Gramineae are common. 35 blocks of sedimentary rocks were collected from the ice-cored We infer therefore a Neogene, possibly Pliocene, age for these moraines for the analyses in order to determine the hydrodynamic sedimentary clasts (Fang et al., 2007). environment during the formation of these boulders, especially the These young exotic sedimentary clasts in the Grove Mountains young weakly consolidated diamicton. All these samples consist of non- have some common lithologic features that compare with other sorted clasts of various sizes and sandy mud matrix and showing a Pliocene strata in East Antarctica. Comparable strata include the massive texture. We chose the matrix of 11 consolidated to half- Pagodroma Group outcrops in the Prince Charles Mountains (White- consolidated samples making thin sections and statistically counted the head and McKelvey, 2001; McKelvey et al., 2001), and the Sфrsdal mineral components of their sandy particles. According to the EMS Formation outcrops in the Vestford Hills (Quilty, 1992; Quilty et al., analysis upon the compositions of the fine cementing minerals in the 2000). They are all composed of a series of glaciogenic and matrix, there are two kinds of cementation in these debris. One is nonglaciogenic units of compact sedimentary rocks overlapping the cemented mainly by calcite, the other is cemented by clay. A total of 7 Tertiary strata with unconformity surface, and possess an amount of samples were selected for geochemical analysis. There are two groups reworked micro-fossils. Preliminary results of our pollen assemblage that can be subdivided according to the major element contents. One analyses of these young clasts resemble those from the Meyer Desert has a high content in CaO, low in SiO2,Al2O3 and FeO, cemented by Formation, non-marine glaciogenic deposits forming the upper part of calcite. The other has contrary compositions, cemented by clay. Such the Sirius Group in the Transantarctic Mountains, with a biostrati- differences show it attributed not only to the different source rocks, it graphic age of less than 3.8 Ma based on reworked marine diatom may also result from the differences in their sedimentary environments data (Ashworth and Cabtrill, 2004). and weathering conditions, i.e. they may come from different Further evidence of a Pliocene retreat of the ice sheet margin in lithological units. Quartz grains of similar sizes (0.25 to 0.5 mm in this sector of Antarctica comes from the Vestford Hills and Larsemann diameter) were picked out from two samples and observed under a Hills (mainly marine deposits with deep water foraminifera (Quilty, scanning electron microscope (SEM). Quartz grain shapes from the 1992; Quilty et al., 2000; Harwood et al., 2000)), and fiord deposits in diamicton matrix are angular (70–80%), subangular (10–20%), sub- the northern Prince Charles Mountains (Mabin, 1992; Adamson et al., rounded (5–10%) and rounded (∼5%). More than 50% of the grains 1997; Hambrey and McKelvey, 2000a,b). Whitehead and others display surface textures (such as oriented striae, V-shaped pits, or (Barker et al., 1998; Whitehead, 2000) have recently reported that the conchoidal fractures) that suggest mechanical abrasion typical of glacial sedimentary facies of Pliocene glacial deposits in the southern Prince transport. 4 were chosen for the grain sizes characteristics Charles Mountains is continental, and Baker et al. (Whitehead and analysis under an IBAS-2000 image analysis instrument. The values of McKelvey, 2002) indicate that the corresponding strata in the Prydz mean size (Mz) are similar, and their skewness (SK1) all belongs to Bay derive from deep water ooze. The gradual transition from deep minus, showing the feature of subglacial tillites. As for the standard water to continental facies, from Prydz Bay, via the Vestford Hills and deviation (σ1) and kurtosis (Kg), there are differences between the Larsemann Hills, to the Prince Charles Mountains shows significant samples, which show different degrees of sorting, and may represent paleo-topographic features, and supports the idea that the margin of the post-depositional modifications of the grain size distribution by the EAIS was once located some distance upstream of the Grove flowing water or changes of their medium dynamics during the Mountains, at more than ∼450 km south of the margin's present processes of transportation and sedimentation. Such deduction is position. supported by the frequency curve and log-probability plot of samples (Fang et al., 2004; Fang et al., 2005a). 5. Desert soils From the analysis upon their textural characteristics, mineral components of clasts as well as matrix, geochemical compositions, The southern slope of Mount Harding exposes several cold desert and surface textures of quarts grains, coupled with the grain size soil patches, distributed along a sub-horizontal zone higher than the characteristics, indicate that these sedimentary rocks have very bad regional wind–ice erosion limit of ∼100 m above the recent ice textural and compositional maturity, they suffered post-depositional surface (Fig. 1)(Fig. A5). The preservation of such soils indicates that modifications by subglacial streams or melt-water action. Such the regional ice sheet's surface has not overtopped this limit elevation complex multi-media dynamic conditions exist only on the margin since the soil formed. These soil layers are characteristics of the of warm-based ice sheets. We interpret these diamicton clasts as widespread occurrence of surface desert pavement, abundant water- products of a marginal regime of wet-based ice sheet, rather than soluble salts, strongly stained upper portion of soil profile, slightly from the cold interior of the EAIS. These clasts could have originated i) acid and negligible organic matter content. A 1:5 soil–water extracts prior to the formation of a persistent EAIS in the middle Miocene, or ii) analysis indicates that the dominant cations are Mg2+ and Na+, 2+ + 2− − during a subsequent period of large scale collapse and retreat of the followed by Ca and K , and the main anion is SO 4, then Cl and − EAIS margin. NO3 . The water laid structures and water-soluble salts indicate ion Our study has identified more than 25 species of Neogene plant transportation in the frigid and arid environment. The distribution of microfossil from the diamicton clasts. These spore and pollen the salts is related to the maximum content of moisture and clay assemblages derive from a suite of glaciogenic strata hidden beneath mineral. Clay fraction migration occurs in the soils, which is different the EAIS, and therefore provide useful information on the evolution of from that of other cold desert soils in Antarctica. The upper horizons of the EAIS. Most of the pollen and spores appear to be in situ some soil profiles are generally stained, a process called rubification, 284 X. Liu et al. / Earth and Planetary Science Letters 290 (2010) 281–288

Table 1 Occurrence of spore–pollen from the glacial erratics in the Grove Mountains.

Spore and pollen types S1501 S1507 S1509 S1514 S1604 S1605 S1606 Nj02 Nj04 Nj05 Nj07 Nj08 Total

Cryptogam spores 28 Toroisporis (Lygodiaceae) 4 1 5 Granulatisporites (Pteridaceae?) 1 1 2 Osmunda (Osmundaceae) 5 1 4 1 11 Polypodiaceae 1 1 2 Magnastriatites (Parkeriaceae) 1 4 Deltoidospora 3 1 Trilete spore 31 3 Gymnospermous pollen 99 Araucardiaceae 3 1 2 6 Podocarpus (Podocarpaceae) 9 3 1 6 1 1 Dacrydium (Podocarpaceae) 1 2 1 1 20 Taxodiaceae 1 5 Pinus (Pinaceae) 24 3 7 2 1 3 40 Keteleeria (Pinaceae) 1 1 Picea (Pinaceae) 22 1 2 25 Tsuga (Pinaceae) 11 Angiospermous pollen 64 Chenopodiaceae 22522 12 16 Artemisia (Asteraceae) 2 2 2 1 7 Asteraceae 11 2 Gramineae 22 Fraxinoipollenites (Oleaceae) 2 2 Oleoidearumpollenites (Oleaceae) 1 1 Oleaceae 11 Operculumpollis 1 1 Nothofagidites (Nothofagus)2111 5 Rhus 21 3 Quercus (Fagaceae) 1 1 3 1 3 3 12 Juglans (Juglandaceae) 1 1 Pterocarya (Juglandaceae) 11 Liquidambar (Hamamelidaceae) 2 2 Ulmus (Ulmaceae) 1 1 1 3 Ulmoidepites (Ulmaceae) 1 1 Tilia 1 1 Proteacidites(Proteaceae) 1 12 Tricolpopollenites 1 1 Total grains 33 6 17 4 81 9 16 2 1589191

primarily because of the weathering of iron-bearing minerals. The belong to the third–fourth weathering stages, mainly the fourth stage. reddish hues of cold desert soils have been attributed to relatively Surface boulders of the soils are rare and show very strong cavernous high concentrations of dithionite-extractable Fe (Fed)(Li et al., 2003). weathering or rounding, but most boulders are small, polished, and According to the available data observed at present time, the daily well stained. Some desert varnish and ventifacts are also found in this temperature in the Grove Mountains ranges from −13.1 to −22.6 °C zone. Within the profile, soil colors are 5YR5/4 in the upper horizons, in January, i.e. the highest temperature in summer is far below zero. paler in lower horizons. In the profile of this area, salts are often As the formation of soils requires melt-water, their presence indicates abundant beneath surface boulders and also through the profiles. a warmer climate event, with at least seasonal ice-thawing, once They are most commonly found in a continuous or discontinuous existed in the area. Because of the lack of fossils and organic carbon, the weathering stage analysis of cold desert soils has become a common chronological tool in Antarctica. The weathering processes can be indicated in the soil profiles by a number of distinctive morphological properties, related to a number of factors, including topography, climate, and time. With the characteristics of surface rocks, soil colors, horizon differentiation, salt in soils and soil depth, it is possible to recognize up to six weathering stages in soils formed on moraine and regolith along the Transantarctic Mountains (Campbell and Claridge, 1975; Camp- bell and Claridge, 1987). Different weathering stages correspond to different soil ages. With the increase of weathering, the configuration of surface rocks varies from fresh, unstained, coarse and angular of stage 1 to very strongly stained, rounded, crumbled bedrock of stage 6; the soil color varies from pale olive to light grey of stage 1 to strong brown to yellowish red and dark red of stage 6; the soil horizon development varies from nil of stage 1 to very distinct of stage 6; the soil salts absent from stage 1, present in a few flecks in stage 2, form a continuous horizon 20–30 cm below the surface in stage 6. Based on Fig. 2. Plot of 26Al/10Be vs. 10Be concentrations for R9204, R9207, R9213 and R9216. 10Be standard soil weathering characteristics from stage 1 to stage 6, the concentrations have been normalized to sea level and high latitude according to the weathering feature of the soils in the Mount Harding is considered to scaling method of Lal (Lal, 1991), modified by Stone (Stone, 2000). X. Liu et al. / Earth and Planetary Science Letters 290 (2010) 281–288 285

Fig. 3. Minimum 10Be exposure ages vs. altitudes. The ages of samples 9204, 9207, 9213 and 9216 have been corrected, compared to the previous minimum exposure ages (Huang et al., 2008)(Table 2 reflects the new values).

horizon 5–10 cm below the surface of the soil. All these weathering 5.1 and 31.1 atoms g−1 yr−1 for 10Be and 26Al, respectively, for sea level features suggest that they formed between 0.5 and 3.5 Ma before and high latitude in the calculation. present (Li et al., 2003). Despite such large range of ages, this provides Preliminary exposure ages for twelve bedrock samples indicate 10Be some solid evidence constraining the timing of local ice surface minimum exposure ages ranging from 3.29 to 0.57 Ma, and the 26Al fluctuations. minimum exposure ages ranging from 2.13 to 0.18 Ma (Huang et al., 2008). Samples from the crests of Zakharoff Ridge and Mount Harding 6. Exposure ages have minimum 10Be ages of 2.00±0.22 and 2.30±0.26 Ma, respec- tively. These dates suggest that the crests have been above the ice sheet To provide a chronology, we sampled bedrock profiles on Zakhar- surface since at least the Plio-Pleistocene boundary. Adopting a off Ridge (Fig. A6) and Mount Harding (two typical nunataks in the ‘reasonable’ erosion rate of 5–10 cm/Ma increases the exposure ages Grove Mountains), and calculated surface exposure dates using in situ of these two samples into the mid-Pliocene. Thus the bedrock exposure cosmogenic nuclides 10Be and 26Al (Huang et al., 2008). The two ages indicate an ∼200 m decrease of the ice sheet surface elevation in bedrock slopes sampled are topographically smooth with stable low the Grove Mountains since mid-Pliocene time. Analyses of seven angles of about ∼5°. They consist of granitic , with thin foliation samples from the two profiles resulted in inconsistent 10Be and 26Al (exfoliation) layers parallel to the slope surfaces. minimum exposure ages, and the five lower elevation samples The cosmogenic nuclide lab at the Institute of Geology and suggested complex exposure histories. The higher two samples Geophysics, Chinese Academy of Sciences, carried out chemical (R9204 and R9207), which have old 10Be minimum exposure ages preparations. Samples were first crushed to 0.1–1.0 mm size, after (N2.9 Ma), but much younger 26Al minimum exposure ages (b1.8 Ma) which, each sample underwent magnetic separation. Quartz samples (Huang et al., 2008), presented a dilemma for interpretation. were purified by leaching 4 or 5 times in a hot ultrasonic bath with a To determine the cause of the inconsistent 10Be and 26Al minimum mixed solution of HF and HNO3 (Kohl and Nishiizumi, 1992), and were exposure ages for R9204 and R9207, we re-treated the four samples completely dissolved together with ∼0.5 mg 9Be carrier. Beryllium and from the south Mount Harding profile that had appeared inconsistent. aluminum were separated by ion chromatography, their hydroxides This time we modified our procedures to avoid errors in the Al precipitated, and then baked to oxides at 850 °C. Procedure blanks were concentration measurement. In the new procedure, we isolated used to correct the measured values. Total Al concentrations were aliquots for Al measurement by ICP-OES before fume off HF in the quantified in aliquots of the dissolved quartz by ICP-OES, and 10Be and process of quartz dissolution. We checked the new procedure by 26Al concentrations were measured by the accelerator mass spectrom- comparing multi aliquot Al concentrations with those for identical 1 g etry (AMS) at the Australian Nuclear Science and Technology quartz fractions that showed consistency within 10%. Table 1 lists the Organisation (ANSTO). The minimum exposure ages shown in Table 1 measured values and calculated ages for the four samples. Fig. 2 shows were calculated using the scaling method of Lal (Lal, 1991), modified by the 26Al/10Be ratios vs. 10Be concentrations. With this new procedure, Stone (Stone, 2000)forAntarctica.Thisstudyusedproductionratesof the four samples in the south Mount Harding profile show consistent

Table 2 Elevations and minimum exposure ages of bedrock samples from south Mount Harding, Grove Mountains.

Sample Elev Quartz 10Be 26Al 26Al/10Be Minimum 10Be age Minimum 26Al age (m) (g) (106 atoms/g) (106 atoms/g) (Ma) (Ma)

R9204 2275 17.10 74.44±2.00 249.63±8.42 3.35±0.23 3.29±0.51 2.51±0.77 R9207 2250 12.71 74.83±2.24 254.34±8.16 3.40±0.22 3.47±0.58 3.02±1.29 R9213 2200 16.05 34.09±0.96 130.20±17.99 3.82±0.27 1.03±0.09 0.71±0.16 R9216 2175 25.36 32.96±1.01 157.03±7.21 4.76±0.30 1.01±0.09 0.98±0.12

Minimum 10Be and 26Al exposure ages calculated using the scaling method for Antarctica from Lal (Lal, 1991), modified by Stone (Stone, 2000). Errors of minimum 10Be and 26Al exposure ages include: 2% from AMS, 6% from production rate, 1% from Be carrier, and 4% from ICP-OES for Al. 286 X. Liu et al. / Earth and Planetary Science Letters 290 (2010) 281–288

Fig. 4. History of ice surface fluctuations in the Grove Mountains region since the Pliocene. Bold line represents the ice surface fluctuation through time, the solid line is supported by bedrock exposure ages, and the dashed lines on both sides of the solid line are inferred based on erosion, topographic features, and soils. The ice margin retreat, probably before the Middle Pliocene Warmth (3.29–2.97 Ma), is based on the sedimentary boulders and their spore–pollen assemblages.

10Be and 26Al minimum exposure ages, and their 10Be minimum 2. The elevation of the ice surface in the Grove Mountains region rose exposure ages remain the same as the previous results. This change in to more than 200 m higher than today at about the mid-Pliocene, procedure indicates that the apparently inconsistent 10Be and 26Al and remained high even during the Middle Pliocene Climatic Warm ages of the four samples (Huang et al., 2008) resulted from Al Event (3.29–2.97 Ma), based on the sea surface temperature and concentration errors. Our data now show that all samples in this deep-sea bottom-water temperatures from ostracode Mg/Ca ratios profile have simple exposure histories. The following discussion is (Dowsett et al., 1996; Cronin et al., 2005). based on the minimum 10Be exposure ages because we believe them 3. The ice surface then progressively lowered with some minor to be fairly accurate and closer to the actual exposure ages than fluctuations during the general global cooling trend based on minimum 26Al ages, which may have been influenced by erosion. reconstructions from deep-marine oxygen isotope records, Overall, the re-calculated bedrock exposure ages tend to be older resulted probably from less precipitation inland of the EAIS. for samples at higher elevations and younger at lower elevations, and 4. Exposure ages of the lowest samples, from the ice/bedrock contact the minimum exposure ages of R9204 and R9207 are ∼3.29 and line, are Middle to Late Pleistocene. This indicates two possibilities: 3.47 Ma. These ages suggest that deglaciation in the Grove Mountains one is that the EAIS surface have remained at the current level for a began no later than mid-Pliocene and lasted no longer than mid to late long period, the other is that the ice surface has re-raised to the Pleistocene. current limit of today from a lower level, with a complex history of Fig. 3 shows the minimum 10Be exposure ages vs. elevation on the fluctuation during the Quaternary Epoch (Fig. 4). two profiles. The oldest minimum 10Be exposure age on the Mount Harding profile is N3.0 Ma. But the age on the pinnacle (∼200 m above Our results, obtained with a combination of methods, support a current ice surface) is 2.30 Ma. On the top of the Zakharoff Ridge profile dynamic model of the interior EAIS since the Pliocene. Because the (∼150 m above current ice surface), the oldest minimum 10Be 10Be and 26Al concentrations have not yet reached secular equilibrium exposures age is 2.00 Ma. All of our samples exhibit minimum 10Be at the tops of the two profiles we sampled, the highest ice surface level exposure ages that decrease downwards to the Pleistocene. The in the Grove Mountains may have been still much higher. The onset of relationship between the 10Be minimum exposure ages and the lowering of the ice surface might have occurred earlier than we infer elevations of samples does not display a linear trend, but does show here. the general tendency of a steadily decreasing ice sheet. The cause of this There have been similar efforts at looking at exposure age dating in non-linear appearance could be due to the nature of the rock samples various age ranges in the EAIS. D. Fink et al. have reported the history (different erosion rates) and/or errors in the cosmogenic nuclide geo- of ice surface fluctuation from the Pliocene–Pleistocene boundary up chronology. to the beginning of the Holocene in the Region. The The relationship between the cosmogenic nuclide concentrations ages cluster into three groups that decrease in age with both and the elevations of the samples shows that the 10Be and 26Al decreasing altitude (from 1260 to 70 m above sea level) and concentrations have not yet reached secular equilibrium, indicating a increasing proximity to the Lambert Glacier-Amery Ice Shelf drainage continuously decrease of the ice surface elevation from the mid- system (Fink et al., 2006). A. Mackintosh et al. presented that the Pliocene. Furthermore, the oldest ages on the profiles would imply coastal area thinned by at most 350 m in this region during the past that ice covered the summits before the mid-Pliocene Epoch. The 13 ka (Mackintosh et al., 2007). S. Strasky et al. have concluded that a boundary observed in the field between wind and glacial erosion major ice advance reaching elevations of about 500 m above present mentioned above (about 100 m higher than today's ice level) exhibits ice levels occurred between 1.125 and 1.375 Ma before the present in ages near the Pliocene/Pleistocene boundary, and provides no with erosion-corrected 21Ne and 10Be exposure ages. evidence of a higher subsequent ice surface (Fig. 3). Subsequent ice fluctuations were of lesser extent with complex exposure histories (Strasky et al., 2009). L.D. Nicola et al. reported 7. Conclusion and discussion relict landscape features eroded by extensive ice overriding the whole coastal area before at least 6 Ma in the region. Since Based on geomorphic evidence, inferred ages of soil formation, then, summit surfaces were continuously exposed and well preserved lithologic analyses of sedimentary boulders, and spore–pollen assem- under polar conditions with negligible erosion rates. The combination blages, coupled with in situ cosmogenic nuclide exposure ages, we of stable and radionuclide isotopes of drifts documents complex suggest that: multiple Pleistocene glacial cycles (Nicola et al., 2009). P. Oberholzer et al. have yielded minimum ages of 11 to 34 ka, 309 ka, and 2.6 Ma. 1. The margin of the EAIS may have been south of the Grove Taking erosion into account, the oldest surface is 5.3 Ma old. The ice Mountains region, some 450 km south of its present coastal surface lowered before the Mid-Pliocene. Subsequent glacial events position, before the mid-Pliocene Epoch. reached an elevation of hundreds of meters, but were restricted to the X. Liu et al. / Earth and Planetary Science Letters 290 (2010) 281–288 287

Fig. 5. Summary of combine time vs. relative ice elevation change for other studies of the EAIS (Oberholzer et al., 2003; Fink et al., 2006; Mackintosh et al., 2007; Strasky et al., 2009; Nicola et al., 2009).

valley. Such advances took place at least twice throughout the Appendix A. Supplementary data Pleistocene (Oberholzer et al., 2003)(Fig. 5). Our evidence from the sedimentary boulders supporting a Pliocene Supplementary data associated with this article can be found, in ice sheet collapse remains weak. The age of the comparable Meyer the online version, at doi:10.1016/j.epsl.2009.12.008. Desert Formation remains unresolved. Data from reworked marine diatoms and age estimations for paleosols on glacial deposits suggest a Pliocene age (3 Ma). However, new cosmogenic dating and dating of References volcanic ash deposits in the Dry Valleys (Transantarctic Mountains) Adamson, D.A., Mabin, M.C.G., Luly, J.G., 1997. Holocene isostasy and late Cenozoic indicate that the Sirius Group deposits and correlated deposits on the development of landforms including Beaver and Radok Lake basins in the Amery – Beardmore Glacier are Middle Miocene, appropriate for the transition Oasis, Prince Charles Mountains, Antarctica. Antarctic Sci. 9, 299 306. Ashworth, A.C., Cabtrill, D.J., 2004. Neogene vegetation of the Meyer Desert Formation from wet-to-cold based glaciation (Marchant et al., 2002; Francis (Sirius Group) Transantarctic Mountains, Antarctica. Palaeogeogr. Palaeoclimatol. et al., 2007; Lewis et al., 2007, 2008). Herbaceous pollen and the Palaeoecol. 213, 65–82. position distribution of Pliocene strata outcropping around the Barrett, P.J., 1996. Antarctic palaeoenvironment through Cenozoic times — a review. Terra Antartica 3, 103–119. Lambert Glacier region reveal a rapid ice sheet collapse at a young Barker, P.F., Barrett, P.J., Camerlenghi, A., et al., 1998. Ice sheet history from Antarctic age as Pliocene. The chronology of sedimentary boulders needs to be continental margin sediments: the ANTOSTRAT approach. Terra Antarct. 5, 737–760. corroborated by further pollen assemblage studies and the offshore Barrett, P.J., 2007. Cenozoic climate and sea level history from glacimarine strata off the Victoria Land coast, , Antarctica. In: Hambrey, M.J., Christoffersen, P., sediment record in Prydz Bay. Our scenario corresponds roughly with Glasser, N.F., Hubbart, B. (Eds.), Glacial Processes and products, International Association climatic events on a global scale. For example, an extensive ice sheet of Sedimentologists Special Publication 39, 259–287. started to form in the during the Pliocene. More Burckle, L.H., Potter, N., 1996. Pliocene–Pleistocene diatoms in and Mesozoic fl sedimentary and igneous rocks from Antarctica: a Sirius problem solved. Geology frequent, but lower amplitude, uctuations of the ice surface during 24, 235–238. the Quaternary correspond with the glacial cycles recorded in the Campbell, I.B., Claridge, G.G.C., 1975. Morphology and age relationship of Antarctica northern hemisphere. Overall, our new land-based data imply that the soils. In: Suggate, R.P., Cresswell, M.M. (Eds.), Quaternary Studies: Royal Society of – EAIS ice surface fluctuation depended not simply on the global surface New Zealand Bulletin, vol. 13, pp. 83 88. Campbell, I.B., Claridge, G.G.C., 1987. Antarctica: Soils, Weathering Processes and temperature, but possible other factors such as large scale tectonic Environment: Amsterdam, Elsev. Sci. Pub., pp. 1–368. events have to be considered. Cronin, T.M., Dowsett, H.J., Dwyer, G.S., Baker, P.A., Chandler, M.A., 2005. Mid-Pliocene Despite the report of Cenozoic up-lifting (presumed to be deep-sea bottom-water temperatures based on ostracode Mg/Ca ratios. Mar. Micropalaeont. 54, 249–261. isostatic) of bedrock along the west coast of the Lambert Graben DeConto, R.M., Pollard, D.A., 2003. Rapid Cenozoic glaciation of Antarctica induced by (Hambrey and McKelvey, 2000a,b), we did not detect evidence of declining atmospheric CO2. Nature 421, 245–249. regional tectonic activity in the Grove Mountains region from our Dowsett, H., Barron, J., Poore, R., 1996. Middle Pliocene sea surface temperatures: a global reconstruction. Mar. Micropaleontol. 27, 13–25. geological investigations. The bedrock shows a stable high meta- Fang, A.M., Liu, X.H., Lee, H.I., Li, X.L., Huang, F.X., 2004. Sedimentary environments of morphic crust without active structures such as faults, the Cenozoic sedimentary debris found in the moraines of the Grove Mountains, thrusting, young sediments or Cenozoic volcanism and earth- east Antarctica and its climatic implications. Prog. Nat. Sci. 14, 3, 223–234. Fang, A.M., Liu, X.H., Li, X.L., Huang, F.X., Yu, L.J., 2005a. Cenozoic glaciogenic sedimentary quakes, etc. Consequently, our conclusions do not consider regional record in the Grove Mountains of East Antarctica. Antarct. Sci. 17 (2), 237–240. gravitational isostatic correction during the fluctuation of the ice Fang, A.M., Liu, X.H., Wang, W.M., Li, X.L., Yu, L.J., Huang, F.X., 2005b. Preliminary study sheet. on the spore–pollen assemblages found in the Cenozoic sedimentary rocks in Grove Mountains, East Antarctica and its climatic implications. Chin. J. Pol. Res. 16 (1), 23–32. Acknowledgements Fang, A.M., Liu, X.H., Wang, W.M., Lee, J.I., 2007. Spores and pollen from glacial erratics in the Grove Mountains, East Antarctica. U.S. Geological Survey and The National This work was supported by grants from National Natural Science Academies; USGS OF-2007-1047. Extended Abstract 122. Fink, D., McKelvey, B., Hambrey, M.J., Fabel, D., Brown, R., 2006. Pleistocene deglaciation Foundation of China (40631004 and 40506003), and field logistics chronology of the Amery Oasis and Radok Lake, northern Prince Charles Mountains, was supported by the Chinese Polar Research Administration grants Antaectica. Earth Planet. Sci. Let. 243, 229–243. during 1998–2005. We thank Professor Paul Tapponnier and Kenneth Francis, J.E., Haywood, A., Ashworth, A.C., Valdes, J., 2007. environments in the Neogene Sirius Group, Antarctica: evidence from the geological record and coupled J. Hsü for their constructive comments. We thank Professor Peter atmosphere–vegetation models. J. Geol. Soc. London 164, 1–6. Barrett and Tim Naish, together with Editor Claude Jaupart for their Hambrey, M.J., McKelvey, B.C., 2000a. Neogene fjordal sedimentation on the western insightful and patient comments that helped to improve this study. margin of the Lambert Graben. East Antarctica, Sedimentology 47, 577–607. Hambrey, M.J., McKelvey, B.C., 2000b. Major Neogene fluctuation of the East Antarctic We thank Dr. Bill Isherwood for assistance with the wording of this Ice Sheet, Stratigraphic evidence from the Lambert Glacial region. Geology 28 (10), manuscript. 887–890. 288 X. Liu et al. / Earth and Planetary Science Letters 290 (2010) 281–288

Harwood, D.M., Webb, P.N., 1998. Glacial transport of diatoms in the Antarctic Sirius little change in ice-sheet thickness since the Last Glacial Maximum. Geology 35 (6), Group: Pliocene refrigerator. GSA TODAY 8, 1–8. 551–554. Harwood, D.M., McMinn, A., Quilty, P.G., 2000. Diatom biostratigraphy and age of the Marchant, D.R., Lewis, A., Phillips, W.C., Moore, E.J., Souchez, R., Landis, G.P., 2002. Pliocene Sorsdal formation, Vestfold Hills, east Antarctica. Antarct. Sci. 12 (4), Formation of patterned-ground and sublimation till over Miocene glacier ice in 443–462. , Antarctica. Geol. Soc. Am. Bull. 114, 718–730. Haywood, A.M., Valdes, P.J., Sellwood, B.W., Kaplan, J.O., 2002. Antarctic climate during McKelvey, B.C., Hambrey, M.J., Harwood, D.M., 2001. The Pagodroma Group — a the middle Pliocene: model sensitivity to ice sheet variation. Palaeogeogr. Cenozoic record of the in the northern Prince Charles Palaeoclimatol. Palaeoecol. 182, 93–115. Mountains. Antarct. Sci. 13 (4), 455–468. Hicock, S.R., Barrett, P.J., Holme, P.J., 1996. Fragment of an ancient outlet glacier system Naish, T.R., Woolfe, K.J., Barrett, P.J., Wilson, G.S., et al., 2001. Orbitally induced near the top of the Transantarctic Mountains. Geology 31, 821–824. oscillations in the East Antarctic ice sheet at the Oligocene–Miocene boundary. Holmlund, P., Näslund, J.O., 1994. The glacially-sculptured landscape. Dronning Maud Nature 413, 719–723. Land, Antarctica, formed by wet-based mountain glaciation and not by the present Nicola, L.D., Strasky, S., Schlüchter, C., Salvatore, M.C., Akçar, N., Kubic, P.W., Christl, M., ice sheet, Boreas, vol. 23, pp. 139–148. Kasper, H.U., Wieler, R., Baroni, C., 2009. Multiple cosmogenic nuclides document Huang, F.X., Liu, X.H., Kong, P., David, F., Ju, Y.T., Fang, A.M., Yu, L.J., Li, X.L., Na, C.G., 2008. complex Pleistocene exposure history of glacial drifts in Terra Nova Bay (northern Fluctuation history of the interior East Antarctic Ice Sheet since mid-Pliocene. Victoria Land, Antarctica). Quatern. Res. 71, 83–92. Antarct. Sci. 20, 197–203. Oberholzer, P., Baroni, C., Schaefer, J.M., Orombelli, G., Il Ochs, S., Kubik, P.W., Baur, H., Ingolfsson, O., Hjort, C., Berkman, P., Bjorck, S., Colhoun, E., Goodwin, I.D., Hall, B., Wieler, R., 2003. Limited Pliocene/Pleistocene glaciation in Deep Freeze Range, Hirakawa, K., Melles, M., Moller, P., Prentice, M., 1998. Antarctic glacial history since northern Victoria Land, Antartica, derived from in situ cosmogenic nuclides. the last glacial maximum. Antarctic Sci. 10, 326–344. Antarct. Sci. 15 (4), 493–502. Jonsson, S., 1988. Observations on physical geography and glacial history of the Quilty, P.G., 1992. Late Neogene sediments of coastal east Antarctica — an overview. In: Vestfjella nunataks in western Droning Maud Land, Antarctica. Naturgeografiska Yoshida, Y. (Ed.), Recent Progress in Antarctic Earth Science. Terra Scientific Ins. Stockholms Univ. Rapport, vol. 68. 57 pp. Publishing Company, Tokyo, pp. 699–705. Kennett, J.P., Hodell, D.A., 1993. Evidence for relative climatic stability of Antarctica Quilty, P.G., Lirio, J.M., Jillett, D., 2000. of the Pliocene Sorsdal Formation, during the early Pliocene: a marine perspective. Geogr. Ann. 75A, 205–220. Marine Plain, Vestfold Hills, East Antarctica. Antarct. Sci. 12 (2), 205–216. Kohl, C.P., Nishiizumi, K., 1992. Chemical isolation of quartz for measurement of in situ- Stone, J.O., 2000. Air pressure and cosmogenic isotope production. J. Geophys. Res. 105, produced cosmogenic nuclides. Geochim. Cosmochim. Acta 56, 3583–3587. 23753–23759. Lal, D., 1991. Cosmic ray labelling of erosion surface: in situ nuclide production rates Strasky, S., Nicola, L.D., Baroni, C., Salvatore, M.C., Baur, H., Kubik, P.W., Schlüchter, C., and erosion models. Earth Planet. Sci. Lett. 104, 424–439. Wieler, R., 2009. Surface exposure ages imply multiple low-amplitude Pleistocene Lewis, A.R., Marchant, D.R., Ashworth, A.C., Hemming, S.R., Machlus, M.L., 2007. Major variations in East Antaectic Ice Sheet, Ricker Hills, Victoria Land. Antarct. Sci. 21 (1), middle Miocene global climate change: evidence from East Antarctica and the 59–69. Transantarctic Mountains. Bull. Geol. Soc. Am. 119 (11/12), 1449–1461. Stroeven, A., Prentice, M., 1997. A case for Sirius Group alpine glaciation at Mount Lewis, A.R., Marchant, D.R., Ashworth, A.C., Hedenäs, L., Hemming, S.R., Johnson, J.V., Fleming, South Victoria Land, Antarctica: a case against Pliocene East Antarctic Ice Leng, M.L., Machlus, M.L., Newton, A.E., Raine, J.I., Willenbring, J.K., Williams, M., Sheet reduction. GSA Bull. 109, 825–840. Wolfe, A.P., 2008. Mid-Miocene cooling and the extinction of tundra in continental Tingey, R.J., 1991. The regional geology of and rocks in Antarctica. Antarctica. PNAS 105 (31), 10676–10680. In: Tingey, R.J. (Ed.), The . Oxford Science Publishers, Oxford, Li, X.L., Liu, X.H., Ju, Y.T., Huang, F.X., 2003. Properties of soils in Grove Mountains, East pp. 1–73. Antarctica. Sci. Chin. (D) 46 (7), 683–693. Whitehead, J.M., 2000. Cenozoic glacial deposits in the Southern Prince Charles Lintinen, P., Nenonen, J., 1997. Glacial history of the Vestfjella and Heimefrontfjella Mountains of East Antarctica. Terra Antarct. 7 (5), 655–656. nunatak ranges in western Dronning Maud Land, Antarctica. In: Ricci, C. (Ed.), The Whitehead, J.M., McKelvey, B.C., 2001. The stratigraphy of the Pliocene-lower Antarctic Region, Geologic Evolution and Processes, Siena Museo Nazionale dell Pleistocene Bard in Bluffs Formation, Amery Oasis, northern Prince Charles Antartica, pp. 845–852. Mountains, Antarctica. Antarct. Sci. 13 (1), 79–86. Liu, X.H., Zhao, Y., Liu, X.C., Yu, L.J., 2003. Geology of the Grove Mountains in East Whitehead, J.M., McKelvey, B.C., 2002. Cenozoic glacigene sedimentation and erosion at the Antarctica—new evidence for the final suture of land. Sci. Chin. D 46 (4), Menzies Range, southern Prince Charles mountains, Antarctica. J. Glaciology 48 (2), 305–319. 207–247. Mabin, M.C.G., 1992. Late Quaternary ice-surface fluctuations of the Lambert Glacier. Wingham, D.J., Siegert, M.J., Shepherd, A., Muir, A.S., 2006. Rapid discharge connects Antarctic Earth Science: Terra Sci. Pub. Com. Tokyo, Japan, pp. 683–687. Antarctic subglacial lakes, Nature, 440. doi:10.1038/nature04660. Mackintosh, A., white, D., Fink, D., Gore, D.B., Pickard, J., Fanning, P.C., 2007. Exposure ages from mountain dipsticks in Mac. Robertson Land, East Antarctica, indicate