Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Cooper, A. K., P. J. Barrett, H. Stagg, B. Storey, E. Stump, W. Wise, and the 10th ISAES editorial team, eds. (2008). Antarctica: A Keystone in a Changing World. Proceedings of the 10th International Symposium on Antarctic Earth Sciences. Washington, DC: The National Academies Press. Landscape Evolution of Antarctica S. S. R. Jamieson and D. E. Sugden1 ABSTRACT shelf before retreating to its present dimensions at ~13.5 Ma. Subsequent changes in ice extent have been forced mainly by The relative roles of ï¬uvial versus glacial processes in shap- sea-level change. Weathering rates of exposed bedrock have ing the landscape of Antarctica have been debated since the been remarkably slow at high elevations around the margin of expeditions of Robert Scott and Ernest Shackleton in the East Antarctica under the hyperarid polar climate of the last early years of the 20th century. Here we build a synthesis of ~13.5 Ma, offering potential for a long quantitative record Antarctic landscape evolution based on the geomorphology of of ice-sheet evolution with techniques such as cosmogenic passive continental margins and former northern mid-latitude isotope analysis. Pleistocene ice sheets. What makes Antarctica so interesting is that the terrestrial landscape retains elements of a record of change that extends back to the Oligocene. Thus there is the INTRODUCTION potential to link conditions on land with those in the oceans Our aim is to synthesize ideas about the evolution of the and atmosphere as the world switched from a greenhouse terrestrial landscapes of Antarctica. There are advantages to to a glacial world and the Antarctic ice sheet evolved to its such a study. First, the landscape evidence appears to extend present state. In common with other continental fragments of back beyond earliest Oligocene times when the ï¬rst ice Gondwana there is a ï¬uvial signature to the landscape in the sheets formed. Thus events on land may potentially be linked form of the coastal erosion surfaces and escarpments, incised with atmospheric and oceanic change as the world switched river valleys, and a continent-wide network of river basins. from a greenhouse to a glacial world and saw the develop- A selective superimposed glacial signature reï¬ects the pres- ment of the Antarctic ice sheet. This helps in establishing ence or absence of ice at the pressure melting point. Earliest correlations or leads and lags between different components continental-scale ice sheets formed around 34 Ma, growing of the Earth system as a way of establishing cause and effect from local ice caps centered on mountain massifs, and fea- in global environmental change. Second, the evidence of tured phases of ice-sheet expansion and contraction. These landscape evolution can be used to reï¬ne models of Earth ice masses were most likely cold-based over uplands and surface processes and their interaction with the wider global warm-based across lowlands and near their margins. For 20 system (e.g., by linking conditions in the terrestrial source million years ice sheets ï¬uctuated on Croll-Milankovitch fre- areas with the marine record of deposition). Third, there are quencies. At ~14 Ma the ice sheet expanded to its maximum analogies in the Northern Hemisphere of similar-size former and deepened a preexisting radial array of troughs selectively Pleistocene ice sheets where the bed is exposed for study. through the coastal mountains and eroded the continental The glaciological body of evidence and theory built on such a base over 150 years is helpful in assessing the nature of the inaccessible topography beneath the Antarctic ice sheet. As such it can illuminate interpretations of subglacial conditions 1 School of GeoSciences, University of Edinburgh, Edinburgh, EH8 9XP, and the dynamics of the present ice sheet. Scotland, UK (Stewart.Jamieson@ed.ac.uk, David.Sugden@ed.ac.uk). Stewart.Jamieson@ed.ac.uk, The crux of any reconstruction of landscape evolution 39
40 ANTARCTICA: A KEYSTONE IN A CHANGING WORLD in a glaciated area is the extent to which ice sheets have and Antarctica took place ï¬rst, while the separation of Aus- transformed a preexisting ï¬uvial topography. For 150 years tralia from Antarctica took place in earnest after 55 Ma. Tas- there has been debate between those highlighting the erosive mania and New Zealand separated from the Ross Sea margins power of ice and those indicating its preservative capacity. In at around 70 Ma. West Antarctica comprises four separate 1848 Charles Lyell, on his way over the formerly glaciated mini-continental blocks: Antarctic Peninsula, Thurston, eastern Grampians of Scotland to receive a knighthood from Marie Byrd Land, and Ellsworth-Whitmore, thought to be Queen Victoria, wrote in his diary, âHere as on Mt Wash- associated with extensional rifting. The drift of the conti- ington and in the White Mountains the decomposing granite nents opened up seaways around Antarctica and changed boulders and the bare surfaces of disintegrating granite are ocean circulation and productivity. A long-held view is that not scored with glacial furrows or polished.â Such observa- this permitted the development of the Antarctic Circumpolar tions subsequently led to the idea of unglaciated enclaves Current, which introduced conditions favorable for glacia- that escaped glaciation completely (e.g., in Britain: Linton, tion (i.e., cooler temperatures and increased precipitation) 1949; and in North America: Ives, 1966). In Antarctica, (Kennett, 1977). In addition, recent climate modeling studies scientists on the early 20th-century expeditions of Scott and have suggested that a reduction in atmospheric greenhouse Shackleton debated the issue, with Taylor (1922) arguing that gases may have played an important role in the triggering the landscape was essentially glacial in origin but cut under of Antarctic glaciation (DeConto and Pollard, 2003; Huber earlier warmer glacial conditions, and Priestley (1909) argu- and Nof, 2006). Critical dates for the development of ocean ing that glaciers had modiï¬ed an existing ï¬uvial landscape. gateways are ~33 Ma, when a signiï¬cant seaway opened up The debate has continued; some argue for dissection of the between Antarctica and Australia (Stickley et al., 2004), and Transantarctic Mountains by glaciers since the Pliocene (van ~31 Ma, when Drake Strait between the Antarctic Peninsula der Wateren et al., 1999) while others point to the important and South America became a signiï¬cant seaway (Lawver role of ï¬uvial erosion at an earlier time (Sugden and Denton, and Gahagan, 2003). 2004). The stepwise glacial history of Antarctica has been This paper contributes to this debate by outlining the pieced together from marine records. Ice sheets ï¬rst built main variables inï¬uencing landscape evolution in Antarc- up at ~34 Ma and their growth was marked by a sudden rise tica and then developing hypotheses about what might be in benthic 18O values (Zachos et al., 1992). On land on the expected from both a ï¬uvial and a glacial perspective and Ross Sea margin of Antarctica, beech forest similar to that the interaction of the two. in Patagonia today gave way to scrub forest and this change was accompanied by a progressive shift in clay minerals from smectite, typical of forest soils, to chlorite and illite WIDER CONTEXT characteristic of polar environments (Raine and Askin, 2001; East Antarctica consists of a central fragment of Gondwana Ehrmann et al., 2005; Barrett, 2007). Glaciation for the next and is surrounded by rifted margins (Figure 1). The initial ~20 million years was marked by ice volume ï¬uctuations breakup of Gondwana around most of East Antarctica took similar in scale to those of the Pleistocene ice sheets of the place between 160 Ma and 118 Ma. The separation of India Northern Hemisphere. These ï¬uctuations are demonstrated Southern Gondwana 00 30 0E 120 Ma Madagascar India 0 W 30 Africa 60 0 E South East Antarctica Australia 60 0W America Ellsworth - 90 E Whitmore 0 Thurston FIGURE 1 The location of Antarctica Tasmania within Gondwana. The reconstruction Antarctic Peninsula Marie Byrd shows the fragmentation of the supercon- New Land Zealand tinent at 120 Ma (modiï¬ed from Lawver 300S 400S 500S 600S 700S 800S 700S 600S 500S 400S 300S et al., 1992).
JAMIESON AND SUGDEN 41 by strata from 34 Ma to 17 Ma cored off the Victoria Land and in response to secondary faults that develop parallel to coast near Cape Roberts (Naish et al., 2001; Dunbar et al., the coast. Observations on other Gondwanan margins sug- forthcoming), high-resolution isotopic records from near gest that most erosion occurs within 10-20 million years Antarctica (Pekar and DeConto, 2006), and by ice-sheet of the rift opening up (Persano et al., 2002). Subsequently, modeling forced by reduced atmospheric CO2 levels and cooling and crustal ï¬exure may cause subsidence and the contemporary orbital insolation changes (DeConto and seaward ends of the valleys are ï¬ooded by the sea. Pollard, 2003). During the same period, the Cape Roberts record shows a progressive decline in meltwater sediments accumulating offshore and a vegetation decline to moss tun- dra, both indicating progressive cooling. A second stepwise A I Main Escarpment I I I I cooling of Paciï¬c surface waters of 6-7Â°C accompanied by I I Initial Fault I I I I a glacial expansion occurred in the mid-Miocene at ~14 Ma I Coastline (new base level) I I I I and is indicated in the marine isotope record (Shevenell et al., I I Antecedent River I I I I I 2004; Holbourn et al., 2005). This event is also recorded in I I I I I I I I I I I I the transition from ash-bearing temperate proglacial deposits Lithological Escarpment Drainage Divide I to diamict from cold ice in the Olympus Range of South I Subsidiary Fault / I Victoria Land at the edge of the South Polar Plateau (Lewis I I I et al., 2007). Geomorphic evidence indicates that the present I Upper Surface I I I I hyperarid polar climate of interior Antarctica was established I I X Y I I I at this time, along with the present structure of polar ocean Rapid Incision I I circulation. Subsequent increases in ice-sheet volume, such I Coastal Plain I I I as occurred in the Quaternary, involved ice thickening at the X Main Escarpment Y coast in response to a lowering of global sea level (Denton Lithological Escarpment Down Throw Cover Rocks et al., 1989). Basement Rocks Uplift (isostatic / thermal) 76o 32'S 76o 32'S idge 163o 12'E 159o 05'E Y ind R FR G 1240 Ea stw B LANDSCAPE EVOLUTION: THE FLUVIAL SIGNAL L A 2214 C y IE le R al V e wl 2270 To Depot Hypothesis C le Sirius vil O group 1720 EVAN S Island NV Va n ey e deposit re I I Flagship Mt. PIEDMONT ll I I I I OY G I I I I I GLACIER 1600 One can predict in qualitative terms the landscape that would 2330 I Mt. Whitcombe Coombs AC IE GL accompany the separation of a long-lived continent such Hills R CA Mt Brooke 2675 Red Butress RIDGE Mt. Black N MB R 76o 50'S Razorback Pudding 76o 50'S O Peak S as Gondwana. There would be integrated continental-scale lle y N Mt. IDG Mt. Gunn Va E a B Brogger 2463 tn ? Ala Mt. E GL Mt. Marston Granite Harbour river networks similar in scale to that of the Orange River in Morrison Mt. Kar u 0 km 10 AC IE Woolnough ea Plat N South Africa and the Murray River in Australia. Presumably LEGEND R Upper surface Mt. Gran 2233 GLAC I E Irregular R these Antarctic rivers would have developed in a semiarid, escarpment front Intermediate surfaces Y ? KA Dissected mountains R IE seasonally wet climate, especially those in the interior of the Valley benches LA C 1205 Mt. C G England Cape A Coastal piedmont 1396 Sperm Bluff N Roberts O M Main escarpment TT supercontinent far from the sea. The rifted margins would I I I II Miller Glacier trough Detour Nunatak 1120 C O Main valleys Gonville e have been subjected to ï¬uvial processes related to the evolu- g and I I I Fault n Caius Range a 1670 I 1490 I I Main divide R I I M IL WILSO N I I tion of passive continental margins. This has been the focus LE I Other mountain divides e I I a r ER I R AC I I I C l I PIEDMONT GL I Generalised outcrop of I I G I I LA Kukri erosion surface 2229 Robison Peak Wheeler 1870 C IE I M GLACIER I Valley A I 2537 R NH of much research in recent years and is well summarized by I 2229 Spotheight (m) I Skew Peak BE I I I I 1343 DE St. John's Range 77o 15'S 77o 15'S Summerï¬eld (2000). Beaumont et al. (2000) developed a coupled surface process-tectonic model of passive continen- FIGURE 2 (A) Typical landscapes of a passive continental margin tal margin evolution at a classic diverging rift margin (Figure as modeled by Beaumont et al. (2000), using a numerical coupled 2A). The main feature is the uplift of the rifted margin inland tectonic-earth surface process model of landscape evolution. The of a bounding fault running parallel to the coast. The uplift is new base level creates a pulse of erosion and associated rock uplift a result of thermal buoyancy and crustal ï¬exure in response that leads to a staircase of coastal erosion surfaces separated by to denudation near the coast and deposition offshore. Erosion escarpments and incised river valleys, some of which may maintain is stimulated by the steep surface gradients created by the their valleys across the escarpment crest. (B) Geomorphic map of the Convoy Range and Mackay Glacier area, showing the staircase lower base level as the rift opens up. Mass wasting and riv- of seaward-facing escarpments, erosion surfaces, dissected moun- ers with steep gradients carve a coastal lowland and valleys tain landscapes near the coast, and the dendritic valley patterns into the upland rim. A seaward-facing escarpment forms at radiating from a high point in the Coombs Hills. Principal faults the drainage divide. Preexisting rivers may keep pace with parallel to and at right angles to the mountain front are taken from the uplift and traverse the escarpment and allow dissection Fitzgerald (1992). to spread behind the escarpment. Subsidiary erosion surfaces SOURCE: Sugden and Denton (2004). The model in (A) simulates may form in response to lithological contrasts in rock type the ï¬eld evidence in remarkable detail.
42 ANTARCTICA: A KEYSTONE IN A CHANGING WORLD The model applies to idealized young rift margins, and such as the western Weddell Sea, Lambert, Wilkes Land, and there are often more complex relationships affecting other Oates Land basins. The dendritic pattern of the network, the continental margins of Gondwana (Bishop and Goldrick, centripetal pattern of ï¬ow from the subglacial Gamburtsev 2000). Factors such as time elapsed since rifting, proximity Mountains, and the topological coherence of the tributaries to the rift, inherited altitude of the margin, tectonic down- are demonstrated by bifurcation ratios that are representa- warping, and crustal ï¬exure can all affect the amplitude of tive of other river basins, such as the Orange River in South the topography and the evolution of any escarpment. Africa (Jamieson et al., 2005). These observations imply a ï¬uvial signature in the subglacial landscape on a continental scale. Antarctic Evidence Such a conclusion is reinforced by investigation of the Figure 3 shows a reconstruction of the continental river pat- valley patterns on mountain areas rising above the present terns inferred to exist if the Antarctic ice sheet is removed and ice-sheet surface. Good examples of former ï¬uvial valley the land compensated for isostatic depression by full ï¬exural systems now occupied by local glaciers exist in the Trans- rebound. The map is based on the BEDMAP reconstruction antarctic Mountains. The valley networks of several basins of subglacial topography with a nominal resolution of 5 km in northern Victoria Land have been ordered according to (Figure 4) (Lythe et al., 2001). Sea level is assumed to lie ï¬uvially based Horton-Strahler rules (Horton, 1945; Strahler, at â100 m to represent the subsidence associated with the 1958) and the dendritic pattern and hierarchical relationships crustal cooling and ï¬exure of mature passive continental between valley segments, and variables such as cumulative margins. The reconstruction is based on hydrological model- mean length are typical of river networks (Baroni et al., ing methods for cell-based digital elevation models whereby 2005). The Royal Society Range shows a similarly dendritic water is deemed to ï¬ow to the lowest adjacent cell. There are network radiating from one of the highest summits in Ant- many uncertainties, not the least of which is that large areas arctica at over 4000 m (Sugden et al., 1999). have little data, but a sensitivity analysis suggests that models The landforms near the coast of the Antarctic passive forced by a range of different assumptions yield a network continental margin are well displayed in the relatively gla- that is essentially similar from run to run (Jamieson et al., cier-free area of the Dry Valleys and adjacent Convoy Range 2005). The reconstruction shows that there are integrated (Figure 2B). Many features expected of a ï¬uvial landscape networks leading radially to major depressions at the coast, are present. These include erosion surfaces rising inland from 40Â°W 30Â°W 20Â°W 10Â°W 0Â° 10Â°E 20Â°E 30Â°E 40Â°E Dronning 60Â°S Maud Land Weddell Sea Shackleton Lambert Range Glacier Gamburtsev Mts T FIGURE 3 Reconstruction of the Wilkes continental-scale river patterns beneath the Land present ice sheet. We assume that the land is isostatically compensated and that sea Dry Valleys Oates level is 100 m lower than today. The model 60Â°S Land uses the BEDMAP subglacial topography at a nominal cellular scale of 5 km (Figure 4). BEDMAP data are from Lythe et al. (2001). 140Â°W 150Â°W 160Â°W 170Â°W 180Â° 170Â°E 160Â°E 150Â°E 140Â°E
JAMIESON AND SUGDEN 43 FIGURE 4 The present-day subglacial topography of Antarctica. BEDMAP data are from Lythe et al. (2001). the coast and separated by escarpments, a coastal piedmont, ï¬ssion track analyses suggest that most denudation occurred an undulating upper erosion surface dotted with 100-450 m shortly after 55 Ma. inselbergs, a seaward-facing escarpment 1000-1500 m in The evidence presented above is powerful afï¬rmation height, and intermediate erosion surfaces delimited by that the normal ï¬uvial processes of the erosion of a pas- lithological variations, notably near horizontal dolerite sills. sive continental margin explain the main landscape features Sinuous valleys with a dendritic pattern can also be found. of three mountain blocks of the Transantarctic Mountains Most run from the escarpment to the sea, but others, such as extending over a distance of 260 km. In these cases it is that occupied by Mackay Glacier, breach the escarpment and the fault pattern and the position of the drainage divide that drain the interior through tributaries bounded by additional are the main controls on topography and determine the dif- escarpments. In detail the valleys have a sinuous planform, ferences in the landscapes of each block. It is not possible rectilinear valley sides with angles of 26-36Â°, and often to apply the model of passive continental margin evolution lower-angle pediment slopes in the valley ï¬oors (Figure 5). to other sectors of East Antarctica without more detailed The ï¬oor of the southernmost Dry Valley, Taylor Valley, is evidence, but similar landscapes occur in the Shackleton below sea level toward the coast, where it is ï¬lled with over Range (Kerr and Hermichen, 1999) and in Dronning Maud 300 m of sediments of late-Miocene to recent age (Webb Land (NÃ¤slund, 2001). One exception is the Bunger Hills and and Wrenn, 1982). Vestfold Hills sector of East Antarctica east of the Lambert The erosion of the coastal margin has been accompanied Glacier, where there is no escarpment. The matching piece by the denudation of a wedge of rock thickest at the coast and of Gondwana is India, which separated early, and a longer declining inland. This is displayed in the coastal upwarping and more complex evolution of the rift margin may explain of basement rocks and in the denudation history indicated the lack of an escarpment today. by apatite ï¬ssion track analysis. Both lines of evidence agree It seems reasonable to argue that the framework of and point to the denudation since rifting of 4.5-5 km of rock passive continental margin evolution applies to Antarctica at the coast falling to ~1 km of rock at locations 100 km and that in many areas a major pulse of ï¬uvial erosion and inland (Fitzgerald, 1992; Sugden and Denton, 2004). The accompanying uplift was a response to new lower base levels
44 ANTARCTICA: A KEYSTONE IN A CHANGING WORLD FIGURE 5 George Denton and David Marchant in Victoria Valley, Dry Valleys, a typically ï¬uvial landscape with rectilinear slopes and a shallow pediment slope leading to the valley axis. The slopes have escaped modiï¬cation by overriding ice. following continental breakup. Since breakup took place at superposition of a continental-scale radial ï¬ow pattern on the different times, the stage of landscape evolution will vary underlying topography. Such patterns are easily obscured by from place to place in Antarctica. Furthermore, where large- local signals but one can identify the following: scale tectonic features such as the Lambert graben disrupt the continental margin, they are likely to focus the ï¬uvial system Erosion in the center and wedges of deposition in a distinctive way (Jamieson et al., 2005). beneath and around the peripheries (Sugden, 1977; Boulton, 1996); Continental shelves that are deeper near the conti- LANDSCAPE EVOLUTION: THE GLACIAL SIGNAL nent and shallower offshore as a result of erosion near the coast, often at the junction between basement and sedimen- Hypothesis tary rocks (Holtedahl, 1958); The beds of former Northern Hemisphere mid-latitude ice Radial pattern of large 10-km-scale troughs that sheets form the basis of understanding how glaciers trans- breach and dissect the drainage divides near the coast and form preexisting landscapes. There are differences in that the may continue offshore (e.g., in Norway, Greenland, and Baf- Antarctic ice sheet has existed for tens of millions of years ï¬n Island) (LÃ¸ken and Hodgson, 1971; Holtedahl, 1967); rather than a few million, but there are important similari- Radial pattern of ice streams with beds tens of km ties in that the North American ice sheet was of similar size wide, streamlined bedforms in bedrock and drift, and sharply and volume as the Antarctic ice sheet, and in that both have deï¬ned boundaries (Stokes and Clark, 1999); and ï¬uctuated in size in response to orbital forcing during their Radial pattern of meltwater ï¬ow crossing regional evolution, albeit to varying extents. interï¬uves, as revealed, for example, by the pattern of eskers At the outset it is helpful to distinguish two scales of in North America (Prest, 1970). feature: Those that reï¬ect the integrated radial ï¬ow of the ice sheet at a continental scale when it is close to or at its Local and regional patterns also display radial conï¬gu- maximum and those that reï¬ect the local and regional ï¬ow rations and reï¬ect multiple episodes of reduced glaciation. patterns as ice ï¬ows radially from topographic highs. Distinguishing features of marginal glaciation are corries or The key to landscape change by large ice sheets is the cirques, the dominant orientation of which, northeast facing,
JAMIESON AND SUGDEN 45 is determined in the Northern Hemisphere by slopes shaded maximum coastal precipitation at sea level is scaled up to 2 from the sun and subject to wind drift by prevailing westerly m per year, four times that of the present day. The maximum winds (Evans, 1969). Stronger local glaciation typically then falls linearly to 0.5 m per year by the end of the model builds ice caps on mountain massifs with ice ï¬ow carving run. Mean annual temperatures fall from 7Â°C at sea level to a radial pattern of troughs, so well displayed in the English present values of â12Â°C through the model run. Melt rates Lake District and Scotland, for example. under warmer climatic conditions are calculated using a Clearly there will be a complex interaction between positive degree-day model (Reeh, 1991) whereby ablation local, regional, and continental modes of ï¬ow depending on is proportional to the number of days where temperature is such factors as climate, ice extent, and topographic geometry. above the freezing point. Diurnal variability is accounted We attempt to model this complexity in Figure 6 by showing for by using a normal distribution of temperature with a 5Â°C various stages of evolution of the Antarctic ice sheet using standard deviation. The pattern of mass balance used to drive GLIMMER, a three-dimensional thermomechanical ice- ice growth is shown in Figure 7. sheet model as described by Payne (1999) and Jamieson et The bed topography is derived from BEDMAP (Figure al. (forthcoming). The intention is to illustrate the range of 4) (Lythe et al., 2001) and is ï¬exurally rebounded to com- different ice-sheet geometries that would be expected at vari- pensate for the lack of an ice sheet. The use of an isostati- ous stages of Croll-Milankovitch glacial-deglacial cycles. cally compensated present-day topography ignores tectonic The model is run for an arbitrary 1 million years with stepped movements and means that the results of the modeling temperature changes every 50 kyr falling from present-day become more uncertain as one goes further back in time. Patagonian values to present-day Antarctic values. This tim- However, there is less risk in East Antarctica, where the main escale is designed to allow the ice to achieve approximate topographic features were established by Oligocene times. equilibrium at all times and to allow the isostatic response For example, geological evidence in the form of basement of the bedrock to reach a balance with these ï¬uctuations clasts in Oligocene strata cored off the Victoria Land coast in ice thickness. Patagonian climate statistics are used to (CIROS-1 drillcore) (Barrett et al., 1989) and Cape Roberts simulate the climate inferred from vegetation associated with (Cape Roberts Science Team, 2000) suggests that the Trans- the initial glaciation of Antarctica (Cape Roberts Science antarctic Mountains had been eroded deep enough to form Team, 2000; Raine and Askin, 2001). Modeled precipitation a signiï¬cant feature by Oligocene times. Furthermore, the follows the pattern of net surface mass balance derived by Gamburtsev Mountains are considered to be a Pan-African Vaughan et al. (1999). At the beginning of the model run, feature with an age of 500 Ma (van de Flierdt et al., 2007). A B FIGURE 6 Model of the Antarctic ice sheet, generated using the GLIMMER 3-D thermomechanical model and the stepped transition from a Patagonian-style climate C D to the present polar climate. The four stages (A-D) illustrate the range of variability to Modeled Ice Thickness (m) be expected as the Antarctic ice sheet ex- perienced many Croll-Milankovitch glacial 0 1000 2000 3000 4000 5000 cycles during its early evolution.
46 ANTARCTICA: A KEYSTONE IN A CHANGING WORLD 40Â°W 30Â°W 20Â°W 10Â°W 0Â° 10Â°E 20Â°E 30Â°E 40Â°E 60Â°S Y X 60Â°S 140Â°W 150Â°W 160Â°W 170Â°W 180Â° 170Â°E 160Â°E 150Â°E 140Â°E FIGURE 7 Present-day accumulation is used to drive a simulated Antarctic ice 0 0.1 0.2 0.3 0.4 0.5 sheet (Vaughan et al., 1999). Proï¬le X-Y Accumulation (m/yr) shows that under a Patagonian-style regime Mass Balance (m/yr) X Y (grey line) there are zones of increased ac- 1 cumulation at high altitudes. Accumulation is discontinuous because of high levels of 0 ablation across much of the continental lowlands. The present-day accumulation is -1 0 1000 2000 3000 4000 shown by the black line. Distance (km) The different stages of growth illustrate the principal paths before being delivered to the coastal margin by the pattern of glaciation of Antarctica. Initial growth is in coastal continental ice sheet. mountains, such as in Dronning Maud Land, along the Trans- antarctic Mountains, in the West Antarctic archipelago, and Antarctic Evidence in the high Gamburtsev Mountains in the interior. The ice spreads out from these mountain centers, ï¬rst linking the The evidence of continental radial patterns of erosion is main East Antarctic centers and then the West Antarctic cen- spectacular. The Lambert trough, which is 40-50 km across ters. The model is deliberately simple but it sufï¬ces to show and 1 km deep, drains 10 percent of the East Antarctic ice that glaciation starts preferentially in maritime mountains sheet. It is coincident with a graben and is comparable to, and in interior mountains if they are high enough. It also but deeper and longer than, the North American equivalents, serves to illustrate the complexity of the changing pattern such as Frobisher Bay in Bafï¬n Island. And then there is the of ï¬ow as different ice centers merge and ice ï¬ow evolves spectacular series of troughs cutting through the Transant- from locally radial to continentally radial. The subglacial arctic Mountain rim. Webb (1994) has previously suggested landscape of Antarctica can be expected to consist of a that the Beardmore trough, 200 km long, 15-45 km wide, palimpsest of landforms related to these local, regional, and and over 1200 m deep, exploited a preexisting river valley. continental stages, while eroded material will experience a Unloading due to glacial erosion may have contributed to complex history of temporary deposition and changing ï¬ow isostatic uplift of the adjacent mountains (Stern et al., 2005). Offshore there are continuations of such troughs incised
JAMIESON AND SUGDEN 47 into the continental shelf of both East and West Antarctica Sorlien et al., 2007). In these latter cases the troughs are (Wellner et al., 2001). Other troughs, such as that running revealed by seismic survey. In the McMurdo Dry Valleys area parallel to Adelaide Island on the Antarctic Peninsula, have of the Transantarctic Mountains a phase of local glaciation is exploited the junction between basement and sedimentary represented by troughs identiï¬ed on the inland ï¬ank of the rocks (Anderson, 1999). A series of ice streams ï¬ow into the mountains (Drewry, 1982) and troughs exploiting sinuous Ross Sea and Weddell Sea embayments. They are underlain river valleys, such as the Mackay Glacier. Finally, the com- by deformable till and, in the case of the Rutford ice stream, pact wet-based glacial deposits (the Sirius Group deposits) by streamlined bedforms (Smith et al., 2007). There is also distributed at high elevations along 1000 km of the Transant- growing evidence of radial outï¬ow of basal meltwater (Evatt arctic Mountains in the Ross Sea sector represent glaciation et al., 2006). Hundred-meter-deep rock channels and massive centered on the mountains (Denton et al., 1991, 1993). The staircases of giant potholes represent large-scale outbursts of deposits are characterized by local lithologies and the till subglacial meltwater across the Transantarctic Mountains components contain striated stones and a matrix typical of rim in the McMurdo area (Denton and Sugden, 2005; Lewis glacial erosion under warm-based ice. Some of these deposits et al., 2006). incorporate remains of Nothofagus (southern beech) forest There is also evidence of local and regional glacial representative of a cool temperate environment. landforms. Early studies of the subglacial Gamburtsev Mountains revealed characteristic trough overdeepening and LANDSCAPE EVOLUTION: THE COMBINED SIGNAL the presence of hanging valleys, pointing to a local glacia- tion (Figure 8). The glacial landscapes of Dronning Maud Hypotheses Land too were created by local mountain glaciation and not the present ice sheet (Holmlund and NÃ¤slund, 1994). Recent It is possible to relate subglacial landscapes to the processes geophysical surveying has demonstrated the presence of by which glaciers modify preexisting topography. A simple lakes in overdeepened troughs radiating from the Ellsworth classiï¬cation scheme recognizes landscapes of areal scour- Mountain core (Siegert, pers. comm., 2007). Overdeepen- ing with abundant evidence of glacial scour; those of selec- ing in the topographically constrained part of a fjord, and tive linear erosion where troughs dissect plateaus; those with the rock threshold at the point when the trough opens out, no sign of glacial erosion and devoid of glacial landforms; are well-known characteristics of fjords in the Northern and depositional landscapes composed of till and meltwa- Hemisphere (LÃ¸ken and Hodgson, 1971; Holtedahl, 1967). ter deposits (Sugden, 1978). The differences are related Similar features are found to radiate from uplands now below to whether the basal ice is at the pressure melting point. A sea level in the Ross Sea embayment (De Santis et al., 1995; key assumption is that ice erodes effectively when the base Valley Bench Hanging Valley FIGURE 8 An early reconstruction of the landscape of the subglacial Gamburtsev Glacial Trough Glacial Trough Mountains based on the analysis of radio- echo sounding data. The arrows pick out diagnostic glacial features such as overdeep- ening and hanging valleys presumed to have been formed by local mountain glaciation (modiï¬ed from Perkins, 1984).
48 ANTARCTICA: A KEYSTONE IN A CHANGING WORLD is at the pressure melting point because sliding takes place ice discharge is highest and generating most internal heat. between the ice and bedrock, permitting several processes These will be the areas of glacial erosion. The inï¬uence of to entrain bedrock and to deposit material. Such a situation topography is clear in that warm-based ice is focused on the explains landscapes of areal scouring, the linear erosion of depressions and major valleys where ice is thicker and ï¬ows troughs, and zones of deposition. The converse is that when faster. These low-lying areas at the pressure melting point are the basal temperature is below the pressure melting point also those in which subglacial lakes might accumulate. What there is no sliding at the ice and rock interface. In such situ- is striking is the way this zone of peripheral warm-based ations ice can be essentially protective and leave little sign ice and erosion is a wave that sweeps across the landscape of erosion. There is debate as to how protective the ice may as the ice sheet grows to its continental maximum. At the be (Cuffey et al., 2000), but recent work on cosmogenic macroscale the zone of warm-based ice is more extensive isotope analysis demonstrating the age of exposure and the around the continental margins, especially in the vicinity time buried beneath ice has shown that the hypothesis holds of the main preexisting drainage basin outlets. Under both in many areas of the Northern Hemisphere (Briner et al., scenarios the ice over the main uplands remains below the 2006; Stroeven et al. 2002). pressure melting point. Armed with these observations it is possible to hypoth- From the above we can predict that the landscape in esize about the landscape beneath the Antarctic ice sheet lowland areas of the Antarctic ice sheet will be underlain by a and suggest local, regional, and continental patterns. The landscape of areal scouring. This relates both to the presence numerical model of ice-sheet growth can be used to predict of ice at the pressure melting point and to the progressive the changing pattern of glacial erosion during glacial cycles. erosion of rock debris by radial outï¬ow of ice at different Figure 9 shows the distribution of basal ice at the pressure scales. Interaction between local and ice-sheet maximum melting point during stages of regional and continental gla- ï¬ow directions and rock structure will determine the rough- ciation. The regional pattern shows how the peripheries of ness and degree of streamlining of landforms. Flow in the the regional ice sheets are warm-based near their margins same direction under both local and continental modes will and inland to the vicinity of the equilibrium line where the favor elongated streamlined bedforms, perhaps with plucked B D Modeled Erosion Pattern Zero Max FIGURE 9 Modeled distribution of basal ice at the pressure melting point during intermediate and full stages of Antarctic glaciation (letters correspond to snapshots in Figure 6). Erosion, thought to be associated with sliding under such basal conditions, is concentrated toward the ice-sheet margins and at the beds of major outlet glaciers. A wave of erosion accompanies the expansion of local and regional ice to a full continental ice sheet.
JAMIESON AND SUGDEN 49 lee slopes, while complex ï¬ow changes will leave an irregu- Land (Sugden and Denton, 2004), and the plateaus of the lar pattern. Areal scouring will be clearest on the lowlands Antarctic Peninsula area (Linton, 1963). As in the Northern and diminish upslope on upstanding massifs where summits Hemisphere such a description also applies on the scale of may show no sign of glacial modiï¬cation. This latter pat- individual massifs. For example, upstanding nunataks in the tern reï¬ects the effects of topography on the basal thermal Sarnoff Mountains of Marie Byrd Land are bounded by lower regime and implies that the ice remained cold-based during slopes with clear evidence of glacial scouring and yet their local, regional, and continental stages of glaciation. Judging summits have retained an upper surface with tors, weathering by the geomorphology of glaciated shields of the Northern pits, and block ï¬elds. In this case cosmogenic isotope analy- Hemisphere, erosion will have removed some tens of meters sis demonstrates that the mountains have been covered by ice of material, much of it initially as weathered regolith. This is of several glacial maxima and that weathering has continued sufï¬cient to modify but not erase the preexisting river land- sporadically in interglacials for ~1 million years (Sugden et scape, as argued for northern Europe (Lidmar-Bergstrom, al., 2005). The selectivity reï¬ects the difference between 1982). However, given the longer duration of glaciation in the thicker, converging ice that scours as it ï¬ows round the Antarctica one would expect a greater depth to have been massif and the thin diverging ice covering the summit that removed. The products of this erosion will be deposited remains cold-based during each episode of overriding. offshore. The implication of the above is that the hypothesis relating glacial modiï¬cation of a preexisting landscape to the presence or absence of warm-based ice may be helpful Antarctic Evidence in describing and understanding the landscape evolution There is no direct evidence of areally scoured landscapes of Antarctica. When the basal thermal regime remains the beneath the ice sheet. However, there are many observa- same beneath both local and full ice-sheet conditions, then tions from around the margins of Antarctica to indicate that the glacial transformation, or lack of it, will be clearest. such landscapes are likely to be widespread. Areas formerly What is exciting about the present time is that cosmogenic covered by an earlier expanded ice sheet display extensive isotope analysis offers the opportunity to quantify such landscapes of areal scouring. This includes oases in East relationships. Antarctica, such as the Amery Oasis bordering the Lambert Glacier (Hambrey et al., 2007); the Ross Sea area where CHRONOLOGY OF LANDSCAPE EVOLUTION the scouring is most prominent near sea level and can be traced to elevations of 1000-2100 m along the Transantarctic A number of dates help ï¬rm up the relative chronology of Mountains front (Denton and Sugden, 2005); the inner parts landscape evolution. Early studies of offshore sediments of the offshore shelf in many parts of West Antarctica; and in Prydz Bay established the presence of ï¬uvial sediments the offshore shelves surrounding islands off the Antarctic below the earliest glacial sediments, the latter dated to ~34 Peninsula and sub-Antarctic islands (Anderson, 1999). In Ma (Cooper et al., 1991; OâBrien et al., 2001). The location, all these situations meltwater channels testify to the activity structure, and nature of the sediments suggest that the depos- of basal meltwater. Whereas all these observations are con- its have been derived from rivers ï¬owing along the Lambert sistent with the view that warm-based ice occurred beneath graben. Probably the deposits began to accumulate when rift thicker parts of the ice sheet when it was more extended and separation began around 118 Ma (Jamieson et al., 2005). in maritime environments, what is surprising is that striated, Some large glacial troughs were cut by late Miocene scoured rock surfaces also deï¬ne trimlines around mountains times. In the Lambert Glacier area the offshore sedimen- protruding above the ice sheet, even in the interior, such as tary evidence points to an ice sheet that discharged from a the Ellsworth Mountains (Denton et al., 1992). The implica- broad front on the offshore shelf until the late Miocene but tion of shallow surface ice at the pressure melting point is experienced a switch to deposition within the overdeepened that surface climatic conditions must have been within a few Lambert glacial trough subsequently. Ice-sheet model- degrees of freezing point and thus several tens of degrees ing suggests that the deepening of the trough changed the warmer than at present. dynamics of glacier ï¬ow and ablation to such an extent that Landscapes of selective linear erosion are common calving velocity could match ice velocity and that the glacier in the mountainous rim of East Antarctica. In many areas was no longer able to advance through the deep water of glacial troughs are clearly delimited and excavated into the trough (Taylor et al., 2004). The implication is that the a landscape-preserving ï¬uvial valley form, often bearing trough was excavated deeply by the late Miocene. Similar diagnostic subaerial weathering forms such as regolith relationships occur in the glacially deepened mouths of the and tors. Such a description would apply to the landscapes Dry Valleys in the McMurdo Sound area. Microfossils at bordering the Lambert Glacier (Hambrey et al., 2007), the the bottom of the Dry Valleys Drilling Project (DVDP-11) Shackleton Mountains (Kerr and Hermichen, 1999), exposed drillcore at the mouth of Taylor Valley are late Miocene in escarpments in Dronning Maud Land (NÃ¤slund, 2001), the age (Webb and Wrenn, 1982). Marine shells deposited in a mountain blocks of the Transantarctic Mountains in Victoria fjord in the glacial trough of Wright Valley are Pliocene in
50 ANTARCTICA: A KEYSTONE IN A CHANGING WORLD age and overlie a till of >13.6 Ma (Hall et al., 1993). These cold-based local glaciation has been reported from the Olym- age relationships in glacial troughs related to ice ï¬ow from pus Range in the McMurdo Sound area of the Transantarctic the interior of Antarctica demonstrate that they were cut by Mountains (Lewis et al., 2007). Here a classic warm-based the late Miocene. till with meltout facies is overlain by weathered colluvium There is evidence from the McMurdo Sound area that that is itself overlain by tills deposited by cold-based glaciers. the Antarctic ice sheet overrode marginal mountains and The minimum date of transition is ï¬xed by volcanic ashes extended to the outer shelf at its maximum in the mid-Mio- interbedded between the two sets of tills and has an age of cene. The case is argued out in a series of detailed papers in 13.94 Ma. Such a transition beneath small local glaciers the Dry Valleys area, the key chronological ï¬xes of which is argued to represent an atmospheric cooling of 20-25Â°C. are: Moreover, the transition occurs before one or more major ice-sheet overriding events in the same area. There is clear evidence that ice overrode all except perhaps the highest mountains in the Royal Society Range DISCUSSION in the form of ice scouring on cols and subglacial meltwater channels that cross the mountains. The direction of ï¬ow is Here we attempt a synthesis of landscape evolution of conformable with models of ice expansion to the outer edge Antarctica, based mainly on terrestrial evidence (Table 1). of the offshore shelf (Sugden and Denton, 2004). Inevitably the hypothesis is based on partial information and 40 Ar/39Ar dating of in situ volcanic ash deposits is biased toward the data-rich McMurdo Sound area of the overrun by such ice and those ashes deposited on till sheets Transantarctic Mountains. Nevertheless it seems helpful to associated with such overriding ice constrains the event to try and generalize more broadly. between 14.8 Ma and 13.6 Ma (Marchant et al., 1993). An early pulse of ï¬uvial erosion was associated with Landscapes of areal scouring molded by the maxi- the breakup of Gondwana and the creation of new lower mum ice sheet in front of the Royal Society are older than base levels around the separating continental fragments. The 12.4 3 Â± 0.22 Ma, the age of the oldest undisturbed volcanic timing of the pulse varied with the time of base level change cone known to have erupted onto the land surface (Sugden around each segment. Erosion removed a wedge of material et al., 1999). from around the margins of East Antarctica and the smaller 40 Ar/39Ar analyses of tephra show that the major continental fragments of West Antarctica. Escarpments meltwater feature represented by the Labyrinth in Wright Valley predates 12.4 Ma and that the last major outburst occurred some time between 14.4 Ma and 12.4 Ma (Lewis et al., 2006). TABLE 1 The Chronology of Landscape Evolution in 3 He ages of individual dolerite clasts in meltwater Antarctica Based Mainly on Terrestrial Evidence deposits from the overriding ice sheet reveal exposure ages >55-34 Ma Passive continental margin erosion of coastal between 8.63 Â± 0.09 and 10.40 Â± 0.04 Ma. Allowing for ero- surfaces, escarpments, and river valleys, removing sion, they are calculated to have been exposed for ~13 Ma 4-7 km of rock at the coast and 1 km inland since (Margerison et al., 2005). rifting. Cool temperate forest and smectite-rich soils, at least at coast. As yet there are few comparable terrestrial dates else- 34 Ma Initial glaciation of regional uplands with widespread warm-based ice, local radial troughs, where in Antarctica, but it is worth drawing attention to and tills. Climate cooling. work on the ï¬anks of the Lambert Glacier in which there is 34-14 Ma Local, regional, and continental orbital ice- biostratigraphical evidence of a pre-late Miocene phase of sheet ï¬uctuations associated with progressive glacial erosion and deepening, followed by a 10-million-year cooling, declining meltwater, and change to period of exposure (Hambrey et al., 2007). It is tempting to tundra vegetation. Local warm-based glaciers in mountains. equate the deepening to the same ice-sheet maximum. ~14 Ma Expansion of maximum Antarctic ice sheet to edge In the McMurdo Sound area it is possible to establish of continental shelf linked to sharp temperature that a phase of warm-based glaciation occurred before the decline of 20-25Â°C. Change from warm-based mid-Miocene overriding ice. The critical evidence is that to cold-based local mountain glaciers. Selective warm-based tills in the high mountains bounding the Dry erosion of continental-scale radial and offshore glacial troughs and meltwater routes. Valleys, and indeed the Sirius Group deposits, have been 13.6 Ma to Present Ice sheet maintains hyperarid polar climate. Slight modiï¬ed by overriding ice. Typically there are erosional thickening of ice-sheet margins during Pliocene patches excavated into a preexisting till with material warming in East Antarctica. Outlet glaciers respond dragged out down-ice (Marchant et al., 1993) and ripple to sea-level change, especially in West Antarctica. corrugations with a spacing of 25-50 m that are a coherent Extremely low rates of subaerial weathering. Glacial erosion restricted to outlet glaciers and part of the overriding meltwater system (Denton and Sugden, beneath thick ice. 2005). One important ï¬x on the switch from warm-based to
JAMIESON AND SUGDEN 51 and erosion surfaces formed and were dissected to varying its associated polar climate is demonstrated on land by the degrees by ï¬uvial erosion. The degree of dissection increased remarkably low erosion rates in the Transantarctic Mountains with the complex geometry and small size of each fragment as revealed by 39Ar/40Ar dating of volcanic ashes and cones, and was more pronounced near the coast. The interior of East by cosmogenic isotope analysis, and by the preservation of Antarctica was characterized by large river basins, presum- fragile deposits and buried ice (Brook et al., 1995; Ivy-Ochs ably more arid in the interior than at the coast. The climate et al., 1995; Summerï¬eld et al., 1999; Marchant et al., 2002). was sufï¬ciently warm to support beech forests around the Offshore the growth and decay of the maximum ice sheet coast. Soils contained the clay mineral smectite, derived from is demonstrated by a widespread unconformity and dating the chemical weathering associated with forests. evidence of a retreat of the ice in the Ross Sea area after 13.5 Around 34 Ma declining atmospheric carbon dioxide Ma (Anderson, 1999). and the opening of signiï¬cant seaways between Antarctica The implication of the above is that ï¬uctuations of and the southern continents were factors in bringing the two the Antarctic ice sheet in the Pleistocene are forced by conditions necessary for glaciers: cooling and increased changes in sea level. In East Antarctica the ï¬uctuations precipitation from circumpolar storms. Glaciation began are relatively minor. Outlet glaciers thicken and advance in a Patagonian-type climate, at least in the Transantarctic seaward in response to a lowering of sea level in the North- Mountains, and was centered on maritime mountains of East ern Hemisphere, as demonstrated in the case of the outlets and West Antarctica and high continental mountains in East ï¬owing through the Transantarctic Mountains (Denton et Antarctica. The record from the Cape Roberts cores of ï¬uctu- al., 1989). Slight thickening occurs in Mac Robertson Land ating ice-sheet extent, which is supported by marine oxygen (Mackintosh et al., 2007), but the ice does not extend far isotope records, points to a dynamic ice sheet responding offshore (OâBrien et al., 2001; Leventer et al., 2006). In to orbital ï¬uctuations in the same way as the Pleistocene agreement with such a limited expansion, the coastal oases ice sheets of the Northern Hemisphere. The preexisting of the Bunger Hills and the Larsemann Hills appear to have regolith was progressively removed from the continent to remained ice-free during the last glacial cycle (Gore et al., create a subglacial landscape of areal scouring, probably in 2001; Hodgson et al., 2001). In West Antarctica the Pleisto- a complex series of ï¬ows as glacier extent and ï¬ow direc- cene behavior is markedly different. Here ice appears to have tions oscillated between an interglacial and ice-maximum extended to the edge of the continental shelf and occupied state. Modeling suggests a wave of erosion was associated deep troughs extending ~100 km from the present coast with each expansion of ice from the mountain centers. The (Bentley and Anderson, 1998; Ã Cofaigh et al., 2005). In presence of warm-based local glaciers in the Transantarctic this case and following Mercer (1978), one can surmise that Mountains suggests relatively warm interglacial periods. the grounded ice streams occupying the topography below The terrestrial record agrees with the marine record in sea level between the individual massifs are especially sus- pinpointing a sharp temperature decline associated with the ceptible to sea-level changes. expansion of the Antarctic ice sheet over its continental shelf It is interesting that the ice sheet achieved its present at ~14 Ma. Perhaps the expansion was triggered by a change proï¬le in Mac Robertson Land 6000 years ago (Mackintosh in ocean circulation or declining atmospheric carbon dioxide. et al., 2007). This is the time when global sea level had Perhaps it too could have been related to the internal dynam- largely completed its recovery following the ï¬nal disap- ics of the ice sheet in that earlier glaciations had deposited pearance of the North American ice sheet. The coincidence shoals on the offshore shelf, reduced calving, and allowed supports the view that Antarctic ice ï¬uctuations in the Pleis- the ice to advance to the outer edge, behavior well known in tocene are a response to sea-level changes driven primarily the case of fjord glaciers (Mercer, 1961). by the Northern Hemisphere ice sheets. The mid-Miocene maximum ice sheet eroded troughs on a continental-scale, cutting selectively through the mountain WIDER IMPLICATIONS rim, deepening the interior parts of the offshore shelf (and subsea basins?) in West Antarctica. Land surfaces covered by The thrust of this overview is that information about the thin diverging ice remained essentially unchanged. Perhaps evolution of passive continental margins and the processes the offshore deepening was such that in cycles of growth and and forms associated with mid-latitude Northern Hemisphere decay the ice could no longer advance to the shelf edge, as ice sheets is a useful guide for reconstructing the evolution of demonstrated by the behavior of the Lambert Glacier. Alter- the landscape in Antarctica. At present there are insufï¬cient natively the change in ocean and atmospheric conditions constraints to do more than outline possibilities in a quali- after the mid-Miocene maximum to a hyperarid polar climate tative way. Nonetheless even a preliminary view provides may have deprived the ice sheet of moisture. But after ~13.6 some insights into the debate about the relative importance Ma the ice retreated to its present continental lair, at least of ï¬uvial and glacial agents of erosion. Further, there seems in East Antarctica, and remained essentially intact. Coastal ample scope for a focused modeling exercise increasingly fjords were ï¬lled with shallow marine Miocene sediments. founded on quantitative ï¬eld data. The retreat and subsequent stability of the full ice sheet and What is also encouraging about the emerging terrestrial
52 ANTARCTICA: A KEYSTONE IN A CHANGING WORLD record of landscape evolution is that the main stages match Beaumont, C., H. Kooi, and S. Willett. 2000. Coupled tectonic-surface pro- the records obtained from deep-sea and inshore cores. At cess models with applications to rifted margins and collisional orogens. In Geomorphology and Global Tectonics, ed. M. A. Summerï¬eld, pp. present, given the uncertainties associated with each dat- 29-55. Chichester: Wiley. ing technique, it is not possible to be sufï¬ciently precise to Bentley, M. J., and J. B. Anderson. 1998. Glacial and marine evidence for the establish cause and effect and thus understand better the links ice sheet conï¬guration in the Weddell Sea-Antarctic Peninsula region between the ocean, atmosphere, and ice sheet in inï¬uencing during the Last Glacial Maximum. Antarctic Science 10:307-323. or responding to environmental change. Bishop, P., and G. Goldrick, 2000. Geomorphological evolution of the East Australian continental margin. In Geomorphology and Global Tectonics, One important implication arising from this overview ed. M. A. Summerï¬eld, pp. 225-254. Chichester: Wiley. is the realization that fragile features in the landscape can Boulton, G. S. 1996. Theory of glacier erosion, transport and deposition as be very old, whether they have been protected beneath ice a consequence of subglacial sediment deformation. Journal of Glaciol- or subject to a hyperarid climate with minimal erosion. It is ogy 42:43-62. possible for striations and moraines to survive for millions of Briner, J. P., G. H. Miller, P. Davis, and R. C. Finkel. 2006. Cosmogenic radionuclides from fjord landscapes support differential erosion by years. Thus, it is possible that features such as trimlines with overriding ice sheets GSA Bulletin 118:406-420. striations and associated moraines may date from glaciation Brook, E. J., E. T. Brown, M. D. Kurz, R. P. Ackert, G. M. Raisbeck, and F. prior to the mid-Miocene. Pleistocene changes in ice eleva- Yiou. 1995. Constraints on age, erosion and uplift of Neogene glacial tion may be indicated only by a sparse scatter of boulders. deposits in the Transantarctic Mountains determined from in situ cos- If so, then reconstructions of former Pleistocene ice thick- mogenic 10Be and 26Al. Geology 23:1057-1152. Cape Roberts Science Team. 2000. Summary of results. In Studies from nesses based solely on trimline and striation evidence can Cape Roberts Project: Initial Report on CRP-3, Ross Sea, Antarctica, be misleading. eds. P. J. Barrett, M. Sarti, and S. Wise. Terra Antartica 7185-203. Siena: It is worth reï¬ecting on the richness of the record of Terra Antarctica Publication. landscape evolution in Victoria Land. At least in part this Cooper, A., H. Stagg, and E. Geist. 1991. Seismic stratigraphy and structure must be due to ease of access and proximity to the perma- of Prydz Bay, Antarctica: Implications from ODP Leg 119 drilling. In Ocean Drilling Program Leg 119 Scientiï¬c Results, eds. J. B. Barron and nent base of McMurdo Sound. If so, there is the prospect of B. Larsen, pp. 5-25. College Station, TX: Ocean Drilling Program. equally rich archives in other parts of Antarctica. The chal- Cuffey, K. M., H. Conway, A. M. Gades et al. 2000. Entrainment at cold lenge is to develop and explore these further to improve our glacier beds. Geology 28:351-354. understanding of landscape evolution and its contribution to De Santis, L., J. B. Anderson, G. Brancolini, and I. Zayatz. 1995. Seismic the wider sciences. record of late Oligocene through Miocene glaciation on the central and eastern continental shelf of the Ross Sea. In Geology and Seismic Stratigraphy of the Antarctic Margin, eds. A. K. Cooper, P. F. Barker, ACKNOWLEDGMENTS and G. Brancolini, Antarctic Research Series 68:235-260. Washington, D.C.: American Geophysical Union. The authors are indebted to the U.K. Natural Environment DeConto, R. M., and D. Pollard. 2003. Rapid Cenozoic glaciation of Antarc- Research Council for support. Author D.E.S. is grateful for tica induced by declining atmospheric CO2. Nature 421:245-249. Denton, G. H., and D. E. Sugden. 2005. Meltwater features that suggest support from the Division of Polar Programs of the U.S. Miocene ice-sheet overriding of the Transantarctic Mountains in Victoria National Science Foundation, the British Antarctic Survey, Land, Antarctica. Geograï¬ska Annaler 87A:67-85. the Carnegie Trust for the Universities of Scotland and ACE Denton, G. H., J. G. Bockheim, S. C. Wilson, and M. Stuiver. 1989. Late (Antarctic Climate Evolution) Program. We are most grate- Wisconsin and early Holocene glacial history, inner Ross Embayment, ful to Peter Barrett, David Elliot, Adrian Hall, and Andrew Antarctica. Quaternary Research 31:151-182. Denton, G. H., M. L. Prentice, and L. H. Burckle. 1991. Cainozoic history Mackintosh for their thoughtful and helpful comments on of the Antarctic Ice Sheet. In The Geology of Antarctica, ed. R. J. Tingey, the manuscript. pp. 365-433. Oxford: Clarendon Press. Denton, G. H., J. G. Bockheim, R. H. Rutford, and B. G. Andersen. 1992. Glacial history of the Ellsworth Mountains, West Antarctica. In Geology REFERENCES and Palaeontology of the Ellsworth Mountains, West Antarctica, eds. Anderson, J. B. 1999. Antarctic Marine Geology. Cambridge: Cambridge G. F. Webers, C. Craddock, and J. F. Splettstoesser. Geological Society University Press. of America Memoir 170:403-442. Baroni, C., V. Noti, S. Ciccacci, G. Righini, and M. C. Salvatore. 2005. Denton, G. H., D. E. Sugden, D. R. Marchant, B. L. Hall, and T. I. Wilch. Fluvial origin of the valley system in northern Victoria Land (Antarctica) 1993. East Antarctic Ice Sheet Sensitivity to Pliocene Climatic Change from quantitative geomorphic analysis. Geological Society of America from a Dry Valleys Perspective. Geograï¬ska Annaler 75A:155-204. Bulletin 117:212-228. Drewry, D. J. 1982. Ice ï¬ow, bedrock and geothermal studies from radio- Barrett, P. J. 2007. Cenozoic climate and sea level history from glacimarine echo sounding inland of McMurdo Sound, Antarctica. In Antarctic strata off the Victoria Land coast, Cape Roberts Project, Antarctica. In Geoscience, ed. C. Craddock, pp. 977-983. Madison: University of Glacial Processes and Products, eds. M. J. Hambrey, P. Christoffersen, Wisconsin Press. N. F. Glasser, and B. Hubbard. International Association of Sedimen- Dunbar, G. B., T. R. Naish, P. J. Barrett, C. R. Fielding, and R. D. Powell. tologists Special Publication 39:259-287. Forthcoming. Constraining the amplitude of late Oligocene bathymetric Barrett, P. J., M. J. Hambrey, D. M. Harwood, A. R. Pyne, and P.-N. Webb. changes in western Ross Sea during orbitally-induced oscillations in 1989. Synthesis. In Antarctic Cenozoic History from CIROS-1 Drillhole, the East Antarctic Ice Sheet. 1. Implications for glacimarine sequence McMurdo Sound, ed. P. J. Barrett. DSIR Bulletin 245:241-251. Wel- stratigraphic models. Palaeogeography, Palaeoclimatology, Palaeoecol- lington: DSIR Publishing. ogy, doi:10.1016/j.palaeo.2007.08.018.
JAMIESON AND SUGDEN 53 Ehrmann, W. U., M. Setti, and L. Marinoni. 2005. Clay minerals in Cenozoic Lawver, L. A., L. M. Gahagan, and M. F. Cofï¬n. 1992. The development of sediments off Cape Roberts (McMurdo Sound, Antarctica) reveal the seaways around Antarctica. Antarctic Research Series 56:7-30. palaeoclimatic history, Palaeogeography, Palaeoclimatology, Palaeo- Leventer, A., E. Domack, R. Dunbar, J. Pike, C. Stickley, E. Maddison, ecology 229:187-211. S. Brachfeld, P. Manley, and C. McLennen. 2006. Marine sediment Evans, I. S. 1969. The geomorphology and morphometry of glacial and record from the East Antarctic margin reveals dynamics of ice sheet nival areas. In Water, Earth and Man, ed. R. J. Chorley, pp. 369-380. recession. GSA Today 16(12):4-10. London: Methuen. Lewis, A. R., D. R. Marchant, D. E. Kowalewski, S. L. Baldwin, and L. E. Evatt, G. W., A. C. Fowler, C. D. Clark, and N. R. J. Hulton. 2006. Subgla- Webb. 2006. The age and origin of the Labyrinth, western Dry valleys, cial ï¬oods beneath ice sheets. Philosophical Transactions of the Royal Antarctica: Evidence for extensive middle Miocene subglacial ï¬oods Society of London 364:1769-1794. and freshwater discharge to the Southern Ocean. Geology 34:513-516. Fitzgerald, P. G. 1992. The Transantarctic Mountains of southern Victoria Lewis, A. R., D. R. Marchant, A. C. Ashworth, S. R. Hemming, and M. L. Land: The application of ï¬ssion track analysis to a rift shoulder uplift. Machlus. 2007. Major middle Miocene global climate change: Evidence Tectonics 11:634-662. from East Antarctica and the Transantarctic Mountains. Geological Soci- Gore, D. B., E. J. Rhodes, P. C. Augustinius, M. R. Leishman, E. A. Colhoun, ety of America Bulletin 119(11):1449-1461, doi:10.1130/B26134. and J. Rees-Jones. 2001. Bunger Hills, East Antarctica: Ice free at the Lidmar-Bergstrom, K. 1982. Pre-Quaternary geomorphological evolu- Last Glacial Maximum. Geology 29:1103-1106. tion in southern Fennoscandia. Sveriges, geologiska undersÃ¶kning C Hall, B. L., G. H. Denton, D. R. Lux, and J. G. Bockheim. 1993. Late 785:1-202. Tertiary Antarctic paleoclimate and ice-sheet dynamics inferred from Linton, D. L. 1949. Unglaciated areas in Scandinavia and Great Britain. surï¬cial deposits in Wright Valley. Geograï¬ska Annaler 75A:239-267. Irish Geography 2:77-79. Hambrey, M. J., N. F. Glasser, B. C. McKelvey, D. E. Sugden, and D. Fink. Linton, D. L. 1963. The forms of glacial erosion. Transactions of the Insti- 2007. Cenozoic landscape evolution of an East Antarctic oasis (Radok tute of British Geographers 33:1-28. Lake area, northern Prince Charles Mountains), and its implications LÃ¸ken, O. H., and D. A. Hodgson. 1971. On the submarine geology along for the glacial and climatic history of Antarctica. Quaternary Science the east coast of Bafï¬n Island. Canadian Journal of Earth Sciences Reviews 26:598-626. 8:185-195. Hodgson, D. A., P. E. Noon, W. Vyverman, C. L. Bryant, D. B. Gore, P. Lythe, M., D. G. Vaughan, and BEDMAP Consortium. 2001. BEDMAP: A Appleby, M. Gilmour, E. Verleyen, A. Sabbe, V. J. Jones, J. C. Ellis- new thickness and subglacial topographic model of Antarctica. Cana- Evans, and P. B. Wood. 2001. Were the Larsemann Hills ice free through dian Journal of Earth Sciences 106:11335-11352. the Last Glacial Maximum? Antarctic Science 13:440-454. Mackintosh, A., D. White, D. Fink, D. B. Gore, J. Pickard, and P. C. Fanning. Holbourn, A., W. Kuhn, M. Schulz, and H. Erlenkeuser. 2005. Impacts of 2007. Exposure ages from mountain dipsticks in Mac Robertson Land, orbital forcing and atmospheric carbon dioxide on Miocene ice-sheet East Antarctica, indicate little change in ice-sheet thickness since the expansion. Nature 438:483-487. Last Glacial Maximum. Geology 35:551-554. Holmlund, P., and J. O. NÃ¤slund. 1994. The glacially sculptured landscape Marchant, D. R., G. H. Denton, D. E. Sugden, and C. C. Swisher III. 1993. in Dronning Maud Land, Antarctica, formed by wet-based mountain Miocene glacial stratigraphy and landscape evolution of the western glaciation and not by the present ice sheet. Boreas 23:139-148. Asgard Range, Antarctica. Geograï¬ska Annaler 75A:303-330. Holtedahl, H. 1958. Some remarks on the geomorphology of continental Marchant, D. R., A. R. Lewis, W. M. Phillips, E. Moore, R. Souchez, shelves off Norway, Labrador and southeast Alaska. The Journal of G. H. Denton, and D. E. Sugden. 2002. Formation of patterned ground Geology 66:461-471. and sublimation till over Miocene glacier ice, southern Victoria Land, Holtedahl, H. 1967. Notes on the formation of fjords and fjord valleys. Antarctica. GSA Bulletin 114:718-730. Geograï¬ska Annaler 49A:188-203. Margerison, H. R., W. M. Phillips, F. M. Stuart, and D. E. Sugden. 2005. Horton, R. E. 1945. Erosional development of streams and their drainage Cosmogenic 3He concentrations in ancient ï¬ood deposits from the basins: Hydrophysical approach to quantitative morphology. Geological Coombs Hills, northern Dry Valleys, East Antarctica: Interpreting Society of America Bulletin 56:275-370. exposure ages and erosion rates. Earth and Planetary Science Letters Huber, M., and D. Nof. 2006. The ocean circulation in the Southern 230:163-175. Hemisphere and its climatic impacts in the Eocene. Palaeogeography, Mercer, J. H. 1961. The response of ï¬ord glaciers to changes in the ï¬rn Palaeoclimatology, Palaeoecology 231:9-28. limit. Journal of Glaciology 3:850-858. Ives, J. D. 1966. Block ï¬elds, associated weathering forms on mountain tops Mercer, J. H. 1978. West Antarctic Ice Sheet and CO2 greenhouse effect: A and the nunatak hypothesis. Geograï¬ska Annaler 48A:220-223. threat of disaster. Nature 271:321-325. Ivy-Ochs, S., C. SchlÃ¼chter, P. W. Kubik, B. Dittrich-Hannen, and J. Beer. Naish, T. R. et al. 2001. Orbitally induced oscillations in the East Antarctic 1995. Minimum 10Be exposure ages of early Pliocene for the Table ice sheet at the Oligocene/Miocene boundary. Nature 413:719-723. Mountain Plateau and the Sirius Group at Mount Fleming, Dry Valleys, NÃ¤slund, J.-O. 2001. Landscape development in western and central Dron- Antarctica. Geology 23:1007-1010. ning Maud Land, East Antarctica. Antarctic Science 13:302-311. Jamieson, S. S. R., N. R. J. Hulton, D. E. Sugden, A. J. Payne, and J. Taylor. Ã Cofaigh, C., R. D. Larter, J. A. Dowdeswell, C.-D. Hillenbrand, C. J. 2005. Cenozoic landscape evolution of the Lambert basin, East Antarc- Pudsey, J. Evans, and P. Morris. 2005. Flow of the West Antarctic Ice tica: The relative role of rivers and ice sheets. Global and Planetary Sheet on the continental margin of the Bellingshausen Sea at the Last Change 45:35-49. Glacial Maximum. Journal of Geophysical Research 110(B11):103, Jamieson, S. S. R., N. R. J. Hulton, and M. Hagdorn. Forthcoming. doi:10.1029/2005JB003619. Modelling landscape evolution under ice sheets. Geomorphology, OâBrien, P. E., A. K. Cooper, C. Richter et al. 2001. Leg 188 summary: doi:10.1016/j.geomorph.2007.02.047. Prydz Bay-Cooperation Sea, Antarctica. Proceedings of the Ocean Kennett, J. P. 1977. Cenozoic evolution of Antarctic glaciation, the circum- Drilling Program, Initial Reports 188:1-65. College Station, TX: Ocean Antarctic Ocean and their impact on global oceanography. Journal of Drilling Program. Geophysical Research 82:3843-3860. Payne, A. J. 1999. A thermomechanical model of ice ï¬ow in West Antarctica. Kerr, A., and W. D. Hermichen. 1999. Glacial modiï¬cation of the Shackleton Climate Dynamics 15:115-125. Range, Antarctica. Terra Antartica 6:353-360. Pekar, S. F., and R. M. DeConto. 2006. High-resolution ice-volume estimates Lawver, L. A., and L. M. Gahagan. 2003. Evolution of Cenozoic seaways for the early Miocene: Evidence for a dynamic ice sheet in Antarctica. in the circum-Antarctic region. Palaeogeography, Palaeoclimatology, Palaeogeography, Palaeoclimatology, Palaeoecology 231:101-109. Palaeoecology 198(1-2):11-37.
54 ANTARCTICA: A KEYSTONE IN A CHANGING WORLD Perkins, D. 1984. Subglacial landscape in Antarctica. Unpublished Ph.D. Sugden, D. E., M. A. Summerï¬eld, G. H. Denton, T. I. Wilch, W. C. thesis. University of Aberdeen. McIntosh, D. R. Marchant, and R. H. Rutford. 1999. Landscape develop- Persano, C., F. M. Stuart, P. Bishop, and D. N. Barfod. 2002. Apatite (U- ment in the Royal Society Range, southern Victoria Land, Antarctica: Th)/He age constraints on the development of the Great Escarpment on Stability since the middle Miocene. Geomorphology 28:181-200. the southeastern Australian passive margin. Earth and Planetary Science Sugden, D. E., G. Balco, S. G. Cowdery, J. O. Stone, and L. C. Sass III. 2005. Letters 200:79-90. Selective glacial erosion and weathering zones in the coastal mountains Prest, V. K. 1970. Quaternary geology of Canada. Economic Geology Report of Marie Byrd Land, Antarctica. Geomorphology 67:317-334. 1:676-764. Summerï¬eld, M. A. 2000. Geomorphology and global tectonics: Introduc- Priestley, R. E. 1909. Scientiï¬c results of the western journey. In The Heart tion. In Geomorphology and Global Tectonics, ed. M. A. Summerï¬eld, of the Antarctic, vol. 2, ed. E. H. Shackleton, pp. 315-333. London: pp. 3-12. Chichester: Wiley. Heinemann. Summerï¬eld, M. A., D. E. Sugden, G. H. Denton, D. R. Marchant, H. A. Raine, J. I., and R. A. Askin. 2001. Terrestrial palynology of Cape Roberts P. Cockburn, and F. M. Stuart. 1999. Cosmogenic isotope data support drillhole CRP-3, Victoria Land Basin, Antarctica. Terra Antartica previous evidence of extremely low rates of denudation in the Dry 8:389-400. Valleys region, southern Victoria Land. Geological Society of London Reeh, N. 1991. Parameterization of melt rate and surface temperature on Special Publication 162:255-267. the Greenland ice sheet. Polarforschung 59(3):113-128. Taylor, G. 1922. The Physiography of McMurdo Sound and Granite Har- Shevenell, A. E., J. P. Kennett, and D. W. Lea. 2004. Middle Miocene bour Region. British Antarctic (Terra Nova) Expedition, pp. 1910-1913. Southern Ocean cooling and Antarctic cryosphere expansion. Science London: Harrison. 305:1766-1770. Taylor, J., M. J. Siegert, A. J. Payne, M. J. Hambrey, P. E. OâBrien, A. K. Smith, A. M., T. Murray, K. W. Nicholls, K. Makinson, G. Aoalgeirsdottir, Cooper, and G. Leitchenkov. 2004. Topographic controls on post-Oligo- A. E. Behar, and D. G. Vaughan. 2007. Rapid erosion, drumlin forma- cene changes in ice-sheet dynamics, Prydz Bay region, East Antarctica. tion, and changing hydrology beneath an Antarctic ice stream. Geology Geology 32:197-200. 35:127-130. van de Flierdt, T., G. E. Gehrels, S. L. Goldstein, and S. R. Hemming. Sorlien, C. C., B. P. Luyendyk, D. S. Wilson, R. C. Decesari, L. R. Bartel, 2007. Pan-African Age of the Gamburtsev Mountains? In Antarctica: A and J. B. Diebold. 2007. Oligocene development of the West Antarctic Keystone in a Changing WorldâOnline Proceedings for the Tenth Inter- Ice Sheet recorded in eastern Ross Sea strata. Geology 35:467-470. national Symposium on Antarctic Earth Sciences, eds. Cooper, A. K., Stern, T. A., A. K. Baxter, and P. J. Barrett. 2005. Isostatic rebound due C. R. Raymond et al., USGS Open-File Report 2007-1047, Extended to glacial erosion within the Transantarctic Mountains. Geology Abstract 176, http://pubs.usgs.gov/of/2007/1047/. 33:221-224. van der Wateren, F. M., T. J. Dunai, R. T. Van Balen, W. Klas, A. L. L. Stickley, C. E., H. Brinkhuis, S. A. Schellenberg, A. Sluijs, U. Rohl, M. Verbers, S. Passchier, and U. Herpers. 1999. Contrasting Neogene M. Fuller, M. Grauert, M. Huber, J. Warnaar, and G. L. Williams. 2004. denudation histories of different structural regions in the Transantarctic Timing and nature of the deepening of the Tasmanian Gateway. Pale- Mountains rift ï¬ank constrained by cosmogenic isotope measurements. oceanography PA4027, doi:10.1029/2004PA001022. Global and Planetary Change 23:145-172. Stokes, C. R., and C. D. Clark. 1999. Geomorphological criteria for identify- Vaughan, D. G., J. L. Bamber, M. Giovinetto, J. Russell, and A. P. R. Cooper. ing Pleistocene ice streams. Annals of Glaciology 28:67-74. 1999. Reassessment of net surface mass balance in Antarctica. Journal Strahler, A. N. 1958. Dimensional analysis applied to ï¬uvially eroded land- of Climate 12:933-946. forms. Bulletin of the Geological Society America 69:279-300. Webb, P.-N. 1994. Paleo-drainage systems of East Antarctica and sedi- Stroeven, A. P., D. Fabel, C. HÃ¤ttestrand, and J. Harbor. 2002. A relict ment supply to West Antarctic Rift System basins. Terra Antartica landscape in the centre of Fennoscandian glaciation: Cosmogenic 1:457-461. radionuclide evidence of tors preserved through multiple glacial cycles. Webb, P.-N., and J. Wrenn. 1982. Upper Cenozoic biostratigraphy and Geomorphology 44:145-154. micropaleontology of Taylor Valley, Antarctica. In Antarctic Geosci- Sugden, D. E. 1977. Reconstruction of the morphology, dynamics and ther- ence, ed. C. Craddock, pp. 117-1122. Madison: University of Wisconsin mal characteristics of the Laurentide ice sheet at its maximum. Arctic Press. and Alpine Research 9:21-47. Wellner, J. S., A. L. Lowe, S. S. Shipp, and J. B. Anderson. 2001. Distribu- Sugden, D. E. 1978. Glacial erosion by the Laurentide ice sheet. Journal of tion of glacial geomorphic features on the Antarctic continental shelf Glaciology 20:367-392. and correlation with substrate: Implications for ice behavior. Journal of Sugden, D. E., and G. H. Denton. 2004. Cenozoic landscape evolution of Glaciology 47:397-411. the Convoy Range to Mackay Glacier area, Transantarctic Mountains: Zachos, J. C., J. Breza, and S. W. Wise. 1992. Early Oligocene ice sheet Onshore to offshore synthesis. GSA Bulletin 116:840-857. expansion on Antarctica, sedimentological and isotopic evidence from the Kergulen Plateau. Geology 20:569-573.