Erosional history of the Prydz Bay sector of East ...

Article Volume 13, Number 11 29 November 2012 Q11015, doi:10.1029/2012GC004364 ISSN: 1525-2027

Erosional history of the Prydz Bay sector of East Antarctica from detrital apatite and zircon geo- and thermochronology multidating Clare J. Tochilin, Peter W. Reiners, Stuart N. Thomson, and George E. Gehrels Department of Geosciences, University of Arizona, Tucson, Arizona 85721, USA ([email protected])

Sidney R. Hemming and Elizabeth L. Pierce Lamont-Doherty Earth Observatory and Department of Earth and Environmental Sciences, Columbia University, Palisades, New York 10964, USA [1] Approximately 98% of East Antarctica is covered by the East Antarctic Ice Sheet (EAIS), which has

covered parts of the continent since the early Oligocene (34 Ma) and obscures evidence about the region’s tectonic and erosional history. To better constrain the subglacial record, we analyzed geo- and thermochronologic dates of Oligocene-Quaternary sediments from Prydz Bay, which drains 16% of the EAIS. We used multidating techniques, measuring U-Pb, fission track, and (U-Th)/He dates on apatite and zircon grains and 40Ar/39Ar dates on hornblende grains to determine crystallization and cooling ages. Apatite and zircon U-Pb dates and hornblende 40Ar/39Ar dates are dominantly 500 Ma, recording Pan-African metamorphism and magmatism. Zircon fission track dates record cooling at 250–300 Ma and 120 Ma from Permian-Triassic (300-201 Ma) rifting and Cretaceous (120 Ma) magmatic resetting. Mean apatite fission track dates decrease from 280-210 Ma in early Oligocene samples, with lag times decreasing from 250-180 My, indicating increasing erosion rates. Miocene-Quaternary (10.7-0 Ma) samples show a smaller range from 180 to 150 Ma. Youngest measured apatite He ages also decrease from 100 Ma to 25 Ma in Oligocene-Miocene samples. These results indicate increasing erosion rates (0.2 km/My) in catchments draining to Prydz Bay in the early Oligocene, with slower erosion since the late Miocene. This erosion was likely achieved by glacial incision into pre-existing valleys, reaching depths of 2.8–3.0 km by the late Miocene. This is consistent with EAIS models showing a transition to less erosive, cold-based conditions following the mid-Miocene climatic optimum. Components: 12,000 words, 10 figures, 1 table. Keywords: East Antarctica; Prydz Bay; geochronology; multidating; thermochronology. Index Terms: 1140 Geochronology: Thermochronology; 1199 Geochronology: General or miscellaneous; 8175 Tectonophysics: Tectonics and landscape evolution. Received 2 August 2012; Revised 11 October 2012; Accepted 17 October 2012; Published 29 November 2012. Tochilin, C. J., P. W. Reiners, S. N. Thomson, G. E. Gehrels, S. R. Hemming, and E. L. Pierce (2012), Erosional history of the Prydz Bay sector of East Antarctica from detrital apatite and zircon geo- and thermochronology multidating, Geochem. Geophys. Geosyst., 13, Q11015, doi:10.1029/2012GC004364.

©2012. American Geophysical Union. All Rights Reserved.

1 of 21

Geochemistry Geophysics Geosystems

3

G

TOCHILIN ET AL.: EROSIONAL HISTORY OF EAST ANTARCTICA

1. Introduction [2] The tectonic and erosional history of East Antarctica is poorly understood due to coverage of about 98% of the bedrock by the East Antarctic Ice Sheet (EAIS) [Barron et al., 1991; Webb, 1990]. The EAIS is the largest and longest-lived ice sheet on Earth and encompasses 90% of the ice in Antarctica [Bamber et al., 2000]. One nucleation point for the onset of glaciation in East Antarctica in the early Oligocene is thought to be the subglacial Gamburtsev Mountains in the interior of the continent, which reach elevations of up to 4 km above sea level, but are entirely covered by ice [Bo et al., 2009; Cox et al., 2010; DeConto and Pollard, 2003a, 2003b; Ferraccioli et al., 2011; Siegert, 2008; Siegert et al., 2008; Taylor et al., 2004; van de Flierdt et al., 2008]. Rapid expansion of the EAIS at 34 Ma is explained in recent models as being largely a result of global cooling related to declining atmospheric CO2 levels [DeConto and Pollard, 2003a], with the development of the Antarctic Circumpolar Current and consequent thermal isolation of Antarctica acting as a potential trigger [Kennett, 1977]. Between 34 and 14 Ma, the EAIS is considered to have undergone large fluctuations in extent with areas of dynamic, warm-based glaciation and selective linear erosion concentrated along its margins, after which time it underwent a shift to more stable, cold-based, and locally less erosive glaciation as a result of further global cooling [Bo et al., 2009; DeConto and Pollard, 2003a, 2003b; Jamieson and Sugden, 2008; Jamieson et al., 2010; Lewis et al., 2007; Naish et al., 2001; Sugden, 1996; Sugden et al., 1993; Young et al., 2011]. A more complete understanding of the tectonic and erosional history of East Antarctica is important in that it provides greater clarity regarding long-term ice sheet behavior, including the expansion of the EAIS at 34 Ma, its fluctuations in extent from 34 to 14 Ma, and its relative stability since 14 Ma. This history is also crucial to the understanding of the interactions between subglacial topography, tectonics, and ice flow patterns in East Antarctica. [3] Approximately 16% of the EAIS is drained by the largest ice stream in Antarctica, the Lambert Glacier, which flows through the north-south trending Lambert Graben and into Prydz Bay on the northeastern margin of the continent (Figure 1) [Bamber et al., 2000; Barrett, 1996; Cox et al., 2010; DeConto and Pollard, 2003a; Hambrey and McKelvey, 2000; Jamieson et al., 2005]. The Lambert Graben was formed primarily by Permo-Triassic

10.1029/2012GC004364

rifting [Cox et al., 2010; Kurinin and Grikurov, 1982; Lisker, 2002; Lisker et al., 2007b] and extends 500 km to the south of Prydz Bay into the interior of the continent [O’Brien et al., 2007]. Local, small-volume mafic magmatism associated with the break-up of Gondwana occurred in the Lambert Graben/Prydz Bay area during the early to mid Cretaceous [Arne, 1994; Coffin et al., 2002; Collerson and Sheraton, 1986; Hambrey and McKelvey, 2000; Jamieson et al., 2005; Kurinin and Grikurov, 1982; Lisker et al., 2003, 2007a, 2007b]. [4] In addition to hosting an ice stream currently draining a large portion of the EAIS, the Lambert Graben was an important drainage route for fluvial systems prior to the onset of glaciation in East Antarctica. Jamieson and Sugden [2008] modeled the preglacial fluvial drainage patterns of this area (Figure 1) and concluded that the Lambert Graben served as the primary path for rivers carrying sediment out of the Gamburtsev Mountains. This drainage system was also fed by sediments from the Prince Charles Mountains and Grove Mountains along the flanks of the Lambert Graben near Prydz Bay [Cox et al., 2010; Jamieson et al., 2005; O’Brien et al., 2007]. The current ice streams draining this part of East Antarctica likely follow these pre-existing pre-glacial drainages within the Lambert Graben, Gamburtsev Mountains, and surrounding areas, ultimately delivering sediment from the East Antarctic interior to Prydz Bay [Bo et al., 2009; Cox et al., 2010; Ferraccioli et al., 2011; Jamieson et al., 2005]. [5] To produce new constraints on the subglacial exhumational history of East Antarctica since the onset of the EAIS, we applied a multidating approach to zircon and apatite grains from sediments recovered from Prydz Bay by Ocean Drilling Project (ODP) Leg 119, Core 739C. These sediments are ultimately derived from the inaccessible subglacial bedrock and can provide geo- and thermochronologic records of their source areas as well as valuable information about the phases of East Antarctic glaciation since its onset at 34 Ma [Barron et al., 1991; Whitehead et al., 2006]. The four dating methods applied to detrital minerals in this study are U-Pb (apatite and zircon), fission track (apatite and zircon), (U-Th)/He (apatite; due to the limited number of zircon grains in our samples, zircon (U-Th)/He analysis was not conducted), and 40 Ar/39Ar (hornblende). The highest temperature systems applied in this study, the zircon and apatite U-Pb systems, are most useful for providing information on crystallization age and provenance. In the 2 of 21


3

G


10.1029/2012GC004364

Figure 1. Isostatically rebounded subglacial topography (80 times vertically exaggerated) of the Lambert Graben/ Prydz Bay region of East Antarctica showing the location of core 739C used in this study, the location of core 1166A used in the study of Thomson et al. [2011], and the modeled pre-glacial drainage into the Lambert Graben [Jamieson and Sugden, 2008] [after Cox et al., 2010].

case of apatite, which has a closure temperature of 450–500 C, this system also provides high temperature cooling history and indicates possible metamorphic events. The 40Ar/39Ar system for hornblende has a similar closure temperature to the apatite U-Pb system and also provides high temperature cooling history. The fission track and (U-Th)/He systems are lower temperature thermochronometers that indicate cooling history through a range of temperatures below 300 C. These thermochronometers are useful for determining exhumational histories. This combination of multiple high and low temperature chronometers is extremely powerful, as it allows for the determination of both crystallization ages as well as a range of cooling ages on single grains, providing a more complete provenance and cooling history than any one method alone. To distinguish between references to age and time span, we consistently use Ma to indicate age and My to indicate time span. [6] ODP Leg 119, Core 739C was specifically chosen to augment the study of Thomson et al. [2011;

submitted to Nature Geoscience, 2012] by targeting the synglacial stratigraphic record, and hence providing us the opportunity to more completely constrain the post-34 Ma subglacial erosional history of the Lambert Glacier and its tributaries. A better appreciation of changes to subglacial topography during glaciation is crucial to better understanding the still poorly known long-term dynamics of the EAIS. In comparison to the longer sedimentary record provided by ODP Leg 188, Core 1166A (Figure 1) studied by Thomson et al. [2011; also submitted manuscript, 2012], Core 739C includes a more extensive section of Oligocene sediments deposited following the first expansion of the EAIS, as well as a more complete upper part of the section that records glacial sedimentation as far back as the late Miocene. [7] In this study, we apply a single grain multidating approach to glacially derived sediments from Prydz Bay to provide more comprehensive constraints on the tectonic and erosional history of East Antarctica. We use these crystallization and cooling ages to develop a greater understanding of the phases and 3 of 21


3

G


behavior of the EAIS since the early Oligocene and to gain insight into the interactions between subglacial topography, tectonics, and glacial dynamics in the Prydz Bay region of East Antarctica.

2. Geologic Setting [8] The East Antarctic Shield is primarily an amalgamation of Archean through early Cambrian crustal fragments, deformed and intruded during multiple episodes of tectonism [Dalziel, 1992; Ferraccioli et al., 2011; Fitzsimons, 1997, 2000a, 2000b]. Grenville (980–1250 Ma) and Pan-African (500–600 Ma) age metamorphic belts and intrusive rocks provide evidence of the continent’s involvement in the assembly of supercontinents Rodinia and Gondwana, respectively [Boger et al., 2002; Fitzsimons, 2000b, 2003; Veevers, 2003; Veevers et al., 2008]. In some areas within the Lambert Graben, Precambrian bedrock is overlain by mostly undeformed sediments and sedimentary rocks, including the Permian-Triassic Amery Group and the Cenozoic Pagodroma Group [Fitzsimons, 2000a; McKelvey et al., 2001; O’Brien et al., 2007; Veevers and Saeed, 2008; Whitehead et al., 2006]. [9] Several studies have documented U-Pb zircon dates of grains from the limited ice-free bedrock exposures in East Antarctica, including samples from the Prydz Bay coast [Hemming et al., 2007; Veevers and Saeed, 2008; Veevers et al., 2008; Williams et al., 2007, 2010], the margins of the Lambert Graben, the Prince Charles Mountains [Boger et al., 2001; Carson et al., 1996, 2000; Liu et al., 2007; Mikhalsky et al., 2006], and the Grove Mountains. These ages and field observations are extrapolated to determine the geology that may be beneath the ice based on limited geophysical evidence [van de Flierdt et al., 2008] (Figure 2). These studies showed a wide range of crustal formation ages in East Antarctica ranging from 3.5 Ga-0.5 Ga, with dominant peaks at 0.4–0.6 Ga, 0.9–1.2 Ga, 1.6 Ga, 2.2 Ga, 2.6–2.7 Ga, 3.0–3.1 Ga, and 3.4–3.5 Ga. Peaks at 0.5–0.6 Ga and 0.9–1.2 Ga from syntectonic intrusive rocks as well as metamorphic rims on some crystals correlate well with large-scale Pan-African (500–600 Ma) and Grenville (980–1250 Ma) orogenic events known to have affected the region [Fitzsimons, 2000b; Fitzsimons et al., 1997]. [10] Additionally, a number of studies have mea-

sured 40Ar/39Ar ages on hornblende and mica from diamictite in Prydz Bay. In contrast with the variation observed in the U-Pb ages, 40Ar/39Ar dates from this

10.1029/2012GC004364

region of East Antarctica are almost exclusively 500 Ma, reinforcing evidence for a widespread tectonothermal event affecting the area during PanAfrican time (500–600 Ma), resulting in a strong metamorphic overprint. 40Ar/39Ar studies also show a minor peak of Grenville age (1.1–1.3 Ga) [Hemming et al., 2007; Phillips et al., 2007; Roy et al., 2007]. [11] Studies measuring apatite fission track and apa-

tite and zircon (U-Th)/He dates of Precambrian basement rocks, Permian and Triassic sedimentary rocks, and sediment cores from the Prydz Bay region have also been conducted. Cox et al. [2010] analyzed grains in Eocene fluvial sediment from cores in Prydz Bay, reporting apatite fission track dates of 207– 612 Ma and (U-Th)/He dates of 97–377 Ma and 197–397 Ma for apatite and zircon, respectively. They interpret these ages to reflect very slow erosion rates in the Gamburtsev Mountains and very little tectonic activity in the interior of the continent since initial rifting of the Lambert Graben in the PermianTriassic. Published apatite fission track dates in Precambrian basement rocks and Permian-Triassic sedimentary rocks in the Prince Charles Mountains and the Lambert Graben show a wider variation in apatite fission track age from 100–300 Ma with an older Permian-Triassic peak (220–260 Ma) and a younger Cretaceous-Paleocene peak (55–110 Ma) [Arne, 1994; Lisker et al., 2003, 2007a, 2007b]. These authors postulate two phases of cooling: the older peak they associate with Permo-Triassic rifting in the Lambert Graben area, whereas the younger peak is proposed to correspond to Cretaceous rifting and exhumation during the breakup of Gondwana consistent with 2–4 km of erosion associated with a Cretaceous rifting event. However, independent geologic evidence for such an event is lacking. The understanding of ages determined from these geoand thermochronologic systems and their implications regarding tectonic and erosional activity in the Prydz Bay/Lambert Graben area is significant not only in terms of the timing of these events, but also in terms of the erosional patterns that can be identified within this graben system over time. [12] In other areas of East Antarctica such as Dron-

ning Maud Land (55 -15 E) and Wilkes Land (155 -90 E), several of the same zircon U-Pb age peaks seen in the Prydz Bay/Lambert Graben region are observed, especially Grenville (980–1250 Ma) and Pan-African (500–600 Ma) age peaks [Fitzsimons, 2000b; Roy et al., 2007]. Fitzsimons [2000b] does make a distinction between slightly different Grenville ages in each area, however. Grenville age peaks in the Prydz Bay region are 4 of 21


3

G


10.1029/2012GC004364

Figure 2. Bedrock geology map of the area surrounding the Lambert Graben and Prydz Bay showing location of core 739C. NPCM (in legend) refers to the Northern Prince Charles Mountains. Black areas indicate outcrops. This data has been compiled primarily from geological mapping and isotopic analyses. Ages in legend are zircon U-Pb ages. [after van de Flierdt et al., 2008; Fitzsimons, 2003; Kamenev et al., 1993].

mainly 990–900 Ma, while age peaks in Dronning Maud Land and Wilkes Land are mainly 1.09– 1.03 Ga and 1.33–1.13 Ga, respectively. In Wilkes Land there is also a dominant zircon U-Pb age peak from 1.6 to 1.8 Ga not seen in the Prydz Bay catchment [Goodge, 2007]. Hornblende 40Ar/39Ar studies show that all regions of East Antarctica are dominated by 500 Ma dates, with the exception of Wilkes Land and Adélie Land, which show primarily Mesoproterozoic and Paleoproterozoic dates. This implies that this area did not experience the Pan-African (500–600 Ma) orogenic event as strongly as the rest of East Antarctica [Di Vincenzo et al., 2007; Duclaux et al., 2008; Ménot et al., 2007; Pierce et al., 2011; Roy et al., 2007].

section of diamictites down to the early Oligocene recovered in this core and other cores in the vicinity provides evidence for a fully developed EAIS by that time [Barron et al., 1991; Solheim et al., 1991]. Volpi et al. [2009] reanalyzed the stratigraphy of core 739C (Figure 3), slightly redefining the stratigraphic and age divisions previously determined by the Shipboard Scientific Party on the basis of new seismic profiles, which they integrate with existing lithologic, biostratigraphic, and magnetostratigraphic data. The stratigraphic ages in Figure 3 and any depositional ages of samples mentioned throughout this study are those reported by Volpi et al. [2009]. Depositional ages were determined primarily through diatom biostratigraphy [Whitehead et al., 2006].

3. Samples and Stratigraphy

[14] Core 739C consists of six main units of dia-

[13] The samples analyzed in this study are early

Oligocene-Quaternary sediments from ODP Leg 119, Core 739C (67 16.57′S, 75 04.91′E) recovered from Prydz Bay in January, 1988 [Shipboard Scientific Party, 1989]. The presence of a thick

mictites (Figure 3). Unit A, the uppermost segment of the core, consists of soft Quaternary diatomaceous sand, clay, and poorly consolidated diamicton. Unit B, 24–80 m below seafloor (mbsf ), is a homogeneous, poorly sorted, late Pliocene diamictite. Unit C (80–110 mbsf) is a thick, stratified 5 of 21


3

G


10.1029/2012GC004364

Figure 3. Stratigraphy of ODP core 739C. Analyzed segments of the core are indicated with arrows along the left side of the column. [Volpi et al., 2009; Shipboard Scientific Party, 1989].

diamictite that is also late Pliocene. Unit D (110– 140 mbsf ) is a highly compacted diamictite of early Pliocene age. Unit E (140–174 mbsf) is a very highly compacted diamictite divided into two sections. E1 consists of diatomaceous, early Pliocene

diamictite, while E2 is late Miocene diamictite with few to no diatoms. Core 739C is the only core drilled in this area to contain Miocene sediments. Below the Miocene unit is an unconformity, underlain by a very thick section of early Oligocene 6 of 21


3

G


Table 1. Sample Names, Stratigraphic Ages, and Depths of the Twenty Analyzed Samples From ODP Core 739C Sample Name

Stratigraphic Age

Depth (mbsf )

001R55 004R21 005R115 013R10 014R29 017R20 019R85 021R70 022R74 023R110 025R10 030R91 031R85 034R80 038R60 041R110 047R55 052R20 056R35 062R0

Quaternary Late Pliocene Late Pliocene Late Pliocene Early Pliocene Early Pliocene Early Pliocene Late Miocene Late Miocene Late Miocene Early Oligocene Early Oligocene Early Oligocene Early Oligocene Early Oligocene Early Oligocene Early Oligocene Early Oligocene Early Oligocene Early Oligocene

0–9.5 24.1–28.7 28.7–38.3 105.9–115.5 115.5–125.2 135.0–140.0 144.7–149.7 154.3–159.3 159.3–164.0 164.0–169.0 173.6–183.2 221.8–231.4 231.4–241.4 260.4–270.0 298.9–308.6 327.8–337.5 385.9–395.6 434.2–439.2 453.5–458.5 482.6–486.8

diamictites (174–486.8 mbsf ) interpreted to represent sediments deposited in a large prograding trough mouth fan [Barron et al., 1991; Hambrey and McKelvey, 2000; Volpi et al., 2009]. We carried out geo/thermochronological analyses on twenty samples from this core. Sample names, stratigraphic ages, and depths for each of the twenty samples are presented in Table 1.

4. Methods 4.1. Fission Track Analysis [15] All samples were processed using standard

mineral separation techniques for zircon and apatite [Gehrels, 2000]. The zircon yield for all samples was very low, and many of the zircon grains were metamict, making them unsuitable for fission track and U-Pb analysis. Zircon and apatite grains from each of the twenty samples were mounted in Teflon and epoxy respectively, and polished and etched using the methods outlined in Hurford et al. [1991]. The samples were analyzed by the external detector method [Tagami and O’Sullivan, 2005] and irradiated at the Oregon State University Triga Reactor in Corvallis, OR, along with European Institute for Reference Materials and Measurements uraniumdosed glasses IRMM 540R for apatite and IRMM 541 for zircon to monitor neutron fluence. Spontaneous and induced tracks were counted using an

10.1029/2012GC004364

Olympus BX51 microscope at 1250 magnification at the University of Arizona Fission Track Laboratory. Central ages were calculated using the zeta calibration approach of Hurford and Green [1983]. The zeta values used for zircon and apatite were 351.99 8.94 and 113.21 1.81, respectively. Analytical data for 1019 analyzed apatite grains and 218 analyzed zircon grains can be found in the auxiliary material (see Tables A1 and A2 in the Table S1 file).1

4.2. U-Pb Analysis [16] Fission track mounts for zircon and apatite were

attached to glass slides adjacent to standards mounted separately in epoxy or Teflon for U-Pb analysis. Primary standards included a Sri Lanka zircon [Gehrels et al., 2008], and Madagascar and McClure Mountain apatites [Thomson et al., 2012]. All grains dated by fission track analysis were located and analyzed at the Arizona LaserChron Center by laser ablation-multicollector-inductively coupled plasma mass spectrometry (LA-MCICPMS). Detailed procedures for zircon analysis are described by Gehrels et al. [2008]; procedures for apatite analysis are described by Thomson et al. [2012]. Three analyses were conducted in each apatite grain, depending on the size of the grain, to account for compositional variation. Concordia ages were later calculated to determine one age for each grain. In cases where very few zircon grains (16 or fewer, in most cases) could be dated by fission track analysis, a small number of additional suitable zircon grains were analyzed to obtain a greater number of U-Pb analyses for those samples. The measured isotopic ratios for zircon are corrected for common Pb using the measured 204Pb, assuming an initial Pb composition according to Stacey and Kramers [1975]. Measured isotopic ratios for apatite are also corrected for common Pb according to Stacey and Kramers [1975]. However, due to the high common to radiogenic Pb ratio in most apatite (204Pb is typically in the range of 1000–3000 cps for apatite in this study, with 206Pb/204Pb ratios ranging from 20 to 200), a five-step iterative process is also applied to correct for this [Thomson et al., 2012]. Full analytical data for 982 apatite grains (multiple analyses per grain) and 262 zircon grains yielding data of acceptable precision (apatite: