New Insights into Arctic Paleogeography and Tectonics from U-Pb Detrital Zircon Geochronology

In order to test existing models for the formation of the Amerasian Basin of the Arctic, detrital zircon suites from 12 samples of Triassic sandstone from the circum-Arctic region were d

New Insights into Arctic Paleogeography and Tectonics from U-Pb

Detrital Zircon Geochronology

Elizabeth L Miller 1 , Jaime Toro 2 , George Gehrels 3 , Jeffrey M Amato 4 , Andrei

Prokopiev 5 , Marianna Tuchkova 6 , Vyacheslav V Akinin 7, Trevor A Dumitru1 , Thomas

E Moore 8, Ashton Embry9 and Michael P Cecile 9

1 Department of Geological and Environmental Sciences, Stanford Univ., Stanford, CA, USA

2 Department of Geology and Geography, West Virginia Univ., Morgantown, WV, USA

3 Department of Geosciences, Univ of Arizona, Tucson, AZ, USA

4 Department of Geological Sciences, New Mexico State Univ., MSC 3AB, PO Box 30001,

Las Cruces, NM, USA

5Diamond and Precious Metal Geology Institute, Russian Academy of Sciences, Yakutsk,

Russian Federation

6Russian Academy of Sciences, Moscow, Russian Federation

7Russian Academy of Sciences, Magadan, Russian Federation

8U.S Geological Survey, Menlo Park, CA, USA

9Geological Survey of Canada, Calgary, Alberta, Canada

In order to test existing models for the formation of the Amerasian Basin of the Arctic, detrital zircon suites from 12 samples of Triassic sandstone from the circum-Arctic region were dated by laser ablation-ICPMS Triassic successions represent the last major phase of regional deposition in the Arctic prior to the formation of the Amerasian Basin Detrital zircon samples from the northern Verkhoyansk fold-and-thrust belt (Siberia), from Chukotka and Wrangel Island (NE Russia), and from northwestern Alaska have similar age

distributions distinguished by Carboniferous and Permo-Triassic peaks In contrast, samples from further east in Alaska and Canada (Sadlerochit Mts and Sverdrup Basin) are dominated

by Cambro-Ordovician, or Proterozoic zircons The samples from Chukotka most closely resemble those from the Verkhoyansk The current most popular reconstruction of the

Amerasian Basin involves counter-clockwise rotation of Arctic Alaska and Chukotka away from the Canadian Arctic margin Although this satisfies many stratigraphic and geophysical constraints for the Alaska portion of the reconstruction, it places Chukotka a great distance away from Siberia and it leads to considerable overlap of continental crust if Chukotka is treated as a rigid block To resolve these discrepancies we propose an alternative model for formation of the Amerasian Basin where Chukotka originates closer to Siberia, but after rotation with Alaska, moves further away from Siberia by extension and right-lateral strike-slip related to the subsequent opening of the Makarov Basin portion of the Amerasian Basin

INDEX TERMS: 1165 Geochronology: Sedimentary geochronology; 8155 Tectonophysics:

Plate motions (3040); 8178 Tectonics and magmatism; 9315 Geographic Location: Arctic region (0718, 4207)

KEYWORDS: Detrital Zircons, Canada Basin, Amerasian Basin, Chukotka, Triassic

“windshield wiper” model (Figures 1 and 2) This model proposes that Early Cretaceous rifting translated a small continental plate, known as the Arctic Alaska-Chukotka microplate, southward from the Arctic margin of Canada The plate rotated about a pole located in the McKenzie Delta region, opening the Amerasian Basin by rifting while closing the

Angayucham and Anyui ocean basins to the south [e.g Grantz et al., 1990] (Figure 2) Data

in support of this model include correlation of the upper Paleozoic-Mesozoic stratigraphy of the North Slope of Alaska to the Sverdrup Basin of Arctic Canada [Grantz et al., 1990; Embry et al., 1990; Toro et al., 2004], and magnetic and gravity anomalies that identify a paleo-spreading center in part of the Amerasian Basin [Laxon and McAdoo, 1996; Brozena etal., 2002] Although the rotational model best satisfies geological and geophysical data from the Canadian and Alaskan portion of the circum-Arctic, reconstructing the Russian portion of the Arctic Alaska-Chukotka plate by this model creates a space problem Geological

constraints suggest that the Arctic Alaska-Chukotka microplate extends as far west as the New Siberian Islands and includes most of the immense and poorly known East Siberian continental shelf (Figure 1) Restoring this piece of real estate against the Canadian Arctic

and Barents margins with the rotation model produces significant overlap of continental crust(Figure 2) [e.g Drachev , 2004; Natal’in, 2004].

Understanding the development of the Amerasian Basin is required if we are to understand the origin and evolution of the vast Russian continental shelves and basins as well

as the origin of several enigmatic bathymetric features in the Arctic Ocean such as the

continental Lomonosov Ridge, the Alpha-Mendeleev Ridge, and the Chukchi Borderland (Figure 1) At present, land-based geologic efforts provide a straightforward and cost-

effective means of addressing this problem In particular, data from NE Russia has the potential for making significant contributions to our understanding For instance, the

reconstruction shown in Figure 2 suggests Chukotka, Russia, should have depositional ties with Arctic Canada and not Siberia Here we present U-Pb dating of single grains from detrital zircon populations from Triassic sandstones of the circum-Arctic, carried out in order

to better constrain the position of the different parts of the Arctic Alaska-Chukotka

microplate prior to the opening of the Amerasian basin Recent advances in laser mass spectrometry and ion microprobe technology have made the dating of large populations

ablation-of detrital zircon feasible and thus is a highly desired tool for provenance studies [e.g

Gehrels et al., ] The utility of such studies for plate tectonic reconstructions and for our understanding of frontier regions is just beginning to be demonstrated

2 The Arctic Alaska-Chukotka Microplate

The key element in the plate tectonic evolution of the Amerasian Basin is the Arctic Alaska-Chukotka microplate Figure 1 portrays the microplate as a single elongate

continental sliver Its northern boundary (present-day coordinates) is the Arctic Alaskan and Russian outer shelf edges Its southern boundary is defined by a belt of ophiolitic rocks that

includes the Angayucham terrane of the southern Brooks Range [e.g Moore et al., 1994] and the South Anyui zone of western Chukotka [e.g Sokolov et al., 2002]

Arctic Alaska-Chukotka is believed to be a unified continental fragment because of : 1) similar-age Neoproterozoic volcanic and plutonic basement rocks [Moore et al.,1994; Amato et al., 2003; Kos’ko et al.,1993]; 2) An early Paleozoic carbonate succession with Siberian faunal affinities that unconformably overlies this basement [Dumoulin et al., 2002; Natalin et al., 1998; Kos’ko et al.,1993]; 3) A suite of Devonian plutons found in the Brooks Range that have counterparts on the Seward Peninsula and along the north coast of Chukotka [Moore et al., 1994; Toro et al., 2002; Kos’ko et al., 1993]; 4) Similarities in the nature and age of Middle Jurassic-Early Cretaceous structural trends including ophiolite belts developed

by collision-related deformation along the southern side of the plate These related structures appear to extend continuously from the Brooks Range to offshore Wrangel Island and into the Chukotka fold belt (where they remain poorly dated) (Figure 1a)

shortening-Important along-strike differences in the stratigraphy of the microplate include the varying nature of Triassic sedimentary successions Northern Alaska is characterized by a relatively thin, clastic passive margin succession, which includes chert and other pelagic deposits in the south and more proximal thin platform sandstones in the north In Chukotka, the Triassic consists of thick turbidite sequences intruded by gabbroic dikes and sills at their base [Gelman, 1963; Ivanov and Milov, 1975] Thus, in contrast to Alaska, Chukotka was paleogeographically linked to a major clastic source and probably experienced a Triassic rifting event that formed the deep-water basins [Tuchkova et al., 2004]

3 Analytical Methods

We separated detrital zircons from twelve samples of Triassic sandstones (Table 1) Zircons were mounted in epoxy and polished to expose the interior of the grains Isotopic

analyses were performed with a Micromass Isoprobe multicollector ICPMS with a laser ablation system Laser beam diameter was ~30 m and yielded ablation pits ~20 m deep Inter-element fractionation was monitored by analyzing fragments of a large concordant zircon with a known (ID-TIMS) age of 564 ± 1 Ma (2 σ error) This standard was analyzed once for every four unknowns Grains were selected randomly from all sizes and

morphologies present, except for avoidance of grains with fractures or inclusion The ablated material is carried in argon gas into the plasma source of a Micromass Isoprobe, which is equipped with a flight tube of sufficient width that U, Th, and Pb isotopes are measured simultaneously All measurements are made in static mode, using Faraday detectors for 238U,

232Th, 208-206Pb, and an ion-counting channel for 204Pb Ion yields are ~1 mv per ppm Each analysis consists of one 20-second integration on peaks with the laser off (for backgrounds),

20 one-second integrations with the laser firing, and a 30 second delay to purge the previous sample and prepare for the next analysis

Common Pb correction is made by using the measured 204Pb and assuming an initial Pb composition from Stacey and Kramers [1975] (with uncertainties of 1.0 for 206Pb/204Pb and 0.3 for 207Pb/204Pb) Our measurement of 204Pb is unaffected by the presence of 204Hg because backgrounds are measured on peaks (thereby subtracting any background 204Hg and 204Pb), and because very little Hg is present in the argon gas

Inter-element fractionation of Pb/U is generally <20%, whereas fractionation of Pb isotopes is generally <5% In-run analysis of fragments of a large zircon crystal (generally every fifth measurement) with known age of 564 ± 4 Ma (2-sigma error) [G Gehrels,

unpublished data] is used to correct for this fractionation The uncertainty resulting from the calibration correction (together with the uncertainty from decay constants and common Pb composition) is generally 3% (2-sigma) for 206Pb/238U and >1.2 Ga 206Pb/207Pb ages

Fractionation also increases with depth into the laser pit The accepted isotope ratios are

accordingly determined by least-squares projection through the measured values back to the initial determination The complete measured isotopic ratios and ages are reported as tables inthe Supplementary Data Repository1 Errors are from the measurement of 206Pb/238U,

206Pb/207Pb, and 206Pb/204Pb and are reported at the 1-sigma level Additional errors that affect all ages include uncertainties from (1) U decay constants, (2) the composition of common Pb (assumed to be ±1.0 for 206Pb/204Pb and ±0.3 for 207Pb/204Pb), and (3) calibration correction These systematic errors add an additional 2% (1-sigma) uncertainty to 206Pb/238U and >1.2 Ga

interpreted ages are presented in Table 2

The U-Pb dates are plotted on a normalized relative-probability distribution diagrams [Ludwig, 2003] The relative height of the peaks corresponds to the significance of that population of ages (Figure 3) Some analyses were not plotted because of one of two

reasons: 1) for ages <1000 Ma, the 206Pb/238U uncertainty was >10%; 2) for ages >1000 Ma, the 207Pb/206Pb uncertainty was >10% or the discordance between the 206Pb/238U and 207Pb/206Pbages was > 20%

1 Insert reference to Supplementary Data Repository here.

4 Geological Setting of Samples and Interpretation of Detrital Data

Triassic sandstones were selected for our study because they represent the youngest stage of widespread marine deposition prior to the onset of rifting that led to the dispersal of terranes during formation of the Amerasian Basin Rifting started in the Late Jurassic along the Alaskan margin [Grantz et al., 1990], but the major episode of sea floor spreading appears to have occurred in the Early Cretaceous [Grantz et al., 1990, but see Drachev, 2004] In

addition, compressional deformation also affected parts of Arctic Alaska starting as early as the mid Jurassic when ophiolitic sequences were emplaced in Alaska and the Kolyma region [Moore et al., 1998; Oxman et al., 1995] Detrital age data from Triassic sedimentary rocks along the length of the Arctic Alaska-Chukotka microplate were collected to test the integrity and continuity of this crustal fragment and constrain its paleogeographic links to source terranes Three additional samples from the northern Verkhoyansk fold-and-thrust belt were used to characterize a “Siberian source”, and two additional samples were dated from the Sverdrup Basin to characterize a “Canadian source” (Figure 1)

Nearly all of the Triassic sandstones analyzed were derived from the erosion of Paleozoic and early Mesozoic source terranes Precambrian zircons are present in all samples but only in limited amounts, with the exception of one sample from the southern Sverdrup basin (Figure 3) Our discussion below focuses on the interpretation of the Phanerozoic ages

in the zircon populations because these are most easily interpreted with respect to possible source areas making them useful for paleogeographic reconstructions The samples are discussed below from west to east

The shallow marine Triassic succession of the Verkhoyansk (Figure 1) is part of an immense passive-margin siliciclastic wedge that ranges in age from Carboniferous to Early Jurassic [Khudoley and Guriev, 1994] A major fluvial system connected the tectonically

active Baikal Mountain region to the Verkhoyansk Siberian margin via the long-lived Vilyui graben system (Figure 1) and other distributary systems [e.g Khudoley and Guriev, 1994]

Earliest Triassic strata of the northern Verkhoyansk contain basalts and tuffs that are believed to be coeval with Siberian Trap volcanism [Andrianova and Andrianov, 1970] Triassic sandstones have petrographic characteristics of “recycled orogen” sources,

compatible with their inferred provenance The age distribution of detrital zircon suites from one Middle Triassic and two Upper Triassic sandstone samples are very similar and are grouped in Figure 3 (a total of 281 ages) Post-Vendian peaks in the probability distribution occur at 288, 482, 517 and 597 Ma These ages are compatible with their inferred source in the Baikal Mountain region along the southern margin of the North Asia craton, known to have been an active Andean margin throughout the Paleozoic and early Mesozoic Major batholiths were intruded in the Carboniferous; Ordovican and Cambrian plutonic rocks are present as well [Wickam et al., 1995; Yarmolyuk et al., 1997; Bukharov et al., 1992]

The Triassic of Chukotka is represented by up to 5 km of mostly distal turbidite sequences that range upwards into shelf deposits of Upper Triassic (Norian) age [Tuchkova etal., 2004] Sandstones are mostly fine-grained with compositions varying from “continental block” to “recycled orogen” [Tuchkova et al., 2004] We analysed one Middle Triassic (Carnian) and two Upper Triassic (Norian) samples Zircon populations from these three samples are very similar and are grouped in Figure 3 (285 grains) Zircon ages show peaks at

247, 299, 370, 440, and 494 Ma with older peaks at 575 and 650 Ma It is notable that the Paleozoic peak ages in these samples are very similar to those found in the Verkhoyansk samples The Permo-Triassic peak at 247 Ma lies within error of the age of well-dated Siberian Trap magmatism (248-253 Ma) [e.g Campbell et al., 1992; Venkatesan et al., 1997; Renne and Basu, 1996] Although Siberian Trap magmatism was dominantly basaltic, lessersilicic intrusions and tuffs have been described In the Taymir Peninsula there are 249-241

Ma high level syenite-granite stocks that have been linked to the mantle plume which also generated the Siberian Traps [Vernikovsky et al., 2003] Tuffs and tuffaceous sediments are also present in the Late Permian and earliest Triassic deposits of both the Verkhoyansk and Chukotka [Andriavova and Andrianov, 1970; Chasovitin and Shpetnyi,1964; Belik and Sosunov,1969; Sosunov and Tilíman, 1960] suggesting that magmatism related to the

Siberian Traps was the most likely origin for zircons of this age

Wrangel Island lies between the sample sites in Chukotka and those in western Alaska (Figure 1) The Triassic strata of Wrangel Island consist of proximal siliciclastic turbidites [Kos’ko et al., 1993] A single small Triassic sample from Wrangel Island (provided by Mike Cecile of the Geological Survey of Canada) yielded 47 datable zircons (Figure 3) Phanerozoic zircon age peaks at 253, 280, and 449 Ma are similar to the peaks seen in the samples from Chukotka (to which Wrangel Island is linked, geographically, stratigraphically and structurally) and the Verkhoyansk (which is quite distant) (Figure 3) A single grain yielded an age younger than 200 Ma, but this is not considered to be statistically significant given that concordance cannot be demonstrated (Figure 3) (Supplementary Data)

The westernmost sample site in Alaska is in the Lisburne Hills fold-and-thrust belt which consists of lower Paleozoic to Mesozoic sedimentary rocks [e.g Moore et al., 2002] Two samples were collected from an unusual sandstone member of the Triassic Otuk

Formation which lies stratigraphically between the shale and chert members described by Blome et al [1988] The sandstone is very fine to fine-grained, massive, and about 5 m thick

Monotis fossils found in the sandstone suggest a Late Triassic age The sandstone is lithic,

consisting of about 40 percent monocrystalline quartz, 20 percent altered feldspar, and 40 percent lithic fragments (argillite, chert, siltstone, granitic fragments and carbonate) The Otuk is thought to represent condensed sedimentation in an outer shelf environment [Moore and others, 1994] Although a predominantly sedimentary provenance is inferred for this

sandstone, the granitic grains suggest that it was shed from a source region that lay to the NW

in the offshore Chukchi platform region [K.W Sherwood, 2000, written communication as cited in Moore and others, 1994]

Because the age spectra of the two samples from the Lisburne Hills are similar, the data are grouped in Figure 3 (187 zircons) Similarities and differences in the ages of the detrital zircons from the Lisburne Hills and NE Russia are evident A Permo-Triassic age peak at 255 Ma is present in the Lisburne Hills Triassic which is also seen in samples from Wrangel Island and Chukotka Paleozoic peaks at 321, 362 and 431 Ma are similar to the Russian samples Differences include a younger Triassic (222 Ma) peak in the Lisburne Hillssandstones Its closeness in age to the depositional age of the sandstone suggests the presence

of an active magmatic source The Lisburne Hills samples also lack the strong Cambrian and Late Proterozoic peaks seen in the Chukotka samples

A sample from the Early Triassic Ledge Member of the Ivishak Formation of the Sadlerochit Mountains in northeastern Alaska (Figures 1, 2) shows the greatest difference in provenance ages from the samples previously discussed Sedimentary structures and burrowsindicate a shallow marine deltaic origin for the Ledge Member [Mariani, 1987] This unit is also notable because it is the main oil reservoir of the Prudhoe Bay field Zircons from the Ivishak Sandstone exhibit age peaks at 466, 531 and 564 Ma Permo-Triassic and Late Carboniferous to earliest Permian ages are conspicuously absent (Figure 3)

Despite their geographic proximity, two samples from the Triassic successions of the Sverdrup Basin in Arctic Canada (Figure 1) have very different zircon age populations from each other Sample AE2 from Axel Heiberg Island on the northern flank of the Sverdrup Basin has a zircon age distribution virtually identical to that of the Sadlerochit Mountains of northeastern Alaska (Figure 3) This lends support to the rotational hypothesis for the

opening of the Canada basin [Grantz et al., 1990], which would place these two samples in

close proximity prior to rifting (Figure 2) In contrast, sample AE1, from the southern flank

of the Sverdrup Basin on Ellesmere Island, is dominated by Proterozoic zircons with major peaks at 1.2 and 1.78 Ga, plus a few zircons of Paleozoic age (Figure 3) An in depth

provenance study utilizing Sm-Nd isotopic data concluded that most of the sediments in the Sverdrup basin were sourced from recycled Franklinian-Caledonian (Early Paleozoic)

sources [Patchett et al., 2004] This is consistent with the detrital ages of sample AE2 and the detrital zircon ages from the Triassic of the Sadlerochit Mountains However, the zircons dated in sample AE1 appear to represent the erosion of older crystalline rocks to the south and east of the Sverdrup Basin, possibly related to uplift of Early Proterozoic and

Greenvillian age basement during the early stages of Atlantic rifting

Alaska-Chukotka have the greatest similarities in ages of source regions with coeval rocks of the Verkhoyansk belt of Siberia (Figure 3) However, Chukotka now lies a great distance from the northern Verkhoyansk and also from the Siberian Traps from which it may have received