Pulsed Miocene Range Growth in Northeastern Tibet: Insights from Xunhua Basin Magnetostratigraphy and Provenance

Downloaded from gsabulletin.gsapubs.org on July 5, 2012

Pulsed Miocene range growth in northeastern Tibet: Insights from Xunhua Basin magnetostratigraphy and provenance

Richard O. Lease1,†, Douglas W. Burbank1, Brian Hough2, Zhicai Wang3, and Daoyang Yuan3 1Department of Earth Science, University of California, Santa Barbara, California 93106, USA 2Department of Earth and Environmental Sciences, University of Rochester, Rochester, New York 14627, USA 3State Key Laboratory of Earthquake Dynamics, Institute of Geology, China Earthquake Administration, Beijing 100029, China

ABSTRACT the north at ca. 22 Ma. Subsequently, growth and deposition initiating in the late Miocene on of the north-trending Jishi Shan occurred both the northern and eastern plateau margins. Sedimentary rocks in Tibetan Plateau at ca. 13 Ma and is highlighted by an accel- Some of the most compelling evidence derives basins archive the spatiotemporal patterns eration in Xunhua Basin accumulation rates from low-temperature thermochronological his- of deformation, erosion, and associated cli- between 12 and 9 Ma, as well as by a signifi - tories in uplifted mountain ranges where age- mate change that resulted from Cenozoic cant change in detrital zircon provenance of depth or time-temperature relationships exhibit a continental collision. Despite growing under- nearby Linxia Basin deposits by 11.5 Ma. Ini- clear transition to rapid cooling. Accelerated late standing of basin development in northeast- tial growth of the WNW-trending Laji Shan Miocene–Pliocene development of the eastern ern Tibet during initial India-Asia collision, in the early Miocene and subsequent growth plateau margin is revealed by detailed thermo- as well as in the late Miocene–Holocene, sur- of the north-trending Jishi Shan ~10 m.y. chronological records from the Longmen Shan prisingly little is known about the interven- later support interpretations of a middle (Godard et al., 2009), Min Shan (Kirby et al., ing period: a time when the plateau may have Mio cene kinematic reorganization in north- 2002), and Gonga Shan areas (Clark et al., 2005; undergone fundamental tectonic changes. To eastern Tibet. Ouimet et al., 2010), as well as fi ssion-track- fi ll this gap, we present new magnetostratig- length modeling from a portion of the Qinling raphy from a >2300-m-thick fl uviolacustrine INTRODUCTION (Enkelmann et al., 2006). In northeastern Tibet, succession that spans ca. 30–9.3 Ma. An inte- thermochronologic age-elevation relationships grated analysis of sedimentology, subsidence, Over the past decade, growing evidence in- also clearly demarcate a late Miocene onset of and provenance from this section reveals the dicates that contractional deformation near the accelerated cooling in the Liupan Shan (Zheng sequential, pulsed erosion of multiple ranges present northern margin of the Tibetan Plateau et al., 2006) and north Qilian Shan (Zheng et al., bordering the Xunhua Basin. Emergence of commenced a few million years after initial 2010), which is corroborated by sedimentologi- the WNW-trending Laji Shan is highlighted India-Asia continental collision (Yin et al., cal and magnetostratigraphic archives adjacent by a doubling of sediment accumulation 2002, 2008a; Yin and Harrison, 2000), despite to these ranges (Fig. 1A; Bovet et al., 2009; rates between 24 and 21 Ma and a transition the >3000 km distance separating northeast- Wang et al., 2011a). Finally, magnetostrati- to coarse allu vial facies at 20.3 Ma. Detrital ern Tibet from the plate boundary at ca. 50 Ma graphic records adjacent to the Riyue Shan zircon U/Pb age spectra show that these (Molnar and Stock, 2009). Along the northeast- (Fang et al., 2005; Lease et al., 2007), Qinghai coarse sediments came from basement ter- ern plateau margin, a regional 25° clockwise ro- Nan Shan (Zhang et al., 2011), and Gonghe Nan ranes within the Laji Shan. Together these tation by ca. 41 Ma (Dupont-Nivet et al., 2008) Shan (Craddock et al., 2011) reveal the late Mio- observations suggest accelerated growth and the onset of localized thrust-induced cool- cene emergence of these ranges. of the Laji Shan and its coupled foreland ing (Clark et al., 2010) and fault gouge (Duvall Despite increasing knowledge of plateau- basin at ca. 22 Ma. The most rapid accumu- et al., 2011) at ca. 45–50 Ma directly support margin deformation during both the Eocene lation rates in Xunhua Basin occur within an Eocene onset of contractional tectonism in onset of continental collision as well as during the fi nest-grained strata and suggest an un- the Xining-Lanzhou region (Fig. 1A). To the the late Miocene, geological knowledge of the derfi lled basin during the fastest interval of west, along the northern plateau margin, initia- interval between these two episodes remains Laji Shan deformation. Growth of the Laji tion of a suite of contractional structures in the sparse. For example, the large body of work re- Shan occurred northward of the contempo- northern Qaidam Basin and adjacent ranges at vealing extensive late Miocene–Pliocene range raneous plateau margin, which had been de- ca. 65–50 Ma (Yin et al., 2008a) corroborates growth in northeastern Tibet (Fig. 1A) stands in fi ned since ca. 45–50 Ma by the West Qinling, this early deformation. Furthermore, the oldest contrast to only meager evidence for Oligocene– lying ~60 km farther south. Hence, following strata within continuous Cenozoic stratigraphic middle Miocene range growth. Proxy data sets, ~20–25 m.y. of apparent stability, the defor- successions in both the Xining (Dai et al., 2006) such as middle Miocene vertical-axis rotations mation front in this region jumped ~60 km to and Qaidam Basins (summarized in Yin et al., (Yan et al., 2006) or decreases in detrital ther- 2008b) are nearly coeval with the initial phase mochronology lag times (Zheng et al., 2003), †Present address: Institut für Geowissenschaft, Uni- of India-Asia continental collision. are often unable to delimit proximate causes versität Tübingen, 72074 Tübingen, Germany; e-mail: On the other hand, several studies document (Zhang et al., 2011) or source areas. In addition, [email protected] an episode of accelerated deformation, erosion, recent thermochronological results that suggest

GSA Bulletin; May/June 2012; v. 124; no. 5/6; p. 657–677; doi: 10.1130/B30524.1; 15 ﬁ gures; 2 tables; Data Repository item 2012078.

Riyue Xiejia Magnetostratigraphic section A B Xining Huangshuihe R. Shan Detrital zircon sample Basin 100°E 105°E 7 Fault Thermochron QS Quaternary 6 9–11 L Late Cenozoic 20 km 8 a j i Fig.13 S h Early Cenozoic QNS LJS a Cretaceous (-Jurassic) 5–7 21 16 n KS H LPS Triassic (-Carboniferous) –24 Ashi- YellowYel R. 7–10 Ganjia 5 Triassic-Permian plutons ~35 RS: 8–10 gong l J Z ow R. Pr2 Silurian (-Devonian) a JS: 12–14 Hualong i m Cambrian (-Ordovician) S GNS 36°N Guide h a s Early Paleozoic plutons 35°N WQ a z Basin a Xunhua Basin h K1 7–10 45–50 n r Pr3 Proterozoic N i i K Q: 4–9 Xunhua 15 W S Fig.14 oww R. e 4x YellYello s h MS: 3–5 t Maogou a 4 14 K2 Q LMS: 8–11 i n Wangjiashan n l Linxia 30°N i Basin GS: 9–13 N n QNS Mountain range g 5–7 Age of growth (Ma) 102°E 103°E

Figure 1. (A) Northeastern margin of the Tibetan Plateau showing ages of onset of accelerated Cenozoic range growth derived from bedrock thermochronology and sedimentary records. Contractional deformation across WNW-trending faults commenced shortly after initial India- Asia continental collision (yellow circles), with a later phase of incremental expansion and accelerated exhumation in the late Miocene (blue circles) occurring toward both the east and north. In this paper, we discuss new evidence for early–middle Miocene range growth (white circles). Data from west to east: the Kunlun Shan (KS) ca. 35 Ma (Clark et al., 2010), Qilian Shan (QS) 9–11 Ma (Bovet et al., 2009; Zheng et al., 2010), Qinghai Nan Shan (QNS) 5–7 Ma (Zhang et al., 2011), Gonghe Nan Shan (GNS) 7–10 Ma (Craddock et al., 2011), Riyue Shan (RS) 8–10 Ma (Fang et al., 2005; Lease et al., 2007), Gonga Shan area (GS) 9–13 Ma (Clark et al., 2005; Ouimet et al., 2010), Laji Shan (LJS) 21–24 Ma and Jishi Shan (JS) 12–14 Ma (Lease et al., 2011; this paper), West Qinling (WQ) 45–50 Ma (Clark et al., 2010; Duvall et al., 2011), Longmen Shan (LMS) 8–11 Ma (Godard et al., 2009), Min Shan (MS) 3–5 Ma (Kirby et al., 2002), Qinling (Q) 4–9 Ma (Enkelmann et al., 2006), and Liupan Shan (LPS) 7–10 Ma (Wang et al., 2011a; Zheng et al., 2006). K—Kunlun fault and H—Haiyuan fault. (B) Geological map of Xunhua Basin study area with magnetostratigraphic section locations (stars) (this paper; Dai et al., 2006; Fang et al., 2003, 2005; Hough et al., 2011), modern detrital zircon samples (numbered circles) (this paper; Lease et al., 2007), and major thermochronology transects (hexa- gons) (Lease et al., 2011; Clark et al., 2010) shown. Geology is modifi ed after Qinghai Bureau of Geology and Mineral Resources (QBGMR) (1991). For each detrital zircon sample, the associated catchment is outlined. sequential, early and middle Miocene growth of Factors that remain unknown are whether For clear answers to these questions, we need the Laji-Jishi Shan complex (Lease et al., 2011) signifi cant basin development and contractional to defi ne the onset, pace, nature, and extent of have yet to be substantiated by stratigraphic ar- deformation took place in the Xining-Lanzhou basin development and range growth. In this chives from adjacent basins. region of northeastern Tibet between 40 and paper, we integrate new magnetostratigraphic, In fact, few sedimentary successions from 10 Ma, and if so, whether these processes were provenance, subsidence, and sedimentological northeastern Tibet provide detailed histories isolated or part of a continuum. Several outstand- analyses from the northern Xunhua Basin with of both erosion and deposition between 40 ing questions concerning the Oligocene–middle extant work to assess the late Oligocene-to-Plio- and 10 Ma. More studies have focused on the Miocene depositional history of northeastern cene depositional and tectonic history of north- widespread Upper Miocene–Pliocene basin fi lls Tibet remain unanswered: (1) Was accumulation eastern Tibet. We document sequential growth that display increases in sediment accumulation slow and steady for tens of millions of years (e.g., of the Laji-Jishi Shan complex in northeastern rates and grain sizes (Fang et al., 2003, 2005; Dai et al., 2006) or was accumulation punctuated Tibet in two discrete phases: punctuated erosion Zhang et al., 2011), new source terranes (Lease by the onset of fl exure (e.g., Fang et al., 2003) of the Laji Shan at ca. 22 Ma and of the Jishi et al., 2007), and changes in stable isotopic com- or sediment ponding (e.g., Métivier et al., 1998) Shan at ca. 13 Ma. positions (Dettman et al., 2003; Hough et al., adjacent to newly uplifted ranges? (2) Was sedi- 2011). Many of these changes occurred ca. ment derived primarily from distal sources in the GEOLOGICAL SETTING 11–8 Ma and have been interpreted to indicate plateau interior to the south or from newly ex- accelerated deformation, basin development, posed sources outboard of the contemporaneous The Tertiary Xunhua Basin (herein simply re- and the emergence of adjacent mountain ranges plateau margin? (3) When did a once-contiguous ferred to as Xunhua Basin) is located ~100 km on the margin of the plateau since this time (e.g., foreland basin become compartmentalized into southeast of Xining, Qinghai province, China Dayem et al., 2009; Molnar and Stock, 2009). several smaller intermontane basins? (Fig. 1B). Xunhua Basin fi ll occupies a surface

658 Geological Society of America Bulletin, May/June 2012 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Oligocene–Miocene basin evolution on the northeastern Tibetan Plateau area of ~3300 km2 with elevations ranging from within the younger deposits. Lower Miocene Facies B deposits, located at 290–1070 m and 1870 m (base level of the Yellow River at the deposits in the southern Xunhua Basin generally 1160–1250 m height, consist of massive red- town of Xunhua) to 3000 m. Headward incision fi ne upward from sandy deltaic/fl oodplain facies brown gypsiferous mudstone with interbedded by the ancestral Yellow River and its tributaries to silt- and clay-rich lacustrine facies (Hough tan sandstone and gray-green siltstone (Fig. 2). through this region starting at ca. 1.7 Ma (Crad- et al., 2011). Upper Miocene–Pliocene deposits The mudstone has abundant gypsum veins and dock et al., 2010; Harkins et al., 2007; Li et al., in southern Xunhua, on the other hand, display a is otherwise similar in character to the mudstone 1997) has exhumed several well-exposed sedi- coarsening-upward trend from lacustrine facies (Fm, Table 1) described in the playa facies , with mentary sections. to braided fl uvial facies. Miocene–Pliocene strati- minor differences. Gypsum decreases in abun- In this paper, we utilize extant Chinese nomen- graphic records from the nearby Linxia and dance up section, with layered gypsum becom- clature to subdivide the range previously referred Guide Basins show similar depositional envi- ing rare above 400 m and gypsum veins rare to simply as the “Laji Shan” (Lease et al., 2007) ronments and coarsening-upward trends (Fang above 840 m. Mudstone found above 550 m, into three segments. This subdivision refl ects the et al., 2003, 2005). on the other hand, is characterized by gray- different structural orientations, uplift/erosion green to blue-green mottling, which increases histories, and geology of the constituent ranges— CENOZOIC STRATIGRAPHY in abundance up section. Finally, 0.3–1-m-thick from northwest to southeast (Fig. 1B), the Riyue interbeds of medium- to coarse-grained, tan Shan, the Laji Shan, and the Jishi Shan. The Xun- We measured and described ~2300 m of Ter- sandstone that commonly exhibits trough cross- hua Basin, the focus of our study, is bordered on tiary sedimentary rocks (Fig. 2) in a continuous stratifi cation (St, Table 1) occur throughout the all sides by ranges bounded by thrust faults: succession (the Hualong section) in the north- entire interval (Fig. 2). Individual lenticular the narrow ranges of the Laji Shan to the north, the eastern part of Xunhua Basin (Fig. 1B) near sandstone interbeds are stacked to form re- Jishi Shan to the east, and the Zamazari Shan to the town of Qiajia (section base: 35.981°N, stricted lenticular bodies that are generally <5 m the west, as well as the wider, more substantive 102.510°E; section top: 36.016°N, 102.505°E). thick and extend laterally for <10 m. Horizontal West Qinling to the south (Fig. 1B). Key aspects The succession lies adjacent to the eastern Laji laminations (Sh, Table 1) characterize the sand- of the growth of several of these ranges have Shan–northern Jishi Shan and records the late stone interlayers found between 450 and 650 m. been previously examined. The Eocene onset of Oligocene–Miocene growth of these ranges. The lithologies we group together as facies reverse faulting in the West Qinling defi ned the Sedimentary rocks in the Hualong section are B (Table 2) are interpreted to represent diverse contemporaneous plateau margin (Clark et al., interpreted to have been deposited within fi ve depositional characteristics. Thick intervals of 2010), whereas Miocene faulting northward of sedimentary environments (playa, marginal dominantly red-brown mudstone with abundant this margin initiated in the WNW-trending Laji lacustrine/deltaic plain, alluvial fan, braided mottling extending for tens to hundred of meters Shan ca. 22 Ma and in the north-trending Jishi fluvial, floodplain) that exhibit a generally of vertical section (Fig. 2) suggest prolonged Shan ca. 13 Ma (Lease et al., 2011) (Fig. 1B). coarsening-upward trend. Next, we describe the fl oodplain, mud fl at, or lacustrine deposition, Erosion of the NW-trending Riyue Shan north of lithologies found in the Hualong section in though ubiquitous gypsum veins obfuscate a Guide Basin intensifi ed after 10 Ma (Fang et al., the context of fi ve depositional facies that make more detailed analysis of mudstone sedimen- 2005; Lease et al., 2007; note, in earlier publi- up three overarching stratigraphic units. tology. Intervals of alternating blue-green and cations, the Riyue Shan were referred to simply At the base of the section (0–290 m; Fig. 2), red-brown siltstones (Fig. 2), on the other hand, as the “Laji Shan”). The response of the Xunhua facies A is composed of massive red-brown suggest periodic subaerial conditions, and rare Basin to uplift of these ranges of various sizes gypsiferous mudstone that is interbedded with primary gypsum indicates occasional evapora- and orientations, however, has been largely un- gray-green laminated siltstone and green-white tive conditions. Additionally, distinctive fl u- constrained. Thus, our study focuses on Xunhua layered gypsum (up to 3 m thick). The gypsif- vial channel sands commonly punctuate the Basin development in the northeastern corner erous mudstone is mildy calcareous to noncal- fi ne-grained deposits (Fig. 2) and are restricted of the basin adjacent to both the Laji Shan and careous with abundant interstitial gypsum and laterally to only a few meters, suggesting rib- J i s h i S h a n r a n g e s . gypsum veins (Fm, Table 1) and local gypsum bon channel bodies incised into a muddy plain The known Mesozoic–Cenozoic depositional fl owers, blades, and gypsarenite layers. Mud- (Friend et al., 1979). Collectively, we inter- record of northeastern Tibet in the Xining-Lan- stone beds are 0.1–30 m thick, tabular, and later- pret this diverse facies assemblage (facies B, zhou region to the north, east, and west of Xun- ally continuous over tens to hundreds of meters. Table 2) to be suggestive of intermittent fl ood- hua Basin is characterized by quite slow rates of Most gypsum beds are 0.5 to 30 cm thick, plain, shallow lacustrine, and distributary chan- accumulation (~40 m/m.y.) in a Lower Creta- tabular, and laterally continuous over meters to nels as part of a deltaic plain/marginal lacustrine ceous (extensional?) basin (Horton et al., 2004), >100 m (E, Table 1). Clasts coarser than very environment (Cabrera et al., 1985; Reading and followed by slower, but relatively steady Eocene- fi ne sand are chiefl y composed of gypsum Collinson, 1996). to-early Miocene accumulation (~20 m/m.y.) in a crystals from either primary or secondary pre- Facies C contains brown and red-brown con- foreland basin (Dai et al., 2006). Accelerated late cipitation. The weathering of layers varies from glomerates and sandstones with interbedded Miocene–Pliocene accumulation (>100 m/m.y.) blocky (0–50 m) to hackly (50–125 m) to cliff- mudstone arranged in several fi ning-upward in the Guide Basin (Fang et al., 2005) has been forming (125–290 m), with a series of cliffs and successions (up to 10 m thick). These depos- interpreted to indicate accelerated local rock up- ledges (125–290 m) defi ned by gypsum-rich its are restricted to the middle of the section lift (Lease et al., 2007). layers. We interpret facies A (Table 2) to have at 1070–1160 m height (Fig. 2). Conglomer- Eocene–Oligocene deposition in the Xining been deposited in a shallow lacustrine environ- ate beds are typically massive, matrix sup- Basin (Dai et al., 2006) occurred primarily ment with playa-like conditions (Schreiber and ported, and have nonerosive bases (Cmm, within playa and marginal lacustrine environ- El Tabakh, 2000; Talbot and Allen, 1996) based Table 1). Commonly, subangular pebbles up to ments (Abels et al., 2011; Dupont-Nivet et al., on the presence of extensive primary gypsum 40 cm diameter are enclosed in a medium- to 2007) that were limited in spatial extent. Clas- beds in combination with both oxidized and re- coarse-grained sand matrix. Sandstone beds tic input and inferred aridity generally increase duced gypsiferous muds (Fig. 2). are 0.5–1.5 m thick, medium grained, and have

Geological Society of America Bulletin, May/June 2012 659 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Lease et al.

Lithology UNIT FACIES 2300 E. Floodplain E E 2200 2295 Floodplain Floodplain D. Braided 2100 Fm

D Fm

2000 2290 Braided

1900 E. Floodplain Fm Ps Cci s f c g c E E 1800 2010 Flood Flood - plain - plain

D. Braided Fm Fm 1700 Ccm St Ps 1790 Unit 3 - FLUVIAL 1600 Sm Cci Fm St Cci 1530 Sl Sm D 1500 Cci

Braided Fm 2000 2000 1400 Fm/Sm Fm Cci Cci

GUIDE GROUP Ps 1300 Sh Fm s f c g c s c c Cci Fm f g St 1780 B B 1200 Cci plain s c g c Deltaic Deltaic St f C. Alluvial fan 1520 Cci Unit 2 - C fan 1100 Fm TRANSITIONAL

Alluvial St Cmm

s f c g c 1115 1000

Cmm 900 B. Deltaic plain Ccm Sl 1110 800 Fm

700 St Ccm Sm 1105 1105 B Fm 600 deltaic plain deltaic plain B. Marginal s f c g c Marginal lacustrine lacustrineMarginal Marginal lacustrineMarginal lacustrine 500 850 Fm Rock color Red-brown

400 475 St Brown Fm Yellow-brown Unit 1 - LACUSTRINE A. Playa Green-blue 300 Fm Coarse-grained Fm 525 470 525 s f c g c St 200 Sedimentary structures 840

XINING GROUP E Fossil horse A

Playa 100 Fm Mottling 170 Gypsum veins

Fm } s f c g c 0 Root traces s f c g c + 90 m silt fine sandcoarsegranule sandcobble E Calcareous nodules E 165 165 Trough cross-bedding Horizontal lamination Fm height (m) height Stratigraphic Stratigraphic s f c g c Figure 2. Lithologies and interpreted depositional facies and units within the measured Hualong magnetostratigraphic section from the northeastern margin of Xunhua Basin. Facies code abbreviations are described in Table 1. Expanded columns A to F illustrate typical facies assemblages for each depositional environment (see Table 2).

660 Geological Society of America Bulletin, May/June 2012 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Oligocene–Miocene basin evolution on the northeastern Tibetan Plateau

TABLE 1. LITHOFACIES AND INTERPRETATIONS USED IN THIS STUDY Lithofacies code cseD r i p t i no I n t e r p r e t a t i no E eyaL r sdebmuspygde imehC ac pl r ce i p i t a t i opavenanino r a t i lev eka Ps Carbonate with pedogenic features including nodules and adjacent root traces Soil formation on fl oodplain Fm Massive mudstone to fi ne-grained sandstone with common gray mottling and Suspension settling in lake and overbank deposits gypsum veins Sm Massive medium- to fin e - g r a i sdnasden t eno dumydnaS fl ows and suspension settling in lake and overbank deposits Sr Fine- to medium-grained sandstone with small, asymmetric, 2D and 3D current Migration of small 2D and 3D ripples under weak (~20–40 cm/s), unidirectional ripples fl ows in shallow channels St Medium- to very coarse-grained sandstone with trough cross-stratifi cation Migration of large 3D ripples (dunes) under moderately powerful (40–100 cm/s) unidirectional fl ows in large channels Sp Medium- to very coarse-grained sandstone with planar cross-stratifi cation Migration of large 2D ripples under moderately powerful (~40–60 cm/s), unidirectional channelized fl ows; migration of sandy transverse bars Sh Fine- to medium-grained sandstone with plane-parallel lamination Upper plane bed conditions under unidirectional fl ows, either strong (>100 cm/s) or very shallow Ccm Pebble to cobble conglomerate, clast-supported, unstratifi ed, disorganized to Deposition from sheetfl oods and clast-rich debris fl ows poorly sorted Cci Pebble to cobble conglomerate, clast-supported, horizontally stratifi ed, Deposition from shallow traction currents in longitudinal bars and gravel sheets imbricated, poorly sorted Cmm Massive, matrix-supported pebble to cobble conglomerate, poorly sorted, Deposition by cohesive mud-matrix debris fl ows disorganized, unstratifi ed Note: Modifi ed after Miall (1978) and DeCelles et al. (1991).

horizontal laminations (Sh, Table 1) and uncom- TABLE 2. DEPOSITIONAL FACIES AND ASSOCIATED FACIES CODES (FROM TABLE 1) mon lenses (<0.2 m thick) of matrix-supported sopeD i t i ano l f ca i se aM j o r f ca i sedocse iM on r f ca i sedocse y P.A l aya mF lF,hS,E,mS pebble conglomerate. Individual conglomer- tsucallanigraM.B led/enir nialpciat mF lF,hS,tS,mS ate and sandstone beds are laterally continuous A.C l l vu i a l f na mmC mF, mcC, tS,mS,hS over several to tens of meters, with stacked beds D. Braided fl laivu C c i , C mc , mS , mF tS,hS,mmC rS, nialpdoolF.E mF cC,hS,mcC,mS i rS,sP, forming larger bodies that are laterally continuous over tens to hundreds of meters. We interpret facies C (Table 2) to have been deposited largely by mass fl ows (Nemec and Steel, 1984) in an Sandstone beds are laterally continuous over and ~2310 m and are laterally continuous over alluvial-fan environment (Blair and McPherson, tens of meters, and bodies composed of several tens of meters (Fig. 2). Interlayers of medium- 1992) based on the general lack of stratifi cation beds are continuous over hundreds of meters. to coarse-grained sandstone and clast-supported and channelization, nonerosive bases, matrix Brown, silty, fi ne-sand interlayers are largely pebble conglomerate are 0.1–2 m thick, massive support, and textural immaturity of the deposits. featureless, with mottling, rare fl oating gran- to horizontally laminated, and laterally continu- Further investigation is needed to ascertain the ules and pebbles, and poorly developed pedo- ous over meters to tens of meters. These coarse types of mass fl ows (i.e., debris versus hyper- genic structures (Fm, Table 1). We interpret deposits exhibit poorly defi ned normal grad- concentrated) that are dominant. this facies association (facies D, Table 2) to ing and defi ne a series of minor channel fi lls. Fluvial deposits of facies D are found in represent a braided fl uvial environment (Miall, We interpret facies E (Table 2) to have been the upper half of the section at 1300–1750 m 1996), based on the widespread channelization deposited in a fl uvial fl oodplain environment and 1900–2110 m height (Fig. 2) and consist of conglomerates and sandstones together with (Collinson, 1996; Nanson and Croke, 1992) in of conglomerates and sandstones with inter- evidence of interfl uvial pedogenesis. The tex- which carbonate-rich paleosols developed on bedded silty fi ne-grained sandstone. The con- tural immaturity of the conglomerates may sug- stable or slowly aggrading fl oodplains (Mack glomerates consist of subangular granules and gest deposition by ephemeral (fl ashy) fl ooding et al., 1993). pebbles that are typically clast-supported and (Nemec and Steel, 1984). Additi onally, some We divide the 2310 m Hualong stratigraphic imbricated (Cci, Table 1). Clasts are generally conglomerate beds may have been deposited by section into three broad depositional categories 0.5–2 cm in diameter within a poorly sorted clast-rich or hyperconcentrated fl ows. based on the facies associations described here. matrix of sand, silt, and clay. Conglomer- Facies E consists of red-brown, brown, and From the bottom to the top of the section, the ate channel scours are commonly 10–50 cm yellow-brown, fi ne- to very fi ne-grained sand- Hualong units are (1) lacustrine, (2) transitional, deep, and fi ning-upward successions (from stone that is prevalent between 1750 and 1900 m and (3) fl uvial. Lacustrine unit 1 constitutes the cobble to pebble and pebble to medium sand) and 2110–2310 m in the section (Fig. 2). Minor bottom half of the section at 0–1070 m height are up to 1–2 m thick. Conglomerate beds are conglomerate and medium- to coarse-grained and is composed of facies A playa deposits 0.2–1 m thick, lenticular, and often pinch out sandstone interlayers also occur in this inter- and facies B marginal lacustrine/deltaic plain over several meters. Individual beds stack to val. The fi ne-grained sandstones are massive, deposits. We defi ne a brief interval of facies C form lenticular bodies that are 1–5 m thick and 1–20-m-thick, tabular beds that extend laterally alluvial-fan and facies B deposits at 1070–1250 laterally continuous over tens of meters. Clast for tens to hundreds of meters and contain abun- height as transitional unit 2. The upper half of roundness increases up section. Sandstones dant root traces and gray mottling with rare bur- the section (1300–2310 m) comprises fl uvial (Fig. 2) are typically medium grained and mas- rows and fl oating granules (Fm, Table 1; Fig. 2). unit 3 and contains alternating facies D braided sive (Sm, Table 1), and they contain common Uncommon calcrete and calcareous nodules stream and facies E fl oodplain deposits. Within horizontal laminations (Sh) and rare trough defi ne several prominent 30-cm-thick, white- a broad regional framework, lacustrine unit 1 cross-stratifi cation (St) and ripple marks (Sr). green beds (Ps, Table 1) at heights of ~1780 m likely correlates with the “Xining Group,”

Geological Society of America Bulletin, May/June 2012 661 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Lease et al. whereas transitional unit 2 and fl uvial unit 3 paleogeographically unique: Rocks of simi- tometer has a background noise of <1 pAm2 and collectively correlate to the “Guide Group” (for lar fi ne-grained and basement lithologies crop is equipped with computer-controlled alternat- descriptions, see Zhai and Cai, 1984; Fang et al., out on all sides of the basin (QBGMR, 1991). ing fi eld (AF) demagnetization coils and an au- 2005; Dai et al., 2006; and references therein). Despite their similar lithologic characteristics, tomated vacuum pick-and-put sample changer Paleocurrent and clast count measurements however, rocks on various sides of the basin (Kirschvink et al., 2008). Thermal demagneti- were obtained for conglomerates in units 2 and have different ages. Therefore, we use detrital zation was performed in a magnetically shielded 3 (Fig. 3); no measurements were obtained from zircon age distributions to precisely characterize ASCTM oven in a nitrogen atmosphere. unit 1 and the lower half of the section. Each of different sediment sources and track their evo- One specimen from each site plus 77 dupli- the seven paleocurrent sites contains measure- lution within the entire stratigraphic succession cates from certain sites were analyzed using ments from only ~10 imbricated pebble clasts. (see Detrital Zircon Provenance section). a combination of thermal and AF demagne- Measurements from four sites between 1360 tization techniques. All 587 specimens were and 1730 m show variable approximately south- MAGNETOSTRATIGRAPHY initially measured for natural remanent mag- westward paleofl ow ranging from west-north- netization (NRM) and then cooled in a LN2 westward to southeastward. Three paleocurrent Methods bath to remove potential multidomain viscous sites located between 1850 and 2050 m, on the magnetizations. Specimens were subsequently other hand, show a more consistent westward We collected samples for magnetic polarity subjected fi rst to AF demagnetization in 2.5 mT paleofl ow direction. stratigraphy in mudstones, siltstones, and un- steps up to 10 mT to remove low-coercivity Observations of 100 clast lithologies from commonly sandstones from the Hualong strati- magnetizations, and then to stepwise thermal each of four sites between 1100 and 2050 m graphic section (Fig. 2). We drilled 1747 cores demagnetization in 8–12 steps between a 150– reveal no major changes in clast provenance. from 510 sites spanning ~1900 m for an average 250 °C starting temperature and a 600–680 °C All sites are dominated by fi ne-grained shale sampling site spacing of 3.7 m/site. All 2.5-cm- ending temperature, typically in (1) 100–150 °C and sandstone (79%–91%) with minor base- diameter core specimens were obtained in situ steps up to 450 °C, and (2) 15–50 °C steps up ment lithologies (Fig. 3). Granite and gneiss using a gas-powered drill. Remanent magneti- to 680 °C (Fig. 4). At least one specimen from clasts together with potassium feldspar pheno- zations were measured using a 2G Enterprises every site was demagnetized, with second and crysts comprise 6%–15% of clasts, with DC SQuID three-axis cryogenic magnetometer third specimens also demagnetized if the initial 0%–6% vein quartz also present as angular to housed in a magnetically shielded room at the specimen yielded unstable directions or was subangular clasts. None of these lithologies is California Institute of Technology. The magne- bracketed by samples from adjacent sites with opposite polarity. Lithology Our 2310-m-thick Hualong magnetostratigraphic section is a composite of several shorter 2300 subsections, as is typical of many stratigraphic studies (see GSA Data Repository Fig. DR1 for Clast counts 2200 Paleocurrents sampling map1). We utilized prominent marker 60 beds to translate over short distances of a few n=100 n=11 40 2100 hundred meters or less. Remeasurement of the 20 section during detailed stratigraphic analysis

% clasts and comparison of magnetostratigraphies on ei- 0 2000 n=10 ther side of the subsections, however, revealed three errors in our initial subsection correlations. 1900 Each of these errors was made in a place where 60 n=9 Figure 3. Clast counts and paleo- n=100 correlations over large distances (~1 km) were 40 current (imbrication) measure- (m) height Stratigraphic 1800 necessary, and the errors are due to the difﬁ culty 20 of correlating over rugged terrain. We discuss

ments from the Hualong % clasts n=11 magnetostratigraphic section. 0 these stratigraphic corrections both for transpar- 1700 Paleo current measurements are ency and so that subsequent studies can reassess corrected for fold plunge fol- these correlations. We discovered an ~150-m- lowed by bedding dip; n = num- 1600 n=10 wide gap in the magnetostratigraphy centered chert ber of clasts measured. k-spar near 450 m height, an ~45-m-wide gap centered Lithology vein quartzvein at 1275 m height, and an additional ~68 m of granite / gneiss granite 1500 ms / shale schist sandstone / qtzite sandstone limestone / marble / limestone overlap at a portion of the section centered at 60 n=100 n=10 ~2000 m. The stratigraphic section that we pre- 40 1400 sent (Fig. 2) reﬂ ects these corrections. The basal 20 200 m interval of our measured section was too

% clasts n=10 0 1300 friable and dissected by gypsum veins to permit s f c g c 60 n=100 reliable magnetostratigraphic sampling. 40 1200 1 20 GSA Data Repository item 2012078, Paleomag- netic sampling map and results, detritial zircon, U/Pb % clasts 0 1100 ages, is available at http://www.geosociety.org/pubs Clast type /ft2012.htm or by request to [email protected].

662 Geological Society of America Bulletin, May/June 2012 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Oligocene–Miocene basin evolution on the northeastern Tibetan Plateau

DM126.3 West specimen (height) A NRM, LTa (vertical – Up) B NORMAL AF2.5, AF5 (1782.5 m) DM42.3 (2124.2 m) Decl. 18.0 Declination AF7.5 AF10, T150 T625, T580, T560, Incl. 32.4 Apparent T650 T600 T540 T250 inclination origin North T350 tilt-corrected T680 T520 (horizontal) REVERSED coordinates T500 T450 T600 T450 AF AF AF Decl. 186.1 T500 T350 AF –6 10 7.5 5 2.5 NRM Incl. –31.4 T520, T540 10 { T560,580 mT T500 HTC –7 T250 T600,625 5 x 10 T450 LTa AF7.5 T350 { mT NRM T500 T650 T350 LTC T600 T680 T250 T250 South origin NRM East T150 AF AF (horizontal) (vertical – Down) AF 7.5 5 2.5

C DM430.1 (657.4 m) D DM375.3 1.0 430.1 West T560, (865.1 m) NRM E (vertical – Up) T250 T580, 126.3 T350 T150 T600, AF5 T625, AF10 0.5 NRM T450 T650 + T500 AF10 T350 T150 North J/Jmax LTa AF Thermal AF7.5 origin T520 T540 (horizontal) T560 AF5 T680 T580 LTC AF2.5 T520, T540 0.0 T600 LTa T450, T500 010200 300 400 500 600 700 T350 T625 Field (mT) Temperature (°C) HTC –8 T650 5 x 10 NRM HTC T250 { mT 1.0 T150 T680 REVERSED F –6 Decl. 219.9 { 5 x 10 T250 mT Incl. –68.9 AF10 NRM 42.3 0.5 AF7.5 + 375.3 T450 NORMAL AF5 J/Jmax LTa AF Decl. 12.8 T600 AF LTC AF2.5 10 AF Incl. 46.4 LTa Thermal 5 LTa NRM 0.0 NRM North East 010200 300 400 500 600 700 origin (horizontal) (vertical – Down) Field (mT) Temperature (°C)

Figure 4. (A–D) Representative orthogonal demagnetization plots for specimens from the Hualong magnetostratigraphic section. Results are from stepwise alternating-ﬁ eld and thermal demagnetization steps. Plots show a low-temperature component (LTC) that is typically removed by 250 °C and interpreted to be a viscous overprint, as well as a high-temperature component (HTC) that is interpreted to be the characteristic remanent magnetization. Note that the intensity scale varies ~100-fold. (E–F) Decay of magnetic moment for the same specimens during demagnetization. NRM—natural remanent magnetization, LTa—“low-temperature” step (samples cooled in liquid nitrogen), AF—alternating ﬁ eld (in mT), T—temperature (in °C).

Results refl ect the characteristic remanent magnetiza- difference in remanent direction is observed tion (ChRM). Progressive unblocking of the when comparing the 250–580 °C portion of The intensity of the NRM for Hualong speci- high-temperature component between 250 °C the unblocking spectrum with the 600–680 °C mens is typically 10−2 A/m, with a range of and 580 °C suggests a magnetite carrier. Com- portion (Fig. 4). This behavior suggests that −1 −3 10 –10 A/m. After LN2 treatment, specimens plete unblocking of the high-temperature com- both magnetite and hematite recorded the same typically displayed a drop in magnetic intensity ponent by 680 °C, however, indicates that a paleo magnetic fi eld when their remanent direc- of <15%, indicating that no appreciable rema- hematite carrier is also present (Fig. 4). Fur- tions became fi xed in the rock. nence was carried by multidomain magnetite thermore, coercivity spectrum analyses of iso- ChRM directions were determined for each or hematite (Figs. 4E and 4F). Our stepwise de- thermal remanent magnetism (IRM; Fig. DR2 specimen using principal component analysis magnetization procedure clearly resolves a low- [see footnote 1]) for selected Hualong speci- (Kirschvink, 1980) as implemented in Paleo- coercivity, low-temperature component and a mens generally show gradual, continual acqui- Mag 3.1.0 b1 (Jones, 2002), typically from nine high-temperature component (Fig. 4). The low- sition of IRM with increasing fi eld strength—a points (minimum 3 points, maximum 20) from temperature component is typically removed by behavior consistent with a hematite carrier. A the stable high-temperature component. The ori- 250 °C (Figs. 4B, 4C, and 4D), but sometimes small drop in remanence at ~580 °C in many gin is only included if the demagnetization path not until 450 °C. The low-temperature compo- specimens (Fig. 4E) and a protracted inter- decays toward it. We report ChRM directions nent does not typically decay toward the origin val of relatively rapid IRM acquisition before in stratigraphic coordinates that are corrected and is interpreted to be a viscous overprint. ~70 mT for some IRM samples (Fig. DR2 [see for fold plunge followed by bedding tilt (Fig. 5) The high-temperature component decays footnote 1]), however, suggest a magnetite car- because we mapped a series of folds within the toward the origin, typically exhibits stable be- rier as well. When both magnetite and hematite Hualong sections with axes that plunge ~25° in havior until 650–680 °C, and is interpreted to are present in the same specimen, no signifi cant the direction of N17°E. Paleocurrent imbrication

Geological Society of America Bulletin, May/June 2012 663 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Lease et al.

Fold-corrected Geographic Fold-corrected + Tilt-corrected Hemisphere: NORMAL: NORMAL: NORMAL: lower Decl: 21.4 Decl: 20.6 Decl: 8.8 upper Incl: 55.2 Incl: 30.2 Incl: 32.3 α95=3.7 α95=3.3 α95=2.9 K=8.1 K=8.0 K=10.5 N=252 N=257 N=256

REVERSED: REVERSED: REVERSED: Decl: 220.9 Decl: 213.3 Decl: 200.7 Incl: -52.7 Incl: -28.8 Incl: -38.4 α95=4.6 α95=4.8 α95=3.8 K=6.0 K=5.81 K=8.4 N=194 N=189 N=190

Figure 5. Stereographic equal-area plots of characteristic remanent magnetization directions from the Hualong section in geographic, fold- corrected, and fold- and tilt-corrected coordinates. Black circles are plotted in the lower hemisphere, and white squares are plotted in the upper hemisphere. Larger circles indicate α95 error around the Fisher mean with mean data listed adjacent to each plot. Different N values reﬂ ect a few low-latitude specimens that change polarity when fold and tilt corrections are applied. The Hualong data fail the reversal test (McFadden and McElhinny, 1990), likely due to an unremoved partial normal overprint that biases reversed directions to more westerly orientations.

measurements are also subject to this two-stage curring at 108% untilting, with 95% confi dence et al. (2008). Given the signifi cant remanence untilting procedure. limits of 97%−122% untilting (Fig. 6). It is retained by some samples even at 680 °C, e.g., Virtual geomagnetic poles (VGP) were then worth noting that without the additional correc- sample DM430.1 (Fig. 4C), a residual overprint calculated and their latitude used to denote tion for the fold plunge, the data set fails the fold is not unexpected. In addition, inclination shal- either a normal polarity or reversed polarity test. Clearly, for folds with plunges >5°–10°, a lowing, common in central Asia sedimentary specimen. Specimens with a noisy or ambigu- two-step unfolding procedure is warranted. strata, e.g., Gilder et al. (2003), is suggested for ous orthogonal demagnetization diagram or a The Hualong paleomagnetic data set, how- the Hualong paleomagnetic data set by mean maximum angular deviation (MAD) >15% are ever, has a negative reversal test (McFadden and corrected inclinations (32°–38°; Fig. 5) that are excluded from further analysis, thus removing McElhinny, 1990), with the paleomagnetic data much lower than expected: 56°–59°, calculated 23% of the measured specimens. The average set (Fig. 5) failing both as a whole and when from late Oligocene–late Miocene Eurasian MAD for resulting specimens is 6°. A positive divided into 10 regularly spaced sub–data sets reference poles in Besse and Courtillot (2002). fold test (Tauxe and Watson, 1994, as imple- (7 of 10 sub–data sets fail). The negative re- Given that inclination shallowing occurs early mented in Tauxe et al., 2009) attests to the reli- versal test suggests an unremoved overprint. during sedimentation, its presence in the Hua- ability of our demagnetization data in defi ning Reversed directions are biased toward more long data set suggests a primary origin for the the depositional remanence (Fig. 6). Progressive westerly directions, and normal directions are measured magnetic remanence. untilting of the individual fold-corrected ChRM biased toward more northerly directions, as ex- Given the success of the fold test, the pres- directions about the local bedding orientation pected from the effect of an unresolved partial ence of inclination shallowing, and the fact that shows the maximum clustering of directions oc- normal overprint, e.g., fi gure 7 of Dupont-Nivet normal and reversed directions all plot within their respective hemispheres (indicating that the sample populations are qualitatively antipodal), we accept our polarity determinations despite Fold test 97 122 the failure of the reversal test. We acknowledge 0.82 Figure 6. Positive fold test for Hualong char- that a partial overprint remains unremoved, acteristic remanent magnetization (ChRM) however, and suggest that it would be unwise directions shown by the 95% confi dence in- to utilize the Hualong paleomagnetic data set to terval for τ overlapping with 100% untilt- 1 0.78 study rotations within the basin. The success of ing of ChRM vectors corrected for strata the fold test and the failure of the reversal test, orientation (Tauxe et al., 2009; Tauxe and 95% when considered together, suggest that the par- Watson, 1994). The τ1 maximum refl ects the (”tightness”) 0.74 confidence tial normal overprint affected many Hualong tightest grouping of the ChRM directions 1 τ interval sites and likely occurred after deposition of the during progressive untilting of the strata youngest unit within the section. and is defi ned by the eigenvalues of the ori- 0.70 We constructed a detailed magnetic polarity entation matrix. 0 40 80 120 160 stratigraphy based on the VGP latitudes from % untilting the Hualong paleomagnetic data set (Fig. 7).

664 Geological Society of America Bulletin, May/June 2012 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Oligocene–Miocene basin evolution on the northeastern Tibetan Plateau

We identifi ed 33 normal magnetozones and 31 GPTS reversed magnetozones in the Hualong sec- Lithology VGP latitude MPS tion, with each magnetozone defi ned by two or UNIT FACIES * 9 more specimens that display the same polarity 2300 * and have VGP latitudes >30°. A jackknife re- * (Ma) Age E E 10 sampling of our magnetostratigraphy affi rms the 2200 C5n. Floodplain robustness of our data set, returning a J value Floodplain Chron of −0.40, which lies within the recommended 2100 11 range of 0 to −0.5 (Tauxe and Gallet, 1991; Fig. C5r. DR3 [see footnote 1]). This result suggests that D 2000 12 we have recovered more than 95% of the true Braided number of magnetochrons. Additionally, a sta- 1900 C5Ar. tistical assessment (Johnson and Mcgee, 1983) 13 of the number of reversals expected with our site E 1800 C5ABn Flood Flood - plain density (63) is commensurate with the number - plain 14 of reversals we detected (65). Questionable re- 1700 C5ADn versals defi ned by only a single specimen are shown with a half-bar (Fig. 7). 15 1600 The correlation of the Hualong magnetic Unit 3 - FLUVIAL C5Br polarity stratigraphy to the geomagnetic polarity 16 time scale (GPTS: Ogg and Smith, 2004) is D 1500 C5Cn. facilitated by our discovery of an in situ fossil Braided jawbone complete with teeth from a Miocene 1400 17 horse. We recovered the fossil within our mea- C5Dn sured stratigraphic section during the course of GUIDE GROUP 1300 18 sample collection. The teeth (Fig. 8) are iden- s f c g c -90 0 90 C5En tifi ed as belonging to the Hipparion weihoense B B 1200 plain plain Deltaic taxon with an age of 9.7–8.7 Ma (early part of Deltaic 19 the late Miocene, NMU 9 of Bahean or MN 10 C6n Unit 2 - C of Valle sian; Deng, 2008, personal commun.). fan 1100 Alluvial 20.3 TRANSITIONAL 20 The presence of the fossil specimen within Ma C6An. our magnetostratigraphic section is fortuitous, 1000 because it eliminates potential error that arises 21 when correlating fossil-bearing strata across C6AAr. 900 several to tens of kilometers, which is a particularly valid concern in the Xunhua region due to 22 800 C6Br the presence of rapid lateral variations in both C6Cn. facies and thicknesses, e.g., Fang et al. (2003). 23 The Hipparion weihoense specimen serves as 700 C6Cr

an independent chronostratigraphic tie point that B anchors the top of the Hualong magnetostratig- 24 600 C7n deltaic plain deltaic raphy to the GPTS (Fig. 7). For our preferred plain plain deltaic Marginal lacustrineMarginal correlation of the Hualong magnetic polarity lacustrine lacustrineMarginal 25 500

26 400 Figure 7. Hualong section magnetic polarity

stratigraphy (MPS) and its correlation to Unit 1 - LACUSTRINE 300 27 the geomagnetic polarity time scale (GPTS; C9n Ogg and Smith, 2004). Stratigraphic logs of virtual geomagnetic pole (VGP) latitudes, 200 28 lith ologies, and depositional units are shown XINING GROUP A on left. Each magnetozone in the magnetic Playa 100 29 polarity stratigraphy is deﬁ ned by two or more specimens (typically from multiple 0 30 sites, but sometimes from one site) of the s f c g c -90 0 90 ? silt fine sandcoarsegranule sandcobble same polarity that have VGP latitudes >30°. Age (Ma) Questionable reversals defined only by a Chron single specimen measured from one site are height (m) height

labeled with half-bars. Stratigraphic

Geological Society of America Bulletin, May/June 2012 665 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Lease et al.

Figure 8. Hipparion weihoense Hipparion weihoense NMU 9 of Bahean or MN 10 of Vallesian rupted until ca. 2 Ma, although the possibility teeth discovered from the top that Hualong was abandoned as a depocenter of the upper Hualong section at ca. 9 Ma as local deformation propagated into ~2250 m. The fi ve cheek teeth the basin cannot be eliminated. If deposition did belong to the same individual: continue in Hualong from 9 until 2 Ma, extrapo- right upper second, third, and M1 P4 P3 P2 lation of accumulation rates and facies from our fourth premolars and first early Late Miocene upper Hualong section and the upper Xunhua molar (P2, P3, P4, M1); and 2 cm p4 9.7-8.7 Ma section (Hough et al., 2011) suggests an addi- the right lower fourth pre molar tional ~1000 m of fl uviolacustrine deposition (p4). The teeth are charac- during this time. terized by strongly convex labial walls, rounded and long double-knots (metaconid and Accumulation rates derived from the two metastylid), more lingually shifted than entoconid, and simple hypoconulid, without clear different decompaction routines (9 Ma versus subdivision into two lobes. Hipparion weihoense is a taxon from the early part of the late 2 Ma) display similar relative values and trends, Miocene (NMU 9 of Bahean or MN 10 of Vallesian) with an age of 9.7–8.7 Ma (Deng, 2008, differing only in absolute values by 11%–29%. personal commun.). Relatively steady accumulation at a rate of ~170 m/m.y. in Hualong between 27 and 24 Ma was followed by a distinct period of accelerated ac- stratigraphy to the GPTS, the magnetostratig- chron, (2) the Hualong succession had depo- cumulation between 24 and 21 Ma with rates of raphy spans 27.8–9.3 Ma. Depositional ages sitional hiatuses during these short chrons, or ~200–310 m/m.y. (Figs. 9A and 9B). By 21 Ma, are assigned to the described lithologic units as (3) samples from these short, reversed polarity rates dropped to ~120 m/m.y. and remained follows: lacustrine unit 1 spans 30–20.3 Ma, intervals were later overprinted with normal po- steady until 12 Ma, when a distinct accelera- transitional unit 2 spans 20.3–18 Ma, and fl uvial larity. Second, we observe three magnetozones tion raised rates to ~200 m/m.y. until the top of unit 3 spans 18–9.3 Ma. Our correlation places within the Hualong magnetic polarity stra- the section at 9 Ma. The average decompacted the boundary between the Xining Group and tigraphy that have no correlative chron on the accumulation rate for the entire section is 150 Guide Group at 20.3 Ma. GPTS: reversed magnetozones within C6AAn m/m.y. (9 Ma routine) to 180 m/m.y. (2 Ma rou- The correlation of the Hualong magnetic po- and C5Dn, plus a normal zone within C6Cn.1r. tine). Sediment accumulation curves that have larity stratigraphy to the GPTS between 27.8 Assuming steady deposition on 1.5 m.y. time not been corrected for compaction also show and 25 Ma contains some uncertainty because scales, the uncorrelated magnetozones have distinct accelerations in rate at 24 and 12 Ma, of the presence of a gap in our magnetostrati- estimated durations of 20–100 k.y. These un- with an average rate for the entire section of graphic sampling (between 370 and 520 m expected magnetic polarity stratigraphy mag- ~110 m/m.y. (Figs. 9A and 9B). We used com- height in the section), plus several questionable neto zones may be invalid and due to errors pacted accumulation rates when comparing the single-site polarity reversals (Fig. 7). Some of during sampling or laboratory analysis, or, Hualong section with sections from adjacent the single-site reversals, however, might cor- alternatively, they might refl ect unrecognized basins to maintain consistency with them (see relate with short-duration “cryptochrons” rec- cryptochrons within the GPTS. Discussion). ognized within C7n and C9n (Cande and Kent, To discriminate between the effects of tec- 1995; Tauxe and Hartl, 1997). Our proposed SUBSIDENCE tonic versus sediment loading, we performed correlation for the 27.8–25 Ma portion of the a one-dimensional Airy backstripping routine section assumes no major accumulation rate Both the detail and duration of the Hualong with exponential reduction of porosity using the changes and no major depositional hiatus: These magnetostratigraphy make it ideal for deter- method of Allen and Allen (2005) and sediment assumptions are supported by the continuity of mining variations in sediment accumulation porosity values of Sclater and Christie (1980) the overlying 25–9.3 Ma magnetostratigraphic rates for the northern Xunhua Basin from 27 to as implemented in the OSXBackstrip program record, as well as the absence of fi eld evidence 9 Ma. We decompacted sediment thicknesses version 2.6 (Cardozo, 2009). We recognize that for an unconformity. according to their observed lithologies follow- basin fl exure was not the only control on Xun- Finally, we acknowledge several, short-lived ing the methods and porosity values of Sclater hua sediment accumulation rates; tectonic dam- (i.e., <100 k.y.) polarity reversals that do not and Christie (1980). We performed two separate ming behind growing structures may also have correlate between the GPTS time scale and decompaction routines—one with deposition been signifi cant, particularly for the 12–9 Ma the Hualong magnetic polarity stratigraphy ending at ca. 9 Ma (the top of our measured Hua- portion of the section. Thus, our backstripping record. We discuss these reversals later herein; long section) and a second with an additional results provide estimates of maximum sediment however, their presence does not change the 1000 m of deposition ending at ca. 2 Ma. Though and tectonic loading rates; damming would magnetostratigraphic correlation. First, there the youngest strata in the Hualong section are diminish loading rates. We assumed sediment are fi ve, short GPTS chrons that have no cor- 9 Ma, deposition in southern Xunhua and adja- loading only (e.g., no water loading) because of relative magnetozone in the Hualong magnetic cent basins continued until ca. 3–1.8 Ma (Fang the terrestrial, arid setting of Xunhua Basin. We polarity stratigraphy: reversed chrons C4Ar.3r, et al., 2003, 2005; Hough et al., 2011), when the also made the simplifying assumption that the C5Ar.2r, C6Bn.1r, and C7n.1r, plus normal Yellow River began incising headward through elevation of the basin has remained constant for chron C5r.2r-1. These uncorrelated GPTS the region (Craddock et al., 2010; Harkins et al., the duration of accumulation, given a lack of in- chrons, which have durations of 30–60 k.y., 2007) and evacuating basins. The separate 9 Ma dependent paleoelevation information. could be due to the following: (1) Our sampling versus 2 Ma decompaction routines illuminate Our backstripping results show acceler- missed these chrons because the spacing be- the range of possible late Miocene–Pliocene ated subsidence until 21 Ma, when peak rates tween superposed samples represents a longer depositional histories for Hualong; it is likely dropped three- to fourfold compared to the maxi- period of time than the short duration of each that Hualong deposition continued uninter- mum rates at ca. 23 Ma (Fig. 9C). A protracted

666 Geological Society of America Bulletin, May/June 2012 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Oligocene–Miocene basin evolution on the northeastern Tibetan Plateau

Figure 9. (A) Age versus thickness for the 0 Hualong magnetostratigraphic section. A 157 Curves depict both compacted and decom- 500 93 167 pacted lithologies, plus separate decompac- Accumulation 111 222 tion routines run for deposition ending at Compactedcompacted ± 4% 1000 rates (m/m.y.) 75 132 9 Ma and at 2 Ma. The latter assumes an 102 112 ± 3% extra 1 km of deposition prior to 2 Ma. Ac- 167 Decompacted-9decompacted-9 Ma 1500 184 cumulation rates are slopes of linear regres- 96 132 Decompacted-2decompacted-2 Ma 286 216 sions with % error indicating the α95 limits ± 15% of regression. Vertical dashed lines are divid- 2000 ing lines between regressions. Note the accel- 162 327 ± 10% Thickness (m) erated accumulation rates at 24–21 Ma and 2500 ± 13% at 12–9 Ma. (B) Age versus accumulation 181 Sediment Accumulationaccumulation 3000 rate for the Hualong section. Blocky data ± 11% refl ect rates for each correlated magneto- 3500 zone, and smoother data refl ect a 2 m.y. 27 25 23 21 19 17 15 13 11 moving average. (C) Age versus subsidence Age (Ma) for the Hualong section from backstripping rapid slow rapid analysis. Subsidence rates are slopes of lin- B 900 ear regressions with % error indicating α 95 800 Sediment Accumulationaccumulation limits of regression. Range of rates for each age section refl ects the separate “9 Ma” and 700 Smoothing “2 Ma” decompaction routines described in Accumulation rates 600 window (2-m.y.) A. (D) Hualong section lithologies and depo- Compacted sitional units versus age for reference with 500 Decompacted-9 Ma accumulation and subsidence plots above: A, B, and C. Note that (1) rapid accumula- 400 Decompacted-2 Ma tion rates at 24–21 Ma are coincident with 300 deposition of fi ne-grained lacustrine unit 1; (2) progradation of coarse-grained units 2 200 and 3 is coincident with slow rates, suggesting that underfi lled foreland deposition was (m/m.y.) rate Accumulation 100 controlled by subsidence and not supply; 0 and (3) thickness scale is nonlinear. 25 23 21 19 17 15 13 11 Age (Ma) 0 C 71–78 100–110 250 53–60 period of rapid subsidence starting at 27 Ma 25–2825 Tectonic subsidence 154 9 Ma 23 intensifi ed from 24 to 22.5 Ma and remained 500 23–26 –171 28 2 Ma 36–45 accelerated, though slower, until 21 Ma. Sub- 750 ± 14% 236 26 sidence rates then decreased and remained stable –264 72 from 21 to 12 Ma before accelerating 50%–75% 1000 130 ± 14% from 12 to 9 Ma. –148 9 Ma 1250 72–82 ± 11% 2 Ma 72 Our interpretations of fi rst-order changes in ± 16% Total subsidence subsidence rates at ca. 24, 21, and 12 Ma are Depth (m) 1500 ± 16% insensitive to modifi cations of the backstrip- 72–83 1750 83 ping routine such as the addition of a water load Subsidence ± 4% Subsidence history ± 4% 107–139 or an overlying Miocene–Pliocene sedimentary 2000 rates (m/m.y.) load. Inclusion of a large, 100-m-deep water ± 4% 2250 load creates only a 1% increase in the magni- 27 25 23 21 19 17 15 13 11 tude of the relative changes in subsidence rate, Age (Ma) although this water load increases the abso- Unit 2 - Unit 3 - Fluvial lute subsidence rates by 43%–45% uniformly UNIT Unit 1 - Lacustrine Transitional throughout the section. Adding an overlying B C B D E D E D FACIES Alluv. Delta Flood Braid Flood 1000-m-thick, 9–2 Ma sedimentary load, as Marginal Lacustrine/Deltaic plain Fan Plain Braided Plain -ed Plain explored with our 2 Ma decompaction routine, Conglomerate gives relative subsidence values and trends that Sandstone are similar to our 9 Ma routine (without the Siltstone 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 Stratigraphic 400 500 600 700 800 900 addi tional load), although the results differ ab- height (m) solutely by 10%–30% (Fig. 9C).

Geological Society of America Bulletin, May/June 2012 667 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Lease et al.

DETRITAL ZIRCON PROVENANCE U-Pb geochronology of detrital zircons was tinctive age signature but instead are dominated conducted by laser ablation–multicollector– by ca. 450 Ma zircons, likely from abundant Methods inductively coupled plasma–mass spectrom- granitic dikes within the unit. etry (LA-MC-ICP-MS) at the University of Modern catchments draining the West Qin- Provenance analyses using single-grain Arizona LaserChron Center (Gehrels et al., ling are characterized by a discrete ca. 250 Ma detrital zircon U-Pb age determinations have 2006). Measured ages and isotopic ratios are zircon population derived from Triassic plutons proven particularly useful in areas where two or reported in the GSA Data Repository (see (Tp) and a broader >1500 Ma population of re- more geographically distinct source areas have footnote 1). cycled zircons derived from Triassic Songpan- similar lithologies but have distinguishable U-Pb Garze fl ysch (Tr; Fig. 10B; Lease et al., 2007). age signatures (DeCelles et al., 2004). The uplift Results I—Modern Source Terranes The Triassic Songpan-Garze unit (with its and erosion of one range relative to other ranges Detrital zircon age populations from modern >1500 Ma Tr zircons) constitutes the majority can be assessed by examining detrital zircon catchments can be tied to discrete source ter- of exposed rock in the West Qinling (Fig. 1B); age spectra from sediments deposited in basins ranes within adjacent ranges. Modern catch- however, Triassic fl ysch is not found immedi- adjacent to these ranges. Age spectra from sev- ments draining the Laji-Jishi-Riyue Shan are ately north, across the south Qilian suture (Yin eral levels within a stratigraphic succession that dominated by a 450 Ma zircon population de- and Harrison, 2000). North of the West Qinling, spans several million years can track the evolu- rived from early Paleozoic plutonic and Cam- Triassic fl ysch is only present in a restricted area tion of source terranes and detect the emergence brian volcanic units (labeled Pp and Cv), with of the Riyue Shan (Fig. 1B), a distal area that is of newly uplifted ranges (Lease et al., 2007). a subsidiary ca. 850 Ma (or 500 to 1000 Ma) not a likely source area for the Xunhua or Linxia We collected ~5 kg sandstone samples from zircon population derived from a Silurian sedi- Basins, which are our focus. Thus, we interpret each of six stratigraphic horizons within our mentary unit (S) that is exposed locally (Fig. >1500 Ma Triassic fl ysch zircons found within measured section from Hualong (Fig. 2). The 10A; Lease et al., 2007). Our new analyses from the late Oligocene–Miocene Xunhua and Linxia age of each horizon is well constrained based catchments draining Proterozoic units (samples Basin fi lls to be derived primarily from a proxi- on our detailed magnetostratigraphy, and overall Pr2 and Pr3, Fig. 10A) show that they lack a dis- mal, West Qinling source. these ages range from 28 to 10 Ma (Fig. 7). We also collected samples from fi ve horizons within the Wangjiashan section in Linxia Basin, which has ages ranging from ca. 25 to 7.5 Ma based A Laji-Jishi-Riyue Shan samples on magnetostratigraphic correlation (2–11 Ma; Pp and Cv - early Paleozoic plutonic and volcanic rocks Fang et al., 2003) and lithostratigraphic correla- 4 (n=95)* tion to the Maogou section (11–25 Ma; Li et al., 1997); samples from two of these horizons were South -east -east analyzed previously (Lease et al., 2007). To pre- 5 (n=54)* clude potential biasing of the zircon distribution K1 (n=97) by a poorly mixed depositional environment

(DeGraaff-Surpless et al., 2003), each sample JISHI consists of several subsamples collected a few Figure 10. Detrital zircon U/Pb S - Silurian strata Pr2 (n=99) meters apart within the same stratum. We report age distributions from samples 6 (n=94)* detrital zircon age spectra from typically >95 within modern catchments (loca- LAJI LAJI single grains per sample. The Hualong strati- tions in Fig. 1B) in the (A) Laji- Pr3 (n=50) graphic succession is well suited for an analysis Jishi-Riyue Shan and (B) West ? of local tectonism because it lies along the Laji- Qinling. Modern Laji-Jishi Shan 8 (n=96)* Jishi Shan range front and, therefore, is more samples are dominated by ca. likely to contain locally derived sediment and 450 Ma and 500–1000 Ma zircon record the growth of nearby ranges. populations, which can be used 7 (n=51)* RIYUE RIYUE To better characterize detrital zircon age to distinguish detritus eroding 16 (n=99)*(n=99)* North -west signatures currently being supplied by various from the range. Modern West North -west Tr Laji-Jishi Shan local rock units (source terranes) in the Laji- Qinling samples, in contrast, are mean ((n=735)n=735) characterized by ca. 250 Ma and Jishi-Riyue Shan, we supplemented published probability relative Normalized B West Qinling samples work (Lease et al., 2007) with 294 additional >1500 Ma populations. Distribu- single-grain analyses from four new samples tions are shown as probability Tp - Triassic plutonic rocks (Fig. 1B). Each sample is composed of ~5 kg of density functions. sand from a modern catchment that drains one 4x (n=48)* or more lithologic units. Our four new samples East K2 (n=48) are from small catchments (<4 km2) that are dominated by Proterozoic or Cretaceous units, 15 (n=93)* neither of which has been documented to have a 14 (n=97)* West distinctive age signature. We sampled along the West Tr - Triassic flysch West Qinling mean (n=286) range front to avoid contamination by recycled zircon population from recently eroded Neo- 0 250 500 1000 1500 2500 gene strata. Age (Ma) * from Lease et al. (2007)

668 Geological Society of America Bulletin, May/June 2012 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Oligocene–Miocene basin evolution on the northeastern Tibetan Plateau

Our new analyses from catchments draining A Source terranes the Cretaceous unit in both the West Qinling Pp and Cv – Early Paleozoic plutonic and volcanic rocks (sample K2; Fig. 10B) and Laji-Jishi-Riyue Figure 11. (A) Age distribu- Shan (sample K1; Fig. 10A) show that this unit tions from source terranes in S – Silurian strata Laji-Jishi Shan exhibits a composite age spectra including all the Laji-Jishi Shan and West four documented local source terranes, and thus Qinling shown for reference. Tp – Triassic plutonic rocks Triassic Tr – flysch West Qinling the Cretaceous unit cannot be distinguished as a (B–C) Detrital zircon U/Pb age unique source terrane. In addition, during min- distributions from magneto- B eral separation of modern detrital zircon sam- stratigraphic horizons within Xunhua (Hualong) magnetostratigraphic section ples, we observed that samples draining only the (B) northern Xunhua (Hua- Stratigraphic age Cretaceous bedrock (samples K1 and K2) had long) Basin and (C) western 10 Ma (n=96) grain yields that were orders of magnitude lower young Young Linxia Basin. Note enhanced 13.5 Ma (n=97) than yields from samples draining plutonic and abundance of Laji-Jishi Shan other lithologies. The much lower abundance of zircons in Xunhua Basin by 19 Ma (n=98) zircon grains from Cretaceous sources relative 19 Ma and in Linxia Basin by to other potential sources suggests that recycled 11.5 Ma, indicating intensifi ed zircons from Cretaceous sources are volumetri- erosion of the Laji Shan and 22.5 Ma (n=95) cally insignifi cant and that late Oligocene– Jishi Shan, respectively. Wide, Miocene basin strata are dominated by other light-gray line in background basement sources with a greater abundance of portrays evolution of dominant 24.5 Ma (n=98) zircon grains. zircon populations and high- Altogether, detritus derived from the Laji- lights the doubling of the ca. old Old 28 Ma (n=99) Jishi-Riyue Shan, which lie to the north and 450 Ma Laji-Jishi Shan popula- east of Xunhua Basin, is best characterized by probability relative Normalized tion by the middle of each sec- C Linxia (Wangjiashan) magnetostratigraphic section a 450 Ma early Paleozoic population that con- tion. Linxia Basin samples were Stratigraphic age stitutes 62% of all modern detrital zircons dated collected from the Wangjiashan 7.5 Ma (n=96)* young from the range (Fig. 10A). This 450 Ma popula- section with age assignments Young tion constitutes 78% of Laji-Jishi-Riyue zircons based on magnetostratigraphy 11.5 Ma (n=101)* when 2 of 9 modern samples are excluded from (2–11 Ma; Fang et al., 2003) 14.5 Ma (n=96) the range average: (1) sample 16 from the Riyue and lithostratigraphic corre- 21.5 Ma (n=99) Shan, which contains Triassic fl ysch zircons that lation to the Maogou section do not represent a viable northern source for the (11–25 Ma; Li et al., 1997). Dis- Xunhua and Linxia Basins, and (2) sample K1, tributions are shown as prob- which contains indistinguishable, volumetri- Old ability density functions. old 25 Ma (n=97) cally insignificant zircons from Cretaceous 0 250 500 1000 1500 2500 strata. In comparison, detritus derived from the Age (Ma) * from Lease et al. (2007) West Qinling and Zamazari Shan, which lie to the south and west of Xunhua Basin, is best characterized by a 250 Ma Triassic plutonic population (43% of all modern detrital zircons) sources in the West Qinling were supplanted by Modeling of the detrital zircon age distribu- and >1500 Ma population (29%) from Triassic sources in the Laji Shan–Jishi Shan (Fig. 11B). tions from Hualong basin strata quantifi es a fl ysch (Fig. 10B). To determine the combination of source ter- signifi cant provenance change after 22.5 Ma ranes that best defi nes the age distribution of (Fig. 12A). Age distributions from 28, 24.5, and Results II—Late Oligocene–Miocene each stratum (Fig. 11B), we used a sediment- 22.5 Ma Hualong strata are statistically indistin- Provenance mixing model (Amidon et al., 2005; Lease et al., guishable (Kolmogorov-Smirnov P > 0.08, 80 In the context of these contrasting source 2007) that includes our four primary source m.y. smoothing window) and are dominated by areas , the emergence of the eastern Laji Shan and terranes (Fig. 11A): Triassic plutonic (Tp; ca. ~75% zircons derived from the two West Qin- Jishi Shan as important source areas can be dis- 250 Ma), Triassic fl ysch (Tr; >1500 Ma), early ling units: Triassic plutonic (Tp) and Triassic cerned by comparing the U/Pb age distributions Paleozoic plutonic and volcanic (Pp and Cv; fl ysch (Tr). A striking provenance change by in strata from the northern edge of the Xunhua ca. 450 Ma), and Silurian sedimentary (S; 500– 19 Ma is highlighted by the doubling of zircons Basin contained within our Hualong magneto- 1000 Ma) units. For the range of possible source derived from the two Laji-Jishi Shan units— stratigraphic section (Fig. 11B). The age spectra combinations, we modeled synthetic age distri- Paleo zoic (Pp and Cv) and Silurian (S)—to for each stratigraphic horizon can be examined butions by mixing the source terranes and then ~50% of the distribution (Fig. 12A). This against a background of ca. 250 and >1500 Ma iteratively calculating the synthetic age distribu- strengthening of the Laji-Jishi Shan as a source zircon age populations from the West Qinling, tion that gave the minimum mismatch to the ob- area for Xunhua Basin is sustained throughout a range which has been eroding at a relatively served age distribution for a given stratum (Fig. the upper Hualong section, with samples from steady rate of 0.1 mm/yr since ca. 45–50 Ma 12). The Kolmogorov-Smirnov test returned 19, 13.5, and 10 Ma strata being statistically (Clark et al., 2010). The Hualong zircon spectra high P values (>0.99) for the best-fi t samples identical ( Kolmogorov-Smirnov P > 0.44, 80 reveal a major transition in the northern Xun- and indicates that the observed and synthetic age m.y. smoothing window). The enhanced contri- hua Basin between 22.5 and 19 Ma as dominant distributions are statistically indistinguishable. bution from the Laji-Jishi Shan starting between

Geological Society of America Bulletin, May/June 2012 669 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Lease et al.

Figure 12. Source-terrane mix- Erosion of Laji Shan Erosion of Jishi Shan tures (Fig. 11A) that best match 100 100 Detrital zircon zircon-age distributions (Figs. source terranes 80 80 11B and 11C) from selected Laji-Jishi Shan magnetostratigraphic horizons 60 S - Silurian seds. 60 from Xunhua (A) and Linxia Pp - Early Paleozoic 40 40 (B) Basins. Mixtures were de- West Qinling termined from our four-com- 20 Tp - Triassic plutons 20 Source terrane (%) Source terrane (%) ponent sediment-mixing model Tr - Triassic flysch 0 0 (Amidon et al., 2005; Lease 2824.5 22.5 19 13.5 10 ~2521.5 14.5 11.5 7.5 et al., 2007), run with a smooth- A Age (Ma) of Xunhua (Hualong) strata B Age (Ma) of Linxia (WJS) strata ing window of 80 m.y. Note in- tensiﬁ ed erosion of the Laji Shan expressed in Hualong (Xunhua) strata after 22.5 Ma and intensiﬁ ed erosion of the Jishi Shan expressed in Linxia strata after 14.5 Ma. WJS—Wangjiashan section.

22.5 and 19 Ma indicates that the Paleozoic and GPTS between 2 and 29 Ma (Fang et al., 2003; determination of zircon age distributions from Silurian source terranes in the Laji Shan were Li et al., 1997). Furthermore, discrepancies be- specifi c strata of known age within the Xunhua exposed and actively eroding by ca. 20 Ma. We tween fossil ages and magnetostratigraphy in and Linxia Basin magnetostratigraphies al- attribute this exposure to accelerated rock uplift Linxia have caused some vertebrate paleontolo- lows us to track the evolution of plateau margin in the Laji Shan at or before this time. gists to suggest errors as great as 4–8 m.y. in the sources and thus tectonism throughout the late Detrital zircon age distributions (Fig. 11C) late Oligocene–Miocene portion of the Linxia Oligocene–Miocene (Figs. 13 and 14). from the Wangjiashan magneto- and lithostrati- magnetostratigraphic correlation (Deng et al., Additionally, the Xunhua Basin lies at a graphic section in western Linxia Basin (Fang 2004; Garzione et al., 2005). Regardless of the crossroads between studied sections in Linxia, et al., 2003; Fig. 1B) show a major provenance details, Linxia Basin does contain an invalu- Xining, and Guide Basins (Fig. 1B), and con- change after 14.5 Ma, with an abrupt doubling of able late Cenozoic stratigraphic record that has sequently analysis of Xunhua deposition illu- Laji-Jishi Shan zircons by 11.5 Ma (Fig. 12B). served as a useful time line for continental cli- minates broader, regional depositional patterns. Age distributions for the 25, 21.5, and 14.5 Ma matic records (Dettman et al., 2003; Fan et al., Our detailed Xunhua magnetostratigraphic Linxia strata are each dominated by ~75% zir- 2007, 2006; Garzione et al., 2005; Wang and (Fig. 7), sedimentologic (Fig. 2), and sub sidence cons from the two West Qinling units—Trias- Deng, 2005). The Hualong section in northern (Fig. 9) analyses, when compared to adjacent sic plutonic (Tp) and Triassic fl ysch (Tr)—with Xunhua Basin, however, provides a contempo- rec ords, clearly defi ne signifi cant lateral varia- only a minor 25% of zircons from the Laji-Jishi raneous record where sub-million-year Miocene tions in both depositional environments (Figs. Shan units. A striking provenance change by stratigraphic age constraints are known with 13 and 14) and accumulation rates (Fig. 15) in 11.5 Ma is highlighted by the doubling of zir- greater fi delity than in Linxia. Consequently, this region of Tibet. Furthermore, detrital zircon cons derived from the Laji-Jishi Shan units: we exploit the high-resolution Hualong mag- provenance records from Xunhua and Linxia Paleozoic (Pp and Cv) and Silurian (S). This neto stratigraphic record to interpret variations Basin stratigraphy document changes in source strengthening of the Laji-Jishi Shan as a source in sediment accumulation along the northeast- terranes that can be tied to newly emergent area for Linxia Basin is sustained throughout the ern Tibetan Plateau margin in the context of ranges (Fig. 12). Collectively, these analyses upper Wangjiashan section, where the 7.5-Ma local range growth (Fig. 9). Relative climatic suggest that Xunhua Basin has been adjacent to stratum have ~40% Laji-Jishi Shan–derived zir- quiescence throughout the late Oligocene–early emerging high terrain in the Laji Shan since ca. cons. We interpret the abrupt increase in Laji- Miocene (Dettman et al., 2003) suggests that re- 22 Ma and Jishi Shan since ca. 13 Ma. The ero- Jishi Shan–derived zircons between 14.5 and gional climate change alone did not induce the sion of these ranges at ca. 22 Ma and ca. 13 Ma 11.5 Ma to indicate that basement terranes in the changes in subsidence rates and source areas is also corroborated by vertical transects of nearby Jishi Shan became a dominant sediment that we detect in the magnetostratigraphic rec- low-temperature thermochronometer ages from source for the Linxia Basin at this time. ord. Furthermore, if climate were the dominant the hanging walls of the range-bounding thrust control on deposition, then a shift to more arid faults (Lease et al., 2011). DISCUSSION conditions in this region starting in middle Mio- cene time (Dettman et al., 2003; Garzione et al., Growth of the Laji Shan ca. 22 Ma The Hualong magnetostratigraphy provides 2005; Kent-Corson et al., 2009) should have led a late Oligocene–late Miocene record having to a decrease in erosion and deposition rates, Growth of the Laji Shan at ca. 22 Ma is ex- better accuracy and higher resolution than con- rather than the increase that we fi nd. pressed in northern Xunhua Basin by a pulse temporaneous records from nearby basins. The Our integration of detrital zircon records of of enhanced sediment accumulation, changes Hualong magnetostratigraphy, with its in situ provenance within magnetostratigraphic sec- in detrital zircon provenance, and coarsening- Hipparion weihoense fossil, provides a rather tions distinguishes this study from most others. upward facies. The Hualong magnetostrati- convincing correlation to the GPTS between Detrital zircon ages allow us to distinguish sedi- graphic record shows a distinct period of rapid 9.5 Ma and 25 Ma, with 55 chronostratigraphic ment sources in the plateau interior to the south sediment accumulation and basin subsidence tie points overlying a more uncertain chronology from sources outboard of the plateau margin to rates during 24–21 Ma (Fig. 9). Peak accumu- between 25 Ma and 30 Ma (Fig. 7). Nearby mag- the north—areas with somewhat similar gross lation rates during the 24–22.5 Ma interval are netostratigraphy from Linxia Basin, on the other bedrock lithologies, though each area has a more than two times higher than rates during the hand, has a more ambiguous correlation to the unique zircon U/Pb age spectrum (Fig. 10). Our subsequent 21–12 Ma period; peak subsidence

670 Geological Society of America Bulletin, May/June 2012 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Oligocene–Miocene basin evolution on the northeastern Tibetan Plateau

Eocene to early Miocene Zamazari Shan 40 km reconstruction ? 10 km WNW

view to Xiejia 0 km Xining Basin 1818 West Qinling the WSW Laji Shan 2020 2222 SSW 40 km L 2424 70 km 2626 Xunhua Basin 2828 0 km Xunhua 3030 20 km P 16 Hualong ? 3535 L 1500 16 10 km 4040 18 F 16 P 4545 18 ? 5050 F ? K 20 1 km 20 18

? (Ma) Age ? 22 20 1 km 22 (Ma) Age ? ca. 23 Ma 24 24 Mesozoic L D sed. 26 26 basement ? ? 28 28 ? F - Fluvial P 30 ? ~30? D - Deltaic ca. 45 Ma ? 35 L - Lacustrine P - Playa

Figure 13. Eocene to early Miocene block diagram reconstruction synthesizing differences in sediment thickness, age, depositional facies, and provenance relationships among magnetostratigraphic sections with >16 Ma strata in Xunhua (this paper; Hough et al., 2011) and Xining (Dai et al., 2006; Dupont-Nivet et al., 2007) Basins. Location is shown in Figure 1B.

rates are 3–4 times higher. Rapid tectonic sub- 17.5 Ma (Figs. 3 and 9) suggests an infl ux of Several independent lines of evidence, how- sidence until ca. 21 Ma suggests enhanced coarse sediment from an emergent or encroach- ever, suggest that the change in detrital zircon fl exural loading by thrust-thickened crust until ing source area. Provenance determinations signature recorded in Hualong by 19 Ma can be this time. This period of accelerated subsidence from 22.5 Ma and 19 Ma strata (Fig. 12A) reliably attributed to growth of the Laji Shan. is corroborated by low-temperature thermo- show that the coarsening-upward facies transi- First, the similarity of oxygen isotopic values chronol ogy from the hanging wall of the eastern tion is accompanied by a change to enhanced between Xunhua Basin and Linxia Basin sedi- Laji Shan thrust fault, which shows the onset erosion of detritus unique to basement terranes ments between 20 and 16 Ma (Hough et al., of rapid cooling at ca. 22 Ma after an extended in the Laji-Jishi Shan and further supports the 2011) suggests that the two basins were linked period of slow cooling since the Late Cretaceous interpretation that this range became a signifi - during this time with a nonextant or subdued (Lease et al., 2011). We interpret that enhanced cant source area at this time. Unfortunately, we Jishi Shan. Second, thermochronology indi- fl exural loading evinced by the rapid tectonic cannot discriminate Laji Shan versus Jishi Shan cates rapid erosion of the Laji Shan, but not the subsidence in Hualong is a response to growth source terranes with detrital zircon ages be- Jishi Shan, in the early Miocene (Lease et al., of the Laji Shan. cause, within the resolution of our data, these 2011). Third, paleocurrents from our oldest site The change in Hualong depositional envi- two ranges have identical age signatures. Addi- at 17.5 Ma until 13.5 Ma show a range of south- ronment from prolonged marginal lacustrine/ tionally, a preliminary examination of Laji erly paleo fl ow directions that are consistent with deltaic plain deposition (unit 1; 30–20 Ma) to Shan versus Jishi Shan zircon grains show that fl ow away from the Laji Shan (Fig. 3). Thus, the alluvial-fan deposition (unit 2) at 20 Ma and zircons from the two ranges have similar, non- change in zircon signature recorded in Hualong then to braided fl uvial deposition (unit 3) by unique rare earth element patterns. by 19 Ma is most reasonably attributed to a

Geological Society of America Bulletin, May/June 2012 671 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Lease et al.

Late Oligocene to Pleistocene reconstruction West Qinling SW 40 km Linxia Basin View to SE Maogou Xunhua 70 km the SSE Xunhua Wangjiashan Basin Jishi 10 km F 2 30 km 6 4 2 Shan Hualong 4 8 6 12 F 4 16 8 20 L 24 10 6 ? 6 28 F age (Ma) 12 D D ? 8 1 km 14? 8

1 km F L 16? 10 ? 10 L 20? 12 early F 10 Paleozoic 24? 14 plutonic ? ca. 13 Ma F ? 16 12

14 16 F 20 2 km 16 L 2 km 18

? (Ma) (Ma) Age 18 F BASEMENT 24 20 20

22 28 L F - Fluvial D ? D - Deltaic 24 3 km L - Lacustrine 3 km 26 P - Playa ca. 45 Ma 28 Cretac. P sed. ~30?

Figure 14. Late Oligocene to Pleistocene block diagram reconstruction synthesizing differences in sediment thickness, age, depositional facies , and provenance relationships among magnetostratigraphic sections in Xunhua Basin (this paper; Hough et al., 2011) and magneto- and lithostratigraphic sections in Linxia Basin (Fan et al., 2006; Fang et al., 2003; Li et al., 1997). Location shown in Figure 1B.

672 Geological Society of America Bulletin, May/June 2012 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Oligocene–Miocene basin evolution on the northeastern Tibetan Plateau

A Compacted accumulation rates (m/m.y.) uncertain whether the 12 Ma acceleration in Hualong sediment accumulation was due to an 26 Xining Range Basin increase in sediment supply, accommodation N Section space, or ponding behind the emergent Jishi Xunhua Rate Linxia 110 21 Guide (m/m.y.) Shan, each of these scenarios is consistent with 13 21 growth of the Jishi Shan at this time. Finally, growth of the Jishi Shan at ca. 13 Ma is consis- 40 km 30–25 Ma 15–10 Ma tent with local climate records from calcareous basin fi lls that show divergence of Linxia and 21 Xunhua oxygen isotopic values; this result sig- nals topographic isolation of these two basins by 118 114 surface uplift of the intervening Jishi Shan by 11 126 9 24 Ma (Hough et al., 2011). Figure 15. (A) Maps of sediment 63 accumulation rates at 5 m.y. 25–20 Ma 10–5 Ma Regional Depositional Environments, intervals from magnetostrati- Provenance, and Accumulation Patterns graphic sections in Xunhua and 29 adjacent basins (Dai et al., 2006; The late Oligocene depositional history of the Fang et al., 2003, 2005; Hough 84 88 209 northern Xunhua Basin and the Xining Basin to et al., 2011). (B) Age versus com- the northwest was characterized by a prolonged 106 18 113 pacted thickness for sections in 22 period of marginal lacustrine, playa, and deltaic Xunhua and adjacent basins plain deposition (Fig. 13). Widespread playa 20–15 Ma 5–2 Ma4000 shown in A. Black lines in back- facies extend as far back as the early Eocene ground denote linear regres- in the Xining Basin (Dai et al., 2006; Dupont- sions for each section at 5 m.y. B 4 Xining Linxia Wangjiashan Nivet et al., 2007). The abundance of evaporitic intervals. facies suggests a semiarid environment. Detrital Maogou 3 zircon age distributions from Upper Oligocene slowSlow accumulation deposits are dominated by 70%–80% West Qin- Rapid accumulation ling zircons in both Xunhua and Linxia deposits 2 (Fig. 12) and indicate that sediment was derived north Xunhua primarily from southern sources in the plateau (Hualong) south interior. Marginal lacustrine deposition contin- Xunhua ued until 20 Ma in the northern Xunhua Basin, 1 Ganjia when it was terminated by a pulse of coarse allu- Guide vial deposition (Fig. 13). Detrital zircon age dis- 0 Ashigong tributions from these Xunhua alluvial deposits Relative compacted thicknessRelative (km) 3025 20 15 10 5 0 show a notable change in source areas, and, for Age (Ma) the fi rst time, they contain >50% Laji-Jishi Shan zircons, thereby indicating a signifi cant sediment source north of the West Qinling (Fig. 12). Chronostratigraphic analysis of Miocene de- Laji Shan bedrock source near the area of rapid 14.5 and 11.5 Ma brackets the emergence of the posits from two magnetostratigraphic sections erosion revealed by low-temperature thermo- Laji-Jishi Shan as an active sediment source for within the Xunhua Basin that are separated by chronology. Finally, regional patterns of accu- Linxia Basin (Fig. 12B). We ascribe this change ~20 km along a north-south transect (Fig. 1B) mulation and depositional facies suggest that the to enhanced erosion of the Jishi Shan (not the defi nes a depositional environment where (1) in Xunhua Basin was an important early Miocene Laji Shan), in part because apatite (U-Th)/He the early Miocene, the northern and southern depocenter (discussed later herein, under Re- and fi ssion-track age-elevation data from the parts of Xunhua Basin received alluvial sedi- gional Depositional Environments, Provenance, Jishi Shan thrust hanging wall show the onset of ment of different origin (Fig. 13), and (2) in and Accumulation Patterns). accelerated uplift and erosion of the range at ca. the middle and late Miocene, fl uvial deposits 13 Ma (Lease et al., 2011). in the north graded into a lacustrine environment Growth of the Jishi Shan ca. 13 Ma On the western side of the Jishi Shan in Xun- toward the south (Fig. 14). Although alluvial hua Basin, magnetostratigraphic records from deposits occur from 20 Ma until 18 or 19 Ma Detailed analyses of magnetostratigraphy the Hualong basin show two notable changes at in both the northern Xunhua (Hualong) section and provenance from basins fl anking both sides 12 Ma that we ascribe to uplift of the Jishi Shan: and the southern Xunhua section (Hough et al., of the Jishi Shan reveal the emergence of this (1) a 50% acceleration in sediment accumulation 2011), these alluvial deposits are of different range at ca. 13 Ma and corroborate thermo- rate at the top of the section (12–9 Ma; Fig. 9) depositional character and were sourced from chronological and climatic records. Based on and (2) consistently westward paleo currents different ranges. Immature, coarse alluvial- synthesis of results described here for the east- throughout this interval in Hualong (Fig. 3), fan deposits in Hualong (Fig. 2) were depos- ern side of the range, an up-section change in which suggest sediment being shed from the ited during the same interval as more mature, Linxia Basin detrital zircon signatures between emergent Jishi Shan to the east. Although it is fi ner-grained fl uvial sediments in the southern

Geological Society of America Bulletin, May/June 2012 673 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Lease et al.

Xunhua basin (Hough et al., 2011). Further- subsidence continued until ca. 6 Ma (Fang et al., subsidence rather than supply was the dominant more, whereas post–20 Ma Hualong deposits 2003). Additionally, a threefold gradient in late control on sediment accumulation rates in the contain detrital zircons showing a dominant Miocene accumulation rates in Linxia Basin late Oligocene to middle Miocene Xunhua- Laji-Jishi Shan source (Fig. 12A), southern from two magnetostratigraphic sections (Fang Linxia foreland. Xunhua conglomerates are dominated by clasts et al., 2003) along an ~20 km transect orthogo- from the West Qinling (Hough et al., 2011). nal to the Jishi Shan (Fig. 15) suggests a strong Regional Tectonic Implications After 18 Ma, the southern Xunhua Basin tran- fl exural control on local deposition. sitioned to a lacustrine to marginal lacustrine Recognition of the erosion and topographic environment with magnetostratigraphic records Subsidence versus Supply development of the Laji Shan at ca. 22 Ma illu- at 18–15.8 Ma and 11–8.75 Ma (Hough et al., mi nates a poorly known episode of the Cenozoic 2011) (Figs. 13 and 14). In the northern Xunhua Models of foreland basins suggest that grain- geological history of the northeastern Tibetan Basin, on the other hand, fl uvial deposition con- size changes and subsidence rates depend on Plateau. The West Qinling defi ned the northeast- tinued uninterrupted from 18 Ma until the top whether deposition is driven by subsidence or ern plateau margin by 45–50 Ma, within a few of the section at 9 Ma. This southward-fi ning supply (e.g., Flemings and Jordan, 1989; Heller million years after initial India-Asia continental lateral facies variation from coarse fl uvial facies and Paola, 1992; Heller et al., 1988; Paola et al., collision, with the implication that stresses were in the northern Xunhua Basin to fi ner lacustrine 1992). Progradation of coarse-grained facies in transmitted instantaneously (from a geological facies in the southern Xunhua Basin also sug- a supply-driven basin should coincide with rela- perspective) over very large (>3000 km) dis- gests the emergence of the Laji Shan to the north tively fast subsidence and sediment accumula- tances (Clark et al., 2010; Dayem et al., 2009). as an important source area. tion rates. Progradation of coarse-grained facies Over the next 25 million years, the deformation Until middle Miocene time, the Xunhua in a subsidence-driven basin, however, is inhib- front on the northeastern plateau margin was and Linxia Basins appear to have shared a ited during periods of relatively fast subsidence apparently stationary until it jumped northward hydro logical connection (Hough et al., 2011), rates and instead coincides with relatively slow ~60 km with pulsed growth of the Laji Shan although Xunhua received 3–8 times thicker ac- subsidence and sediment accumulation rates. at ca. 22 Ma. cumulation of late Oligocene–middle Miocene The Hualong stratigraphic record suggests Observations from bedrock cooling in bound- sediment than Linxia (Fig. 14). Average δ18O subsidence-driven foreland deposition, where ing ranges and basin subsidence suggest that isotopic compositions within calcareous basin the period of most rapid subsidence and ac- slip on the Laji Shan fault was pulsed and defi lls are similar, suggesting that the Xunhua cumulation (27–21 Ma; Figs. 9A, 9B, and celerated after ca. 18 Ma. Modeling of Laji Shan and Linxia Basins were hydrologically linked 9C) corresponds to the deposition of the fi ne- thermochronological data indicates an order-of- until 11–16 Ma (Hough et al., 2011). Average grained muds of unit 1 (Fig. 9D). Furthermore, magnitude slowdown in cooling by 15–18 Ma early and middle Miocene accumulation rates the period of slowest subsidence and accumula- (Lease et al., 2011), and subsidence records adja- in the Xunhua Basin, however, were 5–13 times tion (21–12 Ma) corresponds to the deposition cent to the Laji Shan in northern Xunhua Basin higher than contemporaneous rates in the Linxia of coarse-grained conglomerates and sand- show a three- to fourfold decrease in subsidence Basin to the east and 3–6 times higher than rates stones of units 2 and 3 (Fig. 9). Although the rates after 21 Ma (Fig. 9). On the West Qinling in the Xining Basin to the west (Fig. 15). The Hualong magnetostratigraphic section only pro- fault ~60 km to the south, on the other hand, signifi cantly greater magnitude and rates of ac- vides a one-dimensional record, the synchrony Clark et al. (2010) interpreted a coeval period of cumulation in Xunhua Basin suggest that it was of coarse-grained deposition with the slowest accelerated faulting; this interpretation is based an important depocenter lying near a potential subsidence and accumulation rates suggests on thermochronological data that suggest a two- locus of crustal fl exure in the Laji Shan. Further- that progradation of gravel across the northern fold increase in erosion rates beginning at ca. more, westward paleofl ow within the ca. 20 Ma Xunhua Basin did not occur until after local 18 Ma. We speculate that these observations to- and ca. 14 Ma deltaic units in the Wangjiashan fl exural loading (i.e., tectonically driven sub- gether suggest transfer of fault slip from the Laji section in western Linxia Basin (Garzione et al., sidence) diminished signifi cantly in magnitude Shan to the West Qinling and out-of-sequence 2005) suggest fl ow toward a Xunhua depocenter after 21 Ma. This fi nding is consistent with a rejuvenation of thrusting in the West Qinling. in the west (Fig. 14). subsidence-driven setting modulated by thrust The relatively low-lying area between the West Uplift of the Jishi Shan at ca. 13 Ma (Lease loading of the Laji Shan to the north. Qinling and Laji Shan ranges that is occupied by et al., 2011) disrupted the hydrological connec- Additionally, the considerably greater thick- the Xunhua Basin is devoid of any notable struc- tion between the Xunhua and Linxia Basins nesses and rates of late Oligocene–middle Mio- tures accommodating north-south contraction (Hough et al., 2011), and, accordingly, late cene accumulation in Xunhua relative to Linxia (Figs. 2, 13, and 14). This paucity of structures Miocene sedimentation in these basins likely and Xining indicate significant subsidence suggests that the Xunhua Basin area has not been responded to different controls. Jishi Shan adja cent to the eastern Laji Shan (Figs. 13 and thickened via distributed north-south shortening uplift was the last in a series of range-growth 14). Late Oligocene–middle Miocene accumu- across the basin in Oligocene-to-Holocene time, events that collectively isolated Xunhua from lation rates in Hualong were 3–13 times more but instead that deformation has been persis- surrounding areas and promoted intermontane rapid than contemporaneous rates in Linxia and tently localized across either the Laji Shan or ponding of sediment. The onset of uplift at ca. Xining (Fig. 15). Between 28 and 20 Ma, rela- West Qinling. 13 Ma in the Jishi Shan, ca. 22 Ma in the Laji tively rapid accumulation of fi ne-grained lacus- The development of the north-trending Jishi Shan (Lease et al., 2011), and ca. 45–50 Ma in trine muds in Hualong (~120 m/m.y.) occurred Shan at ca. 13 Ma is signifi cant because this the West Qinling (Clark et al., 2010) defi ned the at the same time as relatively slow accumulation range is oriented obliquely to and developed eastern, northern, and southwestern margins of of coarser-grained fl uvial and deltaic facies in later than both the WNW-trending Laji Shan Xunhua Basin, respectively. In Linxia Basin, on Linxia (~10 m/m.y.) (Figs. 13 and 14). The re- (ca. 22 Ma) and the NW-trending West Qinling the other hand, subsidence rates that increase gional correlation of fi ne-grained deposits with (ca. 45–50 Ma; Clark et al., 2010) nearby. Tem- with time are interpreted to indicate that fl exural the most rapid accumulation rates suggests that poral variations in the orientation of Cenozoic

674 Geological Society of America Bulletin, May/June 2012 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Oligocene–Miocene basin evolution on the northeastern Tibetan Plateau range growth have been interpreted to refl ect Later growth of the Jishi Shan after ca. 13 Ma search–Solid Earth, v. 107, no. B11, 2300, doi:10.1029 a middle Miocene reorganization of deforma- accommodated east-west contraction. Enhanced /2000JB000050. Blair, T.C., and McPherson, J.G., 1992, The Trollheim alluvial tion, from primarily NNE-SSW contraction, erosion of the Jishi Shan is indicated by a change fan and facies model revisited: Geological Society of which mimics the plate convergence direction, in the detrital zircon signature of Linxia Basin America Bulletin, v. 104, no. 6, p. 762–769, doi:10.1130 /0016-7606(1992)104<0762:TTAFAF>2.3.CO;2. to the inclusion of east-west motion (Lease et al., sediments by 11.5 Ma and by an acceleration in Bovet, P.M., Ritts, B.D., Gehrels, G., Abbink, A.O., Darby, 2011). One expected consequence of such a sediment accumulation in Hualong Basin start- B., and Hourigan, J., 2009, Evidence of Miocene reorganization—left-lateral strike-slip motion ing at 12 Ma. Furthermore, the divergence of crustal shortening in the North Qilian Shan from Ceno- zoic stratigraphy of the western Hexi Corridor, Gansu on the West Qinling and Laji Shan faults—has oxygen isotopic ratios between southern Xun- Province, China: American Journal of Science, v. 309, been documented in Quaternary time (Li, 2005; hua and Linxia Basin strata by 11 Ma suggests no. 4, p. 290–329, doi:10.2475/00.4009.02. Yuan et al., 2005). Furthermore, the present topographic development of the intervening Jishi Cabrera, L., Colombo, F., and Robles, S., 1985, Sedimen- tation and tectonic interrelationships in the Paleogene geodetic fi eld in this area, which is dominated Shan and the growing presence of a rain shadow. marginal alluvial systems of the SE Ebro basin. Transi- by N110°E motion oriented orthogonally to Finally, bedrock thermochronology suggests the tion from alluvial to shallow lacustrine environments, in Milá, M., and Rosell, J., eds., Excursion guidebook the Jishi Shan (Duvall and Clark, 2010; Gan onset of rapid cooling and thrust-induced erosion of the 6th European Regional IAS meeting (Lleida, et al., 2007), suggests that this interpretation in the Jishi Shan at ca. 13 Ma. Spain): Barcelona, International Association of Sedi- is plausible. Detailed structural and kinematic The history of the Tibetan Plateau between 40 mentologists and Institut d’Estudis Ilerdencs, Univer- sitat Autònoma de Barcelona, p. 393–492. analyses of range-bounding faults, however, and 13 Ma is characterized by relatively small Cande, S.C., and Kent, D.V., 1995, Revised calibration of await further investigation. The onset of accel- magnitudes of range growth along the plateau’s the geomagnetic polarity time scale for the Late Creta- erated exhumation on north-trending structures northeastern margin. The northeastern limit of ceous and Cenozoic: Journal of Geophysical Research, v. 100, p. 6093–6095, doi:10.1029/94JB03098. in both the Jishi Shan (ca. 13 Ma) and nearby plateau deformation in this area, established in Cardozo, N., 2009, OSXBackstrip v2.6: http://homepage. Liupan Shan (ca. 8 Ma), ~320 km to the east the early Eocene shortly after initial India-Asia mac.com/nfcd/work/programs.html (last accessed July 2009). (Zheng et al., 2006), may signal the start of continental collision (Clark et al., 2010), re- Clark, M.K., House, M.A., Royden, L.H., Whipple, K.X., the modern defor mation regime in northeast- mained largely stationary over the next 25 m.y. Burchfi el, B.C., Zhang, X., and Tang, W., 2005, Late ern Tibet, where contraction is coupled to left- and only jumped 60 km to the north in early Cenozoic uplift of southeastern Tibet: Geology, v. 33, no. 6, p. 525–528, doi:10.1130/ G21265.1. lateral strike-slip motion in a complex regional Miocene time with pulsed uplift of the Laji Shan. Clark, M.K., Farley, K.A., Zheng, D., Wang, Z.C., and Duvall , stepover between the Kunlun and Haiyuan Growth of the north-trending Jishi Shan and Liu- A., 2010, Early Cenozoic faulting on the northern left-lateral faults (Duvall and Clark, 2010). pan Shan beginning ~10 m.y. later suggests that Tibetan Plateau margin from apatite (U-Th)/He ages: Earth and Planetary Science Letters, v. 296, no. 1–2, Widespread middle and late Miocene accel- this sector of the plateau was accommodating p. 78–88, doi:10.1016/j.epsl.2010.04.051. erations in faulting, erosion, and deposition, as east-west motion by middle Miocene time. Collinson, J.D., 1996, Alluvial sediments, in Reading, H.G., well as coarsening-upward facies and climatic ed., Sedimentary Environments: Processes, Facies, and ACKNOWLEDGMENTS Stratigraphy: Oxford, UK, Blackwell Publishing, Inc., records suggest intensifi ed tectonic activity and p. 37–82. changing kinematics in northeastern Tibet since Craddock, W.H., Kirby, E., Harkins, N.W., Zhang, H.P., This work was funded by the U.S. National Science Shi, X.H., and Liu, J.H., 2010, Rapid fl uvial incision this time (see Fig. 1A, and references therein; Foundation (NSF) Continental Dynamics program along the Yellow River during headward basin inte- Kent-Corson et al., 2009; Molnar, 2005; Ritts (EAR-0507431 and EAR-0506575), the National gration: Nature Geoscience, v. 3, no. 3, p. 209–213, et al., 2008; Wang et al., 2011b). Science Foundation of China (40234040), and the doi:10.1038/ngeo777. China State Key Laboratory of Earthquake Dynam- Craddock, W.H., Kirby, E., and Zheng, D.W., 2011, Late ics (LED2008A01). Lease acknowledges support Miocene-Pliocene range growth in the interior of the CONCLUSION from a NSF graduate research fellowship. We thank northeastern Tibetan Plateau: Lithosphere, v. 3, no. 6, Joe Kirschvink and the Caltech Paleomagnetics Lab- p. 420–438, doi:10.1130/L159.1. Dai, S., Fang, X.M., Dupont-Nivet, G., Song, C.H., Gao, Sedimentary successions from the northern oratory, George Gehrels and the Arizona Laserchron J.P., Krijgsman, W., Langereis, C., and Zhang, W.L., Center, Daniel Culling for zircon separation, Laura Xunhua Basin provide a unique view of the late 2006, Magnetostratigraphy of Cenozoic sediments Smith and Huiping Zhang for fi eld assistance, and from the Xining Basin: Tectonic implications for the Oligocene–Miocene history of northeastern Tibet Zhang Peizhen and colleagues at the Chinese Earth- northeastern Tibetan Plateau: Journal of Geophysical because these archives are particularly sensitive quake Administration (Beijing and Lanzhou) for sup- Research–Solid Earth, v. 111, B11102, doi:10.1029 to small increments of growth of nearby ranges port in the fi eld. Many thanks go to Tao Deng and /2005JB004187. Dayem, K.E., Molnar, P., Clark, M.K., and Houseman, G.A., that more distant basins fail to clearly record. The Xiaoming Wang for punctual fossil identifi cation. We acknowledge reviews by Rob van der Voo and An Yin 2009, Far-fi eld lithospheric deformation in Tibet dur- Hualong section displays a more than twofold ing continental collision: Tectonics, v. 28, TC6005, on an earlier version of this manuscript, plus reviews doi:10.1029/2008TC002344. acceleration in sediment accumulation and basin by Guillaume Dupont-Nivet and Carmala Garzione on DeCelles, P.G., Gray, M.B., Cole, R.B., Pequera, N., Pivnik, subsidence during 24–21 Ma and a transition this version. D.A., Ridgway, K.D., and Srivastava, P., 1991, Controls to alluvial facies at 20 Ma that was coincident on synorogenic alluvial-fan architecture, Beartooth REFERENCES CITED Conglomerate: Wyoming and Montana: Sedimentol- with intensifi ed erosion of basement source ter- ogy, v. 38, p. 567–590, doi:10.1111/j.1365-3091.1991 ranes in the Laji-Jishi Shan by 19 Ma. Together, Abels, H.A., Dupont-Nivet, G., Xiao, G., Bosboom, R.E., .tb01009.x. these fi ndings suggest growth of the Laji Shan at and Krijgsman, W., 2011, Step-wise change of Asian DeCelles, P.G., Gehrels, G.E., Najman, Y., Martin, A.J., interior climate preceding the Eocene-Oligocene tran- Carter, A., and Garzanti, E., 2004, Detrital geochronol- 24–21 Ma in a subsidence-driven setting where sition (EOT): Palaeogeography, Palaeoclimatology, ogy and geochemistry of Cretaceous–early Miocene progradation of coarse facies was inhibited dur- Palaeo ecology, v. 299, no. 3–4, p. 399–412, doi:10.1016 strata of Nepal: Implications for timing and diachro- ing the period of most intense subsidence. Addi- /j.palaeo.2010.11.028. neity of initial Himalayan orogenesis: Earth and Plan- Allen, P.A., and Allen, J.R., 2005, Basin Analysis: Principles etary Science Letters, v. 227, no. 3–4, p. 313–330, tionally, bedrock thermochronol ogy suggests the and Applications: Oxford, UK, Wiley-Blackwell, 560 p. doi:10.1016/j.epsl.2004.08.019. onset of rapid Laji Shan cooling at ca. 22 Ma. Amidon, W.H., Burbank, D.W., and Gehrels, G.E., 2005, DeGraaff-Surpless, K., Mahoney, J.B., Wooden, J.L., and U-Pb zircon ages as a sediment mixing tracer in the McWilliams, M.O., 2003, Lithofacies control in detrital Furthermore, an analysis of regional patterns Nepal Himalaya: Earth and Planetary Science Letters, zircon provenance studies: Insights from the Creta- of depositional facies and accumulation rates v. 235, no. 1–2, p. 244–260, doi:10.1016/j.epsl.2005 ceous Methow basin, southern Canadian Cordillera: indicates that the Xunhua Basin was an impor- .03.019. Geological Society of America Bulletin, v. 115, no. 8, Besse, J., and Courtillot, V., 2002, Apparent and true polar p. 899–915, doi:10.1130/B25267.1. tant early Miocene depocenter that lay along the wander and the geometry of the geomagnetic fi eld Deng, T., Wang, X., Ni, X.J., and Liu, L., 2004, Sequence southern margin of the Laji Shan. over the last 200 Myr: Journal of Geophysical Re- of Cenozoic mammalian faunas of the Linxia Basin in

Geological Society of America Bulletin, May/June 2012 675 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Lease et al.

Gansu, China: Acta Geologica Sinica, v. 78, no. 1, Gehrels, G.E., Valencia, V., and Pullen, A., 2006, Detrital Lease, R.O., Burbank, D.W., Clark, M.K., Farley, K.A., p. 8–14. zircon geochronology by laser-ablation multicollector Zheng, D., and Zhang, H., 2011, Middle Miocene Dettman, D.L., Fang, X.M., Garzione, C.N., and Li, J.J., ICPMS at the Arizona Laserchron Center, in Olszew- reorganization of deformation along the northeastern 2003, Uplift-driven climate change at 12 Ma: A ski, T., and Huff, W., eds., Geochronology: Emerging Tibetan Plateau: Geology, v. 39, no. 4, p. 359–362, long delta O-18 record from the NE margin of the Opportunities: Philadelphia, Paleontological Society doi:10.1130/G31356.1. Tibetan Plateau: Earth and Planetary Science Letters, Short Course, p. 1–10. Li, C., 2005, Quantitative research on several prominent v. 214, no. 1–2, p. 267–277, doi:10.1016/S0012-821X Gilder, S., Chen, Y., Cogne, J.P., Tan, X.D., Courtillot, V., faults in the NE margin of the Tibetan Plateau [Ph.D. (03)00383-2. Sun, D.J., and Li, Y.G., 2003, Paleomagnetism of Upper thesis]: Bejing, China, Institute of Geology, China Dupont-Nivet, G., Krijgsman, W., Langereis, C.G., Abels, Jurassic to Lower Cretaceous volcanic and sedimen- Earthquake Administration (in Chinese). H.A., Dai, S., and Fang, X.M., 2007, Tibetan Plateau tary rocks from the western Tarim Basin and implica- Li, J.J., Fang, X.M., Van der Voo, R., Zhu, J.J., Mac Niocaill , aridifi cation linked to global cooling at the Eocene- tions for inclination shallowing and absolute dating C., Cao, J.X., Zhong, W., Chen, H.L., Wang, J.L., Oligocene transition: Nature, v. 445, no. 7128, p. 635– of the M-0 (ISEA?) chron: Earth and Planetary Sci- Wang, J.M., and Zhang, Y.C., 1997, Late Cenozoic 638, doi:10.1038/nature05516. ence Letters, v. 206, no. 3–4, p. 587–600, doi:10.1016 magnetostratigraphy (11–0 Ma) of the Dongshanding Dupont-Nivet, G., Dai, S., Fang, X., Krijgsman, W., Erens, /S0012-821X(02)01074-9. and Wangjiashan sections in the Longzhong Basin, V., Reitsma, M., and Langereis, C., 2008, Timing and Godard, V., Pik, R., Lave, J., Cattin, R., Tibari, B., de Sigoyer, western China: Geologie en Mijnbouw, v. 76, no. 1–2, distribution of tectonic rotations in the northeastern J., Pubellier, M., and Zhu, J., 2009, Late Cenozoic evo- p. 121–134, doi:10.1023/A:1003153717799. Tibetan Plateau, in Burchfi el, B.C., and Wang, E., lution of the central Longmen Shan, eastern Tibet: In- Mack, G.H., James, W.C., and Monger, H.C., 1993, Classifi - eds., Investigations into the Tectonics of the Tibetan sight from (U-Th)/He thermochronometry: Tectonics, cation of paleosols: Geological Society of America Bul- Plateau: Geological Society of America Special Paper v. 28, TC5009, doi:10.1029/2008TC002407. letin, v. 105, no. 2, p. 129–136, doi:10.1130/0016-7606 444, p. 73–87. Harkins, N., Kirby, E., Heimsath, A., Robinson, R., and (1993)105<0129:COP>2.3.CO;2. Duvall, A.R., and Clark, M.K., 2010, Dissipation of fast Reiser, U., 2007, Transient fl uvial incision in the head- McFadden, P.L., and McElhinny, M.W., 1990, Classifi ca- strike-slip faulting within and beyond northeastern waters of the Yellow River, northeastern Tibet, China: tion of the reversal test in paleomagnetism: Geophysi- Tibet: Geology, v. 38, no. 3, p. 223–226, doi:10.1130 Journal of Geophysical Research–Earth Surface, v. 112, cal Journal International, v. 103, no. 3, p. 725–729, /G30711.1. F03S04, doi:10.1029/2006JF000570. doi:10.1111/j.1365-246X.1990.tb05683.x. Duvall, A.R., Clark, M.K., van der Pluijm, B.A., and Li, C., Heller, P., and Paola, C., 1992, The large-scale dynamics of Métivier, F., Gaudemer, Y., Tapponnier, P., and Meyer, B., 2011, Direct dating of Eocene reverse faulting in north- grain-size variation in alluvial basins: 2. Applications 1998, Northeastward growth of the Tibet Plateau de- eastern Tibet using Ar-dating of fault clays and low- to syntectonic conglomerate: Basin Research, v. 4, duced from balanced reconstruction of two deposi- temperature thermochronometry: Earth and Planetary p. 91–102, doi:10.1111/j.1365-2117.1992.tb00146.x. tional areas: The Qaidam and Hexi Corridor Basins, Science Letters, v. 304, no. 3–4, p. 520–526, doi:10.1016 Heller, P.L., Angevine, C.L., Winslow, N.S., and Paola, C., China: Tectonics, v. 17, no. 6, p. 823–842, doi:10.1029 /j.epsl.2011.02.028. 1988, Two-phase stratigraphic model of foreland-basin /98TC02764. Enkelmann, E., Ratschbacher, L., Jonckheere, R., Nestler, sequences: Geology, v. 16, no. 6, p. 501–504, doi:10.1130 Miall, A.D., 1978, Lithofacies types and vertical pro- R., Fleischer, M., Gloaguen, R., Hacker, B.R., Zhang, /0091-7613(1988)016<0501:TPSMOF>2.3.CO;2. file models in braided river deposits: A summar, in Y.Q., and Ma, Y.S., 2006, Cenozoic exhumation and Horton, B.K., Dupont-Nivet, G., Zhou, J., Waanders, G.L., Miall , A.D., ed., Fluvial Sedimentology: Memoirs of deformation of northeastern Tibet and the Qinling: Butler, R.F., and Wang, J., 2004, Mesozoic-Cenozoic the Canadian Society of Petroleum Geologists, v. 5, Is Tibetan lower crustal fl ow diverging around the evolution of the Xining-Minhe and Dangchang Basins, p. 597–604. Sichuan Basin?: Geological Society of America Bulle- northeastern Tibetan Plateau: Magnetostratigraphic Miall, A.D., 1996, The Geology of Fluvial Deposits: Sedi- tin, v. 118, no. 5–6, p. 651–671, doi:10.1130/B25805.1. and biostratigraphic results: Journal of Geophysical mentary Facies, Basin Analysis, and Petroleum Geol- Fan, M.J., Dettman, D.L., Song, C.H., Fang, X.M., and Research–Solid Earth, v. 109, B04402, doi:10.1029 ogy: Berlin, New York, Springer, 582 p. Garzione, C.N., 2007, Climatic variation in the Linxia /2003JB002913. Molnar, P., 2005, Mio-Pliocene growth of the Tibetan Pla- Basin , NE Tibetan Plateau, from 13.1 to 4.3 Ma: Hough, B.G., Garzione, C.N., Wang, Z.C., Lease, R.O., Bur- teau and evolution of East Asian climate: Palaeontolo- The stable isotope record: Palaeogeography, Palaeo- bank, D.W., and Yuan, D.Y., 2011, Stable isotope evi- gia Electronica, v. 8, no. 1, p. 2A. climatol ogy, Palaeoecology, v. 247, no. 3–4, p. 313– dence for topographic growth and basin segmentation: Molnar, P., and Stock, J.M., 2009, Slowing of India’s conver- 328, doi:10.1016/j.palaeo.2006.11.001. Implications for the evolution of the NE Tibetan Pla- gence with Eurasia since 20 Ma and its implications for Fan, M.J., Song, C.H., Dettman, D.L., Fang, X.M., and teau: Geological Society of America Bulletin, v. 123, Tibetan mantle dynamics: Tectonics, v. 28, TC3001, Xu, X.H., 2006, Intensifi cation of the Asian winter no. 1–2, p. 168–185, doi:10.1130/B30090.1. doi:10.1029/2008TC002271. monsoon after 7.4 Ma: Grain-size evidence from the Johnson, N.M., and Mcgee, V.E., 1983, Magnetic polarity Nanson, G.C., and Croke, J.C., 1992, A genetic classifi cation Linxia Basin, northeastern Tibetan Plateau, 13.1 Ma to stratigraphy—Stochastic properties of data, sampling of fl oodplains: Geomorphology, v. 4, no. 6, p. 459– 4.3 Ma: Earth and Planetary Science Letters, v. 248, problems, and the evaluation of interpretations: Journal 486, doi:10.1016/0169-555X(92)90039-Q. no. 1–2, p. 186–197, doi:10.1016/j.epsl.2006.05.025. of Geophysical Research, v. 88, no. B2, p. 1213–1221, Nemec, W., and Steel, R.J., 1984, Alluvial and coastal con- Fang, X.M., Garzione, C., Van der Voo, R., Li, J.J., and Fan, doi:10.1029/JB088iB02p01213. glomerates: Their signifi cant features and some com- M.J., 2003, Flexural subsidence by 29 Ma on the NE Jones, C.H., 2002, User-driven integrated software lives: ments on gravelly mass-fl ow deposits, in Koster, E.H., edge of Tibet from the magnetostratigraphy of Linxia “Paleomag” paleomagnetics analysis on the Macintosh: and Steel, R.J., eds., Sedimentology of Gravels and Basin, China: Earth and Planetary Science Letters, Computers & Geosciences, v. 28, no. 10, p. 1145–1151, Conglomerates: Calgary, Alberta, Canada, Canadian v. 210, no. 3–4, p. 545–560, doi:10.1016/S0012-821X doi:Pii S0098–3004(02)00032–8. Society of Petroleum Geologists, p. 1–31. (03)00142-0. Kent-Corson, M.L., Ritts, B.D., Zhuang, G.S., Bovet, P.M., Ogg, J.G., and Smith, A.G., 2004, The geomagnetic polarity Fang, X.M., Yan, M.D., Van der Voo, R., Rea, D.K., Song, Graham, S.A., and Chamberlain, C.P., 2009, Stable time scale, in Gradstein, F.M., Ogg, J.G., and Smith, C.H., Pares, J.M., Gao, J.P., Nie, J.S., and Dai, S., isotopic constraints on the tectonic, topographic, and A., eds., Geological Time Scale: Cambridge, UK, 2005, Late Cenozoic deformation and uplift of the NE climatic evolution of the northern margin of the Tibetan Cambridge University Press, p. 65–86. Tibetan Plateau: Evidence from high-resolution mag- Plateau: Earth and Planetary Science Letters, v. 282, Ouimet, W., Whipple, K., Royden, L., Reiners, P., Hodges, netostratigraphy of the Guide Basin, Qinghai Province, no. 1–4, p. 158–166, doi:10.1016/j.epsl.2009.03.011. K., and Pringle, M., 2010, Regional incision of the China: Geological Society of America Bulletin, v. 117, Kirby, E., Reiners, P.W., Krol, M.A., Whipple, K.X., Hodges, eastern margin of the Tibetan Plateau: Lithosphere, no. 9–10, p. 1208–1225, doi:10.1130/B25727.1. K.V., Farley, K.A., Tang, W.Q., and Chen, Z.L., 2002, v. 2, no. 1, p. 50–63, doi:10.1130/L57.1. Flemings, P.B., and Jordan, T.E., 1989, A synthetic strati- Late Cenozoic evolution of the eastern margin of the Paola, C., Heller, P.L., and Angevine, C.L., 1992, The graphic model of foreland basin development: Journal Tibetan Plateau: Inferences from Ar-40/Ar-39 and large-scale dynamics of grain-size variation in allu- of Geophysical Research–Solid Earth, v. 94, no. B4, (U-Th)/He thermochronology: Tectonics, v. 21, 1001, vial basins: 1. Theory: Basin Research, v. 4, p. 73–90, p. 3851–3866, doi:10.1029/JB094iB04p03851. doi:10.1029/2000TC001246. doi:10.1111/j.1365-2117.1992.tb00145.x. Friend, P.F., Slater, M.J., and Williams, R.C., 1979, Verti- Kirschvink, J.L., 1980, The least-squares line and plane Qinghai Bureau of Geology and Mineral Resources cal and lateral building of river sandstone bodies, Ebro and the analysis of paleomagnetic data: Geophysical (QBGMR ), 1991, Regional Geology of the Qinghai Basin, Spain: Journal of the Geological Society of Lon- Journal of the Royal Astronomical Society, v. 62, no. 3, Province: Beijing, Geological Publishing House, 662 p. don, v. 136, no. 1, p. 39, doi:10.1144/gsjgs.136.1.0039. p. 699–718. Reading, H.G., and Collinson, J.D., 1996, Clastic coasts, Gan, W.J., Zhang, P.Z., Shen, Z.K., Niu, Z.J., Wang, M., Kirschvink, J.L., Kopp, R.E., Raub, T.D., Baumgartner, in Reading, H.G., ed., Sedimentary Environments: Wan, Y.G., Zhou, D.M., and Cheng, J., 2007, Present- C.T., and Holt, J.W., 2008, Rapid, precise, and high- Processes, Facies and Stratigraphy: Oxford, UK, Black- day crustal motion within the Tibetan Plateau inferred sensitivity acquisition of paleomagnetic and rock- well Publishing Inc., p. 154–231. from GPS measurements: Journal of Geophysical magnetic data: Development of a low-noise automatic Ritts, B.D., Yue, Y.J., Graham, S.A., Sobel, E.R., Abbink, Research–Solid Earth, v. 112, B08416, doi:10.1029 sample changing system for superconducting rock O.A., and Stockli, D., 2008, From sea level to high /2005JB004120. magnetometers: Geochemistry, Geophysics, Geosys- elevation in 15 million years: Uplift history of the Garzione, C.N., Ikari, M.J., and Basu, A.R., 2005, Source of tems, v. 9, Q05Y01, doi:10.1029/2007GC001856. northern Tibetan Plateau margin in the Altun Shan: Oligocene to Pliocene sedimentary rocks in the Linxia Lease, R.O., Burbank, D.W., Gehrels, G.E., Wang, Z.C., American Journal of Science, v. 308, no. 5, p. 657–678, Basin in northeastern Tibet from Nd isotopes: Impli- and Yuan, D.Y., 2007, Signatures of mountain build- doi:10.2475/05.2008.01. cations for tectonic forcing of climate: Geological ing: Detrital zircon U/Pb ages from northeastern Schreiber, B.C., and El Tabakh, M., 2000, Deposition and Society of America Bulletin, v. 117, no. 9–10, p. 1156– Tibet : Geology, v. 35, no. 3, p. 239–242, doi:10.1130 early alteration of evaporites: Sedimentology, v. 47, 1166, doi:10.1130/B25743.1. /G23057A.1. p. 215–238, doi:10.1046/j.1365-3091.2000.00002.x.

676 Geological Society of America Bulletin, May/June 2012 Downloaded from gsabulletin.gsapubs.org on July 5, 2012 Oligocene–Miocene basin evolution on the northeastern Tibetan Plateau

Sclater, J.G., and Christie, P.A.F., 1980, Continental stretch- the NE Tibetan Plateau, China: Implications for middle province: Earthquake Research in China, v. 19, no. 4, ing—An explanation of the post-mid-Cretaceous Miocene tectonics of the West Qinling fault zone: p. 391–401. subsidence of the central North Sea Basin: Journal of Journal of Asian Earth Sciences (in press), doi:10.1016 Zhai, Y., and Cai, T., 1984, The Tertiary system of Gansu Geophysical Research, v. 85, no. B7, p. 3711–3739, /j.jseaes.2011.06.009. Province, Wang, Y.M., ed., in Gansu Geology: Lan- doi:10.1029/JB085iB07p03711. Yan, M., Van der Voo, R., Fang, X.-M., Pares, J.M., and zhou, China, People’s Press of Gansu, p. 1–40 (in Talbot, M.R., and Allen, P.A., 1996, Lakes, in Reading, Rea, D.K., 2006, Paleo-magnetic evidence for a mid- Chinese). H.G., ed., Sedimentary Environments: Processes, Fa- Miocene clockwise rotation of about 25° of the Guide Zhang, H.P., Craddock, W.H., Lease, R.O., Wang, W.T., cies and Stratigraphy: Oxford, UK, Blackwell Publish- Basin area in NE Tibet: Earth and Planetary Science Yuan, D.Y., Zhang, P.Z., Molnar, P., Zheng, D.W., and ing Inc., p. 83–124. Letters, v. 241, p. 234–247, doi:10.1016/j.epsl.2005 Zheng, W.J., 2011, Magnetostratigraphy of the Neo- Tauxe, L., and Gallet, Y., 1991, A jackknife for magneto- .10.013. gene Chaka Basin and its implications for mountain stratigraphy: Geophysical Research Letters, v. 18, Yin, A., and Harrison, T.M., 2000, Geologic evolution of the building processes in the north-eastern Tibetan Plateau: no. 9, p. 1783–1786, doi:10.1029/91GL01223. Himalayan-Tibetan orogen: Annual Review of Earth Basin Research (in press), doi:10.1111/j.1365-2117 Tauxe, L.P., and Hartl, P., 1997, 11 million years of Oligocene and Planetary Sciences, v. 28, p. 211–280, doi:10.1146 .2011.00512.x. geomagnetic ﬁ eld behavior: Geophysical Journal Inter- /annurev.earth.28.1.211. Zheng, D.W., Zhang, P.Z., Wan, J.L., Li, C., and Cao, J., national, v. 128, p. 217–229, doi:10.1111/j.1365-246X Yin, A., Rumelhart, P.E., Butler, R., Cowgill, E., Harrison, 2003, Late Cenozoic deformation subsequence in .1997.tb04082.x. T.M., Foster, D.A., Ingersoll, R.V., Zhang, Q., Zhou, northeastern margin of Tibet: Detrital AFT records Tauxe, L., and Watson, G.S., 1994, The fold test—An eigen X.Q., Wang, X.F., Hanson, A., and Raza, A., 2002, from Linxia Basin: Science in China, v. 46, p. 266–275. analysis approach: Earth and Planetary Science Letters, Tectonic history of the Altyn Tagh fault system in Zheng, D.W., Zhang, P.Z., Wan, J.L., Yuan, D.Y., Li, C.Y., v. 122, no. 3–4, p. 331–341, doi:10.1016/0012-821X northern Tibet inferred from Cenozoic sedimentation: Yin, G.M., Zhang, G.L., Wang, Z.C., Min, W., and (94)90006-X. Geological Society of America Bulletin, v. 114, no. 10, Chen, J., 2006, Rapid exhumation at ~8 Ma on the Liu- Tauxe, L., Butler, R.F., Banerjee, S.K., and Van der Voo, R., p. 1257–1295, doi:10.1130/0016-7606(2002)114<1257: pan Shan thrust fault from apatite ﬁ ssion-track thermo- 2009, Essentials of Paleomagnetism: Berkeley, Cali- THOTAT>2.0.CO;2. chronology: Implications for growth of the northeastern fornia, University of California Press, 512 p. Yin, A., Dang, Y.Q., Wang, L.C., Jiang, W.M., Zhou, S.P., Tibetan Plateau margin: Earth and Planetary Science Wang, W.T., Zhang, P.Z., Kirby, E., Wang, L.H., Zhang, Chen, X.H., Gehrels, G.E., and McRivette, M.W., Letters, v. 248, no. 1–2, p. 198–208, doi:10.1016 G.L., Zheng, D.W., and Chai, C.Z., 2011a, A revised 2008a, Cenozoic tectonic evolution of Qaidam Basin /j.epsl.2006.05.023. chronology for Tertiary sedimentation in the Sikouzi and its surrounding regions (Part 1): The southern Zheng, D.W., Clark, M.K., Zhang, P.Z., Zheng, W., and Farley , Basin: Implications for the tectonic evolution of the Qilian Shan–Nan Shan thrust belt and northern Qaidam K.A., 2010, Erosion, fault initiation and topographic northeastern corner of the Tibetan Plateau: Tectono- Basin: Geological Society of America Bulletin, v. 120, growth of the North Qilian Shan (northern Tibetan physics, v. 505, no. 1–4, p. 100–114, doi:10.1016 no. 7–8, p. 813–846, doi:10.1130/B26180.1. Plateau): Geosphere, v. 6, no. 6, p. 1–5, doi:10.1130 /j.tecto.2011.04.006. Yin, A., Dang, Y.Q., Zhang, M., Chen, X.H., and McRivette, /GES00523.1. Wang, Y., and Deng, T., 2005, A 25 m.y. isotopic record of M.W., 2008b, Cenozoic tectonic evolution of the paleodiet and environmental change from fossil mam- Qaidam Basin and its surrounding regions (Part 3): SCIENCE EDITOR: A. HOPE JAHREN mals and paleosols from the NE margin of the Tibetan Structural geology, sedimentation, and regional tec- ASSOCIATE EDITOR: AN YIN Plateau: Earth and Planetary Science Letters, v. 236, tonic reconstruction: Geological Society of America no. 1–2, p. 322–338, doi:10.1016/j.epsl.2005.05.006. Bulletin, v. 120, no. 7–8, p. 847–876, doi:10.1130/ MANUSCRIPT RECEIVED 1 APRIL 2011 Wang, Z.C., Zhang, P.Z., Garzione, C.N., Lease, R.O., B26232.1. REVISED MANUSCRIPT RECEIVED 9 AUGUST 2011 Zhang, G., Zheng, D.W., Hough, B., Yuan, D.Y., Li, C., Yuan, D.Y., Zhang, P.Z., Lei, Z.S., Liu, B.C., and Liu, X.L., MANUSCRIPT ACCEPTED 16 AUGUST 2011 Liu, J., and Wu, Q., 2011b, Magnetostratigraphy and 2005, A preliminary study on the new activity fea- depositional history of the Miocene Wushan Basin on tures of the Lajishan mountain fault zone in Qinghai Printed in the USA

Geological Society of America Bulletin, May/June 2012 677