Pre-Miocene birth of the Yangtze RiverHongbo Zhenga,1, Peter D. Cliftb, Ping Wanga, Ryuji Tadac, Juntao Jiad, Mengying Hee, and Fred Jourdanf

aSchool of Geography Science, Nanjing Normal University, Nanjing 210023, China; bDepartment of Geology and Geophysics, Louisiana State University,Baton Rouge, LA 70803; cDepartment of Earth and Planetary Science, University of Tokyo, Tokyo 113-0033, Japan; dSchool of Geosciences, China PetroleumUniversity, Qingdao 266580, China; eSchool of Earth Science and Engineering, Nanjing University, Nanjing 210093, China; and fWestern Australian ArgonIsotope Facility, Department of Applied Geology and John de Laeter Centre, Curtin University, Perth, WA 6845, Australia

Edited by Paul Tapponnier, Earth Observatory of Singapore, Singapore, and approved March 22, 2013 (received for review September 19, 2012)

The development of fluvial systems in East Asia is closely linked tothe evolving topography following India–Eurasia collision. Despitethis, the age of the Yangtze River systemhas been strongly debated,with estimates ranging from 40 to 45Ma, to a more recent initiationaround 2Ma. Here,we present 40Ar/39Ar ages from basalts interbed-ded with fluvial sediments from the lower reaches of the Yangtzetogether with detrital zircon U–Pb ages from sand grains withinthese sediments. We show that a river containing sediments indis-tinguishable from the modern river was established before ∼23Ma.We argue that the connection through the Three Gorges must post-date 36.5 Ma because of evaporite and lacustrine sedimentation inthe Jianghan Basin before that time. We propose that the presentYangtze River system formed in response to regional extensionthroughout eastern China, synchronous with the start of strike–sliptectonismand surfaceuplift in eastern Tibet and fedby strengthenedrains caused by the newly intensified summer monsoon.

Asian monsoon | drainage capture | provenance | Subei Basin | Yangtze gravel

Major river systems are responsible both sculpting the land-scape over large areas of the continental crust, as well as

controlling the development of offshore geology along continentalmargins. Despite this significance it is often unclear why majorriver systems are initially formed and what processes influencetheir development. The Yangtze River is one of the largest in EastAsia and has been the subject of debate for more than a century(1). Despite disagreement on the timing of its establishment mostgeologists agree that the development of this river representsa response to the evolving topography and climate of East Asia (2).Uplift of the Tibetan Plateau and subsidence in eastern China,following the cessation of Cretaceous arc magmatism within theCathaysia block during the Cenozoic (3) have acted to reverse anearlier Cretaceous southeast-to-northwest regional topographicgradient and allow the Yangtze to flow eastward into the EastChina Sea (Fig. 1). Continued debate about the age of highelevation in Asia means that constraining when the modernrivers achieved their present geometry is important because oftheir sensitivity to the changing regional topographic gradient.A proposed Pleistocene age for the Yangtze largely hinges onevidence that the modern delta had only been active since thattime (4, 5). However, extensive petroleum exploration in theSubei–South Yellow Sea Basin (Fig. 1) to the north of the moderndelta has revealed the stratigraphy of Late Cretaceous–Cenozoicsedimentation, indicating that large fluvial systems must have beenfeeding the basin since a much earlier time (6). In addition, U–Pbdating of zircons from the modern delta region shows that a riverindistinguishable from the modern stream was supplying sedimentsto that area before 3.2 Ma (7). Further constraints on the initiationof flow come from upstream where the modern Yangtze is con-trolled by flow through the First Bend (FB) and the Three Gorges(TG; Fig. 1). The age of incision of the Three Gorges has variouslybeen argued to postdate 750 ka (8) at one extreme and to be asearly as 40–45 Ma at the other (9). Here we address this debate bystudy of sediments in the lower reaches of the river in an attemptto determine when the river first achieved its present character.

Geological SettingDownstream of the Three Gorges, the river crosses the JianghanBasin (JHB; Fig. 1 and Fig. S1), entering the East China Sea alongthe southern margin of the Subei–South Yellow Sea Basin. TheJianghan Basin begun rifting in the Late Cretaceous, as did theSubei–South Yellow Sea Basin, and became a well-developedextensional basin during the Paleogene (10, 11). Sedimentary se-quences, up to 7 km thick and spanning the Late Cretaceous topresent, preclude passage of the Yangtze through this regionbefore 36.5 Ma (12). This conclusion is based on the presence oflacustrine and, especially, evaporite sediments (up to 2 km thick)whose depositional age is controlled by well-dated, intercalatedvolcanic rocks (12, 13). The presence of evaporites and organic-rich lacustrine sediments is incompatible with flow of the Yangtzethrough that basin at that time because evaporate require a weaksupply of water. If the middle and upper Yangtze existed before36.5 Ma, then they must have drained in a different direction,likely toward the southeast and into the Red River (14–16).DuringNeogene time, eastern China entered a postrifting phase

characterized by thermal subsidence, forming regional down-warping depressions with basin fills onlapping over the earlierfault-bounded rift sequences (6). The Yangtze gravels, which areobserved along the lower reaches all of the way from the ThreeGorges to the delta, are of fluvial nature and were depositedduring this postrift phase.Basin formation has been linked to volcanic activity, which is

tholeiitic and basaltic during the Paleocene–Eocene in northernChina, but became more intense in eastern China in the middle tolate Miocene when the volcanism switched to alkaline and per-alkaline compositions (17). The Yangtze gravel sediments arepredominantly sands of fluvial facies and have widely been inter-preted as Pleistocene deposits (18), although scattered fossil woodfragments of Miocene age have also been discovered (19). In thisstudy, we examine sections close toNanjing (Fig. 2 andDataset S1)where the fluvial sediments are overlain by and interbedded withbasaltic lavas that provide the opportunity to accurately date thesediments. Sedimentary sequences underlying the dated basaltswere also selected for provenance studies.

40Ar/39Ar Dating of BasaltsThe groundmass approach has been widely used to date the rapidcooling young volcanic rocks immediately after their eruption andhas been successfully applied worldwide (20), because of thegeneral freshness of young rocks that results in groundmass40Ar/39Ar ages that are internally concordant and fully consistentwith other geological constraints and isotopic methods. Dating ofbasaltic lavas was performed at the Western Australian Argon

Author contributions: H.Z. designed research; H.Z., P.W., R.T., J.J., M.H., and F.J. per-formed research; H.Z., P.D.C., and R.T. analyzed data; and H.Z., P.D.C., and P.W. wrotethe paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.1To whom correspondence should be addressed. E-mail: zhenghb@njnu.edu.cn.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1216241110/-/DCSupplemental.

7556–7561 | PNAS | May 7, 2013 | vol. 110 | no. 19 www.pnas.org/cgi/doi/10.1073/pnas.1216241110

Dow

nloa

ded

by g

uest

on

June

24,

202

0

Isotope Facility at Curtin University, operated by a consortiumconsisting of Curtin University and the University of WesternAustralia (Dataset S2). Each sample was step heated in a doublevacuum high-frequency Pond Engineering furnace. This approachis designed to look at the gas released from sites of increasing argonretentivity. When a number of consecutive steps, carrying a sub-stantial amount of the total argon released, give the same age, theresulting average value carries geological significance. Plateau ages(Fig. S2) are given at the 2-σ level and are calculated using themean of all of the plateau steps, each weighted by the inversevariance of their individual analytical error and must definea probability of fit (P)> 0.05. Ages were calculated using the decayconstant of Renne et al. (21) and the Fish Canyon sanidine stan-dard for which an age of 28.305 Ma (±0.13%) was adopted.Uncertainties on the decay constants and the age of the standardare not included in the age calculation.The sample from Guizishan (Fig. S2A) yielded a complex age

spectrum where no straightforward eruption age could be calcu-lated because of the sudden drop in apparent age of three steps inthe middle of the spectrum, which was likely caused by late-stagealteration phases. Nevertheless, assuming alteration is the cause ofthe disturbance, a minimum age can be derived by using all but thethree discordant steps in an age spectrum. Because the individualstep ages are amixture between an older crystallization component

and a younger alteration component, an absolute age cannot beobtained for this sample. Nonetheless, we can say that the crys-tallization age must be older than the age given by the flat sectionsdefined by the oldest steps. This yielded a minimum age of 22.9 ±0.3 Ma (mean square weighted deviation = 1.5; P = 0.16) for theeruption of this lava flow. Samples from Lingyanshan (Fig. S2B)and Xiaopanshan (Fig. S2C) yielded 100% plateau ages of 10.32±0.13 Ma (P = 0.15) and 21.71 ± 0.17 Ma (P = 0.10), respectively,unambiguously indicating the eruption ages of these lava flows,with the 21.7 Ma age being consistent with the 22.9 Ma age fromGuizishan. The trapped 40Ar/36Ar compositions of the Guizishanand Xiaopanshan samples (Fig. S2 D and F) are indistinguishablefrom an atmospheric ratio of ∼299 (22) suggesting that no excess40Ar is present in these samples. No isochron could be obtained forsample Lingyanshan (Fig. S2E) because of a clustering of the datanear the radiogenic axis (and thus the absence of influence of anyminute trapped Ar on the age calculation).

U–Pb Zircon DatingZircon grains from sands in the Yangtze gravels were separatedand dated by the U–Pb method to constrain their provenance.This method dates the age at which each zircon crystal cooledbelow ∼750 °C (20) and is a widely used and accepted sedimentproxy method based on the concept that different crustal blocks

EC

S

EC

SC

SSSSB

SB

EC

SB

YYYSYSYSYSYSSSY

BBHBBHBBBHB

PRB

YB

YB

YB ththSouthth

Ch nanaChinana

Sea

F I CC IFP CCFIAPP

40°

30°

20°

90° 100° 110° 120° 130°

JHBJHBJHB

SBSB

QBQBSPGSPG

RedRedRedR

iver

Riv

er

Riv

er

NanjingNanjing

Yellow SeaYellow SeaYellow Sea

TGTG

FBFB

Pearl

Pearl

Pearl

YR-1

YR-2

Red R

iver Fault

Red R

iver Fault

Xianshuihe Fault

Xianshuihe Fault

Longm

enshan

Longm

enshan

East East

China China

SeaSea

East

China

Sea

Bohai Bohai

SeaSea

Bohai

Sea

T I B E TT I B E T

Yellow

Yellow

Yellow

RiverRiverRiver

River

River

River

RiverRiverRiver

Yangtz

e

Yangtz

e

Yangtz

e

Yalo

ng

Yalo

ng R

iver

Riv

er

WuhanWuhan

YichangYichang

Qinling-Dabie Belt

Qinling-Dabie Belt

Qinling-Dabie Belt

Tan

lu F

au

lt

Tan

lu F

au

lt

Tan

lu F

au

lt

Altyn-T

agh Fault

Altyn-T

agh Fault

Altyn-T

agh FaultTarimTarimTarim

Fig. 1. Topographic map of East Asia, showing ma-jor rivers and the locations mentioned in the text.SBSYB, Subei–South Yellow Sea Basin; BHB, BohaiBasin; ECSB, East China Sea Basin; PRB, Pearl RiverMouth Basin; YB, Yinggehai Basin; SPG, SongpanGarze; QB, Qiangtang Block. Red circles show loca-tions of Yellow River samples (31). Major faultsmarked are taken from Replumaz and Tapponnier(59). The black dashed box indicates the location ofFig. 2.

Paleozoic Mesozoic Neogene Quaternary Basalt

JHB

E1

N+Q

E3

E243.0

39.2

37.7

36.5

47.0

56.065.0

0

1

2

3

4

5

6K2

36.5 Basalt Age (Myr)

Halite GypsumSandstoneGravel Mudstone

Basalt Petrified woodSample

Depth

(km)

(d)LYS XPS GZS

0

5

10

15

20

25

10.3

0

5

10

15

22.90

5

10

15

20

21.7

Depth

(m)

(a) (b) (c)

For A and C:10 20 30 40 50 60

YangtzeLingyanshanXiaopanshan

I I’16080

0masl

km0

River

Nanjing

Luhe

32°3

0'32

°20'

32°1

0'32

°00'

109°00'108°50'108°40'108°30'

I

I’

Lingyanshan

Xiaopanshan

Guizishan

A B

C

Yang

tze

Rive

r

Fig. 2. (A) Simplified geological map of Nanjingarea showing the locations of studied sections. (B)Sedimentary logs from the studied Yangtze gravelsections in the lower reaches of the Yangtze basin incomparison with the lithostratigraphy of JianghanBasin. (a) Lingyanshan; (b) Xiaopanshan; (c) Guizishan;and (d) Jianghan Basin. Locations of sediments fromwhich zircons were dated are shown together withthe age controls. The y axis in the sedimentary logs isin depth; N+Q, Neogene and Pleistocene; E3, Oli-gocene; E2, Eocene; E1, Paleocene; K2, Late Creta-ceous. (C) Sketch showing a cross section (black linein A) of the geology of Nanjing area.

Zheng et al. PNAS | May 7, 2013 | vol. 110 | no. 19 | 7557

EART

H,A

TMOSP

HER

IC,

ANDPL

ANET

ARY

SCIENCE

SEN

VIRONMEN

TAL

SCIENCE

S

Dow

nloa

ded

by g

uest

on

June

24,

202

0

have different ages of formation and that erosion of the bedrockin a large and diverse drainage basin will result in a unique fin-gerprint for the zircon age population in any given river. Becauseof the complexity of the zircon age population in sands fromlarge river basins, samples with identical age populations arelikely derived from the same river system eroding the samebedrocks. Because zircon is relatively resistant to abrasion duringtransport, grains eroded from sources in the headwaters arecommunicated to the lower reaches. As a result all sources in thecatchment, not just those located close to the delta, contribute tothe bedload found in the lower reaches. The zircon age spectrumis controlled by the source exposure area and the concentrationof zircon in the bedrock (23), although focused erosion mayresult in certain parts of the basin yielding more sediment thanothers. The result is that the unique combination of bedrock age,relative abundance, and erosion intensity patterns makes it veryunlikely that two large rivers would have the same spectra ofzircon U–Pb ages.The location of the zircon samples within each stratigraphic

section are shown in Fig. 2. The sediments are of clear fluvialfacies and form cross-bedded sandstone complexes 5 to 10 mthick, which suggests deposition in a significant channel complex.The spectrum of the detrital zircon grain ages (Dataset S3) fromeach of the considered samples is shown in Fig. 3 using thekernel density estimation (KDE) of Vermeesch (24), which plotsthe detrital ages as a set of Gaussian distributions, but does notexplicitly take into account the analytical uncertainties. Vermeeschargues that this is a more statistically robust method for looking atdetrital ages when the number of grains and analytical precision areboth high. This method allows the age ranges and abundances ofthe different age populations to be graphically assessed.Detailed comparison is possible by dividing the zircons up into

population groups associated with the major tectono-thermalevents that have affected the crust of eastern Asia. Specificallywe assign grains to groups with the following age ranges: 100–200 Ma (Dabieshan and lower reaches); 200–250 Ma (Indo-sinian/Qiangtang Block) (25); 400–500 Ma (Songpan Garze) (26);700–1,000 Ma (Yangtze Craton and Songpan Garze) (27); 1,700–2,100 Ma (Songpan Garze); and 2,400–2,600 Ma (Songpan Garze

and North China Craton) (28). We assess the potentially differentprovenance of each of the sands considered in this study by con-sidering the relative abundance of each of these age groups be-tween different samples. Because individual age ranges are notunique to a single source terrain, and indeed because sourcescontain grains of different ages, it is not possible to assign in-dividual grain to a single source based only on the U–Pb age, al-though it is possible to eliminate possible sources. Furthermore,a relatively small number of grains do not fall within these groups,but because little data are lost by removing these and, becausethese grains are not source diagnostic, they are not helpful indemonstrating how the sediment load of the river has changedthrough time. We do not try to separate grains associated with theEmeishan flood basalt province because the age range of thatprovince is narrow (29) and cannot be unequivocally resolved fromthe 200–250 Ma range of the Indosinian Orogeny using the in-ductively coupled plasma mass spectrometry (ICP-MS) methodused here. After excluding grains that do not fall within the di-agnostic groups the budgets were recalculated to 100% andplotted as pie charts to show the overall character and source ofthe detrital zircons (Fig. S3). These figures also allow the threeage populations younger than 500 Ma to be better resolved into100–200 Ma (Mesozoic arc magmatism/Yanshanian Orogeny),200–250 Ma (Indosinian Orogeny), and 400–500 Ma (“Cale-donian” Orogeny) (30). As we note above, these age ranges arenot unique to single sources, but rather to tectono-thermalevents that affected eastern China but which are heterogeneousacross the Yangtze drainage basin, so that changes in erosionpatterns might be expected to change the age spectrum of thezircon sand grains in the river sediment.In addition to the modern river and Yangtze gravel samples

we plot all of the post–3.2-Ma detrital zircons from the YangtzeDelta (7) in an attempt to provide a general image of the sedi-ment reaching the delta since the Late Pliocene (Fig. 3). Jia et al.(7) have shown that the provenance of the sediment in the areaof the modern delta has been approximately constant and similarto the modern river for 3.2 Ma. Our intention is to providea more time-integrated image of recent flux to the delta becausethere are moderate variations over short time periods, as dem-onstrated by the differences between the modern samples fromNanjing and Wuhan that show the river is not of a completelyconstant and homogeneous character but that the degree ofvariation is limited within modest bounds (Fig. 3). Although weaccept that provenance can change over different timescalesbecause of changing climatic patterns and evolving rock uplift, itis clear from the work of Jia et al. (7) that the Yangtze has beenrelatively stable in its provenance, and thus catchment size, for atleast 3.2 Ma. This alone is somewhat surprising given the po-tential influence of the Yellow River in the delta region (Fig. 1).To test for the influence of the Yellow River we plot zircon

U–Pb ages from this system using two samples taken from north ofthe Qinling-Dabie belt (31) (Fig. 1) and which might reflect theexpected composition of the Yangtze gravels if they had actuallybeen deposited from that stream and not the Yangtze River. Twosamples are plotted because each is taken at a potential capturepoint for flow into the Yangtze, and this also allows us to see someof the natural variation within the modern Yellow River. In therecent historical past the lower reaches of the Yellow River haveswitched between two routes, one in the north to the Bohai Sea (asis the case today) and the other to the south to the Yellow Sea (32).This has led to speculation that a switch of the routes has occurredin the geological history, which would have influenced the sedi-ments of the lower Yangtze system. Fig. 3 shows how the YellowRiver age spectra are markedly different from any of the samplestaken from the modern and Pleistocene Yangtze River (althoughthey are quite similar to one another), most notably in havinga small number of grains dating to 700–1,000 Ma and a highabundance of grains dating to 1,700–2,100 Ma. The complexity of

0 500 1000 1500 2000 2500 3000

>22.9 Ma, GuizishanshanN = 99

>21.7 Ma, Xiaopanshan - 1N = 91

>21.7 Ma, Xiaopanshan - 2N = 93

>10.3 Ma, LingyanshanN = 97

Modern Yangtze, NanjingN = 95

Modern Yangtze, WuhanN = 97

Yangtze delta average, 3.2 MaN = 526

Crystallization age (Ma)

Rela

tive

pro

bab

ility

Modern Yellow River -1, N = 84Modern Yellow River -2, N = 99

Fig. 3. KDE diagrams of zircon populations from Yangtze River gravelsshowing the similarity of the provenance to the modern river at Nanjing, atWuhan, and the average composition of the delta 3.2 Ma. Note the differentspectrum provided by the lower reaches of the Yellow River. Locations of thesampled outcrops are shown on Figs. 1 and 2.

7558 | www.pnas.org/cgi/doi/10.1073/pnas.1216241110 Zheng et al.

Dow

nloa

ded

by g

uest

on

June

24,

202

0

the zircon age population demonstrates that the Yangtze gravelscould not represent the deposits of a locally sourced tributary be-cause this would not have sufficient diversity in its bedrock togenerate a complicated age spectrum of this variety.The pie charts (Fig. S3) represent a simplified version of the

KDE diagrams shown in Fig. 3 and are consistent with the visualimpression that all of the Yangtze gravel and river samples havesimilar zircon populations. Not surprisingly the 700–1,000 Magrains associated with the Yangtze Craton dominate in all thesesamples. Some variations with time since 23 Ma are to beexpected because of active tectonics in the headwaters (32, 33)and changing monsoon intensities since that time, which changesthe patterns of erosion within the basin. Nonetheless, it is ap-parent that all of the Yangtze gravel samples have received zir-cons from the same array of sources as the modern river and thatalthough the proportion of any particular age group varies, theextent of that variation is limited. We conclude that at least since∼23 Ma, a river similar to the modern Yangtze flowed in thevicinity of Nanjing and that this must have connected to themiddle and upper reaches of the present drainage, as the riverdoes today Fig. 4). The Yangtze gravels are not the product ofsedimentation from the Yellow River or a single, local Yangtzetributary, which would be dominated by erosion from the DabieShan where rocks dating at 100–200 Ma old dominate (34).To quantify the degree of similarity we here perform additional

statistical tests to assess the degree to which the Miocene sedi-ments differ from themodern river atNanjing, atWuhan, and fromthe 3.2-Ma-to-present record at the delta. Statistics suggest that∼100 grains are required to generate a reliable provenance recordin a large, complicated drainage like the Yangtze (24). Modernsediments from the Yangtze Delta, not surprisingly, contain zir-cons from all major source terrains, including some reworked fromolder sedimentary rocks. We compare that population with thezircon U–Pb ages from the Yangtze gravels and find that there isstriking degree of similarity. Kolmogorov–Smirnov (K–S) statis-tical testing cannot prove that two sands are identical, but it doesindicate when there are significant differences between samples.Our analysis (SI Text and Dataset S4) indicates that the Miocenesands are indistinguishable from the modern sediments or fromthose in the post–3.2-Ma delta, but are different from the YellowRiver.We argue that this provides proof of a river flowing throughthe lower reaches since 23 Ma, and that this stream was collectingsediments from all of the terrains that now feed the Yangtze. Thefact that the same age populations can also be seen before 22.9Ma(Fig. 3) suggests that this river has been in a state of continuousflow from that time until the present day. It appears that theriver evolved close to its modern state at least by the Miocene–Oligocene boundary (∼24 Ma).

Drainage Evolution in East AsiaWe argue that although sedimentation in the region of themodern delta only dates from the Pliocene (4, 5) this does notpreclude an earlier delta located farther north in the Subei–SouthYellow Sea Basin. Moreover, an Early Miocene age for the onsetof Yangtze River flow is consistent with Nd isotopic data from theRed River system that indicate loss of drainage from the middleYangtze into that river close to the Oligocene–Miocene bound-ary (15) (ca. 24 Ma; Fig. 4). Pb isotope characteristics ofK-feldspar grains in the Red River delta show a connection withthe middle Yangtze in the Eocene, a link that had been lost by themiddle Miocene (35), but little firm evidence for a link with theupper reaches of the Yangtze to the Red River. Similarly, U–Pbzircon provenance data from the lower Red River indicates thatthis river had achieved its present geometry before 12Ma (36).Wefollow studies such as Clark et al. (16) in advocating capture andflow reversal of the middle Yangtze (2,140 km between the FirstBend and the Three Gorges; ∼33% of the total trunk streamlength) by connection to the lower reaches before 23 Ma.

It could be argued that the composition of the Yangtze was ableto achieve a modern character by 23 Ma as a result of sedimentdelivery from the Yalongjiang (Fig. 1) rather than from theupper Yangtze, which could still have been connected to the RedRiver. However, the zircon age spectrum in the sediment in theYalongjiang is quite different from that at the First Bend on theYangtze, despite the fact that both rivers have major sources inthe Songpan Garze terrane. The Yalongjiang has a very highabundance of 600–1,000 Ma grains compared with the First Bend(76% versus 19%) and is relatively depleted in grains dated at300–600 Ma and 1,700–2,000 Ma (5% and 3% compared with23% and 25%; Fig. S3). As a result it is doubtful whether theYangtze 23 Ma would show the same zircon age spectrum as itdoes now without a connection to the Yangtze upstream of theFirst Bend. In any case, data from the Red River itself indicatelittle definitive evidence for a connection between the Red Riverand the upperYangtze at any time (36). In theory there could havebeen an early Miocene phase with the upper Yangtze connectedto the Red River and the Yalongjiang connected to the rest of theYangtze, although it would be unlikely that the zircon population

10˚

20˚

30˚

90˚ 100˚ 110˚ 120˚

10˚

20˚

30˚

South ChinaSouth China

North ChinaNorth ChinaNorth China

IndochinaIndochinaIndochina

IndochinaIndochinaIndochina

QiangtangQiangtang

QaidamQaidamQaidamQinling-DabieQinling-DabieQinling-Dabie

South ChinaSouth China

North ChinaNorth ChinaNorth China

QiangtangQiangtangSongpanSongpan

SongpanSongpanSichuanSichuan

SichuanSichuan

QaidamQaidamQaidamQinling-DabieQinling-DabieQinling-Dabie

ASRRASRRASRR

LMS

LMS

LMS

ASRRASRRASRR

IndiaIndiaIndia

IndiaIndiaIndia

HimalayaHimalayaHimalaya

LhasaLhasaLhasa

LhasaLhasaLhasa

?

?

?

?

?A

~32 Ma

B~16 Ma

SBSYBSBSYB

SBSYBSBSYB

ECSBECSB

ECSBECSB

PRBPRB

PRBPRB

YBYB

YBYBSCSSCS

JHBJHB

JHBJHB

Paleo-Red River

Red River

Yangtze

Bangong-Nujiang sutureBangong-Nujiang suture

Bangong-Nujiang sutureBangong-Nujiang suture

Yarlung-Tsangpo suture

Yarlung-Tsangpo suture

Ya rlung-Tsangpo sutureYa rlung-Tsangpo suture

Fig. 4. Simplified maps showing the development of the Yangtze River inresponse to tectonic evolution in East Asia, based on the data presented inthis paper as well as associated studies discussed in the text. (A) 32 and (B)16 Ma. Names of basins as in Fig. 1. Maps reflect the progressive extrusion ofIndochina and the opening of the South China Sea, but recognize that thecontinental blocks east of Longmenshan have remained relatively rigid sincethe Eocene (59). ASRR, Ailao Shan–Red River Fault; LMS, Longmenshan.Shaded river segment indicates the reversed section of the Yangtze River.

Zheng et al. PNAS | May 7, 2013 | vol. 110 | no. 19 | 7559

EART

H,A

TMOSP

HER

IC,

ANDPL

ANET

ARY

SCIENCE

SEN

VIRONMEN

TAL

SCIENCE

S

Dow

nloa

ded

by g

uest

on

June

24,

202

0

of the Yangtze load would not have been disturbed by the captureof the upper reaches.Although Richardson et al. (9) favored a connection of the

Yangtze through the Three Gorges before 40 Ma the presence ofevaporites in the Jianghan Basin indicates that the exhumationof the Three Gorges at 40–45 Ma could not have been caused byinitiation of a throughgoing river flowing in that direction. Al-though a major river could (and indeed does) flow through a lakesystem, it is impossible for a lake in the Jianghan Basin to everbecome so desiccated that evaporites, with a thickness up to2 km, would form because this would be inconsistent with sig-nificant discharge from the river. Although some rivers may formfollowing a process of alternating fluvial and lacustrine phases weknow that if this process had been operating in the Yangtze itmust have ceased before 23 Ma because the provenance ob-served in the lower reaches requires a fully connected river of thetype we see today. A river of this scale is unlikely to be seasonaleven during arid climatic conditions, which in any case are notcharacteristic of the early and middle Miocene in East Asia (37).Sedimentation rates in the East China Sea show a steady in-crease during the Oligocene (38), consistent with our model,however there is no sharp jump in accumulation rate, probablybecause of sediment buffering in depocenters onshore, mostlynotably the Jianghan and Subei basins (39).

Tibetan Tectonics and the Yangtze RiverBirth of theYangtzeRiver around the start of theMiocene can alsobe understood in termsof a progressive or stepwise Paleogeneupliftof the Tibetan Plateau, because recent paleoaltitude studies sup-port the idea of central and southern Tibet being close to modernelevations no later than the middleMiocene (∼16Ma), rather thanindicating a late-stage rapid uplift in the Pleistocene (40).Although (U–Th)/He thermochronometry from the Long-

menshan of western Sichuan indicates that major topographic upliftpredated 8–11 Ma (41, 42), a recent study using a variety of ther-mochronology proxies from this same area indicates significant ex-humation starting during the Oligocene–early Miocene and thepresence of some topography in eastern Tibet even predating India–Eurasia collision (43). These data are broadly consistent with tectonicmodels indicating a stepwise growth of the plateau with southeasternTibet/southwest China uplifting in Oligocene times (44).Strike–slip tectonism strongly affected the geology and topog-

raphy of the eastern plateau margin and must have been in-strumental in helping to guide the reorganization of rivers in theregion. The southeastern flank of the Tibetan Plateau is deformedby large strike–slip faults the activity of which is likely linked to thesurface uplift. Dating of the Red River fault zone shows that mo-tion started ∼34Ma and was most rapid after 27 and before 17Ma(45, 46) and other faults in this region (e.g., Xianshui He Fault)share similar ages of motion and exhumation (47, 48). Lacassinet al. (49) note strong shearing and metamorphism of the moun-tains immediately north of the Yangtze First Bend after 36 Ma,followed by uplift driven by folding around 17 Ma. This daterequires that the river was flowing through that region before17 Ma, although it need not necessarily have been connected tothe present day Yangtze. It can be imagined that surface upliftdriven by this deformation and related to the progressive growthof Tibetan topography toward the southeast might trigger majorreorganization of the rivers flowing through this area and that theinitial reorganization of the rivers would occur before major sur-face uplift (16, 50), i.e., predating 17 Ma.We emphasize that headwater capture does not require major

surface uplift in the First Bend area, but only a gentle retilting ofthe topography toward the east, because water will flow downhilleven if the slope is not very steep. Indeed, the entrenchment of theYangtze and its tributaries into deep canyons during the LateMiocene (51) makes further capture after that time very difficultbecause the topography prevents lateral channel migration. An

alternative view is that the Yangtze and other Southeast Asianrivers originated outside the Tibetan Plateau and then cut into thisflat, high-standing massif as a result of headward retreat, ratherthan incising into a low-altitude peneplain that was then uplifted(52). Our data would also be consistent with this model but wouldrequire most of the head retreat to be complete by 23 Ma.If the current geometry of the Yangtze reflects the evolving

topography, it is hard to understand how this river could post-date the large-scale late Miocene uplift of eastern Tibet. Bydating the Yangtze as being older than 23 Ma we now synthesizeour understanding of the river with what is known of topographicgrowth in the headwaters (Fig. 4). Furthermore, as Tibet con-tinued to grow, sedimentary basins across eastern China expe-rienced the initiation of enhanced early-to-middle Miocenesubsidence (53), which are connected by throughgoing fluvialsystems, as the case of Jianghan and Subei–South Yellow Seabasins today. Within-plate magmatism testifies to a sharp changein the geodynamic regime at that time, likely linked to rollback ofthe subducting Pacific plate (54). Initial uplift of Tibet and re-gional strike–slip faulting driving uplift coupled with subsidenceof basins across eastern China opened a path for the Yangtze tothe east, diverting the upper and middle reaches away from theRed River and South China Sea.We further note that the pre-Miocene uplift of the Tibetan

Plateau implied by our reconstruction and by recent thermo-chronology (43) is also consistent with recent revisions concerningthe time of intensification of the East Asian summer monsoonclose to the start of the Miocene (37, 55) rather than during thelate Miocene ∼8 Ma as had previously been believed (56). Thesenew dates for monsoon strengthening are based on environ-mental and floral proxies and are not model dependent and moreclosely linked to precipitation than had been the case for earliermonsoon indicators. Prell and Kutzbach (57) are among severalstudies that predict a correlation between increasing plateau al-titude and monsoon intensity. Although it is clear that monsoonintensity is controlled by more than simply Tibetan elevation,recent climate models still show that the presence of a plateauintensifies summer rainfall in the Asian region (58). Thus, thetimescale of uplift implied by the start of Yangtze throughflowbefore 23 Ma is consistent with the more recent climate recon-structions that indicate strengthening around the end of theOligocene (37, 55).

ConclusionsIn this study we constrain the age of initiation of theYangtze Riverthrough study of the Yangtze gravels in the region of Nanjing. The40Ar/39Ar dating of basalts interbedded within these sedimentsshows that some of these were deposited before 23Ma, and are notPleistocene as previously assumed. We further analyzed theprovenance of the gravels through application of U–Pb dating tozircon sand grains extracted from these deposits. Statistical anal-ysis shows that lower Miocene sediments are indistinguishablefrom sediments collected from the modern river or from the deltaduring the last 3.2 Ma. This requires that a river, much like themodern one, was in existence before 23Ma. Our data do not allowus to date the oldest possible age of the Yangtze, but we infer thatthe river did not start before the late Eocene because of thepresence of evaporites in the Jianghan Basin before that time.Those sediments lie just downstream of the Three Gorges andpreclude major river flow until after 36.5 Ma. A pre–23-MaYangtze River is consistent with both recent estimates for thetiming of eastern Tibetan uplift, the onset of major strike–slipdeformation, and with the age of headwater capture of the middleYangtze reaches away from the Red River. Our study confirmsa close relationship between the evolving tectonically driven to-pography and drainage systems in East Asia during the Cenozoic.

7560 | www.pnas.org/cgi/doi/10.1073/pnas.1216241110 Zheng et al.

Dow

nloa

ded

by g

uest

on

June

24,

202

0

ACKNOWLEDGMENTS. The authors thank Philippe Leloup, Zhenyu Yang,Lin Ding, and Xiumian Hu for fruitful discussions. P.C. thanks the HanseWissenschaftskolleg for providing financial support to work in Asia. Thiswork was financially supported by the “Strategic Priority Research

Program” of the Chinese Academy of Sciences (XDB03020301), the Na-tional Science Foundation of China (NSFC 40830107, 41111140016, and41102104) and United Nations Educational, Scientific, and CulturalOrganization IGCP581.

1. Willis BJ, et al. (1907) Research in China (Carnegie Institution of Washington, Wash-ington, DC).

2. Wang P (2004) Cenozoic deformation and the history of sealand interactions in Asia.Continent–Ocean Interactions in the East Asian Marginal Seas, eds Clift P, Wang P,Kuhnt W, Hayes D (American Geophysical Union, Washington, DC), Vol 149, pp 1–22.

3. Zhou D, et al. (2008) Mesozoic paleogeography and tectonic evolution of South ChinaSea and adjacent areas in the context of Tethyan and Paleo-Pacific interconnections.Isl Arc 17(2):186–207.

4. Fan D, Li C (2008) Timing of the Yangtze initiation draining the Tibetan Plateauthroughout to the East China Sea: A review. Front Earth Sci China 2(3):302–313.

5. Li C, Chen Q, Zhang J, Yang S, Fan D (2000) Stratigraphy and paleoenvironmentalchanges in the Yangtze delta during late Pleistocene. J Asian Earth Sci 18:453–469.

6. Ren J, Tamaki K, Lu S, Zhang J (2002) Late Mesozoic and Cenozoic rifting and itsdynamic setting in Eastern China and adjacent areas. Tectonophysics 344:175–205.

7. Jia JT, et al. (2010) Detrital zircon U-Pb ages of Late Cenozoic sediments from theYangtze delta: Implication for the evolution of the Yangtze River. Chin Sci Bull 55:1520–1528.

8. Xiang F, et al. (2007) Quaternary sediment in the Yichang area: Implications for theformation of the Three Gorges of the Yangtze River. Geomorphology 85:249–258.

9. Richardson NJ, Densmore AL, Seward D, Wipf M, Yong L (2010) Did incision of theThree Gorges begin in the Eocene? Geology 38(6):551–554.

10. Peters KE, Cunningham AE, Walters CC, Jigang J, Zhaoan F (1996) Petroleum systemsin the Jiangling-Dangyang area, Jianghan basin, China. Org Geochem 24:1035–1060.

11. Dai SZ (1997) Petroleum Geology of Jianghan Saline Basin (Petroleum Industry Pub-lishing House, Beijing).

12. Zheng H, Jia D, Chen J, Wang P (2011) Forum comment: Did incision of the ThreeGorges begin in the Eocene? Geology, 10.1130/G31944C.1.

13. Xu L, et al. (1995) Chronology of Paleogene volcanic rocks in the Jianghan Basin. OilGas Geol 16:132–137.

14. Brookfield ME (1998) The evolution of the great river systems of southern Asia duringthe Cenozoic India-Asia collision; Rivers draining southwards. Geomorphology22(3-4):285–312.

15. Clift PD, Blusztajn J, Nguyen DA (2006) Large-scale drainage capture and surfaceuplift in eastern Tibet-SW China before 24 Ma inferred from sediments of the HanoiBasin, Vietnam. Geophys Res Lett 33(L19403).

16. Clark MK, et al. (2004) Surface uplift, tectonics, and erosion of eastern Tibet fromlarge-scale drainage patterns. Tectonics 23:TC1006.

17. Ma X, Wu D (1987) Cenozoic extensional tectonics in China. Tectonophysics 133(3–4):243–255.

18. Yang H, Xu X, Yang D (1995) Environmental Changes and Ecosystems in the Middleand Lower Yangtze River (Hohai Univ Press, Nanjing, China).

19. Hu D, Xu R, Yang J, Qi G (2005) The geological age of the angiospermous fossil woodgroup and the paleogravel bed in Yangluo of Hubei. Resour Environ Eng 19:12–15.

20. Hodges K (2003) Geochronology and thermochronology in orogenic systems. TheCrust, ed Rudnick R (Elsevier-Science, Amsterdam), pp 263–292.

21. Renne PR, Mundil R, Balco G, Min K, Ludwig KR (2010) Joint determination of 40Kdecay constants and 40Ar*/40K for the Fish Canyon sanidine standard, and improvedaccuracy for 40Ar/39Ar geochronology. Geochim Cosmochim Acta 74:5349–5367.

22. Lee J-Y, et al. (2006) A redetermination of the isotopic abundance of atmospheric Ar.Geochim Cosmochim Acta 70:4507–4512.

23. Amidon WH, Burbank DW, Gehrels GE (2005) Construction of detrital mineral pop-ulations: Insights from mixing of U-Pb zircon ages in Himalayan rivers. Basin Res 17(4):463–485.

24. Vermeesch P (2004) How many grains are needed for a provenance study? EarthPlanet Sci Lett 224:351–441.

25. Roger F, et al. (2003) Geochronological and geochemical constraints on Mesozoicsuturing in east central Tibet. Tectonics 22(4).

26. Weislogel AL, et al. (2006) Detrital zircon provenance of the Late Triassic Songpan-Ganzi complex: Sedimentary record of collision of the North and South China blocks.Geology 34:97–100.

27. Zheng J, et al. (2006) Widespread Archean basement beneath the Yangtze Craton.Geology 34(6):417–420.

28. Diwu C, et al. (2008) U-Pb ages and Hf isotopes for detrital zircons from quartzite inthe Paleoproterozoic Songshan Group on the southwestern margin of the NorthChina Craton. Chin Sci Bull 53(18):2828–2839.

29. Xu YG, Luo ZY, Huang XL (2008) Zircon U-Pb and Hf isotope constraints on crustalmelting associated with the Emeishan mantle plume. Geochim Cosmochim Acta 72:3084–3104.

30. Lu S, Li H, Yu H, Zhao F, Yang C (1999) Neoproterozoic Orogeny in NorthwesternChina. Gondwana Res 2(4):610–611.

31. Yang J, et al. (2009) Episodic crustal growth of North China as revealed by U-Pb ageand Hf isotopes of detrital zircons from modern rivers. Geochim Cosmochim Acta73(9):2660–2673.

32. Chen Z, et al. (1999) GPS monitoring of the crustal motion in southwestern China.Chin Sci Bull 44(19):1804–1807.

33. Wang E, et al. (1998) Late Cenozoic Xianshuihe-Xiaojiang, Red River, and Dali FaultSystems of Southwestern Sichuan and Central Yunnan, China (Geological Society of

America, Boulder, CO).34. Ma LF (2000) The Geological Atlas of China (Geological Publishing House, Beijing) (in

Chinese).35. Clift PD, et al. (2008) Evolving East Asian river systems reconstructed by trace element

and Pb and Nd isotope variations in modern and ancient Red River-Song Hong

sediments. Geochem Geophys Geosyst 9(Q04039).36. Hoang LV, Wu FY, Clift PD, Wysocka A, Swierczewska A (2009) Evaluating the evo-

lution of the Red River system based on in-situ U-Pb dating and Hf isotope analysis of

zircons. Geochem Geophys Geosyst 10(Q11008).37. Clift PD, et al. (2008) Greater Himalayan exhumation triggered by Early Miocene

monsoon intensification. Nat Geosci 1:875–880.38. Clift PD, Layne GD, Blusztajn J (2004) Marine sedimentary evidence for monsoon

strengthening, Tibetan uplift and drainage evolution in east Asia. Continent–Ocean

Interactions in the East Asian Marginal Seas, eds Clift P, Kuhnt W, Wang P, Hayes D,

Geophysical Monograph (American Geophysical Union, Washington, DC), Vol 149,pp 255–282.

39. Wu S, Ni XN, Cai F (2008) Petroleum geological framework and hydrocarbon potentialin the Yellow Sea. Chin J Oceanol Limnol 26(1):23–34.

40. Rowley DB, Currie BS (2006) Palaeo-altimetry of the late Eocene to Miocene Lunpola

basin, central Tibet. Nature 439(7077):677–681.41. Godard V, et al. (2009) Late Cenozoic evolution of the central Longmen Shan, eastern

Tibet: Insight from (U–Th)/He thermochronometry. Tectonics 28(TC5009).42. Kirby E, et al. (2002) Late Cenozoic evolution of the eastern margin of the Tibetan

Plateau: Inferences from Ar-40/Ar-39 and (U-Th)/He thermochronology. Tectonics

21(1):1–19.43. Wang E, et al. (2012) Two-phase growth of high topography in eastern Tibet during

the Cenozoic. Nat Geosci 5:640–645.44. Tapponnier P, et al. (2001) Oblique stepwise rise and growth of the Tibet plateau.

Science 294(5547):1671–1677.45. Leloup PH, et al. (2001) New constraints on the structure, thermochronology, and

timing of the Ailao Shan-Red River shear zone, SE Asia. J Geophys Res 106(B4):

6657–6671.46. Gilley LD, et al. (2003) Direct dating of left-lateral deformation along the Red River

shear zone, China and Vietnam. J Geophys Res 108(2127).47. Akciz S, Burchfiel BC, Crowley JL, Jiyun Y, Liangzhong C (2008) Geometry, kinematics

and regional significance of the Chong Shan shear zone, Eastern Himalayan Syntaxis,

Yunnan, China. Geosphere 4:292–314.48. Wang Y, et al. (2006) Kinematics and 40Ar/39Ar geochronology of the Gaoligong and

Chongshan shear systems, western Yunnan, China: Implications for early Oligocene

tectonic extrusion of SE Asia. Tectonophysics 418(3-4):235–254.49. Lacassin R, et al. (1996) Tertiary deformation and metamorphism SE of Tibet: The

folded tiger-leap décollement of NW Yunnan, China. Tectonics 15(3):605–622.50. Yan Y, et al. (2012) Constraints on Cenozoic regional drainage evolution of SW China

from the provenance of the Jianchuan Basin. Geophys Geochem Geosyst 13(Q03001).51. Clark MK, et al. (2005) Late Cenozoic uplift of southeastern Tibet. Geology 33(6):

525–528.52. Liu-Zeng J, Tapponnier P, Gaudemer Y, Ding L (2008) Quantifying landscape differ-

ences across the Tibetan plateau: Implications for topographic relief evolution.J Geophys Res 113(F04018).

53. Kusky T, Windley BF, Zhai MG (2007) Tectonic evolution of the North China Block:From Orogen to Craton to Orogen. Mesozoic Sub-Continental Lithospheric Thinning

Under Eastern Asia, Special Publications, eds Zhai MG, Windley BF, Kusky TM,

Meng QR (Geological Society, London), Vol 280, pp 1–34.54. Zhang HF, et al. (2003) Secular evolution of the lithosphere beneath the eastern

North China craton: Evidence from Mesozoic basalts and high-Mg andesites. Geochim

Cosmochim Acta 67:4373–4387.55. Sun X, Wang P (2005) How old is the Asian monsoon system? Palaeobotanical records

from China. Palaeogeogr Palaeoclimatol Palaeoecol 222(3-4):181–222.56. Molnar P, England P, Martinod J (1993) Mantle dynamics, uplift of the Tibetan Pla-

teau, and the Indian Monsoon. Rev Geophys 31(4):357–396.57. Prell WL, Kutzbach JE (1992) Sensitivity of the Indian Monsoon to forcing parameters

and implications for its evolution. Nature 360(6405):647–652.58. Huber M, Goldner A (2012) Eocene monsoons. J Asian Earth Sci 44:3–23.59. Replumaz A, Tapponnier P (2003) Reconstruction of the deformed collision zone

between India and Asia by backward motion of lithospheric blocks. J. Geophys Res

108:2285(B6).

Zheng et al. PNAS | May 7, 2013 | vol. 110 | no. 19 | 7561

EART

H,A

TMOSP

HER

IC,

ANDPL

ANET

ARY

SCIENCE

SEN

VIRONMEN

TAL

SCIENCE

S

Dow

nloa

ded

by g

uest

on

June

24,

202

0