Top Banner
Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China Sea: A review of competing proxies Peter D. Clift a,b, , Shiming Wan c , Jerzy Blusztajn d a Department of Geology and Geophysics, Louisiana State University, Baton Rouge, LA 70803, USA b South China Sea Institute of Oceanology, Chinese Academy of Sciences, 164 Xingang Road, Guangzhou, China c Key Laboratory of Marine Geology and Environment, Institute of Oceanology, Chinese Academy of Sciences, Qingdao 266071, China d Department of Geology and Geophysics, Woods Hole Oceanographic Institution, Woods Hole, MA 02540, USA abstract article info Article history: Received 9 October 2013 Accepted 7 January 2014 Available online 15 January 2014 Keywords: Monsoon Erosion Weathering Clay mineralogy Sediment geochemistry South China Sea Reconstructing the changing strength of the East Asian summer monsoon has been controversial because different proxies, many being indirect measures of rainfall, tell contrasting stories about how this has varied over long periods of geologic time. Here we present new Sr isotope, grain-size and clastic ux data and synthesize existing proxies to reconstruct changing chemical erosion in the northern South China Sea since the Oligocene, using the links between weathering rates and monsoon strength established in younger sediments as a way to infer inten- sity. Chemical proxies such as K/Rb, K/Al and the Chemical Index of Alteration (CIA), together with clay proxies like kaolinite/(illite + chlorite) show a steady decline in alteration after a sharp fall following a maximum at the Mid Miocene Climatic Optimum (MMCO; 15.517.2 Ma), probably as a result of cooling global temperatures. In contrast, physical erosion proxies, including bulk Ti/Ca and clastic mass accumulation rates (MAR), show peaks at 2123 Ma, ~19 Ma and 15.517.2 Ma, implying faster run-off in the absence of drainage capture. Rates increase again, likely driven by slightly increased run-off after 13 Ma, but decrease after 8 Ma, which is identied as a period of summer monsoon weakening. Sr isotope composition correlates with hematite/goethite and the spectral proxy C RAT to show stronger weathering linked to more monsoonal seasonality. These proxies argue for a strengthening of the East Asian Monsoon after 2223 Ma, followed by an extended period of monsoon maximum between 18 and 10 Ma, then weakening. There is some suggestion that the summer monsoon may have strengthened since 34 Ma after reaching a minimum in the Pliocene. © 2014 Elsevier B.V. All rights reserved. Contents 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 2. Monsoon Proxies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 3. Geological setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 4. Links between monsoon and erosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 5. Earlier monsoon reconstructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 6. Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 7. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 8. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 8.1. Physical erosion proxies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 8.2. Geochemical proxies for alteration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 8.3. Clay minerals as alteration proxies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 8.4. Exploring the C RAT proxy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 8.5. Sr isotope evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 8.6. Sediment provenance at ODP Sites 1146 and 1148 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 8.7. Environments in the Plio-Pleistocene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 8.8. Weathering in the Late Miocene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Earth-Science Reviews 130 (2014) 86102 Corresponding author at: Department of Geology and Geophysics, Louisiana State University, Baton Rouge, LA 70803, USA. Tel.: +1 225 578 2153; fax: +1 225 578 2302. E-mail address: [email protected] (P.D. Clift). 0012-8252/$ see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.earscirev.2014.01.002 Contents lists available at ScienceDirect Earth-Science Reviews journal homepage: www.elsevier.com/locate/earscirev
17

Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

May 30, 2020

Download

Documents

dariahiddleston
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

Earth-Science Reviews 130 (2014) 86–102

Contents lists available at ScienceDirect

Earth-Science Reviews

j ourna l homepage: www.e lsev ie r .com/ locate /earsc i rev

Reconstructing chemical weathering, physical erosion and monsoonintensity since 25 Ma in the northern South China Sea: A reviewof competing proxies

Peter D. Clift a,b,⁎, Shiming Wan c, Jerzy Blusztajn d

a Department of Geology and Geophysics, Louisiana State University, Baton Rouge, LA 70803, USAb South China Sea Institute of Oceanology, Chinese Academy of Sciences, 164 Xingang Road, Guangzhou, Chinac Key Laboratory of Marine Geology and Environment, Institute of Oceanology, Chinese Academy of Sciences, Qingdao 266071, Chinad Department of Geology and Geophysics, Woods Hole Oceanographic Institution, Woods Hole, MA 02540, USA

⁎ Corresponding author at: Department of Geology andE-mail address: [email protected] (P.D. Clift).

0012-8252/$ – see front matter © 2014 Elsevier B.V. All rihttp://dx.doi.org/10.1016/j.earscirev.2014.01.002

a b s t r a c t

a r t i c l e i n f o

Article history:Received 9 October 2013Accepted 7 January 2014Available online 15 January 2014

Keywords:MonsoonErosionWeatheringClay mineralogySediment geochemistrySouth China Sea

Reconstructing the changing strength of the East Asian summermonsoon has been controversial because differentproxies, many being indirect measures of rainfall, tell contrasting stories about how this has varied over longperiods of geologic time. Here we present new Sr isotope, grain-size and clastic flux data and synthesize existingproxies to reconstruct changing chemical erosion in the northern South China Sea since the Oligocene, using thelinks betweenweathering rates andmonsoon strength established in younger sediments as a way to infer inten-sity. Chemical proxies such as K/Rb, K/Al and the Chemical Index of Alteration (CIA), together with clay proxieslike kaolinite/(illite+ chlorite) show a steady decline in alteration after a sharp fall following amaximum at theMid Miocene Climatic Optimum (MMCO; 15.5–17.2 Ma), probably as a result of cooling global temperatures. Incontrast, physical erosion proxies, including bulk Ti/Ca and clasticmass accumulation rates (MAR), showpeaks at21–23 Ma, ~19 Ma and 15.5–17.2 Ma, implying faster run-off in the absence of drainage capture. Rates increaseagain, likely driven by slightly increased run-off after 13 Ma, but decrease after 8 Ma, which is identified as aperiod of summer monsoon weakening. Sr isotope composition correlates with hematite/goethite and thespectral proxy CRAT to show stronger weathering linked to more monsoonal seasonality. These proxies arguefor a strengthening of the East Asian Monsoon after 22–23 Ma, followed by an extended period of monsoonmaximum between 18 and 10 Ma, then weakening. There is some suggestion that the summer monsoon mayhave strengthened since 3–4 Ma after reaching a minimum in the Pliocene.

© 2014 Elsevier B.V. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 872. Monsoon Proxies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 883. Geological setting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 884. Links between monsoon and erosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 895. Earlier monsoon reconstructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 906. Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 917. Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 918. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

8.1. Physical erosion proxies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 928.2. Geochemical proxies for alteration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 938.3. Clay minerals as alteration proxies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948.4. Exploring the CRAT proxy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948.5. Sr isotope evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 958.6. Sediment provenance at ODP Sites 1146 and 1148 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 958.7. Environments in the Plio-Pleistocene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 968.8. Weathering in the Late Miocene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

Geophysics, Louisiana State University, Baton Rouge, LA 70803, USA. Tel.: +1 225 578 2153; fax: +1 225 578 2302.

ghts reserved.

Page 2: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

87P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

8.9. Erosion at the MMCO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 968.10. Erosion at 23.2–21.3Ma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 988.11. Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99

9. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

1. Introduction

The Asianmonsoon is often portrayed as the classic example of howevolution of the solid Earth is coupled to atmospheric processes, yet itstemporal evolution is not well defined over periods N10 Myr. This is amajor problem because the monsoon and the potentially associatedtectonic evolution of Asia may be one of the most significant processesaffecting global climatic conditions during the Cenozoic (Raymo et al.,1988; Berner and Berner, 1997; Wang et al., 2003b). However, demon-strating any linkage between climate and tectonics is impossible with-out a better understanding of when and how the monsoon developed,which can then be correlated, or not, with changing topography incentral Asia or the tectonics of the Himalaya (Clemens et al., 1991;Prell and Kutzbach, 1992; Molnar et al., 1993; Clift and Plumb, 2008).Reconstructingmonsoon intensity is hampered both by lack of continu-ous sedimentary sequences and by agreement onwhat the best proxiesto measure are, because there are few robust rainfall proxies, beyondsalinity in seawater, which is only well developed in the Bay of Bengal(Kudrass et al., 2001).

The key problem in reconstructing monsoon strength is to pick theright proxies to measure because the monsoon is strongly variableacross Asia, being wetter in some places, windier in others, whilesome places are dominated by winter rather than summer monsoons(Wang et al., 2003a). Proxies linked to wind strength are potentiallythe most reliable, but are often hard to extract at high resolution andover long time periods because they tend to accumulate only in con-densed deep sea sediments (Rea et al., 1998). Moreover, if continentalrainfall is the key variable that we wish to examine wind speed is notnecessarily the ideal process to measure. Rainfall is the process that ismost crucial to those trying to understand how the summer monsoonmay have affected the development of orogens in Asia throughenhanced erosion (Burbank et al., 2003; Wobus et al., 2003; Clift et al.,2008). Care needs to be exercised in equating total rainfall to monsoonintensity despite the fact that monsoon rain dominates that in much ofSouth and East Asia. This is because climate models predict coastal rain-fall inmost scenarios, with rainfall deep in the continental interior beinga feature of monsoons, alongwith a seasonal variation between dry andwet conditions (Webster et al., 1998).Moreover, strong bands of rainfallfollow the seasonal migration of the Intertropical Convergence Zone(ITCZ) so that motion of this belt can result in climate changes thatmimic monsoon strength changes (Armstrong and Allen, 2011). Moreoften the ITCZ and monsoon merge in Asia to act in concert with oneanother so that simple separation of these influences is not alwayspossible.

Alternatively, many paleoceanographic studies have focused onmonsoon-related marine upwelling forced by the winds as a measureof monsoon strength (Kroon et al., 1991; Prell et al., 1992; Clemens,1998; Chen et al., 2003). Unfortunately these proxies also do not alwaysclosely track rainfall in Asia. While studies of salinity near delta mouthscan be useful, such data sets are currently sparse (Steinke et al., 2010)and often of short duration (Kudrass et al., 2001). They may also beaffected by other processes, such as development of oceanic currentsystems.

An alternative approach to reconstructing the monsoon comes fromexamination of changing environmental conditions in continental Asia.Studies of changing flora (Quade et al., 1989; Sun andWang, 2005; Galyet al., 2008), fauna (Dettman et al., 2001; Nelson, 2005; Wang et al.,

2005), and weathering regimes (Derry and France-Lanord, 1996; Cliftet al., 2008) are closely linked to climate andmay be used to reconstructhow these have changed in the past.While their use asmonsoon proxiesonshore is impeded because of lack of incomplete and often unfossilif-erous, hard to date, sedimentary records the sediments accumulatingin deltas and on submarine fans hold out the prospect of a much morecomplete history of continental environmental evolution.

These repositories are also not without their problems because thesediment deposited on continental margins must necessarily lag theerosion/weathering processes onshore because of finite travel times(Goodbred and Kuehl, 1998; Chappell et al., 2006). Recently Hu et al.(2012) showed that changes in the alteration state of sediment depositedon the continental margin of the northern South China Sea largelyreflected the efficiency of reworking rather than a direct reaction toclimate change and that rapid sealevel change also affected the deep-water record. Nonetheless, a link was established between rainfall andthe chemistry and mineralogy of margin sediments.

Sediment transport times can potentially be long, depending on thedrainage system and have been estimated at N20 kyr for the Indus basin(Clift and Giosan, in press) for sands and longer for zircon grains. How-ever, for finer grained, lower density minerals such as clays the lagbetween erosion and deposition on the slope may be much less, as theyare carried rapidly in suspension, often bypassing sediment depocenterson the continental shelf and moving more directly into the deep water(Rego et al., 2010). In any case, this lagmay not be significant to climate-tectonic studies if the lag is b105 years,while tectonic forcing is typicallyover 106 years and longer timescales on the regional scale. Moreover,prior to the onset of Northern Hemispheric Glaciation the influence ofrapid sealevel change in buffering flux to the continental margin musthave been much reduced.

Here we synthesize the evidence for evolving physical erosion andchemical weathering in the region of the South China Sea since theOligocene in order to infer a tectonic-scale reconstruction of monsoonactivity in SE Asia. We focus on sediments recovered at Ocean DrillingProgram (ODP) Sites 1146 and 1148,which are located on the continen-tal slope of southern China in the northern South China Sea (Fig. 1). ODPSite 1144 is at 20°3.18′N, 117°25.14′E at a water depth of ~2037m. ODPSite 1148 is at 18°50.169′N, 116°33.939′E, at a water depth of ~3294 m,at the base of the continental slope. However, ODP Site 1144 is unusualin being located on a major sediment drift (Shipboard Scientific Party,2000a). Seismic and current-meter evidence shows that bottom cur-rents have been important in the recent geological past at 1000–2700m water depth along the northern margin of the South China Sea (Leiet al., 2007). Such currents are affected by history of the gateways tothe east of the drilling area with the Luzon Strait south of Taiwanbeing the gateway for flow from the Pacific Ocean into the South ChinaSea. The gateway is deep (N2000 m) and is not affected by glaciallydriven sealevel variations, unlike the Taiwan Strait, but the history ofbottom current flow is unknown. Close to the Chinese coast long-shore currents are directed towards the west but surface currents alsoflow towards the northeast during the summer monsoon (Hu et al.,2000). Whether these same currents affected the region in the past isquestionable because of the recent initiation of collision betweenLuzon and the Chinese margin, which must have constricted flowfrom the Pacific in the South China Sea.

This region is influenced by the East Asian summer monsoon, andhas been the target of several studies of monsoon intensification since

Page 3: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

Fig. 1. Bathymetric map of the northern South China Sea showing the location of ODP Sites 1144, 1146 and 1148 and their position relative to the delta of the Pearl River. Bathymetry isfrom GEBCO.

88 P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

the Oligocene (Clift et al., 2002; Chen et al., 2003; Jia et al., 2003; Wanget al., 2003c; Hess and Kuhnt, 2005; Wan et al., 2006; Wei et al., 2006;Wan et al., 2007; Clift et al., 2008; Steinke et al., 2010). Unfortunatelymany of these studies yield different conclusions aboutmonsoon history,probably because they each rely on different proxies. We attempt toreconcile these apparently contradictory existing records using newgrain-size and mass accumulation rate data, along with geochemicalSr isotope compositions from ODP Site 1148.

2. Monsoon Proxies

In examining the monsoon's development since 25 Ma we considerthe proxies applied to weathering and erosion studies over the lastglacial cycle (Fig. 2). This is an important starting point because thereis a general consensus concerning the intensity of the summermonsoonsince ~14 ka. Hu et al. (2012)measured a series of proxies spanning thetime since 14 ka at ODP Site 1144, located ~120 km NE of the sites wetarget here. The similar location in deepwater on the continentalmarginin the NE South China Sea suggests that proxies that are effective atresolving environmental change at ODP Site 1144 should also be appli-cable at ODP Sites 1146 and 1148. Since 14 ka it is possible to correlatechanging erosion and weathering intensity proxies with the changingmonsoon strength measured by speleothem records at Dongge cavein SW China, close to the study area (Dykoski et al., 2005). Thespeleothems may be considered robust at the first order level becausethey are in accord with other cave records in China (Yuan et al.,2004), in Nepal (Sinha et al., 2005)and even Oman (Fleitmann et al.,2003), as well as upwelling records for the Oman margin (Gupta et al.,2003), lake records in NW China (Herzschuh et al., 2005) and floral as-semblages in the Loess Plateau (Liu et al., 2005). Thus, although therehave been questions raised about the effects that winter temperatures,as well as summer monsoon rains, have on the isotopic composition ofstalagmites (Clemens et al., 2010)we believe that the general pattern ofsummer monsoon strengthening from 11 to 8 ka, followed by a declineuntil ~3 ka is real during the Holocene.

The sediments at ODP Site 1144 show that intensification of themonsoon in the Early Holocene resulted in a clear shift in a number ofproxies, most notably clastic mass accumulation rates (MAR), clay

mineralogy, hematite/goethite, Ti/Ca, K/Rb and the Sr isotope composi-tion of the clastic fraction (Fig. 2) (Hu et al., 2012). Such responses arepredicted because rates of chemical weathering generally increase withgreater humidity and higher temperatures (West et al., 2005), whichare usually associated with strong summer monsoons. Nonetheless,some of the proxy responses were not always what might have beenpredicted. For example, hematite/goethite increased during a periodof strengthening monsoon, when hematite is normally associatedwith drier conditions (Harris and Mix, 1999). This was interpreted toreflect enhanced reworking from the exposed continental shelf underthe influence of a strong summer monsoon (Hu et al., 2012) and con-trasts with the record from the southern South China Sea that showedlower hematite/goethite ratios during periods of stronger summermonsoon (Zhang et al., 2007). Despite this a clear response in terms ofgeochemistry andmineralogy is detected and is linked to themonsoon,so that these proxies are considered worth examining over tectonictimescales.While some proxies forweatheringmaynot respondquicklyenough to show a change over the short duration of the ODP Site 1144record these proxies are at least a good starting point for looking atthe longer development. Hu et al. (2012) noted that because of thesealevel rise the weathering and erosional response deposited at ODPSite 1144 did not last as long as the apparent monsoon maximum.Because sealevel variations have been especially intense since theonset of Northern Hemispheric Glaciation (Haq et al., 1987) this damp-ening effect is likely less important prior to ~3Ma. More recentwork onthe delta of the Pearl River also shows a response in the discharge fromthat drainage basin in terms of K/Al, K/Rb, and claymineralogy, demon-strating sedimentation of less altered material at times of weak mon-soon (Hu et al., 2013).

3. Geological setting

The sediments deposited on the Chinese continental margin at ODPSites 1146 and 1148 represent amixture of pelagic carbonate and clasticsediments and potentially record variations in the weathering anderosive environment of the landmass from which they were derived.Because the whole region around the South China Sea is affected by theEast Asian Monsoon, reconstructions of paleo-continental environment

Page 4: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

Age

(ka

)

Hematite/Goethite(565/435)

0.12

Ti/Ca

0.060.08

0.018 0.019 0.020

K/Rb

0.714 0.7100.712

87Sr/86Sr

0.8 0.4 0.2 00.6

Smectite/(Illite+Chlorite)

1.4 1.0 0.8 0.6 0.41.2

Dongge Cave δ18O (‰)

-10 -8 -7 -6-9

Mass AccumulationRate (g/cm2/ky)

40 20 060

Stronger monsoon

More alteration

More clastic fluxMore alteration

Fig. 2. Proxy records fromODP Site 1144 showing variations in erosion andweathering-related properties since 14 ka compared to the speleothemprecipitation record fromDongge Cave,SW China (Dykoski et al., 2005). Data are from Hu et al. (2012). Note the clear response in several of the proxies at 8–11 ka during the phase of monsoon intensification.

89P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

can be used to infer the timing of changes in the intensity of the summermonsoon. Moreover, drilling at this location extended to the Oligocene(Shipboard Scientific Party, 2000b), providing an unusually long andcontinuous record of continental conditions compared to the shorterrecords cored on the Bengal Fan and at the Oman margin by ODP,extending to 18 Ma and 14 Ma respectively (Shipboard Scientific Party,1989a; Shipboard Scientific Party, 1989b).

Although there has been debate about the provenance of sedimenton the South China northern margin (Tamburini et al., 2003; Liu et al.,2007b), there is a consensus that the main source of clastic sedimentis from the mouth of the nearby Pearl River and neighboring drainagesthroughout most of the Neogene (Clift et al., 2002; Li et al., 2003). Since~4–5 Ma the rapid uplift and exhumation of Taiwan has resulted in achange in the dominant provenance in the region towards this island(Liu et al., 2010; Wan et al., 2010b; Hu et al., 2012). This change is inde-pendent of the climatic evolution and complicates the interpretation ofthe weathering record since that time.

Mineralogical and geochemical changes after 5 Ma may reflectTaiwanese orogenesis and not climate change alone. Prior to that timetectonic processes are not thought to have had much effect on erosionand weathering in the Pearl River basin because the Pearl River doesnot erode the flanks of the Tibetan Plateau and the passive margin itselfhas largely been in a state of long-term thermal subsidence (Ru andPigott, 1986; Clift and Lin, 2001). The Pearl River contrasts with theneighboring river systems, such as the Mekong or the Yangtze, whereheadwater capture may have been important (Brookfield, 1998; Clarket al., 2004; Clift et al., 2006), although recent finding suggest thatmajor drainage reorganization may have been complete by ~22 Ma in

the Early Miocene (Zheng et al., 2013). Instead the clastic materialshed from the Pearl River and other rivers draining southern Chinacan be used as a proxy for the state of chemical weathering in thisregion, at least before 5 Ma.

4. Links between monsoon and erosion

The relationships betweenmonsoon climate and erosional responsehave been a subject of significant debate. Studies of the monsoon sincethe Last Glacial Maximum (LGM; ~20 ka) show that large volumes ofsediment, formed by enhanced physical erosion in the Himalaya, weredelivered to South Asian deltas during the Early Holocene, at a timewhen the summer monsoon strongly intensified (Goodbred andKuehl, 2000; Giosan et al., 2012). Stronger physical erosion linked to astronger summer monsoon is consistent with what is known aboutthe erosive power of seasonal rainfall (Molnar, 2001) and to linksbetween large-scale mass wasting and monsoon intensity (Bookhagenet al., 2005). Whether this linkage is still true over periods N107 yearsis presently unclear.

Initial sediment budgets from across monsoonal Asia emphasizedrapid clastic flux in the Plio-Pleistocene (Métivier et al., 1999), buthave been modified by later seismically-based studies (Clift, 2006;Hoang et al., 2010) that also highlight the Early and Middle Miocene asbeing periods of rapid clastic sedimentation on the Chinese continentalmargin. Clift et al. (2008) linked Middle Miocene intensified erosion tostrong summer monsoons, but the evidence supporting this hypothesiswas not as robust as it might be. This present study tests this hypothesisby better quantifying changing continental weathering regimes since

Page 5: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

90 P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

the Oligocene.We emphasize the important difference between erosionand chemical weathering. It is well understood that these two processesare linked and that alteration rates increase as erosion also increases(Rasmusson and Carpenter, 1982; Allan et al., 1996) because the processof breaking bedrock into sediment particles provides more area foralteration reactions to proceed. It has been argued that erosion con-trolled by tectonically driven rock uplift may dominate over climate asa control of chemical weathering (Riebe et al., 2001), although in theless tectonically active Pearl River basin this is not a key factor. It isalso noteworthy thatwhile chemical weathering increases with erosionthere is a subsequent reduction in weathering as erosion reaches amaximum and soil thicknesses, where alteration is largely achievedtends to zero (Ferrier and Kirchner, 2008).

5. Earlier monsoon reconstructions

While it has commonly been accepted that the summer monsoonintensified in South and East Asia around 8 Ma (Kroon et al., 1991;Prell et al., 1992; Chen et al., 2003; Wan et al., 2007), this view hasbeen called into question by data that points to an earlier, Middle orEarly Miocene intensification. Clift et al. (2002) used clay mineralogyfromODP Site 1148 to suggest that initial summermonsoon intensifica-tion occurred around 15 Ma, based on the generally accepted under-standing that clay mineralogy in marine sediment is controlled by thestate of chemical weathering in the flood plains of the river supplyingthe basin (Thiry, 2000). In an alternative approach Jia et al. (2003)employed organic chemistry to argue for a strong summer monsoonstarting in the Early Miocene, which is consistent with facies and pollen

More physicalerosion

0.1 0.2 0.3 0.40

Ti/Ca scannedODP Site 1148CRAT

0.80.40

Age

(M

a)

A B C D0

10

5

15

25

20

Clastic MAR(g/cm2/k.y.)

Ple

ist.

Plio

cene

U. M

ioce

neM

. Mio

cene

L. M

ioce

ne

1 3 5 0 105Chlorit

ODP Sit

Start of Taiwan Orogen

Less chloritemore alteration

Fig. 3. Temporal evolution in proxies linked to physical erosion and flux of clasticmaterial to therates as calculated from the carbonate coulometry work presented in Supplementary Table 1(D) Proportion of chlorite in the clay mineral assemblage at ODP Site 1146 (Wan et al., 2007(F) Sedimentation rates on the southern Chinese margin reconstructed from seismic data (Clifttimes of intensified physical erosion.

data from continental China that points to a major change in regionalclimate around the Miocene–Oligocene boundary (Sun and Wang,2005). In another study Wan et al. (2007) applied a series of grain-sizeand clay mineral proxies to sediments recovered at ODP Site 1146 toargue that it was the winter monsoon that strengthened at 15 Ma andat 8 Ma, and with the summer monsoon mostly intensifying around3Ma. In contrast,Wei et al. (2006) andWanet al. (2010a) used chemicalweathering data to argue for a general weakening of alteration and thusof the summer monsoon since the Early Miocene.

In contrast, Clift et al. (2008) used a new method of spectral coloranalysis to propose a peak summer monsoon intensity in the MiddleMiocene, followed by a weakening, especially around 8–10 Ma. Theirnew CRAT proxy (Fig. 3A) was designed to provide a mineralogicalratio of chlorite/(chlorite + hematite + goethite), exploiting the factthat chlorite is the product of physical erosion, while hematite andgoethite reflect chemical weathering processes. This proxy predictedweakening, not strengthening of the summer monsoon at ~8 Ma, butthis is consistent with salinity records from the northern South ChinaSea pointing to reduced run-off starting at that time (Steinke et al.,2010), and explained why clastic sedimentation rates fall across Asiain the Late Miocene (Clift, 2006). As noted above, on millennial time-scales sediment delivery to the sea coincides with periods of wet sum-mer monsoons (Goodbred and Kuehl, 2000).

The East Asian Summer Monsoon now appears to have a muchlonger and potentially multi-staged history compared to what wasunderstood only ten years ago. What has become clear is that some ofthe complexities and apparent contradictions in competing studiesmay be related to the fact that people have mixed up summer and

E F

MMCO

0 5 10 15 20 25Sedimentation rate(1000 x km3/m.y.)

2015e (%)e 1146

0.40 0.300.35

Al/Si scannedODP Site 1148

G

More sand

-1000100200

Relative sealevel(m)

ocean (A) the spectral-based CRAT proxy of Clift et al. (2008), (B) clasticmass accumulation, (C) Ti/Ca at ODP Site 1148 derived from XRF whole core scanning (Hoang et al., 2010),), (E) Al/Si at ODP Site 1148 derived from XRF whole core scanning (Hoang et al., 2010),, 2006), (G) Global eustatic sealevel fromHaq et al. (1987). Horizontal gray bands indicate

Page 6: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

Table 1Sr isotope compositions of selected samples from ODP Hole 1148A.

Leg Site Hole Core T Section Top (cm) Age (Ma) 87Sr/87Sr

184 1148 A 5 H 5 76 0.56 0.717270184 1147 A 8 H 3 45 0.9 0.719374184 1148 A 13 H 2 15 1.95 0.719874184 1148 A 17 X CC 2 3.33 0.714775184 1148 A 19 X 4 92 4.32 0.716593184 1148 A 20 X 7 16 5.21 0.716299184 1148 A 21 X 6 72 5.87 0.716919184 1148 A 22 X 4 64 6.39 0.718453184 1148 A 24 X 3 45 7.5 0.721174184 1148 A 25 X 6 24 8.11 0.717257184 1148 A 27 X 6 16 8.77 0.718006184 1148 A 28 X 6 35 10.2 0.719324184 1148 A 29 X 4 115 11.07 0.718381184 1148 A 31 X 5 104 12.02 0.718701184 1148 A 32 X 7 16 13.51 0.719722184 1148 A 33 X 6 116 15.3 0.719334184 1148 A 35 X 5 105 16.5 0.717087184 1148 A 36 X 6 144 16.37 0.714140184 1148 A 37 X 6 25 16.76 0.715759184 1148 A 38 X 5 55 17.06 0.719824184 1148 A 39 X 7 24 18.25 0.718001184 1148 A 40 X 6 64 19.08 0.716720184 1148 A 43 X 6 104 21.37 0.715259184 1148 A 44 X 7 35 22 0.715545184 1148 A 45 X 6 35 22.97 0.712219184 1148 A 47 X 6 145 25.02 0.713819

91P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

winter monsoon proxies and that each proposed proxy is measuring adifferent aspect of the monsoon. Few of these proxies offer a uniquesignature of rainfall onshore.

6. Methodology

As well as synthesizing a number of existing data sets from ODP Sites1146 and1148wepresent here newdata that help us to resolve changingerosion andweathering regimes. In this studywe constrain variations inclastic sediment flux from the Pearl River into the South China Sea byreconstructing the clastic mass accumulation rate (MAR) and grain-size of sediment recovered at ODP Site 1148 (Fig. 1). We test thehypothesis that stronger summer monsoon rains might cause fasterand stronger run-off and increase both the rate at which sediment isdelivered to the continental margin and the grain-size of that material,assuming that stronger fluvial flux would transport larger sedimentparticles further offshore. Earlier, high-resolution work on grain-sizevariations at ODP Site 1146 (Fig. 1) showed that during the EarlyPleistocene grain size variations correlatedwithmonsoon rainfall inten-sity, specifically that the finest fluvial-derived fraction was increasedrelative to medium grained material when the summer monsoon wasstrong (Boulay et al., 2007). This relationship was however only provensince 1.8 Ma, a time period when monsoon strength was closely corre-lated to sealevel variations. Over longer time scales data from ODP Site1146 shows a steadily decreasing grain-size from20 to ~16Ma followedby a long-term increase from 16 to 9 Ma, which was interpreted asreflecting monsoon intensity (Wan et al., 2007).

Grain-size is a potentially complicated proxy because density, aswell as size, also has an impact on the flux rates of sediment grainsthrough a sedimentary system. Large light particles could travel as fastas smaller heavier particles. Nonetheless, in a large continental riverbasin such as southern China clays, quartz and feldspar grains dominatethe sediment. Provenance is believed to have been largely stablebetween 24Ma and5Ma (Li et al., 2003), after the endofmajor drainagereorganization in eastern Asia (Zheng et al., 2013), so that coherentchanges in grain size in the deep-water, offshore sediments cannot belinked to a simple change to lighter or denser particles, because thesediments are always dominated by the same mineral groups. We canalso test this assumption through use of geochemistry to check forchanges in composition and MAR. Grain size is also potentially compli-cated by variations in sea level, which bring the river mouth closer orfurther away from the drill site as sea level goes down and up. Consid-ered only by itself grain size is therefore a relatively unreliable proxyfor the intensity of physical erosion and sediment delivery rates.

Analyses were performed on sediments taken from ODP Site 1148.The sediments are all mixtures of finer grained clastic material, a mix-ture of clay and silt, together with pelagic nannofossil ooze (ShipboardScientific Party, 2000b). The proportion of carbonate varies in the coreup to a maximum of 67%, but with a median of 32%. Samples wereprepared for grain-size analysis by treatingwith acetic acid (CH3COOH)to remove the carbonate sediment fraction then rinsing with water.The sediment was then treated with a Calgon™ solution to preventflocculation of clays. The samples were analyzed in a laser-basedBeckman–Coulter LS13320 laser particle size analyzer at the WoodsHole Oceanographic Institution, USA. Results of this analysis are shownin Supplementary Table 1.

Samples for carbonate analysiswere processed at theMarineBiologyLaboratory,Woods Hole, USA, using a CO2 coulometer (Model 5011, Sys-tem140with AcidificationModule CM5130),made byUIC Coulometrics.Approximately 10–12 mg of dried, ground sediment was reacted with 2M HCl to liberate CO2, which was then measured to determine the car-bonate content. The results of this analysis are shown in SupplementaryTable 2. The amount of carbonate is expressed as weight percent, and nocorrection was made for the presence of other carbonate minerals sinceshipboard microscopic investigation had demonstrated that these werenot volumetrically significant (Shipboard Scientific Party, 2000b).

Linear rates of sedimentation were calculated using the nannofossil-based age model of Su et al. (2006). More importantly a carbonate MARcould be determined from the percentage of carbonate in each sediment,together with the sediment dry densities measured during ODP Leg 184(Shipboard Scientific Party, 2000b). We assume that the total MARreflects a combination of carbonate and clastic components, becausethere is negligible siliceous sedimentation in this region (Clemenset al., 2008). Consequently a clastic MAR can be readily determinedonce the carbonate MAR has been subtracted from the total.

Decarbonated sampleswere also used for Sr isotope analysis. Sampleswere accurately weighed into Teflon screw-top beakers and dissolvedusing HF-HClO4. Sr was separated in 2.5 N HCl using Bio-Rad AG50WX8 200–400 mesh cation exchange resin and standard columnmethods.Sr samples were analyzed on a Finnigan “Neptune” multi-collectorinductively coupled plasma mass spectrometer (MC-ICP-MS) at WoodsHole Oceanographic Institution, USA. Sample measurements were nor-malized to 86Sr/88Sr = 0.1194 and referenced to a value of 0.710240for NBS987 standard. Results are provided in Table 1.

7. Results

The calculated clastic MAR since 25 Ma is shown graphically inFig. 3B. There is a clear peak inMAR~1.5Ma after a sharp increase begin-ning after ~4 Ma. There are also well defined maxima at 15–17 Ma,18.5–19.2 Ma and 21.3–23.5 Ma. Clastic MAR increased modestly from13 to 11.5 Ma then remained at a moderate level before decliningagain after 8.5 Ma.

Grain-size analyses were only run for the 15–20Ma period (Fig. 4A),largely because the Quaternary data from Hu et al. (2012) shows noclear relationship between grain-size and monsoon strength andbecause grain-size can be affected by many variables. This result con-trasts with the coherently decreasing grain-size trend recorded at ODPSite 1146 from 20 to 16 Ma (Wan et al., 2007). We examined grain-size variation over a period in which monsoon intensity was expectedto show significant variation. Our record shows a dominant silt meansize, which slightly decreases from 20 to 15 Ma. There is a peak inmean grain-size and in the grain-size of the biggest 10% of the popula-tion at 15.8–16.5 Ma, and at 19.1–19.8 Ma. Samples with mean grain-sizes N10 μm comprise 41% of the samples analyzed at 15.8–16.5 Ma

Page 7: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

15

16

17

18

19

20

0 5 10 15

Age

(M

a)

Ti/Ca0.40.2

K/AlODP Site 1148

Clastic MAR(g/cm2/k.y.)

1 2 3 4

Meangrainsize (µm)

More chemicalweathering

A B C D E F G

Al/SiODP Site 1148

Sortable silt

Fine silt

Clay

1.5 2.0δ18O (‰)

Hotter Colder

CRAT

1.00.5

More clasticflux

clay-rich sand-rich

H I767778798081

CIA ODP Site 1146

K/AlODP Site 1146

0.17 0.19 0.21 0.23

3.0 3.4 3.8

0.280.300.32

CIA

Periods of stronger physical erosion

More physicalerosion

Fig. 4. Temporal evolution in weathering and erosion proxies at ODP Site 1148 over the time period 15–20Ma. (A) mean grain-size, (B) clastic MAR, compared to (C) Ti/Ca and (D) Al/Siand (E) K/Al running average of 40 analyses measured by XRF scanner (Hoang et al., 2010), (F) K/Al and (G) CIA from ODP Site 1146 measured by conventional XRF (Wan et al., 2007),(H) the CRAT weathering proxy of Clift et al. (2008), and (I) the δ18O global compilation of Zachos et al. (2001). Horizontal gray band indicates the Middle Miocene Climatic Optimum(MMCO).

92 P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

and 31% of the samples at 19.1–19.8 Ma. In contrast, outside theseperiods samples with coarse silt grains are almost completely absent.The 15.8–16.5 Ma large grain-size sediments correspond to the middleof the period of high clastic MAR. However, the 19.1–19.8 Ma coarsesediments pre-date the 18.5–19.2 Ma period of elevated clastic MAR.

Sr isotopes are seen to vary coherently with time (Fig. 5H). 87Sr/86Srwas low ~24 Ma, then rose to high values by around 18 Ma, remainedhigh until 10 Ma, except for a brief but dramatic reduction in 87Sr/86Srvalues between 16.8 and 15.2 Ma. 87Sr/86Sr values decreased afterreaching a maximum at 7.5 Ma to a Pliocene minimum around 3.3 Ma.Since that time 87Sr/86Sr values have shown a moderate increase to~0.72, similar to both the values found in Taiwanese rivers (Chen andLee, 1990), the Pearl River delta during the Holocene (Hu et al., 2013),in ODP Site 1144 sediments (Hu et al., 2012), but somewhat belowthe range seen in the modern Pearl River (Liu et al., 2007b)(Fig. 6A),although this latter high range likely reflects enhanced reworking ofsoil since ~3 ka cause by anthropogenic influences.

8. Discussion

8.1. Physical erosion proxies

Possible causes for the changes in clastic MAR and grain-size at ODPSite 1148 can be determined by comparing the reconstructed trendswith existing records of climate and environmental conditions. Grain-size can be affected bymany factors, including power of the river trans-port system, the proximity of the drill site to the source rivermouth andthe speed of reworking bottom currents, so that by themselves they arenotmeaningful. Seismic and current-meter evidence shows that bottomcurrents have been important in the recent geological past at 1000–2700 m water depth along the northern margin of the South ChinaSea (Lei et al., 2007). Such currents may have been significant in thedeeper past as well, depending on the opening and closure of gateways

towards the east of the study area. This is despite the fact that the seismicstratigraphy at ODP Site 1148 does not suggest contourite sedimentationat any point in the sampled history (cf., ODP Site 1144) (ShipboardScientific Party, 2000b). Because of the various processes that mayhave affected grain-size variations we do not believe that grain-sizedata provide a robust proxy for erosion intensity and are a poormonsoonproxy at least at ODP Site 1148.

Fig. 4 allows us to compare MAR with other proxies for erosionintensity and to ocean temperatures, as reconstructed by the globaloxygen isotope compilation of Zachos et al. (2001)(Fig. 4I). The increasedclasticMAR (and coarser grain size) at 15.5–17.2Ma correlateswell withthewarmer global temperatures of theMMCO, consistent withwhat hasbeen shown at ODP Site 1146 and the hypothesis of strong continentalerosion during this warm, wet climatic phase (Wan et al., 2009).

In order to achieve high temporal resolution we use XRF scanningdata fromODP Site 1148 fromHoang et al. (2010). Use of XRF core scan-ning data as a tool in paleoceanographic and stratigraphic studies is welldefined and widely accepted (Röhl and Abrams, 2000; Tjallingii et al.,2007; Clemens et al., 2008), but it has rarely been used to reconstructerosion or weathering histories.

Ti/Ca can be used as a proxy for clastic flux, with Ti being dominatedby presence of various heavy minerals, including ilmenite, perovskite,rutile and titanite, while Ca contents in scanned XRF data are dominatedby biogenic carbonate which may be considered a background pelagicsignal. Thus high values in Ti/Ca generally correlate with periods ofintensified clastic delivery from the continent. Maxima are seen after4 Ma, and at 10–11 Ma, 13.2–17 Ma, 18–19 Ma and 20.0–20.8 Ma(Fig. 3C). The high peaks between 15 and 16.5 Ma correlate well withthe MMCO.

One alternative approach to looking at erosional strength involvesAl/Si which can be effectively used as a proxy for the ratio of clay (richin Al) and quartz sand (rich in Si). Low Al/Si values during the MMCOsuggest the presence of greater than normal proportions of sand and

Page 8: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

0.710

CRAT

0.80.40

Age

(M

a)

A B C0

10

5

15

25

20

80 75 70

CIA ODP Site 1148

More chemicalweathering

E G

MMCO

Ple

ist.

Plio

cene

U. M

ioce

neM

. Mio

cene

L. M

ioce

ne0.0150.005

K/RbODP Site 1148

0

5

10

15

20

25

Hematite/goethite

4 2 -20

H I J

87Sr/86Sr

0.725

-11 -10 -9

εNd-14 -13 -12

Clift et al.

Li et al.

0.720 0.715

PearlHolocene

0.010

Taiwanrivers

Start of Taiwan Orogen

F0.8 1.0 1.2

K/Al

D

0.20.40.6

Kaolinite/illite

Pearl RiverHolocene

04812Smectite/kaolinite

More chemicalweathering

More seasonality

D

Fig. 5. Temporal evolution in proxies linked to chemical weathering intensity since 25 Ma. (A) The CRAT proxy of Clift et al. (2008) from ODP Site 1148, (B) hematite/goethite calculatedfrom color spectral data from ODP Site 1148, (C) kaolinite/illite and (D) smectite/kaolinite of the total clay assemblage from ODP Site 1146 (Wan et al., 2007), (E) K/Al, (F) Chemical Indexof Alteration (CIA), (G) K/Rb from ODP Site 1148 from data of Li et al. (2003). (H) 87Sr/86Sr values at ODP Site 1148 and (I) εNd values from Li et al. (2003) and (J) from Clift et al. (2002).Green shading shows the measured range in Nd isotopes in the Pearl River Delta since 14 ka (Hu et al., 2013). Horizontal gray band indicates the Middle Miocene Climatic Optimum(MMCO).

93P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

thus a higher energy transport regime, during a period of intensifiedrun-off. This is also seen during the earlier 18.3–19.0 Ma period whenTi/Ca values were also high (Fig. 4).

A longer history of Al/Si variations is shown in Fig. 3E which shows agradual decrease in values since ~8 Ma, and especially in the last 4 Ma,indicating more sandy sediment, a trend that is consistent with thehigherMAR and high Ti/Ca. There are dramatic high Al/Si values around4 and at 8 Ma, as well as a less extreme but longer duration episodeat 12–13 Ma. The peaks in MAR at 15.5–17.2, 18.5–19.2 and 21.3–23.5 Ma correspond to resolvable but small minima in Al/Si, as mightbe expected if faster erosion resulted in more and coarser grainedsediment flux. These periods are interpreted as times of faster physicalerosion in the source regions.

8.2. Geochemical proxies for alteration

Chemical weathering and alteration changes the bulk sedimentchemistry and clay mineralogy. The terrigenous sediment geochemicaldata of Wei et al. (2006), which was also collected from ODP Site 1148allows chemical weathering indicators to be calculated, in particularthe “Chemical Index of Alteration” (CIA)(Nesbitt and Young, 1982)(Figs. 4G and 5D), which has been a well used chemical weatheringproxy for many years. Li and Yang (2010) noted that while CIA can bea good indicator of alteration state that it necessarily represents anintegrated picture of chemical weathering in a drainage basin, ratherthan the instantaneous state of chemical weathering rates at the timeof sedimentation. We also show the ratios K/Al and K/Rb for the Weiet al. (2006) data (Figs. 4F, 5E and F). Like CIA these ratios have proxypotential based on the fact that alkali elements, including K, are relativelymobile in water, while Al and Rb are less mobile during the breakdownof minerals.

The geochemistry of K is complicated because it is enriched undermoderate degrees of chemical weathering, but then becomes depletedunder more intense weathering as K-feldspars breakdown (Blaxland,1974; Nesbitt et al., 1997). K/Rb, K/Al and CIA show very similar long-term patterns in decreasing chemical weathering since 25 Ma (Fig. 5).CIA shows a deflection to slightly higher values (more altered) duringtheMMCO, with all three indicators showing a steady decrease in alter-ation after that time. Fig. 4F and G show the co-variation in CIA and K/Alin greater detail between 20 and 15 Ma. These plots show very similarforms and a clear deflection to intensified chemical weathering inboth proxies from 17.2 to 15.5 Ma, spanning the MMCO. K/Al showssimilar patterns in both scanned and bulk sedimentXRFdata, emphasiz-ing the rise in K/Al after 15.5 Ma. the two data sets which might beexpected to agree closely seem to diverge in sediments older than~18.5Ma but the number of data points from the bulk sediment analysisis rather low at that point.

The intensity of chemical weathering appears to track the δ18Orecord of Zachos et al. (2001)(Fig. 4I). This is true not only at theMMCO, but to a lesser extent between 18 and 19 Ma. We concludethat chemical weathering falls from 25 Ma to the present, with a peakduring the MMCO. Although Wei et al. (2006) suggested that thelong-term trend reflected a decreasing summer monsoon we find thatinterpretation to be inconsistent with the record of physical erosionand mass flux to the ocean, which is also expected to be positivelycorrelated to summer monsoon intensity. Consequently we interpretthe chemicalweathering trend to bemore closely controlled by temper-atures, as shown by the δ18O record, which mirrors CIA in particular at15–20 Ma (Fig. 4I) and more generally over the entire 25 Ma of study.As the region has become colder after the MMCO the degree of alter-ation in the sediments deposited on the continental margin has gener-ally reduced.

Page 9: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

0.70

0.71

0.72

0.73

0.74

-15 -10 -5 0 5

Pearl River

Luzon

Taiwan

ODP Site 1144

A

87S

r/86

Sr

Pearl River - Holocene

ODP Site 1148

0.708

0.712

0.716

0.720

0.724

-12 -11 -10 -9 -8

ODP Site 1148Lower Miocene

ODP Site 1148Mid Miocene

+ Plioc-Recent

ODP Site 1148~7 Ma

B

87S

r/86

Sr

εNd

Incr

ease

d ch

emic

al w

eath

erin

g

Less continental source

Pearl River - HoloceneODP Site 1144

Modern Taiwan riversModern Luzon rivers

Fig. 6. Sr and Nd isotopic plot showing the variability in Holocene sediments from ODPSites 1144 (Hu et al., 2012)and in Lower Miocene to Recent sediments at ODP Site 1148compared tomodern potential sources around the South China Sea. (A) shows the generalsimilarity of the sedimentwithmodern Taiwanese rivers and bed rock samples (Chen andLee, 1990; Lan et al., 2002) and the differences with sediments in the modern Pearl River(Liu et al., 2007b) and with potential volcanic sources in the Philippine island of Luzon(Zhou et al., 2002). The Neogene sediments at ODP Site 1148 show overlap with bothTaiwanese rivers and the Holocene sediments of the Pearl River estuary (Hu et al., 2013).

94 P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

8.3. Clay minerals as alteration proxies

The clay mineral record at ODP Site 1148 is low resolution and ismostly characterized by a sharp reduction in smectite contents after15.5 Ma (Clift et al., 2002). Fortunately, a much more detailed record isavailable at ODP Site 1146 (Wan et al., 2007). Clay minerals and alsoany affected by climate but also by provenance in the South China Sea.Liu et al. (2007b) demonstrated that Luzon rivers are rich in smectite,while kaolinite is presently abundant in the Pearl River (Liu et al.,2007b). However, Hu et al. (2013) nowshow that the state of themodernPearl River is not representative of the recent geological past and that themodern abundance kaolinite is largely a function of re-working driven bythe establishment of agriculture.

Both smectite and kaolinite are formed by chemical alteration. Soilforms rapidly and to greater depths in tropical and subtropical environ-ments, where chemical weathering by leaching produces kaolin groupminerals (Thiry, 2000), whereas warm, dry regions with less leachingdominantly produce smectite-rich soil. This ratio allows the relativestrength of tropical to more monsoonal, seasonal weathering to beassessed. Furthermore, the kaolinite/illiite ratio in marine sediments

constitutes a reliable indicator of chemical hydrolysis versus physicalprocesses in continental weathering profiles (Chamley, 1989), whichis effectively a proxy for humidity (Thamban and Rao, 2005).

Fig. 5C shows that kaolinite/illite values increased from 20 to 17 Ma(MMCO) then remained variable but high until 8 Ma, after which theyhave declined to the present day. The overall trend is towards weakerchemical weathering, less chemical hydrolysis, even before the start ofsediment flux from Taiwan.

Chlorite may be a useful proxy for physical erosion in the sourcebecause this mineral is derived from erosion of low and medium tem-perature metamorphic rocks (Hurlbut and Klein, 1985). Indeed, claysin Taiwanese rivers comprise on average 41% chlorite (Liu et al., 2008),while those from the more modern chemically weathered Pearl Riverreach only 25–30% chlorite (Liu et al., 2008). Fig. 3D shows the develop-ment of chlorite as a proportion of the total clay mineral population atODP Site 1146 (Wan et al., 2007). The plot shows a gradual increasefrom 18 Ma, with a step increase after the MMCO at 15 Ma and anotherjump in values after 3 Ma, close to the time of initiation of NorthernHemispheric Glaciation. This history would argue for more physicalerosion as the Cenozoic progressed and especially after 3 Ma. Whetherthe younger transition is related to faster erosion following the onsetof North Hemispheric Glaciation (Zhang et al., 2001) or whether this isa response to the exhumation of Taiwan is unclear.

8.4. Exploring the CRAT proxy

We note that the evolution in CRAT, which is proposed to trackchlorite/(chlorite + hematite + goethite) (Clift et al., 2008), does notfollow chlorite abundance very closely (Fig. 3A and D), except for a com-mon decline between 13 and 3Ma. Poor correlation between CRAT valuesand chlorite is especially marked prior to 13 Ma. CRAT values fall sharplyfrom 15.5 to 13 Ma, while chlorite is rising. Assuming that there is notan additional unaccounted mineral phase interfering with the CRATvalue calculation this proxy appears to be strongly linked to the relativeabundance of hematite versus chlorite, with the influence of goethiteharder to isolate. High CRAT values generally correspond to periods ofrelatively strong physical erosion compared to chemical weathering.Although both proxies track towards higher values from 13 to 7 Maand to lower after 4 Ma the major excursion to higher CRAT values atthe MMCO (15.5–17.2 Ma) is not observed in the chlorite abundance.If this represents a reduction in hematite and goethite then the lowhematite/goethite ratios at that time (Fig. 5B) would require that thisfall is largely driven by reduced hematite, implying high humidity duringthe MMCO.

Comparison with the other erosion-linked proxies in Fig. 3 showsthat CRAT values show an association with peaks in clastic MAR, atleast prior to 5 Ma, consistent with the proposal that this records thedominance of physical erosion over chemical weathering. CRAT valuesand clastic MAR are high at 21.5–23.8, at 18.8–20.0, as well as duringthe MMCO (15.2–17.0 Ma).

In order to better understand what the CRAT proxy is measuring wecompared itwith commonly acceptedmeasures of chemicalweathering(Fig. 5). Fig. 5B shows variations in hematite/goethite as measured bythe relative strength of 565 and 435 nmwavelengths in the color spectraof the fresh cut core (Farquhar et al., 1989). Because hematite is prefer-entially formed in drier andwarmer conditions,while goethite is favoredby wetter and cooler climates (Schwertmann, 1971; Schwertmann,1988), the color spectral data can provide a first order measure ofweathering conditions in the drainage basin of the Pearl River.

Earlier studies linked higher hematite/goethite ratios to drier condi-tions (Harris andMix, 1999; Ji et al., 2004; Zhang et al., 2007) andwouldbe expected to be linked to a weaker summer monsoon with less rain.However, this ignores the role of seasonality compared to tropicalweathering. In a seasonal monsoonal climate the environment experi-ences a dry winter season as well as a wet summer one, so a monsoonclimate may appear drier than a tropical one in the sedimentary record.

Page 10: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

KaoliniteKaolinite(%)(%)

SmectiteSmectite(%)(%)

Illite + Chlorite (%)Illite + Chlorite (%)

Modern Pearl RiverModern Pearl River

Taiwan riversTaiwan rivers

Taiwan RiversTaiwan Rivers

Pearl RiverPearl Riveroffshoreoffshore

LuzonLuzonRiversRivers

ODP Site 1146ODP Site 1146

Luzon riversLuzon rivers

Pearl River offshorePearl River offshore

Pearl River HolocenePearl River Holocene

Fig. 7. Plot showing how clays at ODP Site 1146 largely overlap with the Holocenesediments of the Pearl River delta (Hu et al., 2013), but range fully between Luzon andTaiwan (Liu et al., 2010). They do not show much similarity with the modern PearlRiver as measured by Liu et al. (2007b).

95P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

The hematite/goethite proxy bears close similarity with the CRAT

history at the first order level (Fig. 5). The most biggest discrepanciesbetween the two proxies are the sharp rise in CRAT values between 5.5and 4.0Ma, compared to amuchmoremodest fall in hematite/goethite,and the gradual decrease in CRAT values from 4Ma to the present, whilehematite/geothite is more stable. Low CRAT values at 10–15 Ma wouldindicate relatively strong chemical weathering, which is consistentwith K/Rb, CIA, and clay mineralogy at that time. However, this is atime when hematite/goethite is high, which would normally be linkedto drier conditions that might be expected to cause less chemicalweathering. Although Hu et al. (2012) showed higher hematite/goethite values during a period of stronger summer monsoon in theEarly Holocene these workers interpreted that trend in terms ofreworking of previously weathered material rather than as a directimpact of monsoon weathering. It is possible that enhanced reworkingis also causing the high hematite at 10–15 Ma. Whether that could besustained over periods of ~5 Myr, rather than the millennial scale iden-tified by Hu et al. (2012)when climatewas less variable than during theQuaternary is another question, although our data would be consistentwith that model.

Hematite may be forming as a result of the annual dessication of thePearl River basin during dry winter monsoons. While summer rainsdrive erosion and weathering the dry season also affects the develop-ment of soil mineralogy. Although inmost settings hematite productionis linked to dry conditions these have usually been regions of moreequitable climate, not the strong seasonality of East Asia that is a keyfeature of the monsoon. A proxy that traced seasonality rather thanaverage humidity would be a good monsoon indicator. On the evidencepresented here and byHuet al. (2012) showing commonhighhematite/goethite and strong chemical weathering at the same time we concludethat hematite/goethite may be such a proxy in the northern SouthChina Sea. The close relationship between CRAT values and hematite/goethite indicates that CRAT is also closely related to summer monsoonintensity.

8.5. Sr isotope evolution

Sr isotopes have also previously been used to reconstruct chemicalweathering intensities. Derry and France-Lanord (1996) argued thatwhen provenance was stable 87Sr/86Sr values rose when chemicalweathering became more intense. Such an interpretation is also sup-ported by the observation that the modern Pearl River has both higher87Sr/86Sr values and greater degrees of chemical alteration as shownby CIA and K/Al values compared to the Holocene River (Hu et al.,2013). ODP Site 1148 appears to be a good place to use this isotopesystem because Li et al. (2003) indicate that Nd isotope ratios, whichare a common provenance proxy, have been largely stable since around24 Ma (Fig. 5I). There is however some dispute about the Nd isotopecomposition of sediment at ODP Site 1148 because the εNd valuesreported by Li et al. (2003) were consistently ~2 points more negativethan those reported by Clift et al. (2002)(Fig. 5J). Comparing the twohistories with new Nd data from the Pearl River Delta (Hu et al., 2013)and from ODP Site 1144 (Hu et al., 2012) shows that the Clift et al.(2002) data, though more sparse, have a closer consistency with theHolocene values. Consequently we use these data to assess the prove-nance control over the Sr reported here.

Ideally we would show a cross plot of Sr versus Nd isotope ratios todetermine any linkage. Unfortunately only five samples have beenanalyzed for both isotopes, making a definitive answer hard to achieve,although any correlation appears to be weak for these samples. FallingεNd values from 10 to 7 Ma in both isotopic systems matches rising87Sr/86Sr values,which is the reverse trend fromwhatmight be expectedin a provenance controlled system.We follow the conclusions of Huet al.(2012) in suggesting that weathering is the primary control on 87Sr/86Srvalues. We further note that there is reasonable overall correlation ofCRAT and hematite/goethite with the 87Sr/86Sr evolution (Fig. 5),

especially in terms of maximum weathering from 18 to 10 Ma and astrong fall at the MMCO when no response is seen in Nd isotopes(Fig. 5I, J), supporting the role of alteration in governing 87Sr/86Srevolution.

8.6. Sediment provenance at ODP Sites 1146 and 1148

The combined Sr and Nd allow a simple test of the provenance of thesediments at ODP Site 1148 to be made. Fig. 6A shows the field intowhich all analyses from this site fall, compared to data from themodernPearl River, from Holocene sediments from the Pearl River estuary (Huet al., 2013), from Taiwanese rivers and sedimentary rocks in Taiwan(Chen and Lee, 1990; Lan et al., 2002), as well as from volcanic rocksfrom Luzon (Zhou et al., 2002). The ODP Site 1148 sediments fallbetween the Pearl River Holocene and ODP Site 1144 data, plottinggenerally at slightly higher εNd values than Taiwanese Rivers. This isconsistent with their dominant source being southern China withsome potential influx in Taiwan. Fig. 6B shows a simplified Sr–Nd co-evolution for ODP Site 1148. The Lower Miocene sediments plot withthe lowest 87Sr/86Sr values and the highest εNd values, lying closest toODP Site 1144. Below the sediments from ODP Site 1144 are fromTaiwan this cannot be the case for Lower Miocene sediments at ODPSite 1148 because Taiwan had not even started to form at this time, atleast in the region of its present exposure. The peak 87Sr/86Sr value isat ~7 Ma, but most of the sediment deposited in the period 18 to 7 Mafalls into a field overlapping with Holocene sediments in the PearlRiver estuary, strongly suggestive of a source in southern China, if notfrom the Pearl River itself. After 5 Ma the influence of Taiwan makes asimple weathering interpretation of the Sr isotope data impossible.

Clay mineralogy can also act as a guide to sediment provenance.Fig. 7 shows a triangular plot displaying the different clay assemblagesfor the sediment at ODP Site 1146 compared to modern rivers in theregion as well as to recent sediments on the continental shelf offshorethe Pearl River mouth, as well as from the modern Pearl River itself(Liu et al., 2007b). The assemblages seemed to span the entire rangebetween Taiwan and Luzon, but show a heavy overlap with sedimentsin the Holocene Pearl River Delta. After the uplift of Taiwan it is logicalthat some of the sediment may be coming from this island and detailedstudies of ODP Site 1144 do indicate that most of the sediment in thatregion is coming from Taiwan. We do note however that ODP Site1144 does ~200 km closer to Taiwan than the two sites consideredhere. The lack of correlation between the modern Pearl River and

Page 11: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

96 P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

offshore sediments any reinforces the interpretation that this river isnow in a state of anthropogenic disruption and is not representative oftheflux to the ocean during the geological past. The clayminerals supporta dominant provenancewithin southern China. It is questionablewhetherthe sediments that plot closest to the Luzon field are truly derived fromthis source. These sediments are the oldest in the cored section, largelypredating 12 Ma when Luzon and the associated, now subducted, partsof its arc lay somewhat to the south of the drill site (Hall, 2002). Alterna-tively, smectite-rich and high εNd sediment could be derived fromweathering of volcanic rocks along the southern coast of China, prior tothe expansion of the Pearl River into the continental interior. The largevariation in smectite content seen even within the Holocene suggeststhat the regions of southern China are quite capable of producing largevolumes of this clay mineral and do not require erosion from Luzon.

8.7. Environments in the Plio-Pleistocene

It has been recognized for some time that the Plio-Pleistocene ischaracterized by powerful physical erosion driven by the high variabilityof global climate during the onset of Northern Hemispheric Glaciation(Métivier et al., 1999; Zhang et al., 2001; Molnar, 2004). Our data andthat of Wan et al. (2007) at ODP Site 1146 broadly support this idea.Clastic MARs and whole core Ti/Ca rose steeply during this period(Fig. 3B and D), as did the proportion of chlorite (Fig. 3F), indicatingstrong physical erosion at that time. This change comes against a back-ground of weakening chemical weathering (Fig. 5F, G).

However, this area is poorly suited to examining changes in Plio-Pleistocene erosion driven by climate change because this time periodcoincides with the emergence of Taiwan (Huang et al., 2006) and allproxies after ~5 Ma are likely influenced by that process. Although thePlio-Pleistocene 87Sr/86Sr values at ODP Site 1148 are more positivethan those at ODP Site 1144 (Hu et al., 2012), indicating less influencefrom Taiwan and more from southern China, it is likely that there stillis some influx from that source. Although, the modern Pearl River isrich in kaolinite (Liu et al., 2007a) we note here that kaolinite hasbeen decreasing as a proportion of the clay mineral assemblage at ODPSite 1146 since ~7 Ma (Fig. 5C). Data from the Holocene of the PearlRiver Delta shows that prior to the Anthropocene (~3 ka) the PearlRiver itself was depleted in kaolinite (Hu et al., 2013). Nonetheless,kaolinite/illite ratios were typically 0.4–0.5, greater than the valuesseen at OPD Site 1146 since 3 Ma when they average 0.2. This requiresinput from an illite-rich source, such as Taiwan.

Weakening chemical weathering in the Plio-Pleistocene may reflectboth the influx of Taiwanese sediment and the effect of long-termglobalcooling, rather than amajor drying of the climate. Experiments indicatethat higher temperatures are associated with higher rates of chemicalweathering, especially in the presence of high humidity (White et al.,1999; West et al., 2005). The strong physical erosion we record in thePlio-Pleistocene (Fig. 3B, C, D andE) canbe interpreted to bepartly drivenby some periods (likely interglacial times) of strong summer rainsbecause in the southern South China Sea, which is not influenced byTaiwanese uplift, clastic mass flux rates also increase at that time(Wan et al., 2006). Bedrock erosion and sediment transport are moreefficient in regions with greater climatic seasonality (Molnar, 2001;Molnar, 2004). This tendency is accentuated by the known weakeningof summer monsoons during glacial times and strengthening duringinterglacial periods. We thus propose that summer rains intensifiedduring interglacial periods after 4 Ma, but that the cooling climate anddrier glacial periods reduced the overall chemicalweathering comparedto the Late Miocene (Fig. 5C, E, and F).

Fig. 8. Compilation of some of the more robust erosion and weathering proxies spanning 25(B) K/Rb as a measure of chemical weathering intensity from the data of Wei et al. (2006), (Cbiomass (Jia et al., 2003), (D) Scanned Ti/Ca from ODP Site 1148 (Hoang et al., 2010) with thform ODP Site 1148, (F) the CRAT proxy of Clift et al. (2008) tracking the relative influence o(Sun and Wang, 2005).

8.8. Weathering in the Late Miocene

The Late Miocene (~7–8 Ma) has long been considered as a time ofsummermonsoon intensification, yet this does easily match the compi-lation presented here. Clastic MAR falls slightly after 8 Ma, as do Ti/Cavalues indicating less clastic run-off, normally associatedwith less rainfall.In contrast, chlorite shows a steady increase in concentration suggestingmore physical erosion, which could be linked to a stronger monsoon,and kaolinite/illite shows a marked reduction between 8 and 7 Ma,indicative of stronger physical erosion relative to chemical weathering.CIA, CRAT values, K/Rb and 87Sr/86Sr are consistent with that shift inshowing reducing chemical weathering after that time, suggestive ofless humidity. The decrease in 87Sr/86Sr values is especially marked.While decreasing hematite/goethite usually indicates a wetter climatethis is inconsistentwith the other chemicalweatheringdata,we suggestthat the lower hematite/goethite reflects decreased seasonality and aweaker monsoon, similar to the situation in the Early Holocene (Huet al., 2012).

Global δ18O values do not change greatly in this period (Zachos et al.,2001), so we do not think that global cooling caused the changes inchemical weathering over the period 10 to 5 Ma. Instead a weakeningsummer monsoon with reduced run-off at 7.5 Ma (Steinke et al., 2010)would explain the reduced physical erosion. A drier climate might alsobe expected to cause less chemical weathering in the way observed.

8.9. Erosion at the MMCO

It is clear from Fig. 4 that the period of higher clastic MAR and higherTi/Ca at 15.5–17.2 Ma corresponds to a time of enhanced physical ero-sion. Although the increase in Ti/Ca correlates well with an apparentsharp sea-level fall at 16.5 Ma (Haq et al., 1987), this is short lived andcould not account for the longer lasting clastic MAR and Ti/Ca maxima.We interpret the high Ti/Ca ratios to reflect greater fluvial input fromsouthern China from 16.5 to 15.5 Ma, diluting the pelagic carbonatebackground.

Al/Si values were depressed during part of the clastic MAR peak,especially at 15.9–16.4 Ma, which coincides with the grain-size maxi-mum, as might be anticipated. These data point to faster physical erosionat 17.2 until 15.5 Ma, i.e. during the MMCO. We propose that this fastererosion is caused by greater run-off and heavier precipitation at that time.

Lower hematite/goethite ratios indicate that15.5–17.2 Ma was aperiod of increasing humidity. Hematite/goethite ratios behave in theway they are normally interpreted in tropical settings, i.e. with fallinghematite during wet periods (Harris and Mix, 1999). Chemicalweathering proxies generally indicate that this period was one ofmore altered sediment, consistent with faster weathering under theinfluence of a hot, wet, climate coinciding with the climatic optimum(Wanet al., 2009). Our data do not require a strongly seasonal,monsoonclimate but rather a more tropical environment. It is the lack of season-ality that would inhibit the formation of hematite.

CRAT values were high at that time interpreted to indicate high phys-ical erosion outstripping even the elevated chemical weathering. HighCRAT values at ~15.5–17.0Ma are not driven by increased flux of chloritebut by lower hematite. It is noteworthy that prior to 17.2Ma CRAT valueswere low (i.e., low chemical weathering; Fig. 4H), at a timewhen grain-sizes, clastic MAR and Ti/Ca were also reduced (low physical erosion;Figs. 4A, 5B and C). Together these indicate that physical erosion wasslower before 17.2 Ma compared to at the MMCO (15.5–17.2 Ma). Themonsoon may still have been strong then, but rainfall was less intensethan during the MMCO.

Ma at ODP Sites 1146 and 1148. (A) Kaolinite/(illite + chlorite) from Wan et al. (2007),) Carbon isotopes from black carbon from ODP Site 1148 representing regional terrestriale vertical scale expanded to eliminate the influx from Taiwan since 5 Ma, (E) Sr isotopesf chemical weathering versus physical erosion, and (G) Pollen proxies from central Asia

Page 12: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

Pollen Proxies from Central Asia

Ephedripitesgymnosperm

NitrariaShrub

PotomogetonPondweed

Pinus pollenitesPine

Picea pollenitesSpruce

>40%

>30%

20-40

20-30 15-20 10-15 4-10 2-4 1-2 <1%

11-19 7-10 3-6 1-2 <1%upper plotlower plot

0 5 10 15 20 25 30

Xin

ing–

Min

he B

asin

Cen

tral

Chi

na

G

Age (Ma)

0

2

4

OligoceneE. MioceneMid. Mioc.L. MiocenePlio.Pl.

0 5 10 15 20 25 30

K/A

l

Less

che

mic

alw

eath

erin

g

0 15 20 25 30

Less

che

mic

alw

eath

erin

g

10 15 20 25 300

87S

r/86

Sr 0.7195

0.7135

0.7165

0.7115

Less

sea

sona

lcl

imat

eM

ore

leac

hing

Sm

ectit

e/K

aolin

ite

0 15 20 25 30

Ti/C

aS

ite 1

148

0.05

0.10

0.15

0 15 20 25 30

CΔ1

3

15

17

19

21

23

1315 20 25 300

C3

C4

C

B

A

D

E

F

0

Mor

e se

ason

al

1.3

1.1

0.9

0.7

5

5

5

5

5

MM

CO

Humid floraArid flora

Mor

e cl

astic

flux

to s

ea

0

4

8

12 Seasonal

Tropical

-2

Hem

atite

/Goe

thite

Carbon isotopes

Mass flux, Ti/Ca

Sr isotopes

Hematite/Goethite

Clay mineralogy

Tropical not monsoonalMonsoonMonsoonVariablemonsoon Dry

Chemical weathering, K/Al

Wea

keni

ngM

onso

on

10

10

10

10

Transition toHumid Climate

97P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

Page 13: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

Heavy seasonal precipitationHightemperature

Burial, diagenesis

and authigenesis

Current

re-distribution

5

Physical weathering

Chemical alteration

climate dependant

River

Transport 10s to 100s of km

Soil formation

Palaeoclimatesignal preserved

in slope sediment?

High smectiteLow K/Al, Low CRAT

Bottom curre

nt

Early Miocene, 20 Ma

High sealevel

Narrow shelf

A

Strong erosion

C3 dominantflora

Chemical alteration

climate dependant

Bottom curre

nt

Weaker precipitationLowertemperature

Late Miocene, 6 Ma

Lower sealevel

Wide shelf

High K/AlHigh chlorite, High CRAT

B

Weaker erosion

C4 dominantflora

Fig. 9. Cartoon showing themany processes and factors that must be consideredwhen trying to extract paleoclimatic information frommarine sediment, whichmay be linked directly orindirectly to climatically modulated surface processes. (A) Middle Miocene conditions with high sealevel, humidity and rainfall, contrasting with (B) the Pliocene when rainfall hadreduced as sealevel fell and temperatures declined resulting in more arid weathering on wide exposed shelves. Modified after Fagel (2007).

98 P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

8.10. Erosion at 23.2–21.3 Ma

The earliest identified period of enhanced mass flux, as indicated byTi/Ca and Al/Si values is dated at 21.3–23.2 Ma (Fig. 3B). Sediment

deposited at this time is also quite altered, as tracked by K/Al and CIA.Sea level appears to have been relatively high and more stable than at15.0–17.2 Ma, with just a small, progressive fall in level (Haq et al.,1987). Thismeans that the clastic pulse probably reflected faster erosion

Page 14: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

99P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

onshore rather than relative proximity of the drill site to the Pearl Rivermouth, or exposure of the shelf. CRAT values are high, indicating morephysical erosion relative to chemical weathering. Unfortunately thereis no XRD mineralogy from ODP Site 1146 spanning this time period,but the low hematite/goethite ratios measured at ODP Site 1148 wouldsuggest that the climatewas not very seasonal, i.e. aweak summermon-soon but quite tropical. The global oxygen isotope curve indicates thatthe clastic pulse occurred during a generally warm and stable periodcompared to the following colder stage (Zachos et al., 2001).

8.11. Synthesis

Fig. 8 shows a number of the key proxies ground-truthed in the last14 ka by Hu et al. (2012; 2013) and discussed above. These showthe development of weathering and erosion in southern China since25Ma. The records are shownwith a line indicating the start of influencefrom the Taiwan orogeny. What is clear is that chemical weatheringpeaked early in the record, especially at 21–23 Ma and at the MMCO.Sediment became less altered after 15Ma, then stabilized until a furtherdecline after 8.5 Ma, as tracked by K/Al (Fig. 8A). This long-term trendwas also revealed by CIA values (Fig. 5F)(Wei et al., 2006).We interpretthis trend to be largely controlled by the long-termdecline in global tem-peratures, slowing alteration rates rather than continuous weakening ofthe summer monsoon, because the latter mechanism would also causereducing physical erosion that we do not observe. Mass flux measuredby Ti/Ca shows a pulse at 21 Ma, with another at 19 Ma and a majorpeak between 17 and 15 Ma, coinciding with the MMCO. This is consis-tentwith the idea that theMMCOwas especially erosive on a global scale(Wan et al., 2009), but does not equate to a strong monsoon but rathermore tropical conditions. After a brief period of lower Ti/Ca there is arise after 12Ma followed by a fall to aminimumafter 8.0Ma.We suggestthat the peak periods represent periods of stronger erosion duringperiods of heavy rains.

A third group of proxies show a common form and are generallyinterpreted to indicate relative strength of chemical weathering. CRAT,and Sr isotopes (Fig. 8E) show a change towards more chemicalweathering between ~23 and 17 Ma. Fig. 9A shows in cartoon formhow high temperatures and heavy monsoon rainfall in the EarlyMiocenewould result in fast erosion and the delivery of strongly alteredmaterial to the continental margin. High hematite/goethite valuessuggest that this alteration occurred under more seasonal, rather thantropical conditions. However, this is not consistent with the smectite/kaolinite, CIA and the K/Al data. Sr isotopes, CRAT and hematite/goethiteare not only dependent on temperature-modulated chemicalweatheringandwe suggest are better proxies for summermonsoon conditions. Theyshow a sharp fall at the MMCO, probably because physical erosionincreases relative to chemical weathering at that more tropical time.They show a consistent long-lived period of high values from ~16 to10 Ma and then decline into the Late Miocene, reaching minima in thePliocene. In doing so they parallel regional enhanced mass flux to theocean reconstructed from seismic data (Clift, 2006). Fig. 9B shows thesituation with weaker seasonal monsoon rainfall, colder temperaturesand lower sealevel than seen in the Early Miocene.

We compare these recordswith the carbon isotope record of Jia et al.(2003) that supposedly represents the carbon isotope composition ofterrestrial biomass in the continental flood plain. This can be used todistinguish the transition from grassland C4 flora to C3 woodlandflora, with Jia et al. (2003) placing the boundary at a CΔ13 value of 18(Fig. 8C). C4flora are generallymore tolerant of seasonal, dry conditionsthan C3. This record shows an initial change in flora starting at 22 Ma,followed by a shift back to more C3-dominated, tropical conditionsfrom 18 to 15 Ma during the MMCO. The importance of 15 Ma is againhighlighted as the flora become more C4-dominated at that time, onlyshifting back to more C3 influence after 10 Ma. From 8.5 Ma the carbonisotopes show a progress increase to more C4 flora until 5 Ma. Drying ofthe climate matches the low degrees of alteration seen in sediment of

that age and is consistent with pollen and chemical weatheringevidence from central China that shows drier, cooler conditions after11–12 Ma (Jiang et al., 2008).

This pattern bears some similaritywith pollen data from central Asiacompiled by Sun andWang (2005) (Fig. 8G), aswell as indications fromthe Loess Plateau that the monsoon may have initiated prior to 22 Ma(Guo et al., 2002). The pollen record does not have fine temporal resolu-tion, but shows a shift starting at the base of theMiocene frommore aridflora into more humid assemblages, especially in the Mid Miocene, butdeclining to less humid flora in the Late Miocene. This is consistent withthe data presented here for an increase in summer monsoon intensitystarting after ~22 Ma. The Middle Miocene appears to be the mosthumid phase, which might be expected given the influence of theMMCO. The progressive drying in central Asia may partly reflect therain shadow effect of a rising Tibetan Plateau (Zheng et al., 2010), butcombined with a gradual weakening of the summer monsoon in thecontext of a cooling Earth.

9. Conclusions

This study shows that a combination of several geochemical andmineral proxies can be an effective way of reconstructing continentalweathering and erosion histories and allow these processes to be com-pared and used to constrain variations in climate and thus in summermonsoon strength. Chemical weathering and physical erosion intensityare decoupled from one another, but are not polar opposites, both beingstrong at 15.5–17.2 Ma during the Mid Miocene Climatic Optimum. Incontrast, physical erosion is strong but chemical weathering was weaksince 4 Ma, because of greater seasonality, colder temperatures andweak summer rains during glacial periods. However, the influence ofTaiwan after 5 Ma makes it hard to uniquely define Pliocene–Recentmonsoon intensity from this region.

Clastic MARs provide important information on erosion rates, butare susceptible to switches in depocenters and to a lesser extent to sealevel variations. They act as better proxies for continental erosion inten-sities when coupled with Ti/Ca and with Al/Si. Scanned geochemicaldata appear to yield useful and robust erosion records. We infer aclose association of rapid erosion with the intensity of the summermonsoon if tectonic forces are not highly variable, or if major drainagecapture as not occurred, as appears to be the case prior to 5 Ma in thisregion.

CIA and K/Al are shown to be robust indicators of chemicalweathering intensities, but are closely coupled to global temperaturesand are less dependent on summer rainfall intensity. They showa gradualdecrease from a maximum at the MMCO. Kaolinite/illite shows a similartrend to CIA and K/Al but remained high until around 8Ma after which ithas declined to the present day, partly driven by temperatures and partlyby reducing humidity as the monsoon weakened after 8 Ma. 87Sr/86Srdoes not appear to be controlled by provenance and is interpreted to belinked to weathering processes. It shows a similar history as hematite/goethite with both of them favoring a long period of seasonal climaticconditions from around 18 to 10 Ma, with an intermission for moresteady humid conditions during the MMCO. Hematite productionrequires at least intermittent dry conditions and is more likely to beformed when the climate is seasonal with long, dry periods betweenthe summer rains rather than more tropical. The new spectrally basedCRAT proxy has close parallels with the more established hematite/goethite spectral-based proxy and with 87Sr/86Sr at ODP Site 1148.These proxies are more weathering oriented but differs significantlyfrom the trends in CIA and K/Al. They reflect seasonality as well as tem-perature. Because seasonality is a key feature of the monsoon they areconsidered here to be important paleo-monsoon proxies.

This study indicates that chemical weathering weakened after apeak in the Middle Miocene, largely tracking the global cooling afterthat time. At the same time physical erosion increased at least over sig-nificant periods and is interpreted to track the intensity of summer

Page 15: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

100 P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

monsoon rains, which are now responsible for much of the erosionaldischarge fromAsian river systems. Erosion rates andmonsoon strengthare at their lowest following the decline that started after ~8 Ma(Fig. 9B), but may have become stronger since 5Ma, although the influ-ence from Taiwan makes this hard to demonstrate in the study region.The timing of initial monsoon intensification would appear to be~21 Ma, following a warm wet period of more tropical conditions.Such an evolution is consistent with terrestrial pollen data (Sun andWang, 2005) and the black carbon isotope record (Jia et al., 2003). Thelack of a continuous record prior to 25 Ma makes it impossible to testfor earlier periods of monsoon activity.

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.earscirev.2014.01.002.

Acknowledgements

PC wishes to thank Maureen Conte and J.C. Weber at the MarineBiology Laboratory, Woods Hole for their help with the carbonatecoulometry work. Ursula Röhl and Vera Lukies provided support forthe XRF scanning at MARUM, Bremen. Analyses were funded by agrant from JOI-USSAC and by the University of Aberdeen. PC thanksthe Hanse Wissenschaftkolleg in Delmenhorst, Germany for the timeto work on this topic. SW thanks the China NSF (41076033) for support.The work benefited from discussions with Ryuji Tada, Liviu Giosan,Steve Clemens andWarren Prell. Steve Clemens and an anonymous re-viewer provided important comments on an earlier version of this man-uscript. Jun Chen and an anonymous reviewer helped improve thepresent manuscript.

References

Allan, R., Lindesay, J., Parker, D., 1996. El Nino: Southern Oscillation andClimatic Variability.CSIRO Publishing, Canberra (416 pp.).

Armstrong, H.A., Allen,M.B., 2011. Shifts in the Intertropical Convergence Zone, Himalayanexhumation, and late Cenozoic climate. Geology 39 (1), 11–14.

Berner, R.A., Berner, E.K., 1997. Silicate weathering and climate. In: Ruddiman, W.F. (Ed.),Tectonic Uplift and Climate Change. Springer, New York, pp. 353–365.

Blaxland, A.B., 1974. Geochemistry and geochronology of chemical weathering, Butler HillGranite, Missouri. Geochim. Cosmochim. Acta 38 (6), 843–852.

Bookhagen, B., Thiede, R.C., Strecker, M.R., 2005. Abnormal monsoon years and theircontrol on erosion and sediment flux in the high, arid Northwest Himalaya. EarthPlanet. Sci. Lett. 231 (1–2), 131–146.

Boulay, S., Colin, C., Trentesaux, A., Clain, S., Liu, Z., Lauer-Leredde, C., 2007. Sedimentaryresponses to the Pleistocene climatic variations recorded in the South China Sea.Quat. Res. 68, 162–172.

Brookfield, M.E., 1998. The evolution of the great river systems of southern Asia duringthe Cenozoic India–Asia collision; rivers draining southwards. Geomorphology 22(3–4), 285–312.

Burbank, D.W., Blythe, A.E., Putkonen, J., Pratt-Sitaula, B., Gabet, E., Oskins, M., Barros, A.,Ojha, T.P., 2003. Decoupling of erosion and precipitation in the Himalayas. Nature426, 652–655.

Chamley, H., 1989. Clay Sedimentology. Springer–Verlag, Berlin (267 pp.).Chappell, J., Zheng, H., Fifield, K., 2006. Yangtse River sediments and erosion rates from

source to sink tracedwith cosmogenic 10Be: sediments frommajor rivers. Palaeogeogr.Palaeoclimatol. Palaeoecol. 241, 79–94.

Chen, C.H., Lee, T., 1990. A Nd–Sr isotopic study on river sediments of Taiwan. Proc. Geol.Soc. China 33 (4), 339–350.

Chen, M., Wang, R., Yang, L., Han, J., Lu, J., 2003. Development of East Asian summermonsoon environments in the late Miocene; radiolarian evidence from Site 1143 ofODP Leg 184. Mar. Geol. 201, 169–177.

Clark, M.K., Schoenbohm, L.M., Royden, L.H., Whipple, K.X., Burchfiel, B.C., Zhang, X., Tang,W., Wang, E., Chen, L., 2004. Surface uplift, tectonics, and erosion of eastern Tibetfrom large-scale drainage patterns. Tectonics 23, TC1006.

Clemens, S.C., 1998. Dust response to seasonal atmospheric forcing: proxy evaluation andcalibration. Paleoceanography 13 (5), 471–490.

Clemens, S., Prell, W., Murray, D., Shimmield, G., Weedon, G., 1991. Forcing Mechanismsof the Indian-Ocean Monsoon. Nature 353 (6346), 720–725.

Clemens, S.C., Prell, W.L., Sun, Y., Liu, Z., Chen, G., 2008. Southern Hemisphere forcing ofPliocene δ18O and the evolution of Indo-Asian monsoons. Paleoceanography 23,PA4210.

Clemens, S.C., Prell, W.L., Sun, Y., 2010. Orbital-scale timing and mechanisms driving LatePleistocene Indo-Asian summer monsoons: reinterpreting cave speleothem ∂18O.Paleoceanography 25, PA4207.

Clift, P.D., 2006. Controls on the erosion of Cenozoic Asia and the flux of clastic sedimentto the ocean. Earth Planet. Sci. Lett. 241 (3–4), 571–580.

Clift, P.D., Giosan, L., 2014. Sediment fluxes and buffering in the post-glacial Indus Basin.Basin Res. http://dx.doi.org/10.1111/bre.12038 (in press).

Clift, P., Lin, J., 2001. Preferential mantle lithospheric extension under the South Chinamargin. Mar. Pet. Geol. 18 (8), 929–945.

Clift, P.D., Plumb, R.A., 2008. The Asian Monsoon: Causes, History and Effects. CambridgeUniversity Press, Cambridge (288 pp.).

Clift, P., Lee, J.I., Clark, M.K., Blusztajn, J., 2002. Erosional response of south China to arcrifting and monsoonal strengthening; a record from the South China Sea. Mar. Geol.184 (3–4), 207–226.

Clift, P.D., Blusztajn, J., Nguyen, D.A., 2006. Large-scale drainage capture and surface upliftin eastern Tibet–SW China before 24 Ma inferred from sediments of the Hanoi Basin,Vietnam. Geophys. Res. Lett. 33 (L19403).

Clift, P.D., Hodges, K., Heslop, D., Hannigan, R., Hoang, L.V., Calves, G., 2008. GreaterHimalayan exhumation triggered by Early Miocene monsoon intensification. Nat.Geosci. 1, 875–880.

Derry, L.A., France-Lanord, C., 1996. Neogene Himalayan weathering history and river87Sr/86Sr; impact on the marine Sr record. Earth Planet. Sci. Lett. 142, 59–74.

Dettman, D.L., Kohn, M.J., Quade, J., Ryerson, F.J., Ojha, T.P., Hamidullah, S., 2001. Seasonalstable isotope evidence for a strong Asian monsoon throughout the past 10.7 m.y.Geology 29 (1), 31–34.

Dykoski, C.A., Edwards, R.L., Cheng, H., Yuan, D., Cai, Y., Zhang, M., Lin, Y., Qing, J., An, Z.,Revenaugh, J., 2005. A high-resolution, absolute-dated Holocene and deglacial Asianmonsoon record from Dongge Cave, China. Earth Planet. Sci. Lett. 233 (1–2), 71–86.

Fagel, N., 2007. Marine clay minerals, deep circulation and climate. In: Hillaire-Marcel, C.,Vernal, A.d. (Eds.), Paleoceanography of the Late Cenozoic, Volume 1: Methods.Elsevier, Amsterdam, pp. 139–184.

Farquhar, G.D., Ehleringer, J.R., Hubick, K.T., 1989. Carbon isotope discrimination andphotosynthesis. Annu. Rev. Plant Physiol. Plant Mol. Biol. 40, 503–537.

Ferrier, K.L., Kirchner, J.W., 2008. Effects of physical erosion on chemical denudation rates:a numerical modeling study of soil-mantled hillslopes. Earth Planet. Sci. Lett. 272,591–599.

Fleitmann, D., Burns, S.J., Mudelsee, M., Neff, U., Kramers, J., Mangini, A., Matter, A., 2003.Holocene forcing of the Indian monsoon recorded in a stalagmite from southernOman. Science 300 (5626), 1737–1739.

Galy, V., François, L., France-Lanord, C., Faure, P., Kudrass, H., Palhol, F., Singh, S., 2008. C4plants decline in the Himalayan basin since the Last Glacial Maximum. Quat. Sci. Rev.27, 1396–1409.

Giosan, L., Clift, P.D., Macklin, M.G., Fuller, D.Q., Constantinescu, S., Durcan, J.A., Stevens, T.,Duller, G.A.T., Tabrez, A., Adhikari, R., Gangal, K., Alizai, A., Filip, F., VanLaningham, S.,Syvitski, J.P.M., 2012. Fluvial Landscapes of the Harappan Civilization. Proc. Natl. Acad.Sci. 109 (26), 1688–1694.

Goodbred, S.L., Kuehl, S.A., 1998. Floodplain processes in the Bengal Basin and the storageof Ganges–Brahmaputra River sediment; an accretion study using 137Cs and 210Pbgeochronology. Sediment. Geol. 121 (3–4), 239–258.

Goodbred, S.L., Kuehl, S.A., 2000. Enormous Ganges–Brahmaputra sediment dischargeduring strengthened early Holocene monsoon. Geol. (Boulder) 28 (12), 1083–1086.

Guo, Z.T., Ruddiman, W.F., Hao, Q.Z., Wu, H.B., Qiao, Y.S., Zhu, R.X., Peng, S.Z., Wei, J.J.,Yuan, B.Y., Liu, T.S., 2002. Onset of Asian desertification by 22 Myr ago inferredfrom loess deposits in China. Nature (London) 416 (6877), 159–163.

Gupta, A.K., Anderson, D.M., Overpeck, J.T., 2003. Abrupt changes in the Asian southwestmonsoon during the Holocene and their links to the North Atlantic Ocean. Nature421, 354–356.

Hall, R., 2002. Cenozoic geological and plate tectonic evolution of SE Asia and the SWPacific: computer-based reconstructions and animations. J. Asian Earth Sci. 20, 353–434.

Haq, B.U., Hardenbol, J., Vail, P.R., 1987. Chronology of fluctuating sea levels since theTriassic. Science 235, 1156–1167.

Harris, S.E.,Mix, A.C., 1999. Pleistoceneprecipitation balance in theAmazon Basin recordedin deep sea sediments. Quat. Res. 51, 14–26.

Herzschuh, U., Zhang, C., Mischke, S., Herzschuh, R., Mohammadi, F., Mingram, B.,Kuerschner, H., Riedel, F., 2005. A late Quaternary lake record from theQilianMountains(NW China); evolution of the primary production and the water depth reconstructedfrom macrofossil, pollen, biomarker, and isotope data. Glob. Planet. Chang. 46 (1–4),361–379.

Hess, S., Kuhnt, W., 2005. Neogene and Quaternary paleoceanographic changes in thesouthern South China Sea (Site 1143); the benthic foraminiferal record. Mar.Micropaleontol. 54, 63–87.

Hoang, L.V., Clift, P.D., Schwab, A.M., Huuse, M., Nguyen, D.A., Zhen, S., 2010. Large-scaleerosional response of SE Asia to monsoon evolution reconstructed from sedimentaryrecords of the Song Hong–Yinggehai and Qiongdongnan Basins, South China Sea. In:Clift, P.D., Tada, R., Zheng, H. (Eds.), Monsoon evolution and tectonic-climate linkagein Asia. Special Publication. Geological Society, London, pp. 219–244.

Hu, J., Kawamura, H., Hong, H., Qi, Y., 2000. A review on the currents in the South ChinaSea: seasonal circulation, South China Sea Warm Current and Kuroshio Intrusion.J. Oceanogr. 56, 607–624.

Hu, D., Böning, P., Köhler, C.M., Hillier, S., Pressling, N., Wan, S., Brumsack, H.-J., Clift, P.D.,2012. Deep sea records of the continental weathering and erosion response to EastAsian monsoon intensification since 14 ka in the South China Sea. Chem. Geol.326–327, 1–18.

Hu, D., Clift, P.D., Böning, P., Hannigan, R., Hillier, S., Blusztajn, J., Wang, S., Fuller, D.Q.,2013. Holocene evolution in weathering and erosion patterns in the Pearl Riverdelta. Geochem. Geophys. Geosyst. 14.

Huang, C.Y., Yuan, P.B., Tsao, S.H., 2006. Temporal and spatial records of active arc–continentcollision in Taiwan: a synthesis. Geol. Soc. Am. Bull. 118, 274–288.

Hurlbut, C.S., Klein, C., 1985.Manual ofMineralogy. JohnWiley andSons,NewYork (324pp.).Ji, J., Chen, J., Balsam, W., Lu, H., Sun, Y., Xu, H., 2004. High resolution hematite/goethite

records from Chinese loess sequences for the last-glacial–interglacial cycle; rapid

Page 16: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

101P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

climatic response of the east Asianmonsoon to the tropical Pacific. Geophys. Res. Lett.31 (3).

Jia, G., Peng, P., Zhao, Q., Jian, Z., 2003. Changes in terrestrial ecosystem since 30Ma in EastAsia: stable isotope evidence from black carbon in the South China Sea. Geology 31,1093–1096.

Jiang, H., Ji, J., Gao, L., Tang, Z., Ding, Z., 2008. Cooling-driven climate change at 12–11 Ma:multiproxy records from a long fluviolacustrine sequence at Guyuan, Ningxia, China.Palaeogeogr. Palaeoclimatol. Palaeoecol. 265 (1–2), 148–158.

Kroon, D., Steens, T., Troelstra, S.R., 1991. Onset of Monsoonal related upwelling in thewestern Arabian Sea as revealed by planktonic foraminifers. In: Prell, W., Niitsuma,N. (Eds.), Proceedings of the Ocean Drilling Program, Scientific Results. Ocean DrillingProgram, College Station, TX, pp. 257–263.

Kudrass, H.-R., Hofmann, A., Doose, H., Emeis, K.C., Erlenkauser, H., 2001. Modulation andamplification of climatic changes in the Northern Hemisphere by the Indian summermonsoon during the past 80 k.y. Geology 29, 63–66.

Lan, C.Y., Lee, C.-S., Shen, J.J.-S., Lu, C.Y., Mertzman, S.A., Wu, T.-W., 2002. Nd–Sr isotopiccomposition and geochemistry of sediments from Taiwan and their implications.West. Pac. Earth Sci. 2 (2), 205–222.

Lei, S., Li, X., Geng, J., Pang, X., Lei, Y., Qiao, P.,Wang, L.,Wang, H., 2007. Deep water bottomcurrent deposition in the northern South China Sea. Sci. China Ser. D Earth Sci. 50 (7),1862–2801.

Li, C., Yang, S.Y., 2010. Is Chemical Index of Alteration (CIA) a Reliable Proxy for ChemicalWeathering in Global Drainage Basins? Am. J. Sci. 310, 111–127.

Li, X., Wei, G., Shao, L., Liu, Y., Liang, X., Jian, Z., Sun, M., Wang, P., 2003. Geochemical andNd isotopic variations in sediments of the South China Sea; a response to Cenozoictectonism in SE Asia. Earth Planet. Sci. Lett. 211 (3–4), 207–220.

Liu, W., Huang, Y., An, Z., Clemens, S.C., Li, L., Prell, W.L., Ning, Y., 2005. Summer monsoonintensity controls C4/C3 plant abundance during the last 35 ka in the Chinese LoessPlateau; carbon isotope evidence from bulk organic matter and individual leafwaxes. Palaeogeogr. Palaeoclimatol. Palaeoecol. 220 (3–4), 243–254.

Liu, Z., Colin, C., Huang, W., Chen, Z., Trentesaux, A., Chen, J.F., 2007a. Clay minerals insurface sediments of the Pearl River drainage basin and their contribution to theSouth China Sea. Chin. Sci. Bull. 52 (8), 1101–1111.

Liu, Z., Colin, C., Huang, W., Le, K.P., Tong, S., Chen, Z., Trentesaux, A., 2007b. Climatic andtectonic controls on weathering in south China and Indochina Peninsula: clay miner-alogical and geochemical investigations from the Pearl, Red, and Mekong drainagebasins. Geochem. Geophys. Geosyst. 8, Q05005.

Liu, Z., Tuo, S., Colin, C., Liu, J.T., Huang, C.-Y., Selvaraj, K., Chen, C.-T.A., Zhao, Y., Siringan,F.P., Boulay, S., Chen, Z., 2008. Detrital fine-grained sediment contribution fromTaiwan to the northern South China Sea and its relation to regional ocean circulation.Mar. Geol. 255 (3–4), 149–155.

Liu, Z., Colin, C., Li, X., Zhao, Y., Tuo, S., Chen, Z., Siringan, F.P., Liu, J.T., Huang, C.-Y., You,C.-F., Huang, K.-F., 2010. Clay mineral distribution in surface sediments of thenortheastern South China Sea and surrounding fluvial drainage basins: source andtransport. Mar. Geol. 277, 48–60.

Métivier, F., Gaudemer, Y., Tapponnier, P., Klein,M., 1999. Mass accumulation rates in Asiaduring the Cenozoic. Geophys. J. Int. 137 (2), 280–318.

Molnar, P., 2001. Climate change, flooding in arid environments, and erosion rates. Geol.(Boulder) 29 (12), 1071–1074.

Molnar, P., 2004. Late Cenozoic increase in accumulation rates of terestrial sediment: howmight climate change have affected erosion rates? Annu. Rev. Earth Planet. Sci. 32,67–89.

Molnar, P., England, P., Martinod, J., 1993. Mantle Dynamics, Uplift of the Tibetan Plateau,and the Indian Monsoon. Rev. Geophys. 31 (4), 357–396.

Nelson, S.V., 2005. Paleoseasonality inferred from equid teeth and intra-tooth isotopicvariability. Palaeogeogr. Palaeoclimatol. Palaeoecol. 22 (1–2), 122–144.

Nesbitt, H.W., Young, G.M., 1982. Early Proterozoic climates and plate motions inferredfrom major element chemistry of lutites. Nature 299 (5885), 715–717.

Nesbitt, H.W., Fedo, C.M., Young, G.M., 1997. Quartz and feldspar stability, steady andnon-steady-state weathering, and petrogenesis of siliciclastic sands and muds.J. Geol. 105 (2), 173–191.

Prell, W.L., Kutzbach, J.E., 1992. Sensitivity of the Indian Monsoon to forcing parametersand implications for its evolution. Nature 360 (6405), 647–652.

Prell, W.L., Murray, D.W., Clemens, S.C., Anderson, D.M., 1992. Evolution and variability ofthe Indian Ocean Summer Monsoon: evidence from the western Arabian Sea drillingprogram. In: Duncan, R.A., Rea, D.K., Kidd, R.B., von Rad, U., Weissel, J.K. (Eds.),Synthesis of results from scientific drilling in the Indian Ocean. GeophysicalMonograph.American Geophysical Union, Washington, DC, pp. 447–469.

Quade, J., Cerling, T.E., Bowman, J.R., 1989. Development of Asian monsoon revealed bymarked ecological shift during the latest Miocene in northern Pakistan. Nature 342(6246), 163–166.

Rasmusson, E.M., Carpenter, T.H., 1982. Variations in Tropical Sea Surface Temperatureand Surface Wind Fields Associated with the Southern Oscillation/El Niño. Mon.Weather Rev. 110, 354–384.

Raymo, M.E., Ruddiman, W.F., Froelich, P.N., 1988. Influence of Late Cenozoic mountainbuilding on ocean geochemical cycles. Geology 16 (7), 649–653.

Rea, D.K., Snoeckx, H., Joseph, L.H., 1998. Late Cenozoic eolian deposition in the NorthPacific: Asian drying, Tibetan uplift, and cooling of the northern hemisphere.Paleoceanography 13, 215–224.

Rego, J.L., Meselhe, E., Stronach, J., Habib, E., 2010. Numerical modeling of theMississippi–Atchafalaya Rivers' sediment transport and fate: considerations for diversion scenarios.J. Coast. Res. 262, 212–229.

Riebe, C.S., Kirchner, J.W., Granger, D.E., Finkel, R.C., 2001. Strong tectonic and weakclimatic control of long-term chemical weathering rates. Geology 29, 511–514.

Röhl, U., Abrams, L.J., 2000. High-resolution, downhole and non-destructive coremeasurements from Sites 999 and 1001 in the Caribbean Sea: application to the

Late Paleocene Thermal Maximum. Proc. Ocean Drill. Program Sci. Results 165,191–203.

Ru, K., Pigott, J.D., 1986. Episodic rifting and subsidence in the South China Sea. AAPG Bull.70 (9), 1136–1155.

Schwertmann, U., 1971. Transformation of hematite to goethite in soils. Nature 232,624–625.

Schwertmann, U., 1988. Occurrence and formation of iron oxides in various pedo-environments. In: Stucki, J.W. (Ed.), Iron in soils and clayminerals. D. Reidel, Norwell,pp. 267–308.

Shipboard Scientific Party, 1989a. Site 718. In: Cochran, J.R., Stow,D.A.V. (Eds.), Proceedingsof the Ocean Drilling Program, Part A: Initial Reports. Ocean Drilling Program, CollegeStation, TX, pp. 91–154.

Shipboard Scientific Party, 1989b. Site 722. In: Prell, W.L., Niitsuma, N. (Eds.), Proceedingsof the Ocean Drilling Program, Part A: Initial Reports. Ocean Drilling Program, CollegeStation, TX, pp. 255–317.

Shipboard Scientific Party, 2000a. Site 1144. Proceedings of the Ocean Drilling Program,Part A: Initial Reports, pp. 1–97.

Shipboard Scientific Party, 2000b. Site 1148. Proceedings of the Ocean Drilling Program,Part A: Initial Reports.

Sinha, A., Cannariato, K.G., Stott, L.D., Li, H.-C., You, C.-F., Cheng, H., Edwards, R.L., Singh,I.B., 2005. Variability of southwest Indian summer monsoon precipitation duringthe Bolling–Allerod. Geology 33 (10), 813–816.

Steinke, S., Groeneveld, J., Johnstone, H., Rendle-Bühring, R., 2010. East Asian summermonsoon weakening after 7.5 Ma: evidence from combined planktonic foraminiferaMg/Ca and δ18O (ODP Site 1146; northern South China Sea). Palaeogeogr.Palaeoclimatol. Palaeoecol. 289 (1–4), 33–43.

Su, X., Xu, Y., Tu, Q., Arnold, E.M., Buehring, C.J., Chen, M.-P., Clift, P.D., Colin, C.J.G., Farrell,J.W., Higginson, M.J., Jian, Z., Kuhnt, W., Laj, C.E., Lauer-Leredde, C., Leventhal, J.S., Li,A., Li, Q., Lin, J., McIntyre, K., Miranda, C.R., Nathan, S.A., Shyu, J.-P., Solheid, P.A., Su,X., Tamburini, F., Trentesaux, A.,Wang, L., 2006. Early Oligocene–Pleistocene calcareousnannofossil biostratigraphy of the northern South China Sea (Leg 184, Sites 1146–1148).Proceedings of the Ocean Drilling Program, Scientific Results (CD-ROM), 184, p. 24.

Sun, X., Wang, P., 2005. How old is the Asian monsoon system? Palaeobotanical recordsfrom China. Palaeogeogr. Palaeoclimatol. Palaeoecol. 222 (3–4), 181–222.

Tamburini, F., Adatte, T., Foellmi, K., Bernasconi, S.M., Steinmann, P., 2003. Investigatingthe history of East Asian monsoon and climate during the last glacial–interglacialperiod (0–140 000 years); mineralogy and geochemistry of ODP sites 1143 and1144, South China Sea. Mar. Geol. 201, 147–168.

Thamban, M., Rao, V.P., 2005. Clay minerals as palaeomonsoon proxies: evaluation andrelevance to the late Quaternary record from SE Arabian Sea. In: Rajan, S., Pandey,P.C. (Eds.), AntarcticGeoscience:Ocean–atmosphere Interaction andPaleoclimatology.National Centre for Antarctic & Ocean Research, Goa, India, pp. 198–215.

Thiry, M., 2000. Palaeoclimatic interpretation of clay minerals in marine deposits; anoutlook from the continental origin. Earth Sci. Rev. 49 (1–4), 201–221.

Tjallingii, R., Röhl, U., Kölling, M., Bickert, T., 2007. Influence of the water content on X-rayfluorescence core scanning measurements in soft marine sediments. Geochem.Geophys. Geosyst. 8 (2).

Wan, S., Li, A., Clift, P.D., Jiang, H., 2006. Development of the East Asian summermonsoon;evidence from the sediment record in the South China Sea since 8.5 Ma. Palaeogeogr.Palaeoclimatol. Palaeoecol. 241, 139–159.

Wan, S., Li, A., Clift, P.D., Stuut, J.-B.W., 2007. Development of the East Asian monsoon:mineralogical and sedimentologic records in the northern South China Sea since20 Ma. Palaeogeogr. Palaeoclimatol. Palaeoecol. 254 (3–4), 561–582.

Wan, S., Kürschner, W.M., Clift, P.D., Li, A., Li, T., 2009. Extremeweathering/erosion duringthe Miocene Climatic Optimum: evidence from sediment record in the South ChinaSea. Geophys. Res. Lett. 36 (L19706).

Wan, S.M., Li, A., Wu, S., Wang, X., Li, T., 2010a. Increased contribution of terrigenoussupply from Taiwan to the northern South China Sea since about 3 million yearsago. Mar. Geol. 278, 115–121.

Wan, S., Clift, P.D., Li, A., Li, T., Yin, X., 2010b. Geochemical records in the South China Sea:implications for East Asian summer monsoon evolution over the last 20 Ma. In: Clift,P.D., Tada, R., Zheng, H. (Eds.), Monsoon Evolution and Tectonics–Climate Linkage inAsia. Special Publication. Geological Society, London, pp. 245–263.

Wang, B., Clemens, S.C., Liu, P., 2003a. Contrasting the Indian and East Asian monsoons:implications on geologic timescales. Mar. Geol. 201 (1–3), 5–21.

Wang, P., Tian, J., Cheng, X., Liu, C., Xu, J., 2003b. Carbon reservoir changes precededmajorice-sheet expansion at the mid-Brunhes event. Geology 31, 239–242.

Wang, P., Zhao, Q., Jian, Z., Cheng, X., Huang, W., Tian, J., Wang, J., Li, Q., Li, B., Su, X., 2003c.Thirty million year deep-sea records in the South China Sea. Chin. Sci. Bull. 48,2524–2535.

Wang, Y., Cheng, H., Edwards, R.L., He, Y., Kong, X., An, Z., Wu, J., Kelly, M.J., Dykoski, C.A.,Li, X.D., 2005. The Holocene Asianmonsoon; links to solar changes and North Atlanticclimate. Science 308, 854–857.

Webster, P.J., Magana, V.O., Palmer, T.N., Shukla, J., Tomas, R.A., Yanai Y., M., Yasunari, T.,1998. Monsoons: processes, predictability, and the prospects for prediction, in theTOGA decade. J. Geophys. Res. 103, 14,451–14,510.

Wei, G., Li, X.-H., Liu, Y., Shao, L., Liang, X., 2006. Geochemical record of chemicalweathering and monsoon climate change since the early Miocene in the SouthChina Sea. Paleoceanography 21, PA4214.

West, A.J., Galy, A., Bickle, M.J., 2005. Tectonic and climatic controls on silicate weathering.Earth Planet. Sci. Lett. 235, 211–228.

White, A.F., Blum, A.E., Bullen, T.D., Vivit, D.V., Schulz, M., Fitzpaterich, J., 1999. The effectof temperature on experimental and natural chemical weathering rates of granitoidrocks. Geochim. Cosmochim. Acta 63 (19/20), 3277–3291.

Wobus, C.W., Hodges, K.V., Whipple, K.X., 2003. Has focused denudation sustained activethrusting at the Himalayan topographic front? Geology 31 (10), 861–864.

Page 17: Reconstructing chemical weathering, physical erosion and … · Reconstructing chemical weathering, physical erosion and monsoon intensity since 25 Ma in the northern South China

102 P.D. Clift et al. / Earth-Science Reviews 130 (2014) 86–102

Yuan, D., Cheng, H., Edwards, R.L., Dykoski, C.A., Kelly, M.J., Zhang, M., Qing, J., Lin, Y.,Wang, Y., Wu, J., Dorale, J.A., An, Z., Cai, Y., 2004. Timing, duration, and transitionsof the last interglacial Asian monsoon. Science 304 (5670), 575–578.

Zachos, J., Pagani, M., Sloan, L., Thomas, E., Billups, K., 2001. Trends, rythms andabberations in global climate 65 Ma to Present. Science 292, 686–693.

Zhang, P., Molnar, P., Downs, W.R., 2001. Increased sedimentation rates and grain sizes 2–4Myr ago due to the influence of climate change on erosion rates. Nature 410, 891–897.

Zhang, Y.G., Jia, J., Balsam,W.L., Liu, L., Chen, J., 2007. High resolution hematite and goethiterecords from ODP 1143, South China Sea: co-evolution of monsoonal precipitationand El Niño over the past 600,000 years. Earth Planet. Sci. Lett. 264 (1–2), 136–150.

Zheng, H., Tada, R., Jia, J., Lawrence, C., Wang, K., 2010. Cenozoic sediments inthe southern Tarim Basin: implications for the uplift of northern Tibet andevolution of the Taklimakan Desert. In: Clift, P.D., Tada, R., Zheng, H. (Eds.),Monsoon evolution and tectonic-climate linkage in Asia. Special Publication.GeologicalSociety.

Zheng, H., Clift, P.D., Tada, R., Jia, J.T., He, M.Y., Wang, P., 2013. A pre-Miocene birth to theYangtze River. Proceedings of the National Academy of Sciences, pp. 1–6.

Zhou, W., Lu, X., Wu, Z., Deng, L., Jull, A.J.T., Donahue, D., Beck, W., 2002. Peat recordreflecting Holocene climatic change in the Zoige Plateau and AMS radiocarbon dating.Chin. Sci. Bull. 47 (1), 66–70.