Glacial–interglacial variation in denudation rates from interior Texas, USA, established with cosmogenic nuclides

Earth and Planetary Science Letters 390 (2014) 209–221

Contents lists available at ScienceDirect

Earth and Planetary Science Letters www.elsevier.com/locate/epsl

Glacial–interglacial variation in denudation rates from interior Texas, USA, established with cosmogenic nuclides Alan J. Hidy a,∗ , John C. Gosse a , Michael D. Blum b , Martin R. Gibling a a b

Department of Earth Sciences, Dalhousie University, Halifax, Nova Scotia, B3H 4R2, Canada Exxonmobil Upstream Research Company, Houston, TX 77252, United States

a r t i c l e

i n f o

Article history: Received 18 September 2013 Received in revised form 8 January 2014 Accepted 9 January 2014 Available online 31 January 2014 Editor: T.M. Harrison Keywords: 10 Be denudation rates Gulf of Mexico climate change sediment flux coastal plain TCN

a b s t r a c t The Brazos, Colorado, and Trinity rivers of Texas drain a tectonically quiescent, non-glaciated, and lowrelief landscape inland from the Gulf of Mexico, where long-term [103 –105 a] changes in denudation rates are probably driven largely by climate change. Here, we use cosmogenic 10 Be to obtain spatially averaged denudation rates for these river catchments, primarily from terrace deposits associated with glacial or interglacial intervals over the past half million years. The denudation rates are ∼30–35% higher during interglacial periods than during glacial periods, and correlate broadly with temperature. The results are consistent with predictions from the BQART sediment flux model, and support the hypothesis that increased weathering rates associated with warmer climates will accelerate landscape erosion. Furthermore, by analyzing 26 Al/10 Be in these deposits, we can estimate the bed load sourced from upcatchment surfaces. The stored coastal plain fraction varies from ∼10% to 30%, and is greater during times of relatively lower sea level. The results indicate that although sediment flux is moderated by coastalplain storage, increased up-catchment flux during warmer interglacial periods outpaces evacuation of stored sediment during glacial periods, resulting in a net increase in sediment flux to the ocean during warm intervals. If this temperature–sediment flux relationship is valid beyond the Plio-Pleistocene transition, then global sediment flux to the ocean from passive, non-glaciated, and low-relief landscapes would have been greater during the Pliocene than in the cooler Quaternary. © 2014 Elsevier B.V. All rights reserved.

1. Introduction The relative importance of mechanisms driving global sediment flux over the last 5 Ma is uncertain. The first-order mechanisms are tectonic uplift (Raymo and Ruddiman, 1992) and global climate (Molnar, 2004; Molnar and England, 1990; Peizhen et al., 2001), with secondary, related factors including threshold changes in vegetation (Antinao and McDonald, 2013; Bull, 1991; Hay et al., 2002), permafrost (Mason and Knox, 1997), and aridity (Bull, 1991; Molnar, 2001; Pederson et al., 2000), changes in the global proportions of fluvial and glacial erosion (Koppes and Montgomery, 2009; Montgomery, 2002), and the average gradients of continental surfaces (Willenbring et al., 2013). The interconnected nature of these processes often precludes the isolation of a single mechanism, and has led to an inability to resolve the role of climate in controlling global sediment flux. Furthermore, globally integrated sediment flux estimates have remained elusive. Modern sediment flux from a single river cannot be measured with high accuracy, and the variation in sediment flux with time is even less well known. Instead, proxy data such

*

Corresponding author. Tel.: +902 494 1211; fax: +902 494 6889. E-mail address: [email protected] (A.J. Hidy).

0012-821X/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.epsl.2014.01.011

as mass reconstructions based on preserved basin fills (Hay et al., 1989; Peizhen et al., 2001), isotopic signatures of oceanic sediment (Willenbring and von Blanckenburg, 2010), and numerical models trained on modern rivers that implement space-for-time substitutions (Milliman and Syvitski, 1992; Syvitski and Milliman, 2007) are utilized. However, these proxies have produced conflicting results. In fact, one can infer that net sediment flux to the oceans may have increased (Molnar, 2004; Peizhen et al., 2001), decreased (Syvitski and Milliman, 2007), or remained the same (Willenbring and von Blanckenburg, 2010) over the Plio-Pleistocene transition. To improve our understanding of the landscape response to climate change a study area must (1) be removed from effects of tectonics, changes in sediment routing, and direct glaciation so that the climatic component can be isolated, and (2) have robust measurements of paleo-sediment flux that can be traced to a specific land surface. Toward this goal we compare catchment-wide average denudation rates during glacial and interglacial periods based on Quaternary fluvial deposits from three river systems that drain the Texas interior and enter the Gulf of Mexico. These rivers drain a large [∼3 × 105 km2 ] area of low-relief and a gently-sloping landscape. This type of landscape has recently been suggested as a significant contributor to global sediment flux (Willenbring et al., 2013) [although not the dominant contributor as initially

210

A.J. Hidy et al. / Earth and Planetary Science Letters 390 (2014) 209–221

suggested (J.K. Willenbring, 2014)], and therefore the erosional history of these rivers is broadly relevant to global sediment budgets. The river catchments have maintained an approximately constant drainage area, have never been glaciated, and drain a passive margin with minimal tectonic influence over the duration of interest. The aim of this study is to determine how denudation rates in these regions respond to glacial and interglacial climate. We obtain our catchment-wide denudation rates using in situproduced 10 Be in fluvial sediment (Bierman and Steig, 1996; Brown et al., 1995; Granger et al., 1996). This approach can be used to determine both modern and past denudation rates (e.g. Charreau et al., 2011; Fuller et al., 2009; Matmon et al., 2012; Refsnider, 2010; Schaller et al., 2002). Our denudation rates are integrated over ∼15–45 ka, a sufficiently short period to capture potential changes in rates due to the eccentricity-driven 100 ka climate cycle, but sufficiently long to minimize anthropogenic input for modern rates. We test and support the hypothesis that, where tectonic activity is minimized and glaciers are absent, long-term variation in denudation rate correlates broadly with mean catchment temperature, as inferred from the model predictions of Syvitski and Milliman (2007) and Kettner and Syvitski (2008). 2. Study area 2.1. Overview The Colorado, Brazos, and Trinity Rivers (Fig. 1) drain the majority of inland Texas [∼110,000 km2 , ∼115,000 km2 , and ∼45,000 km2 , respectively] and supply most of the sediment to the Gulf Coast. The coastal plain extends inland ∼200 km from the present Texas shoreline, and represents the landward portion of a succession of coalescing alluvial-deltaic deposits associated with Neogene progradation of the continental margin (DuBar et al., 1991; Galloway et al., 2000). Tectonic activity in source terrains for these rivers has been minimal since the Pliocene (DuBar et al., 1991). Although salt tectonics and growth faults are ubiquitous on the outer coastal plain and shelf (Diegel et al., 1995; Ewing, 1991), their effect on erosion rates of upstream source terrains is assumed to have been negligible, and any local faultinduced coastal plain incision would have contributed little to the total sediment supply. The Texas drainage basins have experienced isostatic adjustment in response to incision and sediment loading and may have been marginally affected by post-Laramide broad-wavelength epeirogenic uplift (sensu Pazzaglia and Gardner, 1994) associated with the Rocky Mountains (McMillan et al., 2002, 2006). However, changes in neither of these effects would have significantly affected sedimentation rates over tens of thousands of years, the period over which our 10 Be denudation rates have been integrated. Furthermore, post-depositional tilt measurements of the Miocene Ogallala Group bordering the Rocky Mountains indicates a minimal change in slope at distances >200 km from the foothills (McMillan et al., 2002), where the catchment headwaters reside. Long-term regional tectonic stability, a virtually continuous record of sedimentation along the coastal plain, and a pre-existing geochronological framework [discussed below] make the Texas coastal plain a suitable setting for studying the effect of long-term climate change on sediment supply. The drainage basins are dominated by flat-lying siliciclastic rocks, with modern river sediment consisting almost exclusively of mature, clean, and quartz-rich sand. Exceptions to this include Cretaceous carbonates that crop out in the Edwards Plateau province of west-central Texas, within the Colorado and Brazos river catchments, and exhumed Precambrian igneous and metamorphic rocks [Llano region] in the Colorado catchment. The Colorado and Brazos rivers were sourced in the southern Rocky Mountains until their headwaters were captured by the Rio Grande sys-

tem in the Pliocene (Galloway et al., 2011; Gustavson and Finley, 1985), but their catchments have maintained roughly the same configuration since then (Galloway et al., 2011; Reeves, 1976; Winker, 1979). Thus, we assume that modern river drainage areas are similar to the catchment areas over the Quaternary. The study area has experienced climate-driven variations in temperature, precipitation, and vegetation cover that likely influenced these river systems. Pollen, fossil vertebrate, and plant macrofossil data (Bryant and Holloway, 1985; Hall and Valastro, 1995; Toomey III et al., 1993), atmospheric noble gases in ground water (Stute et al., 1992), and speleothem growth rates (Musgrove et al., 2001) have all been used to locally reconstruct glacial climate in central Texas, and collectively indicate that glacial periods were cooler [∼5 ◦ C] and wetter than present-day interglacial conditions. 2.2. Stratigraphy and chronology of the Texas coastal plain Within the inner coastal plain upstream of the hinge zones (Fig. 1), denoted by the transition between long-term net erosion and net aggradation, the rivers flow through mixed bedrockalluvial valleys (Blum et al., 2013; Blum and Aslan, 2006; sensu Howard et al., 1994), with flights of down-stepping terraces that consist of mostly sandy point-bar deposits. Downstream, on the outer coastal plain, each river emerges onto aggradational alluvial plains where channel-belt sand is flanked by thick flood-plain muds with numerous paleosols formed during successive Quaternary interglacial sea-level highstands (Blum and Price, 1998). At a larger scale, alluvial plains from the individual rivers have coalesced to create composite and nearly seamless topographic surfaces. In the shallow subsurface, beneath aggradational highstand deposits, channel-belt sands represent glacial periods of lowered sea level, when rivers were extended across the emergent shelf and discharged to deltas at the shelf margin (Blum and HattierWomack, 2009; Blum and Törnqvist, 2000). Pleistocene alluvial plains and their associated subsurface glacial-period channel belts are identified as the Early Pleistocene Lissie and Middle to Late Pleistocene Beaumont Formations (after Doering, 1935) (Fig. 1). Channel-belt deposits from the last glacial period are informally referred to as the Deweyville units (Blum et al., 1995), and post-Deweyville strata represent the post-glacial period of sea-level rise and the present sea-level highstand. Pleistocene and Holocene fluvial deposits on the outer coastal plain are of primary interest to this study because (1) the depositional record is well preserved and mapped, (2) an independent chronostratigraphic framework exists (Blum and Price, 1998; Blum and Valastro, 1994; Garvin, 2008), and (3) the glacial or interglacial context is known for targeted Beaumont and postBeaumont [Deweyville and Holocene] deposits based on mapping of stratigraphic relationships from cores and cutbank outcrops, soil development, topographic projection of surfaces, and supporting geochronological data (see Blum and Aslan, 2006 and references therein). Due to the composite nature of the coastal-plain deposits and lack of topographic expression, resolution of regional depositional ages beyond the glacial–interglacial cycles cannot be inferred from an age obtained elsewhere within a mapped unit. Thus, all sampling sites used for paleo-denudation rate measurements require chronological analysis (Table 1). 2.2.1. Lissie Formation Precise age control for the Lissie Formation has not yet been obtained. Kukla and Opdyke (1972) measured reversed magnetic polarity for Lissie floodplain deposits, and suggested deposition during the Matayama [0.79–2.5 Ma] polarity chron. Upstream in the Colorado River catchment, the Lava Creek “B” ash (∼640 ka; Lanphere et al., 2002) was observed in deposits

A.J. Hidy et al. / Earth and Planetary Science Letters 390 (2014) 209–221

211

Fig. 1. Shaded relief map of study area with outlines of the Brazos, Colorado, and Trinity catchments. Inset geologic maps detail relevant coastal plain units within the Colorado, Brazos (a) and Trinity (b) catchments. Cross section A–A , acquired perpendicular to the Texas coastline, shows a general increase in dip of coastal plain deposits with age. A sharp increase in gradient of the Willis surfaces has formed cuestas that provide a significant topographic break defining the boundary between the inner and outer coastal plains. Geologic maps were acquired from compiled Bureau of Economic Geology quadrangles (Aronow and Barnes, 1982; Proctor et al., 1974), with additional detail for Deweyville deposits in the Trinity catchment after Garvin (2008) and for Lissie deposits in the Colorado catchment after Winker (1979) and Blum and Valastro (1994). Deweyville-equivalent deposits in the Colorado and Brazos catchments are not shown on the map due to complete onlapping of the Holocene floodplain. Sample locations are shown on inset graphs and referenced in Table 1.

Table 1 Summary of chronology for the units targeted for sampling for this study. Location of sites is indicated on Fig. 1. OSL = optically stimulated luminescence; TCN = terrestrial cosmogenic nuclide; TL = thermoluminescence. Ages are approximated for modern pre-dam samples. To ensure sampling of material that was older, but not significantly older, than the earliest dams [∼70 a] samples were acquired from sites in modern scroll bars that (1) were ∼100 m inland from the active point bar along the scroll axis, and (2) exhibited incomplete, yet extensive (moderate) vegetation growth. In contrast, modern post-dam samples were collected from the active point bar. Surface

Site #

Estimated age (ka)

Method of chronology

Source

Brazos modern post-dam

1

0.0

Inferred; active point bar

this work

Colorado modern post-dam modern pre-dam Holocene (CBA-1) Deweyville Beaumont (U) Beaumont (L) Lissie

6 8 5 9 7 2 3

0.0 0.1 ± 0 .1 8.5 ± 3.5 35 ± 2 119 ± 9.5 155 ± 15 560 +190 /−130

Inferred; active point bar Inferred; modern scroll bar with moderately developed vegetation 14 C; organic detritus in fluvial sediment TL; quartz TL; quartz TL; quartz TCN depth profile; fluvial sand

this work this work Blum and Durbin et Blum and Blum and this work

Lissie

4

560 +190 /−130

TCN single shielded sample; fluvial sand

this work

12 14 15 11 11 10 13

0. 1 ± 0 . 1 20 ± 1 20 ± 1 25 ± 5 29 ± 5 33 ± 2 >520 ± 150 590 +110 /−310

Inferred; modern scroll bar with moderately developed vegetation OSL; single-grain aliquot on quartz OSL; single-grain aliquot on quartz OSL; single-grain aliquot on quartz OSL; single-grain aliquot on quartz OSL; single-grain aliquot on quartz OSL; single-grain aliquot on quartz TCN depth profile; fluvial sand

this work Garvin (2008) Garvin (2008) Garvin (2008) Garvin (2008) Garvin (2008) this work this work

Trinity modern pre-dam L. Deweyville L. Deweyville M. Deweyville (U) M. Deweyville (L) H. Deweyville Lissie

Valastro (1994) al. (1997) Price (1998) Price (1998)

212

A.J. Hidy et al. / Earth and Planetary Science Letters 390 (2014) 209–221

underlying a dissected terrace (Mandel and Caran, 1992), projected to an elevation between the Lissie and Beaumont surfaces (Blum and Valastro, 1994), and therefore interpreted to represent cut Lissie deposits. Downstream, the slope of the Lissie surface on the outer coastal plain has been projected to seismic reflections offshore in the well constrained marine record of the Gulf of Mexico (DuBar et al., 1991; Winker, 1979). The first downhole appearance of the foraminifer Trimosina denticulata, dated at ca. 600 ka by Armentrout and Clement (1990), appears just above this projected Lissie surface. Thus, the data do not precisely constrain Lissie deposition (Blum and Aslan, 2006; Blum and Price, 1998), but converge on an Early to Middle Pleistocene age. 2.2.2. Beaumont Formation and post-Beaumont deposits More precise chronological control exists for Beaumont and post-Beaumont strata. Blum and Price (1998) and Blum and Aslan (2006) used cross-cutting stratigraphic relations and thermoluminescence [TL] methods to differentiate three valley fills that formed during the Middle to Late Pleistocene [oxygen isotope stages (hereafter OIS) 10–5] within Beaumont alluvial plains of the Colorado River. The youngest [Bay City] and oldest [Lolita] valley fills yielded TL ages of 115–155 ka, and 307–350 ka, respectively, and stratigraphically bracket the undated El Campo fill. Each valley fill was interpreted to represent a 100 ka glacio-eustatic cycle, where channels incised basinward during relative sea-level fall, forming a succession of downward-stepping channel-belt sand bodies, and then back-stepped and aggraded during relative sea-level rise. Glacial [OIS 6] and interglacial [OIS 5] deposits of the Bay City fill are wellexposed along the lower reaches of the modern Colorado River. Post-Beaumont strata reside within valley fills that cross-cut Beaumont alluvial plains, but represent a similar pattern of incision and cross-shelf extension followed by back-stepping and aggradation related to relative sea-level. Blum et al. (1995) differentiated three Deweyville units and deposits from the last glacial period [OIS 4–2] based on stratigraphic relations, with the youngest suite constrained by minimum 14 C ages of 23–16 ka. Aggradational post-Deweyville strata are assigned to the latest Pleistocene and Holocene on the basis of numerous 14 C ages (Blum and Valastro, 1994). From the lower Nueces River, Durbin et al. (1997) reported TL ages of 53–31 ka for a suite of Deweyville deposits and 92–72 ka for the youngest Beaumont deposits. In the Trinity River, Garvin (2008) obtained optically stimulated luminescence [OSL] ages for Deweyville unit in the lower Trinity valley and inferred depositional ages of 65–33 ka, 33–23 ka, and 23–16 ka for three High, Middle, and Low Deweyville fills. The existing chronological controls suggest that Deweyville deposition represents OIS 4–2, and the post-Deweyville fill represents the last 15 ka, with all post-Deweyville deposits exposed along modern rivers being of Holocene age (Blum and Valastro, 1994).

∼220 m3 s−1 for the Brazos, Colorado, and Trinity rivers, respectively, over the same period. The measurements represent the best minimum estimates of sediment fluxes and river discharges from the unobstructed rivers. However, because of extensive irrigation and agricultural development, particularly in the Southern High Plains areas of the Colorado and Brazos catchments, these modern rates may include a significant anthropogenic component. More recently, modern sediment flux estimates of 27 Mt a−1 [Brazos], 13.3 Mt a−1 [Colorado], and 5.5 Mt a−1 [Trinity] were calculated by Blum and Hattier-Womack (2009), and Garvin (2008), respectively, by using the BQART sediment-flux model of Syvitski and Milliman (2007):

Q s = 0.0006B Q 0.31 A 0.5 R T

for T  2 ◦ C,

(1)

and

Q = 0.075 A 0.8 ,

(2)

where Q s is sediment flux [Mt a−1 ], B is a geologic factor determined by catchment lithology, human influence, glacial coverage, and trapping efficiency of lakes and reservoirs, Q is the mean annual water discharge [km3 s−1 ], A is the area of the catchment [km2 ], R is the relief of the catchment [km], and T is the average catchment-wide temperature [◦ C]. This empirical sedimentflux model has been trained with a global database of 488 rivers that explains 96% of the variance in suspended-sediment yields. As noted by Blum and Hattier-Womack (2009), the BQART model suggests that changes in sediment flux through time are proportional to temperature changes, if relief, area, and the geologic factors that contribute to B are constant. With respect to pre-historic sediment supply, Anderson (2005) suggested, based on volumes of sediment in shelf-margin deltas of the Brazos and Colorado rivers, that supply was significantly greater during the last glacial period. However, using the BQART model, Blum and Hattier-Womack (2009) calculated that sediment delivered to the Gulf of Mexico by the Brazos, Colorado, and Trinity rivers during the last glacial maximum was ∼25–30% less than modern [pre-dam] values, and higher supplies were not required to explain volumes in shelf-margin deltas (see also Garvin, 2008). Prather et al. (2012) and Pirmez et al. (2012) subsequently used sediment volumes in the Brazos–Trinity shelf-margin deltas and downdip slope minibasins to suggest that glacial sediment supply was lower than today, supporting the calculations of Blum and Hattier-Womack (2009). Nevertheless, these results contrast with interpretations of climate-induced sediment flux changes elsewhere, and demonstrate the need for actual measurements of sediment flux that are independent of records based on deposit mass. 3. Methods 3.1.

10

Be-derived paleo-denudation rates

2.3. Sediment flux from major Texas rivers The major rivers that drain Texas have been dammed since the 1930s, significantly reducing sediment delivery to the coastal plain. Estimates of modern sediment flux are therefore restricted to pre-dam deposits or derived from numerical models (Table 2). Prior to and during the construction of dams on the Brazos, Colorado, and Trinity rivers, suspended-sediment measurements obtained by Bloodgood and Meador (1941) and Bloodgood (1955) from gauging stations on the coastal plain yielded time-averaged mass fluxes of 31 Mt a−1 , 8.1 Mt a−1 , and 5.5 Mt a−1 , respectively. For the Colorado, Kanes (1970) obtained a mean flux of 13 Mt a−1 by recompiling these data and including estimates of flux during years absent from the main-stem record. These fluxes accompanied mean river discharges of ∼200 m3 s−1 , ∼120 m3 s−1 , and

Cosmogenic 10 Be was measured in quartz from fluvial deposits on the Texas coastal plain to estimate spatially-averaged [catchment-wide] denudation rates. For modern rates, the measured 10 Be concentration in recent sediment is inversely proportional to the catchment’s average denudation rate (Bierman and Steig, 1996; Brown et al., 1995; Granger et al., 1996; von Blanckenburg, 2005). This is because faster surface erosion exhumes deeper regolith with lower concentrations due to the absorption of the cosmic ray flux through mass. Estimating catchment denudation rates from fluvial sediment requires: (1) an average 10 Be production rate for the catchment, which is weighted by catchment hypsometry and the distribution of quartz in the catchment upstream from the sample site (e.g. Safran et al., 2005), and (2) the 10 Be concentration in the sediment at the time of deposition, which,

A.J. Hidy et al. / Earth and Planetary Science Letters 390 (2014) 209–221

213

Table 2 Summary of sediment flux measurements from the Brazos, Colorado, and Trinity Rivers acquired from suspended sediment measurements, BQART model estimates, and estimates from this work. Source

Sediment flux (Mt a−1 )

10

Be

Recorded dates

Total integrated time

Method of measurement

Location

Type

1923–1941 1941–1954 1923–1954

17.3 a 13.0 a 30.3 a

suspended sediment suspended sediment suspended sediment

Richmond, station 1 Richmond, station 1 Richmond, station 1

pre-dam post-dam mixed

Brazos Bloodgood (1955) 37.1 22.3 30.8

Blum and Hattier-Womack (2009) after Syvitski and Milliman (2007) 26.5

NA

NA

modified BQART model

whole catchment

pre-dam

7.67

NA

30 ka

TCN catchment-wide denudation rate

whole catchment

post-dam

1929–1933 1937–1940 1929–1940

3.2 a 2.7 a 5.9 a

suspended sediment suspended sediment suspended sediment

Columbus, station 18 Columbus, station 18 Columbus, station 18

pre-dam post-dam mixed

5a 10 a 15 a

suspended sediment with interpolation suspended sediment with interpolation suspended sediment with interpolation

Columbus, station 18 Columbus, station 18 Columbus, station 18

pre-dam post-dam mixed

This work

Colorado Bloodgood and Meador (1941) 9.82 6.08 8.09

Kanes (1970) after Bloodgood and Meador (1941) 17.7 10.7 12.7

1931–1935 1936–1945 1931–1945

Blum and Hattier-Womack (2009) after Syvitski and Milliman (2007) 13.3

NA

NA

modified BQART model

whole catchment

pre-dam

6.49 6.59 6.07

NA NA NA

29 ka 28 ka 33 ka

TCN catchment-wide denudation rate TCN catchment-wide denudation rate TCN catchment-wide denudation rate

whole catchment whole catchment whole catchment

pre-dam pre-dam post-dam

1935–1948 1948–1954 1935–1954

12.1 a 6.00 a 18.1 a

suspended sediment suspended sediment suspended sediment

Romayor, station 48 Romayor, station 48 Romayor, station 48

pre-dam post-dam mixed

This work

Trinity Bloodgood (1955) 6.97 2.58 5.52

Garvin (2008) after Syvitski and Milliman (2007) 5.50

NA

NA

modified BQART model

whole catchment

pre-dam

4.53 4.75

NA NA

17 ka 17 ka

TCN catchment-wide denudation rate TCN catchment-wide denudation rate

whole catchment whole catchment

pre-dam pre-dam

This work

for older deposits, requires correction for production and decay of 10 Be. In older deposits, post-depositional 10 Be production becomes increasingly significant and must be subtracted from the measured concentration (Schaller et al., 2002). Because 10 Be production is depth-sensitive, modeling post-depositional 10 Be production requires knowledge of depositional age and the surface erosion or aggradation that affected the target deposit, as well as the depositional ages and erosion or aggradation history of all overlying deposits. Thus, each catchment-wide estimate of average denudation rate based on older deposits requires a sample-specific model constrained by absolute chronology and stratigraphic measurements (S.1). Paleo-denudation rate estimates assume that the response time of the 10 Be signal to re-equilibrate to changes in landscape erosion rates is short compared with the driving frequency of glacial– interglacial change. Models of the response time of 10 Be-derived denudation rates to episodic landscape denudation suggest that, for rapid denudation rates [5–15 cm ka−1 ], TCN concentrations in eroded sediment adjust within 10–20 ka (Bierman and Steig, 1996; Schaller and Ehlers, 2006). This response time is inversely proportional to the denudation rate, and buffers denudation-rate variation by causing an overestimate for low denudation rates and an underestimate of high denudation rates. Because denudation rates in this study are on the order of 1–4 cm ka−1 , the response time is slightly longer [∼25 ka]. However, a 25 ka response time is rapid

enough to examine denudation rates for glacial and interglacial periods which, during the Late Pleistocene, are driven primarily by a ∼100 ka orbital cycle (Pisias and Moore, 1981). In this study, pre-existing chronology was utilized for most sites (Table 1); for other sites, 10 Be depth profiles were modeled to obtain new chronology. Profiles of 10 Be through a column of sediment can be used to simultaneously model exposure age, surficial erosion rate, and depositional 10 Be concentration [i.e. the 10 Be present at time of deposition, prior to post-depositional exposure and decay] for a given sediment package (Anderson et al., 1996; Braucher et al., 2009; Hidy et al., 2010). New exposure ages were obtained from the Hidy et al. (2010) depth-profile calculator (see S.2). In the Trinity catchment, an OSL sample was acquired in conjunction with a 10 Be depth profile (see S.3). 3.2. Sediment recycling with 26 Al and 10 Be Previous studies suggest that floodplain storage reintroduced into the river via bank erosion has a negligible effect on catchmentwide denudation rates as measured with long-lived TCN such as 10 Be and 26 Al (Wittmann and von Blanckenburg, 2009; Wittmann et al., 2009). This is because flux derived from erosion of the upstream drainage basin is an order of magnitude greater than that which can be derived from erosion within the valley and channel belt from bank erosion (Blum et al., 2013; Blum and Törnqvist,

214

A.J. Hidy et al. / Earth and Planetary Science Letters 390 (2014) 209–221

2000), and because much of the material eroded from the downstream floodplain is finer than the medium- to coarse-grained material targeted for 10 Be measurements. However, measuring both 26 Al and 10 Be can provide a check on whether a portion of the sediment experienced significant burial between its exposure on hillslope surfaces and its final deposition because the 26 Al/10 Be ratio decreases predictably with burial duration (Granger and Muzikar, 2001). Additionally, if the source of the stored sediment can be determined and its average 26 Al/10 Be can be assumed, the contributions from stored [coastal plain] sediment and from surficial hinterland [up-catchment] regolith erosion may be distinguished. 26 Al and 10 Be measurements were acquired for chronologically distinct depositional intervals in order to quantify this storage effect, adjust our calculated denudation rates, and determine the relative abundance of up-catchment and stored sediment in river sediment yield. We model recycled sediment storage using a binary mixing algorithm after Wittmann et al. (2011b)

Ald

=

Bed

Alnb · (1 − f b ) + Alb · f b Benb · (1 − f b ) + Beb · f b

(3)

,

where the subscripts d, nb, and b describe isotopic concentrations [atoms g−1 a−1 ] that are depositional, from sources that were not buried [i.e., up-catchment], and from sources that experienced burial, respectively, and f b is the fraction of total sediment from the buried source. In order to remove the buried component from our depositional 10 Be concentrations [Bed ], we solve for Benb using

Alnb

= 6.75,

Benb Alb Beb

(4)

in order to obtain a total time-averaged sediment flux. Furthermore, this type of storage estimate relates only to material that has been stored at a depth >3–4 m such that it has been shielded significantly from the secondary cosmic radiation responsible for 26 Al and 10 Be production. Fortunately, this is the approximate minimum depth where significant medium- to coarse-grained sand is found in point-bar packages throughout the Texas coastal plain, and we only used medium- to coarse-grained sand in our analysis. In many river systems, including low-relief drainages such as the Texas catchments, the sediment source can vary over time; thus, we propose that the multiple-isotope approach described above may be suitable and perhaps necessary for many TCN denudation rate studies. 3.3. Sample collection TCN samples consisted of ∼2 kg of medium- to coarse-grained sand and were collected from point-bar facies within active or abandoned channel deposits on the coastal plain (see S.4). The samples were washed and sieved to extract the 355–500 μm size fraction, with additional fractions utilized in samples lacking adequate mass in this range. 10 Be and 26 Al chemical preparation and extraction were performed at the Dalhousie Geochronology Centre (see http://cnef.earthsciences.dal.ca), and AMS measurements for both isotopes were made at the Center for Accelerator Mass Spectrometry at Lawrence Livermore National Laboratory [CAMSLLNL]. 4. Results 4.1. Depositional 10 Be and 26 Al

= Rb,

(5)

and

fb =

6.75 −

Ald Bed

(6)

6.75

where up-catchment 26 Al/10 Be is assumed to be at the production ratio because surface exposure interrupted by burial is less likely (see S.1.3), the buried sediment has some average ratio R b (assumed to be 3 for this study based on measurements of pre-Lissie storage; see Table S.1), and the buried fraction is weighted by the difference between the depositional ratio and the production ratio. Finally, we assume the average buried 10 Be concentration is

Beb = Beav · exp(−λBe · t )

(7) 10

where Beav is estimated as the average depositional Be observed in each catchment [atoms g−1 a−1 ], and t is the time [a] taken for 26 Al/10 Be to decrease from the production ratio to R b , which is calculated from Eq. (6) of Granger and Muzikar (2001). Although this assumes that the depositional concentrations in the storage are of the same order as those during the Late Pleistocene, this calculated correction is insensitive to Beav because of the long time required for the production ratio to decay to 3 [∼2.4 Ma]. Additionally, the flux of mass derived from buried sources can be estimated by assuming Benb as the depositional 10 Be concentration, and multiplying the resulting flux by the factor

fb

(1 − f b )

.

(8)

This approach estimates the instantaneous storage component at time of sample deposition. It does not represent a time-averaged flux as calculated from the up-catchment denudation rate. Thus, the storage flux signal cannot be added to the up-catchment flux

Calculated values for depositional 26 Al and 10 Be concentrations of modern and older sediment are shown in Table 3. Depositional 10 Be concentrations range from 1.3–2.9 × 105 atoms g−1 in the Colorado catchment and 0.7–1.0 × 105 atoms g−1 in the Trinity catchment. A single modern [post-dam] measurement in the Brazos catchment yielded a value of 2.1 × 105 atoms g−1 . For duplicate measurements acquired from samples of different grain size [n = 4], in all cases the coarser sample returned a lower 10 Be or 26 Al concentration, consistent with the results of previous studies where hillslope processes were a first-order erosion mechanism in the source regions (Belmont et al., 2007; Codilean et al., 2012; Matmon et al., 2003; Wittmann et al., 2011a). A pair of measurements from the Low Deweyville separated spatially by 105 km [sites 14 and 15] returned statistically identical depositional 10 Be and 26 Al concentrations (Table 3). This is consistent with recent work that suggests minimal influence of floodplain recycling on depositional TCN concentrations (Wittmann and von Blanckenburg, 2009; Wittmann et al., 2009). 4.2. Catchment-wide denudation and sediment flux Depositional 10 Be and 26 Al concentrations and uncertainties were used to calculate catchment denudation rates and equivalent sediment fluxes from the upstream catchment, sediment fluxes from storage, and the time spans over which the up-catchment rates and fluxes are integrated (Table 3; see S.1). In the Brazos catchment, the modern [post-dam] sample yielded an average denudation rate of 1.73 ± 0.06 cm ka−1 , corresponding to a sediment flux of 5.29 ± 0.18 Mt a−1 averaged over the last 35 ka. A suite of modern [pre- and post-dam] and Late Pleistocene samples were measured from the Colorado and Trinity rivers. In the Colorado catchment, denudation rates and sediment fluxes were calculated at 1.33–3.20 cm ka−1 and 3.87–9.20 Mt a−1 , with average integration times of 19–45 ka. Measurements from the Trinity catchment

Table 3 Calculated results for depositional TCN concentrations, up-catchment denudation rates, up-catchment sediment flux, and sediment from storage. Int. time refers to the integration time of the denudation rates; f b refers to the fraction of buried material in the sample. Total 1σ errors in concentrations include AMS errors, an estimated 2% uncertainty for sample preparation and, for 26 Al samples, a 5% uncertainty from ICP-MS native Al measurements. For samples in the 10 Be depth profiles, depositional 10 Be and uncertainty were acquired directly from the modified Hidy et al. (2010) model output (after Mercader et al., 2012); for all other 10 Be samples, and for all 26 Al samples, depositional concentrations were calculated as described in S.1.2. Spallogenic 10 Be and 26 Al production is based on the Stone (2000) after Lal (1991) scaling scheme with a 10 Be reference production rate of 4.76 atoms g−1 a−1 (Stone, 2000, recalibrated according to Nishiizumi et al., 2007), a spallogenic 26 Al/10 Be ratio of 6.75 (Balco et al., 2008), and a mid-latitude neutron attenuation length of 160 g cm−2 (Gosse and Phillips, 2001). Muon production rates were reduced to half those of Heisinger et al. (2002a, 2002b) since these rates significantly overestimate deep muon production (Balco et al., 2005; Braucher et al., 2003, 2011; Hidy et al., 2013). This is supported by preliminary data from the Beacon Heights bedrock core in Antarctica that indicate Heisinger et al. (2002a, 2002b) systematically overestimate muon production by a factor of ∼2 (pers. comm. J. Stone). Calculated concentrations do not include the unknown systematic error in post-depositional [primarily muon] production rates. Grain size (μm)

Dep. [10 Be] (105 atoms g−1 )

Dep. [26 Al] (105 atoms g−1 )

Dep.

[10 Be]

±1σ

[26 Al]

±1σ

26

Al/10 Be

26

Al/10 Be

±1σ

Brazos modern post-dam

355–500

2.14

0.06

11.66

0.70

5.44

0.36

Colorado modern post-dam modern pre-dam modern pre-dam CBA-1 Deweyville Beaumont (U) Beaumont (U) Beaumont (L) Lissie

355–500 355–500 355–500 355–500 355–500 250–355 355–500 355–500 355–500

2.52 2.23 2.20 2.58 2.19 2.94 2.26 2.62 1.74

0.08 0.06 0.07 0.09 0.05 0.04 0.06 0.09

14.89 13.88

1.00 0.84

5.90 6.22

0.44 0.41

12.72 11.69 16.18 13.26 12.12

0.76 0.71 0.72 0.94 0.77

4.94 5.33 5.50 5.87 4.63

0.34 0.35 0.26 0.45 0.33

1.88

+0.07 / −0.11

Lissie

355–500

+0.23 / −0.18

Lissie

500–850

1.33

+0.24 / −0.18

Trinity modern pre-dam modern pre-dam L. Deweyville L. Deweyville M. Deweyville (U) M. Deweyville (U) M. Deweyville (L) M. Deweyville (L) H. Deweyville H. Deweyville Lissie

250–355 250–355 355–500 355–500 150–250 250–355 150–250 150–250 250–355 250–355 250–500

0.84 0.80 0.95 0.99 0.73 0.70 0.83 0.88 0.98 0.94 1.02

0.03 0.02 0.03 0.03 0.04 0.05 0.04 0.03 0.05 0.04

+0.02 / −0.04

7.40

5.13

Be denudation rate (cm ka−1 )

raw

corr

±1σ

0.19

2.02

1.73

0.06

0.13 0.08 0.08 0.27 0.21 0.18 0.13 0.31 0.17

1.85 2.10 2.13 1.81 2.12 1.57 2.06 1.76 2.69

1.67 1.98 2.01 1.41 1.79 1.33 1.86 1.32 2.38

0.17 +2.40 / −1.96

10

fb

2.49

corr

±1σ

storage

35

6.18

5.29

0.18

1.49

0.06 0.06 0.06 0.06 0.05 0.02 0.06 0.06

+0.31 / −0.32

36 30 30 43 34 45 33 46 25

5.31 5.98 6.07 5.20 6.14 4.56 5.96 5.15 7.75

4.80 5.65 5.73 4.07 5.18 3.87 5.38 3.85 6.85

0.17 0.16 0.18 0.18 0.15 0.06 0.17 0.17

+0.90 / −0.92

0.76 0.51 0.52 1.91 1.63 1.03 0.89 2.35 1.62

2.19

+0.16 / −0.09

28

7.16

6.30

+0.45 / −0.26

1.50

19

10.19

9.20

+1.61 / −1.57

2.13

19 18 25 26 15 15 20 22 20 20 34

4.11 4.31 3.50 3.53 4.79 4.96 4.17 3.96 3.55 3.70 3.32

3.83 4.03 2.80 2.82 NA NA 3.54 3.32 3.55 3.70 2.04

0.13 0.12 0.11 0.10 0.26 0.32 0.20 0.13 0.18 0.16

0.42 0.44 1.20 1.16 NA NA 1.13 1.07 0.00 0.00 0.64

5.58

0.17

3.54

3.20

+0.56 / −0.55

0.40

6.12

0.51

4.77 5.00

0.30 0.34

5.03 5.07

0.35 0.38

0.09 0.09 0.26 0.25

4.43

0.28

5.31

0.42

7.25

0.46

7.40

0.60

+1.10 / −3.50

5.66

+1.08 / −3.50

3.23 3.40 2.46 2.35 NA NA 2.95 2.77 2.96 3.08 1.75

0.11 0.10 0.09 0.09 0.22 0.27 0.17 0.11 0.15 0.13

5.77

3.47 3.64 3.08 2.94 3.99 4.13 3.48 3.29 2.96 3.08 2.85

0.16

Be sediment flux (Mt a−1 )

raw

+2.07 / −1.66

0.21 0.21

10

Int. time (ka)

+0.12 / −0.06

+0.14 / −0.06

A.J. Hidy et al. / Earth and Planetary Science Letters 390 (2014) 209–221

Surface

215

216

A.J. Hidy et al. / Earth and Planetary Science Letters 390 (2014) 209–221

Fig. 2. 26 Al/10 Be burial plots showing depositional 10 Be and 26 Al data for (A) Colorado [red] and Brazos [blue], and (B) Trinity [green] catchments. Material derived from vertical erosion of catchment surfaces should plot within the “erosion banana” defined by the shaded grey area. Material derived from Plio-Pleistocene storage should plot in the shaded yellow region; this region represents TCN concentrations from Plio-Pleistocene sources assuming (1) depositional concentrations equivalent to a catchment-wide denudation rate >0.5 cm ka−1 , and (2) a storage depth of <60 m, which is a conservative estimate for maximum relief of inner coastal plain cuestas. Depositional concentrations plotting between these regions indicate a mixture of storage and up-catchment material. 10 Be denudation rates are corrected for storage by assuming binary mixing of Plio-Pleistocene storage [mean value indicated by a black square] and up-catchment surface material (see S.1.3). TCN concentrations from pre-Lissie sediment at site 13 are indicated by a black dot with error ellipse [mean value, n = 2] and lie on the secular equilibrium curve [solid brown line], which represents the saturation ratio for sediment at a given depth assuming a density of 2 g cm−3 [ratio inflection at ∼4 m due to muon-dominated production]. Post-Lissie concentrations are indicated by dots and error ellipses; most probable values for Lissie concentrations are indicated by squares. All ellipses represent 1σ error. Burial contours [dotted gray lines] denote minimum burial age of the stored sediment component. [For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.]

yielded denudation rates and sediment fluxes of 1.75–3.40 cm ka−1 and 2.04–4.03 Mt a−1 , with average integration times of 18–34 ka. The errors listed in Table 3 for these values do not include the large uncertainty in the scaling scheme used to calculate site production rates, which may be up to ∼10% [1σ ] due to uncertainty in the paleomagnetic field and reference production rate (Balco et al., 2008). However, due to the long integration time of the denudation rates [15–45 ka], time-varying TCN production due to changes in Earth’s magnetic field will not differ significantly among samples. Thus, the production-rate uncertainty is considered systematic for samples within the same catchment and, importantly, only affects the magnitude of the calculated rates, not their relative values. 5. Discussion 5.1. Sediment recycling on the Texas coastal plain The catchment-wide denudation rates are high enough to suggest that the depositional 26 Al/10 Be ratio should approximately equal the spallogenic production ratio [6.75]. However, with the sole exception of the High Deweyville, all measured 26 Al/10 Be ratios are significantly and variably depressed (Fig. 2). This indicates that remobilization of previously stored sediment added significantly to the up-catchment sediment flux, and that its relative contribution fluctuated during the Late Pleistocene. For both the Colorado and Trinity rivers, depositional 26 Al/10 Be was higher for units deposited during relative sea-level high-stands than for units deposited during relative sea-level fall. A comparison of our 26 Al/10 Be data with a relative sea-level time series shows a consistent correlation between sea-level and depositional 26 Al/10 Be (Fig. 3), implying that falling sea level enhanced erosion of sediment in long-term [∼105 a] storage. Stored sediment sources probably exclude much of the outer coastal plain for several reasons. Firstly, within the lower Trinity Valley, similar depositional 26 Al/10 Be ratios were measured in the youngest Deweyville along ∼100 km of the valley, agreeing with

Fig. 3. Calculated depositional 26 Al/10 Be ratios plotted against age for all post-Lissie Brazos, Colorado, and Trinity samples. Ratios calculated for Lissie deposits are not shown since depositional age uncertainties preclude assigning a sea-level context. Black curve indicates relative sea level over the last 200 ka after Miller et al. (2005). Based on the range of 10 Be catchment-wide denudation rates, the expected depositional ratio is ∼6.75 or slightly higher (see erosion banana in Fig. 3) and is indicated by the grey shaded region. Deposits containing material eroded solely from upcatchment surfaces should plot in this region. Lower ratios indicate an increasing contribution of stored sediment in the sampled deposit.

studies that suggest a negligible influence of floodplain erosion on 10 Be denudation measurements (Wittmann and von Blanckenburg, 2009; Wittmann et al., 2009). If erosion of Beaumont and post-Beaumont sediment [comprising most of the outer coastal plain] caused the lowered 26 Al/10 Be, the ratio should have decreased systematically downstream. Secondly, the minimum burial time [neglecting post-depositional production] needed to depress 26 Al/10 Be from the production ratio [6.75] to the measured values [∼5] is >600 ka (see Fig. 2). The bulk of the outer coastal plain, however, is considerably younger than this [<350 ka], and erosion of its stored sediment could not have lowered the ratio sufficiently. Furthermore, 26 Al/10 Be for the eroded sediment should have been significantly lower than the observed ratios to

A.J. Hidy et al. / Earth and Planetary Science Letters 390 (2014) 209–221

217

Fig. 4. Calculated 10 Be denudation rates vs. time from (A) raw depositional 10 Be concentrations from the Colorado catchment, (B) 10 Be concentrations from the Colorado catchment corrected using depositional 26 Al/10 Be ratio, (C) raw depositional 10 Be concentrations from the Trinity catchment, and (D) 10 Be concentrations from the Trinity catchment corrected using depositional 26 Al/10 Be ratio. Even-numbered oxygen isotope stages are indicated by yellow shaded rectangles; the global relative temperature curve, based on the Lisieki and Raymo (2005) LR04 benthic stack, is indicated by the solid blue line. Shaded gray rectangles denote integration intervals of individual denudation rate measurements. [For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.]

offset higher up-catchment values, unless the latter component was negligible. Thirdly, the depositional 26 Al/10 Be within highstand deposits is relatively high, so cannibalizing a previous highstand prism during relative sea-level fall would be insufficient to depress the ratio. Finally, the outer coastal plain is dominated by material finer than that used for this study [150–850 μm] so such modification of the ratio would have required erosion of a very large volume of storage. The most likely source of stored sediment is from Pre-Pleistocene formations above the hinge zone in the inner coastal plain, or from the Lissie Formation. These formations are of the right depositional age such that their 26 Al/10 Be has decayed significantly, but not completely. Additionally, they are rich in medium- to coarsegrained sand targeted for sampling, particularly the Pliocene Willis Formation (Doering, 1935; Heinrich and McCulloh, 2000). They also dip more steeply than other younger coastal-plain deposits, with parts of the Willis forming heavily dissected cuestas that grade into the hinge zone and provide access to deeply buried sediment (Winker, 1979). An important implication of this recycling is that, during times of lower sea-level, uncorrected 10 Be-derived sediment flux from catchments may be significantly overestimated [by ∼10–30%]. Since 10 Be in buried Plio-Pleistocene sediment has decayed substantially since deposition, its concentration should be much lower than that of sediment from up-catchment. Thus, measurements from sediments with components of reworked sediment represent maximum estimates of sediment flux, and the effect cannot be addressed without knowing the precise burial ages and proportions of the sediment sources. An exception would be if the

Plio-Pleistocene sediments were laid down during times when catchment-wide denudation rates were several orders of magnitude lower [i.e., TCN concentrations were higher] so that, even after substantial decay, the 10 Be concentrations in the stored sediment would have been greater than those from up-catchment. However, this is unlikely because depositional concentrations calculated from the Lissie Formation are similar to, or less than, those from younger deposits. Furthermore, 10 Be measurements from undated Willis sediment stratigraphically below the Lissie profile at site 13 (sample IDs 2257 and 2258 in S.4) are an order of magnitude less than all other calculated depositional concentrations. 5.2.

10

Be catchment-wide denudation rates

5.2.1. Comparison with modern sediment fluxes In order to provide context for our paleo-denudation rate record, robust modern rates were needed. This required circumventing the potential impact of modern dams on 10 Be-denudation rates by targeting pre-dam sediment. These rates are then converted to sediment fluxes to compare with modern sediment trap data. Our sediment flux estimates from modern pre-dam sediments are reproducible, but are lower than minimum estimates based on pre-dam sediment traps (Kanes, 1970; LeBlanc and Hodgson, 1959) while in the same relative order for the three river systems. Our Trinity catchment data compare favorably with previous data (see Table 2), but our Colorado and Brazos data are significantly lower than previous estimates. Since 10 Be-based sediment flux values are related to surface exhumation, they represent an integrated sediment flux independent of transport style

218

A.J. Hidy et al. / Earth and Planetary Science Letters 390 (2014) 209–221

[i.e., dissolved, suspended, and bed load]. However, this apparent discrepancy with the sediment trap data may be because: (1) 10 Be rates are integrated over tens of thousands of years whereas data from sediment traps represent very short periods (Kirchner et al., 2001), (2) the sediment-trap data, although acquired prior to dam construction, were measured after the onset of late 19th and early 20th century agriculture with the probability of anthropogenicallyenhanced erosion rates, and (3) the 10 Be sediment flux rates are estimated from medium- to coarse-grained sand [mostly bed load], which may represent catchment areas that are very different for suspended load grain sizes. Although most of our modern samples were acquired from deposits that predate dam construction, we also obtained a single sample from an active Colorado River point bar to test the sensitivity of up-catchment sediment flux rates based on 10 Be measurements to dam construction and sediment trapping. The post-dam sediment flux was estimated to be ∼15% lower than the pre-dam flux (Table 2), and well outside the standard deviation of multiple pre-dam samples, suggesting that the ∼70 yr existence of these dams has had a significant effect on 10 Be concentration downstream. Since these dams effectively shut down the up-catchment supply of the coarser bed load used for this study, we suspect that the lower apparent sediment flux reflects increased reworking of stored sediment in the active bed downstream from the dam. This is consistent with general downstream scouring that occurs following dam construction (Petts, 1979) and with extensive dam-induced scouring observed below the lowest dam on the Trinity River (Phillips et al., 2005). The post-dam sediment has a 10 Be concentration higher than the pre-dam sediment, but intermediate in the range measured for older coastal plain deposits within the Colorado catchment. However, its 26 Al/10 Be ratio is similar to that of the pre-dam sample — indicating a maximum difference in apparent burial age of ∼200 ka. Thus, the scoured material probably consisted primarily of Beaumont or post-Beaumont sediment stored downstream from the dam. More TCN measurements are needed to test the validity of this potentially useful approach to sediment dynamics associated with dams. 5.2.2. Paleo-denudation rates and climate change Our 10 Be-derived denudation rates from the Colorado and Trinity rivers (Table 3) indicate a positive correlation with the global marine δ 18 O record, suggesting that glacial-to-interglacial progressions induce a landscape erosional response (Fig. 4). For both rivers, up-catchment denudation rates correlate with changes in relative temperature (Fig. 5), and significantly increased [∼30–35%] from glacial to interglacial times — a magnitude consistent with sediment flux estimates calculated by Blum and Hattier-Womack (2009) and Garvin (2008) using the temperature-dependent BQART model of Syvitski and Milliman (2007). For the Colorado River, an increase is observed even prior to correcting our denudation rates for differential burial using the 26 Al/10 Be ratio. For the Trinity River, however, an increase is not apparent from the raw values. This suggests that sediment recycling had a greater effect in Trinity’s low-yield system. The relative contribution from stored sediment was calculated at ∼10–30% in both rivers. This observation reinforces the importance of correcting 10 Be denudation rate estimates for sediment recycling. Denudation rates for the mid-Pleistocene Lissie Formation are considerably different from those measured over the last few glacial–interglacial cycles. Values for the Colorado catchment are ∼20% more than any of our estimates for the last ∼200 ka, but values for the Trinity catchment are ∼40% lower than the lowest rates measured over the last ∼50 ka. Uncertainties in chronology aside, this apparent discrepancy could be the result of integrating rates over OIS 15 and OIS 16 in the Colorado and Trinity rivers, respectively. Such an explanation is consistent with the

Fig. 5. Calculated 10 Be denudation rate plotted against time-integrated δ 18 O, after the LR04 benthic stack of Lisiecki and Raymo (2005). δ 18 O data are normalized to the modern value, and negative values correspond to times of greater average global ice mass. Linear regression curves [R 2 values shown on plot] for paleo-denudation rates from Colorado and Trinity catchments show a trend of decreased denudation rate with general decrease in average temperature. Plot shows all 26 Al/10 Becorrected denudation rates except for those acquired from Lissie sediments; these were excluded since their poor temporal resolution precluded a reasonable calculation of time-integrated δ 18 O.

relative chronology at the two Lissie sites, and also with global temperature records that indicate these isotope stages represent exceptionally extended cold and warm periods (e.g. Lüthi et al., 2008). Furthermore, in the Trinity catchment, low TCN concentrations from channel-belt sand underlying the Lissie profile at site 13, along with an erosive basal contact [i.e. no preserved paleosol], suggest that considerable coastal plain incision occurred, possibly in response to base-level change during a relative cold period. If so, the low denudation rate from the Lissie at this site may represent a glacial period, probably OIS 16 based on chronological considerations. The Gulf of Mexico study area has a context of minimal tectonic activity, lack of glacial cover, and stable drainage areas during the Quaternary. Further, because our paleo-denudation rates are obtained from low-relief and gently-sloping landscapes that dominate the Earth’s surface (Willenbring et al., 2013), we infer this record to be of global significance. Because the climate of central Texas had more available moisture during cold [glacial] intervals (Musgrove et al., 2001), we assume that the observed increase in sediment flux during warm [interglacial] intervals is not due to higher net river discharge. Instead, other temperature-dependent processes such as chemical weathering, precipitation intensity and frequency (e.g. Molnar, 2001), changes in vegetation cover (e.g. Hay et al., 2002), or changes in soil thickness may account for the variation in sediment discharge from these systems. If we extrapolate this trend to Pliocene times, when global temperatures and sea levels were significantly higher, our observed temperature–denudation relationship suggests that landscape erosion rates in our study area were considerably higher than during the Pleistocene. We recognize that in addition to temperature, threshold changes in vegetation, moisture, and the increase in amplitude and frequency of climate change from the Pliocene to the Pleistocene may invalidate this temperature–sediment flux relationship in the Pliocene. However, if this simplistic inference is true, the prospect of low-relief, non-glaciated and tectonically inactive regions eroding more rapidly during warm periods has implications for global sediment budgets. Supposing a constant global flux of terrestrial material to the ocean over the last ∼12 Ma,

A.J. Hidy et al. / Earth and Planetary Science Letters 390 (2014) 209–221

as argued by Willenbring and von Blanckenburg (2010), an increase in flux from mountainous or glaciated regions would require an opposing reduction in flux from low-relief and non-glaciated regions — which has indeed been observed in 10 Be denudation rate studies (Fuller et al., 2009; Schaller et al., 2002). Several researchers infer a global increase in net Quaternary sediment flux to the oceans (Herman et al., 2013; Métivier et al., 2002; Molnar, 2004; Peizhen et al., 2001); since those studies are almost exclusively of mountainous, glaciated and/or tectonically active landscapes, they are not necessarily representative of a global average. 6. Conclusions (1) Denudation rates from catchments that drain the tectonically inactive, non-glaciated, and low-relief landscape of interior Texas indicate a 30–35% increase in up-catchment denudation from glacial to interglacial intervals during the Late Pleistocene, consistent with predictions of temperature-driven models for paleosediment flux. Our results suggest these landscapes erode more rapidly during interglacial periods with higher temperatures. (2) During glaciations and times of low sea-level, we observe an increase in the proportion of reworked, previously stored coastal plain sediment relative to up-catchment sediment in the total sediment flux from the Colorado and Trinity rivers. For these systems, about 10–30% of the total medium- to coarse-grained sediment was reworked, probably from Plio-Pleistocene coastal plain strata. This enhanced erosion of stored sediment was contemporaneous with decreased up-catchment denudation, which buffers the climatically-controlled differences between glacial and interglacial sediment delivery. (3) We obtain the first numerical ages for the Lissie Formation using TCN and OSL techniques. The results (∼520–590 ka) support existing stratigraphic frameworks and provide a means to improve Gulf of Mexico sediment models that rely on onshore–offshore correlations and rates of long-term shelf migration. (4) The magnitudes of up-catchment 10 Be sediment flux estimates from the Brazos, Colorado, and Trinity rivers are of the same order as modern pre-dam measurements acquired from sediment traps. They are also relatively consistent with those expected from modern measurements (greatest to least flux by river, respectively). (5) 10 Be catchment-wide denudation rates from pre- and postdam sediment indicate that trapping of sediment behind dams affects 10 Be concentrations over decadal time scales. If the goal is to study temporal changes in 10 Be catchment-wide denudation rates without an anthropogenic effect, pre-dam deposits should be targeted to obtain a reliable modern datum. (6) Recent data that suggest constant global erosion rates from the Pliocene onwards appear to be at odds with mass-balance calculations that indicate a significant increase in global erosion rates entering the Quaternary. To resolve this apparent discrepancy, we hypothesize that the landscape response to climate is counterbalanced between mountainous and glaciated regions and low-relief and non-glaciated regions. Acknowledgements We thank M. Garvin for field assistance during multiple sample trips, and for thoughtful discussions of coastal plain sedimentology; G. Yang for support at the Dalhousie Geochronology Centre cosmogenic nuclide lab; R. Finkel, D. Rood, and S. Zimmerman for long AMS measurement times to improve precision of the 26 Al/10 Be; T. Rittenour and J. Pederson at the Utah State University Luminescence Lab for assistance with interpreting our OSL data; and two anonymous reviewers who improved the manuscript. J. Gosse acknowledges support from the 2004 Petro-Canada Young

219

Investigators Award, NSERC Discovery Grant, and NSERC Major Resources Support Grant for field travel, chemical preparation, AMS measurements, and OSL date for the Lissie Formation; A.J.H. appreciates research funds from the Geological Society of America (GSA), American Geophysical Union (AGU), and Gulf Coast Association of Geological Societies (GCAGS) for travel and accommodation during field work. Appendix A. Supplementary material Supplementary material related to this article can be found online at http://dx.doi.org/10.1016/j.epsl.2014.01.011. References Anderson, J.B., 2005. Diachronous development of late quaternary shelf-margin deltas in the Northwestern Gulf of Mexico: Implications for sequence stratigraphy and deep-water reservoir occurrence. In: River Deltas — Concepts, Models, and Examples. In: SEPM Special Publication, vol. 83, pp. 257–276. Anderson, R.S., Repka, J.L., Dick, G.S., 1996. Explicit treatment of inheritance in dating depositional surfaces using in situ 10 Be and 26 Al. Geology 24, 47. Antinao, J.L., McDonald, E., 2013. A reduced relevance of vegetation change for alluvial aggradation in arid zones. Geology 41, 11–14. Armentrout, J., Clement, J., 1990. Biostratigraphic calibration of depositional cycles: a case study in High Island–Galveston–East Breaks areas, offshore Texas. In: Armentrout, J.M., Perkins, B.F. (Eds.), Sequence Stratigraphy as an Exploration Tool: Concepts and Practices in the Gulf Coast, Proceedings of the 11th Annual Research Conference. Gulf Coast Section of the Society of Economic Paleontologists and Mineralogists Foundation, pp. 21–51. Aronow, S., Barnes, V., 1982. Houston sheet. In: Geologic Atlas of Texas. The University of Texas at Austin, Bureau of Economic Geology. Scale 1. Balco, G., Stone, J.O.H., Mason, J.A., 2005. Numerical ages for Plio-Pleistocene glacial sediment sequences by 26 Al/10 Be dating of quartz in buried paleosols. Earth Planet. Sci. Lett. 232, 179–191. Balco, G., Stone, J.O., Lifton, N.A., Dunai, T.J., 2008. A complete and easily accessible means of calculating surface exposure ages or erosion rates from 10 Be and 26 Al measurements. Quat. Geochronol. 3, 174–195. Belmont, P., Pazzaglia, F., Gosse, J., 2007. Cosmogenic 10 Be as a tracer for hillslope and channel sediment dynamics in the Clearwater River, western Washington State. Earth Planet. Sci. Lett. 264, 123–135. Bierman, P., Steig, E.J., 1996. Estimating rates of denudation using cosmogenic isotope abundances in sediment. Earth Surf. Process. Landf. 21, 125–139. Bloodgood, D.W., 1955. Sixteenth annual report of silt load of Texas streams for water year 1953–1954. Texas Water Development Board, p. 54. Bloodgood, D.W., Meador, A.A., 1941. The silt load of Texas streams, part II, p. 23. Blum, M.D., Aslan, A., 2006. Signatures of climate vs. sea-level change within incised valley-fill successions: Quaternary examples from the Texas Gulf Coast. Sediment. Geol. 190, 177–211. Blum, M.D., Hattier-Womack, J., 2009. Climate change, sea-level change, and fluvial sediment supply to deepwater depositional systems. In: SEPM Special Publication. Tulsa, Oklahoma, USA, pp. 15–39. Blum, M.D., Price, D.M., 1998. Quaternary alluvial plain construction in response to glacio-eustatic and climatic controls, Texas Gulf coastal plain. In: SEPM Special Publication, vol. 59, pp. 31–48. Blum, M.D., Törnqvist, T.E., 2000. Fluvial responses to climate and sea-level change: a review and look forward. Sedimentology 47, 2–48. Blum, M.D., Valastro, S.J., 1994. Late Quaternary sedimentation, lower Colorado River, Gulf Coastal Plain of Texas. Geol. Soc. Am. Bull. 106, 1002–1016. Blum, M.D., Morton, R.A., Durbin, J.M., 1995. Deweyville terraces and deposits of the Texas Gulf coastal plain. Gulf Coast Association of Geological Societies Transactions 45, 53–60. Blum, M., Martin, J., Milliken, K., Garvin, M., 2013. Paleovalley systems: Insights from Quaternary analogs and experiments. Earth-Sci. Rev. 116, 128–169. Braucher, R., Brown, E., Bourles, D., Colin, F., 2003. In situ produced 10 Be measurements at great depths: implications for production rates by fast muons. Earth Planet. Sci. Lett. 211, 251–258. Braucher, R., Del Castillo, P., Siame, L., Hidy, A., Bourlès, D., 2009. Determination of both exposure time and denudation rate from an in situ-produced 10 Be depth profile: A mathematical proof of uniqueness. Model sensitivity and applications to natural cases. Quat. Geochronol. 4, 56–67. Braucher, R., Merchel, S., Borgomano, J., Bourlès, D., 2011. Production of cosmogenic radionuclides at great depth: A multi element approach. Earth Planet. Sci. Lett. 309, 1–9. Brown, E.T., Stallard, R.F., Larsen, M.C., Raisbeck, G.M., Yiou, F., 1995. Denudation rates determined from the accumulation of in situ-produced 10 Be in the Luquillo Experimental Forest, Puerto Rico. Earth Planet. Sci. Lett. 129, 193–202.

220

A.J. Hidy et al. / Earth and Planetary Science Letters 390 (2014) 209–221

Bryant, V.M.J., Holloway, R.G., 1985. A Late-Quaternary paleoenvironmental record of Texas: an overview of the pollen evidence. Pollen records of Late-Quaternary North American sediments. American Association of Stratigraphic Palynologists Foundation, Dallas (TX), pp. 39–70. Bull, W.B., 1991. Geomorphic Responses to Climatic Change. Oxford University Press, Oxford, UK. Charreau, J., Blard, P.-H., Puchol, N., Avouac, J.-P., Lallier-Vergès, E., Bourlès, D., Braucher, R., Gallaud, A., Finkel, R., Jolivet, M., 2011. Paleo-erosion rates in Central Asia since 9 Ma: A transient increase at the onset of Quaternary glaciations?. Earth Planet. Sci. Lett. 304, 85–92. Codilean, A.T., Fenton, C.R., Fabel, D., Bishop, P., Xu, S., 2012. Discordance between cosmogenic nuclide concentrations in amalgamated sands and individual fluvial pebbles in an arid zone catchment. Quat. Geochronol.. http://dx.doi.org/10.1016/j.quageo.2012.1004.1007. Diegel, F., Karlo, J., Schuster, D., Shoup, R., Tauvers, P., 1995. Cenozoic structural evolution and tectono-stratigraphic framework of the northern Gulf Coast continental margin. In: Jackson, M.P.A., Snelson, S. (Eds.), Salt Tectonics: A Global Perspective. Amer. Assoc. of Petroleum Geologists, pp. 109–151. Doering, J., 1935. Post-Fleming surface formations of coastal southeast Texas and south Louisiana. Am. Assoc. Pet. Geol. Bull. 19, 651–688. DuBar, J., Ewing, T., Lundelius Jr., E., Otvos, E., Winker, C., 1991. Quaternary geology of the Gulf of Mexico coastal plain. In: Quaternary Non-Glacial Geology of the Conterminous United States: Boulder, Colorado. In: The Geology of North America, vol. 2. The Geological Society of America, pp. 583–610. Durbin, J.M., Blum, M.D., Price, D.M., 1997. Late Pleistocene stratigraphy of the lower Nueces River, Corpus Christi, Texas: glacio-eustatic influences on valley-fill architecture. Gulf Coast Assoc. Geol. Soc. Trans. 47, 119–129. Ewing, T.E., 1991. Structural framework. The Gulf of Mexico Basin. In: Geology of North America. Geological Society of America, pp. 31–52. Fuller, T.K., Perg, L.A., Willenbring, J.K., Lepper, K., 2009. Field evidence for climatedriven changes in sediment supply leading to strath terrace formation. Geology 37, 467–470. Galloway, W.E., Ganey-Curry, P.E., Li, X., Buffler, R.T., 2000. Cenozoic depositional history of the Gulf of Mexico basin. Am. Assoc. Pet. Geol. Bull. 84, 1743. Galloway, W.E., Whiteaker, T.L., Ganey-Curry, P., 2011. History of Cenozoic North American drainage basin evolution, sediment yield, and accumulation in the Gulf of Mexico basin. Geosphere 7, 938. Garvin, M.G., 2008. Late Quaternary Geochronologic, Stratigraphic, and Sedimentologic Framework of the Trinity River Incised Valley: East Texas Coast. Louisiana State University. Gosse, J.C., Phillips, F.M., 2001. Terrestrial in situ cosmogenic nuclides: theory and application. Quat. Sci. Rev. 20, 1475–1560. Granger, D., Muzikar, P., 2001. Dating sediment burial with in situ-produced cosmogenic nuclides: theory, techniques, and limitations. Earth Planet. Sci. Lett. 188, 269–281. Granger, D.E., Kirchner, J.W., Finkel, R., 1996. Spatially averaged long-term erosion rates measured from in situ-produced cosmogenic nuclides in alluvial sediment. J. Geol. 104, 249–257. Gustavson, T.C., Finley, R.J., 1985. Late Cenozoic geomorphic evolution of the Texas Panhandle and northeastern New Mexico: case studies of structural controls on regional drainage development. Bureau of Economic Geology, University of Texas at Austin. Hall, S.A., Valastro, S., 1995. Grassland vegetation in the southern Great Plains during the last glacial maximum. Quat. Res. 44, 237–245. Hay, W.W., Shaw, C.A., Wold, C.N., 1989. Mass-balanced paleogeographic reconstructions. Geol. Rundsch. 78, 207–242. Hay, W.W., Soeding, E., DeConto, R.M., Wold, C.N., 2002. The Late Cenozoic upliftclimate change paradox. Int. J. Earth Sci. 91, 746–774. Heinrich, P.V., McCulloh, R.P., 2000. Pliocene surface stratigraphy in the Fort Polk region: Implications for Louisiana surface geology. In: Basin Research Institute Bulletin, vol. 9. Louisiana Geological Survey, pp. 51–74. Heisinger, B., Lal, D., Jull, A., Kubik, P., Ivy-Ochs, S., Knie, K., Nolte, E., 2002a. Production of selected cosmogenic radionuclides by muons: 2. Capture of negative muons. Earth Planet. Sci. Lett. 200, 357–369. Heisinger, B., Lal, D., Jull, A., Kubik, P., Ivy-Ochs, S., Neumaier, S., Knie, K., Lazarev, V., Nolte, E., 2002b. Production of selected cosmogenic radionuclides by muons: 1. Fast muons. Earth Planet. Sci. Lett. 200, 345–355. Herman, F., Seward, D., Valla, P.G., Carter, A., Kohn, B., Willett, S.D., Ehlers, T.A., 2013. Worldwide acceleration of mountain erosion under a cooling climate. Nature 504, 423–426. Hidy, A.J., Gosse, J.C., Pederson, J.L., Mattern, J.P., Finkel, R.C., 2010. A geologically constrained Monte Carlo approach to modeling exposure ages from profiles of cosmogenic nuclides: An example from Lees Ferry, Arizona. Geochem. Geophys. Geosyst. 11, Q0AA10. Hidy, A.J., Gosse, J.C., Froese, D.G., Bond, J.D., Rood, D.H., 2013. A latest Pliocene age for the earliest and most extensive Cordilleran Ice Sheet in northwestern Canada. Quat. Sci. Rev. 61, 77–84. Howard, A.D., Dietrich, W.E., Seidl, M.A., 1994. Modeling fluvial erosion on regional to continental scales. J. Geophys. Res., Solid Earth 99, 13971–13986.

Kanes, W.H., 1970. Facies and development of the Colorado River delta in Texas. Deltaic sedimentation: modern and ancient. Spec. Publ. - Soc. Econ. Paleontol. Mineral. 15, 78–106. Kettner, A.J., Syvitski, J.P.M., 2008. HydroTrend v. 3.0: A climate-driven hydrological transport model that simulates discharge and sediment load leaving a river system. Comput. Geosci. 34, 1170–1183. Kirchner, J.W., Finkel, R.C., Riebe, C.S., Granger, D.E., Clayton, J.L., King, J.G., Megahan, W.F., 2001. Mountain erosion over 10 yr, 10 ky, and 10 my time scales. Geology 29, 591–594. Koppes, M.N., Montgomery, D.R., 2009. The relative efficacy of fluvial and glacial erosion over modern to orogenic timescales. Nat. Geosci. 2, 644–647. Kukla, G., Opdyke, N., 1972. American glacial stages in paleomagnetic time scale. In: GSA Abstracts with Programs, vol. 4, pp. 569–570. Lal, D., 1991. Cosmic ray labeling of erosion surfaces: in situ nuclide production rates and erosion models. Earth Planet. Sci. Lett. 104, 424–439. Lanphere, M.A., Champion, D.E., Christiansen, R.L., Izett, G.A., Obradovich, J.D., 2002. Revised ages for tuffs of the Yellowstone Plateau volcanic field: Assignment of the Huckleberry Ridge Tuff to a new geomagnetic polarity event. Geol. Soc. Am. Bull. 114, 559–568. LeBlanc, R.J., Hodgson, W., 1959. Origin and development of the Texas shoreline. Gulf Coast Assoc. Geol. Soc. Trans. IX, 197–220. Lisiecki, L.E., Raymo, M.E., 2005. A Plio-Pleistocene stack of 57 globally distributed benthic 18 O records. Paleoceanography 20, 522–533. Lüthi, D., Le Floch, M., Bereiter, B., Blunier, T., Barnola, J.-M., Siegenthaler, U., Raynaud, D., Jouzel, J., Fischer, H., Kawamura, K., 2008. High-resolution carbon dioxide concentration record 650,000–800,000 years before present. Nature 453, 379–382. Mandel, R.D., Caran, S.C., 1992. Late Cenozoic Alluvial Stratigraphy and Prehistory of the Inner Gulf Coastal Plain, South-central Texas, Draft guidebook to the 10th annual meeting of the South-Central Friends of the Pleistocene. Lubbock Lake Quat. Res. Center Ser. 4. Mason, J.a., Knox, J.C., 1997. Age of colluvium indicates accelerated late Wisconsinan hillslope erosion in the Upper Mississippi Valley. Geology 25, 267. Matmon, A., Bierman, P., Larsen, J., Southworth, S., Pavich, M., Finkel, R., Caffee, M., 2003. Erosion of an ancient mountain range, the Great Smoky Mountains, North Carolina and Tennessee. Am. J. Sci. 303, 817–855. Matmon, A., Stock, G.M., Granger, D.E., Howard, K.A., 2012. Dating of Pliocene Colorado River sediments: Implications for cosmogenic burial dating and the evolution of the lower, Colorado River. Geol. Soc. Am. Bull. 124, 626–640. McMillan, M.E., Angevine, C.L., Heller, P.L., 2002. Postdepositional tilt of the Miocene-Pliocene Ogallala Group on the western Great Plains: Evidence of late Cenozoic uplift of the Rocky Mountains. Geology 30, 63. McMillan, M.E., Heller, P.L., Wing, S.L., 2006. History and causes of post-Laramide relief in the Rocky Mountain orogenic plateau. Geol. Soc. Am. Bull. 118, 393–405. Mercader, J., Gosse, J.C., Bennett, T., Hidy, A.J., Rood, D.H., 2012. Cosmogenic nuclide age constraints on Middle Stone Age lithics from Niassa, Mozambique. Quat. Sci. Rev. 47, 116–130. Métivier, F., Gaudemer, Y., Tapponnier, P., Klein, M., 2002. Mass accumulation rates in Asia during the Cenozoic. Geophys. J. Int. 137, 280–318. Miller, K.G., Kominz, M.A., Browning, J.V., Wright, J.D., Mountain, G.S., Katz, M.E., Sugarman, P.J., Cramer, B.S., Christie-Blick, N., Pekar, S.F., 2005. The Phanerozoic record of global sea-level change. Science 310, 1293–1298. Milliman, J.D., Syvitski, J.P., 1992. Geomorphic/tectonic control of sediment discharge to the ocean: the importance of small mountainous rivers. J. Geol. 100, 525–544. Molnar, P., 2001. Climate change, flooding in arid environments, and erosion rates. Geology 29, 1071–1074. Molnar, P., 2004. Late cenozoic increase in accumulation rates of terrestrial sediment: How might climate change have affected erosion rates?. Annu. Rev. Earth Planet. Sci. 32, 67–89. Molnar, P., England, P., 1990. Late Cenozoic uplift of mountain ranges and global climate change: chicken or egg?. Nature 346, 29–34. Montgomery, D.R., 2002. Valley formation by fluvial and glacial erosion. Geology 30, 1047–1050. Musgrove, M., Banner, J.L., Mack, L.E., Combs, D.M., James, E.W., Cheng, H., Edwards, R.L., 2001. Geochronology of late Pleistocene to Holocene speleothems from central Texas: Implications for regional paleoclimate. Geol. Soc. Am. Bull. 113, 1532–1543. Nishiizumi, K., Imamura, M., Caffee, M.W., Southon, J.R., Finkel, R.C., McAninch, J., 2007. Absolute calibration of 10 Be AMS standards. Nucl. Instrum. Methods Phys. Res., Sect. B, Beam Interact. Mater. Atoms 258, 403–413. Pazzaglia, F.J., Gardner, T.W., 1994. Late Cenozoic flexural deformation of the middle US Atlantic passive margin. J. Geophys. Res., Solid Earth 99, 12143–12157. Pederson, J., Pazzaglia, F., Smith, G., 2000. Ancient hillslope deposits: Missing links in the study of climate controls on sedimentation. Geology 28, 27–30. Peizhen, Z., Molnar, P., Downs, W.R., 2001. Increased sedimentation rates and grain sizes 2–4 Myr ago due to the influence of climate change on erosion rates. Nature 410, 891–897. Petts, G.E., 1979. Complex response of river channel morphology subsequent to reservoir construction. Prog. Phys. Geogr. 3, 329–362.

A.J. Hidy et al. / Earth and Planetary Science Letters 390 (2014) 209–221

Phillips, J.D., Slattery, M.C., Musselman, Z.A., 2005. Channel adjustments of the lower Trinity River, Texas, downstream of Livingston Dam. Earth Surf. Process. Landf. 30, 1419–1439. Pirmez, C., Prather, B.E., Mallarino, G., O’Hayer, W.W., Droxler, A.W., Winker, C.D., 2012. Chronostratigraphy of the Brazos–Trinity depositional system, western Gulf of Mexico: implications for deepwater depositional models. In: Prather, B.E., Deptuck, M.E., Mohrig, D., Van Hoorn, B., Wynn, R.B. (Eds.), Application of the Principles of Seismic Geomorphology to Continental-Slope and Base-of-Slope Systems: Case Studies from Seafloor and Near-Seafloor Analogues. In: SEPM Special Publication, vol. 99, pp. 111–143. Pisias, N.G., Moore Jr., T., 1981. The evolution of Pleistocene climate: a time series approach. Earth Planet. Sci. Lett. 52, 450–458. Prather, B.E., Pirmez, C., Winker, C.D., 2012. Brazos–Trinity system, western Gulf of Mexico. In: Prather, B.E., Deptuck, M.E., Mohrig, D., Van Hoorn, B., Wynn, R.B. (Eds.), Application of the Principles of Seismic Geomorphology to ContinentalSlope and Base-of-Slope Systems: Case Studies from Seafloor and Near-Seafloor Analogues. In: SEPM Special Publication, vol. 99, pp. 83–109. Proctor Jr., C., Brown, T., Waechter, N., Aronow, S., Barnes, V., 1974. Seguin sheet. In: Geologic Atlas of Texas. The University of Texas at Austin, Bureau Economic Geology. Scale 1. Raymo, M., Ruddiman, W.F., 1992. Tectonic forcing of late Cenozoic climate. Nature 359, 117–122. Reeves, C., 1976. Quaternary stratigraphy and geologic history of southern High Plains, Texas and New Mexico. In: Quaternary Stratigraphy of North America: Stroudsburg, PA. Dowden, Hutchinson and Ross, pp. 213–223. Refsnider, K.A., 2010. Dramatic increase in late Cenozoic alpine erosion rates recorded by cave sediment in the southern Rocky Mountains. Earth Planet. Sci. Lett. 297, 505–511. Safran, E.B., Bierman, P.R., Aalto, R., Dunne, T., Whipple, K.X., Caffee, M., 2005. Erosion rates driven by channel network incision in the Bolivian Andes. Earth Surf. Process. Landf. 30, 1007–1024. Schaller, M., Ehlers, T., 2006. Limits to quantifying climate driven changes in denudation rates with cosmogenic radionuclides. Earth Planet. Sci. Lett. 248, 153–167. Schaller, M., Von Blanckenburg, F., Veldkamp, A., Tebbens, L., Hovius, N., Kubik, P., 2002. A 30,000 yr record of erosion rates from cosmogenic 10 Be in Middle European river terraces. Earth Planet. Sci. Lett. 204, 307–320.

221

Stone, J.O., 2000. Air pressure and cosmogenic isotope production. J. Geophys. Res. 105, 23753–23759. Stute, M., Schlosser, P., Clark, J., Broecker, W., 1992. Paleotemperatures in the Southwestern United States derived from noble gases in ground water. Science 256, 1000–1003. Syvitski, J.P.M., Milliman, J.D., 2007. Geology, geography, and humans battle for dominance over the delivery of fluvial sediment to the coastal ocean. J. Geol. 115, 1–19. Toomey III, R.S., Blum, M.D., Valastro Jr., S., 1993. Late Quaternary climates and environments of the Edwards Plateau, Texas. Glob. Planet. Change 7, 299–320. von Blanckenburg, F., 2005. The control mechanisms of erosion and weathering at basin scale from cosmogenic nuclides in river sediment. Earth Planet. Sci. Lett. 237, 462–479. Willenbring, J.K., 2014. Personal communication. Willenbring, J.K., von Blanckenburg, F., 2010. Long-term stability of global erosion rates and weathering during late-Cenozoic cooling. Nature 465, 211–214. Willenbring, J.K., Codilean, a.T., McElroy, B., 2013. Earth is (mostly) flat: Apportionment of the flux of continental sediment over millennial time scales. Geology 41, 343–346. Winker, C.D., 1979. Late Pleistocene fluvial-deltaic deposition, Texas coastal plain and shelf. M.S. thesis. University of Texas at Austin, Austin. 187 p. Wittmann, H., von Blanckenburg, F., 2009. Cosmogenic nuclide budgeting of floodplain sediment transfer. Geomorphology 109, 246–256. Wittmann, H., von Blanckenburg, F., Guyot, J.L., Maurice, L., Kubik, P., 2009. From source to sink: Preserving the cosmogenic 10 Be-derived denudation rate signal of the Bolivian Andes in sediment of the Beni and Mamoré foreland basins. Earth Planet. Sci. Lett. 288, 463–474. Wittmann, H., von Blanckenburg, F., Maurice, L., Guyot, J.L., Filizola, N., Kubik, P.W., 2011a. Sediment production and delivery in the Amazon River basin quantified by in situ–produced cosmogenic nuclides and recent river loads. Geol. Soc. Am. Bull. 123, 934–950. Wittmann, H., von Blanckenburg, F., Maurice, L., Guyot, J.L., Kubik, P.W., 2011b. Recycling of Amazon floodplain sediment quantified by cosmogenic Al-26 and Be-10. Geology 39, 467–470.