Earth and Planetary Science Letters 475 (2017) 44–57

Contents lists available at ScienceDirect

Earth and Planetary Science Letters

www.elsevier.com/locate/epsl

Enhanced provenance interpretation using combined U–Pb and

(U–Th)/He double dating of detrital zircon grains from lower Miocene

strata, proximal Gulf of Mexico Basin, North America

Jie Xu a,b,c,∗, Daniel F. Stockli b, John W. Snedden c

a Now at School of Ocean Sciences, China University of Geosciences (Beijing), Beijing 100083, Chinab Department of Geological Sciences, Jackson School of Geosciences, The University of Texas at Austin, 2275 Speedway C9000, Austin, TX, 78712, USAc Institute for Geophysics, Jackson School of Geosciences, The University of Texas at Austin, 10100 Burnet Road, Austin, TX, 78758-4445, USA

a r t i c l e i n f o a b s t r a c t

Article history:Received 8 November 2016Received in revised form 13 July 2017Accepted 14 July 2017Available online 9 August 2017Editor: A. Yin

Keywords:zircon U–Pb geochronologyzircon (U–Th)/He thermochronometrydouble datingsediment provenanceMiocene Gulf of Mexico

Detrital zircon U–Pb analysis is an effective approach for investigating sediment provenance by relating crystallization age to potential crystalline source terranes. Studies of large passive margin basins, such as the Gulf of Mexico Basin, that have received sediment from multiple terranes with non-unique crystallization ages or sedimentary strata, benefit from additional constraints to better elucidate provenance interpretation. In this study, U–Pb and (U–Th)/He double dating analyses on single zircons from the lower Miocene sandstones in the northern Gulf of Mexico Basin reveal a detailed history of sediment source evolution. U–Pb age data indicate that most zircon originated from five major crystalline provinces, including the Western Cordillera Arc (<250 Ma), the Appalachian–Ouachita orogen (500–260 Ma), the Grenville (1300–950 Ma) orogen, the Mid-Continent Granite–Rhyolite (1500–1300 Ma), and the Yavapai–Mazatzal (1800–1600 Ma) terranes as well as sparse Pan-African (700–500 Ma) and Canadian Shield (>1800 Ma) terranes. Zircon (U–Th)/He ages record tectonic cooling and exhumation in the U.S. since the Mesoproterozoic related to the Grenville to Laramide Orogenies. The combined crystallization and cooling information from single zircon double dating can differentiate volcanic and plutonic zircons. Importantly, the U–Pb–He double dating approach allows for the differentiation between multiple possible crystallization-age sources on the basis of their subsequent tectonic evolution. In particular, for Grenville zircons that are present in all of lower Miocene samples, four distinct zircon U–Pb–He age combinations are recognizable that can be traced back to four different possible sources. The integrated U–Pb and (U–Th)/He data eliminate some ambiguities and improves the provenance interpretation for the lower Miocene strata in the northern Gulf of Mexico Basin and illustrate the applicability of this approach for other large-scale basins to reconstruct sediment provenance and dispersal patterns.

© 2017 Elsevier B.V. All rights reserved.

1. Introduction

Detrital zircon (DZ) U–Pb geochronology has become a widely applied tool for provenance analyses elucidating sediment origin, sediment dispersal, drainage basin evolution, as well as provid-ing maximum depositional age constraints (e.g., Dickinson and Gehrels, 2009a; Lawton et al., 2009; Leier and Gehrels, 2011; Gehrels, 2014). This methodology has been increasingly applied to the understanding of Cenozoic sediment delivery into the Gulf of Mexico (GOM) Basin (e.g., Mackey et al., 2012; Craddock and Kylander-Clark, 2013; Blum and Pecha, 2014; Wahl et al., 2016;

* Corresponding author at: School of Ocean Sciences, China University of Geo-sciences (Beijing), Beijing 100083, China.

E-mail address: jiexu@utexas.edu (J. Xu).

http://dx.doi.org/10.1016/j.epsl.2017.07.0240012-821X/© 2017 Elsevier B.V. All rights reserved.

Xu et al., 2017). Despite the undeniable power of DZ U–Pb dat-ing, the approach has some inherent limitations that can introduce ambiguities that limit the resolution of sediment provenance re-construction. While DZ U–Pb ages are useful to trace sediment to their ultimate crystalline terranes, they provide little information to differentiate zircons derived from monotonous crystallization provenances from those originating from multiple potential source terranes with similar crystallization ages (Reiners et al., 2005). For example, Grenville basement rocks, which can be found ubiqui-tously from NE Canada to Mexico, are characterized by high zircon fertility (Moecher and Samson, 2006; Dickinson, 2008). Hence, Grenville-aged DZ U–Pb ages dominate provenance signals in Pa-leozoic through Cenozoic strata throughout eastern (e.g., Park et al., 2010) and western North America (e.g., Dickinson and Gehrels, 2003; Dickinson and Gehrels, 2009b). Given this omnipresence,

J. Xu et al. / Earth and Planetary Science Letters 475 (2017) 44–57 45

Fig. 1. North American crustal terrains, orogenic belts, and sample locations in the northern Gulf of Mexico Basin. The map is adapted from Gehrels et al. (2011), Blum and Pecha (2014), and Fildani et al. (2016). Abbreviations: GRG – Greater Rio Grande Embayment; H – Houston Embayment; MS – Mississippi Embayment; EGOM – Eastern Gulf of Mexico Embayment; M – Marysvale volcanic field; SJ – San Juan volcanic field; TP – Trans-Pecos volcanic field; SMO – Sierra Madre Occidental volcanic field; A–W – Amarillo–Wichita.

Mesoproterozoic DZ U–Pb ages without additional differentiation do not represent a diagnostic signal. In addition, zircon is very stable and can survive multiple erosion–deposition cycles. Such recycling leads to homogenized DZ U–Pb signatures and thus provenance ambiguities, as zircon can be derived from primary crystalline bedrock sources or from multi-cycle sedimentary strata, yielding a similar pattern of U–Pb ages (e.g., Grenville zircons in North America; Thomas, 2011). Amid these complexities and limi-tations, interpretation of detailed sediment provenance and disper-sal of continental-scale drainage systems, such as the Mississippi River, presents a challenge as these large rivers drain vast portions of the continent hinterland and transport sediments derived from various crystalline terranes and recycled sedimentary strata.

The resolving power of DZ U–Pb dating, however, can be im-proved through combination with additional provenance proxies such as sandstone petrology or heavy mineral analysis (Dickinson and Suczek, 1979; Dickinson, 1985) or isotopic or geochemical analysis of the same zircon, including Hf or O isotope analyses (e.g., Hawkesworth and Kemp, 2006; Gehrels and Pecha, 2014; Howard et al., 2015), trace element fingerprinting (Stockli, 2017; Marsh and Stockli, 2015), or fission-track and (U–Th)/He low-temperature thermochronometry (e.g., Reiners et al., 2005; Painter et al., 2014). Hf isotope derived model ages constraining the tim-ing of mantle melt extraction for basement provinces, or stable isotope fingerprints provide additional constraints on the deriva-tion of both first-cycle and multi-cycle zircons (e.g., Howard et al., 2015; Clements et al., 2012). Similarly, trace element and REE mea-surements on zircons by laser-ablation-split-stream ICP-MS analy-sis provide powerful insights into petrologic environment and can allow for differentiation between magmatic and metamorphic zir-

cons or the determination of crystallization temperatures (Marsh and Stockli, 2015). While both isotopic or trace element analyses have the potential to differentiate possible bedrock source terranes, they do not have the ability to address recycling and homogeniza-tion complexities related to post-crystallization tectonic processes. This problem can be addressed through the combination of U–Pb and zircon (U–Th)/He (ZHe) by double dating (U–Pb–He) of single zircon grains, as this approach can reveal tectonic exhuma-tion/cooling in the hinterland. In addition to improved provenance interpretations, this is also important for understanding the cou-pling between tectonic unroofing in source areas and sediment dis-persal and delivery into basins in terms of source-to-sink analysis.

The Cenozoic drainage basins feeding sediment into the GOM are particularly well suited for this U–Pb–He double-dating ap-proach, as the long-lived crustal growth of North America is characterized by distinct and well-studied basement provinces with very different U–Pb crystallization ages as well as a num-ber of Phanerozoic tectonic belts with specific cooling and ex-humation histories (Fig. 1). The resulting pairing of zircons with distinct and different U–Pb and ZHe ages makes this approach particularly powerful for differentiation between multiple possi-ble crystallization-age sources on the basis of their subsequent tectonic evolution. This is particularly important in the case of the northern GOM, as the hinterland contains numerous poten-tial sources for recycled zircons from Paleozoic–Mesozoic sedimen-tary strata, for example in the Appalachian and Cordillera foreland basins or Laramide intermountain basins (Fig. 2). While two zircon U–Pb–He double dating studies were carried out on the modern Mississippi river (Reiners et al., 2005) and Pleistocene Mississippi fan (Fildani et al., 2016), no U–Pb–He double dating has been ap-

46 J. Xu et al. / Earth and Planetary Science Letters 475 (2017) 44–57

Fig. 2. Detrital zircon U–Pb ages from (A) Paleozoic–Mesozoic sedimentary cover in Sevier foreland basin and Paleocene–Eocene strata in Laramide basins, and (B) Paleozoic strata in Appalachian–Ouachita foreland basins. Please note time scale change in 300 Ma. Superscripts indicate zircon U–Pb ages from the hinterland sedimentary sources: 1 – Bush et al. (2016); 2 – Dickinson and Gehrels (2003); 3 – Dickinson and Gehrels (2008); 4 – Dickinson and Gehrels (2009b); 5 – Lawton and Bradford (2011); 6 – Sharman et al. (2017); 7 – Park et al. (2010); 8 – Xie et al. (2016). n = number of analyses.

plied to understand the complex drainage system of the lower Miocene system in the GOM Basin. In this work, 144 new ZHe ages are combined with >2000 zircon U–Pb ages (Xu et al., 2016) for double dating to reduce source ambiguity (e.g., Grenville zircons), and to refine provenance interpretations.

2. Geologic background

Several previous studies have investigated the northern GOM in the Cenozoic and proposed configurations for a continental-scale drainage system delivering sediment from the North Amer-ican hinterland (e.g., Galloway et al., 2000, 2011; Galloway, 2008;Blum and Pecha, 2014). In order to further refine these reconstruc-tions using DZ double dating, however, it is necessary to sum-marize the continental crustal growth and tectonic exhumation history of North America before correlating zircon U–Pb age and ZHe age to the potential sediment source terranes.

2.1. Crustal assembly of North America

The cratonic core of North America is primarily composed of various Archean basements (e.g., Wyoming, Slave, Superior) which were amalgamated by continent-to-continent collisions during the Trans-Hudson Orogeny (2.0–1.8 Ga; Hoffman, 1988; Fig. 1). After assembly, the Yavapai (1.8–1.7 Ga) and Mazatzal (1.7–1.65 Ga) provinces, which are dominantly composed of juvenile arc crust, accreted to Laurentia along the southern margin (Whitmeyer and Karlstrom, 2007). A-type magmatism ensued between 1.48 and 1.35 Ga (here termed Mid-Continent; Bickford et al., 1986;Whitmeyer and Karlstrom, 2007; Fig. 1). A protracted period of tectonism (1.3–0.95 Ga) in eastern and southern Laurentia fol-lowed during formation of the supercontinent of Rodinia. Laurentia was dominated by rift-related magmatism during Rodina breakup (0.95–0.5 Ga) and was followed in the Paleozoic (500–250 Ma; Hatcher, 2010; Fig. 1) by a series of orogenic events during the Appalachian–Ouachita orogeny along its eastern and south-

J. Xu et al. / Earth and Planetary Science Letters 475 (2017) 44–57 47

ern margins. From the Permian to the Cenozoic, western North America was dominated by subduction, arc magmatism, and retro-arc shortening during the Sevier and Laramide orogenies in the Cordillera (post-250 Ma; Dickinson, 2004; DeCelles, 2004). In con-trast, eastern and southern North America remained relatively stable after Pangean rifting in the Mesozoic, leading to the opening of the Atlantic and Gulf of Mexico (250–150 Ma; Weislogel et al., 2015).

2.2. Tectonic evolution and exhumation of the United States

Before the Mesozoic, eastern and southeastern Laurentia con-tained the most tectonically active regions in North America. Eastern and southern Laurentia were impacted by the Grenville Orogeny (1.3–0.95 Ga), providing sediment to paleorivers that flowed more than 3000 km across the Laurentian craton to the Neoproterozoic Shaler Group in NW Canada (Rainbird et al., 1992). Eastern Laurentia was subsequently rejuvenated during several orogenic events that collectively define the Paleozoic Appalachian–Ouachita Orogeny (500–250 Ma), resulting in constructional ex-humation within the orogens and the development of foreland basin systems (e.g., Park et al., 2010) as well as trans-continental transport to the rifted continental margin along the western edge of Laurentia (e.g., Dickinson and Gehrels, 2003; Dickinson and Gehrels, 2009b). Meanwhile, the Ancestral Rocky Mountains linked to Alleghenian–Ouachita tectonism in the western U.S. (Kluth, 1986; Dickinson and Lawton, 2003) exhumed Yavapai–Mazatzal and Mesoproterozoic (1480–1340 Ma) crystalline basement rocks (Dickinson and Gehrels, 2003). In the Late Paleozoic to Cenozoic, major tectonism and magmatic events dominated western North American in response to subduction, arc magmatism, and retro-arc shortening and extension in the Cordillera (Dickinson, 2004). The major tectonic events, including the Sevier–Nevadan, Laramide oro-genies and mid-Cenozoic volcanism, strongly influenced hinterland erosion, sediment dispersal, and deposition in the northern GOM Basin throughout the Cenozoic (Galloway et al., 2011).

2.3. Early Miocene evolution of the Gulf of Mexico basin

The GOM basin in the Cenozoic represents the ultimate sed-iment sink, recording sediment supply from the North American interior and providing unique opportunities to analyze the link-ages between tectonic changes in the hinterland source areas and sediment deposition in the basinal sink. More than 700,000 km3

of siliciclastic sediment are estimated to have been deposited on the continental passive margin in the form of large deltas and deepwater submarine fans (e.g., Galloway et al., 2011). The Early Miocene is a transitional period in terms of drainage reorgani-zation in the western GOM, with diminished input from the Rio Grande and Houston–Brazos River and the emergence of the Red River from Oligocene to middle Miocene times (Galloway et al., 2011). Meanwhile in the eastern GOM, increased erosion of the Appalachian Mountains due to widespread uplift and incision in the late early Miocene–middle Miocene resulted in the ancestral Mississippi becoming the main sediment carrier in the eastern GOM (e.g., Boettcher and Milliken, 1994; Galloway et al., 2011;Dutton et al., 2012).

In the hinterland, the thermal updoming induced by Eocene–late Oligocene magmatism caused significant erosion on the south-ern Great Plains, the southern Colorado Plateau, parts of northeast-ern Mexico, and central to western Texas (e.g., Cather et al., 2012). The southern Rocky Mountains were deeply eroded during Late Oligocene to Early Miocene times, including major exhumation in the Sangre de Cristo and Sandia Mountains in New Mexico (e.g., Kelley and Chapin, 1995; Pazzaglia and Kelley, 1998). Meanwhile, the Rio Grande Rift started to create extensional basins and to trap

Fig. 3. Petrographic data of lower Miocene strata in the western-central Gulf of Mex-ico. Modified from Dutton et al. (2012). Provenance fields of QtFL ternary plots are after Dickinson (1985).

sediment and also prevent sediment transport from the Colorado Plateau and western Cordillera to the GOM (Chapin and Cather, 1994). Locally, the Llano Uplift and Edwards Plateau were uplifted and eroded in early Miocene (Ewing, 2005).

Previous petrographic studies have suggested regional varia-tions in sandstone composition from the Rio Grande Embayment to the Mississippi Embayment (Fig. 3; Dutton et al., 2012). Sandstone compositional maturity increases from the western to the central GOM, indicating a sediment source shift from volcanic sources in the western GOM to more compositionally mature sediments from the cratonic interior in Mississippi Embayment.

3. Methodology

In this study, 16 outcrop samples and 4 subsurface cores were selected from Early Miocene strata in the northern GOM for zir-con U–Pb–He analyses (Fig. 1). All analyzes were conducted at the UTChron Geo- and Thermochronometry Laboratories at the Univer-sity of Texas at Austin. After mineral separation, at least 120 de-trital zircons were randomly selected for depth-profile U–Pb anal-yses on a Laser Ablation Inductively Coupled Plasma Mass Spec-trometry (LAc-ICP-MS) using procedures described in Marsh and Stockli (2015) to obtain a statistically representative provenance dataset (Vermeesch, 2004). Depth-profile ICP-MS analysis enables ages from multiple zircon growth zones (e.g., core and rim ages) to be resolved, since the laser ablates perpendicular to growth zoning from the exterior of an unpolished grain, and preserves the zir-con for subsequent conventional whole-grain ZHe dating (Stockli, 2017). The U–Pb data from these analyses were then reduced using the Iolite data reduction software with VizualAge (Petrus and Kam-ber, 2012). For zircons younger than 850 Ma, 206Pb/238U ages are reported, whereas 207Pb/206Pb ages are reported for zircons older than 850 Ma. This age division was chosen so as not to artificially divide a population, as well as to maximize the precision of the ages. All ages reported use 2σ absolute propagated uncertainties, 207Pb/206Pb ages are less than 30% discordant, and 206Pb/238U ages are less than 10% discordant (Gehrels, 2011). The discordance re-ported is calculated using the 206Pb/238U and 207Pb/235U ages if <850 Ma and the 206Pb/238U and 207Pb/206Pb ages if >850 Ma.

After depth-profile U–Pb analysis, 144 zircons representative of each major U–Pb age component from individual samples, were

48 J. Xu et al. / Earth and Planetary Science Letters 475 (2017) 44–57

Fig. 4. Detrital zircon U–Pb age data of 20 lower Miocene samples, northern Gulf of Mexico. (A) Normalized probability plot of composited zircon U–Pb ages in each embayment. Only limited grains have U–Pb ages older than 2000 Ma (<5%); thus, they are omitted here. (B) Cumulative percentage plot of zircon U–Pb ages of each sample. n = number of analyses. Sample locations are shown in Fig. 1.

selected for ZHe analysis following the laboratory procedures of Wolfe and Stockli (2010). Unfractured and euhedral zircons with width >70 μm were morphometrically measured for alpha ejec-tion correction and packed into 1 mm platinum tubes for diode laser heating (10 minutes at ∼1300 ◦C) to extract 4He and reheat-ing for ∼10 minutes at ∼1300 ◦C until 4He yield dropped to <1%. Degassed zircons were dissolved after removal from Pt tubes, using a two-step HF–HNO3 and HCl pressure–vessel digestion procedure and analyzed for U, Th by ID-ICP-MS analysis using a Thermo El-ement2 ICP-MS. Reported ages are alpha ejection corrected (F T ) and uncertainties are standard errors (∼8%, 2σ ) on the basis of intra-laboratory reproducibility of the Fish Canyon Tuff zircon.

4. Results

In order to demonstrate the utility of the double-dating ap-proach for enhanced provenance interpretation, this work presents 144 new ZHe ages from 10 samples collected from the northern GOM (Fig. 1; Table A1). Published U–Pb ages from 19 samples (GOM1–19, n = 2192; Xu et al., 2017) and 112 new DZ U–Pb ages from an offshore sample (GOM20; Table A2) were combined with ZHe ages to refine sediment provenance interpretation. DZ U–Pb

data are characterized by five major age components: (1) 250–24 Ma, (2) 500–250 Ma, (3) 1300–950 Ma, (4) 1500–1300 Ma, and (5) 1800–1600 Ma, with minor contributions from zircons having U–Pb ages of 700–500 Ma or greater than 1800 Ma (Fig. 4). Based on zircon U–Pb age signatures and sample geographic locations, the samples can be divided into four distinct groups: (1) Greater Rio Grande (GRG) Embayment, (2) Houston Embayment, (3) Missis-sippi Embayment, and (4) Eastern GOM Embayment (Figs. 1 and 4; Xu et al., 2017). The zircon U–Pb ages display a shift in relative age abundances from the western to eastern GOM as shown in the probability curve and cumulative percentage plot (Figs. 4A and 4B).

ZHe data display a wide spectrum of cooling (exhumation) ages ranging from Mesoproterozoic to Cenozoic, with the majority scat-tering between 500–24 Ma (Fig. 5). Paleoproterozoic zircons (U–Pb age: 1800–1600 Ma) are an important component in most sam-ples from the western GOM (7%–38%; Fig. 4) and 28 grains were selected for ZHe analyses. Overall, ∼80% of ZHe ages for Paleo-proterozoic grains record cooling from 210–55 Ma (n = 22), ∼10% from 400–250 Ma (n = 3) and the remaining ∼10% from 1000–600 Ma (n = 3) (Fig. 5; Table A1). Mesoproterozoic zircons (U–Pb: 1500–1300 Ma) are another important U–Pb age component for

J. Xu et al. / Earth and Planetary Science Letters 475 (2017) 44–57 49

Fig. 5. Detrital zircon U–Pb–He age of the lower Miocene strata in the northern Gulf of Mexico Basin.

samples in GRG, Houston and Mississippi embayments (7–23%; Fig. 4). Analyses of 22 Mesoproterozoic zircons (U–Pb: 1500–1300 Ma) that occurred frequently in DZ samples yield similar ZHe age populations, of which ∼70% are 220–40 Ma (n = 16), ∼15% are 400–250 Ma (n = 3) and the remaining ∼15% are 1100–600 Ma (n = 3) (Fig. 5).

Mesoproterozoic to Neoproterozoic zircons (U–Pb age: 1300–950 Ma) form the dominant component in samples from the Mis-sissippi (52–54%) and Eastern GOM embayments (50–75%; Fig. 4) and 46 grains were selected for ZHe analyses (Fig. 5). More than half of the grains (n = 24; ∼50%) have Paleozoic ZHe ages (500–260 Ma; Fig. 5). Another ∼35% zircons yield Mesozoic ZHe ages (n = 15; 230–50 Ma). The rest of the zircons yield Precam-brian (1250–600 Ma; n = 6) ZHe ages.

Neoproterozoic zircons (U–Pb age: 700–500 Ma) only account for 0–5% grains in each sample and only two grains were selected for ZHe analyses, which yield late Paleozoic ZHe ages (Fig. 5). Pa-leozoic (U–Pb age: 500–260 Ma) zircon is secondary important component in the Mississippi (5–14%) and Eastern GOM Embay-ment samples (20–36%; Fig. 4) and 14 grains were picked to obtain ZHe ages. Three grains yield late Paleozoic ZHe ages, and most of the ZHe ages are Mesozoic (n = 10). Only one grain has a Cenozoic ZHe age (Fig. 5; Table A1).

Mesozoic to Cenozoic zircon (U–Pb age: <250 Ma) is the most prevalent component of samples in the GRG Embayment and Houston Embayment (28–68%; Fig. 4). Thirty-two grains of those grains were analyzed for ZHe and mostly yielded ZHe ages that are within errors of their U–Pb ages (Fig. 5).

5. Interpretation and discussion

5.1. Summary of detrital Zircon U–Pb provenance analysis

DZ U–Pb age distributions for the GRG Embayment samples are dominated by Mesozoic to Cenozoic ages (260–24 Ma; Fig. 4) likely originating from the Permian–Triassic Cordilleran magmatic arc, Laramide magmatic rocks, and mid-Cenozoic volcanic rocks in western North America (Chen and Moore, 1982; Dickinson and Lawton, 2001; Ducea, 2001; DeCelles, 2004; Chapin et al., 2004). Mid-Continent (1500–1300 Ma) and Yavapai–Mazatzal (1800–1600 Ma) DZ grains are second most prominent components

(Fig. 4), indicating a possible source from the uplifted Laramide basement blocks (Xu et al., 2017). Grenville (1300–950 Ma) zir-cons could be sourced either from A-type granites (e.g., Pike Peak) or more locally from the Llano uplift that was erosionally un-roofed in early Miocene (Ewing, 2005; Stockli, 2017). Alternatively, Paleozoic–Cenozoic sedimentary strata in the western U.S. contain similar DZ U–Pb age populations that could have been recycled in Cenozoic times (Fig. 2).

DZ U–Pb age spectra in the Houston Embayment samples show prominent DZ peaks of Mid-Continent and Yavapai–Mazatzal com-ponents, indicating increased sediment supply from Laramide up-lifts (e.g., the Front Range). Alternatively, Pennsylvanian–Cretaceous strata of the Rocky Mountains (Dickinson and Gehrels, 2003; Dickinson and Gehrels, 2009b; Bush et al., 2016), or Paleocene–Eocene strata in Laramide basins characterized by similar zircon U–Pb age distributions, could also be possible recycled sources for these 1800–1300 Ma zircons (Fig. 2). Western Cordillera magmatic components are still present, but are slightly decreases in Houston Embayment compared to the GRG (Fig. 4).

The DZ U–Pb age spectra of the Mississippi Embayment sam-ples differ distinctly from the age spectra of Houston Embayment or GRG samples and is characterized by the predominance of Grenville zircons (Fig. 4). The Western Cordillera DZ component is secondary and combined with the Mid-Continent and Yavapai–Mazatzal components, and constitutes about 30% of the total DZ U–Pb age population in the Mississippi Embayment samples. The DZ U–Pb age spectra indicate sediment supply from both high-lands in the western U.S. and Appalachian terranes in eastern North America (Xu et al., 2017). The EGOM samples are dominated by Appalachian–Ouachita (500–260 Ma) and Grenville components, indicating dominant sources from the southern Appalachian Moun-tains and associated foreland basins (Becker et al., 2005; Park et al., 2010; Xie et al., 2016).

These DZ U–Pb results reveal zircon provenance from the Yavapai–Mazatzal, Mid-Continent, Grenville, Appalachian–Ouachita, and Western Cordillera Arc terranes. However, many of these zir-cons could also be derived from recycled sedimentary strata in North America or simply are not diagnostic given the spatial ex-panse of some of these terranes. For example, Yavapai–Mazatzal, Mid-Continent, and Grenville detrital zircons are commonly docu-mented both in sedimentary basins of western North America (e.g.,

50 J. Xu et al. / Earth and Planetary Science Letters 475 (2017) 44–57

Dickinson and Gehrels, 2003, Dickinson and Gehrels, 2009b; Fig. 2) as well as Appalachian–Ouachita foreland basins in the east (e.g., Park et al., 2010; Xie et al., 2016; Fig. 2). For a large continental-scale drainage system, the combination of zircons derived directly from basement, and the mixing with zircons from the recycled sedimentary sources make provenance interpretation ambiguous. Therefore, additional ZHe ages on the sample grains should yield powerful additional constraints to eliminate some ambiguities and improves the provenance interpretation.

5.2. Zircon U–Pb–He double dating

5.2.1. Source differentiation constrained by U–Pb–He age5.2.1.1. Differentiation of Grenville zircons by U–Pb–He double datingGrenville zircons represent as an important and common compo-nent in all lower Miocene samples of the proximal northern GOM (Fig. 4). However, provenance for these Grenville zircons is hard to be assigned to a particular source area based on U–Pb age alone, given the spatially widespread occurrence of Grenville-aged ter-ranes and their high zircon fertility (e.g., Moecher and Samson, 2006; Reiners et al., 2005; Dickinson, 2008). Furthermore, sedi-mentary basins contain abundant Grenville zircons, as have been documented in Appalachian–Ouachita foreland basins (Fig. 2; Park et al., 2010; Xie et al., 2016), Paleozoic–Mesozoic sedimentary cover in Cordillera foreland basins (Fig. 2; Dickinson and Gehrels, 2003, 2008; Dickinson and Gehrels, 2009b; Lawton and Bradford, 2011) and Paleocene–Eocene strata in Laramide basins (Fig. 2; Bush et al., 2016; Sharman et al., 2017). Recycling of these sed-imentary strata likely contribute multi-cycled zircons to Miocene strata of the GOM basin. Therefore, zircon U–Pb ages alone are not diagnostic and insufficient to discriminate between zircons sourced directly from a primary igneous rock and multi-cycled zircons de-rived from a sedimentary rock.

Fifteen Grenville zircons from the GRG Embayment were se-lected for ZHe analysis and display a wide range of cooling ages from 1250 to 50 Ma (Fig. 6A). Grenville zircons cooled between 1250 and 700 Ma (n = 4; Fig. 6A) probably record the Grenville assembly and subsequent breakup of the supercontinent Rodinia. Previous work on the Neoproterozoic Shaler Group in northwestern Canada documented an abundance of Grenville zircons, suggesting that large volumes of detritus were derived from the Grenville oro-genic belt in eastern North America (Rainbird et al., 1992). There-fore, Grenville zircons in the Shaler Group may have ZHe ages of 1250–700 Ma. However, in comparison with the abundant Trans-Hudson (2000–1800 Ma) and Superior–Wyoming (>2.5 Ga) zircons in the Shaler Group, the lower Miocene samples lack a large num-ber of grains >1800 Ma, precluding recycled Shaler Group as a ma-jor Grenville zircon source. The Llano Uplift in Central Texas could be a more likely Grenville source as zircons from the Llano Up-lift basement record ZHe ages of 850–750 Ma (Stockli, 2017). The Grenville basement in the Llano Uplift was only shallowly buried and not thermal reset during the Phanerozoic, yielding Neoprotero-zoic ZHe ages recording Rodinia rifting, and became subjected to erosion in the early Miocene (Ewing, 2005; Stokli, 2017).

Five double-dated Grenville zircons from GRG Embayment yielded Paleozoic ZHe ages (500–260 Ma; Fig. 6A), correspond-ing to the Appalachian–Ouachita orogeny. This suggests an ulti-mate origin from the Appalachian terranes. However, the paleo-Mississippi River is thought to have formed between Late Cre-taceous and early Paleocene time and thus preventing zircons from the east getting into the western U.S. in the Miocene (Blum and Pecha, 2014). However, zircon U–Pb–He dating of the Juras-sic Navajo Sandstone in Utah has documented the presence of zircons with Grenville crystallization and Appalachian–Ouachita exhumation ages, which are similar to U–Pb–He ages for the lower Miocene strata (Rahl et al., 2003). Therefore, these grains are

most likely eroded and recycled from the Paleozoic–Mesozoic or Paleocene–Eocene sedimentary cover in the western U.S. (Fig. 6B).

Grenville zircons from the GRG Embayment (n = 3) also ex-hibit Mesozoic ZHe ages (Fig. 6A), but the source of such Mesozoic cooling ages remains enigmatic. These Grenville zircons might be partially reset and derived from Paleozoic–Mesozoic strata during Sevier thrusting. Another three Grenville zircons from the GRG Em-bayment yielded ZHe ages of 73–54 Ma (Fig. 6A), suggestive of tectonic exhumation related to the Laramide Orogeny. They could either be sourced from Grenville basement exposure in western Texas (e.g., the Franklin Mountains) and northern Mexico, or likely from recycled from reset Paleozoic strata exhumed in Laramide up-lifts (Fig. 6B).

In contrast to the variable ZHe ages of Grenville zircons from the GRG Embayment, most Grenville zircons in the Houston Em-bayment (6 of 8) are characterized by Paleozoic ZHe cooling ages (Fig. 6A). This indicates an origin from the Appalachian Mountains or associated foreland basin in eastern North America. However, these Grenville zircons also could not be derived directly from the east given Miocene Mississippi and are thus likely recycled from Paleozoic–Mesozoic strata in the western U.S., which received Grenville zircons with Paleozoic cooling ages from eastern North America (Rahl et al., 2003; Blum and Pecha, 2014; Fig. 6B).

Grenville zircons from the Mississippi Embayment diaplay sim-ilar ZHe ages to those from the Houston Embayment. The Miocene Mississippi River was interpreted to tap both the Rocky Mountains and Appalachian Mountains (Galloway et al., 2011; Xu et al., 2017)and thus Grenville zircons could be derived from either the west-ern or eastern U.S. (Fig. 6B). All Grenville zircons (n = 7) observed in the EGOM display prominent Mesozoic ZHe ages (250–100 Ma; Fig. 6A). While the two older ZHe ages (215 and 224 Ma) could reflect exhumation due to post-Alleghenian erosion or Triassic rift-ing, the remaining grains (n = 5) record Jurassic cooling likely cor-responding to Pangea rifting and exhumation along the Atlantic rift margin (Fig. 6A).

5.2.1.2. Differentiation of older Meso- and Paleoproterozoic zirconsDouble dating of Yavapai–Mazatzal zircons (n = 28) yields a wide spectrum of ZHe ages, ranging from 1100 to 55 Ma (Fig. 7A). Precambrian ZHe ages (1100–624 Ma, n = 3) might be cooled during the assembly and breakup of Rodina in southern Lau-rentia. Paleozoic–Triassic ZHe ages (366–267 Ma; n = 3) ap-pear to be linked to the Ancestral Rocky Mountains or Ouachita Orogeny (Dickinson and Lawton, 2003). These zircons could be sourced from the Pennsylvanian–Permian strata in the Ances-tral Rocky Mountain basins (Fig. 2; Dickinson and Gehrels, 2003;Bush et al., 2016) and are probably recycled into the lower Miocene strata. Triassic to Early Cretaceous ZHe ages (n = 4) could be derived from Sevier–Nevada orogenic exhumation, while Late Cretaceous to Paleogene ZHe ages (n = 4) indicate Laramide in-volvement (Fig. 7A).

Mid-Continent Granite–Rhyolite zircons (n = 22) show a simi-lar ZHe pattern to that of Yavapai–Mazatzal zircons, with cooling ages of 1100 to 40 Ma (Fig. 7A). Three Precambrian ZHe ages ap-pear to be related to the Rodina assembly and breakup (Fig. 7A), whereas other Paleozoic cooled zircons (390–250 Ma; n = 3) are likely related to the Ancestral Rocky Orogeny and exhumation. Similar combination of U–Pb ages (1500–1300 Ma) and ZHe ages (1000–250 Ma) was documented in Jurassic Navajo sandstone in Utah (Rahl et al., 2003), lending support to the notion that these zircons are likely recycled from Paleozoic–Mesozoic sedimentary strata in the western U.S. Most Mid-Continent zircons exhibit Late Triassic–Late Cretaceous ZHe ages (220–90 Ma; n = 11) as well as Late Cretaceous–Paleogene (80–40 Ma; n = 5) that correspond to either Sevier–Nevadan or Laramide orogenic events in western North America, respectively.

J. Xu et al. / Earth and Planetary Science Letters 475 (2017) 44–57 51

Fig. 6. (A) U–Pb–He ages of Grenville zircons in the lower Miocene strata of the Gulf of Mexico basin, (B) Sediment routing of Grenville grains. Pre-Middle Cretaceous paleoriver was proposed by Rahl et al. (2003), Dickinson and Gehrels (2009b), and Blum and Pecha (2014). RGR – Rio Grande Rift. n = number of analyses. Color bars indicate different orogenic events. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

52 J. Xu et al. / Earth and Planetary Science Letters 475 (2017) 44–57

Fig. 7. (A) U–Pb–He ages of Yavapai–Mazatzal and Mid-Continent Granite–Rhyolite zircons. (B) U–Pb–He ages of Pan-African, Appalachian–Ouachita, and Western Cordillera Arc zircons. n = number of analyses. Color bars indicate different orogenic events. Please note the volcanic zircons (�t ∼ 0) are removed from this plot. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Both Mid-Continent and Yavapai–Mazatzal zircons show an abundance of ZHe ages related to the Sevier–Nevadan Orogeny (Fig. 7A), indicating an original source from the Nevada and Se-vier thrust belts. Drainages linking Cordillera thrust terranes to the GOM (Lawton, 2008) supplied Late Cretaceous to Paleogene sedimentary strata on the Great Plains (Fig. 2). However, Ceno-zoic extension in the Basin and Range and Rio Grande rift re-versed early Miocene paleostreams on the Colorado Plateau that collected materials from the Sevier–Nevada terranes to flow west-ward to the Pacific margin and no longer to the GOM (e.g., Cather et al., 2012). As a result of this drainage disruption, the Sevier–Nevada terranes no longer can be considered as a direct zir-con source for the GOM in the early Miocene time. Erosion of these Late Cretaceous to Paleogene strata and remobilized zircons represent a likely intermediate source (e.g., Cather et al., 2012;Xu et al., 2017).

5.2.1.3. Differentiation of Neoproterozoic to Cenozoic zircons Neo-proterozoic–early Paleozoic zircons (700–500 Ma) are very rare in the GRG Embayment samples (Fig. 4) and thus only two grains were analyzed for ZHe dating (Fig. 7B). These Neoproterozoic zir-cons (641 and 533 Ma; n = 2) from the GRG Embayment yield Pa-leozoic Appalachian–Ouachita ZHe ages (373 and 254 Ma; Fig. 7B), indicating an origin from the Peri-Gondwanan terranes (e.g., Car-olina terrane) in eastern North America. For the western GOM, direct sourcing of these Gondwanan–Affinity sources from eastern North America would be implausible in the early Miocene. Hence they are likely recycled from Permian–Jurassic sandstone on the Colorado Plateau with documented U–Pb age peaks of 500–700 Ma and 1000–1300 Ma, suggesting westward sediment flux of Appalachian detritus before the Jurassic (Dickinson and Gehrels, 2003; Dickinson and Gehrels, 2009b).

Paleozoic zircons (500–260 Ma) selected from EGOM Embay-ment exhibit a small number of Alleghenian ZHe age (327–259 Ma; n = 3; Fig. 7B). Mesozoic cooled zircons (240–120 Ma; n = 10) are likely related to post-Alleghenian erosion or Atlantic or GOM rifting (Pazzaglia et al., 2015; Weislogel et al., 2015; Fig. 7B).

Thirty-two Mesozoic–Cenozoic zircons (U–Pb age: 230–24 Ma) exhibit ZHe ages that record cooling at 170–80, 80–40, and

Fig. 8. Identification of volcanic zircons using zircon age differences between U–Pb and (U–Th)/He. (A) Plot of U–Pb age (<250 Ma) versus their corresponding (U–Th)/He age. (C) Plot of U–Pb age (<100 Ma) versus their corresponding (U–Th)/He age.

40–24 Ma, respectively, corresponding to tectonic or magmatic cooling events of the Cordilleran arc, Sevier–Nevadan, Laramide orogenies, and mid-Cenozoic volcanism (Figs. 7B and 8). Volcanic originated zircon is characterized by a very short time differ-ence between crystallization and eruption and thus has U–Pb ages and ZHe ages within errors (�t ∼ 0), whereas plutonic zir-cons will cool gradually after crystallization and thus produce a

J. Xu et al. / Earth and Planetary Science Letters 475 (2017) 44–57 53

Fig. 9. Multicomponent analysis of the U–Pb and (U–Th)/He data from samples in the Greater Rio Grande Embayment, illustrating the complex sediment sources in the northwestern Gulf of Mexico. CMB – Colorado Mineral Belt.

large �t . Many of these grains represent volcanic zircons (�t ∼ 0) eroded from Cretaceous–Cenozoic volcanic rocks in the western U.S. (n = 14; Fig. 8). The 80–40 Ma ZHe ages are likely related to Laramide magmatism in northern Mexico and southwestern New Mexico, or the Colorado Mineral Belt. The grains cooled be-tween 170–80 Ma are probably either derived from igneous rocks in the Sevier–Nevadan Thrust Belt or recycled from the Mesozoic–Cenozoic sedimentary strata in western U.S. Triassic–Early Jurassic zircons cooled in Sevier–Nevadan (n = 3; Fig. 7B) could be from the Late Triassic–Jurassic arc in the Chihuahua Fold Belt of north-ern Mexico (Villarreal-Fuentes et al., 2014) or alternatively sourced from the recycling of the Triassic to Jurassic strata of the Colorado Plateau where abundance of Triassic–Jurassic zircons have been documented (Dickinson and Gehrels, 2008; Dickinson and Gehrels, 2009b).

5.2.2. Refined provenance interpretationIdentification of multiple cooling histories for each zircon U–Pb

age component, provides critical tectono-thermal constraints for an improved interpretation of sediment source. A multi-component analysis of zircon U–Pb–He ages was developed as a workflow to investigate complex sediment sources for each embayment in the northern GOM (Fig. 9).

5.2.2.1. Greater Rio Grande Embayment provenance In the GRG Em-bayment, 58 of 792 zircons (∼7%) have zircon U–Pb ages of 40–24 Ma, indicating a likely source from the contemporary mid-Cenozoic volcanic arc terranes in western North America (Chapin et al., 2004; Xu et al., 2017; Fig. 9). An alternative source could be the volcanic-rich Eocene–Oligocene strata on the southern Great Plains, where deep erosion (>1 km) occurred during the early Miocene (Reiners et al., 2005; Cather et al., 2012). The volcanic-source hypothesis is also supported by the abundance of volcanic rock fragments in the lower Miocene Oakville Formation of the Texas coastal plain (Dutton et al., 2012).

A few grains (n = 6) showing 80–40 Ma crystallization and cooling ages (Fig. 10A) that appear to be related to Laramide mag-

matism in the western U.S., likely sourced from the Colorado Min-eral Belt (Chapin, 2012) or the Laramide porphyry copper province in southern New Mexico and northern Mexico (Chapin et al., 2004). Two Triassic–Jurassic grains (229 and 198 Ma; Fig. 10A) could be from the Late Triassic–Jurassic Arc in the Chihuahua Fold Belt in northern Mexico (Villarreal et al., 2014). Alternative source would be recycled from the Triassic–Jurassic strata of the Colorado Plateau, such as the Triassic Chinle Formation (Fig. 2).

Most of the Mid-Continent and Yavapai–Mazatzal zircons record ZHe ages related to Sevier–Nevadan and older orogenies (Fig. 10A) and are likely recycled from the Paleozoic–early Cenozoic sedi-mentary cover in the western U.S. (Fig. 2). The remaining Mid-Continent and Yavapai–Mazatzal zircons that cooled during the Laramide Orogeny might be from the southern Rocky Mountains. Grenville zircons can be sourced from the basement of Llano up-lift (ZHe age >500 Ma) and northern Mexico (ZHe: 40–80 Ma). Grenville zircons cooled in Paleozoic–early Mesozoic are likely re-cycled from the Paleozoic–early Cenozoic sedimentary strata in the western U.S. Cordillera. In total, 13 different cooling episodes can be identified for the four major zircon U–Pb components in GRG Embayment, resulting in less ambiguous source area identifica-tion and thus higher resolution of sediment source interpretation (Fig. 9). When coupled U–Pb and (U–Th)/He zircon data from basement and supracrustal rocks within the Miocene Gulf of Mex-ico catchment become available, an integrated work would reveal more unambiguous sediment delivery pathways from source to sink.

5.2.2.2. Houston Embayment provenance Zircon U–Pb–He double-dating on Houston Embayment samples yields a similar overall pattern as in the GRG Embayment, but with some important dif-ferences. Grenville zircons are mostly cooled during the Paleozoic and are likely recycled from the sedimentary strata in western U.S. or the Ouachita foreland basin. Only one Grenville records a Pre-cambrian ZHe age (817 Ma; Fig. 10B) while no Laramide cooling ages (80–40 Ma) were found for Grenville zircon, suggesting that there was no or very subordinate drainage connection to Grenville

54 J. Xu et al. / Earth and Planetary Science Letters 475 (2017) 44–57

Fig. 10. Sedimentary provenance interpretation of lower Miocene strata, northern Gulf of Mexico, based on detrital zircon U–Pb and (U–Th)/He double dating. Sample locations are shown in Fig. 2. The color bar highlights major crustal crystallization and cooling events of source terranes in North America. CZ – Cenozoic. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

exposures in the Llano Uplift or in western Texas and northern Mexico. This is consistent with previous paleogeographic recon-structions that suggested the sediment supply in this region was mostly fed by the Red River carrying sediments from the south-ern Rocky Mountains (Galloway et al., 2011). Paucity of carbonate rock fragments found in this region also supports limited to no-existing sediment input from the Edwards Plateau in central Texas (Dutton et al., 2012). An increased in Mid-Continent and Yavapai–Mazatzal zircons in the Houston Embayment (Fig. 4), associated

with Laramide ZHe cooling ages (Fig. 10B), strongly points to an increasing sediment supply from the southern Rocky Mountains compared to the GRG Embayment.

5.2.2.3. Mississippi Embayment provenance Mississippi Embayment samples have distinctly different U–Pb–He age combinations com-pared to samples in the GRG and Houston embayments (Fig. 10C). Most of the ZHe cooling ages for Grenville grains are in the range of 500–260 Ma. Such a combination of Grenville U–Pb age and

J. Xu et al. / Earth and Planetary Science Letters 475 (2017) 44–57 55

Appalachian–Ouachita ZHe age (Fig. 10C) strongly points to a pri-mary source from the Appalachian Mountains and its foreland basin strata. However, a few of these grains might have been recycled from the Paleozoic–Cenozoic sedimentary strata in the western U.S. (e.g., Navajo sandstone; Rahl et al., 2003), while still ultimately sourced from the Appalachian terranes.

The Yavapai–Mazatzal and Mid-Continent grains yield ZHe ages predominantly related to the Ancestral Rocky Mountains, Sevier–Nevadan, and Laramide orogenies in the western U.S. (Fig. 10C). If the limited data are broadly representative, it appears that the Laramide Orogeny provided most of the Yavapai–Mazatzal and Mid-Continent zircons, whereas the Ancestral Rocky Mountains and Sevier–Nevada orogenies provided fewer grains for the sed-iment transported to the Mississippi Embayment during Miocene time

5.2.2.4. Eastern Gulf of Mexico Embayment provenance Given the dominance of zircons with Grenville and Appalachian–Ouachita U–Pb ages in the Eastern GOM Embayment samples (Fig. 4), ZHe dating was mainly performed on these grains (Fig. 10D). East-ern GOM Embayment samples exhibit makedly much younger ZHe ages, ranging from 250 to 150 Ma. These ages are likely related to the rifting of Pangea and the opening of the Central Atlantic (Weislogel et al., 2015). This seems to be supported by several sed-iment flux episodes recording high rates of sedimentation in the Mesozoic both in the GOM and in the Atlantic margins (Poag and Sevon, 1989; Galloway, 2008).

6. Conclusions

This study presents the first regional application of detrital zircon U–Pb and (U–Th)/He double-dating to unravel sediment routing from both primary and sedimentary sources to the deposi-tional sink for the Miocene system of GOM. Lower Miocene strata are dominated by detrital zircons from the Yavapai–Mazatzal, Mid-Continent, Grenville, Appalachian–Ouachita, and Western Cordillera Arc terranes, as well as very few zircons sourced from the Wyoming–Superior and Pan-African terranes. In addition to the zir-con U–Pb signatures, ZHe analyses reveal sediment generated from exhumation during Proterozoic Grenville, Paleozoic Appalachian–Ouachita and Ancestral Rocky Mountains, Mesozoic Sevier–Nevada, Mesozoic–Cenozoic Laramide orogenies, and Cordilleran and mid-Cenozoic volcanism.

The U–Pb–He double-dating technique is especially powerful for differentiating the multiple sources of cosmopolitan Grenville zir-cons in North America. Four distinct U–Pb–He age combinations were recognized for Grenville zircons. Sources for Grenville zir-cons that are possibly derived directly from crystalline terranes include the Llano Uplift in central Texas (ZHe age >600 Ma), Grenville exposures in western Texas and northern Mexico (ZHe age: 80–40 Ma), and Appalachian terranes (ZHe age: 250–150 Ma). Multi-cycled Grenville zircons (ZHe age: 500–260 Ma) originated from Grenville exposures in eastern North America, either de-posited in Paleozoic strata in the Appalachian–Ouachita foreland basin or transported and deposited in Paleozoic–Cenozoic strata of the mid-continent and western interior, and then eroded and recy-cled again during Miocene times.

The combined U–Pb–He approach provides both crystallization and cooling ages of source areas, adding a critical tectonic dimen-sion to provenance analysis. As coupled U–Pb and (U–Th)/He zir-con results become available from basement rocks and supracrustal cover present within the Miocene catchment of the Gulf of Mex-ico, it will become possible to confirm the thermal histories of sediment sources outlined in this study (Fig. 9). Integrated study of this nature will greatly increase understanding of the complex

sediment source and dispersal in large passive margin basins like the Gulf of Mexico Basin.

Acknowledgements

This research was funded in part by an internal Jackson School of Geosciences Energy Theme Seed Grant to Stockli and Snedden. We are grateful for an Ed Picou Fellowship from Gulf Coast Section SEPM (GCSSEPM), which helped fund our fieldwork. We also thank the Institute for Geophysics at The University of Texas at Austin, for providing an Ewing–Worzel Graduate Fellowship. Field sample col-lection was facilitated by Drs. Gary Kinsland and Robert Hatcher. The Florida Geological Survey and UT’s Bureau of Economic Ge-ology allowed us to sample subsurface cores. We also appreci-ate the assistance of Lisa Stockli, Spencer Seman, Daniel Arnost, Adam Goldsmith, Edgardo J Pujols, and Shanping Liu for their help with U–Pb and (U–Th)/He double dating analyses. We thank the members of the Gulf Basin Depositional Synthesis (GBDS) Indus-trial Associates Program and project manager Patricia Ganey-Curry. Constructive comments by two anonymous reviewers are appreci-ated and helped greatly to improve this manuscript.

Appendix A. Supplementary material

Supplementary material related to this article can be found on-line at http://dx.doi.org/10.1016/j.epsl.2017.07.024.

References

Becker, T.P., Thomas, W.A., Samson, S.D., Gehrels, G.E., 2005. Detrital zircon evidence of Laurentian crustal dominance in the lower Pennsylvanian deposits of the Al-leghanian clastic wedge in eastern North America. Sediment. Geol. 182, 59–86.

Bickford, M.E., Van Schmus, W.R., Zietz, I., 1986. Proterozoic history of the midcon-tinent region of North America. Geology 14, 492–496. http://dx.doi.org/10.1130/0091-7613(1986)14<492:PHOTMR>2.0.CO;2.

Blum, M., Pecha, M., 2014. Mid-Cretaceous to Paleocene North American drainage reorganization from detrital zircons. Geology 42, 607–610. http://dx.doi.org/10.1130/G35513.1.

Boettcher, S.S., Milliken, K.L., 1994. Mesozoic–Cenozoic unroofing of the southern Appalachian Basin: apatite fission track evidence from Middle Pennsylvanian Sandstones. J. Geol. 102, 655–668. http://dx.doi.org/10.1086/629710.

Bush, M.A., Horton, B.K., Murphy, M.A., Stockli, D.F., 2016. Detrital record of initial basement exhumation along the Laramide deformation front, southern Rocky Mountains. Tectonics 35, 2117–2130.

Cather, S.M., Chapin, C.E., Kelley, S.A., 2012. Diachronous episodes of Cenozoic ero-sion in southwestern North America and their relationship to surface uplift, paleoclimate, paleodrainage, and paleoaltimetry. Geosphere 8 (6), 1177–1206.

Chapin, C.E., Cather, S.M., 1994. Tectonic setting of the axial basins of the northern and central Rio Grande rift. In: Keller, G.R., Cather, S.M. (Eds.), Basins of the Rio Grande Rift; Structure, Stratigraphy and Tectonic Setting. In: Geological Society of America Special Paper, vol. 291, pp. 5–25.

Chapin, C.E., Wilks, M., McIntosh, W.C., 2004. Space–time patterns of late Cretaceous to present magmatism in New Mexico – Comparison with Andean volcanism and potential for future volcanism: Socorro. N. M. Bur. Geol. Miner. Resour. Bull. 160, 13–40.

Chapin, C.E., 2012. Origin of the Colorado Mineral Belt. Geosphere 8, 28–43. http://dx.doi.org/10.1130/GES00694.1.

Chen, J.H., Moore, J.G., 1982. Uranium-lead isotopic ages from the Sierra Nevada Batholith, California. J. Geophys. Res. 87 (B6), 4761–4784. http://dx.doi.org/10.1029/JB087iB06p04761.

Clements, B., Sevastjanova, I., Hall, R., Belousova, E.A., Griffin, W.L., Pearson, N., 2012. Detrital zircon U–Pb age and Hf-isotope perspective on sediment provenance and tectonic models in SE Asia. In: Rasbury, E.T., Hemming, S.R., Riggs, N.R. (Eds.), Mineralogical and Geochemical Approaches to Provenance. In: Geologi-cal Society of America Special Paper, vol. 487, pp. 37–61.

Craddock, W.H., Kylander-Clark, A.R.C., 2013. U–Pb ages of detrital zircons from the Tertiary Mississippi River Delta in central Louisiana: insights into sediment provenance. Geosphere 9, 1832–1851. http://dx.doi.org/10.1130/GES00917.1.

DeCelles, P.G., 2004. Late Jurassic to Eocene evolution of the Cordilleran thrust belt and foreland basin system, western USA. Am. J. Sci. 304, 105–168. http://dx.doi.org/10.2475/ajs.304.2.105.

Dickinson, W.R., 1985. Interpreting provenance relations from detrital modes of sandstones. In: Zuffa, G.G. (Ed.), Provenance of Arenites. D. Reidel Publishing Company, Boston, pp. 333–361.

56 J. Xu et al. / Earth and Planetary Science Letters 475 (2017) 44–57

Dickinson, W.R., 2004. Evolution of the North American Cordillera. Annu. Rev. Earth Planet. Sci. 32, 13–45. http://dx.doi.org/10.1146/annurev.earth.32.101802.120257.

Dickinson, W.R., 2008. Impact of differential zircon fertility of granitoid base-ment rocks in North America on age populations of detrital zircons and implications for granite petrogenesis. Earth Planet. Sci. Lett. 275, 80–92. http://dx.doi.org/10.1016/j.epsl.2008.08.003.

Dickinson, W.R., Gehrels, G.E., 2003. U–Pb ages of detrital zircons from Per-mian and Jurassic eolian sandstones of the Colorado Plateau, USA: paleo-geographic implications. Sediment. Geol. 163, 29–66. http://dx.doi.org/10.1016/S0037-0738(03)00158-1.

Dickinson, W.R., Gehrels, G.E., 2008. U–Pb ages of detrital zircons in relation to pa-leogeography: Triassic paleodrainage networks and sediment dispersal across southwest Laurentia. J. Sediment. Res. 78, 745–764. http://dx.doi.org/10.2110/jsr.2008.088.

Dickinson, W.R., Gehrels, G.E., 2009a. Use of U–Pb ages of detrital zircons to in-fer maximum depositional ages of strata: a test against a Colorado Plateau Mesozoic database. Earth Planet. Sci. Lett. 288 (1–2), 115–125. http://dx.doi.org/10.1016/j.epsl.2009.09.013.

Dickinson, W.R., Gehrels, G.E., 2009b. U–Pb ages of detrital zircons in Jurassic eolian and associated sandstones of the Colorado Plateau: evidence for transcontinen-tal dispersal and intraregional recycling of sediment. Geol. Soc. Am. Bull. 121, 408–433. http://dx.doi.org/10.1130/B26406.1.

Dickinson, W.R., Lawton, T.F., 2001. Carboniferous to Cretaceous assembly and frag-mentation of Mexico. Geol. Soc. Am. Bull. 113, 1142–1160. http://dx.doi.org/10.1130/0016-7606(2001)113<1142:CTCAAF>2.0.CO;2.

Dickinson, W.R., Lawton, T.F., 2003. Sequential intercontinental suturing as the ulti-mate control for Pennsylvanian Ancestral Rocky Mountains deformation. Geol-ogy 31, 609–612. http://dx.doi.org/10.1130/0091-7613(2003)0312.0.CO;2.

Dickinson, W.R., Suczek, C.A., 1979. Plate tectonics and sandstone compositions. Am. Assoc. Pet. Geol. Bull. 63, 2167–2182.

Ducea, M., 2001. The California arc: thick granitic batholiths, eclogitic residues, lithospheric-scale thrusting and magmatic flare-ups. GSA Today 11 (11), 4–10. http://dx.doi.org/10.1130/1052–5173(2001)011<0004:TCATGB>2.0.CO;2.

Dutton, S.P., Loucks, R.G., Day-Stirrat, R.J., 2012. Impact of regional variation in detri-tal mineral composition on reservoir quality in deep to ultradeep lower Miocene sandstones, western Gulf of Mexico. Mar. Pet. Geol. 35, 139–153.

Ewing, T.E., 2005. Phanerozoic development of the Llano Uplift. S. Tex. Geol. Soc. Bull. 45, 15–25.

Fildani, A., McKay, M.P., Stockli, D., Clark, J., Dykstra, M.L., Stockli, L., Hessler, A.M., 2016. The ancestral Mississippi drainage archived in the late Wisconsin Missis-sippi deep-sea fan. Geology 44 (6), 479–482.

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, 1743–1774.

Galloway, W.E., 2008. Depositional evolution of the Gulf of Mexico sedimentary basin. In: Hsü, K.J. (Ed.), Sedimentary Basins of the WorldIn: Miall, A.D. (Ed.), The Sedimentary Basins of the United States and Canada, vol. 5. Elsevier, The Netherlands, pp. 505–549.

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, 938–973. http://dx.doi.org/10.1130/GES00647.1.

Gehrels, G., 2011. Detrital zircon U–Pb geochronology: current methods and new opportunities. In: Busby, C., Azor, A. (Eds.), Recent Advances in Tectonics of Sed-imentary Basins. Blackwell Publishing, Hoboken, New Jersey, pp. 47–62.

Gehrels, G.E., Blakey, R., Karlstrom, K.E., Timmons, J.M., Dickinson, B., Pecha, M., 2011. Detrital zircon U–Pb geochronology of Paleozoic strata in the Grand Canyon, Arizona. Lithosphere 3, 183–200. http://dx.doi.org/10.1130/L121.1.

Gehrels, G., 2014. Detrital zircon U–Pb geochronology applied to tectonics. Annu. Rev. Earth Planet. Sci. 42 (1), 127–149.

Gehrels, G., Pecha, M., 2014. Detrital zircon U–Pb geochronology and Hf isotope geochemistry of Paleozoic and Triassic passive margin strata of western North America. Geosphere 10 (1), 49–65. http://dx.doi.org/10.1130/GES00889.1.

Hatcher Jr., R.D., 2010. The Appalachian orogen: a brief summary. In: Tollo, R.P., Bartholomew, M.J., Hibbard, J.P., Karabinos, P.M. (Eds.), From Rodinia to Pangea: The Lithotectonic Record of the Appalachian Region. In: Geological Society of America Memoir, vol. 206, pp. 1–19.

Hawkesworth, C.J., Kemp, A.I.S., 2006. Evolution of the continental crust. Nature 443, 811–817.

Hoffman, P.F., 1988. United plates of America, the birth of a craton: Early Proterozoic assembly and growth of Laurentia. Annu. Rev. Earth Planet. Sci. 16, 543–603. http://dx.doi.org/10.1146/annurev.ea.16.050188.002551.

Howard, A.L., Farmer, G.L., Amato, J.M., Fedo, C.M., 2015. Zircon U–Pb ages and Hf isotopic compositions indicate multiple sources for Grenvillian detrital zircon deposited in western Laurentia. Earth Planet. Sci. Lett. 432, 300–310.

Kelley, S.A., Chapin, C.E., 1995. Apatite fission-track thermochronology of southern Rocky Mountain–Rio Grande rift–western High Plains provinces. In: Bauer, P.W., et al. (Eds.), Geology of the Santa Fe Region. In: New Mexico Geological Society Field Conference Guidebook, vol. 46, pp. 87–96.

Kluth, C.F., 1986. Plate tectonics of the Ancestral Rocky Mountains. In: Peterson, J.A. (Ed.), Paleotectonics and Sedimentation in the Rocky Mountain Region,

United States. In: American Association of Petroleum Geologists Memoir, vol. 41, pp. 353–369.

Lawton, T.F., 2008. Laramide sedimentary basins. In: Miall, A.D., Hsü, K.J. (Eds.), Sedi-mentary Basins of the World, vol. 5. The Sedimentary Basins of the United States and Canada. Elsevier, The Netherlands, pp. 429–450.

Lawton, T.F., Bradford, B.A., 2011. Correlation and provenance of Upper Cretaceous (Campanian) fluvial strata, Utah, U.S.A., from zircon U-Pb geochronology and petrography. J. Sediment. Res. 81, 495–512.

Lawton, T.F., Bradford, I.A., Vega, F.J., Gehrels, G.E., Amato, J.M., 2009. Provenance of Upper Cretaceous–Paleogene sandstones in the foreland basin system of the Sierra Madre Oriental, northeastern Mexico, and its bearing on fluvial dispersal systems of the Mexican Laramide Province. Geol. Soc. Am. Bull. 121, 820–836. http://dx.doi.org/10.1130/B26450.1.

Leier, A.L., Gehrels, G.E., 2011. Continental-scale detrital zircon provenance signa-tures in Lower Cretaceous strata, western North America. Geology 39, 399–402. http://dx.doi.org/10.1130/G31762.1.

Mackey, G.N., Horton, B.K., Milliken, K.L., 2012. Provenance of the Paleocene–Eocene Wilcox Group, western Gulf of Mexico basin: evidence for integrated drainage of the southern Laramide Rocky Mountains and Cordilleran arc. Geol. Soc. Am. Bull. 124, 1007–1024. http://dx.doi.org/10.1130/B30458.1.

Marsh, J.H., Stockli, D.F., 2015. Zircon U–Pb and trace element zoning characteristics in an anatectic granulite domain: insights from LASS-ICP-MS depth profiling. Lithos 239, 170–185.

Moecher, D.P., Samson, S.D., 2006. Differential zircon fertility of source terranes and natural bias in the detrital zircon record: implications for sedimentary prove-nance analysis. Earth Planet. Sci. Lett. 247, 252–266. http://dx.doi.org/10.1016/j.epsl.2006.04.035.

Painter, C.S., Carrapa, B., DeCelles, P.G., Gehrels, G.E., Thomson, S.N., 2014. Ex-humation of the North American Cordillera revealed by multi-dating of Up-per Jurassic–upper Cretaceous foreland basin deposits. GSA Bull. 126 (11/12), 1439–1464.

Park, H., Barbeau Jr., D.L., Rickenbaker, A., Bachmann-Krug, D., Gehrels, G., 2010. Application of foreland basin detrital-zircon geochronology to the reconstruction of the Southern and Central Appalachian orogen. J. Geol. 118 (1), 23–44. http://dx.doi.org/10.1086/648400.

Pazzaglia, F.J., Kelley, S.A., 1998. Large-scale geomorphology and fission-track ther-mochronology in topographic and exhumation reconstructions of the South-ern Rocky Mountains. Rocky Mt. Geol. 33, 229–257. http://dx.doi.org/10.2113/33.2.229.

Pazzaglia, F.J., Carter, M., et al., 2015. Geomorphology, active tectonics, and land-scape evolution in the Mid-Atlantic region. In: Brezinski, D.K., Halka, J.P., Ortt Jr., R.A. (Eds.), Tripping from the Fall Line: Field Excursions for the GSA An-nual Meeting. Baltimore, 2015. In: Geological Society of Field Guide, vol. 40, pp. 109–169.

Petrus, J.A., Kamber, B.S., 2012. VizualAge: a novel approach to laser ablation ICP-MS U–Pb geochronology data reduction. Geostand. Geoanal. Res. 36, 247–270. http://dx.doi.org/10.1111/j.1751-908X.2012.00158.x.

Poag, C.W., Sevon, W.D., 1989. A record of Appalachian denudation in post-rift Mesozoic and Cenozoic sedimentary deposits of the U.S. middle Atlantic con-tinental margin. Geomorphology 2, 119–157. http://dx.doi.org/10.1016/0169–555X(89)90009–3.

Rahl, J.M., Reiners, P.W., Campbell, I.H., Nicolescu, S., Allen, C.M., 2003. Combined single-grain (U–Th)/He and U/Pb dating of detrital zircons from the Navajo Sandstone, Utah. Geology 31, 761–764. http://dx.doi.org/10.1130/G19653.1.

Rainbird, R.H., Heaman, L.M., Young, G., 1992. Sampling Laurentia: detrital zircon geochronology offers evidence for an extensive Neoproterozoic river system originating from the Grenville orogen. Geology 20, 351–354.

Reiners, P.W., Campbell, I.H., Nicolescu, S., Allen, C.M., Hourigan, J.K., Garver, J.I., Mat-tinson, J.M., Cowan, D.S., 2005. (U–Th)/(He–Pb) double dating of detrital zircons. Am. J. Sci. 305 (4), 259–311.

Sharman, G.R., Covault, J.A., Stockli, D.S., Wroblewski, A.F., Bush, M.A., 2017. Early Cenozoic drainage reorganization of the United States Western interior – Gulf of Mexico sediment routing system. Geology 45 (2), 187–190.

Stockli, D.F., 2017. Texas three Step – the exhumation and exposure history of the Llano Uplift along the Laurentia and Gulf of Mexico rift margins. Geol. Soc. Am. Abstr. Progr. 49 (1).

Thomas, W.A., 2011. Detrital–zircon geochronology and sedimentary provenance. Lithosphere 3, 304–308. http://dx.doi.org/10.1130/RF.L001.1.

Wahl, P.J., Yancey, T.E., Pope, M.C., Miller, B.V., Ayers, W.B., 2016. U–Pb detrital zircon geochronology of the upper Paleocene to lower Eocene Wilcox Group, east–central Texas. Geosphere 12, 1517–1531. http://dx.doi.org/10.1130/GES01313.1.

Vermeesch, P., 2004. How many grains are needed for a provenance study? Earth Planet. Sci. Lett. 224, 441–451. http://dx.doi.org/10.1016/j.epsl.2004.05.037.

Villarreal-Fuentes, J., Levresse, G., Nieto-Samaniego, A.F., Corona-Esquivel, R., 2014. New geological and geochronological data of the Placer de Guadalupe uplift, Mexico: a new piece of the Late Triassic–Jurassic Nazas Arc? Int. Geol. Rev. 56 (16), 2000–2014.

Weislogel, A.L., Hunt, B., Lisi, A., Lovell, T., Robinson, D.M., 2015. Detrital zircon provenance of the eastern Gulf of Mexico subsurface: constraints on Late Juras-sic paleogeography and sediment dispersal of North America. In: Anderson, T.H.,

J. Xu et al. / Earth and Planetary Science Letters 475 (2017) 44–57 57

Didenko, A.N., Johnson, C.L., Khanchuk, A.I., MacDonald Jr., J.H. (Eds.), Late Juras-sic Margin of Laurasia—A Record of Faulting Accommodating Plate Rotation. In: Geological Society of America Special Paper, vol. 513.

Whitmeyer, S.J., Karlstrom, K.E., 2007. Tectonic model for the Proterozoic growth of North America. Geosphere 3, 220–259. http://dx.doi.org/10.1130/GES00055.1.

Wolfe, M., Stockli, D.F., 2010. Zircon (U–Th)/He thermochronometry in the KTB drill hole, Germany, and its implications for bulk He diffusion kinetics in zircon. Earth Planet. Sci. Lett. 295, 69–82. http://dx.doi.org/10.1016/j.epsl.2010.03.025.

Xie, X., O’Connor, P.M., Alsleben, H., 2016. Carboniferous sediment dispersal in the Appalachian–Ouachita juncture: provenance of selected late Mississippian sandstones in the Black Warrior Basin, Mississippi, United States. Sediment. Geol. 342, 191–201.

Xu, J., Snedden, J.W., Stockli, D.F., Fulthorpe, C.S., Galloway, W.E., 2017. Early Miocene continental-scale sediment supply to the Gulf of Mexico Basin based on detrital zircon analysis. Geol. Soc. Am. Bull. 129 (1–2), 3–22.