The Cretaceous Cordilleran foreland basin strata exposed in the Book Cliffs of eastern Utah and western Colorado have motivated important concepts linking thrust belt deformation and foreland basin evolution largely on the basis of sequence stratigraphy, stratal architecture, and sediment provenance evolution. However, these methods and approaches generally cannot provide critical insights into the temporal or causal linkages between foreland basin architecture and thrust belt deformation. This is in part due to discrepancies in age resolution and lack of evidence with which to directly couple sediment supply and basin-fill evolution to thrust belt unroofing. New detrital zircon (DZ) geothermochronometric data from Upper Cretaceous proximal to distal foreland basin strata in the Book Cliffs provide new quantitative insights into sediment origin and dispersal in relation to thrust belt deformation and exhumation. Detailed DZ U-Pb and (U-Th)/He double dating reveals that the Book Cliffs foredeep detritus was mainly delivered by transverse routing systems from two major sources: (1) Neoproterozoic and Lower Paleozoic strata from the central Utah Sevier thrust belt, and (2) Permian–Jurassic and synorogenic Cretaceous strata recycled from the frontal part of the thrust belt. A dramatic increase in Sierran magmatic arc and Yavapai-Mazatzal DZ U-Pb ages, as well as Paleozoic DZ He ages, in the deeper marine portions of the foreland basin points to axial fluvial and littoral sediment input from the Sierran magmatic arc and Mogollon highland sources. Both transverse and axial transport systems acted contemporaneously during eastward propagation of the Late Cretaceous thrust belt. DZ He depositional lag time estimates reveal three distinct exhumation pulses in the Sevier thrust belt in the Cenomanian and Campanian. The exhumation pulses correlate with shifts in sediment provenance, dispersal style, and progradation rates in the foreland basin. These new data support conceptual models that temporally and causally link accelerated exhumation and unroofing in the thrust belt to increases in sediment supply and rapid clastic progradation in the foreland basin.

While tectonic models of foreland basins directly link flexural subsidence to thrust belt loading, stratigraphic models based on lithostratigraphic and sequence stratigraphic concepts are commonly ambiguous about a direct and genetic association with thrust belt deformation. This is largely due to limitations in constraining and correlating the timing of crustal deformation and tectonic unroofing to changes in sediment supply, clastic progradation, and basin-fill evolution. Moreover, external forcing factors such as tectonic activity and loading, sediment supply, flexural subsidence, eustasy, climate, and autogenic processes can result in the generation of similar stratal architecture and stacking patterns. Hence, the mechanisms modulating foreland basin evolution and stratigraphic architecture are not easily unraveled (e.g., Flemings and Jordan, 1989; Dickinson, 1996; Aschoff and Steel, 2011a). As a consequence, the impact of tectonics on foreland stratigraphic architecture has been debated for decades, and the interplay of these forcing factors and their competing and compensatory relationships remain poorly understood (e.g., Burbank et al., 1988; Heller et al., 1988; Heller and Paola, 1989; Kamola and Huntoon, 1995; Houston et al., 2000; DeCelles, 2004; Horton et al., 2004; Aschoff and Steel, 2011a).

Two contrasting hypotheses have been postulated for the temporal and causal feedbacks between tectonic activity within the fold-and-thrust belt and clastic progradation in the foreland basin. (1) Heller et al. (1988) proposed a two-phase model in which coarse clastic sediment accumulation synchronous with active thrusting is restricted to the proximal basin foredeep due to significant flexural subsidence. This pulse in thrust belt deformation is followed by reduced tectonic activity, major isostatic compensation, thrust belt erosion, and long-distance coarse clastic sediment progradation into the foreland basin. This second phase is considered to be an interval of enhanced erosion and sediment dispersal during tectonic “quiescence” associated with a slowdown in flexural subsidence and accommodation (Heller et al., 1988; Aschoff and Steel, 2011a). (2) In a contrasting model, detailed studies in the Cordilleran foreland basin of central Utah have attributed parasequence cyclicity in the stratigraphic stacking patterns, abrupt changes in shoreline trajectory, and stratal architecture directly to flexural pulses driven by synchronous deformation and thrust belt advances (Burbank et al., 1988; Kamola and Huntoon, 1995; Houston et al., 2000; Aschoff and Steel, 2011a). However, this causal linkage between thrust belt deformation and foreland sedimentation in these models heavily relies on the assumptions of constant sediment supply and accommodation controls external to the sediment routing system. These assumptions are difficult to corroborate but of fundamental importance for understanding the evolution of stacking patterns and the influence of tectonism on foreland architecture (Wehr, 1993; Kamola and Huntoon, 1995; Leva López et al., 2014).

Petrographic and isotopic provenance studies have documented syntectonic coarse clastic sediment deposition in the proximal foredeep by associating major changes in sediment composition with tectonic unroofing in the frontal thrust belt (Jefferson, 1982; Lawton, 1986b; Horton et al., 2004). However, evidence for this linkage remains elusive in the more distal, nonmarine facies, partially because they lack high-resolution temporal constraints. In addition, multiple studies have shown that fluvial and marine facies can be heavily influenced by axial or littoral sediment transport, sediment recycling, and drainage reorganization, adding spatial intricacy to sediment derivation and dispersal patterns (e.g., Lawton et al., 2003, 2014; Jinnah et al., 2009; Szwarc et al., 2015).

This study employed detrital zircon (DZ) U-Pb and (U-Th)/He (He) double-dating methods on Late Cretaceous foredeep strata in the Book Cliffs portion of the Cordilleran foreland basin to outline the temporal and causal relationships among sedimentation, stratal architecture (distal fluvial and marine), and thrust belt unroofing. These new geo- and thermochronometric data provide new insights into (1) sediment provenance and temporal changes in sources, (2) the extent of transverse and axial fluvial and marine sediment transport, and (3) the temporal and causal linkages among frontal thrust belt deformation, unroofing, and clastic wedge progradation into the distal foreland basin.

Tectonic Setting

The Mesozoic to early Cenozoic evolution of the Sevier foreland basin was primarily associated with the development and eastward migration of contractional retroarc deformation and orogenesis along the convergent western margin of North America (Armstrong, 1968; Currie, 1998; Dickinson, 2004). The Cordilleran orogeny formed in response to the Farallon oceanic plate's subduction beneath North America, resulting in nearly 100 m.y. of retroarc shortening (DeCelles, 2004). The Cordilleran foreland basin developed mainly in response to flexural loading during the propagation of the thin-skinned Cordilleran (Sevier) fold-and-thrust belt and was partitioned in the Late Cretaceous by Laramide basement-involved uplifts (Dickinson et al., 1988; Lawton, 2008; Lawton and Bradford, 2011). The early stages of Laramide deformation were concomitant with late-stage Sevier deformation and have been interpreted as the beginning of flat-slab subduction of the Farallon oceanic plate under the North America plate (Dickinson and Snyder, 1978; Miller et al., 1992; Saleeby, 2003). Cretaceous deformation in the frontal part of the thin-skinned Sevier fold-and-thrust was accommodated by the Canyon Range, Pavant, Paxton, and Gunnison thrusts and duplexes in central Utah (Fig. 1; Mitra et al., 1994; Mitra and Sussman, 1997; DeCelles, 2004; DeCelles and Coogan, 2006) and the Sheeprock, Tintic Valley, and Midas thrusts and the Charleston-Nebo salient in northern Utah (Mitra, 1994, 1997; Constenius et al., 2003; Kwon and Mitra, 2004). Both the central and northern portions of the Sevier fold-and-thrust belt are intricately contacted in the Leamington accommodation zone, north of Canyon Range (Kwon and Mitra, 2006).

In the Early to early Late Cretaceous, the Canyon Range, Pavant, Sheeprock, Tintic Valley, and Midas thrusts and duplexes accommodated over 100 km of shortening (e.g., Mitra, 1994; Mitra and Sussman, 1997; Kwon and Mitra, 2004; DeCelles and Coogan, 2006; Guenthner et al., 2014). Late Cretaceous foreland basin sedimentation was synchronous with the Paxton and Gunnison deformational phases, responsible for ∼30 km of shortening (DeCelles and Coogan, 2006). In northern Utah, palinspastic restorations bracket shortening on the Nebo and Charleston thrusts to 35 and 50 km, respectively (Constenius et al., 2003), or a combined <30 km (Kwon and Mitra, 2004), reflecting uncertainties in footwall cutoffs and styles of deformation. These thrusts combined with the Santaquin culmination and backthrusting in the frontal orogenic wedge were responsible for ∼100 km of Late Cretaceous shortening in central Utah (Constenius et al., 2003).

Stratigraphic Setting

The Late Cretaceous Book Cliffs strata in east-central Utah to western Colorado have been paramount for developing fundamental foreland basin and sequence stratigraphic concepts. Furthermore, this area has served as a key natural laboratory for promoting our understanding of the interplay between fold-and-thrust belt tectonics and foreland basin stratigraphy, including stacking patterns, stratal geometries, and the extent and expression of sequence-parasequence boundaries (e.g., Van Wagoner, 1991a, 1991b; Kamola and Huntoon, 1995; Van Wagoner, 1995; Yoshida et al., 1998; Houston et al., 2000; Yoshida, 2000). The strata exposed in the Book Cliffs are composed of a range of fluvial to marine sedimentary successions deposited in the foredeep of the Cordilleran foreland basin as part of the Western Interior Seaway and record fluctuations in sediment input, eustasy, and thrust-induced flexural subsidence (Fouch et al., 1983; Kamola and Huntoon, 1995; Houston et al., 2000; McLaurin and Steel, 2000, 2007). The overall Late Cretaceous foreland basin stratigraphy preserves a shallowing-upward and eastward-thinning progradational clastic wedge (Lawton, 1983). The base of this wedge is marked by a major early Turonian marine transgression, followed by late Turonian and Santonian tongues of fluvial to shallow-marine sandy facies known as the Ferron Sandstone and Emery Sandstone, respectively (e.g., Fisher et al., 1960; Hale, 1972; Fouch et al., 1983; Ryer, 1983; Edwards et al., 2005; Ketzer and Morad, 2006; Olariu et al., 2010; Fielding, 2011; Li et al., 2011). Detailed studies on the basin architecture, shoreline stacking trajectories, and ammonite biozones have helped to subdivide the Campanian clastic progradational wedge into three subwedges (Fouch et al., 1983; Cobban et al., 2006; Aschoff and Steel, 2011a). The lower clastic wedge (wedge A) is composed of all six members of the Blackhawk Formation and the Castlegate Sandstone (Fig. 2; Van Wagoner et al., 1990; Van Wagoner, 1991b; Aschoff and Steel, 2011a). The middle clastic wedge (wedge B) includes the Middle Castlegate and Buck Tongue units (including the distal Sego Sandstone and Neslen Formation), as well as the Corcoran and Cozzette Members of the Iles Formation, Mount Garfield, Colorado (Aschoff and Steel, 2011a). The upper clastic wedge (wedge C) includes the Bluecastle Tongue of the Castlegate Formation, the Lower Williams Fork Formation, and the Rollins Sandstone. The spatiotemporal framework delineated by Aschoff and Steel (2011a) suggests that wedge B had the fastest progradation rate (163–192 km/m.y.) compared to wedges A (30–38 km/m.y.) and C (56–81 km/m.y.). They hypothesized that the fast progradation and flat stacking trajectory of wedge B were the result of a reduction in regional accommodation space induced by the onset of Laramide basement-core uplift. This, coupled with an active thrust belt and high sediment supply, may explain the discrepancies in the style and rate of progradation between the wedges (Aschoff and Steel, 2011a).

Foreland Basin Sediment Composition and DZ Provenance

In the Cordilleran foreland basin of central-northern Utah, detrital compositional and DZ geo- and thermochronology studies have elucidated the unroofing history of the Sevier fold-and-thrust belt. The upper section of the Indianola Group (Funk Valley and Sixmile Canyon Formations) and the Canyon Range Conglomerate (Leamington Canyon and Pass Canyon Members) represent the coarse clastic proximal equivalents of the Book Cliffs foredeep strata (Fig. 2; e.g., Lawton, 1982; Lawton et al., 2007, and references therein).

The Indianola Group is characterized by a gradual up-section compositional transition from carbonate- to quartz-rich strata. The up-section increase in quartz clasts has been attributed to the progressive unroofing of Paleozoic Sevier fold-and-thrust belt passive-margin strata (e.g., Armstrong, 1968; Lawton, 1986b; DeCelles et al., 1995). However, due to the complex stratigraphic and structural evolution of the Sevier fold-and-thrust belt, not all proximal (wedge-top) Indianola Group deposits record clear clastic compositional trends indicative of progressive unroofing (e.g., Jefferson, 1982).

Progressive unroofing of the Canyon Range, Pavant, and Nebo thrust sheets is also reflected in the DZ U-Pb and DZ He age trends recorded in the Indianola Group (Pujols et al., 2020). The Sanpete and Funk Valley Formations (lower Indianola Group) are characterized by sediment mixing of Neoproterozoic to Mesozoic strata, likely sourced from Neoproterozoic–Paleozoic sections in the Canyon Range thrust sheet and Mesozoic–Paleozoic strata in the easternmost Pavant and Nebo thrust sheets (Pujols et al., 2020). Moving up section into the Sixmile Canyon Formation, the DZ U-Pb age spectra become dominated by Neoproterozoic–Paleozoic DZ U-Pb ages in response to unroofing of Neoproterozoic–Paleozoic strata of the Canyon Range and potentially the Sheeprock and Tintic thrust sheets in central and northern Utah, respectively (Pujols et al., 2020). The progression of unroofing is corroborated by a progressively younger trend of DZ He cooling ages from Permian–Carboniferous in the Sanpete Formation to Mesozoic DZ He ages in the Funk Valley and Sixmile Canyon Formations (Pujols et al., 2020). Active shortening and thrust belt unroofing in the Campanian are inferred from very short DZ He lag times in the Sixmile Canyon Formation.

The wedge-top deposits of the Canyon Range Conglomerate are defined by up-section alternations in clast composition with three petrofacies—quartzose, carbonate, and mixed (quartzose-carbonate). Lawton et al. (2007) subdivided the conglomerate into the Cow Canyon (oldest), Wild Horse Canyon, Leamington Canyon, Pass Canyon, and Wide Canyon (youngest) Members based on clast compositional variations, pedogenic intervals, and intraformational unconformities. Quartzose petrofacies increase up section and record the unroofing of Paleozoic and Neoproterozoic strata in the Canyon Range thrust sheet (DeCelles et al., 1995; Lawton et al., 2007). This is supported by DZ He ages that get younger up section, and by an up-section decrease in Paleozoic DZ U-Pb ages above the Cow Canyon Member (Pujols et al., 2020). The DZ He ages also chronicle a polyphase and progressive unroofing history of Paleozoic to Neoproterozoic thrust belt strata from the Turonian to the Campanian.

In the Wasatch Plateau and Book Cliffs (Fig. 1), detailed petrographic sandstone analysis of the Star Point Sandstone and the Blackhawk (Fishman et al., 2013) and Castlegate Formations (Lawton, 1983, 1986a; Horton et al., 2004) described sandstone compositions that are dominated by quartz with subordinate lithics fragments and minor feldspars (Lawton, 1983; Fishman et al., 2013). An abrupt increase in quartz grains marks the most significant petrographic difference between the Castlegate and Blackhawk Formations. The quartz-rich nature of the Castlegate Formation is in agreement with proximal equivalents, including the Sixmile Canyon and Castlegate–Price River (proximal conglomerates) Formations, and the Leamington Canyon and Pass Canyon Members (Lawton, 1986b; Horton et al., 2004; Lawton et al., 2007). Although compositional and paleocurrent data in the Wasatch Plateau and Book Cliffs clearly indicate an overall western derivation (e.g., Fouch et al., 1983; Lawton, 1986a; Horton et al., 2004; Sahoo and Ganid, 2015), the source of the major mineral fractions in the distal foreland basin deposits cannot be further resolved, as both the Canyon Range and Charleston-Nebo salient contain compositionally equivalent strata (Lawton, 1986b; Lawton et al., 2003; Horton et al., 2004; Lawton et al., 2007).

Numerous other regional provenance studies have employed DZ U-Pb geochronology to more distal Cordilleran foreland strata on the Colorado Plateau, including the Book Cliffs, with the purpose of tracing sediment sourcing from the Sevier orogenic belt, the magmatic arc, and Mogollon Highland sources (e.g., Dickinson and Gehrels, 2008a, 2009a; Lawton and Bradford, 2011; Dickinson et al., 2012; Bartschi et al., 2018; Pettit et al., 2019).

Detrital Zircon U-Pb and He Dating

Over the past decade, DZ U-Pb dating has seen an explosive growth in its application to provenance studies, with particular emphasis on reconstructing sediment sourcing and routing, paleodrainages and paleogeographic evolution, and the interplay between tectonics and sedimentation in source-to-sink systems. This approach is based on the correlation of DZ U-Pb age signatures with U-Pb crystallization ages from different, distinct source terranes (Reiners et al., 2005). In addition, the youngest DZ U-Pb age mode or grains can also be used to determine maximum deposition age (MDA) estimates (e.g., Dickinson and Gehrels, 2009c; Spencer et al., 2016). This is particularly powerful for strata in basins adjacent to active magmatic arcs and in the absence of a high-resolution biostratigraphic record, especially in nonmarine rocks (Stockli and Najman, 2020).

DZ U-Pb provenance analysis also has limitations due to either monotonous or complex source terrane U-Pb signatures and sediment recycling. Hence, we combined DZ U-Pb ages with (U-Th)/He dating on individual zircons to investigate the low-temperature cooling history of the sediment source region and to spatially and temporally link sedimentation to fold-and-thrust-belt exhumation (e.g., Rahl et al., 2003; Reiners et al., 2005; Pujols et al., 2018). With a nominal closure temperature of ∼180 °C and partial retention zone of ∼140–220 °C (Reiners et al., 2002; Stockli, 2005; Wolfe and Stockli, 2010; Guenthner et al., 2013), zircon (U-Th)/He ages are well suited to reconstruct major upper-crustal tectonic exhumation, but they are also sufficiently shielded from resetting due to burial heating (Stockli and Najman, 2020). Zircon He thermal sensitivity is mainly controlled by the mineral-specific He diffusion kinetics and cooling rates and modulated by the accumulated alpha dosage (damage) and grain size (e.g., Nasdala et al., 1995, 2001; Reiners and Farley, 2001; Palenik et al., 2003; Nasdala et al., 2004; Shuster and Farley, 2009; Goldsmith et al., 2012; Guenthner et al., 2014). In cases where DZ He ages are derived from detrital rocks, age inheritance and partial resetting can complicate source identification (Guenthner et al., 2014). The time span between cooling through closure temperature, exhumation, and deposition is determined by calculating the difference between the youngest DZ He ages and stratigraphic age (i.e., lag time estimate; Ruiz et al., 2004; Stockli and Najman, 2020). However, DZ grains with equal U-Pb and He ages within error (first-cycle volcanic grains) must be excluded in order to determine a meaningful lag time estimate and reconstruct source exhumation histories (Reiners et al., 2005; Saylor et al., 2012; Pujols et al., 2018). The DZ He ages and resulting lag time estimates provide insight into the timing and magnitude of active deformation and unroofing during syntectonic deposition (Saylor et al., 2012; Painter et al., 2014). DZ He ages indistinguishable from depositional age signal major tectonic cooling related to rapid thrust sheet emplacement and exhumation, likely associated with movement up a thrust ramp. Hence, the exhumation and progressive unroofing history of orogenic systems and accompanying syntectonic sediment dispersal processes can be delineated by employing DZ (U-Th)–(He-Pb) double-dating techniques within their proper structural and stratigraphic framework.

Zircon U-Pb and (U-Th)/He Analytical Procedure

Zircon (U-Th)/(He-Pb) double-dating analyses were performed at UTChron facilities at the University of Texas at Austin using the procedures detailed in Pujols et al. (2020). Zircons were separated using standard mineral separation methods. The unpolished zircons were mounted on acrylic disks for in situ laser-ablation U-Pb dating depth profiling to recover multiple U-Pb ages and the youngest growth zone from each grain. Zircon grains from each sample were analyzed at random using a Photon Machines Analyte G2 ATLex 300si ArF 193 nm excimer laser with a two-volume HelEx ablation cell and Element 2 inductively coupled plasma–mass spectrometer (ICP-MS). For the laser ablation, we used a 30 µm spot size diameter and 300 ablation pulses at a rate of 10 Hz, producing an ablation depth of ∼15–17 µm. To achieve a 95% confidence that we had incorporated DZ components representing more than 5% of the total population, we analyzed ∼120 zircon grains per sample (Vermeesch, 2004). The primary and secondary reference standards were GJ1 (Jackson et al., 2004) and Pak1 (in-house zircon standard ca. 42 Ma from Pakistan). U-Pb data were reduced using Iolite software and VizualAge (Paton et al., 2011; Petrus and Kamber, 2012). Only DZ ages with a percentage of discordance <30% were included in the detrital analysis. The 206Pb/207Pb ages were chosen over 206Pb/238U ages when older than 850 Ma to improve DZ U-Pb age precision (Gehrels et al., 2008).

Grains for DZ He double dating were selected as a function of DZ U-Pb age mode or components following the procedures of Wolfe and Stockli (2010) and Pujols et al. (2020). DZ U-Pb ages with a percentage of discordance greater than 10% and geometrically inadequate for standard FT alpha ejection correction were excluded (i.e., flat, broken, or rich in inclusions; Farley et al., 1996). U-Pb dated zircon grains were removed from the acrylic mounts and wrapped in Pt foil tubes for in-vacuum helium extraction by diode laser heating at temperatures of ∼1050 °C. The released gas was spiked with 3He for isotope dilution and purified using a Janis cryogenic trap, and the 4He/3He ratio was analyzed with a Blazers Prisma QMS-200 quadrupole mass spectrometer. Laser heating was repeated until 4He gas extractions were <1% of the total extracted 4He. All zircons were unpacked from Pt tubes and spiked with enriched and calibrated 235U/238U, 230Th/232Th, and 149Sm/147Sm tracer prior to two-step HF-HCl pressure vessel dissolution and analysis by solution ICP-MS using a Thermo Element2 ICP-MS with a 50 mm microconcentric nebulizer. Zircon (U-Th)/He ages were calculated using isotope dilution and age equations described in Vermeesch (2008). For DZ U-Pb and DZ He isotopic data, see Supplemental Material Tables S2 and S3.1

Sampling Strategy

Coarse- to medium-grained sandstones were sampled from multiple stratigraphic units and Book Cliffs localities in Utah and Colorado (Fig. 1; Table S1 [see footnote 1]). The sampling was performed in proximal (Price Canyon, Utah) to distal (Grand Junction, Colorado) fluvial and marine deposits at an almost million-year stratigraphic resolution using the stratigraphic framework of Cobban et al. (2006). In ascending stratigraphic order, the stratigraphic units sampled included the Ferron and Emery Sandstones and the Star Point, Blackhawk, and Castlegate Formations and corresponding members (Fig. 2). In the southern part of the Piceance Basin, north of Grand Junction, Colorado, the Campanian Corcoran, Cozzette, and Rollins Sandstones from the Mount Garfield Formation (Iles Formation of Aschoff and Steel, 2011a) were sampled to correlate to the very distal time-equivalent DZ U-Pb populations (Fig. 1). For the same purpose, we collected lowstand tidal-dominated Meeker and Marapos units north of Meeker City, Colorado. Sample stratigraphic names and general location are displayed in Figures 1 and 2. For exact sample coordinates, see Table S1.

The DZ U-Pb and (U-Th)/He data in this study are grouped into stratigraphic transects from numerous geographic locations (Figs. 1 and 3). The formations, members, or units composing each stratigraphic transect and their DZ U-Pb and He modes are discussed in their corresponding section below. They are organized from proximal (Price River Canyon, Utah) to distal (Grand Junction, Colorado) locations (Fig. 3). Figure 3 shows kernel density estimator (KDE) representations of all DZ U-Pb and He age components for each sample and stratigraphic transect. KDEs and histograms were plotted using DensityPlotter 7.3 (Vermeesch, 2012) with histogram bin width and KDEs bandwidth indicated in the captions. All zircon U-Pb and (U-Th)/He data are tabulated in the Supplemental Material. Available youngest DZ U-Pb age modes or single ages are shown in Figure 3 and interpreted as the MDA.

The DZ U-Pb age results are discussed within the context of major North American tectonic provinces as described by Dickinson and Gehrels (2009a). These provinces include the Archean craton (3500–2420 Ma); cratonic Paleoproterozoic belts (2015–1810 Ma); Yavapai-Mazatzal orogenic belts (1810–1535 Ma); anorogenic Mesoproterozoic plutons (1535–1300 Ma); Grenville orogeny (1300–905 Ma); peri-Gondwanan terranes (725–510 Ma); Appalachian plutons (510–285 Ma); and Cordilleran-Sierran volcanic arcs (285–60 Ma). DZ (U-Th)/He age data are described in terms of geologic time periods and epochs as defined by the International Stratigraphic Chart version 2015/01 (Cohen et al., 2013, updated).

In general, six to eight DZ U-Pb age modes could be recognized in all samples and attributed to known magmatic phases in North America (e.g., Gehrels et al., 1995; Dickinson and Gehrels, 2003, 2009a, 2009b; Lawton et al., 2010; Lawton and Bradford, 2011). DZ grains from each major U-Pb mode were double-dated by (U-Th)/(He-Pb) for tectonic provenance identification, thermal analysis, and lag time estimation. First-cycle volcanic zircon grains were excluded from exhumation histories. No grains were rejected on the basis of alpha dosage because the DZ thermal history of each grain and alpha dosage influence on age could not be determined independently from the source area. Only a maximum dosage could be calculated.

Price River Canyon DZ U-Pb-He Age Transect

Samples from deltaic facies of the Santonian–Campanian Panther Tongue, Storrs Sandstone, and Spring Canyon Member contained all major North America DZ U-Pb age modes and were dominated by Grenville (19%–26%), Yavapai-Mazatzal (16%–20%), Appalachian plutons (12%–15%), and Cordilleran-Sierran volcanic arc (2%–12%) DZ U-Pb age modes (Fig. 3A). Higher in the stratigraphy, the coastal plain deposits of the Aberdeen Member to the fluvial Bluecastle Tongue yielded almost identical DZ U-Pb age modes, except for a reduction in Cordilleran-Sierran volcanic arc sources (0%–4%) and a slight increase in Grenville (23%–35%), and Yavapai-Mazatzal (19%–28%) ages (Fig. 3A).

The Panther Tongue deposits recorded Paleozoic to Mesozoic DZ He ages with no clear modes, whereas the Storrs Sandstone contained ∼34% of both Jurassic and Cretaceous DZ He ages, with modes at 180 and 122 Ma, respectively. The Spring Canyon and Aberdeen Members of the Blackhawk Formation and the Castlegate Sandstone yielded a larger Paleozoic DZ He age fraction (63%–58%) with distinct Carboniferous–Permian modes. From the Middle Castlegate to the Bluecastle Tongue, Mesozoic DZ He ages composed 52%–85% of the age mode, and late Paleozoic DZ He ages constituted 7%–39% (Fig. 3B), with similar modes at ca. 300, 218, and 90 Ma.

North of San Rafael Swell Zircon U-Pb-He Ages

Both the Ferron and the Emery deltaic sandstones showed prominent Cretaceous DZ U-Pb ages representing 33% and 18% of the populations, respectively, with lesser proportions of Yavapai-Mazatzal (15%–20%) and Grenville DZ U-Pb ages (17%) than younger units in the Book Cliffs. The Kenilworth Member samples were mostly dominated by three Precambrian DZ U-Pb age modes: Grenville (30%), anorogenic Mesoproterozoic plutons (22%), and Yavapai-Mazatzal (13%). The combined Appalachian Paleozoic plutons and Cordilleran-Sierran volcanic arc DZ U-Pb ages in this member represent less than 5% of the entire sampled population (Fig. 3B).

Samples north of the San Rafael Swell yielded predominantly Mesozoic and Paleozoic DZ He ages with a small percentage of Neoproterozoic DZ He ages. The Ferron Sandstone was characterized by Jurassic (37%) and Permian DZ He ages (16%). The Emery Sandstone yielded 26% Jurassic and Triassic DZ He ages and 12% Permian and Carboniferous DZ He ages. The Kenilworth Member had 25% Jurassic, 19% Carboniferous, and 22% Cretaceous DZ He ages. Two DZ He age modes were also identifiable, one that spanned from Jurassic to Cretaceous (ca. 146 Ma peak) and a second that was Permian–Carboniferous (ca. 302 Ma peak) in age. The Ferron and Emery Sandstones yielded between 6% and 10% Cretaceous DZ He ages (Fig. 3B).

Sunnyside Canyon Zircon U-Pb-He Ages

Overall, the DZ U-Pb age populations from this area contained Grenville (17%–38%), Yavapai-Mazatzal (11%–25%), anorogenic Mesoproterozoic plutons (8%–18%), Appalachian Paleozoic plutons (4%–13%), and Cordilleran-Sierran volcanic arc (4%–14%) ages. The Middle Castlegate and both Castlegate Sandstone samples were dominated by Yavapai-Mazatzal ages (20%–25%), whereas the rest of the stratigraphic section was dominated by Grenville DZ U-Pb ages (23%–38%; Fig. 3C). Both Castlegate Sandstone samples contained the largest DZ fraction of accreted peri-Gondwanan U-Pb ages (6%–14%), and the Middle Castlegate sample had the smallest fraction of Late Jurassic–Cretaceous DZ U-Pb ages (4%).

Samples from the Sunnyside Canyon contained mostly Mesozoic DZ He ages (47%–95%). The Kenilworth and Sunnyside Members included equal proportions: 22%–23% of Jurassic and Triassic DZ He ages and 17% of Early Cretaceous DZ He ages. In contrast, the Castlegate Sandstone preserved a smaller Triassic (7%–11%) and a larger Permian DZ He age fraction (22%–27%), while still containing 20%–22% Jurassic DZ He ages. The Middle Castlegate and Bluecastle Tongue incorporated the most Early Cretaceous DZ He ages (31%–32%), and the Middle Castlegate yielded the most Late Cretaceous DZ He ages (26%; Fig. 3C).

Green River Zircon U-Pb-He Ages

The most noticeable difference in these samples was the percent difference in Cordilleran-Sierran volcanic arc (3%–25%) and Grenville (23%–42%) U-Pb ages (Fig. 3D). The Bluecastle Tongue and Sunnyside Members recorded 18% and 25% of Cordilleran-Sierran volcanic arc DZ U-Pb ages, respectively. The Castlegate Sandstone and Grassy Member contained a higher percent of Grenville DZ U-Pb ages, 42% and 35% in that order.

DZ He ages from Green River samples predominantly fell within the Paleozoic and Mesozoic. Mesozoic ages dominated DZ He populations, except for in the Bluecastle Tongue and Sunnyside Member (13GRT01) samples. All DZ He age populations had a significant Jurassic fraction (20%–31%), except for the Sunnyside Member (13GRT01) sample, which yielded 15% Jurassic fraction. The relative proportions of Early and Late Cretaceous DZ He ages in the Blackhawk members ranged 5%–25% and 0%–20%, respectively. The Castlegate Sandstone and Bluecastle Tongue contained 10% and 5% of Early Cretaceous DZ He ages and 5% and 10% of Late Cretaceous DZ He ages, respectively. Overall, Triassic and Permian DZ He ages represented 6%–20% and 10%–23% of all populations correspondingly. Carboniferous DZ He ages represented ∼20% of the Sunnyside Member and 14% of the Bluecastle Tongue samples (Fig. 3D).

Colorado DZ U-Pb-He Ages (Mount Garfield/Iles Formation, Marapos, and Meeker Sandstones)

The Colorado DZ U-Pb populations predominantly contained Yavapai-Mazatzal (18%–31%), Cordilleran-Sierran volcanic arc (13%–31%), and Grenville ages (11%–28%). Appalachian Paleozoic pluton DZ U-Pb ages represented 3%–8% of the populations, except for the Loyd Sandstone, which incorporated 14% (Fig. 3E). The Meeker and Marapo Sandstone samples were dominated by Cordilleran-Sierran volcanic arc (28%–30%) and Yavapai-Mazatzal (22%–26%) DZ U-Pb ages. The Loyd Sandstone, Sego Sandstone, and Neslen Formation samples contained Grenville (20%–28%), Yavapai-Mazatzal (18%–26%), Cordilleran-Sierran volcanic arc (13%–19%), and Appalachian (7%–17%) DZ U-Pb ages.

DZ He ages from Colorado samples were Paleozoic and Mesozoic in age. The Carboniferous (10%–33%), Permian (6%–25%), and Cretaceous (14%–43%) DZ He ages dominated the overall sampled distribution, followed by Jurassic (7%–25%) DZ He ages. The Meeker, Marapos, Sego, and Rollings sandstones had the largest fraction of Carboniferous (>20%) and Permian (>12%) DZ He ages. The Meeker and Neslen units contained the highest percentage of Late Cretaceous ages at 33% and 32%, respectively (Fig. 3E).

Maximum Depositional Ages

The youngest DZ U-Pb age or age mode for each sample was compared against its established depositional age and arc-magmatic age source (Figs. 3 and 4). For this purpose, we used the stratigraphic and biostratigraphic age framework of Fouch et al. (1983), Edwards et al. (2005), Cobban et al. (2006), Lawton et al. (2007), Aschoff and Steel (2011a), and Steel et al. (2012). Most samples were characterized by syndepositional first-cycle volcanic zircon and showed no systematic correlations between age and U/Th or U concentration, indicative of potential metamorphism or Pb loss (e.g., Dickinson and Gehrels, 2009c). In addition, depth laser-ablation ICP-MS profile analysis allowed us to exclude grains or grain zones characterized by Pb loss or incorporation of common Pb (Marsh and Stockli, 2015). MDAs were determined from the weighted average of the youngest two or more grains that overlapped at 2σ error, denoted YC2σ(2+) (Fig. 3), which represents a slight deviation from the YC1σ(2+) approach of Dickinson and Gehrels (2009c) because it included ages that overlapped at 2σ analytical error. In cases where no two DZ U-Pb ages overlapped at 2σ in the Late Cretaceous depositional age range, we displayed only the youngest single grain age (YSG; Fig. 3) (Dickinson and Gehrels, 2009c). Although most MDAs in this study agree with previously established ammonite ages and stratigraphic correlations (e.g., Cobban et al., 2006; Aschoff and Steel, 2011a), four Late Cretaceous units yielded MDAs that were significantly younger than their corresponding biostratigraphic ages using the above MDA method. The maximum depositional age constraint for the ca. 90.5–89 Ma (ammonite age) Ferron Sandstone is 87.5 ± 0.7 Ma; for the 85.7–83.9 Ma Emery Sandstone, it is 83.5 ± 2.3 Ma; for the ca. 75–74 Ma Bluecastle Tongue, it is 67.6 ± 1.8 Ma in Green River, Utah; and for the ca. 80 Ma Marapos Sandstone, it is 76.8 ± 1.3 Ma north of Meeker, Colorado (Fig. 3; e.g., Edwards et al., 2005; Cobban et al., 2006; Steel et al., 2012). While the age discrepancy between the YC2σ(2+) DZ U-Pb ages (2σ error) and ammonite ages is <5 m.y., YSG ages tend to be systematically younger than the corresponding biostratigraphic ages for Turonian and Campanian strata (Fig. 4). When including all the youngest grains that overlap within 2σ error for each of the samples mentioned before, we obtained MDAs for the Ferron and Emery Sandstones that are in better agreement with ammonite ages (Fig. 4). However, the Bluecastle Tongue (68.2 ± 0.8 Ma) and Marapos Sandstone (76.8 ± 0.7 Ma) samples yielded consistently younger ages than ammonite ages (Fig. 4). These age discrepancies are not easily explained, as Pb loss is not apparent, and they require further examination of correlative stratigraphic intervals and high-resolution geochronology.

DZ (U-Th)/(Pb-He) Sevier Fold-and-Thrust Belt

Detailed DZ (U-Th)/(Pb-He) data from Neoproterozoic to Mesozoic strata in the frontal Sevier fold-and-thrust belt (Pujols et al., 2020) exhibit considerable spatial differences in terms of provenance and thermal history. These differences are critical for correlations with more distal strata in the Book Cliffs and neighboring strata and to establish a provenance signal directly linking the active thrust belt and the distal foreland basin facies.

Three distinct DZ U-Pb and He signatures were identifiable in the central and northern Utah segment of the Sevier fold-and-thrust belt. In central Utah, the Upper Neoproterozoic to Lower Paleozoic strata of the Canyon Range and Pavant thrust sheets contained only Precambrian DZ U-Pb modes and were dominated by Mesozoic (∼75%) and subordinate Paleozoic (∼25%) DZ He cooling ages (Fig. 5). In northern Utah, Neoproterozoic to Mississippian strata in the Charleston-Nebo salient contained uniquely late Albian to Turonian DZ He cooling ages with a well-defined Cenomanian mode (Fig. 5). Pennsylvanian to Lower Permian strata of the Oquirrh Group exhibited mainly Paleozoic (51%) and Jurassic–Cretaceous (38%) DZ He cooling ages and were dominated by Yavapai-Mazatzal DZ U-Pb ages (31%). Lower Permian to Lower Jurassic strata showed predominantly Devonian and Carboniferous (both 22%) DZ He ages and an increase in Appalachian (18%) and peri-Gondwanan (14%) DZ U-Pb ages. Paleozoic DZ U-Pb and DZ He ages were significantly more abundant in Pennsylvanian through Jurassic strata than in older strata and can be linked to Appalachian and Ancestral Rockies mountain-building events (Fig. 5; e.g., Rahl et al., 2003). Carboniferous–Permian DZ He cooling ages have also been documented in Jurassic eolian strata of the Colorado Plateau (Rahl et al., 2003). DZ U-Pb ages attributed to Cordilleran magmatic arc and peri-Gondwanan sources are primarily restricted to Triassic, Jurassic, and Cretaceous synorogenic strata within the frontal Sevier fold-and-thrust belt (Dickinson and Gehrels, 2003, 2008b, 2009a; Lawton et al., 2010; Pujols et al., 2020).

Late Cretaceous DZ Provenance and Link to the Sevier Fold-and-Thrust Belt in Central and Northern Utah

The identification of Late Cretaceous Sevier fold-and-thrust belt source signals was accomplished on the basis of detailed bedrock DZ U-Pb and DZ He analyses and DZ data from more proximal coarse-grained synorogenic deposits (Pujols et al., 2020). Composite DZ U-Pb and DZ He signatures were created for various Late Cretaceous stratigraphic intervals from the Canyon Range to the Book Cliffs. We resorted to time-equivalent DZ U-Pb and DZ He composite (multisample) populations because the DZ U-Pb provenance is nearly identical for most Upper Cretaceous strata in the Book Cliffs (Fig. 6). The multidimensional scaling plot in Figure 6 compares all the Book Cliffs strata DZ U-Pb age populations against those in the Sevier fold-and-thrust belt and proximal foreland basin strata using Kolmogorov-Smirnov (K-S) statistics. As the plot shows, meaningful sediment provenance distinctions are complex or nonresolvable with the current sample size. Only a few distal foreland basin samples from Colorado appear to show statistically significant dissimilarities with P values <0.05 (Table S4a; see footnote 1) and similarities with Pennsylvanian and Permian strata in the Sevier fold-and-thrust belt (Fig. 6). K-S tests were carried out by both including and excluding all volcanic DZ U-Pb ages (younger than 128 Ma) (Tables S4b and S4c). The K-S comparisons excluding all volcanic grains showed an ∼13% increase in sample similarity, further corroborating the Sevier fold-and-thrust belt provenance yet inhibiting any additional provenance differentiation.

The new DZ U-Pb and DZ He record of the Upper Cretaceous proximal and distal foreland basin strata reveals composite provenance signals that appear to be influenced by the structural evolution of either the central or northern Utah portions of the Sevier fold-and-thrust belt and sediment dispersal patterns. Transverse sediment transport from Neoproterozoic to Cambrian strata in the Canyon Range and Pavant thrust sheet to the distal foredeep during Campanian times is evidenced by the presence of Jurassic to Cretaceous DZ He ages in Precambrian zircons (Figs. 3 and 7). Similar DZ He ages have been reported from proximal Canyon Range Conglomerate and Indianola Group samples (Pujols et al., 2020). Eastward sediment transport is also supported by NE-, ESE-, and ENE-directed paleocurrent data from the Blackhawk, Castlegate Sandstone, and Bluecastle Tongue units, respectively (e.g., Lawton, 1986b; Yoshida, 2000; Edwards et al., 2005; Lawton et al., 2007; Sahoo and Ganid, 2015). Derivation of sediment from the Charleston-Nebo salient in northern Utah is supported based on the following lines of evidence: (1) There is a relatively high percentage of Paleozoic DZ He ages (30%) in the Book Cliffs (up to 40% in the Price River Canyon), which is higher than that observed in the Canyon Range bedrock (<25%) and Campanian Pass Canyon Member wedge-top deposits (<10%) (Fig. 7). (2) Carboniferous–Permian Oquirrh Group and Mesozoic strata in the Charleston-Nebo salient, containing mostly Paleozoic DZ He ages (Fig. 5), underwent deformation and rapid unroofing in the Campanian, resulting in the growth strata of the Castlegate–Price River formations (Constenius et al., 2003; Horton et al., 2004). The absence of Upper Paleozoic to Middle Mesozoic strata in the Canyon Range and Pavant thrust sheets also makes the Charleston-Nebo salient a more likely source for these Paleozoic DZ He ages (DeCelles and Coogan, 2006; Lawton et al., 2007). (3) Paleozoic and Yavapai-Mazatzal DZ U-Pb ages suggestive of Upper Paleozoic and Lower Mesozoic Charleston-Nebo salient sources appear in larger proportions in the Price River Canyon transect compared to time-equivalent proximal Campanian strata in the Gunnison Plateau, Hop Creek, Sixmile Canyon, and Canyon Mountains (Pujols et al., 2020). These different lines of evidence for Charleston-Nebo salient sourcing are also supported by paleocurrent directions indicative of north-northwest derivation (Lawton, 1983; Horton et al., 2004).

Previous studies suggested possible northern sources for the Book Cliffs deposits, including the Canada passive margin, because it contains Paleoproterozoic and Archean DZ U-Pb ages commonly absent in the Cordilleran passive-margin section exposed in the Sevier fold-and-thrust belt strata (e.g., Laskowski et al., 2013; Bartschi et al., 2018). However, this and other studies (e.g., Laskowski et al., 2013; Yonkee et al., 2014) document that Paleoproterozoic and Archean DZ U-Pb modes are present in Sevier belt thrust sheets and proximal foreland basin deposits of the Canyon Range Conglomerate and Indianola Group (Pujols et al., 2020). Moreover, the lack of cratonic Paleozoic belt and NW Laurentian (Wopmay orogen) DZ U-Pb ages, abundant in northern provinces and/or Canada passive-margin deposits (Laskowski et al., 2013; Yonkee et al., 2014), precludes major input from those sources into the Campanian foredeep. Hence, local recycling from the Sevier fold-and-thrust belt can readily explain those Paleoproterozoic and Archean DZ U-Pb components.

In summary, the DZ U-Pb and He ages require derivation from both (1) Neoproterozoic and Lower Paleozoic strata in central Utah that cooled during the exhumation of the Sevier fold-and-thrust belt (Fig. 7), and (2) Upper Paleozoic–Mesozoic strata containing recycled DZ grains with Ancestral Rocky Mountains, Antler, and Appalachian cooling signatures. Based on structural reconstructions and proximal provenance studies in the Sevier fold-and-thrust belt in northern and central Utah (Constenius et al., 2003; Horton et al., 2004), we propose that foredeep strata in the Price Canyon area and eastward were synchronously fed by two major transverse sources: (1) Neoproterozoic and Lower Paleozoic strata from the central portion of the Sevier fold-and-thrust belt (Canyon Range and Pavant thrust sheet), and (2) Permian–Jurassic and synorogenic Cretaceous strata exhumed by folding and backthrusting in frontal Charleston-Nebo salient horses (Constenius et al., 2003).

Axial Sediment Transport and Delivery to Proximal Marine Foreland Basin

Differentiating between transverse and axial transport in foreland basins can be difficult, but it is critical for understanding sediment dispersal patterns, facies distributions, and their potential relation to thrust belt deformation. Detailed sample comparison of proximal Canyon Range Conglomerate to distal Campanian strata in the Grand Junction area revealed a systematic downdip reduction (∼30%) of Precambrian DZ U-Pb ages over a distance of ∼330 km (Fig. 7). This systematic shift supports a model of dilution through the influence of an axial transport system or additional drainages feeding into the more distal sections of the basin. The sample comparison also revealed significant differences in the downstream modal percentages of major DZ U-Pb age components, including Grenville, Yavapai-Mazatzal, and Cordilleran-Sierran volcanic arc sources (Figs. 7 and 8). The most striking difference was the higher abundance of Cordilleran-Sierran volcanic arc DZ grains in shallow-marine facies compared to proximal alluvial and fluvial equivalents (Fig. 8A). Fluvial strata contained on average <3% Cordilleran-Sierran volcanic arc grains, while distal marine and lowstand facies contained on average >20% Cordilleran-Sierran volcanic arc DZ U-Pb grains. Yavapai-Mazatzal U-Pb ages exhibited a similar trend, increasing in proportion basinward. In contrast, Grenville DZ U-Pb ages became significantly less abundant in the distal foreland basin strata (Fig. 8B). The constant presence of Cordilleran-Sierran volcanic arc DZ grains, laterally and up section, shows that modal changes were not restricted to episodes of fluvial drainage reorganization, as documented in southern Utah (e.g., Lawton et al., 2014). The DZ age modes were also not limited to different delta types, as these Cordilleran-Sierran volcanic arc and Yavapai-Mazatzal DZ grains were equally dominant in distal facies of the river-dominated (Ferron Sandstones), wave-dominated (Emery Sandstone and Blackhawk Formation), and tide-dominated (Middle Castlegate) deltas.

The DZ spatiotemporal percentage variation precludes a simple Sevier fold-and-thrust belt derivation and transverse sediment dispersal scenario. Rather, the DZ data suggest that the foreland basin was simultaneously fed by three distinct large drainage systems that spatially coevolved from the Cenomanian to the Campanian as the Sevier fold-and-thrust belt propagated eastward: two transverse drainage systems emerging from the central and northern Utah segments of the Sevier fold-and-thrust belt (see the previous section), and an axial fluvial or littoral delivery system from the south, delivering both Cordilleran-Sierran volcanic arc and Yavapai-Mazatzal (Mogollon Highland) DZ grains to the distal marine portion of the foreland basin. Axial fluvial sediment transport in the Late Cretaceous foreland basin is also supported by stratal architecture, paleocurrents, sandstone petrography, and DZ U-Pb data from numerous locations in SW Utah and as far south as Nevada-Arizona (e.g., Fillmore, 1991, 1993; Lawton et al., 2003, 2014; Jinnah et al., 2009; Szwarc et al., 2015). Detrital evidence for NE-directed axial fluvial transport comes from the Campanian Book Cliffs strata (Bartschi et al., 2018; Pettit et al., 2019) and from the Kaiparowits Plateau of southern Utah (Jinnah et al., 2009; Lawton et al., 2014; Szwarc et al., 2015). The Straight Cliffs, Wahweap, and lower Kaiparowits formations, which are time-equivalent to the Blackhawk and Castlegate formations in the Book Cliffs, are characterized by abundant Yavapai-Mazatzal and subordinate Cordilleran-Sierran volcanic arc DZ U-Pb age modes. The sediment source components support a southern derivation from the Mogollon Highlands basement and Cordilleran magmatic arcs (Jinnah et al., 2009; Laskowski et al., 2013; Lawton et al., 2014; Szwarc et al., 2015).

The trend of younger ages up section for a small fraction of syndepositional Cordilleran-Sierran volcanic arc grains chronicles renewed Late Cretaceous volcanism and records the input of airborne first-cycle volcanic zircon (Fig. 4). The bulk of Cordilleran-Sierran volcanic arc ages are Albian to Santonian, with subordinate Triassic to Early Cretaceous DZ U-Pb ages (Fig. 4). The Cenomanian DZ U-Pb ages are the dominant Cordilleran-Sierran volcanic arc component, indicating mostly southern Sierran volcanic arc derivation (Fig. 4; e.g., Dickinson et al., 2012; Lawton et al., 2014). The Triassic to Early Cretaceous DZ U-Pb ages further support a southern arc derivation, as the earliest record of magmatism in potential northern arc sources, such as the Idaho Batholith, did not commence until the late Early Cretaceous (125–105 Ma) and did not peak until the Late Cretaceous (83–67 Ma; e.g., Hyndman, 1983; Manduca et al., 1993; Lee, 2004; Gaschnig et al., 2010).

The observed discrepancy between nonmarine and marine DZ U-Pb age signatures, particularly Cordilleran-Sierran volcanic arc DZ grains, supports a component of littoral/longshore sediment transport redistribution, potentially by strong tide- and/or wave-induced forces. Given the regional north-to-south marine circulation patterns proposed for the Cretaceous Western Interior Seaway (e.g., Ericksen and Slingerland, 1990), it is unlikely these sediments were directly delivered by long-distance shoreline-parallel transport from the southern arc and Yavapai-Mazatzal sources. Hence, we propose that Cordilleran-Sierran volcanic arc DZ grains were mainly delivered by axial river systems draining southern volcanic and Yavapai-Mazatzal provinces. Strong tidal currents, waves, and littoral currents subsequently redistributed the sediment, producing the observed DZ age discrepancy between marine strata and transverse fluvial systems from the Sevier fold-and-thrust belt.

Hydrodynamic grain shape and size sorting in distal fluvial and marine facies could factor in the observed age differences (e.g., Li et al., 2019). However, this is unlikely because Yavapai-Mazatzal DZ U-Pb ages—a population with a wide range of size and grain morphology—also increase basinward, while Grenville grains—more similar in size and shape to the volcanic arc zircons—do not exhibit a basinward increase. Airborne dispersal of volcanic zircons across the drainage basin can also introduce bias. In this scenario, one would expect higher proportions of volcanic ages in the proximal Sevier fold-and-thrust belt deposits and a downstream reduction in volcanic grains due to dilution and increased distance from the western arc. Still, this is not the case, as syndepositional volcanic ages only constitute a small percentage of the Cordilleran-Sierran volcanic arc (285–60 Ma) zircon population in the proximal foredeep and increase dramatically in the distal marine foreland basin strata. Hence, the spatial trends and incorporation of these two DZ U-Pb components into the distal facies of the basin are best explained through a provenance shift, adding both of these components from a southern source.

DZ He Lag Time Estimates, Sevier Fold-and-Thrust Belt Exhumation, and Clastic Wedge Migration

Thermochronometric lag times are commonly used to link cooling events associated with unroofing and faulting in the thrust belt to basin evolution (Bernet et al., 2001; Ruiz et al., 2004; Rahl et al., 2007; Saylor et al., 2012; Stockli and Najman, 2020). In the Campanian foreland basin, three discrete clastic wedges (A, B, and C) with different progradational rates have been documented on the basis of sedimentary facies, depositional environment, ammonite biozones, and shoreline stacking trajectories (Fouch et al., 1983; Cobban et al., 2006; Aschoff and Steel, 2011a). Variations in the style and rate of clastic wedge progradation have been attributed to reductions in accommodation space ascribed to intrabasin Laramide uplift (wedge B), active thrusting, and high sediment supply from the Sevier fold-and-thrust belt (Aschoff and Steel, 2011a). However, until now, the deformation history of the Sevier fold-and-thrust belt has had an insufficient temporal resolution to support interpretation of a tectonic driver for sediment supply and wedge progradation. For this purpose, the youngest DZ He lag time estimates for the Late Cretaceous foreland basin were compiled from correlative strata to determine this relationship and test the causality between thrust belt cooling and exhumation and progradation of the clastic wedges at near million-year-scale resolution.

The DZ He lag time results (Fig. 9) indicate that variations in the clastic wedge depositional architecture and progradation rates in the Campanian foreland basin can be correlated with episodes of rapid thrust belt exhumation in the central and northern Utah portions of the Sevier fold-and-thrust belt. DZ He ages are indistinguishable from depositional ages (i.e., near zero lag time) in the Blackhawk Formation (wedge A) and the Bluecastle Tongue (wedge C), indicating very rapid exhumation in the Sevier fold-and-thrust belt. These near-zero lag times are also coeval with the documented stacking patterns showing shoreline rising trajectory (progradational and aggradational stacking) of Campanian clastic wedges A and C (Aschoff and Steel, 2011a). Thus, we propose that these rising stacking trajectories and likely the wave-dominated shoreline migration associated with both wedge A and wedge C were controlled by episodes of rapid thrust belt deformation, exhumation, and associate flexural accommodation. In contrast, wedge B, characterized by a flat shoreline stacking trajectory (progradational stacking) and tidally influenced deltas, lacks short lag time DZ He ages and shows limited coarse clastic progradation (Fouch et al., 1983; Aschoff and Steel, 2011a, 2011b). Hence, the flat trajectory and extent of the tide-dominated deltaic facies of wedge B are best explained, as proposed by Aschoff and Steel (2011a), by flexural attenuation (reduction in accommodation) due to intrabasin Laramide uplift and tidally influenced sediment dispersal (e.g., Rossi et al., 2016) and were not driven by a deformation pulse within the frontal Sevier fold-and-thrust belt. The lack of evidence for a cessation of intrabasinal uplift and migration of the crustal load during deposition of wedge C suggest accommodation was likely driven by a combination of dynamic and flexural subsidence.

Rapid sediment progradation rates with relatively flat shoreline stacking trajectories and short depositional time span (<2 m.y.) have also been documented for the Turonian river-dominated Ferron and Santonian wave-dominated Emery deltaic shorelines (Edwards et al., 2005, and references therein). Similar to wedge B, both Ferron and Emery sandstone strata lack near-zero DZ He lag times and contain a high percentage of Cordilleran-Sierran volcanic arc and Yavapai-Mazatzal DZ U-Pb sediments sourced from the south, likely the Mogollon Highland and the volcanic arc (Figs. 3 and 10). This suggests that major thrust belt exhumation and increased sediment supply were not the prime drivers for these clastic progradational pulses.

Although rapid sediment progradation and DZ age signatures in the Ferron and Emery sandstones are similar to those in clastic wedge B (Fig. 10), there is no clear similarity in depositional facies. For example, wave-dominated shorelines and/or deltas (i.e., wedges A, C, and Emery Sandstone) are not exclusively linked to major exhumation phases in the Sevier fold-and-thrust belt, as indicated by DZ He zero lag time estimates. The lack of correlation between deformation and depositional facies (i.e., shoreline and delta types) attests to the intricate nature of sediment transport and suggests that although progradation is largely driven by thrust activity, the style of dispersion is not.

In summary, the causal relationship between thrust belt exhumation and clastic migration recorded by short DZ He lag time ages supports thrust belt deformation as the primary driver for clastic progradation and aggradation in high-frequency cycles (∼1 m.y.; Fig. 9). The short-lived, rapid progradation of clastic wedges such as in wedge B and the Ferron and Emery sandstones was not driven by major thrust belt deformation, but rather a combination of factors such as sediment availability from within or outside the Sevier fold-and-thrust belt, accommodation, and forcing factors such as minor deformation, flexure, eustasy, and climate. It is worth mentioning that while major exhumation pulses in the Sevier fold-and-thrust belt can be linked to major clastic progradational wedges, the analytical resolution and uncertainty of ZHe ages are not sufficient to conclusively discard the influence of shorter-lived and lower-magnitude deformational events in the Sevier fold-and-thrust belt. This is problematic when considering generation of depositional slope, which has been postulated as a primary control on progradation direction and successive shoreline migration (e.g., Edwards et al., 2005). Nonetheless, DZ He lag time estimates can help infer the causal linkages and differentiate between major wedge progradation driven by thrust belt deformation and that driven by other factors.

Composite DZ Provenance of Late Cretaceous Clastic Wedges

The composite DZ U-Pb signatures of the clastic wedges A, B, C, and pre-Campanian strata (i.e., Ferron and Emery Sandstones and Panther Tongue) are contrasted by the variations in percentage of Cordilleran-Sierran volcanic arc, Grenville, and Yavapai-Mazatzal DZ U-Pb components (Fig. 10). The relative abundances of Grenville and Yavapai-Mazatzal DZ U-Pb ages can be used to differentiate the clastic wedges and divide them into two groups. Wedges A and C contain more Grenville (32%–34%) than Yavapai-Mazatzal (18%) ages, and pre-Campanian wedges and wedge B are defined by similar Grenville (24%–26%) and Yavapai-Mazatzal (25%–28%) DZ U-Pb age proportions (Fig. 10). These two groups have distinct differences in DZ He ages, with wedges A and C dominated by Late Jurassic–Early Cretaceous DZ He ages and wedge B dominated by Late Cretaceous DZ He ages. This difference appears to reflect a change in source terrane cooling histories (Fig. 10). The DZ He and DZ U-Pb ages in wedges A and C likely reflect derivation from Neoproterozoic and Lower Paleozoic Sevier fold-and-thrust belt bedrock or recycled upper Canyon Range Conglomerate and Indianola Group strata. In contrast, wedge B and pre-Campanian strata, characterized by increased Yavapai-Mazatzal DZ U-Pb ages and Late Cretaceous DZ He ages, were likely derived from southern sources in addition to the Sevier fold-and-thrust belt sediment contributions (e.g., Dickinson and Gehrels, 2008a; Lawton et al., 2014). We interpret the decrease in the youngest DZ U-Pb modes (younger than 900 Ma) up section (Fig. 10) and renewed DZ He ages to record the progressive unroofing of Neoproterozoic–Paleozoic strata in the Sevier fold-and-thrust belt. These DZ age relationships, unroofing, and short DZ He lag times reinforce the interpretation of dominant transverse transport from Sevier fold-and-thrust belt strata during progradation of the clastic wedges.

Basin-Scale Source-to-Sink Model

DZ U-Pb and He age signatures in the Book Cliffs require the input from different sources and cannot be explained only by transverse sediment delivery from the progressive unroofing of the Canyon Range and Pavant thrust sheets (e.g., Van Wagoner, 1995; Robinson and Slingerl, 1998; DeCelles, 2004; Hampson et al., 2005; DeCelles and Coogan, 2006). Transverse sediment supply and transport to the Book Cliffs foredeep appear to have been modulated by the synchronous unroofing of different stratigraphic units within the central and northern Utah Sevier fold-and-thrust belt. Based on proximal foreland strata composition, geothermochronometric and paleocurrent data, and Sevier fold-and-thrust belt structural reconstructions in central and northern Utah, we conclude that the most likely provenance for zircons with Paleozoic DZ He cooling ages was Upper Paleozoic to Mesozoic strata exhumed from shallow levels (<5 km) during backthrusting and exhumation of the frontal Charleston-Nebo salient. This explanation is preferred to frontal Pavant or Paxton thrusting derivation, which also exhumed and deformed Mesozoic strata during this time, because DZ geothermochronology evidence in the wedge-top and foredeep-equivalent strata adjacent to the Pavant and Paxton thrust sheets is mainly characterized by Precambrian–Paleozoic zircons yielding Jurassic–Cretaceous DZ He cooling ages (Pujols et al., 2020). In addition to the transverse sediment transport and Sevier fold-and-thrust belt derivation, the provenance of distal fluvial to shallow-marine Book Cliffs deposits seems to have been controlled by axial transport (Fig. 8) derived from the Mogollon Highlands and Sierran magmatic arc. The axial fluvial and longshore marine sediment transport coevolved with the eastward migration of Sevier fold-and-thrust belt deformation and major transverse sediment delivery from the Turonian to the Campanian with no apparent interruption (Fig. 8).

Near-zero lag time signatures are characteristic of clastic wedges A and C and thus establish a causal link between rapid exhumation, major unroofing in the Sevier fold-and-thrust belt due to active thrusting, and clastic wedge progradation (Figs. 9, 10, and 11). The rising shoreline trajectories in wedges A and C are coeval with active and rapid thrust belt deformation. Composite DZ U-Pb and DZ He age signatures support a dominant Sevier fold-and-thrust belt derivation for wedges A and C. These observations point to sediment supply exceeding accommodation space induced by flexural and potentially dynamic subsidence (e.g., Painter and Carrapa, 2013), as well as eustatic effects during active thrusting and major exhumation (Fig. 11). The migration of the clastic wedges was propelled by massive sediment generation and transverse delivery from the Sevier fold-and-thrust belt. The magnitude of flexure limited the dispersal of sediments, resulting over time in a rising shoreline trajectory (Fig. 11). In contrast, rapid progradation of deltas within wedge B is not characterized by near-zero lag times indicative of rapid exhumation and hence was more likely driven by reduced accommodation (Aschoff and Steel, 2011a). A strong tidal influence might have additionally modulated the extent of this clastic migration of wedge B (e.g., Rossi et al., 2016). The punctuated sediment progradation pulses of the Ferron and Emery sandstones, similar to the Middle Castlegate strata (clastic wedge B), also do not coincide with short lag time DZ He ages indicative of rapid thrust belt exhumation. Therefore, the progradation of these smaller clastic wedges was likely driven by other forcing factors such as eustasy, climate, or smaller-magnitude exhumational events not recorded by the DZ He ages. An up-section systematic reduction in Paleozoic DZ U-Pb age modes within the Turonian to Campanian section confirms progressive unroofing of lower structural levels of Paleozoic through Neoproterozoic strata in the Sevier fold-and-thrust belt. The sources of zircons recording short lag times were likely associated with structural culminations, duplexing, and backthrusting in the Sevier fold-and-thrust belt (Fig. 11; e.g., Constenius et al., 2003; DeCelles and Coogan, 2006; Lawton et al., 2007).

New DZ U-Pb and He data from Turonian through Campanian sandstone units in the Book Cliffs and vicinity allow us to distinguish provenance, identify and track major sediment routing systems, and quantify the magnitude of transverse and axial sediment delivery during Sevier fold-and-thrust belt activity. These data point to primary transverse sediment sourcing from both the central and northern Utah Sevier fold-and-thrust belt and track the progressive erosional unroofing of Neoproterozoic to Mesozoic sedimentary units within the different thrust sheets (e.g., Pujols et al., 2020). More specifically, rapid erosional unroofing of Upper Paleozoic to Mesozoic sedimentary sections, characterized by mostly Paleozoic DZ He ages, occurred in northern Utah during backthrusting in the Charleston-Nebo salient. In contrast, Neoproterozoic–Paleozoic strata, characterized by Mesozoic DZ He ages, were tectonically exhumed in the Canyon Range and Pavant thrust sheets by duplexing and activity of the Paxton thrust (Pujols et al., 2020).

Comparison of new DZ U-Pb data from the distal foreland basin strata in the Book Cliffs with published data from the proximal foredeep illustrates the importance of transverse sediment delivery from the frontal Sevier fold-and-thrust belt. It also underscores the critical component of along-strike sourcing from the south and the role of axial sediment transport and delivery for the marine portions of the Sevier foreland basin. In particular, DZ U-Pb signatures corresponding to the Mogollon Highlands and Sierra Nevada arc sources document an important long-lived axial input that has fundamental implications for sediment dispersal patterns and stratal architecture in the distal marine portions of the foreland basin. Spatial differences in DZ U-Pb and He ages in time-equivalent foreland basin strata attest to the intricate and multicomponent nature of sediment delivery, dispersal, and mixing from the Sevier thrust belt.

DZ He data tracked the exhumation of the frontal Sevier fold-and-thrust belt in Utah. Depositional lag time estimates and coupled DZ He and DZ U-Pb ages allow for the investigation of the temporal interplay between clastic wedge progradation and active tectonic hinterland exhumation. Our new data allow for a more critical and differentiated view of clastic progradation in foreland basins and support conceptual models that have tied pulses of sediment dispersion to active thrusting (Burbank et al., 1988; Kamola and Huntoon, 1995; Houston et al., 2000; Horton et al., 2004). DZ He lag times for clastic wedge A (the Blackhawk Formation) and wedge C (Bluecastle Tongue of the Castlegate Formation) record very rapid exhumation in the Sevier fold-and-thrust belt and point to a dominant tectonic driver for rapid clastic progradation. In contrast, wedge B, exhibiting a purely progradational, flat shoreline trajectory and tide-influenced deltas, lacks short lag time DZ He cooling ages and is characterized by limited coarse clastic progradation (Fouch et al., 1983; Aschoff and Steel, 2011a, 2011b). The flat shoreline trajectory and tide-dominated deltaic facies extent of wedge B are best explained by intrabasin flexural attenuation due to basement-core uplifts rather than frontal Sevier fold-and-thrust belt deformation as describe by Aschoff and Steel (2011a).

While rates and extents of clastic wedge progradation can be controlled by numerous factors such as fold-and-thrust belt deformation, availability of accommodation space, sediment supply, eustasy, and potentially climatic factors, these data show that DZ U-Pb-He double dating and DZ He lag time estimates can elucidate the relationship between tectonic unroofing in the frontal thrust belt and clastic wedge progradation. These high-resolution DZ geo- and thermochronometric data allow for a careful evaluation of the interplay among sediment provenance, sediment mixing, hinterland tectonics, and intrabasinal clastic progradation. The results demonstrate that while some clastic wedges can be temporally correlated and attributed to thrust belt activity, not all wedges are necessarily driven by tectonic activity, allowing for a more differentiated assessment of clastic progradation in foreland basins.

This work was supported by UTChron Laboratories, the Society for Sedimentary Geology (SEPM), and Geological Society of America student research grants to Pujols, and the Chevron (Gulf) Centennial Professorship to Stockli. We also would like to acknowledge field support from Peter Flaig, Dolores van der Kolk, and Ron Steel, and enlightening discussions and insights from Brian Horton, Kurt Constenius, and Tim Lawton. This work would not have been possible without the analytical support of Desmond Patterson and Lisa Stockli. We would like to also thank reviewers Tyson Smith, Peter Clift, and Andrew Laskowski for comments that improved the manuscript. All DZ He and DZ U-Pb isotopic data, ages, and sample global positioning system (GPS) locations are provided in the Supplemental Material and Geochron global database, http://www.geochron.org/dataset/html/geochron_dataset_2021_05_18_dXSD8.

1Supplemental Material. Table S1 contains sample locations and formation names. Tables S2 and S3 contain the isotopic data for all samples DZ U-Pb (Table S2) and DZHe ages (Table S3). Table S4a-b shows the K-S test p and D values between all Book Cliffs DZ U-Pb samples before and after excluding volcanic grains. Table S4c shows K-S test p and D values between all the DZ U-Pb age populations from the Book Cliffs (this study), Sevier fold-and-thrust belt, and proximal strata (Pujols et al., 2020). Please visit https://doi.org/10.1130/GEOS.S.15145527 to access the supplemental material, and contact editing@geosociety.org with any questions.
Science Editor: Andrea Hampel
Associate Editor: Jason W. Ricketts
Gold Open Access: This paper is published under the terms of the CC-BY-NC license.