Abstract

The Paleocene–Eocene strata of the rapidly subsiding Hanna Basin give insights in sedimentation patterns and regional paleogeography during the Laramide orogeny and across the climatic event at the Paleocene–Eocene Thermal Maximum (PETM). Abundant coalbeds and carbonaceous shales of the fluvial, paludal, and lacustrine strata of the Hanna Formation offer a different depositional setting than PETM sections described in the nearby Piceance and Bighorn Basins, and the uniquely high sediment accumulation rates give an expanded and near-complete record across this interval. Stratigraphic sections were measured for an ∼1250 m interval spanning the Paleocene–Eocene boundary across the northeastern syncline of the basin, documenting depositional changes between axial fluvial sandstones, basin margin, paludal, floodplain, and lacustrine deposits. Leaf macrofossils, palynology, mollusks, δ13C isotopes of bulk organic matter, and zircon sample locations were integrated within the stratigraphic framework and refined the position of the PETM. As observed in other basins of the same age, an interval of coarse, amalgamated sandstones occurs as a response to the PETM. Although this pulse of relatively coarser sediment appears related to climate change at the PETM, it must be noted that several very similar sandstone bodies occur with the Hanna Formation. These sandstones occur in regular intervals and have an apparent cyclic pattern; however, age control is not sufficient yet to address the origin of the cyclicity. Signs of increased ponding and lake expansion upward in the section appear to be a response to basin isolation by emerging Laramide uplifts.

INTRODUCTION

From the latest Cretaceous through early Eocene, the Hanna Basin in south-central Wyoming was one of the most rapidly subsiding basins of the Laramide foreland (Fig. 1; Roberts and Kirschbaum, 1995; Wroblewski, 2002; Jones et al., 2011). As a result, a thick, relatively continuous succession of shallow marine, fluvial, paludal and lacustrine strata is preserved in the center of the basin. This thick and continuous record allows to fill gaps in understanding the regional paleogeography and basin-fill history of nearby Laramide basins, such as the Denver Basin and North Park–Middle Park Basin, where sections are not as complete, and hiatuses are well documented within the section (Raynolds and Johnson, 2003; Cole et al., 2010; Dechesne et al., 2013). Earliest work focused on coal resources in the area (Veatch, 1907; Dobbin et al., 1929, Glass, 1980; Glass and Roberts, 1984; Flores et al., 1999a, 1999b) and showed either short sections for individual coal beds or generalized stratigraphic sections. The extensive work of Lillegraven (1994), Lillegraven and Snoke (1996), and Lillegraven et al. (2004) mostly focused on vertebrate occurrences and structure in The Breaks area and the northern basin margin. The occurrence of the Paleocene–Eocene Thermal Maximum (PETM, 56 Ma; Jaramillo et al., 2010) within the Hanna Basin had only generally been identified via mollusks, palynology, and isotopes (Kirschner, 1984; Flores et al.,1999a, 1999b; Higgins, 2012; Pew, 2014).

It has been hypothesized that climate change across the PETM caused changes in basin sedimentation by increased seasonal differences that produced larger flood events (Foreman et al., 2012; Foreman, 2014; Plink-Björklund, 2015). Terrestrial PETM sections in nearby Laramide basins (Foreman et al., 2012; Foreman, 2014) and in the Spanish Pyrenees (Colombera et al., 2017) have all been associated with amalgamated sheet sandstones. However, the Paleocene–Eocene deposits of the nearby Piceance and Bighorn Basins include predominantly fluvial deposits and floodplains with abundant paleosols (Foreman et al., 2012; Foreman, 2014; Kraus et al., 2015). Background sedimentation in the Hanna Basin is typified by a more ponded, fluvial to shallow lacustrine environment, thus allowing comparison and determination as to whether a similar sedimentation response is present across the PETM in a generally wetter depositional environment.

Geologic Setting

The Hanna Basin is especially noteworthy for its thick package of >12,500 m (40,000 ft) of Phanerozoic sedimentary strata (Fig. 1; Blackstone, 1993; Roberts and Kirschbaum, 1995; Lillegraven and Snoke, 1996; Wroblewski, 2003). Most rapid subsidence and sediment accumulation occurred from the Campanian through Eocene (Ypresian?) (Wrobleski, 2003) due to the Hanna Basin’s position at the distal end of the Laramide flat slab present under the North American continent (Cross and Pilger, 1978; Dickinson and Snyder, 1978; Jones et al., 2011; Copeland et al., 2017).

The sedimentary succession of the Hanna Basin resembles nearby basins of Upper Cretaceous through Eocene age. However, because of its very high subsidence rates and sediment input, the Hanna Basin preserved a uniquely thick and relatively complete sedimentary record in its basin center. Main retreat of the Western Interior Seaway from this area in the late Maastrichtian is marked by the beach and shore-face deposits of the Fox Hills Sandstone. Leaf margin analysis and several marine incursions documented in the Upper Cretaceous Medicine Bow and Upper Cretaceous–Paleocene Ferris deposits suggest that the Hanna Basin remained at or near sea level until at least the early or middle Eocene (Dunn, 2003; Wroblewski, 2003, 2004; Boyd and Lillegraven, 2011; Cather et al., 2012; Lillegraven, 2015). Deposits progressively change from marginal marine and coastal plain to increasingly fluvial, paludal, and shallow lacustrine (Fig. 2). The Hanna Formation is most known for its coal beds and carbonaceous shales, which alternate with successions of siltstones, sandstones, and some conglomerates (Veatch, 1907; Bowen, 1918; Dobbin et al., 1929; Ryan, 1977; Glass, 1980; Flores et al., 1999a, 1999b; Wroblewski, 2002; Lillegraven et al., 2004). Formation and preservation of extensive coal beds indicate a densely vegetated landscape and high freshwater influx, causing extensive ponding and peat production that kept pace with the high accommodation and rapid burial (McCabe, 1984; Cecil, 1990; Bohacs and Suter, 1997; Flores et al., 1999b). The organic material that makes up the coals is interpreted to have accumulated in low-lying peat swamps tied into a fluvial system based on the relatively common centimeter-scale very fine sand and siltstone layers, which indicate intermittent fine-grained influx from distant rivers (McCabe, 1984; Pierce, 1996; Flores et al., 1999b).

Initially, the Hanna Basin was part of the Greater Green River Basin, later isolated into smaller subbasins by localized uplifts of the Laramide orogeny during the mid to late Paleocene (Fig. 1; Ryan, 1977; Blackstone, 1993; Perry and Flores, 1997; Secord; 1998; Kraatz, 2002; Wroblewski, 2003; Lillegraven, 2015; Smith et al., 2015b; Loope and Secord, 2017). The emergence of these uplifts influenced provenance and sedimentation patterns indicating predominantly local sources after this time (LeFebre, 1988; Blackstone 1993; Perry and Flores, 1997; Lillegraven et al., 2004; Peyton and Carrapa, 2013). Synsedimentary deformation within the Hanna and underlying Ferris Formations is well documented, especially along the basin’s edges (Secord, 1998; Kraatz, 2002; Wroblewski, 2003; Lillegraven et al., 2004; Loope and Secord, 2017).

The base of the Hanna Formation is considered conformable with the Ferris Formation in the center of the Hanna Basin (Knight, 1951; Blackstone, 1993; Flores et al., 1999b). However, the basin margins show disconformable to angular contacts, because active underlying structures influenced accommodation and disrupted sedimentation (Knight, 1951; Love and Christiansen, 1985; Blackstone, 1993; Perry and Flores, 1997; Secord, 1998; Lillegraven et al., 2004; Loope and Secord, 2017). Hanna Formation reported thickness is therefore variable and greater in the center of the basin, and several maximum thickness values have been reported, ranging from ∼3500 m (11,000 ft) to 2150 m (7000 ft) (Dobbin et al., 1929; Gill et al., 1970; LeFebre, 1988; Lillegraven and Snoke, 1996; Wroblewski, 2003; WOGCC, 2016).

General age information for the Hanna Formation was based on palynostratigraphy and intermittent occurrences of diagnostic mammalian fossils. Pollen zones P1 and P2 of Nichols and Ott (1978) are present in the Ferris Formation (Flores et al., 1999b; Dunn, 2003) along with Lancian and Puercan (Pu1–Pu3) mammals (Late Cretaceous–Paleocene, North American Land Mammal Age (NALMA); Eberle and Lillegraven, 1998a, 1998b). P3 pollen are found at the base of the Hanna Formation in the center of the basin, indicating onset of Hanna sedimentation in the mid-Paleocene or later, and P5 pollen are found at the basin margin in The Breaks corroborating uplift before and during deposition (Perry and Flores, 1997; Dunn, 2003; Lillegraven et al., 2004). A succession of vertebrate localities spans the latest Torrejonian (To3) through middle Tiffanian (Ti3) land mammal ages on the northeastern margin (Higgins, 2003), and a single tooth of Hyracotherium grangeri was found near the top of the Hanna Formation (Lillegraven et al., 2004), confirming a post-PETM Wasatchian age (early Eocene, NALMA) for its youngest beds (Gingerich and Clyde, 2001).

Paleoclimatic Setting

The North American continental interior was warm and equable, with minimal frost, during the Paleocene and early Eocene (Wing and Greenwood, 1993). Globally, temperatures gradually warmed across this interval to a sustained Cenozoic maximum from 52.6 to 50.3 Ma (Zachos et al., 2008; Payros et al., 2015). The PETM, an abrupt global perturbation to the carbon cycle at the Paleocene–Eocene boundary, had a profound effect on Earth surface systems (McInerney and Wing, 2011). In less than 20,000 years, thousands of petagrams of isotopically light carbon were released into the atmosphere and ocean as evidenced by a 3–5 per mil (‰) global decrease in carbon isotope values, referred to as a negative carbon isotope excursion (CIE) (Kennett and Stott, 1991; Koch et al., 1992; Cui et al., 2011; Bowen et al., 2015).

Independent paleoclimate proxies document global warming of 5–8 °C accompanying this carbon release (McInerney and Wing, 2011). Bowen et al. (2004) and Winguth et al. (2010) proposed that increased pCO2 and temperature intensified the hydrologic cycle and enhanced monsoonal circulation in the Western Interior during the PETM. PETM vegetation records exhibit varying degrees of floral change, with most records for temperate and high latitudes showing an abundance decrease in conifers and increase in thermophilic angiosperms (Wing and Currano, 2013). Changes to the hydrologic cycle and vegetation structure during the PETM, in turn, impacted river systems and sediment accumulation in the Bighorn and Piceance Basins (Foreman et al., 2012; Foreman 2014).

FIELD STUDIES AND ANALYTICAL METHODS

Stratigraphy

Detailed stratigraphic sections were measured on both flanks of the northeast syncline from which a stratigraphic framework for the upper Hanna Formation was developed (Figs. 3 and 4; Supplement S11). Two of the measured sections span the entire studied interval: one in Hanna Draw (∼1250 m; ∼4100 ft) on the west limb of the syncline and one on the east limb, in The Breaks (TB; 940 m; 3083 ft). The Hanna Draw composite section was measured southeast of Hanna Draw road (HD; Fig. 3). The Breaks section approximately corresponds with the location of Leg 17, originally measured by Lillegraven (1994), and later worked by Higgins (2012) and Pew (2014), who both focused on identifying the PETM in The Breaks by isotopes and palynology. The Beer Mug Vista (BMV) section (278 m; 912 ft) captures proximal-to-distal facies relations from near the basin margin to the relatively more distal and shallow lacustrine section in The Breaks. The Big Channel Lateral (BCL; 193 m; 633 ft) and Big Channel Axis (BCA; 82 m; 270 ft) sections capture lateral changes between across the PETM section in Hanna Draw.

The stratigraphic framework was tied to the coal-bed nomenclature of Dobbin et al. (1929); this framework was established in Hanna Draw near our measured section and was maintained by later workers, so that previous work could be more easily integrated (e.g., Glass, 1980; Glass and Roberts, 1984; Flores et al., 1999a, 1999b). Dobbin et al. (1929) did not extend correlations toward the active basin margins because increased clastic input interfingers with the paludal deposits, and coal beds are not as well developed. However, in the center of the basin, coal beds and carbonaceous shale intervals are laterally continuous over at least 5–10 km. Dobbin maps only the thickest and laterally most continuous coal beds and shales (Coals 78, 79, 80, 81, 87, 88, and 89) across the also faulted syncline between Hanna Draw and The Breaks. To enhance correlation certainty, this study did not exclusively use coal beds and carbonaceous shales, but also paleobotany results and stratigraphic marker beds, such as the gastropod limestone bed within Coal 87, aerial photography (NAIP, 2012), Google Earth, and walking beds laterally. The here-developed framework spans the interval between Coal 77 and an originally unnamed coal bed just above Coal 89 of Dobbin et al. (1929), labeled Coal 90 here (Figs. 2 and 4). It is important to note that Lillegraven et al. (2004) mapped 600 additional meters of increasingly more sand-rich strata above Coal 90 in the Hanna Formation that were not incorporated in this framework.

Paleobotany

Organic-rich sandstone, siltstone, mudstone, and shale beds along Hanna Draw and in The Breaks were surveyed for plant macrofossils. Occurrences of biostratigraphically indicative taxa were recorded and plotted on the stratigraphic sections. Of particular importance are the early Eocene index taxa Platycarya (Juglandaceae), Salvinia preauriculata Berry (Salviniaceae), Lygodium kaulfussi Heer (Schizaeaceae), and Cnemidaria magna (Cyatheaceae) (Brown, 1962; Nichols and Ott, 1978; Manchester and Zavada, 1987; Wing, 1998). The occurrence of Cyclocarya fruits and Cornus swingii leaves in the study area indicates a Paleocene age (Manchester, 1987; Manchester et al., 2009). Voucher specimens of identifiable plant parts from each locality are curated at the University of Wyoming Geological Museum. GPS coordinates are available from the authors, pending permission from the Bureau of Land Management (BLM) or property owner.

Palynological sampling was undertaken to determine the age of strata in the measured sections based on the well-established Paleocene–Eocene palynostratigraphic scheme (P-biozones) of Nichols and Ott (1978) and Nichols (2003). Samples were collected from fine-grained units including carbonaceous shales and coals. Palynological preparations were made using two different methods: (1) standard palynological preparations were performed by Global Geolabs Ltd. in Medicine Hat, Alberta, Canada; and (2) HF-free preparations (following O’Keefe and Eble, 2012) were performed by Jen O’Keefe at Morehead State University, Morehead, Kentucky, USA. Prepared microscope slides were scanned using light microscopy, and the occurrences of diagnostic pollen grains were noted. To pinpoint the PETM interval, first appearance data were recorded for known PETM and Eocene taxa described from other Western Interior basins and the Gulf Coast, USA (e.g., Harrington, 2003a, 2003b).

Invertebrate Fossils–Continental Mollusks

The study of Hanna Formation continental mollusks is part of a larger project to revise and evaluate taxonomy and biostratigraphy of western United States Cretaceous and Paleogene taxa2. An evaluation of previous continental molluscan records from the Ferris and Hanna Formations by Lesquereux (1874), Glass (1980), and Kirschner (1984) is supplemented with new, stratigraphically controlled specimens to confirm the Paleocene–Eocene boundary, to refine early Eocene stratigraphy and enhance paleoenvironmental interpretations. Study included examination of material at the Smithsonian Institution (Department of Paleobiology, U.S. National Museum of Natural History, Washington, D.C.), U.S. Geological Survey (USGS, Denver, Colorado), and the University of Wyoming Geological Museum (Laramie, Wyoming). Species of interest to determine the Paleocene–Eocene boundary include Micropygus, Elimia, and Paludotrochus.

Bulk Organic Carbon

Bulk organic carbon samples for each major coal or lignite interval were collected while measuring section. Each sample was pulverized in the lab, after which the sample was treated with hydrochloric acid to remove siderite and carbonate overprints. Samples were then dried, weighed, and analyzed by mass spectrometer for δ13Corg. Two labs were used for processing and analyzing. About half the samples were processed by the Stable Isotope Ratios in the Environment, Analytical Laboratory (SIREAL) at the University of Rochester, New York, using a DeltaPlus XP mass spectrometer interfaced to a Costech 1410 Elemental Analyzer via a Thermo Electron Conflo III. All analyses were reported in permil (‰) relative to VPDB (Vienna Pee Dee Belemnite) and normalized (Coplen, 1994) so that the carbon isotopic value of standard reference material (SRM) 8539 is −29.73‰, of SRM 8542 is −10.45‰, and of SRM 8541 is −16.05‰ following the methods outlined in Higgins (2012). The other samples were processed at the USGS in Lakewood, Colorado, following Johnson et al. (2018) using a Carlo Erba NC2500 elemental analyzer coupled to a Micromass Optima isotope ratio mass spectrometer (IRMS) via a custom-built, open-split interface. All δ13Corg samples were standardized to VPDB.

Zircon Geochronology

Several centimeter-scale, white, weathered, clay-rich mudstones to fine-grained sandstone beds, suspected of being ash beds or tonsteins, are present within the coal and carbonaceous shale intervals. Samples were collected for isotope dilution–thermal ionization mass spectrometry (ID-TIMS) U-Pb zircon dating to better constrain the age of the Hanna Formation and specifically the absolute age of strata interpreted to represent the PETM. Zircons were concentrated from these beds by ultrasonic deflocculation (mudstones), standard crushing (sandstones), and density and magnetic separations. Each sample yielded zircons with a variety of morphologies, including minor populations of mechanically rounded grains that are interpreted to be older, detrital zircons. Zircons were dated by laser ablation–inductively coupled mass spectrometry (LA-ICPMS) as a screening tool and by high-precision, chemical abrasion (adapted from Mattinson, 2005), isotope dilution, thermal ionization mass spectrometry (CA-ID-TIMS). Dating efforts by TIMS focused on euhedral grains with sharp crystal vertices, elongate tips, and longitudinal bubble tracks, which are diagnostic of volcanic, ash-fall origin. A detailed description of TIMS and LA-ICPMS analysis techniques is in Supplement S2a (footnote 1).

RESULTS

Stratigraphy

The stratigraphic framework documents facies changes across the study area and allowed positioning of all sample, paleobotany, and mollusk localities leading to a more precise identification of the PETM (Fig. 4; detailed framework with fossil and sample localities in Supplement S1 [footnote 1]). Sandstone, shale, and coal intervals are typical for fluvial, paludal, and lacustrine environments, and lateral facies changes are gradational and influenced by relative position within the basin, e.g., basin-axial (Hanna Draw Section [HD]), distal to marginal lacustrine near the east side of the study area (The Breaks Section [TB]), or basin margin (Beer Mug Vista [BMV]). Lithofacies are summarized in Table 1.

Fluvial Sandstones

The largest sandstone bodies within our study area, here informally named “Big Channel” and “Camp Channel,” can be traced between Hanna Draw and more distally into The Breaks (Figs. 3 and 4). They contain multiple amalgamated and internally incised channel sandstone bodies with meter-scale barforms, low-angle, planar, trough, and climbing-dune cross stratification. Soft-sediment deformation of sedimentary structures is relatively common, especially in the upper parts of the channels and corroborates rapid deposition (Fig. 5). Coarsest clasts (cobbles up to 10 cm) are found in the Big Channel section (BC) in Hanna Draw. Clasts match compositions of the units present in nearby uplifts, ranging from chert (Cloverly Formation), angular, white to light-gray shale parts (Mowry Shale), quartz sandstones with well-rounded frosted grains (Tensleep Formation), and occasional granitic or metamorphic basement clasts.

In Hanna Draw, Big Channel is not one channel but a clustering of three axial channel complexes of each 10–20 m (30–65 ft) thick, separated by equally thick intervals of carbonaceous shales with root casts, slickensides, and light-gray to orange and orange-to-red mottled paleosols. The orange-to-red paleosols are best developed in the fine-grained sand, silt, and shale deposits between 121 and 136 m (stories 2 and 3) in BC in Hanna Draw and are not common in the rest of the section in Hanna Draw. Lateral offsets between individual channel axes are between ∼30 and 125 m (100–400 ft). In BCL and away from its axis, Big Channel complex splits up into multiple individual sandstone bodies that are less coarse, indicating that the location of BC is more axial (Fig. 4).

Big Channel’s distal expression in The Breaks (Fig. 6A) is at least 34 m thick (110 ft; top not exposed) and at least 3 km wide, of yellow sandstone with clasts up to 4 cm (1.5 in) diameter, abundant crossbedding (Fig. 6B), and soft-sediment deformation. It is correlated to Hanna Draw based on paleobotanical information, overall stratigraphic patterns, and lithologic characteristics. Camp Channel, almost equally as large as Big Channel, is also correlative into The Breaks. It is made up of similar sedimentary features; however, soft sediment deformation is even more abundant, as are up to 2 m concretions (Fig. 6C).

In contrast, between Coal 88 and 88 upper and in general vertical sequence with the occurrence of the other larger sandstone complexes, is an 18-m-thick sandstone unit encased carbonaceous shales (Bird Plug Fig. 2; Figs. 5 and 7). This sandstone is also correlative between Hanna Draw and BMV; however, it is only expressed as a thin, rippled sandstone bed in TB (at 505 m; Supplement S1 [footnote 1]). The carbonaceous shales both below and above this sandstone contain large metasequoia tree trunks and dark-brown mudstones with fish scales (Fig. 7C). Internal foresets dip in multiple directions, and climbing ripples are abundant. Coarse basal pebble lags are absent; however, soft clast rip-up lags and mud-injection features are common. Oversteepened bedding (Fig. 7D), rotated blocks, and slumping suggest a less consolidated stratum and possibly deeper water depth more typical for a (lacustrine-) deltaic than fluvial origin. Distally, foresets become unidirectional more distally and resemble the tangential foresets common between Coal 81 and 82 (Figs. 7E and 7F). These features suggest distal conditions compared with the other large amalgamated sandstones, likely by the increasingly lacustrine nature of the section above Coal 87.

Large sandstone complexes with similar facies and channel dimensions occur at regular intervals in the Hanna Formation—for example, between Coal 80 and 81 and Coal 76 and 77 (called Doug; Fig. 2) and at the base of the Hanna Formation (Sand Ridge, between Coal 66 and 67, Fig. 2). These pulses of relatively coarse sediment appear repetitive. Comparison of internal lithofacies and dimensions between these sandstones shows that there is an overall decrease in large-scale, high-energy bedforms and increase in ripples and soft sediment deformation within the channels (Fig. 8). The sandstones of Big Channel do stand out as a non-amalgamated cluster of channels and coarser clasts and larger bedforms than observed in the channels directly below.

Isolated channels with accretion sets, abundant climbing ripples, and sometimes coarse to gravelly lags are common within coal and carbonaceous shale intervals (most commonly between Coals 77 and 79 and Coals 83–85 in Hanna Draw). Channel sizes vary between 1 and 5 m in thickness, and widths range from 8 to 85 m. They are associated with laterally continuous, tabular sandstones with internal scouring and climbing ripples encased in carbonaceous shale intervals that are interpreted as splays and splay channels.

Basin Margin

Closer to the basin margin, at Beer Mug Vista (BMV; Fig. 4), coarse sandstone-to gravelly beds of 10–30 cm with common angular clasts of Mowry Shale, siderite, and iron oxide–cemented sandstone occur (Fig. 9A). These beds are especially common in the lower half of BMV, but some of them extend south toward TB. Coarse grain size and angularity indicate a higher depositional gradient and limited transport distance because BMV is presently (after Laramide shortening) ∼6 km from the Freezeout Mountains, the northern basin margin. These shallow lag deposits might represent fan rather than true fluvial deposits. The iron oxide–cemented sandstone clasts appear similar to ones described in the nearby Carbon Basin by Loope and Secord (2017), who proposed that these concretions originally formed as siderite nodules in the Ferris Formation. These nodules were then exposed by erosion from active uplift by underlying structures and resedimented in the Hanna Formation. Dips within BMV shallow upward and range from 30° at its base to 15° near its top, which supports the presence of an actively growing structure (Fig. 9B).

The lower part of BMV section is characterized by orange-to-red mottled paleosols, suggesting well-drained rather than swampy conditions near the basin margin (Fig. 9C). As observed in the other sections, above Coal 87, carbonaceous shales and carbonate concretions increase and suggest regional lake and swamp expansion (Fig. 4).

Paludal and Floodplain Deposits

Coal and carbonaceous shale intervals occur in all sections and vary from 5 to 40 m (15–130 ft) thick and are laterally correlative over kilometers. Carbonaceous shales can range from dark gray to black and dark brown and commonly contain plant remains, gypsum crystals (selenite), and siderite concretions. Highest-grade coals are black, vitreous, and crumbly when encountered in outcrop. Light- to dark-gray immature paleosols have commonly developed in carbonaceous shales. Carbonaceous shales and siltstones alternate with light-gray, carbonate-cemented, fine-grained sandstone beds that are capped with oscillation ripples or vertical burrows and bedding-parallel feeding traces (Planolites; Fig. 10). Small vertical burrows up to 1 cm (0.5 in) are also common in rippled, fine-grained sandstone beds.

Presence of extensive coal and carbonaceous shale intervals indicate permanently inundated floodplains, ponds, or swamps with ample vegetation, which is corroborated by the abundance of Metasequoia tree trunks and foliar remains of Averrhoites affinis, Equisetum sp., Zingiberopsis isonervosa, and various ferns. Crumbly, light- to dark-gray, structureless, sandy and silty shales and siltstones with slickensides, mottling, and root casts indicate immature pedogenesis and changes between inundated and less wet conditions (Glass and Roberts, 1984). The orange- to red-mottled paleosols found near the basin margin (BMV) and in the fluvial sandstone interval at BC indicate low groundwater levels and oxidizing conditions that are generally related to better soil drainage (Kraus et al., 2013). However, these better drained soils are rare within the study area.

Crayfish traces, ranging from 3 to 8 cm in diameter and up to 1 m deep, occur in light-gray shale to fine-grained sandstones with immature paleosols between Coal 83 and Coal 85 in The Breaks (Fig. 10). Crayfish trace fossils have only been found in The Breaks and are typical for areas of fluctuating water tables and seasonality in groundwater levels (Hasiotis and Honey, 2000).

Lacustrine Sandstones and Shales

Deepest lacustrine deposits in the study area consist of chocolate- to dark-brown- to gray shales that often contain fish fragments such as scales and bones, insect wings, gastropods, and bivalves. These deposits alternate with siltstones, carbonaceous shales, and coals of paludal origin, and the boundary between shallow lake, swamp, and permanently inundated floodplain is not always clear. Mollusk preservation appears more prevalent in predominantly lacustrine intervals as water level deepened, and often fish fragments are found nearby.

A millimeter-scale laminated, fine-grained limestone bed with abundant gastropods of which the taxa are unidentifiable and show only few signs of transport occurs as a characteristic 8–15 cm bed in Coal 87, both in Hanna Draw and The Breaks. Limestone beds like this typically indicate freshwater conditions and are common in shallow ponds or littoral lacustrine environments with vegetation (Carroll and Bohacs, 1999; Alonso-Zarza, 2003; Bowen et al., 2008). The limited transport damage suggests that this bed was not a true coquina but deposited in a low-energy littoral lacustrine environment. Carbonate cementation and concretions occur throughout the section but significantly increase above Coal 87, suggesting that the basin became more enclosed. Some of the carbonate-cemented beds have been described as stromatolites in earlier work (Fig. 10; Davis, 2006).

Upward coarsening and thickening packages of light-gray, fine-grained, rippled sandstones suggest lacustrine shore-face environments (Fig. 10C). These are particularly prevalent in Coal 81–82, but also Coal 77–78. Thin, parallel-bedded sandstones with isolated and stacked ripples are encased within shales and occur in both The Breaks and Hanna Draw. Swaley cross bedding indicating wave action is not common but does occur in The Breaks and near Coal 76 in Hanna Draw. Siltstones and shales associated with these coarsening-up packages are remarkably void of organic material, which might indicate higher depositional energy near the lake margin.

Most common between Coal 81 and Coal 82 are extensive packages of unidirectional, up to 2- and 3-m-high foresets. Internal grain sizes fine toward foreset bottoms, and inversely graded internal laminae suggest that these are small deltas or mouth bars, like ones found in deposits of similar age in the nearby North Park–Middle Park Basin (Flores, 1990; Dechesne et al., 2013).

Noteworthy in Coal 90 are thin beds (1 cm thick) of gypsum within dark-brown to gray shales alternating with siltstones containing fish fragments (Fig. 10). The gypsum layers are bedding parallel and interpreted as primary deposit. They also corroborate more restricted conditions in this part of the section as indicated by the increase in carbonate concretions and cement.

Paleobotany

At Hanna Draw, plant localities below the Big Channel Complex are dominated by Metasequoia occidentalis (Newberry) Chaney (Cupressaceae), Platanites raynoldsii (Newberry) Manchester (Platanaceae), Trochodendroides genetrix (Newberry) Manchester (Cercidiphyllaceae), Zizyphoides flabela (Newberry) Crane, Manchester and Dilcher (Trochodendraceae), Archeampelos acerifolia (Newberry) Mciver and Basinger (Cercidiphyllaceae), and Macginitiea gracilis (Lesquereux) Wolfe and Wehr (Platanaceae), all of which are abundant throughout the Rocky Mountain basins during the early Paleogene, particularly in riparian environments. Taxa that are restricted to the Paleocene, including Cyclocarya brownii Manchester and Dilcher and Cornus swingii Manchester, Xiang, Kodrul, and Akhmentiev confirm that these sites predate the Paleocene–Eocene boundary. The first definitively Eocene macrofossils in the Hanna Draw section are Platycarya sp. leaves and catkins, which occur at the 707 and 728 m level in the Hanna Draw Section (Coal 84 of Dobbin et al., 1929; localities EC1501 and EC1506 and 1509, respectively). The first appearances of Lygodium kaulfussi (locality EC1509) and Salvinia preauriculata (locality EC1502) occur close by, respectively at 45 and 60 m higher in the section (Fig. 4; Supplement S1 [footnote 1]).

Early to middle Paleocene plant fossils in The Breaks were described and censused in Dunn (2003). Leaflets of Wing and Currano’s (2013) “Dicot sp. WW004,” a Bighorn Basin morphospecies that is common in the PETM interval and does not occur in any other part of the section, were recovered at the 120 m level in The Breaks section. Platycarya sp. and Cnemidaria magna leaves first occur laterally equivalent to 318 m in The Breaks section (150 m [500 ft]) above the top of Big Channel), and Lygodium kaulfussi occurs at ∼333 m, laterally projected into The Breaks section (165 m [540 ft] above Big Channel, Figs. 11A–11C). Leaves and pollen cones of Alnus sp. were recovered below the first appearance of Platycarya sp., at site ECHB 1609. Because Alnus leaves are restricted to the Eocene in the Greater Green River and Bighorn Basins (Wing, 1998; Wilf, 2000) and only occur above the first occurrence of Platycarya sp. in the Hanna Draw section, this stratigraphic level can also be classified as Eocene.

Palynologic occurrences of Platycarya platycaryoides (Fig. 11D) precede the leaf fossil occurrences in both The Breaks and Hanna Draw sections and help more broadly constrain the proposed PETM interval. Platycarya pollen, typical for the recovery phase of the carbon isotope excursion (CIE) in the Bighorn (Wing et al., 2005) and Powder River Basins (Wing et al., 2003), first occurs 54.5 m (180 ft) below the base of the first channel of the Big Channel Complex (sample 18BC-2.57 m) and again 32.6 m (106 ft) below the base of the Big Channel Complex (sample 18BC-24.4 m). It is then absent in all subsequent samples until it appears again in Coal 83, 39 m (128 ft) above the uppermost channel story of the BCC (sample 16BCr-155). A similar pattern is found in The Breaks where Platycarya platycaryoides first occurs 24 m (78 ft) below the base of the Big Channel sand (106 m in the measured section) and appears again in the mostly covered interval near the top or within the Big Channel at the 176 m level. Platycarya platycaryoides pollen is then more commonly found in samples beginning at the 206 m level in The Breaks section. Our first occurrence datum of Platycarya platycaryoides in The Breaks occurs lower in the section than previously reported (Lillegraven et al., 2004; Pew, 2014).

In the Big Channel Lateral (BCL) section, we were unable to detect the first pulse of Platycarya pollen below the Big Channel bed. Instead, it first appears 19 m (62 ft) above the Big Channel Complex, 170 m high in the local section (sample HB-RD17-007). At this sample location, it occurs abundantly as it does in The Breaks and Hanna Draw sections above the Big Channel Complexes at each of these sites.

In the Hanna Draw section, another key Eocene pollen taxon, Intratriporopollenites cf. instructus (Fig.11E), first appears with Playtcarya platycaryoides in samples 18BC-2.57m and 18BC-24.4m. Brosipollis striatus (Fig. 11F) occurs in several samples between 24.4 m and 129 m, and Retristephanocolpites sp. (Fig. 11G) occurs fleetingly at 29.6 and 36.1 m. These two taxa are known to occur only during the PETM in the Gulf Coast and Rocky Mountain region (Harrington et al., 2015). The fern spore Granulatisporites sp. is another Eocene indicator in the Gulf Coast (Harrington, 2003b) that also occurs in the Hanna Draw section at 24.4 m (Fig. 11H). Palynological samples between 50 and 123 m had generally poor palynomorph preservation. Following this interval, two new Eocene taxa from the Gulf Coast appear at 123 m in the section, Corsinipollenites psilatus (Fig. 11I) and Interpollis microsupplingensis.

Continental Mollusks

There are over 200 Paleogene continental molluscan localities in the Hanna Basin, of which the hydroboids, viviparids, and sphaeriids could be used to demarcate the Paleocene–Eocene boundary. Mollusk occurrence in our section also indicates more lacustrine than fluvial or paludal conditions (Fig. 5; Supplement S1 [footnote 1]). Unfortunately, preservation varies between localities, and species are not always easy to identify because the continental shells are (1) simple in form, (2) considered variable (in time, space, and environment), (3) frequently deformed, and (4) preserved without quality shell material. On top of this, some of the localities are in coarse-grained sandstone granule conglomerate interpreted as local lag accumulations (e.g., L7043 in The Breaks, 26 m above the top of Big Channel).

The hydrobioid Micropygus minutulus (Meek and Hayden), which was recognized from the late Tiffanian in the Williston Basin (type occurrence) and the Clarkforkian in central Utah (La Rocque, 1960), occurs in Coal 80 in Hanna Draw (L7306; Supplement S1 [footnote 1]). The younger molluscan assemblage at and below Coal 82 (L7358, just below the PETM; Supplement S1) represents conventional late Paleocene morphologies similar to the Bighorn and Powder River Basins. However, compared to the diverse assemblages there, species richness is limited in the Hanna Basin (Hartman and Roth, 1997, 1998). In Coal 82 in The Breaks, one small incomplete pleurocerid specimen was found with axial sculpture most commonly present on species of Elimia (L7528; Coal 82, TB; Supplement S1); however, exact age designation remains uncertain due to preservation and correlation issues. Wasatchian pleurocerid Elimia tenera (Hall) was discovered in Coal 85 in The Breaks (L7524; Supplement S1; see also Fig 12A).

Transported viviparids are not uncommon in the section, such as at L7523 in The Breaks (278 m; Supplement S1 [footnote 1]; at 92 m (300 ft) above the PETM; Fig 12B). Hanna Basin viviparid specimen identifications are most likely to be compared to Paludotrochus uniangulatus? (Hall), Paludotrochus paludinaeformis (Hall), and P. aff. P. wyomingensis (Meek) after Hartman (1984), which otherwise have an early Eocene distribution. Although incomplete and/or variously compressed, a few specimens bear traits that indicate assignment to the early Eocene Paludotrochus paludinaeformis (L7350, Coal 87, ∼245 m above the PETM in Hanna Draw), which is common in lower Eocene lacustrine lithic units elsewhere in Wyoming. T.W. Stanton (1916) identified a questioned occurrence (L3511) of the taxon, along with another early Eocene species, Physa pleromatis White, near Coal 90. The reason for absence of other early Eocene continental mollusks needs further study.

Bulk Carbon Isotope Curves

The bulk organic carbon isotope data are plotted on Figure 4. The main δ13Corg signal between Coal 77 and Coal 82 is constant along Hanna Draw and ranges from −25.1‰ to −26.6‰ VPDB (Supplement S3 for data and sample coordinates [footnote 1]). A large shift in δ13Corg occurs between the last occurrence of Paleocene plants and the first appearance of earliest Eocene plants in the interval between Coal 82 upper and Coal 83. Both the most positive (−23.5‰) and most negative (−30.8‰) values occur in this part of the section, and likely reflect the carbon isotope excursion (CIE) globally found at the PETM. Above Coal 83, isotope values range from −25‰ to −28.3‰.

Bulk δ13Corg isotopes in The Breaks vary from −25.2‰ to −27.8‰ and do not reveal a clear CIE. Exposure is poor in the Big Channel interval of The Breaks, precluding isotope samples from this part of the section. In The Breaks, the δ13Corg isotope signal appears to correspond with the pattern of a series of coarsening-up sequences, containing ripple beds and immature paleosols, often capped with thin gravel lags. The gravel lags, which often include reworked siderite clasts (such as in the Carbon Basin; Loope and Secord, 2017) and 2–5 cm angular clasts of organic, carbon-rich Cretaceous Mowry Shale (or other Cretaceous shales) derived from the emerging Seminoe Mountains and Freezeout Mountains.

Zircon Ages

Five of the very fine-grained mudstones samples suspected to be tonsteins were run for ID-TIMS U-Pb zircon dating. One of the five samples, RD0814-36, collected in Coal 88 in The Breaks, contains zircons that are ca. 54 Ma (Supplement S2a, Table S2-1 [footnote 1]). However, this sample also contained numerous euhedral zircon grains that are much older than the expected volcanic age (Supplement S2a, Figs. S2-1, S2-2, and S2-3). Nine zircons yielded dates of ca. 55 Ma (Supplement S2a; Fig. S2-4a), with a spread in 206Pb/238U dates from 57.0 to 53.8 Ma. The best estimate on the age of deposition comes from the Concordia Age (Ludwig, 1998) from the four youngest, overlapping, concordant single-grain analyses, 54.42 ± 0.27 Ma (95% confidence including decay constant errors; Supplement S2a, Fig. S2-4b).

Surprisingly, these young grains are from the euhedral, nearly equant population rather than the elongate grains that exhibit more characteristic ash-fall morphologies. The youngest grain dated from the “ash-fall” morphology (euhedral, elongate grains) sub-population in this sample is 56.0 ± 0.7 Ma (Table S2-1, Supplement S2a [footnote 1]). The five slightly older zircons, including two with apparent ash-fall morphologies, have 206Pb/238U dates that range from to 57.1 ± 1–55.2 ± 0.3 Ma. They are interpreted to reflect either pre-eruptive, antecrystic zircons from the magma chamber or slightly older volcanic grains from earlier events that were entrained during the latest eruption. Single-grain analyses from the other four samples are all older than 74 Ma (Supplement S2a), despite displaying euhedral morphologies, and are interpreted to be detrital.

To address the largely inherited component of zircon in the samples and to screen for more zircons that may represent the age of deposition of the strata, four of the mudstones plus four additional, fine-grained sandstone samples were analyzed by LA-ICPMS (Supplement S2a and S2b [footnote 1]). In general, the sandstone samples have very low zircon yields (fewer than 100 grains in all samples and generally less than 40) and few concordant grains that are Cenozoic in age. Sample RD0814-35, which is from nearly the same stratigraphic horizon in The Breaks as the sample that provided the ca. 54 Ma concordant zircon (RD0814-36), had a single young grain at ca. 54 Ma (Supplement S2b), which is the best limit on deposition age from the zircon data. Based on the zircon age distributions, it appears that there are very few if any ash-fall zircons, and it is possible that the sampled units are not directly derived from volcanic ash but rather fine-grained siliciclastic to clay-rich sediments in very low-energy environments, possibly fluvial floodplain deposits.

DISCUSSION

PETM

Diagnostic late Paleocene and early Eocene paleobotanical remains, incorporated with geochemistry, and stratigraphic correlations, allow us to more precisely place the PETM within the Hanna Formation. Occurrence of late Paleocene plants in Coal 82 and the presence of Eocene indicator taxa above this level best constrains the Paleocene–Eocene boundary in our data set both in Hanna Draw and The Breaks (Fig. 4). Platycarya is thought to be an immigrant taxon that first arrived in North America during the earliest Eocene from Europe where it occurs in the latest Paleocene (e.g., Jolley, 1997). As such, Platycarya pollen has historically been used as a marker taxon in defining early Eocene strata in the Western Interior (Nichols and Ott, 1978). Further refinements of when Platycarya pollen first appears in relation to the PETM have shown it occurs in the recovery phase of the CIE in the Bighorn and Powder River Basins (Wing et al., 2003, 2005) and during pre-CIE warming in the North Sea (Eldrett et al., 2014) and Gulf Coast (Sluijs et al., 2014). The first appearance of Platycarya platycaryoides in the Hanna Basin below the CIE, consistent with the higher-resolution records of the Gulf Coast and North Sea marine cores, confirms that Platycarya first occurs in the latest Paleocene in North America as well. The slightly later occurrence of Brosipollis striatus and Retristephanocolpites sp. ∼20 m higher than the first appearance datum (FAD) of Platycarya platycaryoides in the Hanna Draw section, is consistent with PETM occurrences of these taxa in the Gulf Coast and in other basins in Wyoming (Harrington et al., 2015). In total, in the Hanna Draw section where the palynology is best constrained to date, we find the first appearances of six taxa known from the PETM or earliest Eocene between 24.4 m (30 m below the first sand channel) and 123 m high in the section. Additionally, the presence of a Bighorn Basin PETM leaf morphotype within the Big Channel sequence in The Breaks provides additional evidence that the PETM event is captured at this stratigraphic level in multiple places in the basin.

Our geochemical data were collected over the entire studied interval to locate the CIE that marks the PETM. In the Hanna Draw section, bulk δ13Corg isotopes start deviating from a near-constant background value of about −26‰ (VPDB) below Coal 82, to values both more negative (minimum −30.8‰ VPDB) and more positive (maximum −23.5‰ VPDB) between Coal 82 and 83 (see Fig. 5), concurring with the PETM from our paleobotanical data. Just above Big Channel in Coal 83, the bulk carbon isotope signal stabilizes to more-or-less pre-CIE background values (Fig. 5); however, the signal remains slightly more erratic than before the CIE in both Hanna Draw and The Breaks (also Higgins, 2012).

Higgins (2012) originally identified an area with erratic δ13Corg values within Coal 84 and 85 as the CIE in The Breaks; however, our new isotope data, stratigraphic correlations, and paleobotanical data indicate that the CIE must occur lower than originally placed. Identifying a CIE with bulk δ13Corg in The Breaks has proven harder than in Hanna Draw because Big Channel is poorly exposed, and our sampling density has not been tight enough (Fig. 4). Contamination with modern soil or influx of allochthonous carbon due to weathering and erosion of older carbonaceous material in the surrounding drainage areas could be additional explanations for erratic bulk δ13Corg signals (Baczynski et al., 2016; Lyons et al., 2017).

Continental mollusks of the Hanna Formation are less diverse than in neighboring basins, and poor preservation of specimens limits us from using this data set to better constrain the PETM and early Eocene stratigraphy. However, the mollusk data set shows a change around the PETM (localities L7358, L7528, L7043, and L7523) and supports previous work on ostracods and mollusks (Glass, 1980; see Kirschner, 1984, p. 59). Preserved morphologies of caenogastropods (Viviparidae, Pleuroceridae, and Hydrobiidae), veneroids (Sphaeriidae), and unionioids (Unionidae) are consistent with taxa for a late Paleocene age in localities below Big Channel. Likewise, sparse specimens from localities above Big Channel indicate lower Eocene from different taxa in same families. A significant loss of taxa (caenogastropods, in particular), occurs across the Paleocene–Eocene boundary, but unlike the Bighorn and Powder River Basins, the Hanna Basin does not acquire any pulmonate taxa (Hartman and Roth, 1998).

In contrast with the significant changes in sedimentary facies documented in other basins that are generally more arid in setting (Foreman et al., 2012; Foreman 2014; Colombera et al., 2017), it is remarkable that facies distributions and channel dimensions within the PETM interval in the Hanna Basin are not as different between major channel complexes (Fig. 8). Comparison of facies within each of the large sandstone complexes shows a general decrease in the occurrence of large barforms and lag deposits upward in our studied section, plus an increase in soft sediment deformation, ripple, and climbing rippled sandstones. Big Channel does interrupt this trend and occurs at the identified PETM, between Coal 82upper and 83. Lags and large barforms are more abundant than in the channels directly below or above. Also, three channel intervals occur within the here-identified CIE, rather than one large amalgamated sandstone complex. The second and third stories are separated by an interval with abundant orange-to-red mottled paleosols, which is rare within the generally paludal deposits of the Hanna Basin.

Paleogeography

Regional compilations of paleoriver systems by Galloway et al. (2011), Smith et al. (2014b), and Sharman et al. (2017) indicate the presence of west- to east-directed rivers through the Hanna Basin and possible connections with a larger (California) River during the Paleocene. Paleoflow directions within the study area indicate a dominance of east- and northeast-directed flow, which corroborates this flow pattern and suggests basin-axial flow (Figs. 4 and 5; Ryan, 1977; Wroblewski, 2002).

Progressive basin isolation by the emergence of surrounding uplifts such as the Rawlins uplift, the Freezeout Mountains, Laramie Range, Medicine Bow, and Sierra Madre, coincides with the relative upward increase in ponded facies above Coal 87, and southward transport directions in TB and BMV indicate very limited basin-axial transport and increased influx from the northern basin margin. (Fig. 13; Blackstone, 1993; LeFebre, 1988; Perry and Flores, 1997; Wroblewski, 2002; Flores, 2003; Loope and Secord, 2017). Emerging uplifts caused drainage reorganization and possibly climatic rain shadows that changed water balance and sedimentation patterns in the basin. The gastropod limestone bed in Coal 87 and the evaporites in Coal 90 indicate more restricted conditions often seen at the late-stage fill of an overfilled lake basin and the transition into a more restricted basin fill (Carroll and Bohacs, 1999; Alonso-Zarza, 2003). The evaporites indicate that evaporation became greater than freshwater input, suggesting internal drainage and possibly the onset of drainage reversal from east- to west-directed rivers during the Eocene.

The repetitive pattern of laterally extensive sandstones that are similar in appearance (Figs. 3 and 4) throughout the entire Hanna Formation beyond the PETM, invites speculation on the origin of this cyclicity. Factors such as topography building, channel migration, and avulsions that are internal to a depositional system have been identified to cause an autocyclic succession of channel and floodplain deposits in the underlying Ferris Formation near the study area, however, at a smaller scale (Hajek et al., 2012). Aswasereelert et al. (2013), Smith et al. (2014b), and Noorbergen et al. (2017) link pronounced cyclicity in the Paleocene and Eocene sediments in the Bridger and Powder River Basins to climate and orbital signals such as 100 k.y. short eccentricity. Alternatively, the fluvial sandstone intervals at Big Channel and Camp Channel could reflect sedimentary responses to climate effects of early Eocene hyperthermals, because Big Channel coincides with the PETM based on our data. If the ca. 54.4 Ma zircon age from Coal 88 is indeed detrital, the timing of deposition of Camp Channel could coincide with the next large Eocene hyperthermal, ETM2 at ca. 54 Ma (Westerhold et al., 2007).

Provenance assessment derived from our zircon analysis does not reflect the ages of local Precambrian basement. The Hanna Basin is located near the junction of the >2.5 Ga Wyoming craton, 2.0–1.8 Ga Mojave province, and 1.8–1.7 Ga Yavapai province (e.g., Amato et al., 2008), a Proterozoic suture zone termed the Cheyenne belt (e.g., Duebendorfer et al., 2006). Since the majority of zircon grains analyzed appear to be inherited and/or detrital in nature, we created a composite plot of all samples with greater than ten individual grains analyzed, a total of 156 analyses. The composite detrital zircon Hanna Basin sample has a major peak at ca. 1.7 Ga, which is typical of basement terranes to the south, but has only few grains older than 1.8 Ga (Supplement S2a, Fig. S2-5 [footnote 1]). Other detrital zircon age populations present in the composite sample include ca. 650–500 Ma (peri-Gondwanan terranes) and ca. 480–270 Ma (Appalachian terranes) and a minor ca. 1100–900 Ma (Grenville) peak that suggests detrital zircon was sourced from the Appalachians. Appalachian detrital zircons first appear in the western United States during the Pennsylvanian–Permian (e.g., Ingleside Formation in Colorado Front Range; Nair et al., 2018). The Appalachian-derived detrital zircon could have been recycled from nearby Paleozoic and younger sediments, and our data are not conclusive in distinguishing local from more distant sediment sources.

CONCLUSIONS

This study constrains the Paleocene–Eocene boundary in the rapidly subsiding Hanna Basin and provides a detailed stratigraphic framework of fluvial, paludal, and lacustrine facies across the upper part of the Hanna Formation. Plant fossils and palynology have so far proved to be the most reliable tools to constrain the PETM. Bulk δ13Corg isotopes from carbonaceous shales show most negative values between Coal 82 and 83 in Hanna Draw corresponding with the here-defined PETM; however, bulk δ13Corg isotopes in The Breaks are harder to interpret and need denser sampling. The fluvial response to the PETM in the Hanna Basin as an increased pulse of coarse sediment coincides with sedimentary responses described in other Laramide basins where it is typically attributed to a more seasonal climate. Besides the occurrence of a large sandstone at the PETM, a repetitive pattern of similar large sand bodies alternating with coals exists throughout the Hanna Formation. It remains unknown if this cyclicity results from autocyclicity common to avulsive systems, drainage capturing, climatic signals such as orbital forcing, or other causes. The only date in our stratigraphic framework is the location of the PETM between Coal 82 and 83. Attempts to add time control with zircons remained inconclusive since the zircons were likely detrital and therefore report depositional ages, rather than absolute ages. Additional efforts to increase resolution in correlations and environmental interpretations, especially in the Eocene part of the section by using mollusks, have so far been unsuccessful due to the poor preservation of specimens and uncertainty tying into regional correlations. More age control would better document timing and rates of sedimentation in the basin.

The identification of the Paleocene–Eocene boundary in the Hanna Basin adds an important spatial data point for comparison of PETM sections in terrestrial basins across the western United States. The abundance of carbonaceous shales improves preservation of paleobotanical material and offers more detailed documentation of changes across the PETM compared to more arid basins, such as the Piceance and Bighorn Basins.

ACKNOWLEDGMENTS

This study was part of the mapping efforts by the U.S. Geological Survey (USGS) National Geologic Mapping Program and National Science Foundation grant EAR-145031 to Currano. Field assistance and discussions by Jenna West, Kristi Zellman, Robin Swank, Lauren Schmidt, Kymbre Skersies, Sarah Fanning, Lukas Lindquist, and Michael Loveland were highly appreciated. Burt and Kay Lynn Palm, the Q-Creek Ranch, Anadarko, Wyoming State lands, and the Bureau of Land Management (BLM) are thanked for land access. Work on BLM land was conducted under permit WY-197. We are thankful to Craig Johnson and Cayce Gulbransen of the USGS for help in the lab and insights with isotope analysis. Jen O’ Keefe is thanked for processing and interpreting part of the palynological material. Jeremy Havens is thanked for drafting support and suggestions. Chamberlain was partially supported from Mega-Grant 14.Y26.31.0012 and Science Research project 18-17-00240 of the government of the Russian Federation. Joseph Hartman thanks with appreciation his Hanna Basin museum and field crew Aaron Sobbe and Danielle Zinsmaster of the Harold Hamm School of Geology and Geological Engineering, and Don McCollor and the Energy & Environmental Research Center, at the University of North Dakota (UND). Access to the Gary Glass collections was provided by Laura Vietti, Collections Manager of the Geological Museum, University of Wyoming (UW), and Donald Boyd, Professor Emeritus (UW). Unpublished documents were generously provided by Chris Carroll of the Wyoming State Geological Survey (WGS) and Gary Glass (formerly of WGS). High-resolution images of published material were kindly provided by Jay Lillegraven, Emeritus Professor (UW). Support for this project was provided by the Bud and Mardi Paleontology Fund (UND) and John Reid Fund (UND).

USGS reviewers Debra Higley and Julie Herrick are thanked for helpful comments. Anonymous GSA reviewers and the editors of Geosphere are thanked for their helpful and constructive comments to improve this document for publication.

Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. All pictures were taken by M. Dechesne, unless otherwise noted.

1Supplemental Materials. Supplement S1: Full stratigraphic diagram for the upper Hanna Formation between Coal 77 and 90 between Hanna Draw and The Breaks. Locality and sample data are plotted. Supplement S2a: Methods, figures and results for the thermal ionization mass spectrometry (TIMS) and laser ablation–inductively coupled mass spectrometry (LA–ICPMS) zircon analysis. Supplement S2b: Data tables for the laser ablation zircon analysis. Supplement S3: Tables listing bulk carbon isotope data conducted for this study. Please visit https://doi.org/10.1130/GES02118.S1 or access the full-text article on www.gsapubs.org to view the Supplemental Materials.
Science Editor: David E. Fastovsky
2J.H. Hartman has maintained a North American continental molluscan locality numbering system (L-numbers [Lnos]) since 1978. Lnos reported in Supplement S1 text (footnote 1).
Gold Open Access: This paper is published under the terms of the CC-BY-NC license.