Middle Miocene faulting and basin evolution during central Basin and Range extension: A detailed record from the upper Horse Spring Formation and red sandstone unit, Lake Mead region, Nevada, USA

Miocene basins of the Lake Mead region (southwestern United States) contain a well-exposed record of rifting, structural deformation, diverse depositional settings, and the evolving paleogeography of the eastern central Basin and Range. Basin fill spans the entire Miocene, from 24 to 5 Ma, and records evidence of pre-extensional paleogeography as well as the structures and depositional environments during the initial, main, and waning periods of Basin and Range extension. The stratigraphic and structural architecture of sedimentary basin fill is the product of complex interactions between tectonics, climate, and autogenic processes (e.g., spontaneous channel avulsion events and alluvial-fan lobe switching; Straub et al., 2020). It is rare, however, to be able to tease apart these controls with much confidence due to limitations of exposure and chronostratigraphic control. The Miocene Horse Spring Formation and informally named red sandstone unit, exposed east of Las Vegas (Nevada) and northwest of Lake Mead, provide a unique opportunity for beginning to understand these controls for a rift basin (Fig. 1). Geologic mapping at ~1:3000 scale, numerous measured sections, and radiogenic and tephrochronologic dating of common tuffs have allowed us, with a high degree of confidence, to determine the spatial and temporal evolution of fault sets that created and deformed sedimentary basins, hypothesize the influence of the three major controls on sedimentation, and construct a refined picture of the paleogeographic evolution of the region.

Over the span of several decades, this region, from Las Vegas to the western edge of the Colorado Plateau (Fig. 1), has been the focus of several seminal structural studies that challenged and advanced our understanding of continental rifting (e.g., Anderson, 1971, 1973; Angelier et al., 1985; Çakir et al., 1988; Faulds and Varga, 1998; Wernicke et al., 1988; Wernicke and Axen, 1988; Brady et al., 2000). These studies focused on either the largest faults or the complexity of faults in a localized area and led to numerous competing models. By examining the stratigraphic and structural record of deformation in detail across the Lake Mead region, we set out to address conflicting models on the evolution of faulting, drivers of extension, and the link between the two. Anderson and Barnhard (1993) and Anderson et al. (1994) proposed a model of continuous deformation driven by tectonic extrusion to the west due to southward crustal flow from the elevated northern Basin and Range. They noted that the Lake Mead area is a transition zone between the lower topography of the central and southern Basin and Range and the higher topography of the northern Basin and Range. In their model, north-south constriction created local folds, strike-slip faults, and east-west extension. Other researchers proposed that extension took place in discrete stages with different sets of faults. Duebendorfer and Simpson (1994) analyzed fault data on either side of the Las Vegas Valley shear zone and proposed that faulting occurred in four stages. Umhoefer et al. (2010) argued for discrete stages of extension across the Lake Mead region in a two-stage model. This model includes initial rapid extension by detachment and high-angle normal faulting from 17 to 14.5 Ma associated with the South Virgin–White Hills detachment fault, followed by a transition to more diffuse, transtensional deformation with a jump in faulting to the west after 14.5 Ma.

To address these hypotheses and fill in major data gaps, we studied the timing and kinematics of fault sets and folds as well as the location and rates of sedimentation through time in many of the well-exposed sedimentary basins across the Lake Mead region. We mapped lateral and vertical stratigraphic facies variations at scales of 1:5000 and 1:3000 and collected and analyzed hundreds of tuff samples. Using our new data, including mapping, tens of detailed measured sections, numerous individual mapped marker beds, ⁴⁰Ar/³⁹Ar dates, and a few hundred geochemical analyses for tephrochronology, we were able to create a new chronostratigraphic framework at a detailed level that is not achievable in most basins. This work and new framework provide an opportunity for building on previous structural work and address the following questions: (1) Did extension in the eastern central Basin and Range proceed in discrete stages from detachment faulting to transtensional faulting (Umhoefer et al., 2010) or in a continuous process due to the gravitational collapse of the northern Basin and Range into the Lake Mead region accompanied by westward tectonic escape (Anderson et al., 1994; Anderson and Beard, 2010)? (2) If it proceeded in discrete stages, what was the timing and nature of deformation during each stage? (3) Are folds and faults in the Lake Mead region associated with regional constriction, or did they form during transtensional faulting? We present our new stratigraphic and structural data and show that faulting occurred in different stages and that many of the folds formed as part of this faulting.

To better show our interpretation of the stratigraphic and structural evolution of the White Basin, we present paleogeographic interpretations from 14.5 to ca. 11.4 Ma. This is one piece in our larger ongoing study to determine the paleogeographic evolution of the Lake Mead region from 25 to ca. 11 Ma and thus contribute to better understanding of the evolution of the Colorado River and Colorado Plateau uplift. The paleogeographic evolution of the Lake Mead region over the last 25 m.y. is inextricably linked to the uplift of the Colorado Plateau and development of the Colorado River. Understanding the timing of extension and development of topographic changes in the Lake Mead region helps constrain competing models of river integration and elevation uplift (e.g., Wernicke, 2011; Flowers and Farley, 2012; Lee et al., 2013; Karlstrom et al., 2014, 2013). Although it is not a goal of this paper to address these competing models, we note that the new data presented here, combined with past publications (e.g., Lamb et al., 2015, 2018) and planned future work, will contribute to a more complete understanding of the Lake Mead landscape evolution during the Miocene.

All of these new data also allow us to begin to hypothesize on the influence of tectonics versus additional controls on the complex stratigraphy of the Horse Spring Formation. Given that the stratigraphic fill of a basin can never record every event and is commonly partially or mostly covered, it is typically very difficult to sort out controls on sedimentation. In addition, many basins lack adequate age control or contain facies that do not hold clues to climate variations. Tectonics affects the topography of basins, climate changes affect precipitation and evaporation, and autogenic processes can change the locus of deposition. Thus, using the stratigraphic record to interpret past tectonic and climate events is difficult at best. In the Lake Mead area, the fortuitous combination of exposure, numerous tuffs, and variable facies holds key clues regarding changing depositional environments and the input of climatic factors. The sedimentary facies include (1) thick sequences of microbialite carbonates that record stable-isotopic changes and (2) exposures of varying strata that represent both lacustrine and fluvial conditions. We present an initial comparison of our data with global solar fluctuations, consider the role of autogenic processes, and, as the first step in an ongoing study, pose a preliminary hypothesis on the ways in which climate may have impacted the sedimentary fill of the White Basin.

The first major stage of faulting in the Lake Mead region produced the Thumb Member (Horse Spring Formation) basin in the hanging wall of the South Virgin–White Hills detachment fault (Duebendorfer et al., 2010; Lamb et al., 2010; Umhoefer et al., 2010; Anderson, 2012; Samra, 2013). In this study, we focus on the next stage of extension, from 14.5 to ca. 11.5 Ma, when transtensional processes split the Thumb Member basin into separate smaller basins. We focus on the White Basin (Figs. 1–3), where strata recording the development and demise of the Bitter Ridge Limestone Member (Horse Spring Formation) lake as well as changes in faulting and the resulting new depositional environments of the Lovell Wash and red sandstone are especially well exposed. We present some new data from the Lovell Wash area, but a future paper is planned to address the southern part of the Bitter Ridge Limestone lake and its evolution along the Las Vegas Valley shear zone.

The oldest rocks of the Lake Mead region are Proterozoic crystalline rocks, exposed in the eastern and southern part of the Lake Mead area (Fig. 1). Paleozoic sedimentary units record marine passive-margin deposition and are overlain by lower Mesozoic nonmarine strata (Bohannon, 1984; Beard, 1996; Beard et al., 2007). Cretaceous-age Sevier thrusting north and west of Lake Mead placed Paleozoic strata on Mesozoic rocks (Bohannon, 1983a, 1983b; Wernicke et al., 1988), and Laramide deformation affected strata south of Lake Mead (Bohannon, 1984; Faulds et al., 2001; Beard and Faulds, 2011). The long period of contraction was followed by tectonic quiescence and erosion and locally the deposition of the lower part of the upper Oligocene–lower Miocene Rainbow Gardens Formation (Bohannon, 1984; Lamb et al., 2015, 2018). This unit also records initial extension in the region at 20 Ma (Lamb et al., 2018; Conrad, 2018). The main phase of east-west–directed extension began ca. 17 Ma and continued until at least 10 Ma. This resulted in the development of numerous basins, as recorded by the 17 Ma to ca. 13 Ma Horse Spring Formation (Bohannon, 1984; Beard, 1996; Lamb et al., 2005) and the younger, overlying, informally named red sandstone unit (e.g., Bohannon, 1984; Beard et al., 2007; Duebendorfer and Wallin, 1991). The Muddy Creek Formation and Quaternary alluvial deposits overlie the extensional deposits (Bohannon, 1984; Beard et al., 2007).

The Horse Spring Formation is exposed across the Lake Mead region (Figs. 1 and 2; Bohannon, 1984; Beard, et al., 2007). The lowest Thumb Member records the initial and most rapid phase of extension from ca. 17 Ma to ca. 14 Ma (Figs. 2 and 3; Bohannon, 1984; Beard, 1996; Castor et al., 2000). This unit contains a wide range of facies that represent a mix of terrestrial depositional environments, including alluvial fan, fluvial, and lacustrine deposits; fluvial facies dominate. The upper Horse Spring Formation consists of the 14.5–13.86 Ma Bitter Ridge Limestone Member, a thick-bedded, uniform, microbialite lacustrine carbonate, and the 13.86–12.73 Ma Lovell Wash Member (Figs. 2–4). Like the Thumb Member, the Lovell Wash Member comprises a mix of evaporites, clastics, carbonate, and tuffaceous units representing a range of environments (Bohannon, 1984; Lamb et al., 2010).

The informally named red sandstone (Bohannon, 1984) is found mainly across the western Lake Mead region (Figs. 1 and 2) in separate basins. Duebendorfer and Wallin (1991) and Cassidy (2016) discussed red sandstone deposits east of Frenchman Mountain and south of the Las Vegas Valley shear zone. We focus here only on red sandstone deposits within the White Basin and build on the work of Bohannon (1984). This unit contains finer-grained facies, both clastic and gypsiferous, throughout the western basin and coarser-grained clastic and carbonate faces along the basin margin.

Tuff samples were collected throughout the region for both geochronologic and geochemical analysis. Samples were prepared for electron microprobe analysis at the New Mexico Institute of Mining and Technology (Socorro, New Mexico, USA), and results were used to determine tephrochronologic correlations and suitability for ⁴⁰Ar/³⁹Ar dating (see Item S1 in the Supplemental Material¹ for full details and additional data). All six new ⁴⁰Ar/³⁹Ar ages were determined at the New Mexico Geochronology Research Laboratory (Socorro, New Mexico) using primarily sanidine, with some samples also containing anorthoclase. Based on a revised date of the Fish Canyon Tuff of Kuiper et al. (2008), we recalculated two previously dated tuffs presented in Hickson et al. (2010; see Item S1 for details), and all new data reported here use the Kuiper et al. (2008) calibration as well as the total ⁴⁰K decay constant of 5.463e^–10/a of Min et al. (2000). Complete argon geochronology methods along with additional data are in Item S1. Forty-four samples were collected and analyzed for geochemical correlations (see Item S1 for details). These were used with field-based stratigraphic analyses and field mapping to correlate sections across faults.

We measured more than 30 detailed stratigraphic sections across the White Basin. We mapped at scales of 1:3000–1:5000 to identify lithofacies, lithofacies associations, and marker beds (Item S2, see ^{footnote 1}). Marker beds are mainly unique tuffs, identified by both field and geochemical characteristics, or distinctive sedimentary beds. A few tuffs are very distinctive in terms of their composition and weathering characteristics, while others are part of a distinctive stratigraphic package, making ideal marker beds. These marker beds were walked out in the field as part of the detailed mapping.

Based on our new map data, we delineated five structural domains based on the orientation and location of major faults, dip of bedding, and structural style of deformation in each area (Fig. 3). We analyzed bedding attitudes within each domain using the Stereonet program of Allmendinger et al. (2012) and faults using FaultKin (Marrett and Allmendinger, 1990).

The sedimentary fill of the White Basin and Lovell Wash syncline areas is stratigraphically complex: there are widespread lateral and vertical facies changes, commonly on a scale of a few meters or tens of meters. Figures 4–8 show simplified stratigraphic sections and correlation diagrams based on a subset of more than 30 detailed measured sections. The detailed versions are available in the Supplemental Material (Items S3–S8, see ^{footnote 1}). Geochronologic and tephrochronologic data from numerous tuffs strengthen and constrain correlations across the basin.

We present six new ⁴⁰Ar/³⁹Ar dates from White Basin tuffs: three from the Lovell Wash Member and three from the red sandstone unit (Table 1; see Item S1, ^{footnote 1}, for probability plots and additional details). The argon ages are based on single-crystal laser fusion data, and the assigned age represents the inverse variance weighted mean of the selected analyses. Between 30 and 11 sanidine and anorthoclase crystals were dated for each sample. The samples give weighted mean ages ranging between ca. 13.4 and 11.8 Ma and in general have slightly to moderately elevated mean squared weighted deviation (MSWD) values. The oldest date at 13.43 ± 0.02 Ma, tuff sample LMLL-258, is from the Lovell Wash syncline area (Fig. 2), southwest of the White Basin. Six of 14 crystals define the preferred age which, based on the K/Ca values, is a bimodal distribution of sanidine and anorthoclase crystals. Two other samples (LMLL-241 and LMLL-243) are from the Lovell Wash Member tuffs in section B in the western White Basin (Fig. 3; Table 1). These samples yield ages of 13.20 ± 0.04 Ma and 12.93 ± 0.02 Ma (samples LMLL-241 and LMLL-243, respectively). Both samples appear to have inherited grains that are moderately older than the juvenile populations that define the preferred weighted mean ages. The oldest red sandstone sample is sample 19-WB-01 with a date of 12.11 ± 0.01 Ma. This sample is coarse grained and gives very precise individual grain dates with a small standard deviation and a tight distribution of K/Ca values. There is one xenocrystic grain that is much older than the main population. The other two red sandstone samples are sample 19-WB-263 from section RR3 with an age of 11.88 ± 0.15 Ma and sample 19-WB-42 from section Y at 11.80 ± 0.01 Ma. Twenty-eight of 30 crystals analyzed from sample 19-WB-263 yield the reported weighted mean age, and individual crystals have relatively low precision due to small crystal size. The distribution of ages is fairly scattered (MSWD = 12.76), and this could be related to minor argon loss, incorporation of inherited crystals, or both. Culling the data into smaller subsets would not significantly change the reported age, and thus the majority of the analyses are used for age determination. Sample 19-WB-42 yields a well-defined population of ages for 18 of the 19 grains dated and it has a bimodal distribution of sanidine and anorthoclase. For samples dated with the ARGUS VI mass spectrometer that have very high precision, it is common to observe moderately scattered distributions. This could be related to small geological factors such as minor excess argon hosted in melt inclusions (e.g., Winick et al., 2001) but can also be related to grain-to-grain variation in neutron flux (Heizler et al., 2021). For this study, the weighted mean ages, despite the somewhat elevated MSWD values, are interpreted to yield robust tephra depositional ages.

We recalculated ages on two tuffs within the Bitter Ridge Limestone. Sample 04-BSQ-416 (Hickson et al., 2010) is from the eastern edge of the Bitter Ridge (Fig. 5A) and has an updated age of 14.41 ± 0.10 Ma. Sample O5-BS-35 is from the eastern White Basin and has an updated age of 13.89 ± 0.20 Ma.

Numerous tuffs were also analyzed and tested for geochemical correlations. Strong geochemical correlations are shown and used in conjunction with the stratigraphy to align the sections (Figs. 5 and 6; see Item S1, ^{footnote 1}, for details on the methods used for geochemical correlation).

The White Basin contains a mix of lithologies that vary both laterally and vertically. Lithofacies are defined as either a single lithology or a specific mix of lithologies found together that indicate deposition in a specific environment. Table 2 depicts the main lithofacies we defined for the White Basin as well as our interpretation of the depositional environment in which each lithofacies formed. Lithofacies A–G (Table 2) all contain limestone and indicate deposition within a lacustrine setting, likely shallow ponds and lakes. Lithofacies H contains a mix of gypsum, mudstone, and marl and also records lacustrine deposition. Lithofacies I is a limey sandstone that is interpreted to represent deposition in a palustrine setting in which streams deposited sands into the margins of a lake. Lithofacies J is a mix of sandstone and mudstone with a tephra component and records fluvial sedimentation with occasional input from distant volcanic eruptions. Finally, lithofacies K is predominantly sandstone formed by fluvial processes. These lithofacies are used in the data and interpretations presented for each member below. For some locations, we subdivided lithofacies with multiple lithologies into a set of more detailed facies associations based on the percentages of lithologies. For example, in Figure 6, lithofacies H encompasses “g” which is more than 80% gypsum, “mg” which is a mix of both mudstone and gypsum, and “mglt” which is mostly mudstone and gypsum with some limestone and tuffaceous material.

The Bitter Ridge Limestone (Figs. 2–5; Table 2) has been mapped east of Las Vegas at Frenchman Mountain (Fig. 1) and on the western, southern, and eastern sides of the Muddy Mountains (Fig. 2; Bohannon, 1983a). Our ongoing mapping and isotopic work on these strata suggest that this formation formed in multiple lakes, not one continuous lake (Schanzenbach et al., 2011; Pomerleau et al., 2014). We focus here on the main lake basin whose deposits are located from ~2 km west of West End Wash in the southern Gale Hills to the Lovell Wash area (section Z on Fig. 2) northeastward to the White Basin (section L on Fig. 2), where the ~250-m-thick Bitter Ridge Limestone Member crops out in a series of fault blocks that form a mostly continuous ridge. The main study area for this paper is the northern part of that belt in the White Basin (Fig. 3).

The majority of each measured section in the Bitter Ridge Limestone is composed of stratiform, crinkly laminated limestone beds with millimeter-, centimeter-, or mixed millimeter-centimeter–scale laminae (lithofacies A–C of Table 2; see details on the measured sections in Item S3, ^{footnote 1}; also see Hickson et al. [2010] for more details on the Bitter Ridge Limestone facies). Lithofacies D–G (Table 2) are also present but to a far lesser degree, and for the purposes of this study, the Bitter Ridge Limestone is vertically and laterally very uniform. Hickson et al. (2010) described the depositional environment of the Bitter Ridge Limestone Member as a persistent shallow lake, with depths of a few meters, in which constant subsidence allowed for the continual accumulation of >200 m of microbially laminated limestone. We interpret the lithofacies of the Bitter Ridge Limestone to have a mixed microbial and clastic origin, most likely controlled by small changes in water depth and proximity to shoreline. Because there are only rare macrophytes and no plant or animal fossils within this member, it is likely that the lake was alkaline.

The strata of the Bitter Ridge Limestone form stratigraphic packages that are visible within the cliffs that make up the Bitter Ridge (Fig. 5) and aid in correlation of this member. We divided the Bitter Ridge Limestone Member into six mesoscale units. Mesoscale units are defined as packages of beds that are similar in thickness, composition, color, and weathering characteristics. They are typically tens of meters thick and comprise a uniform layer distinct from units above and below. These are analogous but not identical to lithofacies associations. We were able to trace these six mesoscale units along the cliff face for the entire 11 km of the ridge in the southern White Basin. The thicker centimeter-scale laminated facies (lithofacies A, Lcrc, in Table 2) weather as resistant blocky beds, as in mesoscale unit M1 (Figs. 5C, 5D, and 5F). The thinner millimeter-scale laminated facies (lithofacies B, Lcrm, in Table 2) weather more readily, forming crumbly layers or recessive, covered intervals, such as in mesoscale units M2 and M4 (Fig. 5). The strong weathering variations are visible from a distance and, when combined with detailed section measuring, allow for long-distance correlations across the entire outcrop belt. The mesoscale units of the Bitter Ridge Limestone at the Lovell Wash syncline location (Fig. 2) are harder to decipher because the outcrop is in a slot canyon where the rock is polished and unweathered. The limestones here display incredible detail of the laminations and sedimentary features but do not show the strong differences in weathering that stand out on the cliff facies. Tuffs in the upper part of the member, however, allow for correlation across the basin (Fig. 5A). The continuity of these mesoscale units across the outcrop belt as well as the detailed measured sections all point to a single lake with remarkably homogenous conditions throughout its existence.

Southeast of the East Longwell Ridge there is a thick unit of gypsum that may be coeval with the Bitter Ridge Limestone (purple dashed line on Fig. 2). Bohannon (1984) assigned these beds to the upper Thumb Member. Recently dated tuffs and sedimentation rates of units stratigraphically below the gypsum indicate that the gypsum could be the same age as the limestone. Evaporitic lakes commonly produce a bullseye pattern in plan view, with gypsum deposition in the center and limestone on the margins (e.g., Warren and Kendall, 1985). This pattern is common in older Thumb Member deposits and may have persisted during deposition of the Bitter Ridge Limestone.

The Lovell Wash Member conformably overlies the Bitter Ridge Limestone Member and contrasts sharply with it. Whereas the Bitter Ridge Limestone is fairly uniform both laterally and vertically, the Lovell Wash Member is extremely variable (Figs. 4 and 6; Table 2; see Items S4–S6, ^{footnote 1}, for detailed measured sections and location of samples). Figure 6 presents three stratigraphic correlation diagrams based on the 30 measured sections from the Lovell Wash syncline area and across the White Basin. These measured sections document a range of evaporitic, carbonate, and siliciclastic lithologies, including almost all of the lithofacies in Table 2. Gypsum is the main evaporitic mineral, but locally borate is present and has been mined (Castor, 1993). The limestones display a range of textures, as depicted in Table 2.

Individual beds are grouped into facies associations based on the lithologies present, shown as colored polygons on Figure 6. Although they vary laterally, the facies associations record major trends and transitions through time, and thus the facies associations can be grouped into 10 stratigraphic packages, I to X (Figs. 4 and 6). Every package except one contains multiple facies associations, but the facies associations are typically related and represent deposition in similar environments. Correlations using physically distinctive beds, tephrochronology, and dated tuffs were critical for defining these stratigraphic packages.

The lowest unit of the Lovell Wash Member, package I, consists of red-weathering sandstone and mudstone, informally called the “redbeds” by Castor (1993), and local conglomerate beds. Package 1 rests conformably on the Bitter Ridge Limestone across the basin except in the far northeast where it lies directly on upper Paleozoic strata at the basin margin. The transition from Bitter Ridge Limestone to Lovell Wash Member is fairly sharp in most of the basin but becomes gradational toward the southwestern basin margin, where the upper part of the Bitter Ridge Limestone contains red-weathering sandstone and mudstone interbedded with limestone (i.e., section B in Fig. 5). Closer to the Muddy Mountains, the ratio of sandstone to limestone in the uppermost Bitter Ridge Limestone increases. Package I records a major basinwide change in the depositional environment after 200+ m and 600–700 k.y. of lacustrine limestone deposition. The long-lived stable lake recorded by the Bitter Ridge Limestone appears to have dried up and/or been swamped with clastic input as fluvial conditions were established.

Package II contains a mix of thinner limestone beds, gypsum, and borate, with some mudstone (Fig. 6A). In the eastern part of the basin, gypsum and limestone predominate over mudstone and sandstone. Package II is thickest at section F where the most gypsum is found (Fig. 6A). Section D is the westernmost section that contains gypsum. Between sections D and F, there are also areas that contain borate (Castor, 1993). Westward, the basin contains more mudstone and sandstone. We interpret that package II records a transition back to widespread lacustrine conditions but the lake is more variable across the basin when compared to the Bitter Ridge Limestone lake and likely had more varying depths laterally and through time compared to the Bitter Ridge Limestone lake. Clastics entered the lake via streams on the lake margins.

Package II is gradationally overlain by package III, which is characterized by generally finer-grained carbonate units across the basin and a lack of gypsum (Fig. 6). Mudstone and marl are the main lithologies with minor limestone in each section, except for the southwesternmost section A, which contains mudstone and sandstone (Fig. 6A). In the far northeastern corner of the basin, section R (Fig. 6A) contains gypsum, but this may be a small, separate basin. Many of the mudstone units are not well exposed, making it difficult to interpret a depositional environment in detail. We suggest this package represents mixed fluvio-lacustrine conditions. Package III at section Z (Fig. 6C) contains a mix of sandstone and carbonate units, suggesting a shallowing of water depths and transition to the clastic-rich package IV. Package III at other locations (Fig. 6A) appears transitional between packages II and IV as well.

Package IV records a major shift back to sandstone deposition with a significant coarse-grained clastic input (Fig. 6A). Throughout the eastern half of the basin, there is a distinctive olive green, tuffaceous coarse sandstone unit that is as much as 45 m thick. The unit is more extensive than is implied by Figure 6A due to complexity of faulting. Individual beds contain planar lamination or cross-stratification, indicative of unidirectional currents. Section A contains tuffaceous coarse sandstone and granule conglomerate beds as much as a few tens of meters thick. Section Z (Fig. 6C) records the coarsest clastic input within the basin. We interpret package IV deposits as fluvial. Two thick beds in section Z containing cobbles and boulders, including two blocks of Bitter Ridge Limestone 1–2 m wide, are highly disorganized and chaotic and are interpreted as debris-flow units. Such deposits indicate close proximity to the basin margin and paleotopography. Large (1 m wide, 10 m tall) sandstone injection features, sand-volcano necks, interrupt this clastic package and may be evidence of paleoseismicity.

In the eastern versus western halves of the basin, packages V–VII contain very different lithofacies (Fig. 6A). In the west, package V is similar to the underlying package IV, made up of mostly sandstone and mudstone with tuffaceous input. In section A it is more tuffaceous, whereas in section B it contains several meters of red-weathering mudstone. Both suggest the continuation of fluvial conditions. Package VI, in the western part of the basin, contains thin limestone beds, suggesting the development of intermittent but relatively widespread shallow ponds or wetlands. Package VII contains clastic lithofacies and records a return to fluvial conditions. Packages V–VII on the western side of the basin contrast sharply with their counterparts in sections G, F, and S in the eastern half of the basin. Packages V–VII in the eastern sections contain mostly gypsum with some limestone, suggesting the existence of a mostly evaporitic lake for more than 250,000 years.

Package VIII is a unique microbialite-rich unit. In general, it is a 10–15-m-thick limestone unit found across the basin with many domal stromatolites and thrombolites. In detail, it varies widely from fault block to fault block and contains highly unusual and variable microbialite facies and, in a few fault blocks, is a sandstone-rich unit. In Figure 7, we show examples of lithologic and thickness variations across a series of faults in the western part of the basin around section B and suggest these variations are due to syn-depositional growth faults creating varying water depths within shallow ponds. In one location at section B, there are domal stromatolites as much as 1 m wide within the 15-m-thick limestone unit of package VIII (Table 2, lithofacies E; Fig. 7). Just to the west across a fault, the same unit is only 4 m thick and there are only crinkly laminated stratiform stromatolites. In a few fault blocks in the center of the basin (between sections G and F of Fig. 6A), the limestone is replaced by orange- and yellow-weathering sandstones before transitioning back to microbialite limestone in the area of section F. At section F, we find a wide diversity of stromatolitic and thrombolitic morphologies and textures, suggesting shallow lacustrine conditions (Hickson et al., 2022). We interpret the depositional environment of package VIII as a series of small lakes and ponds with varying depths. The varying water depths produced different microbial communities that in turn created a variety of textures within the limestone beds (see Hickson et al., 2022).

Above the microbialite-rich package VIII, package IX is a distinctive unit characterized by mostly covered fine-grained units punctuated every 10–20 m with 0.5–1-m-thick resistant limestone beds and 10-cm-to 1-m-thick tuffs, tuffaceous sandstone beds, and reworked sandy tuffs. Variations in the exposed units document both vertical and lateral depositional changes. On the eastern side of the basin, sections F and S contain mostly gypsum and limestone. Section A in the far west contains sandstone and mudstone. In between, in the central and west-central portions of the basin, thin, exposed units of limestone and sandstone are found every 2–5 m within mostly covered fine-grained units. We interpret package IX to record predominantly lacustrine conditions on the eastern side of the basin, fluvial conditions on the western basin margin, and a mix of fluvial and lacustrine conditions in between in the central and west-central portions of the basin. We envision that this central portion of the basin contained changing conditions with occasional, short-lived shallow ponds here and there, interspersed with fluvial deposition. In one well-exposed location, between sections B and C, most of the units are lacustrine limestone, with only one bed of oxidized, reddish-weathering sandstone, suggesting small streams depositing sediment, not a long-lived fluvial system transporting sediment across the basin, like in package IV. At two time periods, ca. 13.12 and ca. 13.03 Ma, slightly thicker microbialite limestones were deposited across the basin (see dark blue lines in package IX on Fig. 6A), suggesting more widespread lacustrine conditions. We note that although the facies associations from packages III and IX on Figure 6 are similar, the sedimentation patterns are different. Within package III in the eastern part of the basin, for example, the system fines upwards from thicker limestone units to predominately fine-grained units, whereas in package IX, there is a cyclic pattern of thin limestone and sandstone beds alternating with fine-grained units.

Package X is similar to package IX in that it contains a few thin resistant beds of limestone and sandstone separated by several meters of mostly covered fine-grained units. Package X, however, contains fewer limestones and more sandstones across the basin. We suggest package X depositional environments were both fluvial and lacustrine but with fewer periods of lacustrine deposition than in package IX.

Overall, Figures 4 and 6A–6C demonstrate the thickness variations across the basin. Sections in the eastern half of the basin are consistently thicker than in the western half. The base of the Lovell Wash Member to the 13.43 tuff in package VI at section F is 287 m whereas at section B it is 197 m. From the 13.43 tuff to the 12.93 tuff in section F there is 203 m of section in contrast to 168 m at section B. Thus, in the lower part of the Lovell Wash Member, the eastern depocenter (section F) is 1.5 times thicker than in the area near section B; in the upper part of the Lovell Wash Member, section F is 1.2 times thicker.

In the White Basin, the red sandstone unit conformably overlies the Lovell Wash Member. The base of the red sandstone unit, defined by Bohannon (1984), is marked by a change to gypsum found across the basin (Fig. 6A), with minor limestone and siliciclastics at the basin margins (Fig. 8). Above this, much of the central part of the basin is covered but contains a mix of mostly red siltstone and gypsum with some tuffs and red sandstone where exposed. The widespread gypsum indicates the establishment of a single continuous evaporitic lake once again. The exposed gypsum at section B (Fig. 8) and mapped in other parts of the basin records evaporative lacustrine conditions for ~500,000 years.

Figure 8 contains several measured sections from the margin of the basin where exposures are better than in the center of the basin and the detailed stratigraphy, including numerous distinctive tuffs, could be measured and walked out, respectively (sections E, Y, and RR1 on Fig. 8; see additional details and sections in Item S8, ^{footnote 1}). Basin-margin facies include sandstone, tuffaceous sandstone, and limestone. On the western side of the White Basin, sandstone and limestone in the vicinity of sections B, Y, and E transition laterally into conglomeratic facies near the Muddy Peak fault. Clasts within the coarse-grained units are predominantly Paleozoic limestones derived from the footwall, as first recorded by Bohannon (1984). The basin-margin units represent deposition from streams and shallow lacustrine settings.

We use detailed stratigraphy and dated tuffs to calculate sedimentation accumulation rates across the basin for the Bitter Ridge Limestone and Lovell Wash Members of the Horse Spring Formation and the red sandstone (Table 3). All three units were either cemented early in their history (Bitter Ridge Limestone) or had minimal burial by other units (Lovell Wash Member and red sandstone), and thus we did not decompact the strata. Sedimentation rates in the upper Horse Spring Formation are used to calculate approximate ages for the base and top of the Lovell Wash Member and 10 packages within the member (Fig. 6A), assuming the interval sedimentation rates were constant.

Within the Bitter Ridge Limestone Member, the sedimentation rate at section L is 377 m/m.y. (Fig. 5A; Table 3). Because the Bitter Ridge Limestone is similar in thickness at sections B and Z for the same time period, we presume the sedimentation rate is ~377 m/m.y. at those locations as well. Given this sedimentation rate, it is possible that some of the millimeter-scale layers within the Bitter Ridge Limestone are annual varves. Similar facies in the Lovell Wash Member limestone may also be varves, but we do not have the needed resolution of age data to make this interpretation.

Within the Lovell Wash Member, we calculate three sedimentation rates for strata in section B. The sedimentation rate from the base of the member to the 13.43 ± 0.02 Ma tuff in the middle of package V is 460 m/m.y. (Fig. 7A; Table 3). The second sedimentation rate of 461 m/m.y. covers the interval from the 13.43 ± 0.02 Ma tuff to the 13.20 ± 0.04 Ma tuff in middle of package IX. The third sedimentation rate, for the upper part of the Lovell Wash Member, from the 13.20 ± 0.04 Ma to the 12.93 ± 0.02 Ma tuff in the middle of package X, drops to 248 m/m.y. To the east, at section F, the rates are higher: from the base of the member to the 13.43 Ma tuff, the rate is 667 m/m.y.; from 13.43 to 12.93 Ma, the rate is 406 m/m.y. To the southwest, within section Z, the sedimentation rate in the lower Lovell Wash Member is double that in section B, at 926 m/m.y. (Table 3; Fig. 6C).

Within the red sandstone, we calculate a rate of 204 m/m.y. that covers the lower half of the section at section B (Fig. 8; Table 3), from the 12.93 Ma ± 0.02 Ma tuff of the upper Lovell Wash Member to the base of a three-part tuff in section E dated at 12.107 ± 0.003 Ma. The next rate, for the middle of the red sandstone, jumps to 308 m/m.y. between 12.107 ± 0.003 and 11.88 ± 0.15 Ma. The youngest red sandstone tuff is dated at 11.801 ± 0.008 Ma. The two youngest red sandstone tuffs are each found at two locations, one on the far western side of the basin in sections B and Y and one in the north-central part of the basin at section RR1 (Fig. 8), allowing us to compute and compare rates from two areas. The western rate is 177 m/m.y. and the east-central one is 114 m/m.y. (Table 3; Fig. 8; see Item S8, ^{footnote 1}, for more details).

Sedimentation rates vary significantly throughout the Horse Spring Formation, and these changes coincide with changes in facies and locations across subbasins. In Table 3, we present sedimentation rates for the Thumb Member from Umhoefer et al. (2010), Anderson (2012), and Cains (2014) for comparison to the upper Horse Spring Formation rates. The Thumb Member records high and variable sedimentation rates of 750, 883, and >1000 m/m.y. Deposition rate of the Bitter Ridge Limestone Member is consistent across the basin, dropping to 377 m/m.y. The transition from the long-lived Bitter Ridge Limestone lake to the Lovell Wash depositional environments is marked by an increase in sedimentation rates as well as a variation in rates laterally, to 460 m/m.y. and 667 m/m.y. in the western and eastern parts of the basin, respectively (Table 3; Fig. 6A). This increased rate persists for much of Lovell Wash deposition and then drops during uppermost Lovell Wash deposition to 248 m/m.y. in the western half of the basin. Within the red sandstone stratigraphy on the western side of the basin, the sedimentation rate goes back up slightly to 308 m/m.y. before dropping down to the lowest rates of the entire Horse Spring Formation and red sandstone unit time: 177 m/m.y. in the west and 114 m/m.y. in the east (Table 3).

We divided the upper Horse Spring Formation within the White Basin into five structural domains based on the location and orientation of major faults and the varying styles of deformation across the basin (Figs. 3 and 9). Given the high degree of stratigraphic variation both laterally and vertically within the Lovell Wash Member and exposures, we were able to map in more detail and record more structural data within this unit than in the other two. For this reason, our five domains focus on the Lovell Wash Member with some data from the underlying Bitter Ridge Limestone Member and overlying red sandstone (Figs. 3 and 9A). Figures 3 and 9A depict the domains and first-order structures, including the Muddy Peak, Borax, and White Basin faults and an unnamed fault in the middle of the basin between domains 2 and 3. Figure 9B presents geologic cross-sections from domains 1–3. Domain 1 is defined by west-dipping strata and a series of small to medium east-west faults (Figs. 9B and 9C). Domain 2 is defined by a mix of northwest- and northeast-trending small to medium faults and a series of folded beds, sandwiched between two major faults that divide the Lovell Wash Member and red sandstone strata of the White Basin into three variably deformed domains (Figs. 9A–9C). Domain 3 consists of southwest-dipping beds repeated in a number of fault blocks created by mainly northwest-trending medium-sized faults (Figs. 9B and 9C). Domains 4 and 5 are smaller domains that contain secondary structures near the major White Basin fault. Domain 4 consists of folded, faulted, and tilted beds just west of the White Basin fault. Domain 5 consists of tilted beds and structural features that are similar to those in both domains 3 and 4.

We present the structural details and our interpretations of each domain below, focusing on Figures 9B–9D, and then discuss additional structural trends and features within the red sandstone unit (Fig. 9E).

Data. Domain 1 in the far southwestern portion of the basin contains mostly Lovell Wash Member beds, with four measurements from the Bitter Ridge Limestone Member. In the southern and central portion of domain 1, bedding uniformly strikes north-northeast and dips shallowly west (e.g., see cross-section A on Fig. 9B and stereonet for domain 1 beds on Fig. 9C). Beds are offset slightly in an apparent right-lateral strike-slip sense by east-west–striking faults. In the northern portion of domain 1, mapped faults strike more west-northwesterly and beds dip west-northwest.

There is also a series of steep north-south–striking faults within several fault blocks. Missing section across these faults suggests down-to-the-west displacement. To the south, these faults seem to tie into the northeast-striking, left-lateral oblique West Bowl of Fire fault with down to the west offset (Fig. 3). Northward, these faults cut out section higher stratigraphically, appear to have less throw, and are dying out. Although mostly inferred, the north-south fault surfaces are exposed in a few places and are almost vertical.

Structural interpretations. We interpret the north-south faults to be secondary splays off the larger, northeast-trending left-lateral oblique West Bowl of Fire fault in the southernmost part of domain 1 (Fig. 9D). Similar secondary horsetail faults have been documented nearby (e.g., Marshall et al., 2010). We interpret that the north-south faults were active prior to the east-west strike-slip faults because they are cut by the strike-slip faults and have the same amount of offset as the tuff marker beds. They do not cut any beds younger than package V.

Anderson (2012) concluded that the West Bowl of Fire fault became active during deposition of the Lovell Wash Member, and we suggest it is likely that the north-south faults were active during Lovell Wash Member deposition as well. The north-south normal faults were then tilted to subvertical during the westward tilt of bedding within the hanging wall of the major, east-dipping Muddy Peak normal fault (Bohannon, 1983a, 1983b), which forms the eastern boundary of the Muddy Mountains and western side of the White Basin (Fig. 3).

The east-west–trending strike-slip faults likely formed to accommodate differential amounts of slip from south to north on the Muddy Peak fault, which in turn created a depocenter for continued red sandstone deposition. Mapping in the red sandstone indicates maximum displacement along the Muddy Peak fault to the north of domain 1. Bohannon (1983b) noted that the strike of the beds in this area is at a high angle to the faults, and thus the east-west faults did not cause the stratal tilting. Instead, the rotation of these beds was caused by the Muddy Peak fault during tilting of the hanging wall.

In the northern part of domain 1, a detailed study of the stratigraphy across the west-northwest–striking faults suggests growth faulting: measured sections record abrupt thickening of section across individual faults. The measured sections of Figures 7A and 7B record the largest increase in section thickness across faults above the 13.43 Ma tuff, suggesting that growth faulting was active by this time (package VI). Figure 7C shows two detailed sections from package VIII that demonstrate thickening and abrupt facies changes across faults. These faults are interpreted similarly to the east-west faults in the southern part of domain 1: as secondary faults to the Muddy Peak fault accommodating progressively greater displacement to the north. This can be seen in map view (Fig. 9C), where the north-striking tuff marker beds are offset farther to the east as you move northward in domain 1.

Bohannon (1983a) mapped fanning dips within the red sandstone west of domain 1 (unit Trs in Fig. 9B) and noted this indicates that the Muddy Peak fault was active during red sandstone deposition. Our additional mapping and the increase in sedimentation rates during red sandstone deposition support this interpretation. The east-west and west-northwest faults were presumably active at this time as well.

Data. In domain 2, the Lovell Wash Member is characterized by northwest- and northeast-dipping beds and a few large, open folds (Fig. 9B, cross-section B; Fig. 9C). Fold axes are roughly parallel to the northeast-striking faults and gently plunge north-northeast (see stereonet for domain 2 folded beds on Fig. 9C). Faults within domain 2 are mostly not exposed but are inferred from offsets in marker beds. The faults typically strike both northwest and northeast. Northeast-striking faults cut out section, with younger stratigraphy on the northwest side. These may be oblique, left-lateral normal faults, similar to northeast-striking faults in other domains. The northwest-striking faults have right-lateral separation and are likely right-lateral, oblique strike-slip faults with a normal component.

The far northeastern corner of domain 2 is dominated by a dense cluster of northeast-striking faults that both repeat and cut out section. It is a structurally complicated area with an unusual sequence of beds whose facies vary along strike. Beds can be traced and correlated with the more typical sequence of beds in the rest of the domain (Fig. 6A, section C), but northward, toward and up to the boundary with domain 3, they change character dramatically, especially the limestone units.

Structural interpretations. The large faults in domain 2 exhibit an orthorhombic pattern created by northwest- and northeast-striking faults (Fig. 9D). This pattern is recognized throughout the Basin and Range (Krantz, 1989, and references therein) and is thought to be the result of three-dimensional strain producing east-west extension and north-south shortening across the region. The large folds in domain 2 may be caused by local stresses due to coeval northwest-striking, right-lateral and northeast-striking, left-lateral normal-oblique faults. Alternatively, they may be folds that formed parallel to a normal fault, as shown in Figures 9B and 9D and discussed by Schlische (1995).

We suggest that the dense cluster of faults in the northeastern portion of the domain combined with the extreme facies variations are the result of growth faulting during deposition of the upper part of the Lovell Wash Member. Figure 3 shows a triangular horst of Bitter Ridge Limestone between domains 2 and 3. The complex stratigraphy at the northeastern corner of domain 2 suggests that the faults creating this horst were active during deposition of the upper Lovell Wash Member, from package VII upwards.

Data. Domain 3 (Figs. 3 and 9B–9D) is dominated by northwest-striking, northeast-dipping faults and southwest-dipping beds. The faults divide the Lovell Wash Member into large fault blocks that contain intact stratigraphic sections, and the section is repeated multiple times (Fig. 9C, cross-section C). Some sections are folded, and fold axes are roughly parallel to the northwest faults (see stereonet for domain 3 folded beds on Fig. 9C).

Exposed faults have kinematic indicators that are either normal or right-lateral oblique normal, down to the southeast. Domain 3 faults plot with an extension direction to the west-southwest (see stereonet for domain 3 faults on Fig. 9C). In some places, northeast- and north-south–striking faults combine with the northwest-trending faults to create a rhombohedral fault pattern (Fig. 9D), similar to but even more pronounced than the one in domain 2. Beds within a few of the rhombohedral fault blocks defined by four faults dip in a direction different than the predominantly southwest-dipping beds. This may be due to increased activity on the northeast-striking faults relative to the northwest-striking faults.

Structural interpretations. A few folds seem to be simple anticlines and synclines right up against and parallel to a normal fault (Fig. 9B, cross-section C; Fig. 9D) and are interpreted as fault-propagation or drag folds. Mapping within the red sandstone to the west of this domain reveals a similar style of deformation dominated by northwest-trending normal faults. This pattern ends 5–6 km west of domain 3 against a larger normal fault that places the red sandstone unit against Cambrian strata (Fig. 3).

We concur with Bohannon (1983b) that these faults began prior to and during deposition of the red sandstone. Note that the northern part of this domain contains the basin margin and stratigraphy thins northward away from the depocenter in packages VII–IX (sections F to P on Figs. 3 and 6B). In the uppermost Lovell Wash Member, however, in package X, there are local abrupt changes in thickness across a fault just south of section T (Fig. 3), indicating the presence of a growth fault at this time. The growth faulting records active deformation within the basin during the deposition of the upper Lovell Wash Member.

Data. Domain 4 runs along and includes the White Basin fault as the eastern boundary. The fault is beautifully exposed, with the majority of kinematic indicators showing left-lateral oblique, hanging-wall down-to-the-southwest movement (Figs. 9A and 9C). The domain 4 fault data stereonet (Fig. 9C) shows an average rake of 64° for the five oblique rakes, whereas two measurements from one location show pure dip-slip motion. Upper Paleozoic strata within West Longwell Ridge make up the footwall (Fig. 3). The hanging wall contains the Bitter Ridge Limestone and Lovell Wash Members, which are deformed by secondary structures near this fault (Fig. 3). Several small faults run parallel to the main fault (Fig. 9C).

Domain 4 contains folds in two different orientations. Folds in the southern part of domain 4 trend northeast-southwest, nearly parallel to the White Basin fault (southern folds stereonet for domain 4 on Fig. 9D). In the northern half, there are two mapped folds whose fold axes are roughly perpendicular to the White Basin fault (Fig. 9C). The trend and plunge of the fold axis defined by the folded beds in the northern part of domain 4 (Fig. 9C) are also at a high angle to the main fault.

Structural interpretations. We infer that the smaller folds and faults next to the White Basin fault are secondary features related to the main fault. Some faults are parallel to the main fault, and others may be splays coming off of this fault. Similarly, the folds likely formed due to the fault. The folds perpendicular to and along the central part of the White Basin fault are consistent with folds that formed in the region of maximum displacement on the fault (Fig. 9D; Schlische, 1995). The observation that the main Bitter Ridge Limestone ridge in this domain (Fig. 3) dips northeast, toward the area with folds perpendicular to the White Basin fault, also supports the interpretation of maximum displacement in this area. Southward, folds are parallel to the main fault, similar to those in domain 2 and 3 (Fig. 9D).

Timing on the White Basin fault is based on changes in sedimentation rates. The change in the basin wide rate from 377 m/m.y. during Bitter Ridge Limestone deposition to an increased but variable rate across the White Basin suggests an increase in faulting. During Lovell Wash Member deposition, the highest sedimentation rate in the White Basin area proper, i.e., excluding the Lovell Was area, is 667 m/m.y. This rate is found in the eastern part of the basin, suggesting activity on the White Basin fault and related secondary faults during Lovell Wash Member deposition.

Data. Domain 5 contains similarities with domains 3 and 4 (Fig. 9C). Like domain 3, domain 5 contains fairly intact sequences of the lower Lovell Wash Member offset and repeated by normal faults, but in domain 5 the faults are both north-south striking and northwest striking. Beds dip predominantly southwest. Like domain 4, domain 5 includes the White Basin fault as the eastern boundary.

Stratigraphic throw diminishes northeastward along the White Basin fault from domain 4 to domain 5. In domain 5, the hanging wall of the White Basin fault contains Lovell Wash Member and upper Paleozoic strata juxtaposed against middle and upper Paleozoic strata (Fig. 3). This observation further supports the interpretation that the maximum displacement along the White Basin fault is in the central portion of domain 4.

Structural interpretations. Because the main faults in this domain strike north-south and occur where there is reduction of stratigraphic offset along the White Basin fault, we interpret these north-south faults as splay faults, secondary to and formed in conjunction with the White Basin fault (Fig. 9D). They are similar in style to faults found in the southern portion of domain 1 splaying off the end of the West Bowl of Fire fault, accommodating east-west extension.

Data. Initial mapping and detailed measured sections show that some of the same structural trends within domains 1–5 continue to the west and northwest where the red sandstone is exposed upsection from the upper Horse Spring Formation. Beds in measured red sandstone section B dip consistently to the west, similar to older beds in domain 1 (Figs. 3, 9B, and 9E). Farther westward, however, the dips get progressively shallower and then become horizontal before dipping east next to the Muddy Peak fault, forming a north-south–trending syncline (Figs. 9B and 9E; Bohannon, 1983a). Our youngest distinctive tuff, found at 85 m in the upper section B, is close to horizontal, thus deposited when local faulting was nearly finished, at ca. 11.4 Ma (Figs. 8 and 9E). The uppermost Lovell Wash Member and overlying red sandstone beds in the far northwestern corner of the White Basin wrap continuously from the Muddy Peak fault hanging wall eastward to the hanging wall of a second northwest-trending fault (beds in section E on Fig. 9E).

Red sandstone beds and faults northwest of domain 2 (between the red dashed lines on Fig. 3, east of the Borax fault) mainly trend northwest and dip southwest but are also folded. This area also contains local areas of complex fault interactions, like that in domain 2. Red sandstone beds and faults northwest of domain 3 are similar in trend and geometry with those in domain 3.

Structural interpretations. The Muddy Peak fault is a growth fault, creating the fanning dips west of domain 1 as the red sandstone was deposited on the hanging wall during fault movement. We interpret the folding as extensional folding related to normal faults as shown by the schematic block diagrams in Figure 9D. Sedimentation rates on the western side of the White Basin (Table 3; Fig. 8) support the interpretation of activity on the Muddy Peak fault during red sandstone deposition but waning toward the end of deposition.

Red sandstone beds in the central White Basin, including in measured sections at RR1, RR2, and RR3, as well as mapped tuffs just to the west of domain 2 (Fig. 3) are deformed similarly to the Lovell Wash Member in domains 2 and 3. Bohannon (1983b) noted that the northwest-trending faults cut the red sandstone unit with less offset than the underlying Horse Spring Formation unit. Thus, deformation in the central part of the basin continued past ca. 11.4 Ma but was slowing down.

In this paper, we focus on the interplay of faulting and sedimentation that created the White Basin and Lovell Wash syncline deposits during the Miocene. We also hypothesize about the contribution of faulting and tectonics versus other processes in creating these deposits. It has long been recognized that the architecture of sedimentary basin fill is a function of many processes, including tectonics, climate, base-level changes, and autogenic processes (e.g., Miall, 1990) and that these processes can be difficult to sort out in the ancient record. We present an analysis of solar insolation data from the time period of Horse Spring Formation and red sandstone unit deposition to evaluate the possible influence of climatic factors in creating the White Basin fill and present our preliminary hypothesis. We also consider autogenic sedimentation processes. Next, we reconstruct the basin evolution and discuss the evolving paleogeography, changing fault patterns, and the role tectonics played in creating the variety of facies preserved within the basin. Finally, we consider regional tectonic questions and the implications of this study.

The Miocene records a time of overall global cooling with some notable climate events, including an early Miocene glaciation and the Mid-Miocene Climatic Optimum from ca. 17 to 14.5 Ma (e.g., Zachos et al., 2001a, 2001b; Kürschner et al., 2008). In North America, Miocene climate changes have been linked to widespread changes in flora and fauna (e.g., Wolfe, 1985, 1994; Kürschner et al., 2008). To examine the possible influence of global climate variations on the highly variable stratigraphy of the Horse Spring Formation, we created a chronostratigraphic diagram to compare changing depositional environments to changes in solar insolation. Figure 10 shows the upper Horse Spring Formation and red sandstone stratigraphic facies, including the 10 stratigraphic packages of the Lovell Wash Member from Figure 6, plotted by age. The horizontal and vertical variability of White Basin lithofacies, especially the Lovell Wash Member, is shown by the colored boxes, with each color representing a different lithofacies association (Fig. 10). The uppermost Thumb Member has a time-transgressive transition to the overlying Bitter Ridge Limestone, with initial carbonate deposition occurring earlier and to the southwest of the White Basin. The estimated time of this first Bitter Ridge Limestone deposition is ca. 14.69 Ma, based on sedimentation rates. In other locations, the uppermost Thumb Member contains gypsum and other lacustrine facies. Dashed red lines in Figure 10 show times at which a lake or series of ponds was present across the basin. On the right side of Figure 10, we show variations in solar insolation due to Milankovitch cycles from 11.5 to 17 Ma, compiled from the model and software of Laskar et al. (2004).

A few patterns stand out. First, we note that solar insolation varied more frequently and with a greater range during deposition of the Lovell Wash Member when compared to the units above and below it. During deposition of the Bitter Ridge Limestone Member and the red sandstone, solar insolation was mostly below 342.3 W/m² (left of the green vertical line), whereas during Lovell Wash Member deposition, it was much more variable with many peaks as high as 342.4 and 342.5 W/m². Second, it appears that during Lovell Wash Member deposition, the times of lacustrine conditions across much of the basin coincide with periods of lower solar insolation (red dashed lines on Fig. 10), similar to the lower insolation during the persistent lacustrine facies of the Bitter Ridge Limestone Member and red sandstone unit. Although the stratigraphy of the lower Thumb Member is not shown, it contains a wide range of highly variable units (e.g., Lamb et al., 2010) and, like the Lovell Wash Member, has both a higher number of swings in solar insolation and a greater range of values of those swings than the Bitter Ridge Limestone Member.

Holbourn et al. (2007) examined global climate changes in detail from 17.1 to 12.7 Ma using two continuous deep-sea cores. They designated three middle Miocene climate phases (Fig. 10): a phase I climate optimum prior to 14.7 Ma, phase 2 cooling from 14.7 to 13.9 Ma, and the phase 3 “icehouse” period after 13.9 Ma. They attributed these distinct climate phases in part to the 400 k.y. and 100 k.y. eccentricity cycles as well as changes in obliquity. Deposition of the uppermost lacustrine Thumb Member and the wholly lacustrine Bitter Ridge Limestone falls exclusively within their phase 2 event from ca. 14.7 to 13.9, a time of global cooling. The end of the Bitter Ridge Limestone deposition coincides with the end of their phase 2, which is recorded in their cores by an abrupt change in δ¹⁸O at 13.87–13.84 Ma. We hypothesize that these correlations of the change in depositional environments in the Lake Mead area with global climatic changes indicates that climatic processes might have influenced the depositional environments in the White Basin and played a role in the variety of facies that was deposited. The consistent lower solar insolation and phase 2 cooler climate may have contributed to the long-lasting stability of the Bitter Ridge Limestone lake by reducing evaporation and/or producing more precipitation in the Lake Mead region. The even more variable solar radiation during deposition of the Lovell Wash Member may have triggered facies changes from one package to another and contributed to the great lateral and vertical variability of this unit. These preliminary hypotheses on the role of climate require additional work and testing.

Autogenic processes, i.e., those that happen completely within a system and without external forcing, can play a role in creating cyclic stratigraphy that alternates between coarse- and fine-grained deposition or between fluvial and lacustrine deposition (Straub et al., 2020, and references therein; Hajek and Straub, 2017). Because the Lovell Wash Member records fluctuations between fluvial and lacustrine systems, we consider the role in which these processes contributed to the stratigraphic variability of this member. Much of the Lovell Wash Member is dominated by lacustrine sedimentation, with carbonate, evaporite, and marl (mixed carbonate and siliciclastic muds) deposition. Some packages contain more sandstone and mudstone. Intermittent sandstone beds, found in almost all packages, are generally tabular and express channel dimensions on the order of 1 m thickness. These features of the basin imply a low-relief basin. Within such a system, there are several possible sources of autogenic forcings within and between strata, all of which may act simultaneously:

• Carbonate facies: Small changes in lake level driven by climate changes induce flooding. The carbonate factory, in this case strongly influenced by microbial processes, can essentially keep up with base-level change with very short-time-scale aggradation rates. As the lake expands aerially and there is a rapid aggradation of microbialite or micritic limestones, the base of this limestone package may be characterized by a basal and very subtle erosion or subaerial exposure surface.

• Siliciclastic facies: Siliciclastic mudstone units with occasional sandstone beds commonly occur as stream channels migrate, fill in low areas, and act to subdue topography. We envision streams moving back and forth on small alluvial fans and entering a playa or lake, shifting the locus of fine-grained and coarse-grained beds throughout the basin.

• Interbedding of carbonate, evaporite, marl, and siliciclastic mudstones, in no particular stratigraphic order: It is not uncommon for carbonates, siliciclastic mudstones, marls, and evaporites to be interbedded over sub-meter-scale stratigraphic intervals, suggesting short-time-scale fluctuations in siliciclastic sediment delivery to the basin. During times of relatively low siliciclastic mud delivery, lake waters are less turbid and the carbonate factory dominates. As mud input increases turbidity, carbonate deposition shuts off and mudstone deposition is activated. Intermediate clastic input may be responsible for the production of marls. This suggests a possible autocyclic shifting of very shallow lacustrine depocenters within a sheetflood-dominated, low-relief (braided?) fluvial system.

As a result of these processes, we would anticipate that, unlike in more robust fluvial systems with deep channels and barforms, the “maximum morphodynamic vertical roughness scale” (Straub et al., 2020), or relief, within the basin itself would be quite low, i.e., ~1 m. Autogenic strata would be characterized by sub-meter-to meter-scale bedding. Sheets et al. (2002) and Straub et al. (2020) determined the “compensation time scale”, T_c, which is the maximum time scale of autogenic organization in a given stratigraphy (Straub et al., 2020), by dividing the vertical roughness scale by the sedimentation rate. We estimate, using the scale of small channels in the Lovell Wash Member (1 m) and our sedimentation rates of ~250–650 m/m.y., compensation time scales between 2000 and 4000 yr. This is in alignment with the results of Straub et al. (2020), who noted that extensional basins commonly have small landscape roughness scales and lower sedimentation rates than foreland basins and passive margins and, thus, have a T_c on the order of 5000 yr of less. Our 10 Lovell Wash Member stratigraphic packages range from 30,000 to 270,000 yr in duration, with most lasting 60,000–100,000 yr. Thus, the time scale at which autocyclic processes affect stratigraphic variability is predicted to be much smaller than the cyclicity we observe between packages and instead may contribute to smaller variations within packages. Packages II, III, IX, and X all contain fine-grained mudstones and marls, alternating with numerous 0.5–1-m-thick limestone and sandstone beds, and these variations in lithologies are likely influenced by autogenic processes. The 10 stratal packages representing substantially longer time scales than the time frame of variations within a single package are essentially not influenced by autogenic processes and instead are likely the result of tectonic and/or long-term climatic forcings.

Finally, we recognize that nearby and distant volcanic eruptions resulted in fallout of tephra within the basin. During deposition of the Bitter Ridge Limestone, these volcanic deposits did not stop microbial deposition. During Lovell Wash Member deposition, however, the influx of material into smaller lakes and local ponds might have buried microbial mats and temporarily halted the carbonate factory, contributing to the stratigraphic variability present in the Lovell Wash Member.

The formation of the Bitter Ridge Limestone in a single, large, stable lake that persisted for more than half a million years stands in stark contrast to the underlying and overlying highly variable Thumb and Lovell Wash Members. More than 200 m of vertically homogenous Bitter Ridge Limestone records uniform conditions laterally across a ~10 × 25 km shallow lake likely only 1–3 m deep (Figs. 5 and 11A; Hickson et al., 2010). Sedimentation rates dropped from Thumb Member rates of at least 750 m/m.y. to 377 m/m.y. across the basin (Table 3; Fig. 12). These observations imply fairly constant subsidence with microbially influenced limestone formation keeping pace with the steady creation of accommodation space. Anderson (2012), Samra (2013), and Cains (2014) documented a time-transgressive, gradual contact between the fining-upward clastic fluvial-to-playa sequence of the Thumb Member and the lacustrine Bitter Ridge Limestone and concluded that deposition of the lake began in the southern Gale Hills area and spread eastward and northward. Figure 11A is a schematic interpretation of the paleogeography during deposition of the Bitter Ridge Limestone from 14.5 to 13.85 Ma. We conclude that the transition from a fining-upward and clastic lake sequence in the upper Thumb Member into the persistent lacustrine carbonates of the Bitter Ridge Limestone is the result of partitioning of the larger, older Horse Spring Formation (Thumb Member) basin into a series of subbasins by a significant increase in faults within the former basin (Fig. 11A). These faults, namely the Las Vegas Valley shear zone and associated secondary faults on the southwest and west, the western Bitter Spring Valley fault on the southeast, and the White Basin fault or another, more eastern fault on the northeast (Figs. 2 and 11A) provided the primary control on the uniform deposition in the Bitter Ridge Limestone lake. We suggest that most of the lake margins were relatively low-relief areas along these young faults. The Muddy Mountains on the northern edge of the lake may have been higher as they were eroded from the long-lived Sevier thrust belt.

Samra (2013) and Anderson (2012) suggested, following Duebendorfer and Simpson (1994), that primarily down-to-the-north activity on the Las Vegas Valley shear zone created the southwestern boundary of the Bitter Ridge Limestone lake basin. They documented activity on the Las Vegas Valley shear zone in the Gale Hills before 14.5 Ma, possibly ca. 15.0–14.7 Ma. Conglomerates shed off the footwall on the south side of the Las Vegas Valley shear zone interfinger with lacustrine facies and are evidence for the southern edge of the Bitter Ridge Limestone lake (Anderson, 2012). On the north side of the Las Vegas Valley shear zone, Anderson (2012) and Samra (2013) mapped a north-south normal fault that formed the western boundary of the Bitter Ridge Limestone lake (Fig. 11A). Samra (2013) suggested that during Thumb Member deposition, this and other active normal faults in the Gale Hills area (Figs. 2 and 11A) created horsts and grabens that disrupted the eastward progradation of coarse clastics from the Muddy and/or Spring Mountains, formed a western boundary to the Bitter Ridge Limestone lake basin, and allowed microbialite deposition to persist.

The southeastern boundary of the Bitter Ridge Limestone lake was likely the western Bitter Spring Valley fault. San Filippo (2008) suggested that the western Bitter Spring Valley fault was first active after 14.8 Ma and may have actually been the eastern continuation of the Las Vegas Valley shear zone before a fault reorganization resulted in its main history as a left-lateral fault after ca. 13.5 Ma.

The northern and northwestern boundaries of the lake are now buried beneath younger Miocene units in the White Basin. The Bitter Ridge Limestone was likely deposited along a buttress unconformity against the Muddy Mountains; a fault-sliver remnant of Bitter Ridge Limestone deposited high on the eastern side of the Muddy Mountains (pink star on Fig. 3) supports this idea.

The eastern boundary of the Bitter Ridge Limestone lake is enigmatic. Lamb et al. (2010) introduced two possibilities for the eastern boundary of the Bitter Ridge Limestone basin: (1) the White Basin fault formed an eastern boundary, or (2) the boundary was even farther to the east and the Bitter Ridge Limestone microbialite facies transitioned laterally to the deeper-water, fine-grained clastics and gypsum sequence found to the east in Echo Wash (Fig. 2). We note that a similar transition, from basin-margin microbialite limestone to basinward, deeper-water gypsum, is found in other nearby locations during other time periods of Horse Spring Formation deposition (Lamb et al., 2010). In the first scenario, a fault reorganization at 14.5 Ma formed the White Basin fault while the Las Vegas Valley shear zone was propagating into the area and forming the southern edge of the lake (Umhoefer et al., 2010). In the second scenario, west-dipping normal faults east of or under the Overton Arm of Lake Mead may have been the eastern boundary of the basin (Figs. 1 and 11A). A third possibility is that as the Lime Ridge fault east of the Overton Arm (Fig. 1) was dying out and displacement on the South Virgin–White Hills detachment fault was slowing down, gravels prograded from east to west and created a distant eastern boundary. At this time, we favor the second scenario. One possible candidate for an eastern basin-bounding fault is an unnamed hanging-wall down-to-the-west normal fault shown on Figure 1 (red star) east of the Overton Arm of Lake Mead. We note that if the offset on the younger Hamblin Bay–Bitter Ridge fault is restored, the western Bitter Spring Valley and Lime Ridge faults (shown by yellow stars on Fig. 1) roughly align, placing the fault noted by the red star in an ideal location to be the eastern boundary. Increased sedimentation rates in strata near the White Basin fault favor the initiation of this fault during lower Lovell Wash Member deposition. As noted earlier, the gypsum belt east of the Longwell Ridges is likely to correlate with Bitter Ridge Limestone and represent a deeper part of the lake, although the gypsum is not directly dated.

Faults created the primary control on the formation and steady subsidence and uniform fill of the Bitter Ridge Limestone lake, but we acknowledge that global climate changes may have played a secondary role. Global solar insolation fluctuations were less frequent and insolation was lower relative to the member above and below the Bitter Ridge Limestone Member overall during this time, as discussed above, and these factors may have contributed to the overall striking homogeneity of this member. It is also possible that increased precipitation and runoff contributed to the stability of this large and long-lived lake during the phase 2 cooling of Holbourn et al. (2007). A more subtle solar insolation forcing signal shows up within stable-isotope data collected every 1.5–2 m in the 200+-m-thick section: Hickson et al. (2010) detected a possible 100 k.y. cycle, implying a possible influence from the eccentricity variation.

The transition from the widespread Bitter Ridge Limestone to the Lovell Wash Member is stratigraphically fairly abrupt except near the basin margin at section A (Fig. 6A). Given that the Bitter Ridge Limestone lake was fairly shallow, small fluctuations in lake depth would have caused cycles of lake retreat and inundation along the edges, thus creating the back-and-forth fluvial to lacustrine beds at section A. Package I of the Lovell Wash Member (Fig. 6) records basinwide deposition of red-weathering fine-grained sandstone and mudstone in fluvial environments that persisted for ~60,000 yr. We suggest two factors contributed to the extinction of the Bitter Ridge Limestone lake and transition to fluvial conditions of package I.

First, global and regional climate changes may have played a role in the demise of the Bitter Ridge Limestone lake. The timing of the transition from the Bitter Ridge Limestone lake to the overall drier conditions of the Lovell Wash Member coincides with other global and regional changes that are summarized on the right side of Figure 12. Chapin (2008) summarized plate movements that led to major ocean-current reorganizations across the globe and linked this to the transition out of the global Mid-Miocene Climatic Optimum across the U.S. southwest. Hickson et al. (2010) pointed out that the Zachos global marine oxygen-isotope curve (Zachos et al., 2001a) records a 0.5‰ increase at 13.85 Ma and noted variations in stable-isotope values from Bitter Ridge Limestone samples to lower Lovell Wash Member limestone facies that coincide with this timing. In addition to global changes, the reorganization of ocean currents in the Pacific may have produced regional climatic changes. Jacobs et al. (2004) and Töpel et al. (2012) linked widespread flora and fauna changes in the western U.S. to the development of the Pacific Ocean upwelling and summer drought that began ca. 15–12 Ma. This aridification may have played a role in the shift from the 640,000 years of the persistent Bitter Ridge Limestone lake to the drier conditions of package I and the mixed facies of other packages within the Lovell Wash Member, but additional work is needed to test this hypothesis.

The second factor that likely contributed to transition from the Bitter Ridge Limestone to the Lovell Wash Member is structural, i.e., an increase in activity on new faults and changes in the rate of faulting on established faults. As shown on Figure 12, sedimentation rates increased across the region from the time of deposition of the Bitter Ridge Limestone to packages I–IV of the Lovell Wash Member. The increase in sedimentation rates on the eastern side of the White Basin (Fig. 12) points to initiation of or increased activity on the White Basin fault or a more eastern fault (Fig. 11A). In the Lovell Wash syncline area, the sedimentation rate more than doubles at this time. Anderson (2012) presented evidence that this is due to increased movement on the Las Vegas Valley shear zone. Relative uplift on the south side of this fault changed the basin configuration. These paleogeographic changes could have included one or more of the following: (1) opening of an outlet and draining the lake; (2) increasing the regional slope and, in turn, allowing for more energetic streams; (3) exposing new clastic source areas; and/or (4) creating pathways for increased clastic input which swamped the lake. The inception of or increased movement on the western Bitter Spring Valley fault (Fig. 2) may also have occurred at about the same time as the transition from the Bitter Ridge Limestone to Lovell Wash Members (San Filippo, 2008). This is another mechanism of disruption to the southern edge of the basin.

From Lovell Wash Member package I upwards, there are two major changes in sedimentation. First, the remaining packages II–X record highly variable lateral facies changes across the White Basin instead of the relatively homogenous depositional conditions of the Bitter Ridge Limestone and package I. Initiation of or increased movement on the White Basin fault or a more eastern fault produced an asymmetric basin with a depocenter in the eastern White Basin (Figs. 11B and 12). In packages II–IV (Fig. 6A), section F is thicker than section B, and when lacustrine environments are present across the entire basin, the eastern facies are deeper-water facies. Both of these observations support an active fault on the eastern margin of the basin. The changes in sedimentation rate across the basin (Fig. 12) further support the idea of a more asymmetric basin compared to the Bitter Ridge Limestone lake basin.

A second change is the vertical alternation between fluvial and lacustrine environments during Lovell Wash Member time. Overall, the sedimentology suggests drier conditions, with more spatiotemporal variation in both perennial and ephemeral water bodies (e.g., Figs. 11B and 11C) than the underlying Bitter Ridge Limestone Member. Many of the limestone deposits are thin and discontinuous, suggesting small local ponds. As noted above, this is a time period of more frequent and wider swings in the eccentricity of the earth's orbit than during deposition of the Bitter Ridge Limestone Member, and there may be a correlation between solar insolation lows and the short-lived lakes represented in the Lovell Wash Member (Fig. 10). These solar shifts may have been enough to impact climate and change local rainfall and/or runoff conditions, leading to shifts between fluvial and shallow, short-lived lake or pond environments.

Another contributing factor to the stratigraphic complexity is a change in faulting style: a transition from a few major faults to more numerous, medium-sized faults that begin to break up the basin. The earliest evidence of this is found in section Z in package IV (Fig. 6C; Anderson, 2012). At this time, two coarse-grained debris flows deposited cobbles and boulders, including several large (1–5 m long-axis dimension) rafted blocks of Bitter Ridge Limestone. This clearly required a local uplift with a steep gradient. This local uplift in the vicinity of section Z is also likely responsible for the input of coarse-grained sandstones to the White Basin at the same time, during deposition of package IV (Figs. 10 and 11B). Anderson (2012) documented additional structural and stratigraphic changes in the southern part of the lower Lovell Wash Member basin near the Las Vegas Valley shear zone that occurred at this time that support this idea of basin breakup.

Deposition of the upper Lovell Wash Member is marked by slowing sedimentation rates, periods of unusual and variable microbialite deposition, and the development of the Muddy Peak fault and localized growth faults (Figs. 6, 7, 11C, and 12). Figure 12 shows the reduced sedimentation rates at sections B and F as well as the activity on growth faults in domains 1, 2, and 3. In Figure 7, we show the presence of numerous small east-west–and ESE-WNW–striking growth faults on the western side of the basin in domain 1 that were active during deposition of the upper Lovell Wash Member. We interpret that these are secondary faults that formed in response to differential offset along the Muddy Peak fault, which became active at this time. Growth faults in domains 2 and 3 may be controlled by activity on the northern part of the Borax fault (Fig. 3) of Bohannon (1983a, 1983b). We suggest that the change in sedimentation rates during and after deposition of package VI is a part of the shift from activity on a few major faults to activity on many, distributed faults as the White Basin broke apart (Fig. 11C). Anderson (2012) documented that the southern Lovell Wash Member basin in the area of section Z was deforming and inverting during deposition of packages V–VII and has no preserved sedimentary record after package VII.

Further evidence for the structural disruption of the basin is the presence of a wide variety of microbialite facies found in two distinct time periods across the White Basin, during deposition of package VIII and at 13.19 Ma during deposition of package IX. The morphology of microbialite bioherms, with a range of synoptic reliefs (a primary indicator of water depth), point to variations in water depth and, possibly, chemistry (Hickson et al., 2020; Frisk and Hickson, 2020). We envision the presence of small, ephemeral ponds and some more-widespread lakes, some of which were fed by hot springs along active faults (e.g., Hickson et al., 2016). Packages IX and X record a mix of thin limestone beds alternating with marls, mudstones, and sandstones found laterally across the basin. Lacustrine deposits dominate the eastern side of the White Basin and clastics the west (Fig. 6A), but in general, the eastern and western sides are more similar to each other at this time than they are in the previous seven packages.

The Lovell Wash Member transitions abruptly to the widespread gypsum deposit of the lower part of the red sandstone unit within the central White Basin (recorded at section B on Fig. 8), with siliciclastic and limestone units found near the basin margins. This transition back to fairly uniform lacustrine and highly evaporative deposition across the basin may mark a major shift in faulting: faulting on the eastern side of the basin may have slowed as the Muddy Peak fault on the west became more active (Fig. 12). Sedimentation rates that increased during deposition of the red sandstone on the western side of the basin further support the interpretation that the Muddy Peak fault increased in activity before starting to wane during latest red sandstone deposition. The other factor that may have contributed to the return of more uniform and dominantly evaporative lacustrine deposition is the switch back to a relatively long period of low solar insolation during deposition of the lower red sandstone unit (Figs. 10 and 12) as well as the mostly dry conditions in the southwest created by the El Niño–Southern Oscillation system combined with overall global cooling (Chapin, 2008).

Toward the end of red sandstone unit deposition, fanning dips in the western part of the basin as well as a drop in sedimentation rates suggest that faulting along the Muddy Peak fault was slowing down. Some fault activity in the central White Basin, however, continued after deposition of the youngest red sandstone beds, as noted by Bohannon (1983a, 1983b), with faults that cross-cut the youngest beds. This deformation is primarily controlled by the northwest-trending faults in the central part of the basin.

The final set of questions we set out to address revolves around tectonic models for the larger region and the style and timing of the breakup of the upper plate. Our work supports that of Umhoefer et al. (2010), who suggested that extension proceeded in discrete steps with varying styles and rates. Figure 12 summarizes our work that demonstrates faulting varied in location, intensity, and style from ca. 14.5 to 11.5 Ma. During Bitter Ridge Limestone deposition, faulting occurred on a few key faults and sedimentation rates suggest slower faulting rates than the previous Thumb Member records. During Lovell Wash Member and red sandstone time, we see a shift from the northeast-trending transtensional White Basin fault to the northwest-trending normal Muddy Peak fault as well as the development of several smaller northwest- and northeast-trending transtensional faults. This corroborates the work of Duebendorfer and Simpson (1994), who argued that the faulting proceeded in steps and presented a similar relative chronology for faulting just north of the Las Vegas Valley shear zone: northeast-trending faults developed first, followed by northwest-trending faults associated with the Las Vegas Valley shear zone. We also document the waning activity of extensional deformation in the White Basin area.

Miocene stratigraphy of the White Basin records the second half of central Basin and Range extension north of Lake Mead. Structural and stratigraphic data record the evolution of faulting to produce a changing landscape. We suggest that the large changes in basin paleogeography and sedimentation are due to tectonic factors. The timing and location of major faults changed the basin configuration as well as sedimentation rates through time. The time period from 14.5 to 13.86 Ma was marked by the shallow, microbialitic lake of the Bitter Ridge Limestone Member of the Horse Spring Formation. The persistent lake was broken up structurally from a reorganization of faulting and followed by deposition of the highly variable Lovell Wash Member from 13.86 to 12.7 Ma in a landscape dominated by streams, ponds, and volcanic fallout. The White Basin fault was most active during Lovell Wash Member time, while the Muddy Peak fault was most active during deposition of the red sandstone unit. During the time of Lovell Wash Member deposition, there was also a shift to more numerous but smaller faults, which produced numerous small, short-lived ponds with carbonate and evaporitic facies. The red sandstone unit, 12.7–11.4 Ma, records a widespread, evaporitic lake rimmed with alluvial fans as well as periods dominated by fluvial deposition. By the end of red sandstone deposition at 11.4 Ma, faulting was waning.

We also suggest that climate may have played a secondary role in creating the stratigraphic complexity of the upper Horse Spring Formation and red sandstone. Global fluctuations in solar insolation as well as regional climatic changes due to changes in oceanic circulation coincide with documented changing environmental conditions in the White Basin. Deposition of the upper Horse Spring Formation began with the transition from the Mid-Miocene Climatic Optimum back to global cooling and also occurred at the same time as aridification in the western U.S., including the development of summer drought and wet winter conditions (e.g., Jacobs et al., 2004). We hypothesize that these regional and global changes may have contributed to the local facies heterogeneity throughout the Lovell Wash Member and red sandstone unit. Ongoing work will explore and test this hypothesis.

Ernie Anderson recognized the value of the Lake Mead region in the 1960s and spent many weekends using his days off to do the fieldwork that led to his seminal work. He championed the importance of this region, freely admitted when his own work was “just the beginning” in complex areas, and encouraged our research group to delve deeper. We are grateful for his encouragement, scientific suggestions, help in the field, and never-ending enthusiasm for the joys of geological exploration. We also give thanks to L. Sue Beard and many undergraduate University of St. Thomas and graduate Northern Arizona University students who participated in field research that contributed to this project. Tim Lawton and two anonymous reviewers greatly improved the manuscript, and we are so appreciative of their time, expertise, and careful, detailed reviews! Funding for much of this work was provided by U.S. National Science Foundation grants EAR-0838340 (Lamb and Hickson) and EAR-0838596 (Umhoefer). It was also supported by the Petroleum Research Fund grant 38010-GB2 to Dr. Lamb and EDMAP grants from the U.S. Geological Survey National Cooperative Geologic Mapping Program to Drs. Lamb and Hickson. Finally, we note the recent passing of our dear friend, colleague, and co-author, Dr. Paul Umhoefer, and give thanks to all that he contributed to this project, this paper, and our lives.