Journal of Sedimentary Research, 2006, v. 76, 1197–1214 Research Article DOI: 10.2110/jsr.2006.096 HIGH-RESOLUTION STRATIGRAPHY OF AN UNDERFILLED LAKE BASIN: WILKINS PEAK MEMBER, EOCENE GREEN RIVER FORMATION, WYOMING, U.S.A. JEFFREY T. PIETRAS* AND ALAN R. CARROLL 1University of Wisconsin-Madison, Department of Geology and Geophysics, 1215 West Dayton Street, Madison, Wisconsin 53706, U.S.A. e-mail: [email protected] ABSTRACT: Lakes tend to respond noticeably to minor changes in sediment and water balance driven by climatic, tectonic, or geomorphic processes. This unique behavior of lacustrine basins can provide a high-resolution record of geologic processes within the continental setting, far from the globally averaged record of marine strata. The Wilkins Peak Member of the Eocene Green River Formation, in Wyoming, USA, is dominated by aggradation of repetitive sedimentary facies successions recording distinct lacustrine expansions and contractions. These lacustrine ‘‘cycles’’ consist of up to four successive facies associations: littoral, profundal–sublittoral, palustrine, and salt pan. Because they comprise disparate facies that may never have been simultaneously deposited in the basin, Wilkins Peak Member cycles are non-Waltherian successions that do not readily equate to any established sequence stratigraphic unit. The completeness of the stratigraphic record in the Wilkins Peak Member varies continuously across the basin. At least 126 cycles are present in the ERDA White Mountain #1 core near the basin depocenter, whereas only about one third as many are recognizable 53 km north, nearer to the basin margin. Cycle boundaries terminate northward by gradual amalgamation into palustrine facies, reflecting the interplay between varying lake levels and a south-dipping deposition gradient. Evidence for complete desiccation and hiatuses is commonplace even near the basin center; a truly complete record may not be present anywhere. Based on recent 40Ar/39Ar geochronology of tuffs, the average cycle duration in the White Mountain #1 core is approximately 10,000 years. However, the true average duration is shorter due to the presence of lacunae, and the time required for deposition of the thinnest cycles may have been less than 1000 years. No external driving mechanism is presently known for Eocene cycles of such short duration, raising the possibility that they are instead autogenic cycles related to geomorphic instability of the surrounding landscape. Evaporite deposition corresponded closely in time and space to maximum differential subsidence, suggesting that tectonic influences on drainage patterns and basin accommodation exerted a primary control. Differential accommodation was most pronounced during deposition of the lower Wilkins Peak Member and became progressively less so after that time. Based on structural and stratigraphic observations, a period of increased tectonic uplift commenced concurrent with deposition of the lower Wilkins Peak Member. INTRODUCTION environments may occur repetitively within relatively short vertical successions (e.g., Gilbert 1882; Glenn and Kelts 1991; Talbot and Allen Much of our knowledge of continental climate and of terrestrial faunal 1996). This is especially true in underfilled basins, where stratal stacking and floral evolution comes from records preserved in lacustrine and patterns record repeated extreme fluctuations in base level (Carroll and related alluvial strata (e.g., Wolf 1978; McCune 1987; Stiassny and Meyer Bohacs 1999). 1999). Furthermore, because most long-lived lacustrine basins are A second problem common to many lacustrine deposits is the absence fundamentally tectonic in origin (Kelts 1988), they potentially provide of cosmopolitan marine fauna, which hinders biostratigraphic correlation highly resolved archives of orogenic processes and related landscape of these strata to the geologic timescale. Milankovitch-period deposi- modification (e.g., Yuretich 1989; Lambiase 1990). However, the tional cyclicity has therefore often been proposed as an alternative means sensitivity of lakes to paleoenvironmental change can also complicate of establishing high-resolution chronostratigraphy, on the basis that their stratigraphy, because of the potential rapidity of these changes orbitally controlled changes in solar insolation directly govern sedimen- relative to the time required for sediment deposition (e.g., Owen et al. tation rates and depositional environments in lakes (e.g., Bradley 1929; 1990; Oviatt et al. 1992; Scholz and Finney 1994; Benson et al. 1995). As Van Houten 1962; Olsen 1986; Fischer and Roberts 1991; Roehler 1993; a result, sedimentary facies representing vastly different depositional Fischer et al. 2004). However, the premise of externally regulated cyclicity has seldom been tested against other independent measures of elapsed or * Present address: BP Sakhalin, Inc., 501 Westlake Park Boulevard, Houston, absolute geologic time (Steenbrink et al. 1999; Pietras et al. 2003a). Texas 77079, U.S.A. Extreme Holocene lake-level changes are known to have occurred over Copyright E 2006, SEPM (Society for Sedimentary Geology) 1527-1404/06/076-1197/$03.00 1198 J.T. PIETRAS AND A.R. CARROLL JSR FIG. 1.— Generalized geologic map of the greater green River Basin and the surrounding highlands showing the location of measured sections and selected Eocene faults (Witkind and Grose 1972; Love 1970; Steidtmann et al. 1983; Bradley 1995). Maximum extent of the Tipton Shale, Wilkins Peak, and Laney members is from Roehler (1993). Extent of evaporites is from Wiig et al. (1995). sub-precessional time scales for reasons unrelated to orbital forcing (e.g., construct the first detailed (centimeter- to decimeter-scale) regional cross Malde 1960; Johnson et al. 1996; Bouchard et al. 1998; Colman 1998; section of the entire Wilkins Peak Member. This cross section extends Reheis et al. 2002). Additionally, many previous proposals of Milanko- from near the basin center in the south to the basin margin in the north vitch cyclicity in lake deposits have been based either on single measured (Fig. 1). The unique basin-scale perspective it provides allows reinterpre- sections or wells (e.g., Olsen 1986; Fischer and Roberts 1991; Roehler tation of the controls on lacustrine evaporite deposition, of the nature of 1993) or else on one-dimensional composites of overlapping but the repetitive facies associations that dominate Wilkins Peak Member geographically separated intervals (e.g., Steenbrink et al. 1999; Olsen stratigraphy, and of the temporal significance of these ‘‘cycles.’’ and Kent 1996). It is therefore largely unknown to what extent geographic variation in sedimentary facies or in cycle preservation might REGIONAL SETTING AND STRATIGRAPHY affect the outcome of such studies. The Wilkins Peak Member of the Green River Formation of The greater Green River Basin (gGRB) formed between the Sevier fold southwestern Wyoming, U.S.A., is ideally suited for addressing these and thrust belt to the west, and several Laramide-style Precambrian-cored problems, due to its excellent exposures and to recent advances in its uplifts on the north, east, and south (Fig. 1; Dickinson 1979; Lawton 40Ar/39Ar geochronology (Smith et al. 2003). The general tectonic and 1985; Dickinson et al. 1988; Yuretich 1989). Eocene deposition of the climatic setting of the Green River Formation has been well characterized lacustrine Green River Formation in Lake Gosiute roughly coincided (e.g., Love 1970; Wolf 1978; Gries 1983; Steidtmann et al. 1983; Roehler with late pulses of shortening in the both the Sevier belt and Laramide 1991; Roehler 1993; Markwick 1994; Wilf 2000). Furthermore, previous uplifts (Love 1970; Steidtmann et al. 1983; Dickinson et al. 1988; DeCelles studies have established a sound fundamental understanding of the 1994). The occurrence of palms, ferns, gingers, and crocodilians indicate depositional environments recorded by these lacustrine strata (Culbertson a subtropical, humid climate with muted, equable seasonality and 1961; Eugster and Hardie 1975; Smoot 1983; Roehler 1992). In this study predominantly frost-free winters (MacGinitie 1969; Markwick 1994). we take advantage of the nearly continuous exposures along White Leaf-margin and leaf-area analyses yield interpreted mean annual Mountain, a cuesta on the western flank of the Rocks Springs Uplift, to temperatures (MAT) of approximately 15u to 23uC and mean annual JSR WILKINS PEAK MEMBER CYCLICITY AND BASIN EVOLUTION 1199 FIG. 2.—North–south stratigraphic cross section of the Green River Formation. Note the unique southward thickening of the Wilkins Peak Member compared to the Tipton Shale and Laney members, and the crosscutting relations between basin bounding faults and the lower Wilkins Peak Member. Modified from Roehler (1989). precipitation (MAP) of 75 to 110 cm/year during deposition of the Green LITHOFACIES ASSOCIATIONS River Formation (Wing and Greenwood 1993; Wilf et al. 1998; Wilf The ten lithofacies below are defined on the basis of lithology, texture, 2000). Eocene climate models for southwestern Wyoming estimate similar and sedimentary and biogenic structures, and are grouped into 5 MAT and MAP values (Matthews and Perlmutter 1994; Morrill et al. lithofacies associations. This classification is broadly consistent with 2001). previous descriptions and depositional-environment interpretations (e.g., The Wilkins Peak Member consists of carbonate-rich lacustrine and Eugster and Hardie 1975; Smoot 1983). However, significant advances