Chemostratigraphic and lithostratigraphic framework of the Eocene Kreyenhagen Formation, Kettleman area, central San Joaquin Basin, California

In mudstone deposits, it is critically important to understand vertical and lateral variations in organic matter, biosiliceous material, and detrital content. Relative enrichment in these components records the interactions among benthic redox conditions, primary productivity, and clastic sedimentation rate (Bohacs et al., 2005). Relative sea level and climate exert fundamental controls on these oceanographic factors and therefore mudstone composition (Arthur and Sageman, 2005f). For example, rising sea level (1) enhances water-column stratification, contributing to bottom water anoxia, and (2) shifts the locus of sedimentation, starving distal environments of diluting terrigenous sediment, together forming organic-rich condensed sections (Creaney and Passey, 1993). Earth’s climate affects pole-to-equator temperature gradients, which control wind strength and coastal upwelling intensity and thus nutrient availability for photosynthetic algae. High rates of primary productivity may result in an increased settling flux of siliceous material and organic matter to the seafloor, driving biogenic silica accumulation and organic matter preservation.

Geochemical proxies have proven to be useful for understanding compositional variations in mudstones in the context of relative sea level (Algeo et al., 2007; Lash and Blood, 2014; Dong et al., 2018) and paleoceanography (Riboulleau et al., 2003; Rimmer et al., 2004; Algeo and Maynard, 2004; Tribovillard et al., 2004, 2006; Ver Straeten et al., 2011; Hancock et al., 2019). Major-element, trace-element, and total organic carbon (TOC) analyses are routinely used for inferring paleoredox conditions, paleoproductivity, and clastic sedimentary components. Integration of these geochemical proxies into a lithostratigraphic framework, often constructed using well logs, provides a way to infer relationships between benthic depositional conditions and composition (Abouelresh and Slatt, 2012; Hemmesch et al., 2014; Kohl et al., 2014; Dong et al., 2018).

This study examined the depositional settings that affected compositional variability in the Kreyenhagen Formation, a middle Eocene (48.6–37 Ma), organic-rich, siliceous and calcareous mudstone of the San Joaquin Basin, California, USA (Hosford Scheirer and Magoon, 2007). The thick, organic-rich unit is a known source rock for several oil accumulations, second only to the Antelope Shale (upper Monterey Formation equivalent) with respect to the amount of oil generated in the basin (Peters et al., 2007). The Kreyenhagen Formation records a major relative sea-level highstand that submerged a broad area of the California margin (Milam, 1984, 1985; Bartow, 1991; Bloch, 1991a, 1991b), which was a fragmented forearc depocenter at the time. Additionally, its deposition coincided with a long-term global cooling trend during the middle Eocene that intensified coastal upwelling on continental margins, as recorded in globally distributed diatom-bearing sections (Barron et al., 2015). Despite its significance to both industry and academia, the Kreyenhagen Formation is remarkably understudied in comparison to the Miocene Monterey Formation and Cretaceous Moreno Formation of California. About 10 studies of the Kreyenhagen Formation were published before 1990, which focused largely on outcrop descriptions and biostratigraphy, with the exception of a regional study by Milam (1985). More recently, U.S. Geological Survey Professional Paper 1713 (Hosford Scheirer, 2007) and Larue et al. (2018) addressed the source and reservoir rock properties of the Kreyenhagen Formation as parts of more regional stratigraphic studies. However, there remains no published high-resolution subsurface stratigraphic study in which the primary focus was the Kreyenhagen Formation.

We integrated geochemical proxies for paleoredox, paleoproductivity, and detrital input conditions into a well log–based lithostratigraphic framework to examine how relative sea level and coeval climate change affected sedimentation of the Kreyenhagen Formation. Emphasis was placed on interpreting compositional variations in three dimensions to decipher spatiotemporal changes in benthic depositional environments. Data included eight wells with bulk- and trace-element geochemistry, 72 wells with varied logging tool suites, and petrophysical derivations of clay volume and TOC. The ~300 m (~1000 ft) section is presented as three informal members. Herein, we demonstrate that the Kreyenhagen Formation is a heterogeneous mudstone succession that records multiple fluctuations in paleoredox conditions, primary productivity of marine plankton, and detrital influx across space and time. These fluctuations can be reasonably tied to known records of global climate, relative sea level, and tectonics.

The Kreyenhagen Formation is located within the San Joaquin Basin, a 700-km-long (435-mi-long), asymmetrical structural trough bordered on the east by the Sierra Nevada, the south by the Tehachapi/San Emigdio Mountains, the west by the San Andreas fault zone and southern Diablo Range, and the north by the Stockton arch (Fig. 1). Sedimentary fill consists of 7620 m (~25,000 ft) of dominantly marine Mesozoic through Cenozoic strata (Hosford Scheirer and Magoon, 2007).

Deposition of the Kreyenhagen Formation occurred during the middle Eocene (48.6–37 Ma), by which time a well-defined forearc basin had developed and extended along the California margin (Dickinson and Seely, 1979; Ingersol1, 1979). Sedimentation coincided with increased accommodation space associated with a major middle Eocene transgression that affected the entire California margin (Milam, 1984, 1985; Bartow, 1991; Bloch, 1991a, 1991b). Data from benthic foraminifera indicate that the Kreyenhagen Formation was deposited at water depths between 200 and 2000 m (656 and 6560 ft; Cushman and Siegfus, 1942; Mallory, 1959; Phillips et al., 1974; Jefferis, 1984; Milam, 1985), which Milam (1985) interpreted to represent upper slope, lower slope, and base of slope environments. Uplift of the Franciscan accretionary prism at the western edge of the basin generated a long, narrow, largely isolated marine embayment bounded by the Sierran magmatic arc to its east (Ingersoll, 1979; Dickinson and Seely, 1979; Dickinson, 1995; Mitchell et al., 2010). Pelagic to hemipelagic accumulation occurred within a forearc setting analogous to the modern-day Chilean margin forearc (i.e., Fildani et al., 2008; Sharman et al., 2015). Coeval deep-water turbidite systems delivered sediment to the basin from the north, south, and eastern entry points during the middle Eocene (i.e., Point of Rocks Sandstone Member; Sharman et al., 2015). East of the basin axis, the Kreyenhagen Formation is time equivalent to shallow-marine and nonmarine facies called the Famoso sand and Walker Formation, respectively (Callaway, 1990; Bloch, 1991b).

Modern stratigraphic definition considers the Kreyenhagen Formation to be a middle Eocene (48.5–37 Ma) bathyal shale deposit composed of fine-grained, biogenic, siliceous and calcareous mudstone facies (Hosford Scheirer and Magoon, 2007; Johnson and Graham, 2007).

The Kreyenhagen Formation is composed primarily of siliceous, calcareous, and organic-rich mudstone with lesser amounts of sandstone, diatomite, porcelanite, limestone, and siltstone (Milam, 1985). The fine-grained sediment is a mixture of calcareous and siliceous microfossils, clay minerals, detrital quartz and feldspar, opal-A and diagenetic opal-CT, biogenic quartz, kerogen, and pyrite (Von Estorff, 1930; Jenkins, 1931; Cushman and Siegfus, 1942; Stewart, 1946; Milam, 1985). Biogenic components include skeletons of diatoms, radiolarians, foraminifera, and coccolithophores (Milam, 1985). The clay mineral component is dominated by chlorite, illite, and mixed-layer illite-smectite (Milam, 1985). TOC values are 0.5%–10%, and organic matter is dominantly type II marine (Milam, 1985; Peters et al., 2007).

In the Kettleman area—the focus of this study—the Kreyenhagen Formation is underlain by approximately time-equivalent transgressive sandstone and siltstone deposits called the Avenal Sandstone (Coalinga), Canoas Siltstone Member of the Kreyenhagen Formation (Reef Ridge), and Domengine Sandstone (Kettleman North Dome) (Fig. 2; Laiming, 1940; Harun, 1984). It is conformably overlain by fine-grained facies of the Eocene Tumey/Wagonwheel formations at Coalinga and Kettleman North Dome (Fig. 2). At Reef Ridge, the type section, the Kreyenhagen Formation is unconformably overlain by sandstone of the Miocene Temblor Formation (Fig. 2).

The type section of the Kreyenhagen Formation at Garza Creek on Reef Ridge is a 305 m (~1000 ft) section of primarily laminated mudstone, but it differs from other outcrop localities in containing more organic carbon, limestone beds, and locally interbedded sandstones (Von Estorff, 1930; Cushman and Siegfus, 1942; Stewart, 1946; Milam, 1985). It contains the 9 m (30 ft) clayey laminated Canoas siltstone member (Fig. 2; Cushman and Siegfus, 1942) and two fine-grained basal sandstone beds, ~1.5 and 9 m (5 and 30 ft) thick (Von Estorff, 1930; Cushman and Siegfus, 1942; Stewart, 1946; Milam, 1985). The lower third of the Kreyenhagen Formation is silty shale with three limestone beds up to 0.9 m (3 ft) thick (Von Estorff, 1930; Stewart, 1946; Milam, 1985). The middle section is described as hard, black, platy, laminated, and oil-impregnated shale (Von Estorff, 1930; Cushman and Siegfus, 1942; Stewart, 1946; Milam, 1985). The upper third of the section is silty opaline and radiolarian shale that contains a local 26-m-thick (85-ft-thick) sandstone body (Von Estorff, 1930; Cushman and Siegfus, 1942; Stewart, 1946; Milam, 1985).

This study was conducted in a 64 × 72 km (40 × 45 mi) area in the central San Joaquin Basin that contains Kettleman North Dome, Kettleman Middle Dome, and several smaller oil fields to the east, and the type section on Reef Ridge (Figs. 1 and 2). A long, narrow, northwest-trending anticlinal ridge called the Coalinga nose structurally controls the Kettleman North Dome and Kettleman Middle Dome oil fields (Fig. 1). The Kreyenhagen Formation is the primary source rock for Temblor Formation reservoirs in the Kettleman Fields and Coalinga (Magoon et al., 2009). Additionally, oil has long been produced directly from the Kreyenhagen Formation at Kettleman North Dome by hydraulic fracturing (Larue et al., 2018). Critical factors for extending this study east of the Kettleman structures were: (1) two deep wells (Zodiac 4-9, 1-10) drilled through the Kreyenhagen Formation in 2010 and 2011 at Kettleman City, and (2) availability of drill cuttings through the Kreyenhagen Formation from the Haven 44x-6 well near Westhaven oil field (Fig. 1; Table 1).

Inorganic geochemical data are valuable environmental proxies for paleoredox conditions (Calvert and Pedersen, 1993; Tribovillard et al., 2006; Algeo and Tribovillard, 2009), detrital input (Caplan and Bustin, 1999; Ver Straeten et al., 2011), and primary productivity (Brumsack, 2006; Tribovillard et al., 2006; Schoepfer et al., 2015) in sedimentary basins. Relative and absolute concentrations of different trace-element geochemical proxies are usually presented as absolute concentrations (ppm), absolute concentrations normalized by aluminum, and enrichment factors to infer paleoenvironmental conditions at deposition. Table 2 provides a summary of the proxies used in this study.

A common technique for accounting for dilution by detrital or biogenic components is to normalize trace-element concentrations to aluminum or to compare the degree of enrichment to that of average shale (Wedepohl, 1971; Taylor and McLennan, 1985; McLennan, 2001). These comparative results are called enrichment factors (EFs) and are calculated as EF_{element X} = (X/Al)_sample/(X/Al)_{average shale}. If EF_{element X} is greater than 1, then element X is enriched relative to average shale.

Aluminum (Al), titanium (Ti), and zirconium (Zr) are common constituents of terrigenous (detrital siliciclastic) sediment and are little influenced by biological and diagenetic processes following burial (Brumsack, 2006; Tribovillard et al., 2006). Aluminum is associated with clay and feldspar, whereas Ti occurs in clay-, sand-, and silt-sized grains such as ilmenite, rutile, and augite (Calvert, 1976; Shimmield, 1992). Zirconium is primarily associated with zircon, which is a heavy mineral indicative of a felsic sediment source.

Barium (Ba) exists in marine sediments primarily as barite (BaSO₄), which is precipitated in association with sinking organic matter (Tribovillard et al., 2006; Algeo and Ingall, 2007). The utility of Ba as a productivity proxy can be limited because barite is easily dissolved and remobilized during early diagenesis under strongly reducing conditions (i.e., sulfate reduction; van Os et al., 1991; Torres et al., 1996; van Santvoort et al., 1996). Consequently, effective use of Ba as a proxy may be restricted to oceanic environments with nonsulfidic bottom waters (Tribovillard et al., 2006).

Phosphorous, an essential nutrient for phytoplankton, is often used as a productivity indicator for ancient marine environments (Schoepfer et al., 2015). Phosphorous concentrations above background typically coincide with elevated organic carbon flux in regions of high biologic productivity, such as upwelling zones (Schenau et al., 2005). In strongly reducing environments, P can be remineralized into authigenic phosphate (Tribovillard et al., 2006; Algeo and Ingall, 2007) near the sediment-water interface and is most concentrated in environments with slow sedimentation rates (Garrison et al., 1994). For example, the phosphatic facies of the Monterey Formation of California are characteristic of condensed intervals and high surface-water productivity (Isaacs, 1985; Föllmi et al., 2005, 2017).

Biogenic silica (originally opal-A or SiO₂ • nH₂O) in the deep-marine environment is chiefly formed by skeletal hard parts (i.e., tests, frustules) of radiolarians or diatom plankton. With burial, time, and heating, opal-A eventually converts to diagenetic opal-CT and then microcrystalline quartz (SiO₂) (Behl, 1999). The abundance of biogenic silica provides fundamental information about surficial nutrient budgets and planktonic productivity.

Molybdenum (Mo) and uranium (U) are useful paleoredox proxies for several reasons, including: (1) low background detrital concentrations (Taylor and McLennan, 1985), (2) long residence times in seawater (Algeo and Tribovillard, 2009), and (3) low concentrations in marine phytoplankton. Uranium in oxic seawater primarily exists as soluble U(VI). However, under suboxic conditions, it is reduced to insoluble U(IV) and becomes enriched in sediment through several diagenetic processes (Zheng et al., 2000; Morford et al., 2001; McManus et al., 2005). Sedimentary U enrichment occurs within centimeters below the sediment-water interface and is enhanced by slow sedimentation rates, which allows more time for diffusion from the water column. Molybdenum in oxic seawater exists primarily within the molybdate ion (MoO₄^2–), which is delivered into sediment by adsorption to Mn and Fe oxides (Barling and Anbar, 2004). Additionally, where seawater is highly sulfidic, the molybdate ion converts to insoluble Mo-sulfur complexes and becomes enriched in sediment, often in association with organic matter and Fe-sulfide minerals such as pyrite (Helz et al., 1996; Erickson and Helz, 2000; Zheng et al., 2000). Unlike Mo, U accumulation in sediment is decoupled from water-column sulfide concentration (Algeo and Tribovillard, 2009).

Chromium (Cr) is primarily present in oxygenated seawater as Cr(IV) within the soluble chromate anion (CrO₄^2–) (Calvert and Pedersen, 1993). Under anoxic conditions, Cr is reduced to Cr(III), which can readily complex with acids or adsorb to Fe- and Mn-oxyhydroxides and be exported to the sediment (Breit and Wanty, 1991; Algeo and Maynard, 2004). The use of Cr as a redox proxy is somewhat limited because it may be lost to the water column under euxinic conditions and, in some cases, can be associated with terrigenous sediment (e.g., chromite, clay minerals; François, 1988; Brumsack, 1989; Hild and Brumsack, 1998).

Seventy-two well logs from the Kettleman area were used for correlation to construct a regional stratigraphic framework for the Kreyenhagen Formation (Table 1; Fig. 1). Of the 72 wells, 56 fully penetrated the Kreyenhagen Formation and 29 contained modern log suites. Correlations were based on spontaneous potential (SP), resistivity/induction, and gamma-ray logs. Subsurface structure, isopach, and property maps were generated on the basis of identified stratigraphic horizons. Contours were manually adjusted to remove geometries interpreted as geologically unreasonable products of gridding algorithms.

Where modern log suites were available, we calculated a clay fraction using the neutron porosity–density porosity separation method. We also estimated TOC from the ∆logR method using sonic and resistivity logs (Passey et al., 1990, 2010). To account for expelled petroleum in ∆logR, we assigned a level of maturity (LOM) value of 8 (early-onset generation) based on transformation ratios from Peters et al. (2007) for the Kreyenhagen Formation in the Kettleman area (0.25%–0.3%; expulsion occurs at ~10%). If the Kreyenhagen Formation has significant spatial variability in thermal maturity within the study area, estimates of TOC by ∆logR may be somewhat in error. Based on Peters et al. (2007), the southernmost portion of the study area (Kettleman Middle Dome) would be most susceptible to this error.

Whole-rock major-, minor-, and trace-element analyses were performed on 702 drill-cutting samples from eight of the 72 wells used in this study. Samples from seven wells were provided by California Resources Corporation and were originally analyzed by Chemostrat, Inc. (Houston, Texas, USA), using X-ray fluorescence with inductively coupled plasma–optical emission spectrometry (ICP-OES) and mass spectrometry (ICP-MS) following the Li-metaborate fusion procedure of Jarvis and Jarvis (1995). Vertical sampling density was generally at ~3, 6, or 9 m (10, 20, or 30 ft) intervals. An additional 17 samples of aggregate cuttings at 9–21 m (30–70 ft) intervals from one well (Haven 44x-6) were analyzed by ICP-OES and ICP-MS after lithium metaborate/tetraborate fusion for bulk- and trace-element geochemistry by Activation Laboratories Ltd. (Ancaster, Ontario, Canada; analysis code WRA+Trace-4 Lithoresearch).

Continuous TOC was modeled in the seven wells analyzed by Chemostrat, Inc., by calculating linear regressions in a sample set between selected trace elements and measured TOC (cf. Ratcliffe et al., 2012).

We estimated sedimentary components using a normative technique that corrected for loss on ignition (LOI) to produce a simplified three-component model of detritus, biogenic SiO₂, and carbonate following the technique of Medrano and Piper (1992), in which the detrital fraction was calculated by subtracting concentrations of biogenic SiO₂ and carbonate from the sum of the concentrations for all major oxides (Giannetta, 2019). The ratio of the residual detrital fraction to Al₂O₃ is given by the slope of the best-fit line between the two and is then used as a multiplier for Al₂O₃ to represent the detrital fraction. Total carbonate concentration was determined by summing calcite, dolomite, and CaO concentration within apatite (carbonate fluorapatite).

To estimate the amount of original biogenic SiO₂, we graphically determined a ratio cutoff of SiO₂:Al₂O₃ in weight percent that indicates the ratio in pure detritus in the Kreyenhagen Formation, (SiO₂/Al₂O₃)_detrital. Then, the value of (SiO₂/Al₂O₃)_detrital was multiplied by Al₂O₃ to determine the amount of detrital SiO₂, which was then subtracted from total SiO₂ to yield biogenic SiO₂, as in the following equation: SiO_2bio = SiO_2total – [Al₂O₃ × (SiO₂/Al₂O₃)_detrital], where SiO_2total and Al₂O₃ are total amounts of SiO₂ and Al₂O₃, respectively.

We subdivided the Kreyenhagen Formation into three informal members—lower, middle, and upper—based on log response and sedimentary composition (Fig. 3). With respect to petrophysical data (Fig. 3A), relative to other members, the lower Kreyenhagen member showed moderate gamma-ray values (API 80–100), moderate-to-high clay volume (45%–60%), low average biogenic SiO₂ (<17%), and low average TOC (<2%). The middle Kreyenhagen member had the highest gamma-ray API (>120; local value up to 180), high clay volume (50%–65%), and highest average TOC (>3%). The upper Kreyenhagen member was broadly characterized by relatively low gamma-ray values (API<150; low of 70), moderate TOC (1.5%–2.5%), low clay volume (<40%), and high biogenic SiO₂ (~35%–40%), and this member displayed more heterogeneity than the others (Fig. 3A).

Sediment component calculations showed that the Kreyenhagen Formation is composed primarily of detrital aluminosilicates and biogenic SiO₂ (Figs. 3B and Fig. 3D). Carbonate content is low (1%–5%), except for discreet intervals of the lower Kreyenhagen member with carbonate content up to 35% (Figs. 3B and Fig. 3D). The middle Kreyenhagen member is dominated by detritus (>60%), whereas the upper Kreyenhagen member has the greatest proportion of biogenic SiO₂ (Figs. 3B and Fig. 3D). The weight percent ratio of total detritus to Al₂O₃ components was found to be equal to 3.6 (Fig. 3C; multiplier for Al₂O₃).

For a detailed examination of the stratigraphic thickness and structure of the Kreyenhagen Formation, see Giannetta (2019); the following is a synopsis. The Kreyenhagen Formation thins from 457 m (1500 ft) at Kettleman Middle Dome to less than 152 m (500 ft) near Westhaven (Figs. 4A and Fig. 4B). A southwest-to-northeast correlation section shows that thinning is accommodated by thinning and eventual termination of the upper member (Figs. 4A and Fig. 4B). From southeast to northwest, the Kreyenhagen Formation thins abruptly north of Kettleman Middle Dome but maintains a relatively constant thickness along the Kettleman North Dome–Coalinga trend of ~305 m (1100 ft). The gross thickness trend identified in this study agrees with that of Hosford Scheirer (2007), who identified ~244–366 m (~800–1200 ft) of Kreyenhagen section at Kettleman North Dome to Kettleman City, thickening at Kettleman Middle Dome and progressively thinning east of the Kettleman area.

A structural contour map of the Kreyenhagen Formation shows that its geometry is dominated by the Coalinga Nose, a south-plunging anticlinal trend that encompasses the Kettleman Middle Dome, Kettleman North Dome, and Coalinga Extension East oil fields (Fig. 4C). There is significant structural relief in the area of the Avenal syncline (also referred to as Pleasant Valley syncline; Figs. 1 and 4C), where, within a 16 km (10 mi) transect, the formation is exposed in outcrop on Reef Ridge and then reaches a depth of more than 4877 m (16,000 ft) within the syncline (Fig. 4C). East of the Kettleman North Dome fold axis, the Kreyenhagen Formation is an asymmetric syncline that dips steeply off-structure and shallows gradually toward the northeast, reaching less than 3048 m (10,000 ft) in the northeasternmost extent of the study area (Fig. 4C).

This section presents geochemical relationships with specific proxies (Fig. 5) and vertical trends in geochemical proxies throughout the Kreyenhagen Formation (Figs. 6–9). For vertical trends, we refer primarily to Figure 6. Despite variations in magnitude, the trends observed in Figure 6, which is a compilation of data from wells Zodiac 1-10 and Zodiac 4-9 (Kettleman City), are generally representative of the Kreyenhagen Formation in the Kettleman area. For reference, Figures 7, 8, and 9 are geochemical logs from Kettleman North Dome, Kettleman Middle Dome, and near the Westhaven oil fields.

The relationship between TiO₂ and Al₂O₃ (wt%) shows strong positive correlation (r² = 0.88; Fig. 5A), confirming the consistency of Al and Ti as proxies of detritus. In contrast, Al₂O₃ and Zr are less correlated (r² = 0.35; Fig. 5B). Thus, Zr is not a consistent proxy for detritus.

Vertically through the Kreyenhagen Formation, detrital indicators (Al₂O₃ and TiO₂) tend to decrease upward, though there is finer-scale cyclicity (Fig. 6). Aluminum decreases upward in the lower member before increasing at the boundary with the middle member. In the lower half of the middle member, Al₂O₃ content is the highest in the formation, and then it gradually decreases into the upper member, where Al₂O₃ is consistently low (Fig. 6). Vertical TiO₂ variation is less pronounced than Al₂O₃, but the proxy displays a similar trend. A difference is that TiO₂ in the middle Kreyenhagen member is less noticeably elevated compared to the lower and upper members (Fig. 6). A general decreasing-upward TiO₂ trend through the entire succession is more readily observable. Clay volume estimated from neutron-density separation aligns closely with Al₂O₃ (Fig. 6), suggesting that aluminosilicate minerals are mostly hydrous clay minerals (fine-grained detritus).

Using the ratio of SiO₂ to Al₂O₃ in “pure” detritus, (SiO₂/Al₂O₃)_detrital (“detrital cutoff” in Fig. 5C), the value of (SiO₂/Al₂O₃)_detrital in the Kreyenhagen Formation is to equal 3.1 (Fig. 5C), which agrees with typical values of average shale 3.1 (Wedepohl, 1971). Most SiO₂ in the Kreyenhagen Formation is above the detrital cutoff ratio of 3.1 and follows the trend of biogenic SiO₂ (Fig. 5C).

The biogenic SiO₂ trend is more consistent across wells than the other paleoproductivity proxies. Biogenic SiO₂ is moderate in the lower member and at the base of the middle member (Fig. 3 and 6–9). Within the middle member, biogenic SiO₂ content falls to almost 0% and then increases upward into the upper Kreyenhagen member, where it remains consistently elevated to 30%–50% (Fig. 3 and 6).

Like biogenic SiO₂, Ba content is typically highest in the upper member, but the proxy is highly variable (Fig. 6–9). Phosphorous enrichment does not display any clear trends or cyclicity, though peak concentrations mostly occur in the upper member (Figs. 6–8). Several wells have spurious P peaks above background concentrations (Figs. 6–8) that do not correlate between wells and are thus considered outliers or nonrepresentative local phenomena. At Kettleman City (Fig. 6; ~4467 m/~14,655 ft) and Kettleman North Dome (Fig. 7; ~2667 m/~8750 ft), the peaks are not associated with high TOC and are only ~0.9% of normalized composition, which suggests the P was not sourced from organic matter or remineralized carbonate fluorapatite, respectively; we interpret these outliers as particulate P, either dead plankton or bioapatite (fish teeth or scales). In contrast, P peaks within the middle member at Kettleman Middle Dome (Fig. 8; ~3658 m/~12,001 ft) and Kettleman North Dome (Fig. 7; ~2743 m;/~8999 ft) are associated with high TOC intervals and may have been derived from organic matter.

All three paleoredox proxies—Mo, U, and Cr—display an increasing-decreasing trend upward through the Kreyenhagen Formation, with the highest concentrations in the middle member (Figs. 6–9). Also, Mo and U concentrations in the Kreyenhagen Formation are consistently above those of post-Archean average shale (PAAS; Taylor and McLennan, 1985; Figs. 6–9). Molybdenum variability is less pronounced than Cr and U (Fig. 6). Concentration of Mo in the lower Kreyenhagen member varies between 1 and 8 ppm and elevates to 7–15 ppm in the middle member, with local concentrations >35 ppm (KMDU 17-29; Fig. 8). Typical upper member Mo concentrations are less than 10 ppm. Uranium content is typically less than 4 ppm in the lower member and increases to >7 ppm in the middle member, where U concentration is highest (Figs. 6–9). In the upper Kreyenhagen member, U decreases to 4–7 ppm, although there are a few peak values in excess of 8 ppm (Fig. 6).

Chromium shows little covariation with Al₂O₃ within the Kreyenhagen Formation (r² = 0.29; Fig. 5D), suggesting Cr can be reasonably applied as a redox indicator. Chromium (Cr/Al₂O₃) shows an upward increasing-deceasing cycle that is similar to but more pronounced than U (Fig. 6). Average chromium enrichment is highest in the middle member (>10 ppm/wt%; Fig. 6) compared to the lower and upper members, which are the least Cr enriched. The curve of TOC resembles Cr concentration more closely than Mo and U (Fig. 8). This resemblance is especially distinct in the middle Kreyenhagen member, where peak enrichment of Cr correlates with peak TOC enrichment (~4496 m/14,750 ft; Fig. 6).

In addition to its vertical variability, the Kreyenhagen Formation varies laterally in its composition, as shown by log-derived TOC, log-derived clay volume, and bulk geochemistry.

TOC estimated by the ∆logR method (Passey et al., 1990) could only be calculated in 19 wells where sonic logs were available (Fig. 10A). Average TOC is highest (3.5%–2.5%) along the Kettleman North Dome and Kettleman Middle Dome anticlines and decreases toward the northeast, where values fall below 1% (Fig. 10A). The wells with the highest average TOC (~3.5%) are just northeast of Kettleman North Dome (Fig. 10A). This eastward-decreasing trend in TOC is consistent with the previously published study of TOC within the Kreyenhagen Formation in the Kettleman area (Peters et al., 2007).

Average clay volume of the Kreyenhagen Formation (based on 23 wells) shows an eastward increase (Fig. 10B). The well with the lowest clay volume, Solimar Energy-1, is located near the Guijarral Hills oil field, where clay volume is less than 35%. Along the Kettleman North Dome trend, clay volume averages 40%–45%, but the formation becomes more clay-rich toward the south at Kettleman Middle Dome, where averages exceed 50%. Wells in the northeast of the study area have clay volume of >60% (Fig. 10B).

The compositions of three key proxies are shown in map view in Figure 11: Al₂O₃ (detritus; Figs. 11A and Fig. 11D), biogenic SiO₂ (productivity; Figs. 11B and Fig. 11E), and EFMo (redox conditions; Figs. 11C and Fig. 11F). To highlight trends along an approximately west-east transect, we present the average proxy content of four oil fields: Kettleman North Dome (three wells), Kettleman Middle Dome (three wells), Kettleman City (one well), and Westhaven (one well).

Aluminum content increases from western to eastern areas (Figs. 11A and Fig. 11D). Wells on Kettleman North Dome have the lowest Al₂O₃ values, averaging 11.5%. Average Al₂O₃ contents at Kettleman Middle Dome, Kettleman City, and Westhaven are 12.8%, 13.4%, and 13.9%, respectively (Figs. 11A and Fig. 11D). Numerical differences in Al₂O₃ content are minimal, as the maximum difference between wells is only 4% (9.9% at 27K-7Q-RD; 13.9% at Haven 44x-6).

Biogenic SiO₂ and Mo (presented as EFMo) have the opposite trends as compared to Al₂O₃ and decrease from west to east (Figs. 11B, 11C, 11E, and 11F). Biogenic SiO₂ content is highest in wells along Kettleman North Dome, where it averages 25.8% (Figs. 11B and Fig. 11E). From west to east, biogenic SiO₂ contents at Kettleman Middle Dome, Kettleman City, and Westhaven are 23.7%, 18%, and 14.2%, respectively. Maximum variations in biogenic SiO₂ content between the four areas are greater than those of Al₂O₃. Specifically, there is variability in weight percent of 13% between the two end-member wells (27K-7Q-RD1 and Haven 44x-6). Average EFMo values from west to east at Kettleman North Dome, Kettleman Middle Dome, Kettleman City, and Westhaven are 11.9, 9.0, 8.9, and 5.1, respectively (Figs. 11C and Fig. 11F).

The following section discusses and interprets the rock types that comprise the three informal members of the Kreyenhagen Formation defined in this study. For support, outcrop studies of the Kreyenhagen Formation were compared to subsurface findings of our study. We then coupled rock-type interpretations with geochemical proxies to interpret variability in depositional environments, primarily with respect to detritus, paleoproductivity, and redox conditions. We found that the Kreyenhagen Formation is lithologically heterogeneous both vertically and laterally, as is the case in most well-studied mudstone deposits (Aplin and Macquaker, 2011). Last, we present a paleogeographic reconstruction and three-stage depositional model of the Kreyenhagen Formation. Table 3 provides a summary of our interpretations.

The lower member contains a high proportion of terrigenous-derived components (clay minerals, Al₂O₃, TiO₂) relative to organic matter (<2% TOC) and biogenic SiO₂ (<10%) (Figs. 3 and 6–9). The lower Kreyenhagen member has the highest Th/U fraction in the formation (Fig. 12A), suggesting a more oxygenated and/or detrital-rich environment (Jones and Manning, 1994; Tribovillard et al., 2006). The ratio of Th/U is a useful environmental indicator because Th is concentrated in clay-rich sediments and tends to be immobile during diagenesis, whereas fixation of U into sediment is accelerated under reducing conditions or in the presence of organic matter (Adams and Weaver, 1958). High TiO₂/Al₂O₃—a proxy for the silt:clay ratio (Fig. 12C; Boyle, 1983)—within the lower Kreyenhagen member indicates that, although fine grained, it is probably siltier than the middle member. Last, the lower Kreyenhagen member is more calcareous than other members, as evidenced by ~3%–8% background carbonate and discreet intervals of ~40% carbonate (Fig. 3). Local carbonate maxima are made of calcite (limestone) rather than dolomite, which contributes a consistent background of ~4% (Fig. 12B). We infer that lower member data points represent a mixed mudstone-marlstone-limestone geochemical array (Fig. 13; approximate cutoffs based on Behl and Gross, 2018).

In sum, the lower Kreyenhagen member across the subsurface consists of an organic-poor (<2% TOC), calcareous silty mudstone and marl with discreet limestone beds. This is consistent with the lower Kreyenhagen member outcrops on Reef Ridge and Oro Loma Creek, which contain limestone lenses up to 0.9 m (3 ft) thick and clay shale with marl (Von Estorff, 1930; Stewart, 1946; Milam, 1985). TiO₂/Al₂O₃ ratios indicate that the silt content of the lower Kreyenhagen member in outcrop is comparable to that in the subsurface. The basal laminated Canoas Siltstone Member of the Kreyenhagen Formation on Reef Ridge overlies the transgressive Avenal Sandstone (Domengine Sandstone equivalent; Von Estorff, 1930; Stewart, 1946) and is likely equivalent to the silt-rich subsurface lower Kreyenhagen member of this study.

The middle member is the most organic-rich interval of the formation (average TOC >2.5%; Figs. 3 and 6–8). Values of Th/U and TiO₂/Al₂O₃ both decrease in the middle Kreyenhagen member (Figs. 12A and Fig. 12C), indicating that the middle member has (1) higher TOC:detrital clay and (2) a lower silt:clay ratio. Thus, clay minerals and organic matter are principal components; this interpretation is also supported by high gamma-ray activity (>150 API; Figs. 6–8). Sedimentary components plot the middle Kreyenhagen member within a clayey mudstone geochemical array (Fig. 13). Trace-element proxies for anoxia (U, Cr) and euxinia (Mo) are also elevated (Figs. 6–9), as is characteristic of organic-rich mudstone deposits. The Fe/Al ratio, often used to aid in the recognition of syngenetic pyrite formation and therefore strongly reducing water-column conditions, is highest in the middle Kreyenhagen member (Fig. 12D).

The middle Kreyenhagen member is an organic-rich clayey mudstone equivalent to middle–lower Kreyenhagen sections in nearby outcrops, such as on Reef Ridge, which consist of ~122 m (~400 ft) of thinly bedded, oil-impregnated, dark purplish-brown argillaceous mudstone that overlies calcareous and silty shales (Von Estorff, 1930; Cushman and Siegfus, 1942; Stewart, 1946).

The upper member is the most siliceous part of the formation, containing elevated biogenic SiO₂ in all wells analyzed (Figs. 3, 6–9, and 13). Increases in Th/U and TiO₂/Al₂O₃ in the upper Kreyenhagen member suggest a decrease in U or organic matter and an increase in grain size, respectively (Figs. 12A and Fig. 12C). Sedimentary components show a nearly bimodal mixture of biogenic SiO₂ and detritus, chiefly representing siliceous mudstone (Fig. 13); several intervals have almost 70% biogenic SiO₂ and are considered porcelanite based on Behl and Gross (2018) (Fig. 13). Additionally, gamma-ray activity is typically lower in the upper Kreyenhagen member (<70 API; Figs. 3, 7, and 8) than in the middle and lower members, indicative of lower clay mineral content and TOC. Last, enrichment of redox-sensitive trace elements is lower than in the middle member (Figs. 3 and 6–9), which is expected for a less organic carbon–rich lithology.

We interpret the upper Kreyenhagen member to be predominantly organic-poor (<2% TOC), silty siliceous mudstone with occasional porcelanite (restricted to the basal portion of the upper Kreyenhagen member; Fig. 3D). Previous field studies described the upper Kreyenhagen member as the most siliceous interval in the formation, calling the upper 91 m (300 ft) “opaline” or “radiolarian” shale (Von Estorff, 1930; Stewart, 1946). Cushman and Siegfus (1942) noted a grain-size increase in the upper Kreyenhagen member, calling it “silty-sandy shale.” Based on these comparisons, the subsurface upper Kreyenhagen member of this study appears to be correlative to the upper Kreyenhagen member facies observed in outcrop.

Deposition of the Kreyenhagen Formation encompasses a transgressive-regressive cycle that likely exerted a key control on detrital input that operated independently from changes in marine biogenic sediment productivity (discussed later). Variation in accommodation space is inferred to primarily reflect local tectonics in this convergent-margin setting rather than eustatic fluctuation. Early deposition of the Kreyenhagen Formation coincided with a regional transgression that began to starve the Eocene forearc of detritus (Milam, 1984, 1985; Bartow, 1991; Bloch, 1991a, 1991b). The underlying shallow-marine Domengine Sandstone reflects the initiation of transgression (Milam, 1985), followed by a fining-upward succession in the lower to middle Kreyenhagen members. A regression at the end of Kreyenhagen Formation deposition is recorded in the north and south margins of the San Joaquin Basin by erosional truncations of the Kreyenhagen Formation in outcrop (Bartow, 1991).

The lower member reflects a period of higher detrital input relative to the middle and upper Kreyenhagen members, likely related to late stages of a lowstand systems tract or initial stages of transgression. Consistently high values of Al₂O₃, TiO₂, and TiO₂/Al₂O₃ (Figs. 6–9 and 12C), especially in the basal portion, suggest either (1) a clastic sediment source close enough to deliver silt-sized particles directly to the depocenter, or (2) depositional energies that were adequate for transporting the silt-sized fraction. Such a sedimentation pattern is consistent with the conformable, silty, fining-upward succession observed in outcrop (Von Estorff, 1930; Stewart, 1946).

The middle member reflects a transitional interval that encompasses the period of lowest detrital input and a potential maximum flooding event associated with a regional sea-level highstand. It contains relatively high detrital content in the lower portion that decreases upward (Figs. 6–8) as TiO₂/Al₂O₃ and Th/U also decrease (Fig. 12C). It is likely that deposition of the middle member coincided with (1) a decrease in overall detrital input, and (2) a transition from silt- to clay-sized particle deposition. This fining-upward succession, also indicated by upward-increasing gamma-ray logs (Figs. 6–8), was likely related to a landward shift in the clastic sediment source that starved the depositional basin of detritus.

Upper member detrital indicators suggest increased detrital delivery to the depocenter. Factors controlling this may have been a regression associated with eustatic sea-level fall or progradation during a highstand. Aluminum and TiO₂ content are low at the base of the upper Kreyenhagen member and increase upward to the top of the formation (Figs. 6–8). Th/U and TiO₂/Al₂O₃ also increase upward in the upper member (Fig. 12A), indicating increased silty detritus. Milam (1984) calculated the highest sediment accumulation rate for the formation on Reef Ridge during the interval that we infer is correlative with the upper member. Last, the local 26 m (85 ft) unnamed sandstone body in the upper Kreyenhagen member outcrop on Reef Ridge suggests episodically high clastic input.

Lateral trends in clay volume and geochemical proxies of detritus are consistent with a principal clastic sediment source east of the study area that supplied fine-grained detritus (Figs. 10B, 11A, and 11D). Average Al₂O₃ content and clay volume both show consistent eastward-increasing patterns from Kettleman North Dome to Westhaven (Figs. 11A and Fig. 11D). The eastward increase in detritus probably reflects progressively more proximal environments, with Westhaven oil field (Haven 44x-6) being closest to the paleoshoreline.

Previous authors also suggested that the central Sierra Nevada to the east was the principal source of clastic Eocene forearc deposits (Clark, 1964; Bartow, 1991; Sharman, 2014). Coeval bathyal coarse-clastic deposits, such as the Point of Rocks Sandstone Member of the Kreyenhagen Formation, have Sierra Nevadan provenance signatures (Nilsen and Clarke, 1975; Graham and Berry, 1979; Sharman, 2014). East of the deep-marine basin, time-equivalent shallow-marine and nonmarine units (Famoso sand and Walker Formation) probably formed a belt of west-flowing fluvial-deltaic systems and associated prodeltas that supplied fine silt and clay basinward.

Deposition of the Kreyenhagen Formation coincided with a period of dramatic climatic cooling and global diatomaceous productivity (Barron et al., 2015) that likely controlled biogenic input in the Kreyenhagen Formation. A cooling trend began ca. 49 Ma following the early Eocene climatic optimum—the warmest period of the Cenozoic—that continued until the early Oligocene (Zachos et al., 2001, 2008). The transition from early Eocene (ca. 56–49 Ma) greenhouse conditions to Oligocene (ca. 34 Ma) icehouse conditions encompasses the Kreyenhagen Formation depositional period (48.6–37 Ma). A brief but punctuated climatic perturbation occurred during the middle Eocene climatic optimum (ca. 40 Ma; Miller et al., 1991; Bohaty et al., 2009; Wilson et al., 2013), and this was immediately followed by intensification of pole-to-equator temperature gradients and coastal upwelling on continental margins (McLean and Barron, 1988; Bosboom et al., 2014; Lazarus et al., 2014). Another factor critical to middle Eocene paleoceanography was the initial opening of the Drake Passage between Antarctica and South America, which led to the onset of the Antarctic Circumpolar Current and widespread diatom deposition in Pacific Ocean regions (Thomas, 2004; Scher and Martin, 2006; Thomas et al., 2008). Diatoms of the Kreyenhagen Formation provide some of the earliest evidence for middle Eocene Pacific upwelling as well as amplified productivity following the middle Eocene climatic optimum (Dumoulin, 1984).

The lower member has the lowest concentration of biogenic SiO₂, Ba, and P, suggesting that, in comparison to the middle and upper Kreyenhagen members, either (1) rates of primary productivity in the overlying water column were low, or (2) clastic dilution reduced the productivity proxy signal. Lower rates of primary productivity are expected during the early middle Eocene, as a reduced pole-to-equator temperature gradient resulted in warm surface and deep waters that were poorly mixed by weak zonal winds (Barron et al., 2015). The lower Kreyenhagen member is also the most enriched in carbonate, especially calcite, likely sourced from calcareous nannoplankton (Milam, 1984, 1985). High CaCO₃ content could have resulted from several oceanographic factors, all of which relate to warmer climates of the early–middle Eocene, including (1) favored productivity of calcareous plankton relative to siliceous plankton, (2) enhanced water-column stratification that isolated the deep, nutrient-rich waters required by diatoms from surface waters, and (3) a deeper carbonate compensation depth prior to the opening of the Drake’s Passage and associated entry of older, more CO₂-rich waters into the Pacific Ocean Basin.

The middle member has low values of biogenic SiO₂ and Ba at its base, though these components increase upward (Figs. 3 and 6–8), likely a reflection of increasing surface-water productivity. High values of TOC in the middle Kreyenhagen member probably relate to marine phytoplankton preservation, as indicated by the high hydrogen index (550–600 HC/g TOC), a predominance of type II “marine” kerogen (Peters et al., 2007), and reduced dilution by siliciclastic detritus. Phosphorous, biogenic SiO₂, and CaCO₃ are low in the most organic-rich intervals (Figs. 6–8). Phosphorus may have been released to the overlying water column during sulfidic conditions where Mo exceeds 25 ppm. Relatively low biogenic SiO₂ and CaCO₃ could have resulted from dissolution of biogenic opal and calcareous tests during slow sedimentation during lower–middle Kreyenhagen member deposition (0.6 cm/1000 yr; Milam, 1985), as similarly interpreted by Isaacs (2001) in the Monterey Formation.

The upper member has the highest abundances of biogenic SiO₂ (up to 45%), Ba, and P, indicating the highest levels of primary productivity (Figs. 3 and 6–9). Primary productivity was likely accelerated following increased availability of surficial nutrients associated with (1) middle Eocene climatic cooling and intensified coastal upwelling following the middle Eocene climatic optimum (Bohaty et al., 2009; Barron et al., 2015), and (2) the initiation of Pacific Ocean diatom blooms following the opening of the Drake Passage (Livermore et al., 2007), which brought (3) nutrient-rich deep water to the Pacific Ocean Basin. The upper Kreyenhagen member equivalent at Devil’s Den (Welcome Shale) is dated by calcareous nannofossils to ca. 37 Ma (Milam, 1985), which postdates the middle Eocene climatic optimum (ca. 40 Ma) and the inferred timing of the Drake Passage opening (ca. 39–40 Ma; Dumoulin, 1984; McLean and Barron, 1988; Barron et al., 2015). Thus, deposition of the upper portion of the Kreyenhagen Formation may have postdated these oceanographic shifts, which were the likely forcing mechanisms for enhanced productivity.

Laterally, biogenic SiO₂ content decreases from Kettleman North Dome toward the eastern (proximal) wells (Zodiac 4-9, Haven 44x-6). We attribute this decrease to dilution of biogenic components by eastern sources of detrital input (Sierra Nevada).

In the lower member, enrichment factors of Mo and U average less than 6 and 2, respectively, indicating no significant enrichment (Wedepohl, 1971; Algeo and Tribovillard, 2009). Additionally, relative to other members, values of Cr/Al₂O₃ (anoxia proxy) and Fe/Al (euxinia proxy) are lowest (Figs. 6–9 and 12D). Lower Kreyenhagen member Th/U values suggest dominantly oxic-suboxic redox conditions (Fig. 14). Low TOC relative to other Kreyenhagen members may also suggest more water-column oxygenation or, possibly, increased dilution by fine detritus. We infer, based on agreement of all these proxies, that benthic environments during lower Kreyenhagen deposition were mostly oxic-suboxic.

The middle member represents a shift to more reducing bottom water conditions, as evidenced by the highest enrichment in paleoredox proxies of oxygen depletion and TOC. The basal portion is moderately elevated in TOC and proxies for anoxia (U, Cr) and euxinia (Mo) (average EFU = 2.25; EFMo = 8.5). The middle portion of the member marks a pronounced increase in TOC, Mo, U, and Cr; average EFMo and EFU are 16 and 3, respectively, reflecting significant and detectable enrichment, respectively. Th/U ratios of the middle Kreyenhagen member are the lowest in the formation and commonly indicate anoxia (Figs. 14A and Fig. 14B). Discrete intervals of the middle Kreyenhagen member exhibit Mo concentrations of 30–85 ppm (Figs. 6–8), which suggest intermittent euxinia (>25 ppm; Scott and Lyons, 2012; Dahl et al., 2013). Supporting this notion, elevated Fe/Al values (Fig. 12D) may reflect syngenetic pyrite formation under euxinic conditions. We infer that the middle Kreyenhagen member represents a transition to anoxic benthic conditions with intermittent periods of euxinia.

The upper Kreyenhagen member reflects a shift back to slightly more oxygenated bottom water conditions. Mo, Cr, and especially U all decrease into the member and remain moderate throughout (Figs. 6–9). Th/U and Fe/Al slightly increase and decrease compared to the middle Kreyenhagen member, respectively (Figs. 12A and Fig. 12D). TOC also decreases into the upper Kreyenhagen member. Th/U ratios of the upper Kreyenhagen member indicate highly fluctuating redox conditions that range from oxic to anoxic, but with lower overall U enrichment than the middle Kreyenhagen member (Fig. 14). We interpret the decrease in trace-element enrichment and TOC in the upper member as a gradual oxygenation of bottom waters with frequent intermittent redox fluctuations.

Patterns of Mo-U covariation (Mo/U) provide an additional tool for interpreting paleoredox conditions. Algeo and Tribovillard (2009) demonstrated that U enrichment that exceeds Mo implies suboxic conditions, whereas Mo enrichment that exceeds U indicates sulfidic bottom waters or the operation of an Mn-Fe-oxyhydroxide particulate shuttle (see “Background: Paleoredox Proxies”; Algeo and Tribovillard, 2009; Tribovillard et al., 2012). This shuttling process occurs when Mn-Fe oxyhydroxides, which form at the chemocline, adsorb molybdate oxyanions during transit through the water column. When oxides reach the sediment-water interface, they may be reductively dissolved, releasing Mo, which can then be retained in sediments (Morford and Emerson, 1999; Morford et al., 2005). Because U is decoupled from this process, the particulate shuttle enhances Mo uptake relative to U. Importantly, the process is augmented by frequent water-column redox fluctuations, as vertical translation of the chemocline leads to Fe and Mn oxidation, which drives the shuttle process (Algeo and Tribovillard, 2009).

Mo/U values of the Kreyenhagen Formation generally fall between 3 and 7 (average = 4.8; standard deviation = 2; Fig. 15), suggesting ephemeral formation of a particulate shuttle (Algeo and Tribovillard, 2009). Mo/U values of the Kreyenhagen Formation resemble those of the modern-day Cariaco Basin, an ~700 km² (270 mi²) silled basin on the Venezuelan continental shelf (Jacobs et al., 1985) that experiences frequent redox fluctuations and vertical translation of the chemocline, and enhanced sedimentary Mo uptake. Based on similar Mo/U ratios in the Kreyenhagen Formation, the middle Eocene forearc may have also been a semirestricted water body that experienced episodic redox fluctuations. Geologic evidence also exists for a bounding uplifted accretionary prism and coastal upwelling along the middle Eocene California margin (Mitchell et al., 2010; Barron et al., 2015).

Spatial trends in trace-element enrichment suggest differing redox conditions across the study area (Figs. 11C, 11F, and 14). Average EFMo is 11.9 at Kettleman North Dome (westernmost area) and 5.1 at Westhaven (easternmost area) (Figs. 15C and Fig. 15F). This represents a deep-to-shallow change from significant enrichment to just above detectable enrichment. Similarly, EFU and Cr/Al₂O₃ decrease from Kettleman North Dome (western area) to Westhaven (eastern area). Ratios of Th/U show progressively oxygenated conditions eastward from Kettleman North Dome to Westhaven, where no Th/U values fall within the anoxic realm (Figs. 14A–14D). Based on these compositional trends, we interpret that, at the time of deposition, reducing conditions were stronger in western environments (interpreted as distal), with more oxygenated conditions in eastern environments (interpreted as proximal). Similarly, Mo/U ratios decrease eastward from Kettleman North Dome through Kettleman Middle Dome and Kettleman City to Westhaven (Fig. 15). This pattern suggests that distal western environments experienced a stronger chemocline with more frequent redox fluctuations and euxinic episodes, which drove increased Mo uptake in the sediment.

A regional transgression during deposition of the lower Kreyenhagen member began to starve the Eocene forearc of detritus (Milam, 1984, 1985; Bartow, 1991; Bloch, 1991a, 1991b). Increasing water depths with still warm Eocene surface waters likely enhanced water-column stratification and initiated suboxic conditions in bottom waters. TOC in the lower member has the strongest correlation with U (R² = 0.58; Fig. 16C). Because U enrichment begins at the onset of suboxia (Klinkhammer and Palmer, 1991; Crusius et al., 1996; Zheng et al., 2000; Morford et al., 2001; Chaillou et al., 2002; McManus et al., 2005), the correlation suggests that suboxia modulated organic matter preservation (not anoxia). Higher rates of clastic dilution due to late stages of a lowstand or early transgression probably limited organic matter concentration to lower TOC values than other members (Fig. 17A).

Organic-rich facies of the middle member are interpreted to reflect the period of the highest relative sea level, most reducing bottom waters, and greatest organic matter preservation (Fig. 17B). Inhibited mixing of surficial oxygen with deep water promoted benthic anoxia and euxina (Fig. 17B). TOC correlates closest with Cr (R² = 0.51; Fig. 16B). Cr(IV) reduction to Cr(III) requires anoxic conditions (Breit and Wanty, 1991; Algeo and Maynard, 2004), indicating that bottom water anoxia modulated organic matter preservation. High Mo concentrations (30–85 ppm) suggest that intermittent euxinia may have preserved organic matter as well (Fig. 16A; Scott and Lyons, 2012; Dahl et al., 2013). The enhanced propensity for anoxia could relate to better isolation of the basin by the then-uplifted Franciscan accretionary wedge (Fig. 17B; Dickinson and Seely, 1979; Mitchell et al., 2010). High Mo/U values in the middle member fall within the “weakly restricted” trend of Algeo and Tribovillard (2009).

We interpret the organic matter sedimentation, preservation, and dilution in the upper member as the result of increased upwelling-related productivity associated with known Eocene climatic cooling and relative sea-level fall (Fig. 17C). The regression at the end of the Eocene, as recorded in outcrops by top-down erosional truncations of the Kreyenhagen Formation (Bartow, 1991), would have shifted the clastic sediment source closer to the basin and increased detrital input. Also, the upper member thins and eventually terminates eastward, which we interpret as an erosional truncation during sea-level fall (Figs. 4A and 17C). TOC in the upper member correlates more closely with Al₂O₃ (R² = 0.45) than any redox-sensitive trace elements or biogenic SiO₂ (Fig. 16E). From this, we infer the forcing function of organic matter preservation to have been high detrital sedimentation rates, which limited the exposure time of organic matter to oxygen and aerobic microorganisms. However, Th/U values in the upper member suggest intermittent periods of anoxia (Fig. 14). It is possible that climate-forced biogenic plankton input increased oxygen demand in the benthos, causing periods of anoxia (Fig. 17C).

The spatial distribution of facies within the Kreyenhagen Formation allows inferences about forearc paleogeography during the middle Eocene (Fig. 18). We integrated the findings of this study with benthic foraminiferal paleobathymetry (Cushman and Siegfus, 1942; Mallory, 1959) and middle Eocene reconstructions from Sharman (2014) and Bartow (1991). Figure 18 is a paleogeographic reconstruction that illustrates the dominant sedimentary environments within the middle Eocene forearc basin. On the basis of the east-west proximal-distal relationship implied from trends of geochemical proxies, we interpret bathyal depositional environments along Kettleman North Dome and progressively shallower and more proximal environments toward the east. Eastern deltas of the time-equivalent Famoso sand and Walker Formation supplied fine detritus, and depositional environments became increasingly clastic-starved toward the west. In the context of existing oil fields, the order from most proximal to most distal environments would be Westhaven, Kettleman City, Kettleman Middle Dome, and Kettleman North Dome. A portion of the Franciscan subduction complex (i.e., the uplifted accretionary prism) likely bounded the basin and restricted oceanic circulation to the forearc from the west (Fig. 18).

The middle Eocene Kreyenhagen Formation consists of heterogeneous siliceous-detrital mudstone with variable thickness and composition. Petrophysical analysis and geochemical proxies indicate a diverse array of lithotypes that vary both vertically and laterally in the Kettleman area of the San Joaquin Basin. Detrital components and clay minerals comprise more than 50% and biogenic silica composes ~25% of the formation, with only local intervals of purer porcelanite and limestone. Subsurface mapping shows that the formation thins eastward from 457 m (1500 ft) to less than 183 m (600 ft).

We identified three informal members on the basis of log responses and geochemical compositions that correlate from known oil fields into the deeper San Joaquin Basin toward the east. The lower member is a calcareous silty mudstone with limestone beds that is relatively depleted in proxies of anoxia and paleoproductivity. The middle member is an organic-rich clayey mudstone that is relatively enriched in proxies of anoxic or euxinic paleoredox conditions. The upper member is a silty siliceous mudstone with porcelanite and high concentrations of paleoproductivity proxies. Vertical facies differences are most pronounced in western areas like Kettleman North Dome, as the detrital fraction throughout the formation increases eastward.

With some exceptions, petrophysical estimates of TOC, clay volume, and geochemical proxies indicate an east-west proximal-distal relationship. Fine-grained detritus was likely sourced from shelfal deltas derived from the Sierran magmatic arc. In contrast, western areas along the Coalinga Nose trend were more distal and anoxic. Vertical compositional changes likely record at least one transgressive-regressive cycle that appears to be correlative with known eustatic sea-level changes. Relative sea-level changes and watermass restriction from the uplifted Franciscan subduction complex likely exerted important controls on lithology and organic matter preservation. Last, we observed a distinct up-section increase in proxies of paleoproductivity, notably biogenic SiO₂ content. We attribute this to increased upwelling intensity associated with climatic cooling during the middle Eocene.

This research was made possible by the support of the affiliates of the California State University–Long Beach Monterey and Related Sedimentary Rocks (MARS) project; thanks go to all project affiliates for continued student support in fundamental and applied sedimentology. Special thanks go to California Resources Corporation for their generous contribution of data, which made this research possible. We thank Schlumberger and TIBCO for providing free student licenses of Petrel and Spotfire, respectively.