Short-term sediment accumulation rates determined from Eocene alluvial paleosols

A new method uses alluvial paleosols to calculate sediment accumulation rates for thin (25 m) stratigraphic intervals and allows the reliable interpolation of ages for stratigraphic levels within a thick stratigraphic interval bounded by established dates. Sediment accumulation rates calculated for a 650 m composite section in the Eocene Willwood Formation of Wyoming span time intervals ranging from only 0.05 to 0.25 may. Important sedimentologic changes coincide with changes in accumulation rate and indicate close and direct relations between the history of basin subsidence and depositional patterns. INTRODUCTION Estimating sediment accumulation rates can improve understanding of sedimentary basin development. In foreland basins, the record of sediment accumulation can be used to determine spatial and temporal subsidence patterns of the basin, and these patterns provide information on thrusting history. Differential subsidence may also influence basinal facies patterns (e.g., Alexander and Leeder, 1987); therefore, resolving accumulation rates is also important to sedimentologic and stratigraphic analyses. Sediment accumulation rates are commonly calculated by dividing the thickness of a stratigraphic section by its known or estimated time span. Because of compaction, this method yields rock accumulation rates. Sediment accumulation rates are readily calculated by decompacting a rock thickness to sediment thickness prior to burial (e.g., Baldwin and Butler, 1985). More difficult in most continental sequences is determining how much time is represented by the stratigraphic section of interest. With few exceptions (e.g., Johnson et al., 1988), the temporal resolution of paleomagnetic dating is still relatively coarse. Radiometric dates cannot commonly be obtained and, even where available, are generally widely spaced in time. Similarly, continental biostratigraphy produces relative dates that are poorly constrained or widely spaced in time. Consequently, these techniques usually yield accumulation rates that are poorly time averaged. Time-averaged rates are of value in large-scale studies; however, sediment accumulation rates calculated for short intervals are needed to detail the complexities of and interrelations between the structural and depositional histories of sedimentary basins. Despite some exceptions (e-g., Johnson et al., 1988), few studies have examined differential subsidence on a small scale and its influence on the sedimentary record. This paper describes a new method to calculate sediment accumulation rates for thin stratigraphic intervals. The technique is based on alluvial paleosols and the recognition that paleosol maturity is inversely related to sediment accumulation rate. This paper builds upon a study (Bown and Kraus, 1993) in which we used paleosols to estimate the percent of total section time represented by subdivisions of a 650 m composite section in the Willwood Formation of Wyoming. Dating of the Willwood Formation since that study allows us here to calculate rock and sediment accumulation rates and changes in those rates over short time scales. Important sedimentologic events are correlated to the record of accumuiation rates, and the factors that controlled the depositional changes are examined. GEOLOGIC SETTING The Eocene Willwood Formation in the southern part of the Bighorn basin (Fig. 1) contains ancient fluvial deposits with abundant paleosols (e.g., Bown and Kraus, 1987). Deposition of the Willwood Formation was coeval with Laramide structural development of the Bighorn basin, which is part of the Rocky Mountain foreland. The deep, westwardly asymmetric basin is adjacent to mountain ranges cored by Precambrian rocks (Fig. 1). To the north, in the subbasin termed the Clark's Fork basin, the Tertiary structural axis of the basin is overridden by the Beartooth Mountains. To the south, in the study area, the axis lies close to or is overridden by the Oregon basin fault (Stone, 1985; Parker and Jones, 1986). Subsurface data (Stone, 1985) show that this fault dips -30" to the west-southwest. Maximum vertical offset is -6000 m, and there is nearly 5 krn of horizontal overhang at the basement level. Figure 1. Map of Bighorn basin, Wyoming, showing major structural features and location of study area (shaded) in southern part of basin. Oregon basin fault is shown by heavy sawloothed line. GEOLQGY, v. 21, p. 743-746, August 1993 TEMPORAL RESOLUTION IN THE WILLWOOD FORMATION Temporal resolution of a stratigraphic sequence depends on reconstructing the amount of time represented by deposition, erosion, and nondeposition (e.g., Wheeler, 1958). We began time stratigraphic restoration of the Willwood Formation (Bown and Kraus, 1993), constructing a composite stratigraphic section extending from the base to the 650 m level of the formation from numerous measured sections. Willwood Formation paleosols record periods of dep3 osition due to overbank floods and, more important in terms of total 2 time, periods of nondeposition when pedogenesis modified alluvium. Paleosols vary in their degree of development or maturity. 2 200 Maturity increases with longer periods of nondeposition (pedogenesis) and thus slower sediment accumulation. Paleosols have been eroded by channel sand bodies and by mud-rock-filled scours that are found in certain parts of the Willwood Formation. Both swurand-fill features represent times of erosion, nondeposition, and deposition. To account for the total time represented by channel sand I I I I I 0 20 40 60 80 100 bodies and mud-rock-filled scours, we (Bown and Kraus, 1993) carefully substituted paleosols developed on laterally equivalent Percent of Section Time mud rocks for these features. In this way, the composite stratigraphFigure 2. plot of percent of section time vs. metre level for Wiiiwood ic section was converted to a vertical sequence consisting entirely of Formation composhe section. paleosols. Temporal reconstruction of the composite section is based on the relative maturities of the paleosols. We (Bown and Kraus, 1993) recognized seven stages of paleosol maturity and estimated their relative times of development on the basis of lateral relations among the seven stages. For example, laterally tracing paleosols showed that four vertically stacked stage 1 paleosols are laterally (and thus temporally) equivalent to a single stage 3 paleosol. Similarly, a stage 4 paleosol is laterally and temporally equal to two stage 3 paleosols. The composite section was subdivided into 25-m-thick intervals, each of which consists of vertically stacked paleosols. The relative amount of time represented by each interval was estimated from the maturity of the included paleosols. Because the section is a composite, some of the 25-m-thick intervals are represented by several vertical sections measured in different locations. The number of paleosols examined per interval ranges from two to 82, and averages 24. Each paleosol was weighted according to its relative time of development (based on maturity), and those values were averaged for each interval to yield a maturation index. For example, the lowest 25 m interval contains eight paleosols ranging in maturity from stage 4 to stage 6 (very mature). On the basis of the relative times of development for those stages, the maturation index for the interval is 12.0, the highest for any interval in the Willwood Formation. Assuming that the 650 m section represents 100% time, we then estimated the percent of Willwood Formation time represented by each 25 m interval (Fig. 2). For example, dividing 12.0 by the sum of the maturation indices for the entire section (131.7) and multiplying by 100 indicates that the lowest 25 m interval occupied 9% of total Willwood time. We emphasize that the paleosol weights are averaged not only over time but also over space, because the composite section was compiled from a number of different measured sections. The consequences of this spatial averaging are considered in a later section. SEDIMENT ACCUMULATION RATES tion (Fig. 2), 52.8 Ma is a reasonable upper age for the section. A lower limit for the section is based on pollen correlation to the marine sedimentaryrewrd (Wing et al., 1991). Strata in the Fort Union Formation 35 m below its contact with the Wilwood Formation correlate with the NP9-NP10 boundary, which has an interpolated age of -55.7 Ma (see Obradovich, 1988). Because the Fort Union Formation had rapid rates of sediment accumulation relative to Willwood rocks (e.g., Gingerich, 1983), 55.7 Ma is a good approximation of the lower age limit for the composite section. Consequently, the time span calculated for the 650 m composite section is -2.9 m.y. On the basis of a time span of 2.9 m.y. for the entire section and knowledge of the percentage of time represented by each 25 m interval, the rock accumulation rate was calculated and plotted against the midpoint of each interval (Fig. 3). Sediment accumulation rates were then calculated by decompacting the Willwood section following the methods of Baldwin and Butler (1985). On the basis of regional studies, the top of the Willwood Formation was probably buried to -1000 m (e.g., Bown, 1982). Measured sections in different areas of the basin typically average -75% mudstone and 25% sandstone. Those values were used for each 25 m increment, because the lithologies are relatively evenly dispersed. A plot of sediment accumulation rate against metre level shows trends similar to those in the rock accumulation rate plot (Fig. 3). The agreement between the plots probably reflects the fact that the section is only 650 m thick; thus, the difference in burial depth of the top and bottom of the section was insignificant. In addition, the lithology is relatively uniform. The rate of sediment accumulation shows an overall increase upward through the wmposite stratigraphic section (Fig. 3). The value of the curve lies in the numerous fine subdivisions of Willwood Formation time shown. The longest 25 m interval represents 9% of formation time, or -260 ka. The shortest interval spans only 53 ka. Thus, short-term fluctuations in accumulation rate can be identified New data have established lower and upper ages for the cornand include (1) relatively steady, slow increase from 0 to 138 m with posite section and allowed us to quantitatively determine rock and a slight decrease at 113 m; (2) steady accumulation followed by a sediment accumulation rates for each 25 m interval. A tuff at the 634 rapid increase between 188 and 238 m; (3) a return to steady accum level is 52.8 20.3 Ma, on the basis of 40Ar/39Ar dating of sanidine mulation followed by a rapid increase between 338 and 413 m; (4) a (Wing et al., 1991). Because the 625-650 m interval occupies such a decrease in accumulation rate from 413 to 513 m with a disruption at small percentage of the total time represented by the composite sec438 m; and (5) an increase in accumulation rate. 744 GEOLOGY, August 1993 Accumulation Rate (cmlyr) Figure 3. Plot of rock accumulation rate (black diamonds) and sediment (decompacted rock) accumulation rate (open squares) vs. metre level in comwsite section of Wiilwood Formatlon. Column A shows stratigraphlc iocations of tabular (straight Ilnes) and lenticular carbonaceousshales. Column B shows stratigraphlc location of mud-rock-filled scours, and column C shows position of thick and laterally extensive sheet sand-body complex. To some degree, the overall increase in accumulation rate reflects change in location rather than an increase at a single locality. The oldest Willwood Formation exposures are in the eastern part of the study area. Progressively younger deposits are exposed toward the Oregon basin fault, in the direction of sediment thickening (Parker and Jones, 1986). Whereas the overall increase may be an artifact of the space-averaged nature of the data, the smaller scale perturbations are interpreted as actual temporal changes. The smaller changes listed above occur over relatively short stratigraphic intervals, and they generally record change within individual vertical sections with& the composite sectionSEDIMENTOLOGIC EVENTS Sedimentologic events can be correlated with the sediment accumulation rate curve (Fig. 3) to provide a more synthetic view of the early Eocene history of the Willwood Formation and to show how the structural development of the basin influenced depositional patterns. Similarly, major mammalian disappearances and appearances have been correlated to the upward changes in paleosol maturity (Bown and Kraus, 1993). The principal sedimentologic features include carbonaceous units, large sheet sand bodies, and extensive scour-and-fill structures. The 0 to 138 m interval contains unusually abundant drab siltstones and shales that are carbonaceous and contain abundant plantcompression fossils (Wing, 1984). Although carbonaceous units typically constitute only 2% of the Willwood section, they make up as much as 20% of the lowest 100 m (Wing, 1984). The carbonaceous units in this part of the section are generally tabular, and their lateral extent and megafloral content indicate accumulation on broad, poorly drained flood plains (Wing, 1984). Wing (1984) attributed their development to comparatively low rates of basin subsidence and sediment accumulation, consistent with the slow accumulation rates calculated for the 0 to 138 m interval. At about the 300 m level, carbonaceous units change from tabular to dominantly lenticular (Wing, 1984). This change just preceded the sharp increase in accumulation rates that characterizes the 238 to 413 m interval. Wing (1984) concluded that lenticular units formed in abandoned channels and indicate rapid rates of flood-plain aggradation as compared to the tabular units. Lenticular carbonaceous units in the upper part of the section are consistent with that interpretation. The 413-513 m interval is notable for scour-and-MI structures filled predominantly by mud rocks on which immature paleosols developed. The scours are up to tens of metres deep and dissect sediments upon which older, better developed paleosols formed. The mud-rock-filled scours are found principally between the 430 and 530 m levels of the composite section and thus conform closely to the one significant period in Willwood Formation time when sediment accumulation rates declined (Fig. 3). The upper half of the 413 to 513 m interval also has an unusually thick and laterally extensive sand-body complex. This 40-m-thick complex contains four sand bodies separated by a total thickness of only 10 m of mud rocks. Numerous mudstone-filled scours are associated with the sandstone zone. The close stratigraphic spacing of the sand bodies and the associated scours indicate that overbank deposits were intensively eroded and reworked by channel systems (e.g., Bridge and Leeder, 1979).