Accretion of Mudstone Beds from Migrating Floccule Ripples

Authors: Juergen Schieber,1* John Southard,2 Kevin Thaisen,1

Mudstones make up the majority of the geological record. However, it is difficult to reconstruct the complex processes of mud deposition in the laboratory, such as the clumping of particles into floccules. Using flume experiments, we have investigated the bedload transport and deposition of clay floccules and find that this occurs at flow velocities that transport and deposit sand. Deposition-prone floccules form over a wide range of experimental conditions, which suggests an underlying universal process. Floccule ripples develop into low-angle foresets and mud beds that appear laminated after postdepositional compaction, but the layers retain signs of floccule ripple bedding that would be detectable in the rock record. Because mudstones were long thought to record low-energy conditions of offshore and deeper water environments, our results call for reevaluation of published interpretations of ancient mudstone successions and derived paleoceanographic conditions.

1. Department of Geological Sciences, Indiana University, Bloomington, IN 47405, USA.
2. Department of Earth and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, MA 02139, USA. * To whom correspondence should be addressed. E-mail:

A century ago, Henry Clifton Sorby, one of the pioneers of geology, pointed to the study of muds as one of the most challenging topics in sedimentary geology (1). Today, with our knowledge clearly expanded, muddy sediments are still considered highly complex systems that may require as many as 32 variables and parameters for a satisfactory physicochemical characterization (2). More research may clarify interdependencies between a number of these parameters and may allow us to consider a smaller number of variables, but the fundamental complexity of muddy sediments is likely to remain. A key issue in mudstone sedimentation is flocculation, a phenomenon in which a number of these parameters, such as settling velocity, floccule size, grain-size distribution, ion exchange behavior, and organic content "come together." A joining of smaller particles to form larger aggregates, flocculation enhances the deposition rate of fine-grained sediments, and its understanding is critical for modeling the behavior of mud in sedimentary environments.

Flocculation is affected by particle concentration within the fluid and intensity of turbulence (3, 4). Over time, floccules enlarge to a maximum equilibrium diameter that is related to the intensity of turbulence (5). Floccule deposition is influenced by turbulence, bed shear stress, sediment concentration, and settling velocity. We currently still miss critical data on floccule formation and on the influence of floccule structure and turbulence on the formation of muddy sediments (6). We will collect data concerning these issues with new instrumentation in the near future, although the importance of our observations will not be affected. The notion is widely held that slow-moving currents or still water are a prerequisite for substantial mud deposition (7, 8) because shear stress in swift-moving currents disrupts previously formed fragile floccules and prevents their deposition, but our observations suggest an alternative mode of mud deposition that apparently left its imprint in the rock record.

Mudstones constitute up to two-thirds of the sedimentary record and are arguably the most poorly understood type of sedimentary rocks (9). Mudstone successions contain a wealth of sedimentary features that provide information about depositional conditions and sedimentary history (10–13), but presently we lack the information that would allow us to link features observed in the rock record to measurable sets of physical variables in modern environments.

Although various small-scale sedimentary structures have been described from modern muds, these have not been observed in the making. This forces us to infer controlling parameters (e.g., current velocity and density of suspension) from temporally and spatially very limited measurements in the overlying water column (14–17). Such measurements (e.g., flow velocity, sediment concentration) in modern environments are commonly considered representative of depositional conditions for the uppermost millimeters to decimeters of the accumulating deposits. However, upon close examination, modern sediments show considerable heterogeneity at the millimeter to centimeter scale (16), an indication that what we observe in surficial sediments is not a direct response to measured conditions in the overlying water column. To improve on this situation, it is essential to conduct experiments that replicate natural conditions and to compare the experimental sediments to the rock record.

Here, we report experimental insights into the sedimentology of mudstones. In past experimental studies, centrifugal pumps were used to recirculate mud suspensions (18–20), destroying the clay floccules that are of such key importance in mud transport and deposition. Therefore, to minimize the risk of shredding clay floccules once formed, we built a racetrack flume that uses a paddle belt (21) for moving the mud suspension.

The racetrack flume (fig. S1) used for these experiments (21) has a 25-cm-wide channel. The effective flow depth was 5 cm. Powdered kaolinite clay (Fig. 1A) was mixed with water and added into the flume, running at 50 cm/s velocity (21). Sediment concentrations ranging from 0.03 g/l to 2 g/l were explored, and suspended sediment concentrations were monitored with an optical turbidity sensor. Experiments were conducted in distilled water, fresh (tap) water, and salt water (35 per mil salinity). In a few experiments, Camontmorillonite and natural lake mud (sieved to 63 µm) was used.

Fig. 1. Flume feed and floccules. (A) SEM image of typical kaolinite clay used in the majority of experiments. Inset shows size and morphology of clay flakes. (B) SEM image of floccules (pointed out by arrows) that were trapped at the flume bottom with a grooved glass slide. (C) Close-up of floccules in (B). (D) SEM image of a floccule (outline marked with white arrows). This floccule measures 0.12 mm along the long axis. [View Larger Version of this Image (210K GIF file)]

Addition of clay to the flume resulted within minutes in the formation of "floccule streamers" that mark boundary-layer streaks (22). Floccules range in size from 0.1 mm to almost 1 mm (Fig. 1, B to D) and were sampled and examined with a scanning electron microscope (SEM). After establishment of a stable suspended clay concentration, the velocity was stepwise reduced (Fig. 2) until the critical velocity of sedimentation was reached (23). At that point, a linear decline of sediment concentration was observed and essentially all sediment settled out of the flow (Fig. 2). By shining strong lights from above through the flow, we were able to photograph and film floccule streamers (Fig. 3A), individual migrating floccules, floccule ripples (Fig. 3B), and fields of floccule ripples (Fig. 3C).

Fig. 2. Example of changes in suspended sediment in the course of an experiment. Vertical axis shows continuously logged suspended sediment concentration, and horizontal axis shows time elapsed (tick marks separate successive days). The critical velocity of sedimentation lies between 25 cm/s and 20 cm/s and coincides with the onset of development of floccule ripples. Its exact determination requires the use of smaller velocity steps. [View Larger Version of this Image (70K GIF file)]

Fig. 3. (A) Floccule streamers photographed through the flume bottom (flow velocity 48 cm/s, initial sediment concentration 0.72 g/liter, lake mud in tap water). (B) Streamlined floccule ripple and floccule streamers. (C) Multiple floccule ripples begin to coalesce. Flow is from top to bottom in images (A), (B), and (C). (D) Oblique view of large barchan-shaped migrating floccule ripple. Flow is toward the right. (E) Side view of the left horn of this ripple [pointed out by black arrow in (D)]. Inset shows that the surface is covered with small, round bodies, the clay floccules that moved across the ripple during current activity. Numbers indicate floccule size. [View Larger Version of this Image (196K GIF file)]

Floccules that give rise to "floccule streamers" (Fig. 3A) form even at small sediment concentrations (0.03 g/l), in both distilled water and fresh water, and increase in abundance as velocity is lowered. Below the critical velocity of sedimentation (Fig. 2), patches of floccules form and organize into streamlined ripples that migrate slowly downcurrent (Fig. 3, B to D). The critical velocity for sedimentation depends on initial sediment concentrations and ranges from 10 cm/s for small sediment concentrations (0.03 g/l) to at least 26 cm/s for sediment concentrations in the 1 to 2 g/l range. In several experiments where flow was stopped suddenly and water was drained and replaced with clear water, floccules were observed on the foreset slopes of floccule ripples (Fig. 3E). We also conducted experiments in which we introduced multiple sediment pulses and allowed the clay to accumulate at the bottom before adding the next pulse (21). A small amount of pulverized hematite (a red powder) was added between clay pulses to mark the tops of successive clay pulses. This addition of clay increments and hematite spikes was repeated until a sediment layer of approximately 2 cm (uncompacted thickness) had accumulated. At the end of the flume run, draining of the water typically revealed that the mud bed carried at its surface elongated ripples that stood up to 3 cm above the flume bottom and were spaced between 30 and 40 cm apart in the downstream direction (Fig. 3D and fig. S2). These experimental mud layers were air-dried to the consistency of soft butter and scraped with a butter knife or spatula to reveal internal layering outlined by hematite drapes. These internal layers were inclined in the downcurrent direction (fig. S3), indicating lateral accretion of clays. Once the clay beds have dried completely, these internal laminae appear to layer parallel on surfaces perpendicular to bedding (Fig. 4A). Drying out, however, also leads to separation of the top portions of layers along bottom-parallel fractures and reveals that the overall deposit is characterized by low-angle, downcurrent-dipping cross-strata (Fig. 4B).

Fig. 4. (A) Laminated flume sediment (between white arrows) that was deposited during an experiment with continuous current flow. Sample was embedded in epoxy and cut perpendicular to bedding. Sample is curved due to desiccation. The darker internal laminae are hematite markers. (B) Textural detail from the interior of the clay layer in (A) (top view, arrow indicates 90° twist). As the layer dried out, its upper portion formed concave desiccation polygons, whereas its lower portion remained attached to the flume bottom. Removing the upper portion exposes a bottom-parallel surface through the clay layer. The curved lines are the upper terminations of broken foreset laminae of floccule ripples. The foreset laminae are inclined to the right, and the circled numbers indicate a succession of overriding ripples (see also fig. S4). The overall texture resembles "rib and furrow" structures as known from current rippled sandstones (24). (C) Parallel-laminated black shale, New Albany Shale, Devonian, Indiana. Lighter laminae are silt-enriched. (D) Cross-laminated shale collected from the same core box as (C). (E) Tracing of silt laminae visible in (D). Arrow marks an internal erosion surface. In the center are inclined (to the left) truncated laminae, forming the outline of a compacted and mud-dominated ripple. The synoptic relief of this ripple is 3 mm, but its original relief would have been 20 mm (assuming 85% water content), the same magnitude as observed in our experiments. [View Larger Version of this Image (165K GIF file)]

These observations show that ripples composed of clay floccules migrated over the flume bottom at the onset of deposition (Fig. 3) and that a rippled bed topography was present at the end of deposition (fig. S2). In addition, the texture produced by the low-angle, downcurrent-dipping cross-strata in Fig. 4B has a direct textural analog in sandstones, where it is known as "rib and furrow" structure (24). The latter is seen on horizontal surfaces cut through sandstone beds that accumulated from migrating ripples. Closer inspection of the surface exposed in Fig. 4B shows the deposits of multiple overriding floccule ripples (fig. S4). Under the microscope, the inclined clay laminae from Fig. 4B show a "bumpy" surface pattern of closely packed ovoid bodies (0.2 to 0.6 mm in size) (fig. S5), the compacted floccules from which the migrating ripples and the accreting clay bed were constructed.

It appears that irrespective of what drives flocculation in a given experiment, flocculation provides deposition-prone particles without fail over a wide range of experimental conditions. Formation of floccule ripples from a variety of clay-size materials (kaolinite, montmorillonite, and lake mud), and over a range of sediment concentrations and salinities (distilled, fresh, and salt water) strongly suggests a universal process at work.

Our observations do not support the notion that muds can only be deposited in quiet environments with only intermittent weak currents (8). Instead, bedload transport of flocculated mud and deposition occurs at current velocities that would also transport and deposit sand (21). Clay beds can accrete from migrating floccule ripples under swiftly moving currents in the 10 cm/s to 26 cm/s velocity range, a range likely to expand as flows with larger sediment concentrations are explored. Whereas the clay beds formed in our experiments consist of downcurrent-inclined laminae, they appear to be parallel-laminated once fully compacted (Fig. 4A). Because floccule ripples are spaced 30 to 40 cm apart, ancient sediments of this origin are likely to appear parallel-laminated (Fig. 4C) as well. Detection of ripple-accreted muds in the rock record will require carefully defined, and yet to be developed, criteria. Things to look for might be subtle nonparallel lamina geometry, as well as basal downlap and top truncation of laminae. Examination of ancient shale units may, for example, yield low-amplitude bedforms (Fig. 4, D and E) as indicators of lateral accretion and ripple migration. Bedding-oblique orientation of larger flat particles, such as spores, microfossils, plant debris, and mica flakes, could be another indicator. If such particles are deposited on the inclined foresets of floccule ripples, they may record the gentle inclination of the latter even when compaction has rendered the depositional fabric of inclined laminae (Fig. 4B) unrecognizable.

In the course of two decades of detailed studies of shales and mudstones, one of us (25–27) has seen comparable low-amplitude bedforms (Fig. 4D) in shale units that were deposited in a wide variety of environments. Examples can be found in the Mid-Proterozoic Belt Supergroup, the Devonian of the eastern United States, the Jurassic Posidonia Shale, the Cretaceous Mancos Shale, and the Eocene Green River Formation. This suggests that mud accretion from migrating floccule ripples probably occurred throughout geologic history. Many ancient shale units, once examined carefully, may thus reveal that they accumulated in the manner illustrated here, rather than having largely settled from slow-moving or still suspensions. This, in turn, will most likely necessitate the reevaluation of the sedimentary history of large portions of the geologic record.

Elucidating the mechanisms of mudstone deposition not only helps to better understand the rock record but also benefits hydrocarbon exploration, hydrogeology, and coastal and shelf engineering. Managing mud is important for the maintenance of harbors, shipping lanes, and water reservoirs, especially given the impact of climate change. How mudstones act as barriers to fluid migration (oil and water) is probably linked to depositional processes that affect mud microfabrics. For example, if a mud accumulated from current-transported floccules, one might expect a network of larger pores, poorer sealing capacity, and easier release of liquid and gaseous hydrocarbons. Conversely, accumulation in still water from dispersed clays and low-density floccules should lower permeability and may produce an oil shale that clings tightly to its generated hydrocarbons. These qualities are also critical for the ability of a mudstone unit to protect aquifers from contamination and to compartmentalize groundwater reservoirs.


Around Sydney Australia is the Hawkesbury sandstone. This is used to build many of the historic buidings in Sydney. Many geologists have marvelled at the widespread thick, pure deposits with their cross bedding. A Dr. Patrick Conaghan of Sydney's Macquarie University believes that the sandstone was indeed formed by a succession of catastrophic massive flood waves possibly 20 meters high and up to 250 kilometers wide. This is massive wave unlike anything we see today. In this case the old assumption "the present is the key to the past" is totally useless and inacurate. At last the creationist and the evolutionists agree on something, the formation of the sandstone. They both agree it was formed by huge catastrophic processes. The only disagreement between the two partys is when it happened. Many creationists would say it happened during the global catastrophe of Noah which was thousands of years ago. The bible gives hints of three big events associated with the catastrophe of Noah. This catastrophe involved 40 days of continuous rain, volcanoes all over the earth erupting, (the biblical "fountains of the deep"), and meteors and comets striking the earth ("windows of heaven were opened"). That is what caused the huge waves say the creationists. The typical evolutionists would say it happened 237 million years ago over a time period of millions of years. Sydney Morning Herald of April 30, 1994.


Russian Journal of Lithological and Mineral Resources Publishes Groundbreaking Research on Sedimentary Rocks: French sedimentologist leads the field

MOSCOW--(BUSINESS WIRE)--The latest edition of Lithological and Mineral Resources, a journal of The Russian Academy of Sciences, has reported details of research directed by French sedimentologist Guy Berthault of the prestigious university, Ecole Polytechnique in Paris, showing that sedimentary rocks form very rapidly - two thousandths of the time attributed to them by the geological time scale!

The experimental research spanning a period of thirty-five years was first performed in France at the Marseilles Institute of Fluid Mechanics and subsequently at the Colorado State University hydraulics Laboratory in the USA. Its application in the field was tested on the Cambrian-Ordovician sandstones of the North-West Russian Platform by a team of Russian sedimentologists led by Dr. A. V. Lalomov of the Russian Academy of Sciences' Institute of Geology of Ore Deposits, Petrography, Mineralogy and Geochemistry. Similar results were obtained from other geological formations in Russia.

For details of the experiments and research, see the current issue of the Lithological and Mineral Resources journal (2011, volume 1) and Contacts: Peter Wilders, 377-935-08834; 336-600-56971 (cell)