The amount of silt and clay available to rivers reflects source-terrain composition and weathering and can be a primary control on the form and dynamics of channel networks. Fine sediment also affects the permeability of buried fluvial reservoirs. Despite this significance, there is currently a lack of methods for reconstructing how much fine sediment was transported by ancient rivers. Mud accumulations in sandy river deposits are often interpreted as indicators of variable flow conditions; however, these deposits may present an opportunity to constrain how much fine sediment was transported through ancient rivers. We report results from a series of experiments designed to evaluate how much clay and silt are preserved in sandy riverbed deposits under constant and variable discharge conditions. Our results demonstrate that (1) mud deposits, including drapes and lenses, form readily under constant, high-discharge conditions, (2) the amount of fine sediment recovered from bed-material deposits increases as fine-sediment supply increases, and (3) fine-sediment retention is higher during bed aggradation than during bypass conditions. These results indicate that the net retention of clay and silt in sandy riverbed deposits may be a simple but powerful proxy for comparing the overall amount of fine sediment supplied to ancient rivers.

The amount of fine sediment (silt and clay) in sand-bed rivers significantly influences channel form and movement and the architecture of fluvial deposits at a wide range of scales (e.g., Peakall et al., 2007; Hampson et al., 2014; Ghinassi et al., 2016; Lapôtre et al., 2019; Dunne and Jerolmack, 2020). The ability to interpret the sedimentary archive of fluvial landscape dynamics and predict subsurface reservoir and aquifer quality is currently limited by a lack of constraints on fine-sediment flux to ancient rivers. Estimates of paleo–fine sediment supply would help resolve outstanding questions about, for example, controls on river form and mobility in Earth’s past (e.g., Davies and Gibling, 2011; McMahon and Davies, 2018; Ganti et al., 2019) or how climate-mediated changes in sediment supply, water discharge, or land cover are recorded in fluvial strata (e.g., Foreman et al., 2012; Foreman, 2014; Colombera et al., 2017).

Mechanisms for mud deposition in alluvial channels are varied and still being explored. Based strictly on particle size, silt and clay have slow settling velocities; consequently, mud deposits in sand-bed channels are commonly attributed to periods of slow or stagnant flow (e.g., Martin, 2000). In contrast, significant mud transport and deposition can occur during high-energy, high-concentration flows, which can be common in tidal or highly seasonal channels (e.g., Dalrymple and Choi, 2007; Plink-Björklund, 2015). Flocculation and mud aggregates allow silt and clay to behave like larger particles and interact with the channel bed (e.g., Rust and Nanson, 1989; Lamb et al., 2020), and advective pumping through bedforms can inject fine sediment into bed material (Packman and MacKay, 2003). Large channel-bed features like bars also create locally variable flow conditions, which can promote suspended-sediment deposition (e.g., Szupiany et al., 2012) and enhance bed-deposit preservation (Ganti et al., 2020).

The role these mud transport and deposition processes play in controlling channel kinematics, floodplain aggradation, and sediment mass balance in fluvial systems remains unconstrained. Consequently, it is difficult to uniquely interpret the factors that controlled mud accumulation in ancient fluvial deposits. Conceptually, rivers fed by muddy source areas should carry and deposit a larger proportion of fine sediment than those with mud-poor sources. To test this principle, we conducted a series of experiments to evaluate whether the amount of fine sediment supplied to sandy rivers could be reflected in the amount of mud retained in their deposits. Our experiments were designed to constrain the amount and character of fine-sediment deposits that accumulate under constant, high-flow conditions and provide insight into how mud deposited with channel-bed material might record overall fine-sediment flux or flow intermittency in ancient systems.

We conducted a series of five flume experiments at the St. Anthony Falls Laboratory (SAFL; Minneapolis, Minnesota, USA; Fig. 1; see details in the Supplemental Material1). Water and sediment discharge were set to aggrade a sand bed via a wedge of sediment that prograded down the flume during each run; this is analogous to a bar with superposed bedforms migrating downstream in a river. Total water discharge for each run was 21 L/s (sufficient to transport sand as suspended load; e.g., Wilkerson and Parker, 2011) and was monitored using an acoustic Doppler velocimeter and by measuring the water depth over the weir at the end of the flume. Sand (median grain size D50 = 0.343 mm) and kaolin clay (D50 = 0.004 mm) were supplied to the flume at a constant rate during each run (sand at 15 g/s sand and clay at various concentrations; Table 1). Weir height was fixed, allowing the bed to aggrade ~6 cm in ~4 h, and each run was continued at bypass (i.e., no net bed aggradation) for 15–30 min.

Four runs had constant water discharge and one had intermittent water discharge (Table 1). The four constant-discharge runs had clay concentrations of 0 (control run), 1000, 4000, and 8500 mg/L. For the intermittent-discharge experiment, clay concentration was low (1000 mg/l), and every hour water and sediment discharge were stopped, allowing fine sediment to settle onto the bed for >1 h during each pause. All runs were equivalent to the fully turbulent flows of Baas et al. (2016; see details the in Supplemental Material). Each run was recorded from the side of the flume with a video camera and with photographs. These images were used to reconstruct bed topography and measure bed aggradation, bedform scale, and bedform migration rates in each run.

Fine-sediment mapping and sampling mimicked what could be accomplished in an outcrop. Fine-sediment accumulations were mapped on photographs of the flume wall (analogous to mapping an outcrop photo panel; Fig. 2). After each experiment, the bed was dried for 2 days, excavated, and sampled (analogous to collecting a hand sample of ancient bed material from an outcrop). Samples were collected from bed deposits that accumulated during the aggradational and bypass phases of the experiment and were wet-sieved to determine the fraction of fine sediment.

Fine-sediment accumulations in experimental bed deposits included lenses, drapes, and interstitial fines (Fig. 2; Table 1). Visible mud accumulations were most prominent in deposits from the high-concentration run, with most of the bed showing interstitial fines along with numerous bedform-scale lenses and continuous drapes of fine sediment. Interstitial fines were less noticeable in the intermediate-discharge run, but bed deposits contained mud lenses and some continuous mud drapes. Bed deposits from the low-concentration run contained some fine-sediment drapes. Deposits from the intermittent run contained discontinuous drapes.

The proportion of fines in bed-material deposits increased with higher fine-sediment concentrations (Table 1). For all but the low-concentration constant-discharge run, the average weight percent of fine sediment in a given sample significantly exceeded what would be expected if fine-sediment retention were due only to interstitial fines in the bed (i.e., fine-sediment concentration × bed pore volume). Additionally, the highest fine-sediment retention occurred during aggradational phase of each run (Table 1). Bed-deposit samples from the intermittent-discharge run showed higher mud retention than the constant-discharge run with the same fine-sediment concentration.

Experiments with constant, high-discharge conditions produced deposits similar to those typically considered diagnostic of variable flow (e.g., mud drapes and flaser-like bedding; e.g., Boggs, 2006). This highlights that the presence of drapes and lenses in channel deposits does not uniquely indicate discharge intermittency in ancient rivers. The intermittent-discharge experiment retained more mud than its constant-discharge counterpart, suggesting that flow variability may enhance fine-sediment deposition to some degree, even in low-concentration flows. However, results of these experiments indicate that the overall flux of mud through a system may be the dominant control on the amount of fine sediment deposited in sandy riverbeds.

Mud deposits were most prevalent on the lee sides of individual bedforms (e.g., Fig. 2). This pattern contrasts with that of other experiments where fine sediment accumulated in the bed on the upstream side of dunes through advective pumping and hyporheic exchange (Packman and MacKay, 2003) and is consistent with examples of systems with mud flocs and aggregates that hydrodynamically behave like coarser (e.g., sand-sized) particles (e.g., Rust and Nanson, 1989; Schieber et al., 2007; Matsubara et al., 2015; Mooneyham and Strom, 2018). The degree to which mud aggregates facilitated fine-sediment deposition in these experiments is unresolved; some sand-sized mud aggregates were seen along the glass wall of the flume near the bed of the experiments, confirming their presence in the flume, and overall retention of clay in the bed is consistent with that reported in the data compilation of de Leeuw et al. (2020) and Lamb et al. (2020) (see the Supplemental Material). However, measured suspended-sediment concentrations were constant with depth, a pattern more characteristic of wash-load rather than suspended-load floc transport (Lamb et al., 2020). The presence of flocs in these experiments underscores the potential importance of flocculation in systems not generally considered strongly prone to floc formation (e.g., freshwater settings or rivers with modest clay concentrations), but further investigations will be necessary to determine whether flocs or aggregates drive the deposition of muddy lenses, drapes, and interstitial fines in sandy riverbed deposits.

Mud accumulations were most prevalent among bedforms deposited during the aggradational phase of the experiments, downstream of the sediment wedge. This result is consistent with field data showing silt and clay accumulations in channel beds downstream of bars in modern rivers and ancient deposits (e.g., Lynds and Hajek, 2006; Hajek et al., 2010). The prograding sediment wedge may have enhanced fine-sediment accumulations during the aggradational phase by locally sequestering sand and decreasing the effective sand flux (thereby increasing the relative fines flux) downstream of the wedge. A lower relative sand flux is reflected by observed bedform-migration rates that were ~8× slower during the aggradation phase (1.1–1.8 cm/s) than the bypass phase (8.6–12.0 cm/s) even though total supplied sediment flux was constant. This slower migration rate could have permitted more fine sediment to settle in the recirculation zone downstream of bedforms (see the Supplemental Material). Preservation during the aggradational phase of the experiments was likely enhanced by an abrupt increase in local aggradation as the sediment wedge passed through the flume. In these experiments and field-scale systems, bar migration can rapidly bury slower-moving bedforms, thereby preserving them entirely (e.g., well preserved cross sets in Fig. 2; Reesink et al., 2015; Ganti et al., 2020).

Overall, when more fine sediment was added to the flume, more fine sediment was incorporated into bed deposits, suggesting that the bulk fraction of fine sediment preserved in ancient bed-material deposits may reflect the amount of fine sediment supplied to an ancient river. While progress has been made quantifying paleo–bedload transport in ancient rivers (e.g., Brewer et al., 2020), it remains difficult to reconstruct the fine-sediment flux. The possibility of comparing, even in a relative sense, the proportion of fine sediment moving through ancient rivers provides an important opportunity to attempt complete mass-balance estimates for ancient source-to-sink systems and understand controls on fine-sediment storage and bypass in sedimentary basins (e.g., Chamberlin and Hajek, 2019).

More work is needed to determine how to quantify fine-sediment flux from ancient fluvial deposits and to understand the relative contributions of flow intermittency, flocculation, and other processes that drive mud deposition. However, in the near term, these results indicate that the amount of mud preserved in bed-material deposits (e.g., cross sets from channel sandstones) may provide a benchmark for normalizing and comparing fine-sediment storage at larger scales. Bed-material samples from channel sand bodies spanning documented alluvial architecture transitions could help determine whether and how the fraction of fine sediment in ancient rivers changed along with trends in, for example, channel-body dimensions, floodplain facies, and the overall proportion of channel sediments preserved at different places and times within a basin (e.g., Foreman, 2014; Hampson, 2016; Chamberlin and Hajek, 2019; Wang and Plink-Björklund, 2019).

Relative comparisons of paleo–fine sediment flux may help answer outstanding questions about changes in hillslope weathering or the role of cohesive sediment in controlling river dynamics through Earth’s history (e.g., Foreman et al., 2012; Ielpi and Lapôtre, 2020). Furthermore, constraining the fraction of fines present in bed-material deposits will be helpful for more accurately predicting heterogeneity and compartmentalization in fluvial reservoirs. Measuring the fraction of fines in ancient bed-material deposits offers a tractable, potentially powerful approach to reconstructing and comparing paleo–fine sediment loads through Earth’s history.

These experiments demonstrate that the proportion of fine sediment trapped in sandy riverbed material can reflect the concentration of clay and silt available in the flow. While discharge intermittency may enhance mud deposition for a given fine-sediment flux, our results show that the amount of mud hosted in riverbed deposits may reflect primarily the total supplied fine-sediment load rather than variable discharge. These results highlight a need for more targeted studies aimed at constraining the role of flocculation and local sorting in mixed sand-mud systems and improving our understanding of how interactions of bedforms and larger features like bars influence fine-sediment deposition and preservation in ancient deposits. Measuring mud fractions preserved in riverbed deposits can provide an important avenue for reconstructing the relative abundance of fine sediment transported in ancient channel networks.

This research was supported by the donors of the American Chemical Society Petroleum Research Fund, U.S. National Science Foundation awards 1455240 and 1935513 to Hajek, and student support from the Geological Society of America, American Association of Petroleum Geologists, and Pennsylvania State University Department of Geosciences to Wysocki. Without the incredible expertise and generosity of St. Anthony Falls Laboratory (Minneapolis, Minnesota, USA) personnel, particularly Ben Erickson and Sara Mielke, this work would not have been possible. We are grateful to E. Chamberlin and Macalester College (St. Paul, Minnesota) students for helping run experiments. We thank C. Paola, R. Slingerland, T. Bralower, R. DiBiase, V. Ganti, S. Alpheus, E. Chamberlin, E. Greenberg, X. Hu, S. Lyster, S. Trampush, and J. Walker for helpful discussions; and A. Fernandes, A. Ielpi, M. Perillo, J. Pizzuto, P. Plink-Björklund, J. Shaw, and an anonymous reviewer for thoughtful and constructive reviews.

1Supplemental Material. Details of experimental conditions and analyses. Please visit to access the supplemental material, and contact with any questions.
Gold Open Access: This paper is published under the terms of the CC-BY license.