Rivers have never been as diverse as they are today, with human activities influencing their morphologies and contents. Catherine Russell discusses recent research that founds a crucial new sub-branch of sedimentology: plastic-sediment interactions
We find plastic particles everywhere that we look, in the deepest ocean trenches, at the top of the highest mountains, even in rainwater and our own bodies. Current management strategies for plastic waste are insufficient. For example, ‘plastic overshoot day’, the day when waste management sites and facilities reach capacity for the annual volume of waste they can hold, comes earlier each year. In 2023, plastic overshoot day was 28 July, which means that the equivalent of all plastic waste produced from then until 31 December is likely to be mismanaged and end up in the environment. Indeed, there is so much plastic in the environment, that it is forming deposits that someday may become future geology.
Geoscience is a relatively young discipline among the sciences, with modern geology often said to have been founded on the principle of uniformitarianism – the idea that changes on Earth throughout geological history have resulted from continuous,uniform processes, or, put simply, that the present is the key to the past. However, as we continue to uncover uncomfortable novelties in the ways humankind has reshaped, and continues to reshape, Earth’s systems, it seems that uniformitarianism is becoming a limited principle.
Rivers are more diverse today than at any point in Earth’s history. In addition to the natural river morphologies (braided, meandering, anastomosing), river channels have been modified by humans; they have been channelised, dammed, diverted, made subterranean, and micro-divided for irrigation. The contents of our rivers are just as diverse in what they carry, and the deposits they create.
Sedimentary deposits and the dynamics of riverbeds are broadly understood, with lowland riverbeds transporting sediment downstream via well-defined natural processes (see box ‘The natural rhythm of riverbeds’, below). When plastics reach the environment, they fall under the whim of these natural processes and, like sediment, they may be readily channelled along rivers to eventually reach the ocean. Yet, the impact of plastic on the mechanics of sediment transport and broader river system remains largely unknown.
Much of our current understanding of plastic in river systems is confined to shallow or surface processes, simply because the plastic that floats is easier to see and access compared with that which lies on or travels along the riverbed. We know that plastic exists on riverbeds from studies that scoop, net, and core samples, however the current central principles for how plastic interacts with sediment are limited to the idea that plastic either does or does not become buried. Of course, the simple premise of burial versus no burial is critical for determining whether a channel provides either a conduit, or storage site, for pollution. However, this simplistic view assumes that plastic is merely a passive component in the system.
Working with a team from universities in the UK and USA, we have shown that plastic influences the volume of sediment in the water column, affects the topography of the riverbed, and becomes heterogeneously distributed within the riverbed (Russell et al., 2023). Our results not only define a new sub-branch of sedimentology – plastic-sediment interactions – but also have significant implications for understanding how plastic influences the landscape and river ecosystems, as well as for determining the volumes of plastic pollution present in our environments.
The natural rhythm of riverbeds
A riverbed is an environment constantly in motion, on which sediment, organic debris, and pebbles transit downstream. Patterns in the flow create riverbed mounds, or bedforms, such as ripples and dunes. A bedform has a shallower upstream side (the stoss side), and a steeper downstream side (the lee side), and the highest point on a bedform, where the stoss and lee sides meet, is known as the crest.
Downstream of the crest, the flow of water separates, such that some continues downstream and some circles backwards to the lee slope, which is known as the recirculation zone. On a riverbed, the flow of water over the bedform causes sediment to become eroded from the stoss side, and to be deposited on the lee side, such that the bedforms will translate downstream and sustain equilibrium if no disruptions occur.
With differing flow speed or flow depth, bedform behaviour and migration can change. For naturally occurring sediments, such as sand, this behaviour is broadly well-known, having been extensively documented in specialised laboratories – a necessity because the sediment and flows that govern bedform morphology are sensitive to disturbances, so a person or camera on a riverbed would itself cause the processes to change, invalidating the findings. However, for mixtures of sediment and human-made particles like plastic, the changes in and controls on bedform behaviour are more complex.
The novelty of plastic
Plastic can have any combination of properties imaginable. It is a vastly abundant human-made component of the modern environment. As we seek to monitor and remove it from the environment, we need to understand how it is transported and deposited. While some components of plastic particles may be relatable to other naturally occurring elements, it is the combination of plastic properties that make this material unique. For example, like organic particles, plastic can have a low density compared to most sediment particles. However, organic particles will oxidise and break down in the environment far more readily than most plastics. The endurance of plastic means that it can accumulate in significant volumes and perhaps be reworked for many centuries to come. Plastic is a human-made component of present and future geology, indelibly part of our environment for the foreseeable future.
The stoss-side surprise
In 2019, using flume tank experiments, I led a study to investigate the interactions between plastic and sand in a sandy bedform, with the goal of seeing what would happen to the deposit when plastic was incorporated; we did not anticipate the findings.
A range of plastic particle types were selected for the experiment, all with different properties, and mixed into the sand. We put the sand-plastic mixture into a large tilting flume tank (a long tub, 10 m long by 0.5 m wide, with clear sides and a pump) that circulates water and sediment around in a continuous loop to truthfully mimic a river. With this set up, we could control and monitor the conditions of the flow, such that we could accurately observe and record our findings. We accelerated the flow for two hours to simulate flood conditions, then slowed it to 0.5 m/s, leaving the flume tank with this set-up overnight to establish bedforms.
We found that the plastic particles in the flow were deposited on the lee side of the bedforms either one at a time, forming an isolated particle within the deposit, or in groups, to form part of a lens of plastic-rich sand (Fig. 1). The shape of the plastic particles influenced how they were incorporated into the deposit. For example, elongate particles aligned with the flow direction.
Generally, the depositional observations were expected because they are similar to what we see with organic particles, however, the combination of behaviours that we observed on the stoss side of the bedform were a surprise. The presence of plastic particles enhanced the removal of sand, such that plastic inclusion was locally increasing the amount of sand being eroded from the bedform! We suggest that three key properties of the plastic particles led to this behaviour:
- The large size of the plastic particles (relative to the sand particles) created a ‘flow shadow’ directly downstream from the plastic particle, ‘hiding’ and protecting other sediment from the flow, but also obstructed the flow directly upstream of the plastic particle, enhancing scouring there.
- The smooth plastic particles have low roughness, meaning that once a part of a plastic particle had been exposed, the lack of friction to retain sand grains on its surface led to rapid exposure of the plastic to the flow.
- The lower density of the plastic particles compared to the sand grains meant that the plastic had a lower erosion threshold and was readily entrained back into the flow of water.
While each of these mechanisms is known about individually, the vast heterogeneity of plastic properties and its uniqueness in the natural system causes these mechanisms to combine and interact in novel ways, creating abundant complexities that are new to the field of sedimentology. Even if only a uniform plastic particle type is introduced to the system, when in a group, the plastic particles have a compelling impact on the flow mechanics, wherein different forcings dominate depending on a wide range of factors.
When plastic particles were exposed or removed from the bedform, we observed an increase in the scouring of sand, first as the plastic particle was eroded and secondly as the bedform adjusted. The exposure or removal of the plastic disrupts the bedform and it must adjust to find a new equilibrium. The readjustment may result in bedforms with lower sand volume, lower amplitude, and more rounded crests, or the bedform could experience continued disruption and erode away entirely.
The processes that plastic induces on the stoss side of the bedform lead to bursts of plastic and sediment travelling downstream. Additionally, despite our initial mixing of the plastic into the sand, we found that much of the plastic became concentrated in the upper active layer of the fluvial bed, meaning that even after running the experiment for many hours, the plastic particles remained present and active in and around the bedforms.
The resulting polluted sedimentary deposit formed during our experiments was remarkably heterogenous, with both extensive lateral and vertical variability (Fig. 2). Some parts of the deposit were free from plastic, whereas others, particularly the lenses, had a higher plastic-to-sand ratio. Each of these plastic-rich zones was delineated by a depositional surface that marked a change in bedform process or particle availability. Importantly, the areas with highest plastic concentration are not on the immediate riverbed surface; they occur as lenses beneath the crest of the bedform.
Imagine these deposits scaled up in a natural river. With environmental monitoring in progress, a sample may be acquired by grabbing a surface sample or coring the riverbed. Yet, our results show that the surface of the riverbed does not contain the highest concentrations of plastic pollution, while lateral variability suggests that an individual vertical core may miss the pollution completely.
We may have been underestimating the volumes of plastic present in the riverbeds themselves, and therefore underestimating the pollution storage capacity for river channels
To target parts of a riverbed with the highest concentrations of pollution, and therefore obtain more accurate representations of the volumes of plastic present, a more appropriate methodology may assist. With the goal to target the areas where the plastic-rich lenses are most likely to be present, an appropriate method could be to always core to at least the depth of a typical bedform height, directly beneath the crest of a bedform.
One implication of our findings is that we may have been underestimating the volumes of plastic present in the riverbeds themselves, and therefore underestimating the pollution storage capacity for river channels. It is also possible that we have neglected to understand the processes by which plastic enhances the local rate of sand and plastic transport on riverbeds, and are therefore underestimating the capacity of river systems to be effective conduits for pollution. We simply do not know yet, and more studies are required.
Clearly, plastic is not passive in riverbeds. Rather, it locally enhances the erosion rate of sand and the plastic itself, with the potential to influence fluvial landscapes, environments, and ecosystems, while our observations of extreme heterogeneity in the plastic-sand deposits raise important questions about the techniques we use to sample pollution in riverbeds.
Pollution from plastics may be increasing the slope of the riverbed, or even flattening the riverbed, in turn leading to changes in sedimentation. As such, plastic particles in riverbeds could be causing landscape-scale changes that we have yet to detect, or perhaps they are causing local impacts that leave landscapes and biota largely unchanged.
The changes in bedform morphology that we observed were all under constant flow conditions, showing that it is the inclusion of plastic particles that results in this remarkable diversity of bedform shapes and behaviours. Sedimentology is founded on our understanding of what happens when two grains interact; now, with plastic, we are overwhelmed with possibilities and we need to do a lot more research.
This study represents the first steps in an emerging and necessary sub-discipline of sedimentology that seeks to enhance our understanding of the relationship of plastic particle transport with natural systems. The research is ongoing and sits at the threshold of a fascinating spectrum of cross-cultural, cross-disciplinary, and cross-societal questions such as: should we simply consider plastic to be a sediment itself? Such questions are a new and critical part of our physical and philosophical environment, set to be ever-more prominent as we look to the future of Earth and its deposits.
Dr Catherine Russell
Fulbright-Lloyd’s of London Scholar, Researcher in Sedimentology and the Anthropocene at the University of Leicester, UK, and Louisiana State University, USA.
(open to conversation and collaboration)
X (Twitter) & bsky – @SeddyRocks
Russell, C.E., Fernández, R., Parsons, D.R. et al. Plastic pollution in riverbeds fundamentally affects natural sand transport processes. Commun Earth Environ 4, 255 (2023); https://doi.org/10.1038/s43247-023-00820-7