Introduction

The world’s coasts and deltas are increasingly influenced by human interventions. Deltas are subsiding at alarming rates1,2,3, the influx of fluvial sediments is reduced because of upstream dams4,5 or sand mining6,7, and the alongshore drift is interrupted by coastal structures8,9. Superimposed on these local interventions are global climate change effects, with sea level rise (SLR) progressively threatening coastlines worldwide10,11,12. In the coming century, both human interventions and climate change will put the livelihoods of millions of people at risk13. Deltas and coastal environments are dynamic systems in which erosion, transport, and deposition of sediments provide a delicate balance. The degree to which coastlines will be able to adapt to future changes, especially in view of SLR, strongly depends on modifications to coastal sediment budgets.

Coastal systems display a wide variety of types, ranging from exposed high-energy beaches to estuaries and back-barrier lagoons. Sediment budgets along sandy open-coast beaches are strongly driven by the wave climate and adapt relatively quickly to SLR or human interventions14,15. Using empirical formulations16,17,18, the sediment fluxes along sandy beaches are relatively well known, and their adaptation to SLR can be reasonably predicted when assuming the coastal profile to remain constant12 (although such a simplification is debated19). Sediment transport in estuaries, tidal basins and lagoon systems is primarily driven by tidal transport of both mud and sand. Adaptation of such relatively sheltered systems to SLR requires enhanced sediment trapping. Predicting this enhanced trapping introduces two challenges: one is related to the delicate balance in erosional and depositional processes within the basins, and the other to the availability of sediments20,21. In many of such basins, the availability of sand will be a limiting factor, but this may be (partly) compensated with fine-grained sediments (i.e., mud)22,23,24,25. In contrast to sand-dominated systems, the alongshore fluxes of fine-grained sediments are poorly known. We aim to develop such a fine-grained sediment flux and highlight the importance of knowing it for the world’s largest uninterrupted system of barrier islands and tidal flats—the Wadden Sea.

The Wadden Sea spans a distance of nearly 500 km along the coastlines of the Netherlands, Germany and Denmark in the Northwestern European shelf (Fig. 1). As a protected UNESCO world heritage site, it provides crucial habitats for numerous species of fish, mammals and birds. It also plays an important role as a sink for CO2, in particular by blue carbon storage in its sediments, seagrass and salt marshes26. Water motion in the Wadden Sea basins is influenced by tides (with tidal ranges from about 1.5 to over 4 metres), offshore waves generated in the North Sea that penetrate through the inlets and smaller locally-generated wind waves (Fig. 1d). Besides, wind-driven currents play an important role in the shallow parts of the basins27,28. The main freshwater sources enter the Wadden Sea through the IJssel Lake sluices (average yearly discharge of ~ 450 m3/s29) and the Ems (~100 m3/s30), Weser (~300 m3/s31) and Elbe (~750 m3/s30) estuaries. SLR has so far been limited, with rates of 1.2–2.3 mm/year in the past century32 (Fig. 1d). Accelerated SLR and subsidence, however, threaten the Wadden Sea’s existence: if the sediment import cannot keep up with increased relative SLR rates, this will result in partial loss or even disappearance of the ecologically valuable intertidal areas33,34.

Fig. 1: Overview map of the Wadden Sea.
figure 1

a Current bathymetry (combined maps of 2015–2021 in m to NAP, Netherlands Ordnance Datum). The white dotted lines indicate the tidal range105. b Bed sediment composition plotted as the mud content in the upper bed (top 4–10 cm). The Danish part of the Wadden Sea is not included in the plot because there is no available data on this area. c Location of the Wadden Sea. d Example of a wave climate on an ebb-tidal delta and in a Wadden Sea basin (both calculated over the period 2015–2018), and average yearly mean sea level (MSL) from tide gauge records in the Dutch basins over the past century.

The sediment bed of the Wadden Sea, as that of many other coasts and deltas (e.g., Yangtze Delta, Mekong Delta, Nakdong Estuary, San Francisco Bay), consists of sand and mud. Despite decades of scientific attention to sandy alongshore sediment budgets35,36,37 and sand-mud budgets of fluvial systems38,39,40, marine mud budgets remain largely unknown. This is probably related to the inherent difficulties with obtaining such a budget. Transport of mud cannot be predicted with empirical formulae because its transport is supply-limited and therefore needs to be based on observed bathymetric changes or detailed measurements of hydrodynamics and sediment concentrations. The Wadden Sea provides herein a globally unique case, with detailed observations of bathymetric changes and grain size distribution41,42,43,44,45,46,47. Such observations can be used to quantify the contribution of sand and mud to infilling48, thereby providing a sediment budget.

In this study, we expand on previous work by developing a complete mud budget for the entire Wadden Sea by carefully quantifying sources and sinks. We synthesise earlier studies to obtain the best estimate for the sources while we compute sinks using extensive bathymetric and grain size distribution data. By demonstrating that this mud budget is nearly closed, we show that mud is, in fact, a commodity that may become a limiting factor under conditions of accelerated SLR.

Results and discussion

Mud sources and sinks

To establish the mud budget, we determine the magnitude of the main sources and sinks, focusing on inorganic particulate matter. The largest mud source is the marine North Sea Continental Flow (NSCF). This flow transports mud from the Dover Strait (in between France and the United Kingdom) along the French, Belgian and Netherlands coasts. Its magnitude at the western end of the Wadden Sea is on average, approximately 10–14.4 million (M) ton/year (see the Methods section). The sediment plume first deflects to the East, after which it turns northward again in the German Bight while interacting with the various tidal basins (Fig. 2a). Although these basins primarily act as sediment sinks, several rivers (from Lake IJssel and the Ems, Weser and Elbe estuaries) constitute an additional sediment source, adding a total amount of mud of 1.5 M ton/year into the Wadden Sea.

Fig. 2: Mud budget of the Wadden Sea.
figure 2

a The blue colours show the basin-average mud deposition, calculated for 1996–2015. The brown arrows indicate the main mud sources, and the green dots the average mud extraction from the Ems, Weser, and Elbe estuaries (in which we assume an uncertainty range of 20% originating from a variation in sediment composition of the dredged material). b Cumulative alongshore sediment sinks per type. c Estimated alongshore mud budget and its uncertainty as a function of distance from Den Helder. The uncertainty initially results from the NSCF flux (10–14.4 M ton/year, darkest brown band) but becomes progressively more influenced by sedimentation in the basins, on salt marshes and through extraction (lighter brown bands). The vertical grey bars indicate the location of sources, and the numbered dotted lines the location of individual basins.

The main mud sinks throughout the Wadden Sea can be divided into four categories: deposition in the basins (7.95 M ton/year) and on the salt marshes (1.84 M ton/year), offshore deposition (0.56 M ton/year), and anthropogenic extraction (1.98 M ton/year, see Fig. 2b). Most basin deposition presently takes place in the intertidal areas, especially along the mainland coast, although abandoned channels also acted as major mud sinks in the past48,49. The average mud deposition in the basins and on the salt marshes largely varies per basin (see also Supplementary Tables 1 and 2 and Figs. 1 and 2 in the Supplementary Information). The main offshore sink is Helgoland, where continuous deposition rates of about 1.6 mm/year50 amount to a sink of approximately 0.5 M ton/year. Anthropogenic sediment extraction provides the second biggest sink, even surpassing salt marsh deposition. Hereby we refer to the removal of sediments by dredging without disposing of the sediment within the Wadden Sea or in its vicinity in the North Sea. Although most dredged sediment is redistributed within the Wadden Sea, some of the sediment dredged from the lower Ems51,52, Weser53,54, and Elbe estuaries55,56,57 is disposed on land (either on disposal sites for contaminated sediments or as coastal protection reinforcement).

The mud budget is nearly closed

The various sediment sources and sinks have been converted into a sediment pathway (Fig. 2c). Starting West (at Den Helder), where the NSCF enters the Wadden Sea, it follows the anti-clockwise rotation of the flow along the barrier islands until the most northern end. With increasing distance from Den Helder, the magnitude of the cumulative sinks increases far beyond the sources along its pathway. With our best estimate for the initial sediment supply and various sources and sinks, there is only about 1.5 M ton/year unaccounted for in the northern end. This can be considered a small amount given the wide range of apparent uncertainties and the possible loss of mud to the offshore areas along its pathway.

The largest uncertainty in our estimate is the magnitude of the flux in the NSCF, followed by the basin deposition, mud extraction and salt marsh sedimentation. The uncertainty in these latter three includes the variability in the mud density and sediment composition over time (see the Methods section). The lower end of the cumulative error band is considered to be unrealistic since this would imply that there is no mud left after the Elbe estuary (basin 22), while mud sedimentation has been observed in the basins and on the marshes to its North. The observation that the mud budget is nearly closed is only limitedly affected by uncertainties in the fluvial sources: even if those would vary by a factor of 2 (which does not seem to be the case when comparing our data to other estimates58), the total sources would still only be slightly larger than the sinks.

The spatial variability of the mud sinks is large throughout the entire domain (see Fig. 3a). To compensate for the basin size, we determine the basin deposition per km2, revealing a distinctive trend (Fig. 3b): mud deposition largely increases with increasing distance from Den Helder until km 224 (basin 14, in East Frisia), after which it sharply drops (especially after the Elbe estuary, km 300). Salt marshes are efficient mud traps. Here, sedimentation rates are up to an order of magnitude larger than in the basins (although their total area is much smaller) and follow the same spatial trend. Most sediment deposits are on the foreland marshes, while a smaller amount accretes on the island- and hallig marshes (Supplementary Fig. 2 in the Supplementary Information). The decline of sedimentation rates in the last 200 km of the Wadden Sea is even more pronounced in these foreland marshes (Fig. 3d). These spatial trends in the basin and marsh sedimentation suggest that there could be a mud shortage in this downdrift part of the Wadden Sea59,60.

Fig. 3: Spatial variation of sinks and indicators of a decreasing mud availability.
figure 3

a, b Scatter plot of the total mud sinks in the basins. Tidal basins are indicated by a dot, and estuaries by a cross. Basin sinks per km2 show a clear spatial trend along the Wadden Sea coast. c Histogram of the average salt marsh sinks per km2. d Histogram of the salt marsh sinks in foreland salt marshes. This type of marsh has the largest contribution to the total marsh sinks (see also Supplementary Fig. 2 in the Supplementary Information). e Average mud content (pmud,mean) in the upper sediment bed of the basins. f Modes of the probability density function (PDF) of the mud content in the sediment bed (pmud,modes, plotted per basin), representing the equilibrium mud content in the bed61. Bimodality is detected up to km 300 (Elbe estuary).

Another indicator of the potential mud shortage in the northern German and Danish parts of the Wadden Sea is the upper bed sediment composition in the basins. Whereas the mean mud content ranges between 5% and 25% in the first 300 km without showing a distinctive spatial trend between the basins, it drops abruptly below 10% afterwards (Fig. 3e). Even more pronounced is the trend in the bimodality of the mud content (Fig. 3f). The mud content in the Wadden Sea tends to be bimodally distributed61, with the sediment bed being either relatively sandy (mode 1) or relatively muddy (mode 2). This bimodality may disappear if suspended mud concentrations become low (only mode 1 remains) or high (only mode 2 remains)61. Figure 3f shows that bimodality disappears after km 300, suggesting that suspended mud concentrations are lower here, and therefore also indicating a potential mud shortage compared to the other Wadden Sea basins. Thus, both the sedimentation rates and the bimodality of the bed sediments indicate that mud sedimentation in the northern parts of the Wadden Sea might be supply-limited, which matches the observation that the cumulative mud budget in these areas is largely reduced. Note, however, that a potential higher energy exposure in these areas could also be a contributing factor to this. The western parts of the Wadden Sea appear to be accommodation space limited (i.e., sediment deposition rates are restricted by available space to deposit and not by the supply of sediments). This also explains the high suspended sediment concentrations observed in many parts of the Western Wadden Sea, exceeding many g/l even on the exposed tidal flats28.

Fine sediments as a resource

Intertidal areas comprise nearly half of the surface of the Wadden Sea62 and provide important functions related to safety, ecology and economy. In combination with salt marshes, the intertidal areas protect the mainland coast63,64. Intertidal areas also provide habitat for benthic communities and higher trophic levels, acting as e.g., fish nurseries65 and feeding ground for birds66. Their disappearance would lead to a large-scale loss of biodiversity67. Their economic importance is also evidenced by other ecosystem services such as fisheries62,68.

These intertidal systems are composed of sand and mud. Mud plays a crucial role in creating specific habitats for benthic species45,69,70, and contributing to the current morphological evolution of the intertidal areas48 as well as their ability to keep pace with relative SLR71. The rate at which sand is transported towards the intertidal areas is namely largely dependent on the transport capacity22 and will become a limiting factor under conditions of accelerated SLR34,72,73. The decline in sand supply will increase with increasing distance from the inlet, resulting in partial drowning of the intertidal areas if the sediment deficit is not compensated by mud. Deposition of fine-grained sediments is supply- or accommodation space limited and hardly depending on transport capacity.

The accretion rates on some of the tidal flats presently outpace SLR22,48,74,75, but this may change in the future for two reasons: First, it is expected that SLR will accelerate in the coming centuries. The global sea level has risen 2.1 mm/year since 1970, but this speed is accelerating76,77. The region-mean SLR in the Wadden Sea has been slower—1.77 mm/year since 1958, most likely due to regional geological, glacio-isostatic and meteorological effects—but projections of SLR rates range from 2.2 to 18.3 mm/year for the year 210078. Secondly, the current accretion rates of the tidal flats do not necessarily represent their ability to keep up with SLR. The present-day development of the Wadden Sea is largely influenced by human interventions, of which the response timescales are long. The observed sedimentation rates are partly a response to these interventions, and given the exponential decay of such a response49, it is likely that sedimentation rates will slow down in the future. Accretion rates will largely depend on the sediment availability in combination with sediment management strategies.

Implications for future sediment management

Our findings are important for future sediment management of the Wadden Sea specifically79, but also in general for muddy environments where fine-grained sediments are often assumed to be abundantly available. Our work illustrates that mud is a finite resource, and therefore circularity is key. Mud is abundantly available in the Western Wadden Sea, leading to regular maintenance dredging. Here, even without interventions, SLR will probably lead to higher sedimentation rates over the supra- and intertidal areas because of the increase in accommodation space. But even more, a likely future sediment management strategy will be to progressively trap fine-grained sediments in anticipation of accelerated SLR. SLR, especially when combined with modified sediment management strategies, will then lead to a downdrift reduction of available fine-grained sediments. With the Wadden Sea spanning three different countries, such dependencies illustrate the importance of cross-border sediment management policies.

With mud being a finite resource and SLR inducing shortages over time, some of the policies related to the extraction of mud should potentially be revisited. In the past, sediments were primarily extracted for the construction of flood safety infrastructure (e.g., dikes, dwelling mounds) and roads80. Nowadays, the main motivation for extraction is the deepening of navigation channels (and the costs associated with disposal). Although the total extracted mud mass has decreased over time53,81, it still amounts to 12–17% of the total mud budget. Sediment extraction may provide a relatively cost-effective methodology of dredging and may also lower the suspended sediment concentrations locally81. However, the negative impact of sediment extraction over longer timescales and larger spatial scales in the context of climate change is, at present, not considered in policy frameworks. Given the expected future demand for sediments in order to keep pace with SLR, these long-term large-scale impacts should be accounted for when planning sediment extraction. Such planning would also need to account for a delayed response resulting from the system’s previous buffer capacity22,82,83.

Consequently, we stress the need for cross-border sediment management strategies fuelled by our findings that not only sand but also mud is often a finite resource. The accuracy of tools to predict a coastal system’s response to SLR is limited, posing a challenge for scientists, coastal engineers, and decision-makers. While numerical models may capture short-term up to decadal morphodynamic evolution—especially after large-scale distortion of the system84—their ability to predict (natural) long-term development remains limited85. Moreover, the absence of calibration and validation data of system response to future SLR conditions introduces major uncertainties. We believe that understanding this response should start with a sediment budget on a system scale, i.e., crossing international borders and national policy frameworks.

Methods

North Sea mud sources

The biggest source of mud transported into the Wadden Sea is the North Sea, where mud transport follows the long-year residual current patterns. The North Sea Continental Flow (NSCF, sometimes referred to as Continental Coastal Water) is the most relevant path to the Wadden Sea’s mud import. We reconstruct an estimate of its magnitude based on data and model results reported in the literature, as explained below.

The mud flux transported by the NSCF originates from the Dover Strait, where a net inflow of water and suspended particulate matter is transported into the North Sea86,87. The magnitude of the mud flux has been widely researched, but the results show a large variety originating from the high temporal and spatial variability of mud fluxes in combination with the methodologies that have been used88. An extensive recent study, where the net flux was calculated based on satellite images as well as on numerical model simulations, indicates a mud flux through the Dover Strait of 22.26–31.74 M ton/year88. Approximately 60% of this flux enters the NSCF, whereas the remaining 40% is transported in the East Anglia Plume (EAP)88.

As the NSCF travels North, it transports mud along the French and Belgian coasts toward the Netherlands. Transport from France to Belgium is estimated at 15.5 M ton/year, and further on to the SW Netherlands at 12.8–14.5 M ton/year89. North of the Netherlands–Belgian border, but before reaching the Wadden Sea, the NSCF encounters mud sinks in the Eastern Scheldt and around the Port of Rotterdam, as well as a fluvial source from the Rhine River. Together, this results in a net sink of 0.1 M ton/year90, resulting in 12.7–14.4 M ton/year of mud reaching the coastal zone just South of the Wadden Sea. Other research91 reports slightly lower values of around 10 M ton/year. We therefore assume a range between 10 and 14.4 M tons/year of mud reaching the Wadden Sea.

The EAP is another North Sea current that transports suspended matter from the Dover Strait, the Thames estuary and cliff erosion to the North East along the southern English coast88,92. So far, the amount of suspended matter travelling with the EAP that can reach the Wadden Sea coast seems limited90,92,93. Instead, observations and modelling studies indicate that it is transported via the Oyster Grounds directly to the Skagerrak and the Norwegian Channel93,94.

Main fluvial mud sources

Similar to the marine mud sources, we have collected data from the literature to estimate the average magnitude of the main fluvial sources. These are the freshwater inputs from Lake IJssel (via the sluices den Oever and Kornwerderzand) and the Ems, Weser, and Elbe estuaries. Their magnitudes are shown in Table 1.

Table 1 Main freshwater sources of mud in the Wadden Sea

Atmospheric deposition

We base our estimate of the atmospheric contribution to the mud sources on previously published data. Studies in the North Sea yielded average atmospheric deposition rates of 1.6–2.2 tons/km2/year95. Applied to the total area of the Wadden Sea (11,400 km2), this results in a contribution of 0.018–0.025 M ton/year. This contribution is orders of magnitude smaller than that of the marine and fluvial sources. We therefore consider it to be negligible.

Tidal basin sinks

To calculate the average deposited mud mass in the basins, we follow a previously developed approach that makes use of bathymetry change data in combination with sediment composition data to calculate the contribution of mud to the observed morphodynamic evolution48. Net sedimentation and erosion volumes (ΔV) are calculated by subtracting measured bathymetries. Next, these volumes are multiplied by the volumetric mud fraction in the upper bed (top 4–10 cm) to obtain the volumetric contribution of mud (ΔVmud). Our sediment composition data provides information about the mass fraction of the sediment, for which we use a density relation for sand-mud mixtures in the Wadden Sea96 to derive the volumetric mud fraction. We subsequently transform the calculated volumetric mud sinks into a mud mass by assuming that pure mud in the tidal basins has an average density of 700 kg/m3, while in the estuaries it is slightly lower (500 kg/m3).

We use Dutch bathymetry data of the Vaklodingen dataset (resolution of 20 × 20 m)41 and German data collected within the EASY-GSH project (resolution of 10 × 10 m)42,46. Sediment composition information is based on the Sediment Atlas (sediment samples taken in the ’90s in Netherlands parts of the Wadden Sea, resolution of 500 × 500 m in the channels, and 1000 × 1000 m on the intertidal flats)43, SIBES (sediment samples taken in 2008–2018 in the Netherlands parts of the Wadden Sea, based on 50,000 surface sediment samples of approximately 7400 locations, with a resolution of 1000 × 1000 m in the channels, and 500 × 500 m on the intertidal flats)44,45 and EASY-GSH datasets (German parts of the Wadden Sea, based on 45,000 surface sediment samples which are interpolated on a 10 × 10 m grid)46.

Calculations are performed for the period 1996–2015 (based on data availability) on a computational grid with a resolution of 250 × 250 m. We determine an uncertainty band by assuming that the calculated volumes might vary 40% (20% higher and 20% lower) because of variations in the sediment density of sand-mud mixtures48 and of sediment composition changes within the observed period (i.e., between Sediment Atlas data and SIBES data). No detailed bathymetric evolution or sediment composition data of the Danish basins is available, for which we have based the results of these mud sinks on existing literature97.

Salt marsh sinks

Similarly, we make use of bed composition data in combination with accretion rates98 to calculate the mud sinks on the salt marshes. The size of salt marshes in the study area was derived from literature98 and is based on surveys carried out between 2004 and 2014. For sedimentation, we follow the terminology for sedimentation where this is expressed as the amount of sediment that is deposited in the marsh per unit area and time, i.e., kg/m2/year99. No adjustments were made for organic matter since Wadden Sea salt marshes are generally minerogenic and thereby low in organic matter. Salt marsh vegetation does contribute to an enrichment of organic matter in the upper sediment layers of the marsh bed, but this organic matter appears to be lost at greater depth when it is buried by new sediment layers100. Similar to the calculations of the basin sinks, we account for an uncertainty range of 40% total in the salt marsh sinks calculations, originating from variations in the sediment density and in the sediment composition.