Main

The importance of fine-grained mud in shaping Earth’s climate history has stimulated broad interest in the geosciences1,2, with focus on the associations between mud and organic carbon (mud–OC; Box 1) beginning in the mid-nineteenth century. Mud is a key medium that integrates the carbon cycle as initiated by rock weathering and soil erosion, followed by transport, transformation and sequestration/burial of mud–OC across diverse landscapes3,4,5. Mud has been linked with microbial evolution6, relating it to past and future changes in weathering, biogeochemical cycles and climate6. Mudrocks represent about 60% of all sedimentary rocks in Earth’s crust, and are the primary archive from which geologists reconstruct Earth’s biotic and climatic history7. Mud’s influence on Earth’s carbon cycling and climate has accentuated since the Palaeozoic era (Supplementary Information).

Anthropogenic disturbances to mud and mud–OC cycling have probably occurred from the mid-to-late Holocene epoch8, with the most rapid and profound effects occurring during the Great Acceleration, or Anthropocene (for example, ref. 9). Human activities alter the production, source-to-sink transport and fate of mud–OC via changes in land use (for example, deforestation, agriculture, mining, road building), water management (for example, damming, levees, groundwater withdrawal), perturbations of the coastal and deep ocean (dredging, trawling, offshore wind farms, aquaculture, mining), atmospheric CO2 and climate (for example, droughts, floods, sea-level rise, glacier retreat, permafrost thaw, terrestrial biosphere productivity). Because the fate of mud–OC can have substantial impacts on greenhouse gas fluxes and carbon sequestration/burial in the biosphere, important questions remain on how these anthropogenic changes will affect the carbon cycle and climate in the twenty-first century10. Humans have altered the transit and hence distribution of mud between its weathering sources and its depositional sinks, dramatically changing its residence time along source-to-sink gradients11,12,13,14,15,16,17,18. Moreover, changing environmental conditions strongly influence the inputs, processing, remineralization and sequestration of mud–OC19.

Here we review the state of knowledge about mobilization and storage of mud and its associated OC through dominant source-to-sink pathways (Fig. 1). We focus on Holocene–Anthropocene effects by human activities, which are altering production, transport and environmental conditions for mud–OC and therefore its exchange with pools such as the atmosphere. We emphasize mud–OC in permafrost, whose thawing mud enhances fluxes from land to sea in the Arctic (for example, ref. 4). Our main objectives are to: (1) assess how the dominant drivers of mud–OC production have changed over the Holocene–Anthropocene; (2) provide an overview of the large spatial and temporal changes in where mud–OC is remineralized and buried along the land-to-sea pathway, with potential consequences for carbon cycling and climate; and (3) explore how understanding of the fate of mud–OC can help predict consequences such as the recently accelerated release of ancient petrogenic and permafrost-derived OC to the biosphere.

Fig. 1: Major pathways of mud movement, POC fluxes and human-induced changes.
figure 1

The primary landscape processes controlling fluxes of mud and mud–OC (black text) influence mud–OC flux (red numbers) and their recent changes (+/–/?) via anthropogenic impacts (red abbreviations) along source-to-sink gradients. See Table 1 for references and further information associated with these fluxes. All fluxes are in TgC yr−1 rounded to ±5 TgC yr−1. Note that the budget is not closed because it is partly constructed from independent estimates, the contribution of aquatic system metabolism to burial remains largely unknown, and the timescales of erosion, transport and burial are not uniformized. For instance, mud–OC burial on the continental shelves is timescale-dependent as shown quantitatively in refs. 100,101. The red text indicates notations present in Table 1. Processes/fluxes without numbers lack valid quantitative evidence.

Full size image

The modern mud–OC cycle on land

Production of mud–OC in terrestrial ecosystems

Mud production requires weathering reactions, which change both particulate and dissolved materials (for example, feldspar to clay minerals and consequent increase in dissolved silica and various ions). Mass budgets of mud are dominated by transfers among pre-existing mud deposits11 (for example, soils and sediments) in different parts of the landscape—not new production of mud-sized minerals. Monitoring of dissolved ions to assess global weathering does not indicate Anthropocene increase in mud-producing reactions11.

Despite little change in total available mud, erosion of mud–OC from ice-free landscapes in temperate regions has been substantially affected by land-use change over the Holocene–Anthropocene. Human agriculture has substantially increased global soil denudation since the late Holocene; an estimated ~36 Pg yr−1 of soil was eroded in 201212. Currently, an estimated 37% of all ice-free land is directly used for agriculture and human settlements13,14. Impacts are most pronounced in the Northern Hemisphere, which hosts more land, human population and gross domestic product12,14. Soil loss and mud–OC mobilization due to anthropogenic land-use changes began >4,000 years ago8,16. In North America, the impact of European colonization on the landscape (via agriculture and river modifications) is readily observed, with rates of surficial sediment movement (and hence mud–OC mobilization) over the past century about ten times higher than pre-colonial rates15. Rates of deforestation and agricultural land expansion are now slowing or even reversing in the Northern Hemisphere, and accelerating in the Southern Hemisphere17. In Europe, climate change rather than land-use change is predicted to be the main driver of modest increases in soil erosivity in the coming decades18. Global land-surface models integrating vegetation dynamics suggest that enhanced plant growth driven by increased atmospheric CO2 could partly mitigate the erosive effects of climate change via soil stabilization20. In contrast, substantial late twentieth- and twenty-first-century increases in land-use changes in South America, Africa and Southeast Asia have made these tropical areas the main loci of soil erosion, with rates predicted to substantially increase in the near future12. Human-induced, extreme precipitation/flooding events are also predicted to increase in many regions across the world (for example, ref. 21), affecting erosion rates and the fate of mud–OC12.

Fate of mud–OC along the global inland water network

Humans clearly affect the mobilization, processing and storage of mud–OC in the transit from soils to the ocean, especially via the residence time of mud in different parts of the system3,22,23,24. Dam proliferation in North America, Europe/Eurasia and Asia since the 1950s is generally believed to be starving the coast of sediment (Supplementary Fig. 1). In contrast, sediment transport in 39% of rivers in South America, Africa and Oceania has increased since the 1980s due to land-use changes, especially deforestation25. Current estimates indicate a 49% global reduction in fluvial sediment reaching the oceans, despite a >200% increase in upstream fluvial sediment loads, between 1950 and 201011,25. Dams can trap mud–OC (60 TgC yr−1, range 20–70 TgC yr−1; Figs. 1 and 2c, and Table 1)26, the magnitude of which depends upon specific environmental and hydrological conditions (for example, refs. 26,27). For instance, dams strongly stimulate phytoplanktonic production within the global inland water network, but also the mineralization of both fresh OM and terrestrial-derived material; thus these systems exhibit a highly variable, net heterotrophic status at the global scale27. Dam construction up to 1970 eliminated 8% of the total riverine OC flux through burial and mineralization, and this removal rate is expected to have more than doubled (to 19% in 2030) with dams either completed or planned after 197027. Furthermore, the interruption of sediment flux by dams increases net downstream erosion (Fig. 2c), which can partially offset mud–OC trapping until the river profile re-equilibrates (for example, ref. 28). Lakes without dams are also important hotspots of mud–OC burial (90 TgC yr−1, range 40–180 TgC yr−1; Fig. 1 and Table 1), although substantially less efficient on an areal basis than reservoirs with higher sedimentation rates and better conditions for OC preservation (for example, anoxia; Box 2)26. Moreover, several regional studies suggest a substantial increase in lake mud–OC burial since pre-industrial times5. Northern Hemisphere lakes have increased OC burial by about 50% over the past century (Fig. 2c), possibly due to the combined effects of climate change and enhanced terrestrial productivity29. Present-day OC accumulation rates in European lakes are double those of the Holocene, mostly attributed to land-use change30.

Fig. 2: Examples of major perturbations of mud–OC pathways.
figure 2

a,b, Accommodation space for vegetated coastal ecosystems (a) and deltas (b) is affected by the changing influence of sea level, sediment accumulation rate and vertical ground motion. c,d, Sediment transport pathways are being altered by anthropogenic alterations, such as dam construction (c), and climate warming and impacts to coastal environments (d). Plus and negative signals represents increments (+) or reductions (−) in mud–OC fluxes from/to accommodation space associated with different processes that have been altered during the Anthropocene.

Full size image
Table 1 The POC cycle in the Anthropocene, probably dominated throughout by mud–OC
Full size table

River floodplains are key areas of storage, processing and release of mud–OC, affecting OC:SSA values and OC reactivity. As with reservoirs, they lead to mixtures of new and ‘aged’ materials3,4. In contrast to lakes and reservoirs for which global assessments of long-term burial are available (for example, ref. 26), the amount of mud–OC sequestered in floodplain systems is highly uncertain but possibly of similar magnitude (190 TgC yr−1, range 60–320 TgC yr−1)5,31 as in lentic bodies (150 TgC yr−1, range 60–250 TgC yr−1; Fig. 1 and Table 1). Natural floodplains store and release mud via overbank sedimentation and river channel migration/river bank erosion, respectively, as well as other climate-driven fluctuations in the hydrological cycle (for example, La Niña and El Niño). Levee construction interrupts this process, often by reducing connectivity between rivers and their floodplains (for example, ref. 32). Increased erosion from human activities may reduce mud residence times in the land–ocean transition (Fig. 2b), but the creation and isolation of floodplains has the opposite effect33. The net effects of human activities on the floodplain mud–OC cycle remain largely unknown at the global scale (see ref. 5). While the composition of mud minerals varies relatively little during the long transit from source to sink, as evidenced by similar oceanic and adjacent terrestrial clay mineral suites34, the composition of mud–OC is more variable35,36. Much mud–OC is decomposed in floodplains in its transit23,37. Along the entire source-to-sink transition, but particularly in the floodplain, petrogenic OM is partially replaced by OC from land plants and river/lake phytoplankton (for example, refs. 23,27). The addition of fresh OM may prime the degradation and replacement observed in floodplains38.

Changes in fluvial morphology affect the fate of mud–OC by altering water dynamics, residence time, redox conditions, turbidity, particle size/density and mineral/OC sources39. Agricultural expansion in river networks in China (1960s–1980s) enhanced erosion and resulted in the loss of high-order rivers to sediment infilling40. During later urbanization (1980s–2010s), when ca. 40% of some natural landscapes reached a status of ‘urbanized’ (for example, extensive dredging and reconstruction of high-order rivers), lower-order rivers experienced sediment infilling. The importance of changing floodplain topography on mud–OC cycling remains largely unexplored (for example, ref. 41). The creation of reclaimed agricultural land such as rice paddies (Supplementary Fig. 1) can enhance OC:SSA ratios several fold42. Greater predictability in the land-use-driven changes in river channel evolution and spatial–temporal dynamics of erosion and sedimentation across watersheds43 will provide consistent frameworks to assess changes in mud–OC44.

Production, transfer and storage of mud–OC at the coast

In addition to impacts on terrestrial sediment and mud–OC fluxes, damming has contributed to coastal erosion (Fig. 2) in many of the world’s larger deltas (for example, those with areas >1,000 km2 such as the Mississippi, Mekong and so on)45, which are sinking several times faster than smaller deltas because they are downstream of more dams, contain higher fractions of compressible mud in areas large enough to induce isostatic subsidence11 and are extensively modified by humans. In contrast, many (mostly smaller) coastal deltas have grown over recent decades, largely due to increases in fluvial-derived sediment linked to deforestation46. Century-long records show a doubling of sediment accumulation rates in most North American coastal depocentres apart from the Mississippi Delta region, facilitated by erosion downstream of dams47 (Fig. 2c).

Coastal deltas and estuaries (Fig. 2a,b) are key depositional and processing environments of mud–OC along the source-to-sink transition, where unidirectional river flow interfaces with tidal and wave processes (for example, ref. 48). Despite complexity among coastal regions, sea-level rise and extensive coastal development have resulted in a net global decrease in mudflat area, primarily in temperate and low-latitude regions (Supplementary Fig. 1)49. Over two decades (1999–2019), an estimated 13,700 km2 of tidal wetlands were lost globally, offset by gains of 9,700 km2, for a global net loss of 4,000 km2 (ref. 50). Coastal wetlands (Fig. 2a), which commonly host ‘blue carbon’ (for example, mangroves, tidal marshes and seagrasses; Supplementary Fig. 1), can have very high OC:SSA values (for example, up to 34 mgOC m−2)51,52 and have some of the highest rates of short-term carbon sequestration and mud–OC burial (for example, ref. 53), with a global assessment reaching 60 TgC yr−1 (range 40–80 TgC yr−1; Fig. 1 and Table 1)5.

In fast-warming pan-Arctic latitudes, permafrost thaw and thermo-erosional features in coastal regions have remobilized soils and changed source-to-sink movement of mud–OC (for example, refs. 54,55). While the range of grain size in permafrost can be quite variable, recent studies have shown that the majority of soil OC in permafrost across the Arctic is in the mud fraction56,57. Warming air and sea temperatures, sea-level rise and longer open-water seasons have enhanced pan-Arctic erosion and mobilization by 14 TgOC yr−1 from permafrost soils to the aquatic continuum54,58 (Figs. 1 and 2d, and Table 1). In turn, the mobilization of this old mud–OC and associated nutrients sustains a substantial fraction of Arctic primary production and supply of new fresh OM59. Much of the permafrost OC comprises silty mud–OC draining from nearshore erosion of retrogressive thaw slumps and bluffs/cliffs60. Mud–OC export from retrogressive thaw slumps, which typically extend farther inland than cliffs, may take decades to hundreds of years, compared with days to months from cliffs, before reaching the Arctic Ocean54. These differences in OC release result in a slower and steady conversion to CO2 from retrogressive thaw slumps, compared with more rapid pulses of cliff-derived mud–OC release54. Pan-Arctic coastal systems also provide more targeted zones for examining the impact of ocean phytoplankton on the fate of permafrost-derived mud–OC in a warming climate. For example, mud–OC in deep waters of non-glaciated fjords of southeastern Alaska is largely undegraded with modern radiocarbon ages (biospheric sources), due to inputs of phytoplankton61. In contrast, nearby glaciated fjords are starved of phytoplankton and bury substantial amounts of petrogenic OC and terrigenous biospheric OC (Fig. 2d). Similar to cliff and retrogressive thaw slump systems, Arctic deltas represent an important land–sea interface, where thawed, millennial-aged, permafrost-derived mud–OC is processed. For example, permafrost-derived mud–OC in the Colville River delta, Alaska, originates from bank erosion in upstream tributaries in the basin55 (Fig. 2d). How glacial retreat in the Arctic affects mud–OC burial will largely depend on regional differences in sedimentation rates, the relative inputs of older terrestrial sources (for example, petrogenic, permafrost) versus younger marine (macro- and microalgal), differential binding of these OC sources to minerals and the response of the microbial community to these changing pools.

Accommodation space in nearshore coastal ecosystems

Accommodation space is space available for vertical mineral and organic material accumulation in nearshore ecosystems (for example, coastal wetlands, deltas, estuaries, inner shelves). It is largely controlled by relationships between sea level, sediment accumulation rate and vertical ground motion (for example, isostatic adjustment, tectonics, subsidence/sediment compaction and fluid withdrawal; for example, ref. 62; Figs. 1 and 2a,b). However, dramatic anthropogenic alterations in the delivery of fluvial sediments to the coast11 and structures from human development63 also change available accommodation space, which affects potential storage and turnover of mud–OC (Fig. 2a,b).

Modelling and empirical data suggest that accommodation space is a key variable determining coastal wetland habitat (Fig. 2a) expansion during sea-level rise over the past few millennia64. Recent models show that, over the past ca. 4,200 years, tidal marshes in regions with more rapidly declining relative sea-level rise (for example, Europe and North America) had greater OC concentration than in regions with slower declining relative sea-level rise (for example, Africa, Australia, China and South America)65. In the case of the Northern Hemisphere, where relative sea-level rise has been decelerating, vertical and lateral accommodation space was created over time66, due to greater inundation frequency that allowed for higher mud–OC accumulation64. Controls on accommodation space are further complicated in large deltaic regions experiencing anthropogenic disturbances (Figs. 1 and 2b). For example, these regions experience high rates of erosion and subsidence, largely due to upstream damming and deltaic activities such as fossil fuel and groundwater extraction, respectively (for example, ref. 67). The synergistic effects of damming and subsidence increase relative sea-level rise in these deltaic regions, and further complicate modelling efforts of changing lateral and vertical accommodation space and associated mud–OC storage68. Damming also enhances accommodation space for storage of mud–OC in reservoirs (Figs. 1 and 2c).

Accommodation space in the Arctic coastal zone is also changing, as many glaciers (Figs. 1 and 2d, and Table 1), especially tidewater glaciers, experience rapid retreat69. This retreat poses new questions of how plant colonization will impact the erosion and development of soils, and hence mud–OC, in newly exposed proglacial deposits (sometimes termed paraglacial)70. For example, the development of coastal landforms (deltas, cliffs, tidal flats, beaches) in proglacial deposits over the past 100 years in Svalbard, Norway71, provides new accommodation space for producing and processing mud–OC in the Arctic (Fig. 2d), and over relatively short periods of time (10−1 to 102 years). Tidal glacier retreat creates proglacial landforms that potentially increase accommodation space, which can then increase residence time (in part, stabilized by shrubification) and microbial processing of mud–OC in source-to-sink transport. Newly exposed glacial sediment can show rapid increase in OC:SSA ratios72. To date, much of what is known about primary succession of plants in these temperate-to-Arctic environments is from dated chronosequences from post-glacial retreat following the Little Ice Age73.

Production, transfer and storage of offshore mud–OC

Muddy ocean deposits dominate longer-term processing and storage of mud–OC22. Organic loadings per unit of mud, as indicated by OC:SSA ratios, vary among depositional sites within an ocean margin region, depending on local ratios of supply versus degradation rates of mud–OC3,74. The burial or oxidation fate of enormous quantities of terrigenous OC depends on local oceanographic conditions. For example, the 1,600-km-long inner shelf mud belt, which moves from the mouth of the Amazon River to the Orinoco Delta, efficiently oxidizes terrigenous mud–OC, as energetic transport lowers OC:SSA ratios several fold via frequent resuspension and re-oxidation of the seabed75. In contrast, the offshore Ganges–Brahmaputra and Congo River outflows exhibit seaward escape of sediment via turbidity currents in submarine canyons and efficient terrigenous mud–OC burial on the adjacent deep-sea fan23. One of the more dramatic examples of human impacts on the distribution of mud–OC is the state change from actively accreting to eroding expansive shelf mud blankets. For example, humans and climatic variations interacted to control Holocene mud flux from mid-latitude Chinese loess hills to the adjacent ocean margin76. Recently, the underwater delta off the Yangtze has been rapidly eroding in response to river damming that captures sediment upstream77, and that will surely impact biogeochemical processes and elemental fluxes for the East China Sea. Future planned dams in the Amazon basin will probably dampen the extensive offshore mobile mud belts and mud–OC oxidation78.

Along the ocean margins, the human impacts on mud–OC burial driven by changes in terrigenous OC deliveries are confounded by anthropogenic perturbations to ocean phytoplanktonic productivity. These include the effects of a changing physical climate and changes in nutrient inputs from atmospheric and riverine sources. Although it has long been advocated that human activities have stimulated ocean productivity and OC burial in the shallow portions of the ocean (for example, ref. 79), only recently have these impacts been quantified using physically resolved ocean biogeochemistry models80. Results suggest that over the Anthropocene, net coastal ocean productivity increased by 14% as a result of nutrient inputs, and was higher in hotspot regions such as the East China Sea, southern North Sea, Louisiana shelf and shelves of the Bay of Bengal80. These results confirm reported widespread increases in biological productivity and eutrophication in coastal regions over the past century81, probably inducing greater export, deposition and burial of mud–OC. With low confidence, the effects of a changing physical climate in the coastal ocean appear limited so far. The confounding effects of changing terrestrial and marine carbon cycles driven by multiple human factors (land-use change, climate, atmospheric CO2, nutrient supply) make quantitative assessment of net changes in OC storage in muddy ocean sediments challenging5. Elevated OC:SSA ratios are evident in smaller eutrophied areas such as Long Island Sound74, but not in larger, more exposed environments such as the East China Sea or Mississippi Delta82. This contrast suggests human impacts on OC decay may be more important than those on OC supply.

Mud–OC reactivity across a changing source-to-sink gradient

First-order degradation rate constants (k for a first-order rate law of the type dG/dt = −kG, with k denoting the first-order rate constant, G the OM concentration of natural OM and t is time) vary by many orders of magnitude, with a strong inverse relationship (over 12 orders of magnitude of time) in which OM is exposed to oxidizing or remineralizing conditions83,84 (Fig. 3). This global observation broadly supports the widely accepted ‘aged OC is refractory relative to recently produced OC’ paradigm. It does not directly provide a mechanistic explanation of the long-term persistence of OC in the environment, although it can be summarized via integrative parameters such as energies of activation85. Superimposed on this broad trend is one to three orders of magnitude variation in reactivity at any given time of exposure. This variation is probably owing in part to varying definitions of time and reactivity arising from different data sources or models. It is also certainly subject to a plethora of different factors such as OM composition, electron acceptor availability, benthic microbial community composition, physical and physicochemical protection, temperature, microbial inhibition by specific chemicals, priming, and macrobenthic activity (for example, ref. 86 and references therein). Nevertheless, apparent reactivity provides an averaging dynamic parameter, accounting for interactions of compositional and environmental effects (Box 2). Interactions between mud and mud–OC seem to be particularly important on degradation timescales longer than 100–101 years (ref. 87) that are particularly important to carbon sequestration85.

Fig. 3: Gradients of mud–OC reactivity.
figure 3

a, Typical mud–OC transport timescales for the transit through different environments as derived from radiocarbon measurements (for example, soil102, land–ocean aquatic continuum (LOAC)103,104,105 and shelf106,107) calculated based on vertical sinking rates (shelf, slope and abyss) and burial rates (marine mixed sediment, marine deep sediment). Vertical sinking zones listed on the y axis are as follows: shelf (0–200 m), slope (200–3,000 m) and abyss (>3,000 m). b, Distribution of apparent OM reactivity as derived from observations, model fitting and/or laboratory experiments for POC and dissolved OC (DOC) across different environments over degradation time (for example, river, catchment, lake, reservoir, wetland108) and ocean (sediment traps109, sediment109,110). The first-order degradation rate constant, k, predicted for the respective exposure times by the reactive continuum model model (k = 0.125 (0.56 years + exposure time)−1) and the power model (k = 0.21 × exposure time(−0.985)) are indicated by the dashed and solid line, respectively. We use ‘degradation time’ rather than ‘time’ as time could in theory pass without degradation proceeding (for example, permafrost or redox fluctuations or temporary mineral protection).

Full size image

Accelerated human activity in the Anthropocene acts on these reactivity controls. Climate shifts can thaw permafrost mud–OC and enhance microbial decay, or stratify water columns that can deplete bottom waters of oxygen and slow microbial attack under quiescent or resuspension conditions (for example, ref. 88). Shunting mud into zones subject to fresh OM inputs—such as eutrophic coastal waters, dammed reservoirs or floodplains—can enhance reactivity of aged mud–OC via priming38. Such upshifts and downshifts of reactivity have potentially large impacts on mud–OC reactivity, coupled to changing mud residence times among different depocentres.

Concluding remarks

Most mud–OC formed over timescales longer than those of the Anthropocene, during which human activities are destabilizing this pool. These recent changes have created a non-steady-state situation that contrasts substantially with the mid-Holocene epoch, when climate conditions (and erosion processes) were more stable8,89. Furthermore, human impacts have led to the ‘release’ of old mud–OC, via destabilization and erosion of the landscape, into the modern carbon cycle4,90. While much has been discussed about the thaw and release of millennial-aged OC in high latitudes and its consequences for climate, only recently have we begun to consider the impact of climate change on the stability of mud–OC in tropical settings. Furthermore, anthropogenic activities also mobilize and destabilize petrogenic mud–OC in catchments with steep terrains and OM-rich metasedimentary rocks91.

Mud holds most sequestered OC and exerts an important control on both OC transport and reactivity. The spatiotemporal history of mud in source-to-sink systems controls their respective net carbon budget over a wide range of timescales. On geologic timescales, the balance between the oxidation of petrogenic mud–OC and the formation/stabilization of biospheric mud–OC can tip source-to-sink systems from net carbon sinks (for example, the Ganges–Brahmaputra67) to net carbon sources (for example, Taiwan91). Over much shorter, human timescales, the Great Acceleration9 has caused substantial shifts in the environments of the mud medium, leading to rapid changes in its mud–OC content and composition. While grain-size normalization illuminates mud–OC changes in any grain-size matrix, future research might emphasize areas where large muddy OC fluxes are especially affected by humans—for example, muddy parts of floodplains (Fig. 1). Determination of changes in mud–OC content, source and composition relative to the conservative medium—rather than simple relocation of mud (for example, ref. 92)—will allow better accounting for dynamic carbon reservoirs such as blue carbon, with implications for our ability to predict the global short-term evolution of OC reservoirs and attendant trends in atmospheric CO2 levels. New analytical methods geared towards obtaining fine-scale characterizations of the interactions between mud and mud–OC (for example, MAOC, ramped pyrolysis oxidation, Fe–OC, split flow thin cell technique) as well as data aggregation tools are leading to global scale quantification of these disturbances to previous values. Finally, we summarize that anthropogenic activities have increased net fluxes and/or burial of mud–OC from mountain glaciers, land erosion, dam and lake reservoirs, river export, permafrost thaw, ice-sheet erosion and coastal margins, with uncertain net changes for tidal flats and floodplains, and probably net decreases in coastal wetlands (Table 1). Whether these impacts predict future trends deserves mechanistic and budgetary study. Whether the anthropogenic mobilization of mud–OC results in more or less sequestration of OC is not known with the current data, as it is dependent on timescales that involve complex transient effects.