Coastal marshes store large amounts of organic carbon in their soils, but erosion can mobilize this carbon to the coastal ocean, creating uncertainty in carbon budgets and climate mitigation efforts. Here we show that erosion of coastal marshes along the United States Atlantic and Gulf Coasts represents a substantial lateral transfer of organic carbon (0.66 Tg C yr−1; 68% confidence interval, 0.46–0.91 Tg C yr−1). Using long-term satellite observations of wetland change and elevation-based estimates of eroded soil depth, we quantified carbon mobilized by abrupt and gradual marsh erosion from 1985 to 2022 and compared it with carbon burial (0.22 Tg C yr−1; 68% confidence interval is 0.09–0.59 Tg C yr−1) in newly formed marshes. We found that erosion exports more carbon than is buried in newly formed marshes (net export = 0.38 Tg C yr−1), with strong spatial and temporal variability. The erosion flux is equivalent to 5.7% of riverine organic carbon inputs and 24.2% of marsh outwelling, with regional values ranging from 3–14% and 6–29%, respectively. These results demonstrate that marsh erosion is an important pathway in coastal carbon cycling and should be incorporated into coastal carbon accounting and management frameworks. Coastal marshes along the United States Atlantic and Gulf Coasts mobilize substantial quantities of buried coastal carbon, representing a lateral flux to the coastal ocean as high as 0.66 Tg carbon per year, based on an erosion and burial carbon flux estimates from 1985−2022.

AimTo develop the palaeoecological and palaeoclimatic history of Maplecrest Fen, Greene County, NY.LocationCatskill Mountains, New York.Time Period13,500 calendar years to present (ybp, present = 1950).TaxonAngiosperms, gymnosperms and bryophytes.MethodsWe used pollen and spore analysis, macrofossil analysis, loss-on-ignition (LOI) and X-ray fluorescence (XRF) analysis of a 7.8 m sediment core from the Fen along with AMS radiocarbon dates retrieved from the identified terrestrial macrofossils.ResultsLate-glacial and Holocene vegetational change from ice withdrawal reveals the development of a shallow lake and then fen to the present. The boreal spruce (Picea), fir (Abies) and paper birch ((Betula papyrifera) pollen signature near the base of the core (Zone M2) suggest a Younger Dryas (YD) signal overtopping warmer Allerod aged basal sediments. The early Holocene white pine (Pinus strobus) is followed by increases in hemlock (Tsuga) and beech (Fagus), ushering in a moister climate. About 5200 ybp drought is indicated by the well-known hemlock biostratigraphic decline in the Northeast. Drepanocladus moss and stonewort (Chara) shift to sedges (Carex) and violets (Viola) locally present. For the next 5000 years, the vegetation will remain relatively stable with lower temperatures suggested by the return of spruce and fir. Ragweed (Ambrosia) rise along with other weedy taxa marks European settlement near the top 50 cm of the core, along with a striking increase in anthropogenic lead, as seen using X-ray fluorescence.Main ConclusionsDeglaciation began in the Catskills at a high elevation about 13,500 years ago and recession of ice from the Laurentide margin proceeded at about 0.1 km/year. Temperature shifts promoted changes in the boreal forest including warming and cooling, and drought and wetter intervals both appear in the Holocene history but do not appear to be easily forecast as climate continues to warm.

Published cosmogenic 10Be exposure ages from the terminal moraine of the Laurentide Ice Sheet (LIS) in northeastern North America have been interpreted to date the start of the retreat of the LIS at the Last Glacial Maximum (LGM) about 25 thousand years ago (ka). In contrast, published 14C accelerator mass spectrometry (AMS) dates for terrestrial plant macrofossils in LIS basal deglacial clay deposits range back to only ~16 calibrated (cal) ka, more consistent with the timing of glacio-eustatic rise and associated meltwater discharge to the North Atlantic and Gulf of Mexico associated with LGM deglaciation. We apply statistical models of geomagnetic secular variation, including dipole moment, to the latitudinal scaling of cosmic ray flux to see how well the age discrepancy can be addressed. A preferred new scaling, which is essentially time-invariant over the relevant LGM age range, shifts the exposure ages only a few thousand years younger. The age discrepancy may thus stem more from potential local biases toward higher 10Be concentrations (older apparent ages) at the terminal moraine sites, such as much higher 10Be production rates at the LIS front, and especially from inheritance. Such biases can be tested by obtaining primary 10Be calibration sites in the LGM time frame, and by more comprehensive sampling strategies for glaciated terrain to discern inheritance.

Tidal wetlands provide critical ecosystem functions for coastal communities including flood protection, water filtration, carbon sequestration and aquatic nursery habitat. However, New York City’s salt marshes, including our study site at Pelham Bay Park’s Turtle Cove, are rapidly disappearing due to accelerating relative sea-level (RSL) rise and coastal development. Field research, mapping and satellite imagery reveal significant loss of this ∼10 hectare (ha) wetland, as perturbations from human activity prevent marsh landward migration, impede tidal flows and threaten marsh survival. We extracted three sediment cores and conducted 20 m transects across a gradient of disturbed marsh areas. We present the analyses of land-use change, X-ray fluorescence (XRF), loss on ignition (LOI), stable carbon isotopes (δ13C), foraminifera, and accelerator mass spectrometry (AMS) radiocarbon dating of terrestrial macrofossils to examine the past and to inform future conditions for this rapidly eroding wetland. Moreover, we reconstruct sea level over a millennium to analyze changes in marsh plant communities in response to RSL rise and coastal development. We found that between 1974 and 2018 CE, ∼65% of marsh disappeared at a rate of 1.5% yr-1 or 800 m2 yr-1. The marsh loss coincided with increasing RSL rates of 3.5 mm yr-1 from 1958-1975 CE to 6.7 mm yr-1 from 1999-2024 CE. Meanwhile, developed areas expanded 568 m2 yr-1 from 1985-2023 CE, replacing wetland areas and disrupting hydrologic processes with hardened shorelines. Marsh loss resulted in the release of soil organic carbon stored over many centuries and a concerning amount of lead (Pb) into Long Island Sound, presenting risks to public health and wildlife. Culvert assessments demonstrated that tidal restriction by built structures contributed to rising tide levels comparable to RSL rise over the past century, which likely exacerbated marsh erosion. Lastly, tidal prism reductions caused enough accumulation of heavy metals to significantly alter peat chemical composition for a century. This study improves our understanding of compounded stressors that prevent the capacity of salt marshes to with stand anthropogenic impacts. Ultimately, our findings inform an adaptive management of these threatened ecosystems in their struggle to keep pace with climate change and urbanization.

The persistence and size of the Greenland Ice Sheet (GrIS) through the Pleistocene is uncertain. This is important because reconstructing changes in the GrIS determines its contribution to sea level rise during prior warm climate periods and informs future projections. To understand better the history of Greenland’s ice, we analyzed glacial till collected in 1993 from below 3 km of ice at Summit, Greenland. The till contains plant fragments, wood, insect parts, fungi, and cosmogenic nuclides showing that the bed of the GrIS at Summit is a long-lived, stable land surface preserving a record of deposition, exposure, and interglacial ecosystems. Knowing that central Greenland was tundra-covered during the Pleistocene informs the understanding of Arctic biosphere response to deglaciation.

Past interglacial climates with smaller ice sheets offer analogs for ice sheet response to future warming and contributions to sea level rise; however, well-dated geologic records from formerly ice-free areas are rare. Here we report that subglacial sediment from the Camp Century ice core preserves direct evidence that northwestern Greenland was ice free during the Marine Isotope Stage (MIS) 11 interglacial. Luminescence dating shows that sediment just beneath the ice sheet was deposited by flowing water in an ice-free environment 416 ± 38 thousand years ago. Provenance analyses and cosmogenic nuclide data and calculations suggest the sediment was reworked from local materials and exposed at the surface <16 thousand years before deposition. Ice sheet modeling indicates that ice-free conditions at Camp Century require at least 1.4 meters of sea level equivalent contribution from the Greenland Ice Sheet.

Coastal marshes are efficient ecosystems providing a multitude of benefits for invertebrates, birds, fish and humans alike. Yet despite these benefits, wetlands are threatened by anthropogenic inputs such as human wastewater which contain high levels of nitrogen (N). Increased nitrogen loads cause eutrophication and hypoxia in estuaries leading to further degradation of these valuable ecosystems that are already stressed by sea level rise and climate change. Policies to protect wetlands via wastewater treatments are reactive rather than proactive and a growing body of research shows that characteristics associated with population health and economic activity can be identified in wastewater. Analysis of a 2-m salt marsh sediment core reveals δN15 signatures indicative of human population rise and connects human impact to ecosystem health. Using key X-ray fluorescence (XRF), pollen, sediment and nitrogen signatures along the core, a robust chronology was produced dating back to 1700. This result was coupled with population data to observe the relationship between δN15 levels and population over three centuries. There is a statistically significant positive correlation between δN15 and population. Other external factors such as federal government policies (regulating clean water) show a clear reduction in this association but the use of synthetic nitrogen fertilizer masks the strength of this relationship. Further research to refine the relationship between population and δN15 could be beneficial in predicting nitrogen loads as human population grows, which in turn would create a proactive system to protect our coastal ecosystems.

What we regard as anomalously old 10Be exposure dates reported from the terminal moraine of the Laurentide Ice Sheet (LIS) in northeastern North America, such as recently published for Allamuchy NJ, ostensibly point to the start of deglaciation at 25 thousand calendar years before present (cal. ka). These dates are well within the conventional age span of the Last Glacial Maximum (LGM) and are in stark contrast with published 14C accelerator mass spectrometry (AMS) dates for earliest terrestrial plant macrofossils found in LIS deglacial clay deposits that range back to only ~16 cal. ka, which more plausibly coincide with the known timing of the glacio-eustatic rise and meltwater discharge to the North Atlantic and Gulf of Mexico that mark the demise of the LGM in the marine record. To explore possible explanations for this inconsistency, we first employed a statistical model of the geomagnetic field that includes secular variation with nondipole terms and can be applied globally. The model results in a decrease in the magnetic shielding factor by about 10% at mid-latitudes compared to oft-used geomagnetic scaling schemes. However, the time-integrated axial dipole moment estimated separately suggests little overall change in average shielding since about 20 cal. ka. This seems to leave cosmic ray flux modulated by a time-varying heliomagnetic field linked to sunspot activity as an underestimated factor in widely used 10Be exposure age calculators. If generally biased by about 23% higher compared to modern levels as reported for the past 9.4 cal. ka, the elevated high cosmic ray flux would make 10Be reference production rates proportionately higher, to about 5.5 at/g/y at sea level-high latitude, and reduce exposure ages to about 3/4 of those that have been previously calculated for LGM and younger rocks (to less than 20 cal. ka in the case of Allamuchy). Varying but generally higher solar modulation will require reevaluation of cosmogenic exposure dates in general, as in the case of Allamuchy, that would allow improved synchronization of marine and terrestrial records of glaciation. Other test cases can result in improved GIA deglaciation models and alternative estimates of effects of shielding in ice-flow models.

Wally Broecker's puzzles required radiocarbon dating in order to solve them.

Due to a general scarcity of archives and a lack of temporally highly resolved records, the response of Central European ecosystems to abrupt climate change during the Late Pleistocene - and notably the Last Glacial - has remained insufficiently explored. To contribute to a better understanding of the impact of short-term climate variability on ecosystems in Central Europe north of the Alps during that time, we present a new, near-continuous pollen record from Furamoos, southern Germany. Except for a hiatus from 28 to 20 ka BP, the record spans the past 130 kyrs in centennial-scale resolution. Our pollen dataset is augmented by sedimentological (i.e., XRF core scanning, loss on ignition, and maceral analysis) and macrobotanical data. This new record represents the yet most complete, temporally highest resolved reconstruction of terrestrial ecosystem dynamics in Central Europe for the Last Glacial period. Our results show that temperate forests thrived at Furamoos during full interglacial conditions (i.e., MIS 5e and MIS 1), boreal forests developed during the early glaciation (MIS 5 d-5a), and a tundra vegetation prevailed during full glacial conditions (MIS 4-2). Fourteen transient expansions (contractions) of tree and shrub populations that are synchronous with increases (decreases) in weathering intensity can be firmly attributed to specific Greenland Interstadials (Greenland Stadials) of the NGRIP ice-core record. By comparing individual expansions/contractions of tree and shrub populations with Greenland delta O-18(ice) ice-core and North Atlantic sea-surface temperature (SST) data, we show that during interstadials vegetation dynamics in Central Europe were closely linked to SSTs in the North Atlantic and decoupled from the climate recorded in Greenland; during stadials, the opposite pattern prevailed. We attribute these patterns to changes in AMOC strength and associated latitudinal shifts of the polar front. (C) 2022 Elsevier Ltd. All rights reserved.

Lago Trasimeno (central Italy), 10 km wide and < 6m deep, fills a basin of tectonic origin, and is one of Europe's few endorheic lakes. We report on a multidisciplinary stratigraphic study based on seismic reflection profiles and two sediment cores, aimed at providing information on the vegetational, lithological, and climate history of this area. Trace elements, palynology, macrofossils, organic carbon, and C isotopes, coupled with AMS C-14 dating, describe environmental changes from Late-Glacial to present. The base of the cores records a Late-Glacial steppic vegetation (Poaceae, Artemisia, Amaranthaceae), with dry and cold conditions and a high charcoal/pollen ratio. An ensuing shift to a warmer and moister climate is shown by the rise of Pinus and Quercus, and shallow water aquatics such as Nitella. Early Holocene warming, indicated by Quercus, Oleaceae, Corylus, and Betula, is followed by a hiatus suggesting one or more severe drought events between 9000 and 3000 yr BP. Late Holocene presence of Alnus and Fagus indicates increased moisture, probably in winter, which would have increased lake level. Heavy metal pollution indicators (Pb, Cu, Zn) in the upper portion of the cores imply industrialization. Due to its location, to intensive and uninterrupted anthropogenic impacts since proto historical times, and to its shallow depths, Lago Trasimeno provides an important observation point for climate and environmental changes in the central Mediterranean region during and before Holocene times.

The Hudson Bay Lowlands (HBL) is a vast contiguous peatland extending over >370,000 km2 in Canada's borealsubarctic, and is the traditional land of the Omushkego Cree. It is currently undergoing climatic warming alongside other anthropogenic stressors, and contains a large below-ground carbon pool. Understanding how climate variability and multiple stressors impact peat accumulation in this region is critical to discerning how northern peatlands will respond to future climate and land-use changes. Pollen- and macrofossil-based paleoecological reconstructions, and analyses of aluminum (Al) and titanium (Ti) fluxes in a Holocene-aged peat core (VM375) taken from a bog in the Attawapiskat watershed were conducted to link ecosystem changes with hydroclimate and long-term carbon storage. Peat initiation is dated to 5780 cal yr B.P., coincident with land emergence driven by glacial isostatic adjustment. From 4500 to 4200 cal yr B.P., apparent rate of carbon accumulation increased, and was linked to more rapid rates of peat accretion and increases in minerotrophic indicators in the pollen record. This increase in peat accretion and shift in vegetation composition co-occur with higher rates of mineral influx as shown by Ti and Al concentrations, which may have supplied nutrients. A fen to bog transition takes place -3300 cal yr B.P., with increases in Sphagnum spores and macrofossils, and a decline in the apparent rate of carbon accumulation relative to the earlier half of the record, where paleoecological proxies indicate treed wetland and fen stages. Since peat inception, the total carbon stock of the 260-cm peat column is 110 kg C m- 2. This multi-proxy record shows an association between changing peat types and variability in apparent rates of carbon accumulation, and supports the hypothesis that mineral nutrients either supplied by surface hydrology or by eolian deposition played a role in Holocene peat carbon accumulation in eastern North American boreal peatlands.

Constraining uncertainty in the global carbon cycle requires valid assessment of both surface and stored carbon in marine and coastal ecosystems (Blue Carbon) as well as terrestrial carbon (forests, peatlands, and soils) [Pendleton et al., 2012]. Quantifying the global carbon stock of coastal salt marshes, potentially the most efficient carbon-burying ecosystems in the world per area, is a key area of further research in both of these fields Pendleton et al., 2012. One of the largest challenges is that despite the fact that salt marshes often sequester carbon several meters deep, nearly all estimates of salt marsh carbon stocks consider only the upper 1 m of sediment (Windham-Myers et al., 2015) [54]. This is particularly concerning because coastal wetlands are increasingly at risk due to climate change, sea level rise, and anthropogenic disturbance and destruction (Deegan et al., 2012) [15]. Using full-depth measurements from marsh cores, we estimate the carbon stock of five salt marshes in the highly urbanized estuary of Jamaica Bay, New York and argue that partial-depth measurements can underestimate carbon stocks. These estimates use calculated carbon content and probe depth data of these marshes collected between 2000 and 2019, applying this data across the full area of the marsh obtained from satellite imagery. Carbon density measurements are then multiplied by the full-depth volume of the marshes to create an estimate of total carbon stock. In addition to calculating present-day estimates, we compare our carbon stock estimates to historical Jamaica Bay imagery to calculate historical carbon stocks and carbon loss. The carbon stock estimates presented here show a 95% carbon stock loss between 1885 and 2019 in Jamaica Bay and highlight the severe underestimation of carbon stocks without full-depth calculations. These findings have important implications for disappearing salt marshes with regard to the global carbon cycle and the incorporation of belowground carbon into global climate models. The findings are increasingly relevant for advocacy efforts aiming to conserve these marshes with sea level rise.

1Lamont-Doherty Earth Observatory, Columbia University, Palisades, NY, USA. 2NASA Goddard Institute for Space Studies, New York, NY, USA. ✉e-mail: jnichols@ldeo.columbia.edu The quantification of northern peatland carbon stocks is vital to our understanding of the global carbon cycle, and we welcome the opportunity to respond to the comments submitted in the preceding Matters Arising1,2. In response to Yu et al.1, we provide additional clarification of how data from the Neotoma Paleoecological Database (NPD) are stored and used. We also repeated our quantification of northern peat carbon stocks using only radiocarbon data vetted by other publications. However, our conclusions are unchanged by the elimination of the additional data. We respond to Ratcliffe et al.2 by addressing the conflict between the ‘time-history’ and ‘inventory’ methods. We calculated northern peatland carbon stocks by the inventory method using our dataset and carefully quantify the uncertainties. We find that while the mean values for carbon stock differ by twofold, the uncertainties around those means are larger than previously acknowledged and are overlapping. This result emphasizes the need for continued work by the carbon cycle science community to understand and quantify uncertainty in terrestrial carbon stock estimates.

SignificanceUnderstanding Greenland Ice Sheet history is critical for predicting its response to future climate warming and contribution to sea-level rise. We analyzed sediment at the bottom of the Camp Century ice core, collected 120 km from the coast in northwestern Greenland. The sediment, frozen under nearly 1.4 km of ice, contains well-preserved fossil plants and biomolecules sourced from at least two ice-free warm periods in the past few million years. Enriched stable isotopes in pore ice indicate precipitation at lower elevations than present, implying ice-sheet absence. The similarity of cosmogenic isotope ratios in the upper-most sediment to those measured in bedrock near the center of Greenland suggests that the ice sheet melted and re-formed at least once during the past million years.

Glacial varves can detail ice-margin positions and provide a proxy for meltwater discharge at resolutions comparable to those of the Greenland ice core archives, and thus they can be critical paleorecords for assessing the response of both ancient and modern ice sheets to climate change. Here we provide an ∼1500 yr varve chronology straddling the Younger Dryas (YD)–Holocene boundary (11.65 cal. kyr B.P.), the first such chronology in North America. The varves are from glacial Lake Agassiz (central North America). The chronology is pinned on accelerator mass spectrometry radiocarbon-dated terrestrial macrofossils at the base of a widespread red-clay bed deposited during flooding from the Lake Superior basin. We illustrate the utility of this record by examining ice-margin retreat and melting through the late Younger Dryas and across the Holocene boundary. The ice margin receded at a constant rate, not only during the late YD, but for at least 300 yr after the onset of the Holocene. In contrast, varve thicknesses increased at the boundary, and a moraine formed over a 50 yr period, perhaps in response to the warming climate. Our expectation is that this time series will continue to be developed, expanded, and refined because it promises to be the longest and most geographically extensive glacial varve data set in North America.

Instrumental records indicate a century-long trend towards drying over western North America and wetting over eastern North America. A continuation of these trends into the future would have significant hydroclimatic and socioeconomic consequences in both the semi-arid Southwest and humid East. Using tree-ring reconstructions and hydrologic simulations of summer soil moisture, we evaluate and contextualize the modern summer aridity gradient within its natural range of variability established over the past 600 years and evaluate the effects of observed and anthropogenic precipitation, temperature, and humidity trends. The 2001–2020 positive (wet east-dry west) aridity gradient was larger than any 20 year period since 1400 CE, preceded by the most negative (wet west-dry east) aridity gradient during 1976–1995, leading to a strong multi-decade reversal in aridity gradient anomalies that was rivaled only by a similar event in the late-16th century. The 2001–2020 aridity gradient was dominated by long-term summer precipitation increases in the Midwest and Northeast, with smaller contributions from more warming in the West than the East and spring precipitation decreases in the Southwest. Multi-model mean climate simulations from Coupled Model Intercomparison Project 6 experiments suggest anthropogenic climate trends should not have strongly affected the aridity gradient thus far. However, there is high uncertainty due to inter-model disagreement on anthropogenic precipitation trends. The recent strengthening of the observed aridity gradient, its increasing dependence on precipitation variability, and disagreement in modeled anthropogenic precipitation trends reveal significant uncertainties in how water resource availability will change across North America in the coming decades.