Abstract River nitrogen (N) pollution drives eutrophication and dead zones in more than two‐thirds of global estuaries and bays. Nitrate (NO 3 − ) is a highly mobile form of reactive N. Elevated NO 3 − concentrations in rivers deteriorate water quality and threaten aquatic biota and human health through toxic and carcinogenic effects. Mitigating river NO 3 − in polluted regions is pivotal to achieving Sustainable Development Goal 6 (SDG 6. Clean Water and Sanitation). However, the contributions of major sources to river NO 3 − loads remain poorly resolved, hindering ongoing mitigation efforts. Prevailing source apportionment models attribute river NO 3 − primarily to atmospheric deposition and agricultural non‐point sources. Here we establish a multi‐isotope mass‐balance framework that integrates natural N and oxygen isotopes ( δ 15 N, δ 18 O, Δ 17 O) with concentration, flux, and isotope effect pertaining to river NO 3 − sources, pools, and fates. Application of this framework to polluted rivers in Europe, Asia, and North America reveals that atmospheric deposition contributes only 5 ± 3%, non‐point sources account for 30 ± 9% (lower than previously reported >60%), wastewater remains a dominant contributor at 37 ± 8%, and previously neglected in‐stream nitrification contributes 28 ± 7%, effectively offsetting the fraction formerly overattributed to non‐point sources. This finding demonstrates the importance of strengthening wastewater treatment and controlling in‐river secondary NO 3 − production to reduce river NO 3 − pollution, and significantly advances our knowledge of river NO 3 − sources, which is essential for optimizing mitigation strategies to achieve SDG 6.
Abstract Nutrient inputs can simultaneously regulate microbial necromass formation and soil organic carbon (SOC) accrual, but whether the contribution of microbial necromass to SOC changes over time under nutrient addition, and whether this temporal dependence differs among residue components remains unclear. We synthesized 303 paired observations of microbial necromass and SOC from 50 global studies to quantify the coupling between SOC responses and the responses of total necromass carbon (TNC), fungal necromass carbon (FNC), and bacterial necromass carbon (BNC), and to test how these relationships changed with experimental duration. Across nutrient input types, combined mineral and organic inputs produced the strongest responses, nitrogen, phosphorus, and potassium fertilization combined with manure increasing TNC, FNC, BNC, and SOC by 42.91%, 40.86%, 51.34%, and 44.01%, respectively. Microbial necromass responses were the strongest predictors of SOC responses, yet necromass increased more rapidly than SOC, indicating that enhanced residue production did not translate proportionally into persistent SOC accrual. The contribution of BNC to SOC was comparatively time independent and was mainly regulated by nutrient type. The contribution of FNC showed a convergent trend over time and was significantly influenced by ecosystem type. In contrast, the experimental duration significantly amplified the contribution of TNC to SOC. These findings indicate that long‐term SOC sequestration is an emergent system property arising from the time‐dependent coordination of distinct stabilization pathways. Accounting for this temporal dependence can refine microbial carbon pump theory and provide a globally relevant basis for improving long‐term soil carbon sink projections and nutrient‐management strategies under global change.
Abstract The magnitude of nitrous oxide (N 2 O) emission pulses during the winter‐to‐spring seasonal transition (ST) can exceed summer peaks, but this component remains poorly constrained in the global N 2 O budget. Here, we conducted a synthesis of 182 observations without fertilizer applications using machine‐learning, hierarchical mixed‐effects models, and structural equation modeling to compare ST‐N 2 O pulse magnitude and identify its dominant drivers. The average rate of ST‐N 2 O pulse was 63.2 μg m −2 hr −1 , approximately 3.5 times higher than that during the growing season. Moreover, the average ST‐N 2 O pulse was 108.2 mg m −2 in non‐croplands, which is 2.4 times greater than that in croplands without nitrogen fertilizer (46 mg m −2 ). ST‐N 2 O pulse was influenced by the microbial vernal dam (MVD, defined as the seasonal release of microbial biomass nitrogen), microbial biomass carbon, soil respiration rate, and ammonium content. Soil MVD emerged as the dominant driver of ST‐N 2 O pulse across ecosystem types, accounting for 21% of the observed variation. Cumulative ST‐N 2 O pulses accounted for 2.9% of the nitrogen derived from MVD. These results highlight a ubiquitous hot moment in terrestrial N 2 O emissions and provide a mechanistic basis for improving annual N 2 O budgets and model parameterization.
Abstract Ocean Alkalinity Enhancement (OAE) is a marine carbon dioxide removal (CDR) strategy with a theoretical sequestration potential of several Gt . The long‐term durability of OAE‐induced carbon storage depends on the persistence of the added alkalinity in the ocean, which is influenced by sedimentary processes and biogeochemical and physical feedbacks. Using the University of Victoria v2.10 Earth System Model of intermediate complexity, we investigated the millennial‐scale durability of OAE‐induced extra alkalinity and associated additional carbon storage and, specifically, the role of deep‐sea calcium carbonate sediments in limiting the durability. We conducted 10,000‐year simulations for four emissions scenarios, each combined with a global alkalinity addition experiment (0.135 Pmol for 50 years) with and without interactive carbonate sediments. Neither the added alkalinity nor the associated oceanic carbon storage remained fully stable on millennial timescales compared to a baseline simulation. By year 10,000, 35%–59% of the initial alkalinity increase through OAE is lost due to dynamics in deep‐sea sediments, resulting in a reduction of the OAE‐induced carbon storage by 14%–27% compared to simulations without sediments. Enhanced alkalinity reduces the natural sedimentary climate feedback associated with dissolution. Additionally, outgassing to the atmosphere as a response to reduced atmospheric has been identified as a limiting factor for the durability of OAE‐induced carbon storage in the ocean, possibly affecting all CDR methods. Episodic Southern Ocean deep convection events further contributed to variability in Earth system responses during global cooling.
Abstract Mapping organic carbon (OC) stores in marine sediments is a management priority to potentially minimize the anthropogenic release of organic carbon from the seabed. Sediment OC content shows high regional variability, shaped by a complex interplay of physical, chemical, biological and anthropogenic factors. In this study, we present an assessment of the drivers and patterns of sediment OC across European regional seas, with a particular focus on the role of large‐scale habitats, substratum characteristics, and environmental conditions. To investigate these relationships, we integrated data sets encompassing sediment OC, benthic habitats, sediment substrata and 11 environmental variables. Applying machine learning techniques, we identified wave exposure (estimated from wave fetch), light availability, seafloor geomorphic features, and maximum bottom water temperature as the most important environmental predictors of sediment OC content across European regional seas. While the model captured broad spatial patterns in sediment OC, it did not resolve local scale variability, making it better suited for regional assessments than for site‐specific predictions. Overall, the highest OC contents were found in sediments underlying vegetated habitats such as Posidonia oceanica and other seagrass meadows as well as in mud, and mixed sediment substrata. OC hotspots were generally located in inshore areas with low wave fetch and temperature; these include areas of the Baltic Sea, North Sea, Adriatic Sea, Barents Sea, and the Black Sea. Our data‐driven approach provides a robust foundation for identifying OC‐rich sediments, which are critical for conservation planning and assessing anthropogenic impacts on marine OC.
Abstract The biological, chemical, and physical processes in the mesopelagic zone that impact carbon flux to the deep ocean are poorly understood, particularly in the Southern Ocean, despite their importance to deep ocean carbon storage. Using a CTD‐mounted underwater imaging system, the Underwater Vision Profiler 5, we observed distributions of large marine particles across two basin‐scale global ocean ship‐based hydrographic investigations program repeat hydrography transects in the Pacific and African sectors of the Southern Ocean. Utilizing deep learning features and unsupervised clustering, 2.6 million images of particles were classified by type into 46 detrital categories within four supercategories (e.g., fluffy aggregates, dense aggregates, fibers, feces) and 61 living categories within four supercategories (e.g., crustaceans, rhizarians, gelatinous zooplankton, Trichodesmium ). Different frontal zones generally have distinct patterns in particles, which illuminate different flux pathways. We observed a “Fluffy Aggregate Flux Pathway”, where high fluffy aggregate abundance was observed from the surface to 3,000 m at the same time as elevated surface chlorophyll a . We also observed a “Zooplankton Mediated Flux Pathway,” where grazers were located primarily in the lower mesopelagic zone, and feces and dense aggregates were in high abundance in the abyssopelagic zone. These results highlight two export pathways in the Southern Ocean, both of which contribute to long‐term carbon storage at depth.
Abstract Global nitrogen deposition has steadily increased since the 1980s, peaking around 2015 before stabilizing. However, atmospheric chemical transport models often underestimate its magnitude, limiting accurate assessments of its impacts on terrestrial gross primary productivity (GPP). In this study, we elucidated the drivers of interannual GPP variability and quantified the contribution of nitrogen deposition from 1980 to 2020 using the latest global nitrogen deposition data set, the TRENDY GPP products, and an interpretable machine learning framework (SHAP). Our findings revealed a consistent expansion in global GPP over the past four decades, averaging 156.95 ± 6.4 Pg C yr −1 . As nitrogen deposition has recently stabilized, its contribution to global GPP has shifted from a stimulatory effect to a source of negative fluctuations. This transition is attributable to the declining sensitivity of GPP to nitrogen deposition, potentially driven by increasing vegetation water stress. Climatic factors, primarily temperature and precipitation, dominate interannual GPP fluctuations across plant functional types (PFTs), and nitrogen deposition explains 8.7 ± 2.9% of global variability. Notably, nitrogen enrichment stimulated GPP in grasslands and croplands but had an inhibitory effect on tropical forests. Moreover, the nonlinear response of GPP to nitrogen deposition exhibited distinct optimal thresholds across PFTs, with the global vegetation optimum identified at 12.5 kg N ha −1 yr −1 . Crucially, the direct effect of nitrogen deposition on GPP outweighed its synergistic interactions with climate and CO 2 concentrations, suggesting that nitrogen availability independently modulates terrestrial carbon sinks. This study underscores the biome‐specific sensitivities to nitrogen loading and highlights the necessity of incorporating nitrogen‐saturation thresholds into predictions of ecosystem feedback to global change.
Abstract Riverine dissolved organic carbon (DOC) plays a vital role in the global carbon cycle, but its spatiotemporal dynamics and controlling mechanisms in alpine permafrost regions remain unclear. Here, by integrating a comprehensive in situ DOC data set, multi‐source environmental variables, and machine learning approaches, we reconstruct the riverine DOC concentration ( C DOC ) and flux ( F DOC ) at five headwater rivers on the eastern Tibetan Plateau (TP) from 2000 to 2024 and investigate their spatiotemporal patterns. Results show that the mean C DOC across the five rivers is 3.27 ± 0.96 mg/L but remains highly heterogeneous, with higher values observed in the permafrost‐dominated headwaters of the Yellow (4.31 ± 0.78 mg/L) and Yangtze Rivers (3.39 ± 0.86 mg/L). The permafrost coverage, soil organic carbon content, and vegetation type shape this spatial pattern. The southeastern rivers (Mekong and Salween Rivers) exhibit declining C DOC over the past 25 years, primarily driven by soil moisture reduction under climate warming. In contrast, significant C DOC increases are detected in Yellow (+0.073 mg L −1 decade −1 ) and Yangtze headwaters (+0.029 mg L −1 decade −1 ), due to the enhanced vegetation conditions associated with climate warming and permafrost thaw. Total F DOC at the outlets of major rivers increases markedly, reaching approximately 1.37 Tg yr −1 in response to the rising river discharge. This study presents the spatiotemporal dynamics of DOC across TP rivers, elucidates the underlying mechanisms, and provides a basis for land‐river carbon transfer and regional carbon budget assessments in high‐elevation permafrost environments.
Abstract The Editors of Global Biogeochemical Cycles express their appreciation to those who served as peer reviewers for the journal in 2025.
Abstract Atmospheric N 2 O, the dominant ozone‐depleting substance and a potent greenhouse gas, has risen notably over recent decades. Using a process‐based model, we simulated the sensitivity of N 2 O emissions from undisturbed natural soils to the representation of historical land‐cover change from 1990 to 2023 by factoring the effects of atmospheric nitrogen deposition, biological nitrogen fixation, rock weathering, and soil uptake. Simulations based on the potential vegetation distribution data show average emissions of 6.24 ± 0.72 Tg N yr −1 with a significant upward trend (+0.021 Tg N yr −1 ) from 1990 to 2023, while European Space Agency Climate Change Initiative land cover estimates were 30% lower, underscoring the sensitivity of N 2 O estimates to land‐cover representation. Most emissions stem from background processes (69.8% ± 4.5%), followed by nitrogen fixation (16.3% ± 2.6%), atmospheric deposition (9.2% ± 1.3%), and rock weathering (4.7% ± 1.1%). The accelerating trend of N 2 O emissions under changing climate conditions and key uncertainties are tied to used land‐cover data sets, limited observation data, and unresolved hydrological extremes.
Abstract Coastal blue carbon (CBC) ecosystems export organic matter offshore, but their contribution to deep‐ocean carbon sequestration remains insufficiently quantified. We measured depth‐resolved particulate organic carbon (POC) fluxes, δ 13 C signatures, and environmental DNA (eDNA) from sediment traps deployed at 150, 500, and 1,000 m in the western North Pacific warm pool. POC fluxes at 1,000 m were two‐ to four‐fold higher than predicted by classical attenuation models, indicating efficient deep‐water transfer driven by both vertical export and lateral inputs. eDNA from sinking particles revealed macroalgae, mangrove, and seagrass taxa throughout the mesopelagic and bathypelagic layers, demonstrating long‐distance transport of CBC material. δ 13 C values showed predominantly marine‐derived POC, but with measurable CBC contributions persisting at depth. These observations identify a previously underappreciated pathway linking coastal vegetation to deep‐ocean carbon storage in oligotrophic gyre systems. They also highlight the need for ground‐truth biological carbon pump (BCP) measurements, particularly for oceanic island countries with small land areas yet large exclusive economic zones (EEZs), where accurate carbon accounting depends on in situ validation of model‐based estimates.
Abstract In this special collection, new studies reveal that fjords serve as vital estuarine transition zones between terrestrial ecosystems and the open ocean. Primarily found at high latitudes, fjords are highly sensitive to climate change, which fundamentally reshapes their land and marine inputs. As glaciers retreat, the influx of freshwater, nutrients, and sediments shifts, altering the chemical and biological makeup of fjord waters. While glaciers currently drive the upwelling of nutrient‐rich waters that support primary productivity, glacial retreat onto land may increase stratification and shift ecosystems toward microbial dominance. Fjords also play a critical role in the global climate system as disproportionate hotspots for organic carbon burial. However, warming and melting permafrost threaten to remobilize this stored carbon, potentially changing fjords from carbon sinks into sources of CO 2 emissions. Because of these complex and interacting stressors, fjords are increasingly viewed as “Aquatic Critical Zones” that require urgent interdisciplinary research and management.
Abstract Multiple global change factors (GCFs) often co‐occur, yet their interactive effects on soil greenhouse gas (GHG) emissions worldwide remain poorly understood. Here, we conducted a global meta‐analysis to assess pairwise effects of five GCFs (elevated CO 2 , nitrogen addition, warming, increased precipitation, and decreased precipitation) on soil CO 2 , CH 4 , and N 2 O emissions. We found that individual GCFs significantly increased CO 2 and N 2 O emissions but had negligible effects on CH 4 emissions. Similarly, pairwise combinations of GCFs enhanced CO 2 and N 2 O emissions, whereas CH 4 responses were weak and directionally inconsistent. Across all GCF combinations, additive interactions (69.7%–82.5%) predominated overall, while synergistic and antagonistic interactions were less frequent and context‐dependent. Notably, N 2 O exhibited the highest proportion of both synergistic (13.1%) and antagonistic (17.2%) effects compared to CO 2 and CH 4 , indicating stronger nonlinear responses to concurrent GCFs. Moreover, roughly 50% of the multifactor responses for N 2 O emissions in wetlands and croplands were non‐additive, underscoring that ecosystem‐specific context strongly modulates interaction outcomes. Climatic and ecosystem characteristics, particularly precipitation, modulated interactive effects. These findings highlight that predicting future soil GHG dynamics under concurrent global change requires accounting not only for generally dominant additive effects but also for the context‐dependent nonlinear interactions that ultimately constrain Earth system projections.
Permafrost degradation under a warming climate has altered hydrological and biogeochemical processes across the Arctic. Although increasing fluxes of weathering‐derived ions (e.g., Ca 2+ , Mg 2+ , and SO 4 2− ) have been reported in Arctic rivers, the underlying mechanisms and hotspots within basins remain poorly understood due to limited analysis of environmental drivers such as climate and permafrost dynamics. We investigated long‐term trends (1980–2022) in Ca 2+ , Mg 2+ , and SO 4 2− concentrations in the Kolyma River in northeastern Siberia, and examined basin‐wide changes in air temperature, precipitation, soil temperature, and active layer thickness. We found significant increases in ion concentrations, which were strongly correlated with rising subsurface soil temperatures ( r = 0.61) and active layer deepening ( r = 0.78) in the Yedoma‐rich Kolyma Lowland. These findings, along with the concentration ratios, suggest that sulfuric‐acid‐driven carbonate weathering has intensified in the deeper part of the active layer—where previously frozen minerals become newly exposed—thereby enhancing ion discharges to rivers. Record‐high concentrations were observed in 2020, when an extreme heatwave occurred, and produced exceptionally high subsurface soil temperatures (5.6°C; average 3.8 ± 0.6°C) during the thawed period (May–October) and active layer thicknesses (116.5 cm; average 100.3 ± 10.8 cm). These results underscore the sensitivity of Arctic river systems to heatwave‐induced permafrost degradation, which rapidly intensifies subsurface weathering and solute mobilization. Given the widespread distribution of Yedoma across the Arctic, similar responses may occur in other watersheds. Continued monitoring of water chemistry and permafrost dynamics is essential to understand changes in Arctic river biogeochemistry.
Abstract Lateral transport of terrestrial carbon—via harvested biomass, bioenergy supply chains, and riverine export—redistributes atmospheric CO 2 uptake and subsequent CO 2 emissions across space and time, complicating regional carbon budgeting and atmospheric inversion estimates. We present Lateral Accounting of Transport in Terrestrial Ecosystems (LATTE), a high‐resolution (5 arc‐min) global gridded data set of annual land‐atmosphere CO 2 fluxes attributable to lateral carbon transfers for 1961–2022. LATTE quantifies paired CO 2 sinks at production locations and compensatory CO 2 sources at receptor locations, for four major lateral transport mechanisms: (a) crop harvest, trade, and consumption; (b) industrial roundwood harvest, trade, and storage in harvested wood product pools with country‐specific decay and delayed emissions, (c) biofuel production, and use, including crop biofuels and fuelwood; and (d) the inland water carbon transport loop, including soil‐to‐river leaching, river and lake CO 2 evasion, aquatic burial, and export to the ocean. National production and trade statistics and biofuel energy balances are converted to carbon units and downscaled using satellite net primary productivity, forest carbon removal maps, population density, livestock distributions, and gridded biofuel combustion proxies. Inland water fluxes are based on published climatology with basin‐scale mass balancing across major river basins. Example maps illustrate distinct spatial patterns of sinks and sources associated with crop products, wood products, biofuels, and inland waters, with global totals consistent with previous assessments. LATTE enables improved representation of lateral carbon transport in atmospheric inversions and supports spatially explicit regional and national carbon budget analyses.
Boreal forests are a major contributor to the global land carbon sink under rising CO2 concentrations and a changing climate. Carbon sink estimates for Northern Eurasia from forest inventories, flux mapping, and remote sensing have moved toward convergence over the past decade, although substantial differences remain. Several bottom-up and top-down estimates exceed Russia's national greenhouse gas inventory values. Here, we combine data from four independent sources-machine-learning-based FLUXCOM, atmospheric inversion models, the TRENDY ensemble of dynamic global vegetation models, and CMIP6 Earth System Models-to assess the spatial and temporal characteristics of the Northern Eurasian net carbon sink and discuss the differences. Our multi-approach assessment yields a regional estimate of 0.47 +/- 0.20 GtC year-1 for the 2001-2015 period, which accounts for one-third of the global land carbon sink. CMIP6 model estimates are broadly consistent with those from other approaches, lending confidence to their future projections, although regional differences persist across individual models. We find a pronounced spatial flux gradient from south to north and west to east along a mean temperature gradient, with stronger carbon sinks in the warmer southern regions and weaker sinks in the cooler northeastern regions. Despite rapid warming in the northern parts of Northern Eurasia, CO2- and warming-induced carbon sink enhancement contributes little to the overall regional sink due to its initially low net productivity. These results underscore the importance of forest productivity in shaping the terrestrial carbon sink and provide a multi-perspective view of its evolution under continued anthropogenic forcing.
We present a data set of macronutrient and dissolved and total dissolvable trace metal concentrations collected in June 2012 across the Nordic Seas. Surface water and depth profiles of manganese, iron, cobalt, nickel, copper, zinc, phosphate, silicic acid, and NOx (nitrate + nitrite) reveal how warm saline Atlantic waters and cold fresh Arctic waters shape regional (micro)nutrient availability and phytoplankton growth conditions. At the surface, Arctic waters, depleted in NOx (3.44 +/- 3.30 mu M) but enriched in Si(OH)4 (2.15 +/- 1.42 mu M) and dissolved trace metals (e.g., up to 1.03 nM dFe during peak to post-bloom), reach the western Nordic Seas via the Transpolar Drift and East Greenland Current. Conversely, the North Atlantic Current transports waters enriched in NOx (up to 10.7 mu M) to the surface high-latitude North Atlantic (HLNA) and eastern Nordic Seas. These contrasting inputs create distinct nutrient-deficiency regimes: The western Nordic Seas were N deficient, whereas the HLNA and eastern Nordic Seas showed Fe deficiency following the spring bloom. Benthic flux and internal cycling processes, including remineralization and scavenging, further shape (micro)nutrient distributions in the deep waters, in addition to their pre-formed supply in water mass formation regions. The (micro)nutrient fluxes across the Greenland-Scotland Ridge, combined with reported estimates from the Fram Strait and Barents Sea Opening, highlight the Nordic Seas as a transitional region in mediating (micro)nutrient exchange between the Arctic Ocean and HLNA. Future changes in Arctic export may alter elemental stoichiometry in the Nordic Seas, with implications for primary production and biogeochemical cycling.
Earth's arid and semi-arid regions have been hypothesized to contribute significantly to interannual variability of the global terrestrial carbon sink. Arid and semi-arid regions such as the Western U.S. also show high vulnerability to climate extremes in the form of droughts, heatwaves, and large forest fires, compelling a need to quantify their climate-carbon responses. We quantify the net ecosystem exchange of (NEE) using a high-resolution regional inverse model with constraints from both in situ and space-based observations during 2015-2016. Posterior fluxes are evaluated against withheld aircraft observations across the North American observation network. Observationally constrained fluxes suggest moisture-driven seasonal compensation in NEE across the 2 years, evidenced by a decrease in NEE of 244-262 [Tg C] during the early carbon uptake period and an increase in NEE of 281-389 [Tg C] later during the carbon uptake period, relative to 2016. Atmospheric constraints on NEE, combined with remote sensing and machine learning-based upscaled gross primary productivity (GPP) products, allow for benchmarking of net and gross fluxes as estimated by a suite of terrestrial biosphere models. These benchmarks allow us to link changes in NEE and GPP constrained by atmospheric and space-based observations to changes in stomatal conductance and water use efficiency. High-precision in situ observations, such as those from the Global Greenhouse Gas Reference Network, are key constraints for bias-free carbon flux estimates.
Understanding the mechanisms that drive the mobilization and fate of organic carbon (OC) in Arctic landscapes is important for modeling the feedbacks among permafrost thaw, carbon cycling, and climate change. While significant progress has been made toward measuring in situ OC decomposition in permafrost soils and bulk particulate organic carbon (POC) export from Arctic rivers, few studies have distinguished the source and lability of POC across Arctic landscapes, limiting our ability to predict whether mobilized POC will be oxidized to CO2 and CH4 or buried in downstream depositional environments. This study uses ramped pyrolysis/oxidation radiocarbon (RPO-14C) analyses to investigate spatial and temporal variations in the thermochemical stability and radiocarbon content of fluvial POC during downstream transport from mountains to the coast in the Canning River (North Slope, Alaska). Fluvial POC in the headwaters is predominantly comprised of high activation energy, thermally recalcitrant petrogenic OC (OCpetro) derived from shale bedrock. Moving into the foothills and low-relief coastal plains, river bank erosion primarily drives mobilization of labile, low activation energy, soil-derived OC (OCsoil). Fluvial POC in mountainous upstream reaches consisted of similar to 70% OCpetro and just similar to 30% OCsoil, while POC in the downstream coastal plain reaches comprised similar to 85% OCsoil and similar to 15% OCpetro. The high relative lability of POC exported to the coast indicates high susceptibility to oxidation and microbial decomposition, which could enhance CO2 release as the Arctic hydrologic cycle intensifies. However, the persistence of refractory OCpetro in the suspended load indicates the potential for long-term burial of rock organic carbon in marine sediments.
We present measurements of silicon isotopes in silicic acid, delta 30Si, from GEOTRACES GP17OCE between 20 degrees S and 60 degrees S along 152 degrees W. The section sampled key water masses, allowing assessment of controls on large-scale delta 30Si features. Observational data paralleled results of a data-constrained Si isotope model and are consistent with nutrient trapping and partial nutrient consumption in polar surface waters leading to the accumulation of light silicon isotopes in the deep Southern Ocean (SO) and transport of the residually heavy fraction northward within Sub-Antarctic Mode Water (SAMW) and Antarctic Intermediate Water (AAIW). Deep-water delta 30Si values were nearly constant along isopycnals in the South Pacific, with vertical gradients driven by an increasing fractional contribution of isotopically light regenerated silicic acid with depth. Isotope values in deep waters from the SO to Tahiti exhibited a greater range than previously observed in the South Pacific or predicted by models, with delta 30Si falling between 1.19 parts per thousand and 1.45 parts per thousand. SAMW and AAIW exhibited relatively heavy, uniform delta 30Si. Model results suggest that these mode waters are composed of preformed silicic acid subducted in the outcropping region and regenerated silicic acid sourced in roughly equal measure from polar waters and northern distal sources. Distal input is mainly through entrainment of underlying Upper Circumpolar Deep Water into AAIW.
