Controls on the abundance, provenance and age of organic carbon buried in continental margin sediments.

Earth and Planetary Science Letters 558 (2021) 116759 Contents lists available at ScienceDirect Earth and Planetary Science Letters www.elsevier.com/locate/epsl Controls on the abundance, provenance and age of organic carbon buried in continental margin sediments Blanca Ausína,b,∗, Elena Brunia, Negar Haghipoura,c, Caroline Weltea,c, Stefano M. Bernasconia, Timothy I. Eglintona aGeological Institute, ETH Zurich, Switzerland bGeology Department, Salamanca University, Salamanca, Spain cLaboratory of Ion Beam Physics, ETH Zurich, Switzerland a r t i c l e i n f o a b s t r a c t Article history: Received 20 March 2020 Received in revised form 7 January 2021 Accepted 9 January 2021 Available online xxxx Editor: Y. Asmerom Keywords: organic carbon radiocarbon age hydrodynamic sorting grain size continental margins Continental margins play a fundamental role in the carbon cycle as primary oceanic locations of organic carbon (OC) burial. However, gaps remain in our understanding of factors controlling the distribution and preservation of organic matter (OM) in these heterogeneous and dynamic systems. In particular, the impact of hydrodynamic processes on the age, abundance, and stable isotopic composition of sedimentary OC is poorly constrained. Here, we characterize the OC present in bulk and grain-size sediment fractions from seven continental margin settings. Our results reveal that hydrodynamic particle sorting processes exert a ubiquitous influence on the radiocarbon age of OC. Both, hydrodynamic characteristics of mineral particles and the nature of their interactions with OM influence sedimentary OC content, whereas no significant influence of either effect is manifested in corresponding δ13C values. Since OC preferentially resides within the fine silt fraction (2-8 μm), and this fraction accounts for a substantial fraction of the bulk sediment mass, translocation and subsequent re-deposition of distant fine silt has the greatest potential to distort local OC signatures relative to those associated with clay or coarse silt fractions. We suggest that the magnitude of differences in 14C-age and OC content among grain-size fractions, determined by the interplay of hydrodynamic sorting and other site-specific processes, allow three different categories of depositional environment to be defined: initial, stable, and mature. Each domain is characterized by different degrees of vertical and lateral OC supply that reflect influences of local biological productivity and carbon export from overlying surface waters and physical forcing that drive hydrodynamic processes. This generic framework may serve as a guide to refine assessment of OC burial and to constrain the magnitude of potential aliasing among co-eval proxy signals in continental margin sedimentary sequences. ©2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 1. Introduction Although continental margins account for only 10-20% of the global ocean floor, these regions store ca. 90% of the organic carbon (OC) preserved in marine sediments (Hedges and Keil, 1995; Premuzic et al., 1982), and thus play a key role in the global carbon cycle as major OC sinks. Moreover, continental margins are characterized by rapid sediment accumulation, and thus organic signatures preserved in marine sediment sequences retrieved from these spatially heterogeneous and highly dynamic regions are of strategic value in studies of continent-ocean interactions and pa- *Corresponding author at: Geology Department, Salamanca University, Salamanca, Spain. E-mail address: [email protected] (B. Ausín). leoclimate variations at centennial or even higher temporal resolution. A detailed understanding of the processes affecting OC deposition and burial in these continent-ocean boundary environments is a prerequisite both from the perspective of understanding the C cycle and interpretation of sedimentary records. Diverse, and frequently coupled, factors contribute to the efficiency of remineralization and nature of organic matter (OM) preserved in marine settings. Sedimentation rate, oxygen exposure time (OET) and microbial activity, to name a few, have been found to influence OC burial efficiency (e.g., Arndt et al., 2013; Müller and Suess, 1979). The observation that OC content is broadly correlated with mineral-specific surface area of continental margin sediments led Mayer (1993)to propose OM sorption on mineral surfaces as a mechanism for its long-term physical protection. Further work highlighted the preferential association of OM with fine-grained, large-surface-area minerals, as these provide stronger https://doi.org/10.1016/j.epsl.2021.116759 0012-821X/©2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). B. Ausín, E. Bruni, N. Haghipour et al. Earth and Planetary Science Letters 558 (2021) 116759 Fig. 1. Locations of the surface sediment samples analyzed in this study and sample SHAK06-5K; 2-3 cm from Magill et al. (2018). The corresponding acronym for each site is shown within brackets. protection against degradation (Hedges and Keil, 1995; Keil and Mayer, 2014; Keil et al., 1994b; Mayer, 1994a,b). However, such associations with fine-grained minerals may expose OC to further remineralization if strong hydrodynamic forcing comes into play (Bao et al., 2016, 2018a,c). In addition to controlling OC content and composition (Bergamaschi et al., 1997), different mineral particles may exhibit cohesive behavior or may be prone to selective mobilization depending on their size, and on prevailing hydrodynamic conditions (e.g., shear stress). Besides, the potential impacts of hydrodynamic sorting processes on OM dispersal and burial over continental margins, these processes raise concerns with respect to the fidelity of organic signatures preserved in continental margin sediment records (Inthorn et al., 2006). Such processes may induce spatial and temporal offsets between sedimentary constituents residing in different grain size fractions. Indeed, lateral sediment transport processes have been invoked to explain older 14C ages of bulk OC relative to codeposited planktonic foraminifera (e.g., Ausín et al., 2019; Kusch et al., 2010; Mollenhauer et al., 2005; Ohkouchi et al., 2002). Definitive evidence for this hypothesis requires in-depth investigations at the mineral grain-size level in order to attribute effects to hydrodynamic sorting. Recent studies have revealed the large extent to which OC transport, (re)distribution and preservation are influenced by hydrodynamic sorting processes in the shallow Chinese marginal seas and on the Washington margin (Bao et al., 2019a, 2016, 2018b). The impacts of these processes on recorded paleoclimate signals in a sediment core from the Iberian margin have also been documented (Magill et al., 2018). However, these studies remain limited to few specific regions, and we currently lack knowledge of the broader relevance of these processes with respect to sedimentation and OC burial on continental margins. Here, we examine geochemical characteristics (OC-content, -radiocarbon age, -carbon isotopes, C/N ratio, mineral surface area and grain size) of discrete sediment grain-size fractions collected from a range of continental margin settings and one deep-sea sediment drift in order to assess the prevalence and nature of hydrodynamic mineral-sorting influences on sedimentary OC. 2. Material and methods 2.1. Sample description We selected sediment samples from continental margin systems spanning a range of hydrographic and depositional settings that vary with respect to surface ocean productivity, sedimentation rate, and oxygen levels. All sites are distal from the influence of major rivers to diminish the potential contribution of terrestrial OC and characterized by rapid sedimentation (>20 cm kyr−1) to minimize bioturbation effects (Fig. 1). The Peruvian margin in the equatorial Pacific is one of the most productive regions of the World Oceans (100–400 g C m−2yr−1, (Suess, 1973)) as a result of wind-driven upwelling. Sample KNR 182/9 MC13 (abbreviated here to “PER”) was retrieved from the outer edge of the continental shelf, where large fluxes of labile carbon and resulting oxygen-deficient waters result in very high bulk OC contents of surface sediments (up to 16%) while OM decomposition hinders carbonate preservation (Arthur et al., 1998). The Namibian margin in the southeast Atlantic Ocean also sustains high biological productivity and an intense oxygen minimum zone (OMZ) over the shelf (100-400 m depth) due to the permanent Benguela Upwelling System (Moffitt et al., 2015; Schimmelmann et al., 2016). Sample Mirabilis 2017 26-90 (“NAM”) was retrieved in oxic waters from the mid-slope, where hemipelagic sedimentation, including substantial lateral transport of aged OM has previously been documented (Inthorn et al., 2006). The NW African margin in the subtropical eastern Atlantic is part of the Canary Upwelling System, characterized by high to moderate rates of biological productivity. Sample OC437 MC31 (“NAF”) was retrieved north of the OMZ that extends south of 21◦N (Moffitt et al., 2015) and selected considering this region may exhibit minimal hydrodynamic biases given synchronous signals have been observed between OC and planktonic foraminifera in underlying sediments (Mollenhauer et al., 2005), contrary to modeling studies (Fischer et al., 2009). Sample OCE326 MC2 (“NAT”) was recovered from a shallow depression located on the southern New England Mud Patch, in the Mid-Atlantic Bight continental shelf. This region sustains high to moderate productivity and is characterized by high rates of finegrained active deposition (Twichell et al., 1981). The Santa Barbara Basin and Santa Monica Basin are located near the coast on the California margin in the northeast Pacific Ocean, within a major eastern boundary upwelling system. High productivity and partially restricted circulation resulting in oxygendepleted bottom waters promote deposition of varved sediments with minimum bioturbation in these borderland basins (Schimmelmann et al., 2016). Sample SBB4 MC43 (“SBB”) was recovered from the partially isolated and silled Santa Barbara Basin, characterized by sub-oxic to anoxic bottom waters and where evidence of sediment focusing and deposition of pre-aged OC has previously been documented (Mollenhauer and Eglinton, 2007). Sample SMB1 MC49 (“SMB”), located in the Santa Monica Basin, was recovered from the well-defined OMZ between 500 and 900 m depth (Moffitt et al., 2015). The Bermuda Rise is a deep-ocean sediment drift in the subtropical northwest Atlantic Ocean. Although this site is not located on a continental margin, sample OCE326 BC9 (“BER”) was selected for this study given that the largest 14C age discrepancies between OC and co-deposited planktonic foraminifera have been reported here (Ohkouchi et al., 2002), implying strong hydrodynamic control on sediment properties in this region. Results previously reported from sample SHAK06-5K; 2-3 cm (“IBE”), retrieved from the so-called “Shackleton Sites” of the southwest Iberian Margin (Magill et al., 2018) were included in this study. This site is located on the lower part of the slope and features high sedimentation rates due to high to moderate primary productivity and sediment focusing. Except for the Iberian margin, which was sampled using a kasten corer, the sediment-water interface was preserved at all sites, sampled using a multi-corer (PER, NAM, NAF, NAT, SBB and SMB) and a box-corer (BER). 2.2. Grain-size fractionation Sediment cores were sectioned on-board every 1 cm, and sealed and stored in glass or plastic containers at −20 ◦C until they were 2 B. Ausín, E. Bruni, N. Haghipour et al. Earth and Planetary Science Letters 558 (2021) 116759 further processed for analyses. Between 2 to 5 top cm from each core (Table 1) were freeze-dried, homogenized, and fractionated into four different grain size fractions, namely sand (>300-63 μm); coarse silt (63-10 μm [CS]); fine silt (10-2 μm [FS]); and clay (<2μm). Additionally, planktonic foraminifera were picked from a sub-fraction (250-300 μm) of the sand fraction. The grain-size fractions were obtained following the protocol introduced by Magill et al. (2018). Only Milli-Q water was used for all steps during sample preparation. Prior to use, all glassware was heat-treated overnight at 450◦C, whereas sieves and other metal tools were cleaned with organic solvents. Between 40-200 g of dry sediment per sample were wet-sieved with a sieving tower using sieve mesh-sizes of 20, 63, 250, 300 μm. The sand fraction containing mostly foraminifera and the CS (20-63 μm) fraction were transferred to glass jars. Material passing the 20 μm was then transferred to centrifuge vials and centrifuged at 160 RCF 2 min at maximum speed and 2 min to slow down. The supernatant containing the clay fraction (<2μm) was recovered and transferred into glass jars. This cycle was repeated until the supernatant was clear. To separate the fine silt fraction (2-10 μm) from the lower end of the CS fraction (1020 μm), the remaining fraction (2-20 μm) was settled in a volumetric flask. Reference settling velocities for the 10 μm cutoff were adopted from Gibbs et al. (1971). Accordingly, the 2-10 μm fraction was pipetted and material between 10-20 μm was combined with the 20-63 μm fraction to obtain the CS fraction (10-63 μm). All fractions were then immediately transferred to glass jars and frozen at −20◦C prior to freeze-drying for further analyses. The sand fraction was excluded from the study as microscopic analysis showed that this fraction contained discrete particulate organic debris not bound to mineral surfaces (Keil et al., 1994b; Ransom et al., 1998) as well as OM bound within planktic foraminifera. As the latter are assumed to settle rapidly without substantial lateral advection, their presence may mask the impact of hydrodynamic sorting on mineral associated OM within this fraction. 2.3. Grain size and mineral-specific surface area (SA) analyses Between 100-200 mg of freeze-dried bulk sediment were heated in an oven at 350 ◦C for 12 h and cooled down to 40 ◦C at 50◦C per hour in order to remove OM. Samples were subsequently suspended in a solution of 4 mL of sodium hexametaphosphate (1 g L−1) in Mili-Q water and kept in a shaking platform until grain-size analyses were performed with a Mastersizer 2000. Polystyrene latex and spherical glass beads of 300 nm-100 μm were used as reference materials. Samples were measured in triplicates under conditions of repeatability and the average and precision are reported. Aliquots of solvent-extracted sediments were heated to 150 ◦C for 24 h to ensure OM removal and cooled to room temperature for ca. 45 min. Nitrogen-based BET (Brunauer–Emmett–Teller) surface area determinations were made on sample aliquots of ∼30 mg on a Quantachrome NOVA 4000e. Degassing was performed with a Quantachrome FLOVAC degasser at 150 ◦C under vacuum overnight. Precision was >±0.04 m2/g and <±1.00 m2/g determined from replicate measurements of lowand high-surface area Quantachrome instruments standards, respectively. 2.4. Radiocarbon analyses of planktonic foraminifera and OC All radiocarbon measurements were performed using gas ion source accelerator mass spectrometry (AMS) on the compact Mini Carbon Dating System (MICADAS) at the Laboratory of Ion Beam Physics, ETH Zürich. An aliquot (12-70 mg) of freeze-dried sediment was fumigated with concentrated HCl (37%, 72 h) to remove inorganic carbon and Table 1 Sample locations. Mean annual net primary productivity (NPP) was based on chlorophyll-a, SST, and PAR satellite measurements of the MODIS mission (2003-2016) data. Average sedimentation rate (SR) is adopted from previous works conducted at identical or nearby locations and water depths. Location [acronym] Sample name Cruise/year Longitude Latitude Water depth [m] NPP [mg C m−2day−1] Average SR [cm kyr−1] Reference Depositional setting/Oxygen conditions Peruvian margin [PER] KNR 182/9 MC13 0-3 cm KNR 182/9 2005 −78.17 −11.00 326 2773 20.8 Wefer et al. (1990)Outer continental shelf/Anoxic (OMZ impingement) Namibian margin [NAM] 2017 26-90 0-3 cm Mirabilis May 2016 13.3 −26 1277 1431 22 Inthorn et al. (2006)Mid-slope/Oxic NW African margin [NAF] OC437 MC31 0-3 cm Cheeta Cruise 2007 −13.74 27.54 1090 1377 20.3 Mollenhauer et al. (2005)Upper continental slope/Oxic NW Atlantic margin [NAT] OCE326 MC2 0-3 cm Bermuda Rise 1998 −70.54 40.46 80 1276 32 Bothner et al. (1981)Shelf depocenter/Oxic Santa Barbara Basin [SBB] SBB4 MC43 0-2 cm New Horizon 2001 −119.87 34.33 340 1172 200 Schaaf and Thurow (1995)Lower flank of the basin/Sub-oxic (OMZ impingement) Santa Monica Basin [SMB] SMB1 MC49 1-2 cm New Horizon 2001 −119.22 33.90 765 1055 200 Balestra et al. (2018)Slightly sloping basin floor/Anoxic (OMZ impingement) Bermuda Rise [BER] OCE326 BC9 2-5 cm Bermuda Rise 1998 −57.61 33.69 4517 374 27 Ohkouchi et al. (2002)Drift deposit/Oxic Iberian margin [IBE] SHAK06-5K 2-3 cm JC089 2013 −10.09 37.34 2646 1030 22 Magill et al. (2018)Drift deposit/Oxic 3 B. Ausín, E. Bruni, N. Haghipour et al. Earth and Planetary Science Letters 558 (2021) 116759 subsequently kept under basic atmosphere (NaOH pellets, 72 h) in a desiccator at 60 ◦C. Samples were then wrapped in tin capsules and measured by elemental analyzer (EA) – AMS. Fossil (in house shale) and modern (in house sediment) reference materials were prepared following the same procedure and used as processing blanks. Oxalic acid reference material (NIST SRM 4990C) was used as normalization standard. Capsule contribution and constant contamination introduced during fumigation and EA – AMS analyses was assessed and corrected for according to Welte et al. (2018). Correction parameters were a carbon mass of 8 ±4μg with a F14C of 0.6 ±0.2. Monospecific samples of Globigerina bulloides or Globigerinoides ruber (40-100 tests; 0.5-1 mg of carbonate) were picked from the 250-300 μm fraction in six out of the eight samples. Mixed planktonic foraminifera were picked from SBB and none for PER due to the scarcity of foraminifera. An automated method for acid digestion was applied to leach the outer shell and remove surface contaminants (Wacker et al., 2014). In short, samples were placed in septum-sealed vials and the outer shell was leached by adding 100 μl of 0.06 M HCl. Subsequently, the remaining sample was acidified after the addition of 100 μl of ultrapure H3PO4(85%). The CO2of both, the leachate and the main fraction, was successively collected on a zeolite trap of the gas interface system and transferred into a syringe, where it was diluted with He to a 5% CO2mixture and measured as two separate samples in the AMS. Fossil and modern reference materials, i.e. marble (IAEA-C1, nominal value 0.0020 ±0.0010) and an in-house coral standard (CSTD, nominal value 0.9447 ±0.0002, G. Dos Santos, personal communication), were prepared in the same way as the samples and used, respectively, for blank correction and validation purposes. All radiocarbon results are reported as fraction modern (F14C), which corresponds to the activity ratio of the sample relative to a modern reference material (Reimer et al., 2004). The radiocarbon content of the OC (14C) contained within grain size samples is also reported and calculated considering F14C and year of sample collection (Stuiver and Polach, 1977). 2.5. OC content and properties The OC content was measured simultaneously with OC-14C using the EA to an accuracy of better than 0.1% based on standards. For stable carbon isotopic analyses, aliquots of up to 200 mg of freeze-dried sediment were acidified by means of acid rinsing with 5 mL of 6M HCl to remove carbonates, subsequently neutralized with deionized water and dried in an oven at 50 ◦C. After sample homogenization, an aliquot (0.5-100 mg) was measured for stable carbon isotopic composition (δ13COC) on an FlashEA elemental analyzer coupled in continuous-flow with a Delta V isotope ratio mass spectrometer (all ThermoFischer Scientific, Bremen, Germany). Carbon isotope values are reported in per mille notation () relative to Vienna Pee Dee Belemnite (VPDB). The precision was better than 0.1(1σ) based on replicate measurements of standards. Similarly, 25-30 mg of non-decarbonated freeze-dried sediment was used to measure total nitrogen concentration (Ntot) and calculate the atomic C/N ratio. 3. Results 3.1. Bulk sediment characteristics The majority of sediment mass is found in the CS and FS fractions, in that order, together accounting for >80% of the bulk sediment in all samples (Table S1). Overall, sand is the least abundant fraction, followed by clay. Mean particle size is within the lower range of CS (∼10.5 μm in PER, ∼30 μm in NAM and ∼20 μm in all the other sites; Fig. S1A). Except for BER, which shows a bimodal size distribution centered at 7 and 30 μm, all samples show a unimodal particle diameter ranging from 10 to 30 μm (Fig. S1B). OC content is highest in sediments from the Peruvian (13.2%) and Namibian (6.3%) margins sites, moderate in sediments from Santa Barbara and Santa Monica Basins (∼2.25%) and low (<1%) in sediments at the other sites, reaching minimum values in Bermuda rise (0.4%) (Table S1). These results are in very good agreement with previous works at similar locations (Arthur et al., 1998; Inthorn et al., 2006; Mollenhauer and Eglinton, 2007; Mollenhauer et al., 2005; Ohkouchi et al., 2005; Rowe et al., 1988). The δ13COC values range from 19.8 to 23.2and, together with C/N ratios between 8.7 and 6.0, indicate OM is predominantly of marine origin (Meyers, 1994). The nearby basins of Santa Barbara and Santa Monica show equivalent TOC, δ13COC and C/N values, evidencing their geographical proximity and similar depositional settings. 3.2. Surface area (SA) The highest surface area is observed for clay and/or FS fractions, while CS shows the lowest values (Fig. 2A). Grain-size fractions and bulk sediment samples from the Peruvian and the Namibian margin exhibit an OC/SA ratio >1 mg OC/m2while samples from the NW African and Californian margins show values between 1 and 0.4 mg OC/m2, and those from the NW Atlantic margin and Bermuda rise are below the 0.4 mg OC/m2threshold. 3.3. OC properties: content and stable isotopic composition The Peruvian margin shows the highest OC content with large variability among grain-size fractions in absolute terms (5-18%) (Table S1, Fig. 3A). The opposite is true for Bermuda Rise, which has the lowest OC content and little variability among fractions (0.1-0.5%). Overall, FS hosts the highest proportion of OC in all the sites followed by the clay fraction in three of them. The CS fraction is usually the poorest fraction, while bulk sediment reflects intermediate OC content values within the range displayed by related grain-size fractions. Clay shows a greater variation in OC content in relation to the other fractions. Corresponding δ13COC values vary between −23.5 and −19.5 (Table S1, Fig. 3B and C). Fractions show relatively homogeneous values within each site, vary within a 1.5range for δ13COC, and no systematic pattern is evident among grain-size sediment samples. With the exception of the Peruvian and the Santa Monica Basin clay fractions, which show C/N ratios of 14.9 and 12.1, respectively, all samples exhibit a C/N ratio between 4 and 10 (Table S1, Fig. 3C). Although no clear patterns among grain-sizes are observed, C/N values for FS and bulk sediment are similar (if not identical) for most sites. 3.4. Radiocarbon content and age of OC and planktonic foraminifera Planktonic foraminifera 14C ages are systematically younger than co-deposited OC for all sites (Fig. 4). Samples from Santa Barbara and Santa Monica basins imply incorporation of bomb 14C (Table S2) (Mollenhauer and Eglinton, 2007), meaning the analyzed foraminifera originated sometime after the thermonuclear weapon testing (1960’s) and cannot be used for an accurate assessment of foraminifera-grain size 14C age relationships. Replicate 14C measurements on OC from bulk sediment samples for 5 out of 8 sites yielded virtually identical results (Table S2). Radiocarbon ages of bulk sediment samples generally represent average values within the range recorded by related fractions. Exceptions are the two replicates of the Bermuda Rise bulk sediment, which lie in the upper range, but within the error margin of CS, 4 B. Ausín, E. Bruni, N. Haghipour et al. Earth and Planetary Science Letters 558 (2021) 116759 Fig. 2. A) Mineral surface area (SA) and B) OC content vs SA of bulk sediment and grain-size fractions. OC loadings (OC/SA ratio) thresholds discriminate between highly productive environments characterized by short OET and large OC input (>1 mg OC m−2) and long OET and efficient OM remineralization relative to supply (<0.4 mg OC m−2), with intermediate values characterizing sediments from typical continental shelf settings (Bianchi et al., 2018; Blair and Aller, 2012). Fig. 3. Organic carbon content and properties in bulk sediment and corresponding grain-size fractions (upper panels) and grain-size fraction values normalized to the bulk sediment value (lower panel) for (A) organic carbon content (wt%), (B) stable carbon isotopes of OC, and (C) atomic C/N ratio. the most abundant fraction at this site, and the NW African margin, where bulk sediment is younger than corresponding grain-size classes. Unfortunately, no replicate is available for this sample. Except for BER, the abundance weighted average F14C and OC content of the analyzed grain-size fractions (i.e., all fractions but sand) is lower than values obtained for bulk sediments, a discrepancy that cannot be solely accounted for the relative contribution of the sand fraction (Table S2). Sample preparation involving MilliQ water was minimized to avoid OM loss. Yet, preferential solubilization of a small fraction of labile (typically younger) OC in MilliQ water is possible. Likewise, removal of carbonates by means of acid fumigation can also lead to decreased F14C in sediment samples as a function of inorganic carbon content, environmental matrix, and proportions of labile and refractory OC in the sample (Bao et al., 2019a,b). The potential impact of either processes on F14C values and OC content cannot be discarded. Nevertheless, our samples represent a wide variety in terms of the factors that determine the magnitude of the influence of these processes, and yet, a strong 14C age-grain size dependence is apparent in sediments from all sites, where clay is the youngest fraction and CS the oldest. 14C age relationships between the oldest (CS) and youngest (clay) fractions vary from 515 yrs for Peru to 2335 yrs for Bermuda Rise, with other sites exhibiting an intermediate spread in ages, most commonly within the 1040-1620 yrs range (Table S2). 4. Discussion 4.1. The effect of hydrodynamics on the characteristics of bulk OC The studied sites reflect a wide array of depositional settings, and yet a strong and systematic 14C age-mineral grain-size dependence is apparent in all of them (Figs. 4and 5). Specifically, OC in the clay fraction is the youngest, with 14C ages increasing in the FS and CS, in that order. We attribute this pattern to the ubiquitous influence of hydrodynamic sorting. McCave et al. (1995)determined that particles within the 10-63 μm size range (i.e., CS, or “sortable silt”) are non-cohesive and therefore exhibit greater propensity to resuspension and redistribution dispersal. Ac5 B. Ausín, E. Bruni, N. Haghipour et al. Earth and Planetary Science Letters 558 (2021) 116759 Fig. 4. Radiocarbon ages and age relationships in bulk sediments and grain-size classes. A) 14C ages of planktonic foraminifera, OC of bulk sediment and related replicates, and OC residing within each grain-size sediment fraction. Open diamonds indicate foraminifera that incorporate bomb 14C. B) Age discrepancy between grain-size fractions, bulk sediment, and foraminifera. cordingly, OM residing within CS is more likely to experience protracted translocation, being exposed to aging and selective degradation, and leading to older OC and lower OC content (Figs. 3A and 4). FS (2-10 μm) behaves in a more cohesive manner (McCave et al., 1995), but can be also winnowed by stronger currents (>10-15 cm s−1) (McCave and Hall, 2006) and deposited in distal locations after flow reduction or interference of cohesive forces. In contrast, clays are cohesive and show a lower propensity for mobilization and dispersal, which explains the observation of the youngest OC 14C ages for this fraction (Figs. 4and 5). Wheatcroft (1992) demonstrated that the effect of bioturbation can be particle-size dependent as finer particles penetrate deeper into the sediment. High sedimentation rates characteristic of all study sites diminish their susceptibility to bioturbation (Table 1). Moreover, sub-oxic to anoxic conditions (e.g., PER, SBB and SMB) and the presence of fine lamination in some sites indicate that bioturbation effect is negligible to absent (Schimmelmann et al., 2016) and yet, the same 14C agemineral-size dependence is evident at all sites. A comparable OC content-grain-size dependence only applies to FS and CS fractions, as clay shows the lowest OC content at PER, NAM and SMB, intermediate values in relation to FS and CS at SBB and IBE, and the highest OC content, comparable to that observed within FS, at NAF, NAT and BER (Table S1). Such dependence may be driven by both hydrodynamically-driven sorting processes and by organo-mineral associations. The former acts through preferential dispersal of CS in relation to FS and clay fractions (McCave and Hall, 2006; McCave et al., 1995) with subsequent remineralization of associated OM (Bao et al., 2016), whereas finer grain sizes have the larger sorptive capacity (greater SA) than coarser grain sizes (Keil et al., 1994a,b; Mayer, 1993; Mayer et al., 1988; Premuzic et al., 1982). The lower and variable degrees of OC association with the clay-size fraction contrast with prior studies where clays are typically observed to host greater proportions of OC relative to other size classes, an observation attributed to the larger surface area of clay-size minerals available for OM association (Bergamaschi et al., 1997; Keil et al., 1994a,b; Premuzic et al., 1982). However, there are various potential modes of OMmineral interaction (Mayer, 1993). For instance, surface roughness features like nanopores may increase the amount of OC embedded within mineral surfaces by one or two orders of magnitude (Mayer, Fig. 5. Radiocarbon age-mineral grain-size relationship with OC content. Maximum and minimum bubble size for OC content are adjusted to the OC content range covered by the grain-size fractions of each study site. 1993, 1994a). This is particularly so for coarser fractions, whereas clay minerals exhibit little potential for generating surface roughness (Ransom et al., 1998). Likewise, clay mineralogy (i.e., density, chemistry and flocculation behavior) has proven to play a key role in OM-clay mineral associations (Mayer et al., 2004; Ransom et al., 1998). Such factors may account for lower-than-expected OC contents of clay fractions from some of the samples. In general, δ13COC and C/N values of bulk sediments and grain size classes are consistent with marine inputs as the primary OM source (Meyers, 1994) (Table S1, Fig. 3A and B). Higher δ13COC values at NAM, NAF and PER agree with larger dust annual deposition from adjacent deserts in these regions (Jickells et al., 2005), whereas more negative δ13COC values are observed at BER, the 6 B. Ausín, E. Bruni, N. Haghipour et al. Earth and Planetary Science Letters 558 (2021) 116759 Fig. 6. Relationship between 14C and SA with OC content (A) and δ13COC (B). most offshore location. Despite significant variability in radiocarbon composition among grain-size fractions at each site, variations in C/N ratios and stable isotopes are relatively small, and do not allow for a more detailed discrimination of sedimentary OM origin. Moreover, such parameters may be influenced by selective degradation of specific compounds as a consequence of differential resuspension and dispersal of different grain-size classes. Selective removal of labile sedimentary OM is expected to lead to lower 14COC (older 14C ages) and δ13COC (via preferential degradation of 13C-enriched, labile organic components) values (Blair and Aller, 2012; Wang and Druffel, 1996). However, the SA-δ13COC relationship is less clear than that of SA with 14COC, indicating that the possible influence of hydrodynamic sorting on OC stable isotopic composition through selective OC degradation within specific fractions does not manifest itself in δ13COC-grain-size/SA relationships (Fig. 6). These results might also reflect the influence of sample treatment (i.e., acid rinsing) on δ13COC values. Removal of inorganic carbon by means of acid rinsing can lead to a larger loss of OC in relation to acid fumigation (Bao et al., 2019b), the latter employed here prior 14C and OC content analyses. It is possible that differences in the amount of OC that can potentially be lost due to the application of each of these acidification treatments led to the preservation of apparent patterns in relation to hydrodynamic controls only on 14C age and OC content. Overall, the isotopic contrasts between more and less labile OM appear insufficient to discern the possible influence of hydrodynamically driven-sorting processes. 4.2. The key role of fine silt in determining bulk OC signatures The OC-14C age of bulk sediment samples investigated in this study systematically exceeds that of co-deposited planktonic foraminifera (Fig. 4A), implying supply of OM that is older than that delivered from overlying surface water productivity. Given an absence of significant inputs of (pre-aged) terrestrial OM, these results provide evidence for a pervasive influence of hydrodynamic sorting on bulk sedimentary OC 14C (see section 4.1). We argue that differential contributions from grain-size classes carrying OM with varying OC contents and 14C signatures are responsible for these discrepancies. The relative impact of each grain-size fraction depends on the interplay between their fractional contribution to the sediment mass, the amount of OC they host, and the specific properties of the latter (e.g., 14C and δ13C). Clay contributes a negligible/minor amount to the bulk sediment mass, shows variable OC content in comparison to related CS and FS fractions within each site, and this fraction always contains the youngest OC (Table S1, Fig. 5). OC associated with FS is systematically older than clay and comprises 34% of bulk sediment on average, while CS is generally the most abundant fraction (50% on average) hosting the oldest OC. However, OC contents of CS are substantially lower, and the majority of OC resides within the FS (Fig. 5). Correspondingly, the latter fraction exhibits the greatest influence on bulk OC characteristics. Moreover, given the propensity of this fraction to undergo mobilization and translocation, FS has the potential to overwrite local/contemporaneous signals recorded in marine sediments. Indeed, 14C ages and C/N ratios of FS are comparable to those of bulk sediment for most sites (Fig. 3D, E, F, Table S1). Our results suggest, therefore, that age offsets previously observed between co-deposited bulk OC and planktonic foraminifera (e.g. Ausín et al., 2019; Kusch et al., 2010; Mollenhauer and Eglinton, 2007; Mollenhauer et al., 2005; Ohkouchi et al., 2002) were most likely due to the influence of pre-aged OC associated with FS that was advected to the study sites. Older-than-foraminiferal 14C ages for FS and CS suggest the occurrence of processes such as postdepositional alteration of sedimentary OC via bioturbation and/or delivery of allochthonous OC via riverine and aerosol transport, remobilization of in-situ sediments, and/or advection of distant material. The influence of bioturbation is expected to be minimal at all sites. Moreover, particle-size dependent bioturbation typically results in youngerthan-foraminiferal 14C ages for smaller particles (Bard, 2001; Wheatcroft, 1992). The C/N and δ13COC values suggest that contributions of terrestrial OC to FS and CS fractions (supplied via fluvial or aeolian transport) are negligible. The measurements performed in the present study do not allow us to discern between remobilization/re-suspension of proximal versus distal sediments. However, prior studies specially designed to identify these processes suggest that long-distance lateral sediment transport is a major translocation mechanism on diverse continental margin systems including off Namibia (Inthorn et al., 2006), Iberia (Magill et al., 2018), Siberia (Bröder et al., 2018), China and the northwestern US margins (Bao et al., 2018b). Deposition of laterally transported OM associated with fine-grained minerals implies the entrainment 7 B. Ausín, E. Bruni, N. Haghipour et al. Earth and Planetary Science Letters 558 (2021) 116759 of allochthonous and pre-aged OC that might perturb the original environmental signals originating from overlying waters (Ausín et al., 2019). Further consideration of these issues, both from the perspective of understanding carbon cycling and interpretation of continental margin sedimentary records is therefore warranted. 4.3. A framework for assessing site-specific variability Although all study regions show a strong dependence of SA, 14C age, and OC content on grain-size, the amount of preserved OC in comparable fractions and the magnitude of 14C discrepancies among grain-size fractions (Figs. 3A and 4B) are highly variable between sites. For instance, the grain-size classes from the Peruvian margin show similar and relatively young 14C ages, but the largest OC content variability among fractions of all sites (Figs. 3and 4). Conversely, fractions from Bermuda Rise show large 14C age variability but uniformly low values of OC content. This observation is ascribed to differences in the regional conditions that characterize each site, such as primary productivity, sedimentation rate, OET, hydrodynamic regime, and depositional setting, among others (Arnarson and Keil, 2007; Müller and Suess, 1979). Indeed, corresponding OC loadings (OC/SA ratios; Fig. 2B) indicate that fresh OM supply predominates on the high productive Peruvian margin despite strong coastal currents, whereas the opposite occurs in Bermuda Rise, characterized by oligotrophic overlying waters, long OETs and strong advective currents and sediment focusing (McCave, 2002). These two sites can be considered as end-members within the studied depositional setting range (Fig. 7), and provide distinct information about the role of OM-mineral associations and hydrodynamic processes on the fate of OC preserved in marine sediments. At sites like the Peruvian margin (and NAM to a lesser extent), the effect of hydrodynamic sorting – while noticeable in the 14C age-grain size relationship – is buffered in the bulk sediment by the large and continuous vertical flux of fresh OM. Here, 14C age-grain size relationship provides information on the potential of each fraction to associate with OC and host it at an early stage. In contrast, the majority of deposited OM at sites such as the Bermuda Rise (and IBE to a lesser extent) has previously been exposed to strong hydrodynamic forcing (via entrainment within turbidity currents or intermediate/bottom nepheloid layers, and/or resuspension cycles), while labile biological debris settling from the overlying water column has experienced extensive degradation during vertical transit and sedimentation. Long-term oxygen exposure may also have played a role though the destruction of organic-mineral aggregates (Arnarson and Keil, 2007). The latter conditions, which may lead to similar OC content in each of the fractions regardless of their initial potential to host OM, represent a “mature” system that may inform about the impact of prolonged and strong hydrodynamic processes on sedimentary OC. Sites exhibiting intermediate 14C age and OC content discrepancies between grain-size classes represent systems that inform about the potential of each fraction to protect OC from the effects of hydrodynamic processes. These locations also show intermediate mineral surface area-normalized OC loadings (0.4-1 mg OC m−1) (Fig. 2B) corresponding to relatively stable OM-mineral associations (Blair and Aller, 2012), implying protection from remineralization (Keil et al., 1994b). However, we argue that these sites represent a stable situation that is not necessarily permanent. Ashift towards stronger hydrodynamic forcing may remobilize these sediments, exposing associated OM to further decomposition and shifting residual materials towards a more mature end-member. Study sites are classified within one of these systems according to OC/SA values and sorted as a function of corresponding PP, OC contentand 14C-discrepancy values. Interestingly, water depth increases continuously from stable to mature systems, in agreement with previous work in the Chinese Marginal Seas that evidence increasing 14C age Fig. 7. Conceptual model summarizing the characteristics of the “Initial”, “Stable”, and “Mature” systems as explained in section 4.3. discrepancies between grain-size sediment fractions with increasing depth (Bao et al., 2019a). Our results confirm such relationship in a broader context for stable and mature systems, whereas for initial systems, high PP and vertical input of fresh OM likely mask the impact of any post-depositional sedimentation process on 14C age discrepancies. Moreover, it is likely that the relative contribution of refractory OC increases progressively from initial to mature systems, due to lower vertical fluxes of labile OC and selective preservation of terrigenous and relict OM during lateral transport (Mollenhauer and Eglinton, 2007). In this context, carbon fluxes (productivity), water depth, OET, and mineral grain size – the latter influencing both the capacity to host OM and propensity for (re)mobilization and lateral transport – each play a key role in dictating the amount, provenance, composition, and vintage of organic carbon signatures preserved in continental margin sediments. 5. Conclusions We find that hydrodynamically-driven processes exert a pervasive influence on the OM content and composition of continental margin sediments, and are manifested in the 14C age variability of OC residing in bulk sediments and corresponding grain-size fractions. Moreover, organo-mineral interactions exert a dual influence on sedimentary OC, with respect to both OM protection and its propensity for mobilization and redistribution. OC is preferentially associated with the fine silt fraction in all sites. Due to its higher OC content and contribution to bulk sediment mass, this fraction strongly influences bulk sediment OC 14C age and C/N of OC. Given its propensity to resuspension and advection under strong currents, translocated (allochthonous) OC associated with fine silt may suppress and/or distort primary signals originating from overlying surface waters preserved within continental margin sediments. Based on observed relationships between productivity, oxygen exposure, water depth, and mineral grain size, we suggest that different depositional environments can be categorized as “initial”, “stable” (quasi-steady state), and “mature” phases. These are distinguished based on 14C-age and OC content relationships among grain-size classes that reflect the interplay of hydrodynamic sorting and other regional conditions. Depositional systems reflecting the “initial” phase include those with high productivity, low oxygen exposure and local deposition (i.e., Peruvian and Namibian margins), and are characterized by small 14C agebut large OC content-discrepancies among grain-size fractions. “Stable” phases of deposition occur in systems that are characterized by high to moderate productivity and sediment focussing (e.g. NW Atlantic, and NW African margins), and feature moderate 14C ageand OC content-discrepancies among grain-size fractions. Additional factors, such as variability in oxygen exposure (e.g., SBB, SMB) contribute to compositional heterogeneity within stable depositional settings. “Mature” systems, which exhibit large 14C age discrepancies but smaller variations in OC content among grain-size classes (i.e., Bermuda Rise and SW Iberian margin to a lesser extent), are 8 B. Ausín, E. Bruni, N. Haghipour et al. Earth and Planetary Science Letters 558 (2021) 116759 characterized by weak vertical OC export and strong advective currents and lateral transport, reflecting the impact of prolonged exposure to hydrodynamic processes on sedimentary OC. This simple interpretational framework may help to assess the extent to which down-core variations in continental margin sedimentary records reflect hydrodynamic versus other environmental changes. CRediT authorship contribution statement B.A. and T.I.E. planned this investigation. N.H. and C.W. assisted with radiocarbon analyses. E.B. assisted with grain-size and surface area analyses. S.M.B. assisted with stable carbon isotope and nitrogen determinations. B.A. prepared the samples, analyzed the results, and wrote the manuscript with contributions by all coauthors. Declaration of competing interest There are no interests to declare. Acknowledgements We would like to thank Madalina Jaggi and Daniel Montluçon for their assistance during stable isotope and surface area analyses. We acknowledge the Regional Graduate Network for Oceanography (RGNO) Discovery Camps, supported by the Agouron Institute, the Simons Foundation, the Scientific Committee for Oceanographic Research (SCOR), the Ministry of Fisheries and Marine Resources (MFMR), the National Marine Information and Research Center (Nat MIRC), the University of Namibia (UNAM), ETH Zurich and the Swiss i-research & training institute, as well as scientists and crew of the R/V Mirabilis for realization of Namibian margin sampling. We greatly appreciate the anonymous reviewer for their careful reading of our manuscript and their many insightful comments. This study was supported by the project “TRAMPOLINE” (200021_175823) funded by the Swiss National Science Foundation, granted to T.I.E. and B.A. Appendix A. Supplementary material Supplementary material related to this article can be found online at https://doi .org /10 .1016 /j .epsl .2021.116759. References Arnarson, T.S., Keil, R.G., 2007. Changes in organic matter–mineral interactions for marine sediments with varying oxygen exposure times. Geochim. Cosmochim. Acta 71, 3545–3556. Arndt, S., Jørgensen, B.B., LaRowe, D.E., Middelburg, J.J., Pancost, R.D., Regnier, P., 2013. Quantifying the degradation of organic matter in marine sediments: a review and synthesis. Earth-Sci. Rev. 123, 53–86. Arthur, M.A., Dean, W.E., Laarkamp, K., 1998. Organic carbon accumulation and preservation in surface sediments on the Peru margin. Chem. Geol. 152, 273–286. Ausín, B., Magill, C., Haghipour, N., Fernández, Á., Wacker, L., Hodell, D., Baumann, K.-H., Eglinton, T.I., 2019. (In)coherent multiproxy signals in marine sediments: implications for high-resolution paleoclimate reconstruction. Earth Planet. Sci. Lett. 515, 38–46. Balestra, B., Quintana Krupinski, N.B., Erohina, T., Fessenden-Rahn, J., Rahn, T., Paytan, A., 2018. Bottom-water oxygenation and environmental change in Santa Monica Basin, Southern California during the last 23kyr. Palaeogeogr. Palaeoclimatol. Palaeoecol. 490, 17–37. Bao, R., Blattmann, T.M., McIntyre, C., Zhao, M., Eglinton, T.I., 2019a. Relationships between grain size and organic carbon 14C heterogeneity in continental margin sediments. Earth Planet. Sci. Lett. 505, 76–85. Bao, R., McIntyre, C., Zhao, M., Zhu, C., Kao, S.-J., Eglinton, T.I., 2016. Widespread dispersal and aging of organic carbon in shallow marginal seas. Geology. Bao, R., McNichol, A.P., Hemingway, J.D., Lardie Gaylord, M.C., Eglinton, T.I., 2019b. Influence of different acid treatments on the radiocarbon content spectrum of sedimentary organic matter determined by RPO/accelerator mass spectrometry. Radiocarbon 61, 395–413. Bao, R., Uchida, M., Zhao, M., Haghipour, N., Montlucon, D., McNichol, A., Wacker, L., Hayes, J.M., Eglinton, T.I., 2018a. Organic carbon aging during across-shelf transport. Geophys. Res. Lett. 45, 8425–8434. Bao, R., Uchida, M., Zhao, M., Haghipour, N., Montlucon, D., McNichol, A., Wacker, L., Hayes, J.M., Eglinton, T.I., 2018b. Organic carbon aging during across-shelf transport. Geophys. Res. Lett. Bao, R., van der Voort, T.S., Zhao, M., Guo, X., Montluçon, D.B., McIntyre, C., Eglinton, T.I., 2018c. Influence of hydrodynamic processes on the fate of sedimentary organic matter on continental margins. Glob. Biogeochem. Cycles 32, 1420–1432. Bard, E., 2001. Paleoceanographic implications of the difference in deep-sea sediment mixing between large and fine particles. Paleoceanography 16, 235–239. Bergamaschi, B.A., Tsamakis, E., Keil, R.G., Eglinton, T.I., Montluçon, D.B., Hedges, J.I., 1997. The effect of grain size and surface area on organic matter, lignin and carbohydrate concentration, and molecular compositions in Peru Margin sediments. Geochim. Cosmochim. Acta 61, 1247–1260. Bianchi, T.S., Cui, X., Blair, N.E., Burdige, D.J., Eglinton, T.I., Galy, V., 2018. Centers of organic carbon burial and oxidation at the land-ocean interface. Org. Geochem. 115, 138–155. Blair, N.E., Aller, R.C., 2012. The fate of terrestrial organic carbon in the marine environment. Annu. Rev. Mar. Sci. 4, 401–423. Bothner, M.H., Spiker, E.C., Johnson, P.P., Rendigs, R.R., Aruscavage, P.J., 1981. Geochemical evidence for modern sediment accumulation on the continental shelf off southern New England. J. Sediment. Res. 51, 281–292. Bröder, L., Tesi, T., Andersson, A., Semiletov, I., Gustafsson, Ö., 2018. Bounding crossshelf transport time and degradation in Siberian-Arctic land-ocean carbon transfer. Nat. Commun. 9, 806. Fischer, G., Karakas, G., Blaas, M., Ratmeyer, V., Nowald, N., Schlitzer, R., Helmke, P., Davenport, R., Donner, B., Neuer, S., Wefer, G., 2009. Mineral ballast and particle settling rates in the coastal upwelling system off NW Africa and the South Atlantic. Int. J. Earth Sci. 98, 281–298. Gibbs, R.J., Matthews, M.D., Link, D.A., 1971. The relationship between sphere size and settling velocity. J. Sediment. Res. 41, 7–18. Hedges, J.I., Keil, R.G., 1995. Sedimentary organic matter preservation: an assessment and speculative synthesis. Mar. Chem. 49, 81–115. Inthorn, M., Wagner, T., Scheeder, G., Zabel, M., 2006. Lateral transport controls distribution, quality, and burial of organic matter along continental slopes in high-productivity areas. Geology 34, 205–208. Jickells, T.D., An, Z.S., Andersen, K.K., Baker, A.R., Bergametti, G., Brooks, N., Cao, J.J., Boyd, P.W., Duce, R.A., Hunter, K.A., Kawahata, H., Kubilay, N., laRoche, J., Liss, P.S., Mahowald, N., Prospero, J.M., Ridgwell, A.J., Tegen, I., Torres, R., 2005. Global iron connections between desert dust, ocean biogeochemistry, and climate. Science 308, 67–71. Keil, R.G., Mayer, L.M., 2014. Mineral matrices and organic matter. In: Treatise on Geochemistry, 2nd ed. Elservier, pp. 337–359. Keil, R.G., Montlucon, D.B., Prahl, F.G., Hedges, J.I., 1994a. Sorptive preservation of labile organic matter in marine sediments. Nature 370, 549–552. Keil, R.G., Tsamakis, E., Fuh, C.B., Giddings, J.C., Hedges, J.I., 1994b. Mineralogical and textural controls on the organic composition of coastal marine sediments: hydrodynamic separation using SPLITT-fractionation. Geochim. Cosmochim. Acta 58, 879–893. Kusch, S., Eglinton, T.I., Mix, A.C., Mollenhauer, G., 2010. Timescales of lateral sediment transport in the Panama Basin as revealed by radiocarbon ages of alkenones, total organic carbon and foraminifera. Earth Planet. Sci. Lett. 290, 340–350. Magill, C.R., Ausín, B., Wenk, P., McIntyre, C., Skinner, L., Martínez-García, A., Hodell, D.A., Haug, G.H., Kenney, W., Eglinton, T.I., 2018. Transient hydrodynamic effects influence organic carbon signatures in marine sediments. Nat. Commun. 9, 4690. Mayer, L.M., 1993. Organic matter at the sediment-water interface. In: Engel, M.H., Macko, S.A. (Eds.), Organic Geochemistry. Principles and Applications. Plenum, New York, pp. 171–184. Mayer, L.M., 1994a. Relationships between mineral surfaces and organic carbon concentrations in soils and sediments. Chem. Geol. 114, 347–363. Mayer, L.M., 1994b. Surface area control of organic carbon accumulation in continental shelf sediments. Geochim. Cosmochim. Acta 58, 1271–1284. Mayer, L.M., Macko, S.A., Cammen, L., 1988. Provenance, concentrations and nature of sedimentary organic nitrogen in the gulf of Maine. Mar. Chem. 25, 291–304. Mayer, L.M., Schick, L.L., Hardy, K.R., Wagai, R., McCarthy, J., 2004. Organic matter in small mesopores in sediments and soils. Geochim. Cosmochim. Acta 68, 3863–3872. McCave, I.N., 2002. A poisoned chalice? Science 298, 1186–1187. McCave, I.N., Hall, I.R., 2006. Size sorting in marine muds: processes, pitfalls, and prospects for paleoflow-speed proxies. Geochem. Geophys. Geosyst. 7. McCave, I.N., Manighetti, B., Robinson, S.G., 1995. Sortable silt and fine sediment size/composition slicing: parameters for palaeocurrent speed and palaeoceanography. Paleoceanography 10, 593–610. Meyers, P.A., 1994. Preservation of elemental and isotopic source identification of sedimentary organic matter. Chem. Geol. 114, 289–302. Moffitt, S.E., Moffitt, R.A., Sauthoff, W., Davis, C.V., Hewett, K., Hill, T.M., 2015. Paleoceanographic insights on recent oxygen minimum zone expansion: lessons for modern oceanography. PLoS ONE 10, e0115246. 9