Richard G. Keil

Sedimentary organic matter preservation: an assessment and speculative synthesis

Throughout Earth history, almost all preserved organic matter has been incorporated in marine sediments deposited under oxygenated waters along continental margins. Given modern oceanic productivity and sediment burial rates of 50 × 1015 and 0.16 × 1015 gC yr−1, respectively, organic preservation in the marine environment is < 0.5% efficient. Although correlative information is often used to suggest that productivity, sediment accumulation rate, bottom water oxicity, and organic matter source are key variables, the mechanisms governing sedimentary organic matter preservation have remained unclear. The factors which directly determine preservation vary with depositional regime, but have in common a critical interaction between organic and inorganic materials over locally variable time scales. More than 90% of total sedimentary organic matter from a wide variety of marine depositional environments cannot be physically separated from its mineral matrix. This strongly associated organic component varies directly in concentration with sediment surface area and thus appears to be sorbed to mineral grains. Sediments accumulating outside deltas along continental shelves and upper slopes characteristically exhibit mineral surface area loadings approximately equivalent to a single molecular covering. These monolayer-equivalent coatings include a fraction of reversibly bound organic molecules that are intrinsically labile, but resist appreciable mineralization as they pass rapidly through oxygenated surface sediments and are preserved within underlying anoxic deposits. The delivery of mineral surface area is the primary control on organic matter preservation within these expansive coastal margin regions where roughly 45% of all organic carbon accumulates. Deltaic sediments account for roughly another 45% of global carbon burial, but often exhibit much less than monolayer-equivalent organic coatings. This pattern is seen in periodically oxygenated sediments off the mouth of the Amazon River, even though the component clastic minerals are discharged by the river with monolayer coatings. Comparably extensive losses of organic matter, including distinct particles such as pollen grains, occur in the surfaces of deep-sea turbidites in which long term reaction with O2 is clearly the causative factor. Sub-monolayer organic coatings also are observed in continental rise and abyssal plain sediments where slower accumulation rates and deeper O2 penetration depths result in increased oxygen exposure times and little (~ 5% of the global total) organic matter preservation. A transition zone between monolayer and sub-monolayer organic coatings apparently occurs on lower continental slopes, and is marked along the Washington coast by parallel offshore decreases in total organic matter and pollen between 2000–3000 m water depth. Sediments underlying highly productive, low-oxygen coastal waters such as off Peru and western Mexico are characteristically rich in organic matter, but account for only ~ 5% of total organic carbon burial. These sediments show a direct relationship between organic matter content and mineral surface area, but at organic loadings 2–5 times a monolayer equivalent. Organic materials sorbed in excess of a monolayer thus also may be partially protected. Such high sedimentary organic contents may result from equilibration with DOM-rich porewaters, or very brief O2 exposure times which allow preservation of extremely oxygen-sensitive organic materials such as pigments and unsaturated lipids. Thus organic matter preservation throughout much of the ocean may be controlled largely by competition between sorption at different protective thresholds and oxic degradation. Future research strategies should be specifically directed at delineating the mechanisms for organic matter preservation in marine sediments. In particular, special effort is needed to determine the amounts and types of sorbed organic materials and the nature of their bonding to mineral surfaces. The extent and dynamics with which organic molecules are partitioned between porewaters and solid phases also should be determined, as well as the effects of these phase associations on their reactivities toward chemical and biological agents. In addition, processes for slow oxic (and suboxic) degradation of organic materials bear investigation in deep-sea sediments, as well as in other extreme environments such as oxidizing turbidites, weathering shales, and soils. Such studies should include characterizations of hydrolysis-resistant organic materials and emphasize the complementary use of biochemical compositions with readily separable particles such as pollen to calibrate and typify the mechanisms and stages of sedimentary organic degradation.

Influence of oxygen exposure time on organic carbon preservation in continental margin sediments

Today, over 90% of all organic carbon burial in the ocean occurs in continental margin sediments. This burial is intrinsically linked to the cycling of biogeochemically important elements (such as N, P, S, Fe and Mn) and, on geological timescales, largely controls the oxygen content of the atmosphere. Currently there is a volatile debate over which processes govern sedimentary organic carbon preservation. In spite of numerous studies demonstrating empirical relationships between organic carbon burial and such factors as primary productivity, the flux of organic carbon through the water column, sedimentation rate,, organic carbon degradation rate, and bottom-water oxygen concentration,, the mechanisms directly controlling sedimentary organic carbon preservation remain unclear. Furthermore, as organic carbon burial is the process that, along with pyrite burial, balances O2 concentrations in the atmosphere, it is desirable that any mechanism proposed to control organic carbon preservation include a feedback buffering atmospheric oxygen concentrations over geological time. Here we compare analyses of sediments underlying two regions of the eastern North Pacific Ocean, one which has oxygen-depleted bottom waters and one with typical oxygen distributions. Organic carbon burial efficiency is strongly correlated with the length of time accumulating particles are exposed to molecular oxygen in sediment pore waters. Oxygen exposure time effectively incorporates other proposed environmental variables, and may exert a direct control on sedimentary organic carbon preservation and atmospheric oxygen concentrations.

Mineralogical and textural controls on the organic composition of coastal marine sediments: Hydrodynamic separation using SPLITT-fractionation

Abstract SPLITT-fractionation was used to sort hydrodynamically surficial sediments from the Washington margin, USA, into sand- (>250, 63–250 μm), silt- (35–63, 17–35, 8–17, 3–8 μm), and clay-sized (1–3, 0.5–1, 64 μm) from the shelf, where terrestrially-derived vascular plant debris accounted for >95% of the organic matter. Organic matter that could not be separated from the inorganic sediment accounted for >90% of the total organic carbon in most fractions, and loadings of organic carbon increased as the surface area of the inorganic particles increased. For the sand- and silt-sized fractions, the observed relationship of 0.81 ± 0.04 mg C m−2 (r = 0.97) was consistent with the hypothesis that a monolayer of organic matter is sorbed to the mineral surfaces. Clay-sized particles had lower organic loadings (0.37 ± 0.07 mg C m−2, r = 0.85), probably because the large interlamellar area of expandable clays was inaccessible to most organic molecules. After correcting for interlamellar area, clay-sized particles have the same organic carbon: surface area relationship as sands and silts (0.78 ± 0.08 mg C m−2, r = 0.96). The relationship over all the particle sizes was 0.76 ± 0.03 mg C m−2, (r = 0.96). While total organic matter concentrations were largely controlled by sediment surface area, the elemental composition of the organic matter appears to be partially affected by sediment mineralogy, and shifted from carbon-rich material (atomic C:N ~ 18.0) in larger, quartz-dominated fractions to N-rich material (C:N ~ 9) in the smaller, clay mineral-dominated fractions. Nitrogen enrichment relative to carbon (atomic N:C) was correlated with the amount of total clay (r = 0.80), smectite (r = 0.79), and the iron content (r = 0.74) of the sediments. Measurements of stable carbon isotopes indicate that clay-sized particles preferentially transport sorbed soil organic matter to deep sites while sand-sized fractions contain terrestrial plant debris (discrete and sediment-associated) that is transported along the shelf. The concentrations of terrestriallyderived organic matter in organic matter from shelf and slope sediments was estimated to be 60–85% and 10–15%, respectively. The quantity, bulk chemical composition, and distribution of marine and terrestrially derived organic matter to Washington margin sediments are influenced by 1. (1) the surface area of the sediment minerals, 2. (2) the mineralogical composition of the sedimentary matrix, and 3. (3) the natural hydrodynamic sorting of sedimentary materials along the continental margin. The major fraction of organic material in these sediments is sorbed to mineral grains. Interactions between organic material and mineral surfaces strongly influence the distribution and elemental composition of the organic material present in marine sediments.

Loss of organic matter from riverine particles in deltas

Abstract In order to examine the transport and burial of terrigenous organic matter along the coastal zones of large river systems, we assessed organic matter dynamics in coupled river/delta systems using mineral surface area as a conservative tracer for discharged riverine particulate organic matter (POM). Most POM in the rivers studied (n = 6) is tightly associated with suspended mineral material; e.g., it is sorbed to mineral surfaces. Average organic loadings in the Amazon River (0.67 ± 0.14 mg C m−2), the river for which we have the largest dataset, are approximately twice that of sedimentary minerals from the Amazon Delta (∼0.35 mg C m−2). Stable carbon isotope analysis indicate that approximately two-thirds of the total carbon on the deltaic particles is terrestrial. The combined surface-normalized, isotope-distinguished estimate is that >70% of the Amazon fluvial POM is not buried in the delta, consistent with other independent evidence (Aller et al., 1996). Losses of terrestrial POM have also been quantified for the river/delta systems of Columbia in the USA, Fly in New Guinea, and Huange-He in China. If the losses of riverine POM observed in these river/delta systems are representative of rivers worldwide, then the surface-constrained analyses point toward a global loss of fluvial POM in delta regions of ∼0.1 x 1015 g C y−1.

The effect of grain size and surface area on organic matter, lignin and carbohydrate concentration, and molecular compositions in Peru Margin sediments

A C-rich sediment sample from the Peru Margin was sorted into nine hydrodynamically- determined grain size fractions to explore the effect of grain size distribution and sediment surface area on organic matter content and composition. The neutral monomeric carbohydrate composition, lignin oxidation product yields, total organic carbon, and total nitrogen contents were determined independently for each size fraction, in addition to sediment surface area and abundance of biogenic opal. The percent organic carbon and percent total nitrogen were strongly related to surface area in these sediments. In turn, the distribution of surface area closely followed mass distribution among the textural size classes, suggesting hydrodynamic controls on grain size also control organic carbon content. Never- theless, organic compositional distinctions were observed between textural size classes. Total neutral carbohydrate yields in the Peru Margin sediments were found to closely parallel trends in total organic carbon, increasing in abundance among grain size fractions in proportion to sediment surface area. Coincident with the increases in absolute abundance, rhamnose and mannose increased as a fraction of the total carbohydrate yield in concert with surface area, indicating these monomers were preferentially represented in carbohydrates associated with surfaces. Lignin oxidation product yields varied with surface area when normalized to organic carbon, suggesting that the terrestrially-derived component may be diluted by sorption of marine derived material. Lignin-based parameters suggest a separate source for terrestrially derived material associated with sand-size material as opposed to that associated with silts and clays. Copyright 0 1997 Elsevier Science Ltd

Impact of suboxia on sinking particulate organic carbon: Enhanced carbon flux and preferential degradation of amino acids via denitrification

Fluxes of particulate organic carbon (POC) through the oxygen deficient waters in the eastern tropical North Pacific were found to be relatively less attenuated with depth than elsewhere in the eastern North Pacific. The attenuation coefficient (b) for the flux was found to be 0.40 versus the composite value of 0.86 determined by Martin et al. (1987). To examine this further, sinking POC was collected using sediment traps and allowed to degrade in oxic and suboxic experiments. Using a kinetic model, it was found that degradation proceeded at similar rates (roughly 0.8 day−1) under oxic and suboxic conditions, but a greater fraction of bulk POC was resistant to degradation in the suboxic experiments (61% vs. 23%). Amino acids accounted for 37% of POC collected at 75m, but following degradation the value dropped to 17% and 16% in the oxic and suboxic experiments respectively. POC collected from 500m was 10% amino acids. The non-AA component of POC collected at 75m was not degraded under suboxic conditions, while under oxic conditions it was. These results suggest that microbes degrading OC under suboxic conditions via denitrification preferentially utilize nitrogen-rich amino acids. This preferential degradation of amino acids suggests that 9% more nitrogen may be lost via water column denitrification than is accounted for when a more “Redfieldian” stoichiometry for POC is assumed.

Biochemical distributions (amino acids, neutral sugars, and lignin phenols) among size-classes of modern marine sediments from the Washington coast

Abstract In order to examine relationships of organic matter source, composition, and diagenesis with particle size and mineralogy in modern marine depositional regimes, sediments from the continental shelf and slope along the Northwest Pacific rim (Washington coast, USA) were sorted into hydrodynamic size fractions (sand: >250, 63–250 μm; silt: 35–63, 17–35, 8–17, 3–8 μm; and clay-sized: 1–3, 0.5–1,

Organic geochemical perspectives on estuarine processes: sorption reactions and consequences

Abstract The conventional paradigm for conceptualizing the behavior of bioactive materials in estuarine systems has been to think of dissolved and particulate organic matter as separate entities that do not readily interconvert or associate with minerals. However, many organic molecules are extensively, and to some extent reversibly, associated with mineral surfaces in rivers and estuaries. The realization that some organic molecules may actively partition between dissolved form and the surfaces of minerals has great geochemical implications for many natural environments. This brief review touches on some possible ramifications of organic–mineral interactions for estuaries, using the well-studied Amazon River/Estuary system as an example.

Mechanisms of pore water organic matter adsorption to montmorillonite

The extent and mechanisms of adsorption of marine pore water organic matter to montmorillonite were studied in a series of batch and sequential adsorption experiments. Pore water natural organic matter (pNOM) and easily extracted natural organic matter (eNOM) were collected from Liberty Bay (Puget Sound, WA, USA) sediments. The pNOM and eNOM were each divided into two size fractions using a 1000 D ultrafilter. Batch adsorption isotherms were approximately linear, and the >1000 D fractions of both pNOM and eNOM had larger partition coefficients (Kd) than the 1000 D pNOM and eNOM, and ∼1.6 l/kg for 1000 D fractions during batch isotherm experiments. Adsorption of NOM was found to decrease with increased temperature, suggesting that hydrophobic effects were not the dominant adsorption mechanisms in this system. Ion exchange was also not an important adsorption mechanism because adsorption increased with ionic strength. The observed enhancement in adsorption with ionic strength indicated that van der Waals interactions were important in the adsorption of NOM. Ligand exchange was found to be a significant mechanism since the presence of SO42− in solution reduced the amount of NOM adsorbed. Ca2+ enhanced adsorption slightly more than Na+, suggesting that cation bridging was involved. The relative contributions of van der Waals interactions, ligand exchange and cation bridging were estimated to be approximately 60%, 35% and 5%, respectively, for adsorption of NOM in a CaCl2 solution.

Abiotic transformation of labile protein to refractory protein in sea water

In order to determine if organic matter dissolved in sea water may undergo abiotic alterations that make it resistant to microbial degradation, the protein ribulose 1,5-bisphosphate carboxylase (RuBPcase) was abiotically aged in sterile sea water and then exposed to natural bacterial assemblages. Rates of protein assimilation decreased when protein was aged as little as 6 h; protein aged for 40 days was degraded 4-fold more slowly than non-aged protein. Abiotic modification rates, calculated from decreases in degradation rate with increasing aging time, were highest during the initial day of aging (0.8–4.8 d−1) and then decreased to 0.03 d−1. No aging effect was observed in organic-free sea water, indicating that organic-organic interactions produced the refractory protein. Amino acids from the protein were fully recovered after acid hydrolysis, indicating that the decreased lability was not caused by acid-stable molecular changes to the protein. Abiotic complexation of labile organic compounds with existing DOM may be a critical first step in the formation of refractory organic materials.