Progress Report II.
Ocean Sources and Sinks

Oceanic Sources and Sinks

(Fred Mackenzie, Karen von Damm, Dave DeMaster, Tom Church, Billy Moore)

I. Introduction

In his classic 1974 book, Chemical Oceanography, Wallace Broecker noted that our knowledge of the major cation cycles is surprisingly primitive; he pointed out that most of the hypotheses proposed rested on inadequate evidence. What has been learned regarding the oceanic sources and sinks of major cations and other chemical elements in the intervening 25 years?

Oceanic, riverine and atmospheric data sets have improved and are vastly enlarged. Many include comprehensive analyses of major, nutrient and trace elements as well as considerations of spatial and temporal variations. Translating concentrations to fluxes and understanding processes controlling concentration and flux have been the goals of many studies. Removal processes that were hypotheses in early 1970 have been demonstrated convincingly. Additional source functions have been discovered. Nevertheless, disagreement remains concerning the relative importance of different processes to the balance of elements in the ocean.

The objective of this paper is to consider advances that have been made in understanding processes responsible for regulating the overall chemistry of the ocean and where possible, to evaluate the extent to which firm conclusions regarding sources, sinks, fluxes and elemental balances can be demonstrated.

II. Terrestrial Sources: The Water Route

A. Global Fluxes

Rivers and streams are presently the dominant purveyors of materials to the ocean and probably have been in the past as well. Since the classic works of Clarke and Livingstone, estimates of the average concentrations of the major dissolved constituents of river water have not changed dramatically. What has changed is our understanding of the sources of dissolved materials and the processes controlling their concentrations in rivers. In recent years, estimates of mean streamflow and the total dissolved load of the world's rivers reaching the oceans have converged on values of about 40,000 km³/yr and 4 billion tons/yr for a total dissolved solids concentration of average river water of 110 mg/L.

There is still considerable uncertainty in estimates of both the total suspended load carried by rivers to the ocean and its composition, particularly in terms of the minor and trace elements. Recent estimates of total suspended flux have varied from about 13.5 to 20 billion tons/yr for pre 1950, when increased dam construction began to reduce these fluxes.

The global mean annual direct groundwater discharge to the ocean is less certain than the river discharge but is probably less than 10% of the surface water flow, with a recent estimate of 2,400 km³/yr. Although the global dissolved salt groundwater flux has been estimated as 1,300 million tons/yr, this flux is even more poorly known, particularly for minor and trace elements, dissolved organic carbon, and the nutrient elements N and P. Evidence for the importance of groundwater input and sea water cycling in coastal aquifers has recently been obtained from measurements of chemical tracers in coastal waters.

Although much has been learned about these global fluxes, considerable room for improvement remains. In the 1973 document Orientations in Geochemistry, many of the problems associated with making these estimates were pointed out. The scientific community and funding agencies were encouraged to narrow the gap in our understanding through proper sample collection and analysis of chemical data using documented analytical procedures for the dissolved and suspended loads of rivers and the chemistry of groundwaters and improved estimates of river and groundwater discharge. To some extent, this recommendation has been followed. However, there is still uncertainty in these fluxes. This uncertainty stems from a number of factors, including the fact that most river gauging stations are located well upstream of river mouths, the flow of ungauged rivers must be determined from assumptions that involve the balance between precipitation and evapotranspiration in the watershed, a lack of chemical data and discharge estimates for rivers during flood (particularly small rivers), the amount of sediment temporarily or permanently stored in the lower reaches of rivers, the indirect methods that must be used to estimate groundwater discharge and the lack of analytically reliable chemical data for groundwaters near coasts, and the changing characteristics of both river and groundwater discharge and flux because of human activities.

B. Regulation of Major and Nutrient Elements

The dissolved load of the world's rivers is derived from the following sources: about 7% from beds of salt disseminated in rocks, 10% from gypsum and sulfate salts disseminated in rocks, 38% from carbonates, and 45% from the weathering of silicate minerals. The primary anion in most river water is HCO₃^-. Two thirds is derived from atmospheric CO₂ or decomposition of fossil organic matter, which is converted to HCO₃^- through weathering of silicate (30%) or carbonate (70%) minerals. One third of river HCO₃^- comes directly from carbonate mineral structures undergoing weathering.

Total organic carbon in rivers in the dissolved and particulate phase has its source in the land biota and soil organic carbon pools. Total nitrogen found in rivers originates from the biological nitrogen cycle on land, and about 85% is organic in origin. Total phosphorus is derived from both organic sources on land and through the chemical weathering of phosphate minerals. Chemical weathering of rock minerals and humus provides variable quantities of trace elements to the dissolved and particulate loads of rivers and for the majority of trace elements, transport in the suspended load of rivers both in organic and inorganic phases is most important.

C. Filters and Sinks near the Land-Sea Interface

Chemical reactions may sequester a substantial component of riverine fluxes near the continent. Ion exchange may increase the dissolved concentrations of minor elements. Water flux times the riverine composition cannot be simply translated into net oceanic source terms without intimate knowledge of biogeochemical and exchange reactions either at the transient saline boundary of a river plume, or within the more permanent mixing zone of a confined estuary and attendant sinks.

Most rivers discharge into bays or estuaries, but some rivers may deliver their loads directly to the open shelf or even the upper slope. Groundwater mostly discharges in the proximal portion of the continental margins. The removal of riverine suspended material allows sun light to penetrate into these nutrient-rich waters. The resulting plankton blooms further alter fluxes of riverine materials to the open ocean by removing nutrients and particle-reactive elements. In some cases removal is only temporary as these elements are released by remineralization and transported to the ocean.

D. Anthropogenic Effects on Composition and Flux

In some river and groundwater systems, anthropogenic activities have substantially modified the chemistry and transport of these systems. These modifications even for the global scale occur on a very short time scale. With a residence time of global rivers of only 16 days with respect to net precipitation over land, there is little doubt that such human activities as the application of fertilizers to croplands, land use changes (including deforestation), sewage discharges into surface water systems, utilization and chemical contamination of water resources, and impoundment of water and sediment because of damming are modifying the characteristics of river flows at rates that make it difficult to obtain average global fluxes.

Most of the world's large rivers are located in two areas of net positive precipitation: 10°N-10°S and 30-60°N and S. Because river runoff and groundwater flow depend to a first approximation on net precipitation on land, the fact that global mean precipitation has not changed in the past 100 hundred years or so implies that without damming and other human activities, these flows would have been relatively stable during this time period. However, during the climatic warming of the past century, global precipitation on decadal to multi-decadal timescales has undergone changes in which extremes in precipitation have become more probable. This implies the possibility that at least for rivers, discharge has also become more variable on these time scales. Because of the utilization of about 54% of the accessible water runoff from the land and because of impoundment behind dams and diversion of waters, there have been regional, if not global, decreases in stream flow to the ocean via rivers in the last half century. Although the conclusion is controversial, it has been estimated that only about 20% of the world's drainage basins have pristine water quality.

Suspended inorganic matter is comprised of the major elements O, Al, Fe, Si, Ca, Mg, Na, and P, and a host of minor and trace elements whose concentrations in suspended matter are not well known on a global scale. A relatively large percentage of the suspended load of rivers enters the oceans from East Asia, Southeast Asia, and the high islands of the Pacific and Indian oceans. There is controversial evidence that the suspended sediment flux to the oceans has doubled during the past two centuries primarily because of land use change and other human activities. Certainly impoundment of sediment behind dams has led to regional, if not global, decreases in sediment transport to the ocean since 1950.

The riverine and groundwater transport of dissolved major and minor inorganic species have also been affected by human activities. Increased denudation due to land use changes and increased chemical transport due to such diverse factors as fertilizer application, acid rain and waste discharge have fundamentally altered many river systems. Global estimates of these changes are difficult to constrain. Of the major elements, Na, Cl, and S have been significantly affected by human activities.

The concentrations of N and P may have doubled in global river discharge during the past two centuries, and certainly have increased in regional river systems and groundwater aquifers. Suspended and dissolved organic matter concentrations in river waters are also changing globally because of human activities. Total organic carbon transport to the ocean by rivers has increased during the past two centuries by as much as a factor of two. Dissolved and particulate trace elements in rivers and groundwaters including Hg, Zn, Cd, B, Cu, Pb, Co, Ni and many others have likewise been affected by human activities.

During the last century, coastal aquifers have experienced considerable change due to increased coastal populations. Dredging of channels has breached confining layers and increased contact between sea water and underlying aquifers. Increased groundwater usage has lowered potentiometric surfaces in coastal aquifers causing infiltration of sea water into these formations and in some cases, reversal of flow direction. The result has been increased salinization of coastal groundwaters. As sea water invades these aquifers, chemical reactions change the composition of the fluids. To emphasize the importance of mixing and chemical reaction in these coastal aquifers, Moore has called them subterranean estuaries. Fluid exchange between subterranean estuaries and coastal waters may be a significant source of nutrients and trace metals to the ocean, in some cases rivaling fluxes from local rivers. How much of the signal seen in the coastal ocean derives from anthropogenic effects and how much from natural exchange in these systems is unknown.

III. Terrestrial Sources: The Atmosphere

A. Introduction

Thirty years ago, our knowledge or appreciation of atmospheric sources to the oceans was limited to recognition of relic wind borne quartz in pelagic sediments. Today, the role of the atmosphere as a source or sink to the oceans is vastly improved. This has resulted from many new data, investigations, and discoveries.

The new data are in the form of several large atmospheric programs which for the first time systematically collected atmospheric data in global marine areas (e.g. SEAREX, AEROCE, MAGE, etc.). The data consist of gases, aerosols, and precipitation that were collected on extended cruises or often synoptically at several latitudinal stations in the Pacific and Atlantic. At the same time, new investigations off major river systems and in estuaries documented that the net fluvial input to the open ocean can be quite small for many dissolved and particulate substances as a result of the "estuarine filter". This raised the importance for atmospheric transport of substances such as iron which are almost quantitatively trapped near shore. In fact, one of the most important new discoveries was that the flux of aeolian iron may potentially limit open ocean primary productivity.

B. Sources

The sources of atmospheric delivery to the oceans have recently been reviewed and shown for trace species to rival riverine input, even without estuarine filtration. The extent and magnitude of atmospheric sources depend on geography and meteorology. Geographically, the North Atlantic, western North Pacific and Indian Oceans, and their nearby inland seas, are subjected to large atmospheric input because of their proximity to both natural dusts (deserts) and industrial sources.

Crustal dust transported to the ocean through the atmosphere is a primary source for elements in sedimentary phases. Dissolved sources are minor due both to limited dust solubility in sea water and the overwhelming inputs from riverine sources. However, for crustal trace elements (Al, Fe, Mn), even the small (10% maximum) dissolution of aeolian dust is the primary source to the open-ocean surface waters. Dust dissolution is enhanced through "atmospheric chemical weathering", where highly saline and acidic conditions during cloud processing of dust increase the soluble component. The crustal trace element input is linked to the highly seasonal inputs of dust and wind transport that accompany aridity cycles and trade or monsoon wind patterns. This can lead to the majority of dust input in only a few large dry depositional events. While these patterns are known to have increased during glacial periods by as much as order of magnitude, there are even shorter-term variations of several fold associated with cycles of aridity on periods of decades to centuries.

Modern industrial emissions dominate atmospheric input of many trace elements to the oceans. The most notorious example is lead from gasoline additives, which on a global scale has overwhelmed natural sources by orders of magnitude. This has produced a transient surface enrichment in ocean waters which is now declining with decreased use of tetraethyl lead. Enrichment is also seen for other elements such as Bi with potential coal burning sources. As many of these trace elements from industrial emissions exist as condensation products on sub-micron aerosols, they are readily solubilized in precipitation.

In contrast to the cycles for P and Si, which essentially balance dissolved riverine input with burial in the seabed, the marine N cycle depends on transformations involving dissolved gas, in this case molecular nitrogen. Although not readily recognized, the main source of fixed N to the oceans is through N fixation reactions, which are especially important in tropical and subtropical environments. This source of N to the oceans is larger than the fixed N riverine supply (dissolved plus particulate) and many times larger than the atmospheric N deposition flux. The rate of industrial N fixation to form fertilizer is only slightly less than the natural global N fixation rate, which supports the perception that human activity can significantly affect N cycling in the terrestrial and marine environments.

Because of the atmospheric transport path and short residence time in water, the flux of N to the ocean, particularly from human activities in the coastal zone, is changing continuously, even on a global scale. The stimulation of productivity in estuaries from excess N can change nutrient fluxes to the shelf and affect ecosystem structure in the coastal zone (e.g. harmful algal blooms). Indeed, with the growth of developing economies, fossil fuel emissions and fertilizer usage will continue to grow, increasing the potential for deterioration of the coastal zone.

The main process removing fixed N from the ocean is denitrification, which releases molecular nitrogen. Approximately half of the denitrification occurs in oxygen-depleted waters and half in the organic rich marine sediments. Burial of organic N in marine sediments represents only one eighth of the fixed N removal from the ocean.

Atmospheric delivery of major nutrients in fixed form (N, P and Si) plus some trace nutrients (e.g. Fe) is hypothesized to affect, and in some cases limit ocean production. For example in the North Atlantic, and at times in the North Pacific, there have been substantial increases in atmospheric fixed nitrogen deposition over the past century of industrialization, on the order of a factor 4-5 or a quarter of the current emissions in North America and Europe. Seasonal or modal upwelling of deep waters is thought to control annual blooms, but the episodic nature of atmospheric delivery during oligotrophic periods may play an important role. For example, in the North Atlantic the amount of allochthinous nitrogen found in the main thermocline largely exceeds that from upwelling and requires a significant source from nitrogen fixation. Because iron is required for N fixation, the supply of dust may play a critical role in this cycle. The dust supply has probably increased in the recent past from agricultural practices in the Sahel region of Africa exacerbated by cycles of aridity.

The atmospheric delivery of carbon in either elemental or organic forms is just now being investigated. In the North Atlantic, elemental carbon comprises about a third of the tropospheric aerosol number, presumably from modern practices of biomass burning and fossil fuel combustion, which now may be about equivalent. The atmospheric deposition of organic carbon reveals the unmistakable compound signature of land plants. The deposition of organic nitrogen is about the same as inorganic nitrogen worldwide, but the source is uncertain. Biomass burning may again be implicated.

C. Air-sea exchange

There have been two discoveries which have recently brought the importance of air-sea exchange on marine chemistry to the forefront. These have been named in honor of the discoverers: the CLAW hypothesis of Charlson, Lovelock, Andrea, and Warren and the Martin hypothesis of John Martin. It is conceivable that the two hypotheses might be linked. The Martin hypothesis states that iron (from aeolian dust) may be a limiting nutrient in large expanses of the surface ocean still replete in major nutrients. This appears particularly true in the upwelling regions of the equatorial ocean, or even in areas starved of dust due to meteorological conditions (e.g. northern Southern Ocean). Subsequent iron addition experiments at sea currently support this hypothesis. The CLAW hypothesis states that ocean productivity in part controls the amount of marine cloud cover through the emission of di-methyl sulfide and its subsequent oxidation which produces sulfate condensation nuclei. Since it has been suggested that iron dust is also reduced, solubilized and scavenged by precipitation, the two hypotheses may be linked. Together, they suggest coupling between the air-sea exchange and surface ocean biogeochemistry in the regulation of greenhouse gases. Besides CO₂ sinks to the ocean from augmented productivity, N₂O from coastal denitrification could have important oceanic sources. Thus the atmosphere appears important in the efficient delivery of elements and contaminants to the ocean, in chemically labile forms, and directly to sensitive euphotic zones. In this manner, particularly in the Northern Hemisphere, the modern effects of human industrialization on surface ocean chemistry are best being documented.

IV. Hydrothermal Processes

A. Introduction

In assessing our progress during the last 30 years, it should be noted that 30 years ago hydrothermal systems had not been discovered, and even the underpinning theory of plate tectonics was not fully accepted in North America. The classic 1966 paper of Bostrom and Peterson on metalliferous sediments on the East Pacific Rise, and the suggestion that they were the result of hydrothermal exhalations from the Earth's mantle, had recently been published. Thus the hypothesis that seafloor hydrothermal systems of some type must exist had been seriously presented, although it would be another decade until the first example was discovered in 1977 at the Galapagos Spreading Center.

The Galapagos site was characterized by warm springs (<25 °C above bottom seawater temperatures) and lush biological communities. These results startled biologists and geochemists. The fluids were hypothesized to be of much hotter origin. Two years later ~350°C fluids were observed and sampled at 21° north latitude on the East Pacific Rise. Presently, ~30 active sites of submarine hydrothermal venting have been discovered world wide. Although many occur on the mid-ocean ridge system (divergent plate boundary), an increasing number of sites are being discovered and characterized in places such as back-arc spreading centers, seamounts, and at other convergent plate boundary zones.

B. Fluxes

After the two initial discoveries of submarine hydrothermal fields, estimates of the global chemical fluxes (both positive and negative) were made quantifying the net flux from hydrothermal emanations. The fluxes were calculated using two independent approaches, heat flow and ³He. As the conductive heat flow anomaly along the global mid-ocean ridge system was known, the elemental anomaly/heat (enthalpy) of the hydrothermal fluids was multiplied by the annual heat loss due to convection to reach an annual flux for a large suite of elements. Second, as the oceanic input to the global budget of ³He was defined, assuming the earth-ocean-atmosphere system was at steady-state, the elemental anomaly/3He content was multiplied by the annual input of ³He from the oceans and a net elemental flux was calculated. As the ³He/heat ratio in both the Galapagos and 21°N fluids was the same, as were their elemental concentrations (once the Galapagos fluids were corrected for loss due to subseafloor mixing with seawater and consequent precipitation of metal-sulfates and -sulfides), the fluxes calculated by both methods were the same. Serious problems with elemental balances were noted almost as soon as the initial fluxes were reported. The best example was potassium, where more was being added to seawater by the hydrothermal flux than was present in all of the oceanic crust formed within a year.

In the ~15 years since then we have learned that the elemental fluxes based on heat flow and ³He are not valid for the following additional reasons: (1) ³He/heat ratios vary between vents and temporally within individual vents, (2) elemental concentrations vary by orders of magnitude between vents and vary temporally within individual vents, (3) water- rock interaction is a major control on hydrothermal vent fluid compositions, (4) phase separation ("boiling" when the phase separation is sub-critical and "condensation" when super-critical) appears to affect virtually all vent fluids and profoundly alters their composition and the conditions for water-rock interaction, and (5) direct magmatic degassing occurs in some cases, but how widespread and important this process is remains virtually unknown.

Chloride is essentially the only significant anion in vent fluids because sulfate is removed and alkalinity titrated. As most of the cations are transported as chloro- complexes, Cl is a master variable. With phase separation, the chlorinity of the hydrothermal fluid undergoes major changes from the starting seawater chlorinity. One cannot simply look at an elemental concentration in a vent fluid and conclude that the fluid is a net source or sink to the ocean based on concentration alone. A better approach is to examine the elemental-ratio-to-Cl in comparison to that ratio in seawater, to determine if a net gain or loss has occurred due to water-rock interaction and phase separation processes, etc. This approach assumes that Cl is conservative on some time scale, i.e., there is no major hydrothermal source or sink for Cl. (On decadal time scales, however, Cl mass balance problems are evident in individual vent systems.)

C. Reactions in Hydrothermal Plumes

In addition to the reactions occurring below the seafloor which are the ones "classically" considered when contemplating hydrothermal fluxes, the role of the hydrothermal plumes is a critical one for the net hydrothermal impact on ocean chemistry. The reduced Fe and Mn contained in the hydrothermal vent fluids are injected into the oceanic water column at depths typically 250-300 m above the seafloor for "chronic" plumes (i.e., those formed by "normal", steady-state hydrothermal venting), but can rise to heights >1000 m above the seafloor in the case of 'event' plumes (i.e., associated with volcanic intrusions or eruptions along the mid-ocean ridge system). The Fe and Mn in the plume slowly oxidize through a combination of chemically and microbially mediated reactions. The resulting Fe- and Mn-oxyhydroxides then act as efficient scavengers for oxyanions such as PO₄, As- and V-species, as well as the rare earth elements and some trace metals. The plumes therefore serve as net sinks in the mid-depth oceanic water column.

D. Net Exchange

The net hydrothermal flux to the ocean ultimately depends on the net heat loss associated with the mid-ocean ridges (and other sites of hydrothermal venting). The heat flux is fairly well constrained by a number of arguments, the real difficulty lies in translating that heat flux to a water flux.

At least three major uncertainties in how the heat and water flux are partitioned severely limit our ability to calculate accurately the (chemical) hydrothermal flux to the ocean.
(1) What is the proportion of the flux that occurs as high temperature axial vent fluid flow versus the proportion that occurs at lower temperature on the ridge flanks? It is only within the last few years that data have been collected that begin to address what the composition of these flank fluids are in both temperature and chemical signal. Even right on the youngest oceanic crust on the ridge axis, the fluid flow is split between high (>200°C) and low (<35°C) temperature venting.
(2) What proportion of the axial fluid flux occurs as high temperature fluids versus low temperature fluids. We now know that although almost all of the low temperature flow appears to contain some high temperature fluid, it is not a simple dilution of high temperature fluids with seawater, but instead various chemical species are gained, and others lost, some due to inorganic processes, but others likely due to biological processes. Some work suggests that the low temperature flow is quantitatively more important for total heat transport, but the chemical anomalies are obviously much less.
(3) Since the early 1990's, we have had the opportunity to study the perturbation and influence of magmatic events (whether eruptions and/or intrusions) on seafloor hydrothermal systems - and they are profound. Characterizing the proportion of heat and water flux associated with this very youngest part of the hydrothermal cycle (compared to the 'steady- state' venting) is very difficult. The amount of water and heat transported in the first year following an 'event' appears to be very large based on observations of three such events. As these very early fluxes have profoundly different compositional characteristics (element/heat; element/³He; element/Cl; total gases), the volume of these will have a significant effect on the sign of the hydrothermal flux (source or sink) for many elements.

The complexities inherent in hydrothermal processes require new approaches to estimating elemental balances. Some approaches center on the oceanic cycles and budgets of Mg, Mn, Sr isotopes, Li isotopes, and Ge/Si ratios. Each approach thus far has thus far yielded information about processes affecting each of these elements, but little quantitative data about hydrothermal fluxes.

Clearly, high temperature hydrothermal activity is a net sink for Mg, PO₄, U, and Mo and a net source for Fe, Mn, Cu, Zn, Si, H⁺ and the alkalis Li, K, Rb, Cs. For the alkalis, this source function may well be balanced by uptake during alteration of the oceanic crust at temperatures <150°C. Fluids exiting from hydrothermal sites in back-arc spreading centers, as well as those from ultramafics on very slow spreading mid-ocean ridges, appear to have some fundamental differences in composition, related to their different source rocks. For seafloor hydrothermal activity in general, the N cycle remains poorly defined, as does the inorganic C cycle, while the organic C cycle is essentially undefined. The direction for S is not clear - although SO₄ is quantitatively lost, H₂S concentrations can be greater than the sulfate concentration in seawater. Na likely has a small sink in high temperature hydrothermal systems, due to albitization which also results in a Ca source (but the direction of the Ca flux could be reversed on the flanks). Alkalinity has a net sink in high temperature hydrothermal venting due to the amount of protons produced during water-rock reactions.

V. Sediment Sinks

The four basic types of deposits accumulating in the ocean include detrital sediments, biogenic sediments, metalliferous sediments, and authigenic minerals. By mass, detrital sediments have the largest accumulation rate (~200 x10¹⁴ gm/yr), followed by biogenic deposits (~40 x10¹⁴ gm/yr), composed of calcium carbonate (~30 x10¹⁴ gm CaCO₃¹⁴ gm SiO₂/yr) and organic matter (~4 x10¹⁴ gm/yr). The rates of accumulation for metalliferous sediments and authigenic minerals have not been accurately quantified, but are certainly small relative to the rates of detrital and biogenic sediment accumulation.

The detrital sediment coming down the major rivers of the world is predominantly suspended material (~80-90%) in the silt- and clay-size range, with the remainder carried as bedload, primarily sand-size material. The silt- and clay-size particles from the river flocculate when they encounter high ionic strength waters during river/ocean mixing, which causes them to settle out primarily in deltaic and shelf sediments. During interglacial times, approximately 75% of the detrital sediment transported to the ocean by rivers is trapped on the continental shelf; much of this is eroded out to the continental slope and rise during glacial low sealevel stands.

Of the biogenic sediments, calcium carbonate is the most rapidly accumulating. During interglacial times approximately half of the calcareous material is buried in continental shelf environments and half is buried in deep-sea and continental slope deposits. When sealevel is lower during glacial times, carbonate accumulation on the shelf is significantly reduced, with much of the calcareous shelf material deposited during interglacial periods transported offshore to the continental slope and rise. The Atlantic Ocean accumulates much more calcium carbonate than the Pacific because the Atlantic deep waters have a higher pH (i.e., less corrosive to CaCO₃) than those in the Pacific as a result of their close proximity to a site of deep-water formation. The surfaces of calcium carbonate shells commonly are coated with Fe minerals, which are effective scavengers of ions such as phosphate.

Biogenic silica primarily accumulates in high-latitude, deep-sea sediments of the Southern Ocean (>50% of the total flux). The production of biogenic silica is not higher in these Antarctic waters than in other oceanic areas; but, the preservation efficiencies are greater than at lower latitudes, because the lower temperatures retard silica dissolution. Many of the high biogenic silica accumulation rates occur beneath the Antarctic Polar Front, which commonly coincides with several oceanic ridge systems. Sediment focusing may lead to erroneously high rates of accumulation in some cases, which can be assessed using ²³⁰Th normalized chronologies. If the Southern Ocean silica budget is corrected for sediment focusing using this radiochemical approach, the deposits may account for only 17-36% of the biogenic silica accumulation occurring in the ocean. A biogenic silica repository that has not been adequately quantified is the upper continental slope. Based on the rate of marine organic carbon accumulation in these slope sediments and a typical slope biogenic silica/organic carbon ratio, as much as 25% of the biogenic silica burial in the marine environment may be occurring in continental slope environments.

In examining organic carbon burial in the marine environment, it is important to distinguish between terrestrial and marine sources. Organic matter is buried the fastest in deltaic environments (~2 x10¹⁴ gm/yr), but most of this organic matter is terrestrial in origin. Continental shelves and upper slope deposits account for another 2 x10¹⁴ gm/yr of organic matter burial, with increasingly larger contributions from marine sources away from the terrestrial dispersal systems. The lower continental slope, the rise, and abyssal sediments only accumulate organic matter at a rate of ~0.6 x10¹⁴ gm/yr, because most of the settling material in these areas is oxidized prior to burial in the sediment column. During the past five years, factors affecting the preservation of organic matter have been examined including bottom water oxygen content and clay/organic matter interactions. Accumulation rate also is an important factor controlling the seabed preservation efficiency of organic carbon as well as the characteristic C/P ratio of a deposit. Global models have been established depicting organic matter burial and oxidation in open ocean sediments for water depths >1000 m. Although organic matter is the primary phase carrying N, P and organic C to the seabed, the fates of the 3 biologically important elements can be quite different. For example, much of the organic carbon reaching the seabed is oxidized back to inorganic carbon that diffuses out of the seabed, whereas regenerated P commonly is reprecipitated as inorganic phosphate minerals. The changes in the rain rate of organic carbon to the sea floor and the extent of calcite dissolution in sediments driven by aerobic oxidation of organic matter may explain in part glacial/interglacial _PCO₂ changes.

Metalliferous sediments primarily originate as a result of hydrothermal emanations. In addition to the trace elements mobilized by the hot circulating brine solutions, metalliferous sediments become enriched in various trace elements because iron and manganese coat the particles near the vent regions, which creates a reactive surface for scavenging of particle-reactive species. Chemical budgets and accumulation rates for these metalliferous deposits have been difficult to establish because of spatial and temporal variations in depositional processes (see section IV). In non-vent areas there are two different types of Mn growths: one that is very rapid (coating historical artifacts such as soda cans and pirate pistols at mm/yr) and one associated with Mn nodule growth that is very slow (mm/my).

It has been hypothesized that the formation of authigenic clay minerals in marine sediments leads to the removal of soluble cations and the release of acidity (CO₂). This was termed "reverse weathering" by Garrels and Mackenzie because it is the opposite of the weathering reaction. Since this hypothesis was proposed, discovery of large-scale hydrothermal cycling (see section IV) of elements at mid-ocean ridges has enlarged the scope of reverse weathering to include these high temperature reactions. The relative roles of high and low temperature reverse weathering in the mass balances of major cations is uncertain.

Conclusive tests of the role of low temperature reverse weathering have been difficult because the abundance of authigenic minerals is small making them extremely difficult to detect in deltaic and other sedimentary environments dominated by clays from continental sources. Evidence for low temperature reverse weathering was first demonstrated in the tropical sediments of Kaneohe Bay, Hawaii. More recently, the formation of significant quantities of authigenic K-Fe-Mg clay minerals has been reported on the Amazon shelf. Although these low temperature reverse weathering reactions have a small effect on the oceanic Si budget, they may be a significant sink for K and possibly other cations. Authigenic mineral formation in sediments also has received significant attention because its formation on biogenic silica surfaces retards dissolution. These surface phases appear to control to a large extent the solubility and dissolution rate of siliceous material.

VI. Net Mass Balance

In an ocean at steady state with respect to composition, the dissolved constituents of river and groundwater entering the ocean would be removed through inorganic and biological reactions. However, for certain dissolved components of river water, like Na and Cl, it may be that they are currently being stored in the ocean. For others, like inorganic carbon, their removal rates may exceed their inputs and the ocean is being depleted in them. Changes in input and removal that occur during glacial-interglacial cycles affect the balances of inorganic carbon, Ca, and nutrients.

Numerous authors have addressed the chemical mass balance between rivers and ocean waters. Some have presented mass balances that closed input and output estimates for the elements Mg, Si, K, Ca, N, P and C within a few to 10%. Balancing riverine inputs of trace metals with outputs to continental margin sediments and open ocean have revealed that as much as 95% of the input of trace metals to the marine environment may accumulate in ocean margins. This accumulation has a significant effect on the oceanic residence times of the trace elements. Because the removal of trace elements in coastal marine environments is strongly coupled to biological activity, the increased riverine nutrient fluxes of the past and those projected for the future imply an increased efficiency of removal of trace elements in ocean margin sediments.

All these balances must be viewed with caution because of the lack of precise input and output data and the assumption of a steady state ocean, which is a major factor in many of these mass balances. Certainly on times scales of human generations to centuries, and glacial/interglacial periods, this assumption may not be valid. On the other hand, oceanic residence times for most elements are considerably longer than 100 years; therefore, the recently altered fluxes have not had time to change oceanic concentrations significantly.

Since 1970 we have discovered new oceanic sources and sinks of elements and learned a great deal about the processes that regulate these and other sources and sinks. We are confident that some combination of known processes controls the chemistry of sea water. This was not true in 1970. However, we cannot yet answer fundamental questions regarding the relative importance of known processes to the mass balance of elements in the ocean. As an undistinguished gourmet once remarked, "I know what's in the soup, but I don't know why it tastes so good".

General References

Berner, K. and R.A. Berner, 1996, Global Environment Water, Air and Geochemical Cycles. Prentice Hall, 376 pp.

Degens, E.T., S. Kempe, and J.E. Richey, 1991, Biogeochemistry of Major World Rivers, SCOPE. Chichester, U. K. John Wiley, 347 pp.

Duce, R., et al, 1991, The atmospheric input of trace species to the world ocean. Global Biogeochemical Cycles, v. 5, p. 193-259.

Gregor, C.B., R.M. Garrels, F.T. Mackenzie and J.B. Maynard, 1988, Chemical Cycles in the Evolution of the Earth. J. Wiley 276 pp.

Humphris, S.E., R.A. Zierenberg, L.S. Mulllineaux, R.E. Thomson, editors Seafloor Hydrothermal Systems: Physical, Chemical, Biological and Geological Interactions. Monograph 91, 1995, Washington, D.C., 466p.

Milliman, J., 1993, Production and accumulation of calcium carbonate in the Ocean: Budget of a nonsteady state. J. Global Biogeochemical Cycles, v. 7, p927-957.

Meybeck, M., 1979, Concentrations des eaux fluviales en elements majeurs et apports en solution aux oceans. Revue de Geologie Dynamique et de Geographie Physique, v. 21, p. 215-246.

Return to FOCUS Progress Report Page

webmaster@joss.ucar.edu
Last modified: 23 January 1998