H43C-0377 1340h
Geomorphic controls on chemical weathering rates in the High Himalayas of Nepal
Evaluating the competing roles of climate and topography in controlling chemical weathering rates is critical to understanding the linkages between tectonic activity, surface processes, and the atmosphere. Rivers issuing from ten watersheds in the Annapurna region of Nepal, with drainage areas ranging from 6 to 2700 sq-km, were monitored during the 2002 monsoon season to estimate discharge and to collect weekly water samples. Chemical denudation rates of bedrock were estimated from the cation and anion load. We find a strong inverse linear relationship (r-sqrd = 0.83) between rates of chemical denudation and average watershed hillslope angle. We propose that this relationship is due to the effects of soil depth (ie. steep slopes = thin soils, gentle slopes = thick soils) on chemical weathering rates in an interface-limited weathering environment. These results also support Gilbert's hypothesis on the relationship between soil depth and the rate of soil production.
H43C-0378 1340h
Surface Area Production: The Impact of Glacial Erosion on Chemical Weathering Fluxes
Glacial erosion produces high sediment yields characterized by a large fraction of fine-grained material. As a consequence, the mineral surface area production due to glacial erosion is extraordinarily high, resulting in a high potential for chemical weathering. The clay and silt-sized fractions comprise as much as 80% of glacial till and, for typical glacial sediment yields, represent on order 10$^{3}$ km$^{2}$ mineral surface area production per km$^{2}$ of glacier per year. Despite this production of fresh mineral surface area, silicate chemical weathering fluxes from glaciers are lower than found in other environments. Here, we demonstrate that the measured silicate weathering flux from a small Alaskan glacier can be predicted from the measured sediment yield, grain size distribution, and mineralogy. We use published mineral weathering rate constants, adjusted for 0\degC conditions found at glacier beds. The success of this calculation shows that silicate-weathering fluxes from glaciers are directly related to mineral surface area production, and that they are low due to depression of the mineral weathering rate constants by low temperatures. A corollary of this calculation is that deposition of fine-grained glacial erosion products in environments conducive to chemical weathering should enhance chemical weathering fluxes. We explore this idea by estimating weathering fluxes from loess deposited in forested regions downstream from active glaciers. Loess is a small fraction of the total glacial sediment yield, but it should be highly reactive owing to its surface area. If, as we expect, weathering fluxes are high from loess-mantled regions, this shows that weathering rates are linked with the products of erosion, rather than the process of physical erosion.
H43C-0379 1340h
Balancing Physical Erosion and Chemical Weathering In and Out of the Steady State
Establishing the nature of a steady state for the earth's surface and criteria for a landscape to be considered to be in such a state presents conceptual and practical challenges. Reconciliation of model predictions of weathering rates with geochemical and geomorphologic data brings this problem to the fore. We present an analytical weathering model that emphasizes the roles of tectonic uplift and weathering profile development in setting the balance between physical and chemical alteration. Ultimately, these are the processes that determine the supply rate of fresh weatherable material to the surface. To facilitate comparisons across tectonic regimes, the model is parameterized in terms of an `effective surface age'. This parameter describes a time scale on which weathering processes can occur in a given environment, and further allows examination of the bounds of steady-state conditions under varying circumstances. Our modeling suggests that physical erosion and chemical weathering, while broadly correlated, do not obey a uniform relationship over the whole spectrum of tectonic environments. Power-law behavior holds over a certain range, but towards extremes of both high uplift and ancient, stable cratons, other descriptions are more appropriate. Non-steady-state events, including slope failures, glaciations, isostatic adjustments and tectonic reorganization, will perturb or overprint the steady-state signal on a variety of temporal and spatial scales.
H43C-0380 1340h
The Chemical Weathering End Member of the Coupled Physical and Chemical Weathering System
It is widely recognized that physical and chemical weathering processes are coupled. In natural systems erosion is constantly removing chemically weathered material making it difficult to decouple physical and chemical contributions. In order to understand these complicated systems it will be necessary to study end-member systems where processes that control chemical and physical weathering can be considered separately. We use 3 cm weathering rinds developed on basalt clasts within 125 ka fluvial terraces along the Pacific coast of Costa Rica as a proxy for saprolite development in the absence of physical weathering. This highly controlled system enables us to examine the weathering interface in detail. The weathering interface is comprised of thin reaction fronts where element concentrations vary from parent rock to rind concentrations. Electron microprobe data indicate a 2 mm thick reaction front of mobile elements (i.e., Ca and Na) at the interface and ~ 4 mm thick reaction fronts of less mobile elements (i.e., Si). These reaction fronts are 3 orders of magnitude narrower than those recognized at the bedrock-saprolite interface in landscape-scale studies. Iso-volumetric weathering of labradorite and augite has produced a rind of gibbsite and iron oxide. This transition is accompanied by a pronounced change in porosity from $<$1% in the core to 50% in the rind. Petrographic and SEM images reveal labradorite dissolution coreward of the rind/core interface producing secondary porosity and increased permeability into the core. Reactive transport modeling allows simulation of rind development using appropriate values of diffusion, mineral surface area, and reaction rate. A controlling feature appears to be porosity of the basalt clast which controls both diffusion of reactants and products into and away from the weathering front. Initial reaction-diffusion simulations were based on the assumption of constant porosity so as to understand important controls on the thickness of the reaction front and the rate at which it advances. Results indicate that the initial rate of rind advancement is a mix of interface- and transport-control. However, once a dissolution front is established, the long-term rate of rind advancement becomes transport-limited and shows the parabolic time dependence characteristic of diffusion processes. Field data, however, indicate that rind advance rates are constant through time, suggesting that more complicated transport within the rind simple diffusion. One possibility is that the increase in rind porosity allows for flow through the rind, thus causing the fixed concentration boundary condition to migrate coreward along with the weathering front. Such a mechanism could result in linear rather than parabolic rind advance rate with time. Diffusion rates and mineral surface area were varied to match initial rind advance rates to 2.4x10-4 mm/year calculated from field data. Effective diffusion coefficients for dissolved aqueous species ranged from 0.51x10-5 cm2/sec to 8.57x10-5 cm2/sec for Fe3+ and H+, respectively. Initial rind advance rate (2.1x10-4 mm/yr) and reaction front thickness (2 mm) were closest to observed values when formation factor, labradorite surface area, and augite surface area were 6.7E-5, 1500 m2/m3, and 500 m2/m3, respectively. Further modeling is ongoing to investigate how porosity changes affect transport within the weathering rind and rind advance rate. These modeling efforts will help clarify the coupling of physical and chemical processes at the reaction front.
H43C-0381 1340h
Regolith Evolution Influences Element Redistribution During Weathering of Volcanic Rocks in Erosional, Sedentary, and Depositional Landscapes: Examples From Hawai'i, Guatemala and Southeastern Australia
This study examines the weathering of volcanic rocks exemplifying each of three landscape/regolith associations (erosional, sedentary/relict/residual, and depositional), and illustrates how the regolith/landscape associations and their geomorphic evolution influence the geochemical evolution of the regolith. In erosional landscapes, the rate of physical erosion exceeds the rate of chemical weathering of rock to altered regolith, and surface materials consist of fresh or minimally weathered bedrock. Recent basalts (<4ka) from Hawai'i have weathered slightly and have accumulated no weathering rinds, saprolite, or allochthonous regolith over their brief exposure history. Whole-rock geochemistry is not affected by the small amount of chemical weathering. Leaching has been insufficient for differential removal of elements, and there are no elemental sources outside of the nearly fresh outcrops from which elements might have been introduced into the exposed volumes. In sedentary/relict/residual landscapes, the rate of chemical weathering equals or exceeds the rate of physical erosion, and surface material consists of deeply weathered saprolite. Some volcanic rocks of Plio-Pleistocene age from Hawai'i and Guatemala have experienced spheroidal or corestone weathering, in which corestones of minimally weathered rock are surrounded by concentric saprolitic shells and saprolite derived from the decomposition of the volcanic rock. Many major elements and some minor elements (REE) are depleted from the saprolitic portions of these regoliths. However, several of these minor elements (REE) are enriched in the inner portions of corestone-shell complexes, suggesting that these minor elements and REE leached from saprolite are transferred within the regolith to secondary minerals formed during incipient weathering of the corestones. In depositional landscapes, the surficial material consists of sediment (colluvial, fluvial/alluvial, or aeolian). Tertiary volcanic rocks of the Monaro Volcanic Province (New South Wales, Australia) were emplaced in fluvial-lacustrine environments and almost immediately covered by fine-grained clastic sediment. The jointed flows weathered spheroidally. Corestones have essentially fresh major element and REE signatures. However, Zr (probably redistributed physically from the fine-grained sediment) exhibits systematic absolute enrichment with progressive weathering in the Monaro corestone-shell complexes. Weathering of volcanic rocks results in geochemical trends that differ systematically with the presence, nature, and extent of development of associated regolith. Geochemical patterns of element depletion and enrichment in individual samples and suites of samples can only be properly interpreted if the regolith/landscape context of the samples is taken into account.
H43C-0382 1340h
A Field Experimental Approach to Understanding Relationships Between Chemical Weathering and Erosion in the Arctic Environment of Swedish Lapland.
A field experimental approach to understanding solutional weathering and erosion processes in the Arctic environment of Kärkevagge Swedish Lapland has been undertaken over the course of the past decade. The study site is located at approximately 68N 18E, with a mean annual temperature of -1.7C and a mean annual precipitation of approximately 1000mm. The experiment involved the installation of polished disks of limestone, dolomite, and granite in soil pits under a diversity of micro environmental settings. Installation sites were selected so as to encompass the major vegetation communities in the valley which display strong altitudinal zonation and strong contrasts in moisture conditions. Disk depth placements varied between 4 and 60cm and were generally located at horizon boundaries. Disks were selectively removed after one year (1995) from each of 5 vegetation communities, and all disks were retrieved from 10 vegetation communities after 5 (1999) and 10 years (2004). At each time period disks were weighed and mass loss determined. Similarly, nylon mesh bags filled with crushed polished granules of dolomite and granite were placed on the soil surface at the same sites as the disks, and weighed at the same time periods. At the end of the 5 year time period processes responsible for mass loss were investigated. This study permits the initial evaluation of an experimental approach to understanding relationships between physical, chemical, and biological processes operating in the "critical weathering zone". Additionally, it permits an initial assessment of early stage weathering and solutional erosion of various lithologies in a cold environment.
H43C-0383 1340h
Investigating an Inverted Soil Column in Northern Tanzania: Could Intense Groundwater Weathering be the Culprit?
Weathering of silicate rocks permanently sequesters a significant amount CO$_{2}$ on our planet (Berner et al., 1983; Dessert et al., 2003; Gaillardet et al., 1999). Therefore, the investigation of soils and their conjugate protoliths have implications for a wide range of disciplines from soil to atmospheric sciences. This study investigates soil formed in Northern Tanzania on the southern slope of the dormant volcano Mt. Kilimanjaro. Our sample site is in the Machame region at an elevation of ~1640 m where the phonotephrite to basaltic bedrock has been dated at 0.4 to 0.5 million years (Evernden and Curtis, 1965). We determined bulk elemental concentrations of soil and bedrock samples from this region using an ICP-MS and XRF. From initial investigations into the bulk soil and bedrock chemistry using a novel mass balance method, we were able to investigate the relative mobility of a suite of elements. Relative abundances of Ta, Nb, Hf, and Zr are constant and therefore these elements are immobile. In contrast, Ti, an element commonly thought to be immobile, is clearly not immobile in our samples. The entire soil column appears to be highly depleted in Si and Ca but enriched in Al. These features indicate extensive weathering and indeed some samples approach bauxite compositions. Surprisingly, however, weathering versus depth is reversed. Si and Ca have been removed by 60 and 70 % respectively from the upper 2 meters, but below 2 m, they have been removed by 95 and 99 %. This means that the soil is more weathered at depth than in the shallowest 2 meters. We believe that ground water weathering is responsible for this inverted soil profile and may increase CO$_{2}$ consumption estimates by 10-30 % for similarly affected basalts.
H43C-0384 1340h
Geochemical Modeling of Soils: A Soil-Component Approach
Quantitative modeling of pedological processes begins by placing measured soil properties within a framework that allows the identification and testing of geochemical and biological mechanisms. For example, mass-balance interpretations of pedogenesis describe gains and losses of chemical elements in a manner that can implicate specific mineral reactions and biological processes. Soil-component modeling is an extension of the mass-balance interpretation of pedogenesis that describes observed gains and losses in chemical elements, ion-exchange capacity, and soil volume with hypothesized sets of soil components such as minerals and soil organic matter. These models are calculated using mole-balance equations for mineral components, associated equations for ion-exchange capacities and volumes of minerals and soil organic matter, and equations that account for the uncertainties of all analytical data. The GRex computer program was written to evaluate these relations and identify combinations of minerals and organic matter that can simultaneously account for the observed chemical and physical properties of a single soil or net-change in properties between two related soils. The power of this approach is that many alternative scenarios of mineral and organic matter assemblages can be quickly defined that may provide insight into the geochemical mechanisms of pedogenesis, or suggest additional field and laboratory work. Examples of this modeling method, as applied to data on soils from sites along a marine terrace chronosequence in Oregon and a basalt climosequence on the island of Hawaii, will be shown.
H43C-0385 1340h
Chemical weathering rates in deep-sea sediments: Comparison of multicomponent reactive transport models and estimates based on $^{234}$U
Chemical weathering rates in natural systems are typically much slower than expected based on experiments and theory. There are several possible explanations. However, because it has been difficult to determine what effects in particular reduce the rates in specific settings, natural rates remain difficult to predict. Silicate-rich deep-sea sediments provide an ideal in-situ laboratory for investigating weathering rates because certain potentially important factors, such as advective transport through heterogeneous media, limitations on the availability of reactive surface area due to low porosity and/or cementation, unsaturated flow conditions, and seasonal variations in fluid flux and temperature, do not occur in this setting. Geochemical profiles from Site 984 in the North Atlantic are modeled using a multi-component reactive transport model (CRUNCH) to determine in-situ rates of plagioclase dissolution and other diagenetic processes, including sulfate reduction and anaerobic methane oxidation. Various possible processes which might contribute to slower rates in the field are considered, including the effect of mineral saturation state, secondary precipitation of clays, inhibition by dissolved aluminum, and the availability of reactive surface area. The reactive transport model includes an isotopic solid-solution formulation that tracks the isotopic composition of precipitating (calcite) and dissolving (plagioclase and calcite) phases, thus allowing the determination of plagioclase dissolution rates. The rate constants for plagioclase determined by geochemical transport modeling of major element profiles are within the same range determined from U-series calculations and suggest that natural weathering rates for this system are on the order of 10$^{-17.5}$ to 10$^{-17.7}$ mol/m$^{2}$/sec assuming estimates of reactive surface area are correct, several orders of magnitude slower than laboratory-derived rates. The slow plagioclase rates are most likely due to the fact that dissolution takes place close to equilibrium, but the close to equilibrium conditions require either slow clay precipitation or precipitation of soluble clays. Unavailability of reactive surface area could also explain the slow rates, but this is considered less likely because of the very high porosity (about 80%) and the low cementation.
H43C-0386 1340h
Sulfur Isotopes in the Rivers of the Mackenzie River Basin: Implication for CO$_{2}$ Consumption
Mass budgets of chemical weathering in hydro-systems usually assume that the dissolution of CO$_{2}$ in rain and soil waters provides most of the protons that attack rock minerals. However, the oxidative weathering of reduced species containing sulfur, such as pyrite, can be a significant source of protons. On a global scale, the origin of sulfate in rivers is still unclear. At least three possibilities can be envisaged: sulfate from sedimentary gypsum, atmospheric pollution and oxidative weathering of sulfide. As shown by previous studies, S and O isotopes of the sulfate molecule can allow deciphering between the different sources. In the aim of constraining the origin of sulfate delivered to the ocean by rivers and to refine CO$_{2}$ consumption budgets for chemical weathering reactions, we have started to measure S and O isotopes in the largest river systems. Among them, the Mackenzie River basin is an ideal case, because it has been recognized by geologists to contain both gypsum and reduced sediments, mainly black-shales, rich in pyrite. The O and S isotopes of the sulfate molecule do show large discrepancies between the two main geomorphic units of the Mackenzie River basin: the Rocky-Mackenzie Mountains to the West and the interior platform to the East. For example, river samples from the lowlands are characterized by values of $\delta^{34}$S ranging from $\ -3.25\permil$ to $\ -18.47\permil$ and those from the mountains varying between 2.06\permil and 9.87\permil. We interpret these values and the relationships between isotopic composition of sulfate and major elements as showing the dominant contribution of sulfide oxidation in the lowlands and gypsum dissolution in the mountains. The details of our mixing model, e.g. end-member choices, will be discussed in detail; but based on our data we calculate that 54 to 96% and 18 to 40% of dissolved sulfate come from sulfide oxidation in lowland rivers and mountain rivers, respectively. The mean value obtained for the Mackenzie River Basin is 32%. Assuming that protons added in water by pyrite oxidation react preferentially with carbonate rocks, we calculate that 30% of the protons reacting with carbonate minerals do not originate from the dissolution of atmospheric CO$_{2}$ in soil water. For a long-term perspective and at a global scale, the oxidative weathering of sulfide coupled to rock weathering can lead to a release in the atmosphere of CO$_{2}$ originating from the carbonate reservoir. This process may be significant in mountainous regions, such as Himalayas, where surface rocks have been recently exhumed and where high physical weathering rates sustain a continuous contact between fresh rocks and water.
H43C-0387 1340h
Sr Fluxes from the Himalayas: Calculation of Carbonate and Silicate Inputs
The marked increase in seawater $^{87}$Sr/$^{86}$Sr ratio since 40 Ma suggests that enhanced physical erosion associated with the Himalayan-Tibetan orogeny impacted on global chemical weathering fluxes. Whether this reflects weathering of high $^{87}$Sr/$^{86}$Sr ratio silicates, high $^{87}$Sr/$^{86}$Sr ratio carbonates or increased silicate chemical weathering is controversial. To resolve this we need to determine the proportions of Sr and $^{87}$Sr which are derived from carbonate and silicate minerals. Here we present a new method for doing this which uses arrays of tributary compositions to define the cation ratios of silicate and carbonate inputs. This reveals the importance of incongruent precipitation and dissolution reactions in controlling water chemistry. The use of end-members defined by the water compositions to calculate Sr sources (but not the partition of major cations) is robust against many of the processes which perturb water compositions. The method is applied to the headwaters and flood plain of the Ganges. In the headwaters we estimate that 50% of the Sr flux and 57 ± 14% of the "excess" $^{87}$Sr flux, which forces changes in seawater Sr-isotopic composition, is derived from silicate. In the flood plain 33 ± 5% of the Sr flux is silicate derived. Overall the 55% of the "excess" $^{87}$Sr flux from the Ganges is silicate-derived. About 60% of the impact of S.E. Asian rivers on seawater $^{87}$Sr/$^{86}$Sr is due to high $^{87}$Sr/$^{86}$Sr ratio silicate or carbonate sources, the remaining ~ 40% being due to the higher than average silicate Sr fluxes.
H43C-0388 1340h
The Dissolved Ca Isotope Composition of Himalayan-Tibetan Waters
Determining the relative proportions of carbonate versus silicate weathering in the Himalaya is important for understanding the long-term atmospheric CO$_2$ budget and the marine Sr isotope record. $^{87}$Sr/$^{86}$Sr is not a straightforward proxy of carbonate to silicate weathering in the Himalaya and up to 50% of the dissolved Ca may be removed by the precipitation of secondary calcite. Ca isotopes have the potential to constrain the relative inputs of carbonates to silicates and incongruent dissolution processes in the weathering environment. Ca is the major cation carried by rivers. Thirty four Himalayan rock and water samples from the Nepal Himalaya and Tibet have been analysed for $^{44/42}$Ca and $^{43/42}$Ca on a Nu-Instruments Multiple Collector -ICP-MS. Unlike the $^{44/40}$Ca ratio the $^{44/42}$Ca is not susceptible to excess $^{40}$Ca production from the decay of K. All samples lie on a single mass fractionation line. There is a total range of 0.4 \permil variation in \delta$^{44}$Ca with values from 0.63 \permil - 0.21 \permil relative to the SRM915a standard. This is comparable to that already reported with \delta$^{44/40}$Ca for small catchments and global rivers. Small first order catchments from each of the main lithotectonic units of the Himalaya have been analysed to examine the effect of lithology on dissolved Ca isotopic composition. In agreement with previous studies elsewhere there is little correlation between source rock and dissolved composition for small rivers spanning a range of source rock from limestone to various silicates and covering a vegetation range from temperate semi-desert to jungle. \delta$^{44}$Ca is not correlated with $^{87}$Sr/$^{86}$Sr or Na/Ca ratios confirming that source rock composition is not the dominant control on the observed range in \delta$^{44}$Ca. A time-series has been examined for the Marsyandi River, central Nepal. In spite of significant systematic variations in major element chemistry including Ca concentration and $^{87}$Sr/$^{86}$Sr the variations in \delta$^{44}$Ca are limited to 0.16 \permil. Either there is only a single isotopic source of Ca or the \delta$^{44}$Ca is controlled by incongruent dissolution processes. The most important incongruent process to affect the Ca budget is the precipitation of pedogenic carbonate. Such incongruent processes should be detectable in the Ca-isotope budget.
H43C-0389 1340h
The Effects of Bedrock age on Radiogenic Marine Isotope Records
Since the discovery of secular changes in radiogenic isotope compositions of seawater in the second half of the last century, a variety of processes capable of forcing such changes has been proposed (glaciations, uplift of mountain ranges, changes in sea level, eruption of flood basalts, submarine hydrothermal activity). Processes affecting continental crust are particularly relevant because continental runoff dominates inputs of most radiogenic isotopes to seawater. Tributaries to the Mississippi and Fraser rivers define positive, linear correlations with different slopes between bedrock age and isotopic composition of the dissolved and particulate Sr, and negative, linear correlations with particulate Nd. These correlations indicate that processes exposing older bedrock to weathering will lead to more radiogenic Sr and less radiogenic Nd fluxes to the ocean. The different slopes are thought to be caused by the different parent/daughter ratios of the end members that define the trends in the respective drainage basins. Though not yet supported by data, similar correlations are likely to exist for other radiogenic isotope systems (Pb, Hf, Os). Based on these observations, I propose that secular changes in the average age of bedrock undergoing weathering are the root cause for changes in the radiogenic isotopic composition of continental runoff. The proposed forcing mechanisms contribute in various ways to changes in the average age of bedrock undergoing weathering. For instance, glaciations and mountain building expose older bedrock through erosion of overlying strata. In contrast, eruption of flood basalts blanket older bedrock with zero-age extrusives. Interestingly, processes that change the globally averaged age of bedrock undergoing weathering often also affect the detrital and dissolved inputs to to the ocean (e.g., glaciation, mountain building), thereby changing both the isotopic composition and the flux of elements to seawater. The proposed hypothesis implies that some processes rarely considered impacting marine isotope records are worthy of closer scrutiny. An example is the formation or destruction of large, internally drained basins, provided the average age of bedrock in such basins differs from the global average. A systematic investigation of global bedrock ages is under way to test this hypothesis and to work towards a reconstruction of changes in bedrock ages throughout the Phanerozoic. The present-day average age of mapped and dated continental crust (600 $\pm$ 125 Myr, 2 S.D.), calculated from large scale bedrock maps, serves as a first Holocene data point for such a reconstruction.
H43C-0390 1340h
Rhenium and Molybdenum in Rivers and Estuaries.
Due to their redox-sensitive nature, the geochemical cycles of Re and Mo are linked to the global organic carbon cycle. Reducing sediments constitute a globally important sink and weathering of organic-rich sediments is responsible for a large portion of the Re and - to a lesser extent - Mo flux to the oceans (Colodner et al., 1993; Jaffe et al., 2002). Riverine concentrations of Re and Mo are a function of the river basin lithology, but are also likely to be affected by anthropogenic contributions (Colodner et al., 1995). Current estimates of global natural riverine Re flux are restricted to single analyses of four major rivers, which characterize only 23%\ of the global freshwater flux (Colodner et al., 1993). Annual variability of Re and Mo concentrations in rivers has not been studied. A single study of Re concentrations along the salinity gradient of the Amazon shelf is suggestive of conservative mixing, but scatter in the data do not allow to exclude the possibility of Re addition in the low-salinity end of the profile (Colodner et al., 1993). Careful evaluation of samples from the Hudson River estuary using a variety of extraction techniques indicates that spike-sample equilibration was not fully achieved using commonly used methods. We have therefore developed a simple, clean and efficient method of extracting Re from filtered water samples. Our method utilizes syringe filtration, prolonged heating to achieve spike-sample equilibration, batch equilibration with TEVA resin, and extraction of Re and Mo using syringe filtration. Rhenium concentrations in the Hudson, Housatonic and Connecticut rivers are 38 pM, 6.6 pM and 14 pM, respectively, much higher than the estimated global average of 2.1 pM (Colodner et al., 1993). Molybdenum concentrations are 4.6 nM, 5.5 nM, 7.8 nM, respectively. These rivers drain basins of Precambrian basement as well as predominantly Paleozoic sediments and have been substantially urbanized. Data for a salinity profile along the Hudson River estuary are suggestive of conservative mixing of Mo. However, data for Re indicate non-conservative mixing with a significant Re source between 5 and 12 psu. This feature could reflect an anthropogenic point source near Croton Point. Alternatively, it could be caused by a process that transfers Re, but not Mo, from the particulate into the operationally defined dissolved phase. Preliminary data for a Mississippi delta salinity profile suggest projected Re and Mo concentrations of about 90 pM and 26 nM for the freshwater end member, respectively. Our results indicate that the surficial Re cycle is more complex then previously thought. It requires further evaluation before a global estimate of natural riverine Re flux can be accurately constrained. A revision of the marine residence time of Re may be required.