V13F-01 INVITED
The changing role of dust in biogeochemical cycling
Dust emission and deposition have the potential to deplete and enrich ecosystems of mineral resources essential to life. In many parts of the world, and particularly in semi-arid settings, wind erosion of soils and the subsequent long-distance transport and deposition of mineral aerosols play a basic role in soil composition and processes, including the production of essential plant nutrients through weathering. Although the long-term role of dust in the development of soils is reasonably well understood, the effects of recent dust emission and deposition on ecosystems are not. Recent work on ecosystems around the world has highlighted the fundamental importance of contemporary wind erosion and dust deposition in biogeochemical cycling. In the western U.S., studies of Sr and Nd isotopes, elemental concentrations, and magnetic properties elucidate the role of dust in recent soil development and soil loss by wind erosion related to land-use change. In the arid landscapes in and around Canyonlands National Park (Utah), these techniques provide insight into the development of soils in stable settings where human activities have been minimal but the loss of soil in areas affected by grazing and recreational activities. In stable settings of the central Colorado Plateau (Utah), dust deposition is responsible for a large proportion (as much as 20 percent) of surface soil mass and elemental content. In contrast, wind erosion is responsible for large losses of nutrients and surface soil of nearby, closely similar geomorphic settings disturbed by human activity. In the San Juan Mountains (Colorado) downwind of the Colorado Plateau, Nd and Sr isotopes in dust and lake sediments provide evidence for large increases in dust deposition during the 19th and 20th century compared to records from the middle to late Holocene. The recent enhancement in dust deposition is also responsible for increased loading of many elements, including essential nutrients that may influence ecological processes. In settings of continued dust accumulation over the past ca. 150 years, geochemical results point to changes in dust composition, particularly in some trace metals and P. The apparent, human-driven change in the amount and composition of emitted dusts has implications for both our understanding and prediction of mineralization processes across a range of landscapes.
V13F-02
Characterizing Dust Inputs to the Caribbean Region Using Radiogenic Isotopes
The long-range transport of mineral aerosols (dust) in the atmosphere influences radiative transfer in the atmosphere and affects ocean productivity via Fe fertilization. Dust transport also affects terrestrial systems by contributing to geochemical fluxes of both sediments and solutes, adding nutrients and nutrient-holding capacity to ecosystems, and neutralizing acidic components of atmospheric deposition. The largest atmospheric dust source is the Sahara-Sahel region of Northern Africa. Of the dust derived from the Sahara-Sahel region, 50 million tons are transported west each year on the trade winds into the Caribbean atmosphere1. Ratios of relatively immobile trace elements provide geochemical evidence that confirms the expected presence of African dust in soils of the Caribbean region. However, estimates of dust deposition fluxes to land in the Caribbean are lacking2. A promising approach for calculating deposition fluxes is to quantify the streamwater Sr flux for a small monolithologic catchment, and then quantify contributions from local substrate, sea salt aerosols, volcanic ash, and long-range transported dust using their unique isotopic signatures. This approach has the advantage of giving a spatially and temporally integrated estimate of the dust deposition flux, which is necessary for assessing the importance of dust to geochemical fluxes and biogeochemical cycling. Many factors will control the importance of dust inputs for a given site. Local soil characteristics will determine the rate at which dust weathers once it enters the soil, and the extent to which dust may contribute to the nutrient budget of the ecosystem. The amount and style of rainfall will affect the rate at which dust particles are scrubbed from the atmosphere. Further, local erosion rates will determine the extent of dust accumulation over time. 87Sr/86Sr ratios are used to calculate the deposition flux of African dust into the small montane Rio Icacos watershed in Puerto Rico of 13 t km-2 yr-1. Nd isotope data show that African dust is incorporated into the regolith to at least 3 m depth. The degree of dust incorporation into the regolith is notable considering the relatively high erosion rates for this mountainous catchment (34 t km-2 yr-1) 3. We discuss the ecological importance of the nutrients (namely phosphorus) provided by the dust to the Rio Icacos ecosystem, as has been shown for dust inputs in Hawaii4, and has been hypothesized for dust inputs in the Amazon basin5. Additionally, considering the Caribbean region as a whole, we attempt to identify those areas where dust inputs may be expected to have the greatest impact on geochemical budgets and biogeochemical cycling. References [1] Colarco et al (2006) JGR-Atmospheres v. 108. [2] Muhs et al (2007) JGR-Atmospheres v. 112. [3] Riebe et al (2003) GCA v.67. [4] Chadwick et al (1999) Nature v. 397. [5] Swap et al (1992) Tellus Series B v. 44.
V13F-03 INVITED
Use of Multiple Isotopic Systems to Interpret Ecosystem Processes in Hawaii
The Hawaiian Islands are an excellent natural laboratory for studying the way in which ecosystems develop and function under varying climates. The mantle-derived basalt parent material provides a constant reaction matrix, the trade winds provide an asymmetric climate pattern that means that the same-age lava flows can be studied under different forcing factors, the relatively few plant species that made it to Hawaii provide a simplified biotic influence on substrate. In essence, we find that the geochemical evolution of basalt weathering provides shifting boundary conditions that constrain ecosystem potentialities, and allows us to apply a number of isotopic systems to enhance the specificity of our interpretation of ecosystem processes. We have applied the following isotopes to assist us in understanding the processes that impact ecosystems: O, C, Sr, Ca, N, Si and Be, and are presently exploring the use of S and Mg. We use these isotopic systems within a matrix of controls that allows us to focus on specific questions. The isotopic signatures from different isotopic systems can define climate- response patterns that are non-linear with each defining different threshold and plateau in rainfall space. Measurement of these isotopic systems allows us to evaluate multiple chemical behaviors at once and to evaluate expected responses to perturbations to any of these tracers in response to past or future changes in climate or other ecosystem drives such as land cover change. For instance, based on deep-soil samples, the plants that grew before humans reached Hawaii have C13 values that drop from -14 per mil to -26 per mil as rainfall increases from 200 mm to 3000 mm. Today the surface-soil values remain close to -14 per mil throughout the rainfall gradient due to the introduction of C4 grasses for pasture. Along the same rainfall gradient, Sr isotopes demonstrate that as C3 plants began to predominate there was a fundmental shift in nutrients supplied from rocks to those supplied by rainfall.
V13F-04 INVITED
Weathering Rates Across Natural Landscape Gradients, Measured Using Cosmogenic Radionuclides
Landscape gradients provide "natural laboratories" for studying how weathering rates respond to environmental conditions. Over the past several years, we have developed methods for measuring long-term (1,000's to 10,000's of years) chemical weathering rates in actively eroding terrain, using measurements of cosmogenic 10Be and the elemental composition of regolith and its parent rock. These methods are applicable to a wide range of field situations, thus facilitating comparative studies of weathering rates and processes. We have used these methods to measure rates of physical erosion and chemical weathering across three elevation transects, spanning up to 1500m in altitude, and across a network of over 40 field sites spanning large gradients in climate and tectonic forcing. Comparisons among these sites show that chemical weathering rates respond to variations in temperature and precipitation, and that they are also strongly coupled to rates of physical erosion. Rates of physical erosion, in turn, respond to variations in hillslope gradients, tectonic forcing, and lithology. Physical erosion rates are also highly episodic, as demonstrated by comparing cosmogenic nuclide measurements with direct measurements of erosion rates over decadal timescales. The view that emerges is of a highly dynamic system in which chemical weathering and physical erosion are coupled to climatic and tectonic forcing, and to each other.
V13F-05
The Impact of Climate and Boundary Conditions on Hillslope Erosion Rates in Northern Chile
Soil production rates on hillslopes (~equivalent to bedrock erosion) are often modeled as a function of curvature and soil depth at the hillslope-scale. However, erosion rates are also dependent on climate and tectonics (both of which affect geomorphology and the soil mass balance), and questions remain regarding the role of climate on rates of soil production. In northern Chile, we measured bedrock erosion rates (using cosmogenic radionuclides 10Be and 26Al) on granitic hillslopes along a climatic gradient (from hyperarid to semiarid) with differing boundary conditions (actively incising channels and stable landforms) to determine how rainfall and hillslope base levels differentially control erosion rates. We found that there is a strong positive correlation between rainfall and average erosion rates due primarily to the shift from slow, abiotic erosive processes on the hyperarid hillslopes to faster biotic processes on the semiarid hillslopes. In the hyperarid region, thick (10-102 cm) soils formed through the accumulation of dust and salt decrease the effectiveness of near-surface processes on the underlying bedrock. In the semiarid region, roots and animal bioturbation drive erosion rates nearly two orders of magnitude higher than in the hyperarid region. The landforms surrounding hillslopes serve as boundary conditions, setting the pace of hillslope processes. As expected, we found that hillslopes bounded by actively incising channels ("active slopes") have higher erosion rates than those bounded by inactive landforms ("stable slopes"). Active slopes appear to be maintaining their grade with channel incision, and thus reflect climatically-driven, regional erosion. In contrast, stable slopes are relaxing towards their stable base levels and away from dynamic steady-state. The climate of northern Chile may have been relatively constant (arid-hyperarid) since the late Pliocene. Given the extremely slow rates of bedrock erosion in the hyperarid and arid regions, most of the hillslope topography is a relict feature from the past, and present day processes are slowly modifying this topography. However, our results suggest that soil thicknesses and erosion rates on the active hillslopes have adjusted to the present climate and there is a positive correlation. This is different from some previous assessments of the impact of climate on denudation, and, when extrapolated to regional landscape evolution, supports the work by others on the potential role of climate in the variation of the Andean landscape.
V13F-06
Cosmogenic 3He in Fe Oxide Weathering Products
We report 3He and 4He concentrations in samples of high-purity goethite and hematite from the uppermost 10 vertical meters of Fe-oxide cement (canga) atop a deep weathering profile at Cajaras, Brazil (6° S, 720m elev.) in the Amazon. 3He concentrations in goethite range from 2x108 atoms/g near the surface to 2x107 atoms/g at 10 meters depth, and in hematite range from 3x108 to 5x107 atoms/g, respectively. With a few exceptions, samples of both minerals show a clear exponential decrease in 3He concentration with increasing depth. However, the samples show no clear relationship between 4He concentration and depth, or between 3He and 4He. Although we consider other potential production mechanisms, we believe the 3He is dominantly cosmogenic in origin and was produced in situ. For instance, nucleogenic 3He from neutron capture by 6Li does not contribute more than ~104 atoms/g of 3He. This low value arises from (i) the low Li content of the Fe oxides (~50 ppb), (ii) the extremely low productivity of neutrons in Fe oxides that are low in light elements and U+Th, and (iii) the comparative youth of the Fe oxides (e.g., goethite (U-Th)He ages < 50 Ma). Assuming a 3He production rate of 100 atoms/g/yr, the surface 3He concentrations require that the Fe oxides have resided near the surface for at least ~3 million years (if the erosion rate is zero), but an exposure age greater than ~10 Myr is implied at erosion rates less than 1 m/Myr. For a mean rock density of approximately 2.5 g/cm3, the apparent attenuation length-scale is much higher than expected for purely spallogenic 3He. Thus it is likely that muon-induced reactions significantly contribute to the measured 3He concentrations at depths greater than ~ 200 g/cm2. Due to their ubiquity as chemical weathering products, cosmogenic nuclides in goethite and hematite offer great potential for the study of surface processes and landscape evolution in environments dominated by chemical weathering.
V13F-07 INVITED
Weathering processes traced with isotopes in a glacial soil chronosequence
There has been numerous studies of isotope tracers to study the processes involved in chemical weathering. In the context of the BigLink project, we have focused on a recent soil chronosequence in the forefield of a retreating glacier in the central Swiss Alps. The bedrock of the soil chronosequence consists of paleozoic granites and gneisses with metamorphic overprint during the Tertiary. The retreat of the Damma glacier has been carefully documented over the past 150 years by swiss scientists such that we now have a fine-scale chronosequence to investigate the early development of soils in a glacial environment. Remarkably, soil construction is extremely rapid and shows the existence of well- developed soils horizons in the first 100 years. Our approach is to sample comprehensively the bedrock, soil horizons, soil solutions, runoff waters as well as plants and soil organic material. Here, we report initial results for Sr, Fe and U-series but the number of studied isotope systems is currently being extended. Our approach allows the derivation of a full mass balance for understanding the cycling of the elements in the soil. Even in the first 100 years of soil development, the soil data shows strong isotope gradients in the secondary phases with an enrichment in the light Fe isotopes in the clay fraction. Our Fe isotope data shows clear evidence for mobilization of iron from the biotite, while the magnetite, the other Fe-carrier in the granite is little affected. Our data shows that light Fe isotopes are preferentially uptaken by plants. We have also explored the stable Sr isotope fractionation in the same soil chronosequence and the most weathered soil horizons are enriched in the heavy strontium isotopes and the waters enriched in the light isotopes, indicating a fractionation of stable Sr isotope during chemical weathering. Although the minerals in the granite do show distinct stable isotope signatures, preferential weathering of one mineral is not consistent with the data. The combination of stable and radiogenic isotope will provide excellent tools for quantifying individual mineral weathering rates. We are currently investigating the source of this fractionation.
V13F-08
The Fingerprint of Present and Past Rainfall on Soil Geochemistry
Research conducted in many locations show that soil weathering rates vary with time and environmental conditions. Here, we assemble long-term (105 to 106 y) chronosequence studies of soil chemistry in sites varying in MAP from ~1 to ~4000 mm y-1 to examine how the rate and magnitude of chemical weathering varies with climate, particularly at the dry end of the spectrum. In humid, vegetated landscapes, soil chemical weathering generally releases an array of rock forming elements that are removed via leaching, causing a subsequent mass loss and volumetric collapse. The rate of this process is non-linear, with instantaneous weathering rates declining greatly with time. In general, the rates and magnitude (for a soil of a given age) of chemical weathering decline greatly with decreasing rainfall. At the arid/hyperarid boundary, where rainfall decreases to the point that biota are essentially absent, chemical weathering nearly ceases, retention of atmospheric solutes and dust increases, and soil chemistry becomes mainly inorganic. The rates of mass gain and volumetric expansion in extremely hyperarid regions is hypothesized to be linear, differing from the non- linearity of processes in more humid regions. While the "fingerprint" of rainfall on soil properties is greatly magnified with increasing soil age, ancient soils commonly bear the imprint of multiple climate changes. In humid regions, detecting these changes in soil chemistry is difficult, whereas an increase in aridity, and a shift from net mass loss to net mass gain, produces a unique geochemical signal, and additionally preserves the weathering signal that occurred during the earlier pluvial episode. In two ancient (Miocene) well-preserved landscapes along a modern (and ancient) rainfall gradient in the Atacama Desert, the geochemical signal of climate change is clearly evident. Paleoclimate reconstructions suggest that northern Chile was under a permanent El Nino-like condition until the late Pliocene, and since that time has generally experienced its present condition of cold upwelling offshore waters and corresponding aridity. Presently, the soils are in arid to hyperarid regions where the main geochemical process is sulfate (north) and carbonate (south) retention, and the soils show enormous gains of these solutes. In stark contrast, the silicate matrix of the soils shows elemental losses of Si, Al, and Fe (and clay formation) that increase with southerly distance (and rainfall), illustrating that past (pre-late Pliocene) geochemical processes differed greatly from modern conditions. These starkly different geochemical signatures (salts over a weathered silicate matrix) reflect soil polygenesis, and have significant implications for Mars soils which contain both weathered silicates and secondary phyllosilicates, and a later stage overlay of sulfates and chlorides.