H41E-0336 INVITED 0800h
Transport and transformation of mercury through soils from contrasting watersheds: Implications for resource management
Watersheds exert a strong influence on Hg cycling, and several studies have shown that characteristics such as soil type, glacial deposits, land use, and land cover control the fate and transport of Hg. Our ongoing research on several Northern Temperate rivers, suggest that more complex influences govern the formation, transport and partitioning of MeHg than previously understood. For instance, forested systems have been considered net sinks for MeHg, mainly delivered via direct atmospheric precipitation. However, forested systems can be a net source of MeHg if both subsurface deposits are highly conductive and proper redox conditions exist, or if discharge to a receiving stream passes through a microbially active hyporheic zone. Particle and colloidal partitioning of MeHg strongly influences bioavailability, and is dependent on surficial deposit types, dissolved organic carbon composition, and the susceptibility of the watershed to erosion. Similarly, inputs to watersheds are predicted to be highly dependent on atmospheric speciation and partitioning of Hg. In the Mercury Experiment To Assess Atmospheric Loading In Canada and the United States (METAALICUS), newly deposited Hg is differentiated from the standing pool using stable isotope amendments. The results will provide the first direct evidence of a watershed response to changing atmospheric inputs of mercury, and will inform pending controls on mercury emissions. Newly deposited mercury is initially retained on the watershed, and subsequent release is dependant on aging and complex soil-solute interactions. Mercury fate assessment models and management scenarios should incorporate these diverse watershed processes.
H41E-0337 INVITED 0800h
Isotopic Tracing of Atmospheric Sulfate in Forest Soils
Stable isotope ratios of sulfur and oxygen are a useful tool for tracing sulfate in terrestrial and aquatic ecosystems. Several studies throughout the last two decades have shown that the isotopic composition of atmospherically deposited sulfate is usually not preserved in aerobic forest soils. Whereas stable sulfur isotope ratios of sulfate in soil solutions tend to remain close to those of atmospheric sulfate, a marked shift to lower oxygen isotope ratios of soil sulfate has been observed between the top of the forest floor and the mineral horizons. This suggests that atmospherically deposited sulfate does not behave conservatively in the unsaturated soil zone, but undergoes redox reactions e.g. via immobilization into organic sulfur compounds and re-mineralization to inorganic sulfate. This implies a significant retardation of atmospherically deposited sulfate in the pedosphere. To identify the sulfur retardation processes in forest soils and to asses their rates, a long-term isotopic tracer study has been conducted. In July 1990, an equivalent of 70 kg S/ha was applied as K2SO4 solution to a forest ecosystem stocked with Norway spruce near Munich (Germany). The sulfur isotope ratio of the deposited sulfate was approximately 25 per mil higher than that of S in soil and seepage water before tracer application. Sulfur isotope ratios for seepage water sulfate collected at 5 cm depth indicated that the tracer passed through the forest floor within four months after application. Mass and isotope balances showed that circa 70 percent of the applied K2SO4 had passed the mineral soil horizons in 20 cm depth 2.5 years after tracer application, but no labeled sulfate was detected in seepage water in 100 cm soil depth during the observation period. Isotope ratio measurements on various sulfur compounds from soil samples obtained in 1991, 1995, 1999 and 2003 confirmed that the majority of the labeled sulfate was retained in the mineral soil for several years. Circa one decade after tracer application some of the labeled sulfur re-appeared in the forest floor. This suggests that substantial amounts of tracer sulfate were taken up by the forest stand and were returned as organic sulfur to the forest floor via needle fall. The study provides evidence that some of the atmospherically deposited sulfate undergoes the immobilization re-mineralization cycle in forested ecosystems, providing a feasible explanation for the observed depletion of oxygen-18 in seepage water sulfate.
H41E-0338 INVITED 0800h
Retention of Atmospheric and Biogeochemically Cycled Nitrogen in a Mediterranean Climate
Southern California has some of the highest rates of atmospheric nitrogen deposition recorded in the world. These high rates of atmospheric deposition have resulted in elevated levels of dissolved nitrogen in some streams in southern California. The levels of nitrogen (overwhelmingly as nitrate) in streams correlate with atmospheric deposition in the region but there is also considerable spatial and temporal variability. The variability in space and time appears to be due to differences in hydrologic flowpath and biogeochemical cycling and how they affect the fate, storage and transport of nitrogen in the environment of the dominant Mediterranean ecosystems of southern California. Over the past several decades catchment scale research in southern California by us and by others has shown several causes for the spatial and temporal differences of nitrogen in theses ecosystems. First, interannual variability appears to be due to nitrate storage within these catchments since wet years following dry years have elevated nitrate concentrations with the reverse also being true. Second, isotopic results recently published indicate that 10% of the nitrate observed at baseflow is direct throughput of atmospherically derived nitrate and during storm events nearly 40% of exported nitrate is throughput of atmospheric nitrate. These high fractions during storm events are likely due in part to direct throughfall into streams. Third, nitrate is well correlated with discharge in any stream in southern California with a significant groundwater flow component, which indicates groundwater storage of nitrate. Fourth, since the water in storm event flows bears a groundwater signature the nitrate observed in stormflows must have undergone some level of storage within the vadose zone/groundwater system. Taken together these results indicate that seasonal and interannual storage of atmospheric nitrogen in southern California catchments is not small and could have a significant control on water quality.
H41E-0339 0800h
A New Twist on the Seasonality of Nitrate Retention and Release in Adirondack Watersheds
Release of nitrate to surface waters in the Northeast has a distinct seasonality that is generally explained by high retention from plant uptake during the growing season, and low retention during the non-growing season, when biological demand is low and soil- water flux is elevated in the absence of transpiration. In the Adirondack region of New York, the highest rates of release, which consistently occur during spring snowmelt, are considered to be the result of nitrate accumulation in the soil and snowpack over the winter. This explanation implies that plants out compete nitrifying bacteria for available ammonium during the growing season. Biweekly and automated high-flow sampling over five years in two tributaries of Buck Creek, in the western Adirondacks, however, has revealed inconsistencies with the conventional view of nitrate retention and release. Although low concentrations of nitrate were measured in stream water during the growing season, concentrations were lowest each year in mid October (near the completion of leaf drop) in the North tributary, and were either the lowest or second lowest each year in mid October in the South tributary. Furthermore, concentrations of nitrate in both watersheds remained elevated throughout the snowmelt periods despite sustained high flows. For example, the concentration in the South tributary on April 9th, 2001, (the initial stage of snowmelt) was 76 micromoles per liter, and on April 24th (following two of the three largest flow events over the 5 years of sampling), was 82 micromoles per liter. Flushing of nitrate stored in the soil over the winter would result in a peak concentration in the stream that would be followed by a rapid decrease. To explain these results we hypothesize a three-way competition that includes heterotrophic non-nitrifying bacteria, as well as plants and autotrophic nitrifying bacteria. Leaf drop in the fall provides a large input of labile carbon with a high C to N ratio (>20) that favors heterotrophs over autotrophs because of the energetic advantage of heterotrophic metabolism. Rapid growth of the heterotrophic community incorporates N into microbial biomass, and leads to low nitrate concentrations in stream water. As fall proceeds into winter, the availability of labile plant-derived carbon decreases, and the populations of active heterotrophs decrease in response, which leads to the release of microbially-derived organic substrate with a low C to N ratio (<10). These conditions are more favorable for nitrifying bacteria, thereby increasing nitrification rates and leaching of nitrate from soils to surface waters. With the onset of snowmelt, soil water flux increases substantially, further lowering the availability of labile carbon (which is primarily in the form of soluble organic matter) and stimulating the nitrifying populations. An inverse relationship between concentrations of dissolved organic carbon and nitrate in stream water of the Buck Creek tributaries supports this interpretation.
H41E-0340 0800h
Controls on inorganic monomeric aluminum release from soils after a clearcut in southeastern New York State, USA
A 24 ha catchment in the Catskill Mountains of southeastern New York was clearcut during the winter of 1996-97. Soil water from the O-, upper B-, and lower B horizons was examined for interactions between inorganic monomeric aluminum (Al$_{im}$) and nitrate (NO$_{3}$$^{-}$), dissolved organic carbon (DOC), pH, and base cations to discern how Al$_{im}$ was released from soils to stream water after the disturbance. Al$_{im}$ at concentrations greater than 2 $\mu$moles l$^{-1}$ can be toxic to some fish species and can inhibit the uptake of calcium by tree roots thereby decreasing tree tolerance to stress. Al$_{im}$ was strongly correlated with NO$_{3}$$^{-}$ in upper and lower B-horizon soil water (r$^{2}$ = 0.67 and 0.68 respectively), but the relation was much weaker in O-horizon soil water (r$^{2}$ = 0.40). O-horizon soil water had the lowest pH values despite having lower NO$_{3}$$^{-}$ concentrations than were measured in the B-horizon; high DOC concentrations in O-horizon soil water suggest that the acidity was partly due to organic acids. The O-horizon also had higher exchangeable base cations than the B-horizon that buffered the inorganic acidity produced by NO$_{3}$$^{-}$ after the clearcut. The high organic content of the O-horizon also allowed for organic complexation of Al as indicated by the strong correlation between DOC and organic monomeric Al (r$^{2}$ = 0.67). Al$_{im}$ concentrations were much higher and DOC concentrations were much lower in B-horizon soil water than in the O-horizon and in the B-horizon the high Al$_{im}$ concentrations persisted for a year longer after the clearcut. Al$_{im}$ concentrations in groundwater seeps were consistently low because mineral dissolution of base cations provided a high buffering capacity; as a result water with high NO$_{3}$$^{-}$ concentration was buffered by base cations rather than by Al$_{im}$. In contrast, B-horizon soil water, which had low buffering capacity and low DOC concentration, contributed large amounts of Al$_{im}$ to stream water, especially at NO$_{3}$$^{-}$ concentrations above 100 $\mu$moles l$^{-1}$, an apparent threshold above which Al$_{im}$ was needed to buffer acidity that exceeded the buffering capacity of base cations within the B-horizon. In conclusion, DOC concentration and base-cation availability were major controls on Al$_{im}$ release from soils after the clearcut, yet, Al$_{im}$ concentrations in stream water draining the catchment greatly exceeded the 2 $\mu$moles l$^{-1}$ toxicity threshold during the first 2 years after the clearcut. An increase in exchangeable Al and a decrease in exchangeable base cations in the soil after the clearcut caused a decrease in the soil Ca:Al ratio; this decrease has been shown to cause tree stress, inhibit regrowth, and result in long-term forest decline.
H41E-0341 0800h
Temporal Shifts in Radiocarbon in Spring Waters: Implications for Decadal Controls on Element Cycling in a Mountain Catchment
Groundwater is a good archive of chemical signals inherited from soil processes. How good it is depends on the residence time of water in aquifers, the amount of mixing of radically different flowpaths, and how well the chemical signals are preserved. Here, we present data collected from nine springs in Sagehen basin, northern Sierra Nevada, California with a focus on the variable influence the soil zone has in regulating the chemistry of recharging waters before they eventually enter the groundwater system. CFC and tritium/3He age-dating methods were used to determine the apparent geochemical ages (or mean residence times) of groundwater emerging from these springs. The apparent ages range between 15 and 43 years and correlate positively with concentrations of rock-derived cations (Ca2+, Na+), conductivity, temperature, and pH, demonstrating the geochemical evolution of a shallow groundwater system. In contrast with the major cations, delta C-13 shows little change with age (values range between -17 and -19 permil), indicating that the carbon chemistry is not evolving. Thus, carbon isotopes record soil processes in this groundwater system. Radiocarbon contents range between 87 and 110 pmC and correlate very well with age, whereby the youngest groundwater has the highest radiocarbon values. This temporal variation reflects the movement of radiocarbon, derived from atmospheric nuclear bomb tests, through the biota and soil zone. Unlike the decadal response time of the atmospheric carbon reservoir, attenuation and lag of the radiocarbon spike recorded in our groundwater data suggest that the turnover time of carbon in the soil zone is relatively long, perhaps 100s to 1000s of years. Additionally, we are investigating other tracers (Ge/Si ratios and silicon isotopes) to elucidate dynamics of element cycling and fractionation in the soil zone, that are in turn recorded in the groundwater chemistry.
H41E-0342 0800h
Atmospheric Nitrate and Limits on the N Cycle in Atacama Desert Soils
Soils of the hyperarid Atacama Desert in northern Chile are renowned for unusually high levels of naturally occurring nitrate and very low levels of organic carbon (OC). Recent work has indicated that the origin of nitrate in the most enriched deposits is largely atmospheric, but these deposits have not been linked to known soil biogeochemical processes, or placed within a global context. Here we investigate the contribution of biological vs. atmospheric sources of nitrate in three soils along a precipitation gradient (25 mm y$^{-1}$ to $<$2 mm y$^{-1}$) in the Atacama, focusing on the links between C and N cycling. A precipitous decline in biological activity accompanies the transition to extreme hyperaridity. As mean annual precipitation (MAP) decreases from 25 mm y$^{-1}$ to $<$2 mm y$^{-1}$ in well-developed (2 My) soils, soil surface OC levels decrease from 40 to 10 $\mu$mol g$^{-1}$, and steady state OC turnover times increase from $\sim$1 y to $\sim$13,000 y. In the same soils, the total nitrate inventory (to $\sim$2 m depth) increases from 0.1 to 17 kg m$^{-2}$, reflecting decreasing rates of loss (leaching or biological) of atmospherically supplied nitrate and ammonium. Nitrate $\Delta$$^{17}$O values in these soils increase from 9.1 to 17.8$\permil$ with decreasing precipitation, approaching the value of atmospherically-derived nitrate: 23.0$\permil$ determined from nitrate collected in passive deposition traps. These data indicate that the biologically unaltered atmospheric portion of soil nitrate increases from 40$%$ to 80$%$ with increasing aridity. Nitrate $\Delta$$^{17}$O values decrease by about 3$\permil$ with increasing (2 m) depth in the most hyperarid soil, suggesting increasingly biological nitrate, possibly from a previous, wetter climate. In the most humid soil (MAP = 25 mm y$^{-1}$), nitrate $\Delta$$^{17}$O values increase with depth, reflecting increased atmospheric nitrate that has been transported below the overlying zone of biological activity. Taken together, nitrate and OC chemistry in these soils reveals that the pervasive nitrate accumulations of the hyperarid core of the Atacama desert are the result of an incomplete soil N cycle, in which slow but continuous inputs of atmospherically-derived nitrate outpace insignificant hydrological losses and biological additions.