H51G-01 INVITED 08:00h
Seasonality of Diel Cycles of Dissolved Trace-Metal Concentrations in a Rocky Mountain Stream
Substantial diel (24-hour) cycles in dissolved (0.1-$\mu$m filtration) metal concentrations were observed during summer low flow, winter low flow, and snowmelt runoff in Prickly Pear Creek in southwestern Montana. The stream was alkaline (pH of 7.65-9.06), and dissolved metal concentrations were relatively low (1.8-7.1 $\mu$g/L for As, 18-57 $\mu$g/L for Mn, and 12-123 $\mu$g/L for Zn). The metals are derived from abandoned mine lands in the stream's headwaters; As also is derived from geothermal sources. During seven diel sampling episodes, each lasting 34-61.5 hours, concentrations of dissolved Mn and Zn increased from minimum values in the afternoon to maximum values shortly after sunrise. The timing of diel cycles of dissolved As concentrations exhibited the inverse pattern. The magnitude of concentration increases during individual 24-hour periods ranged from 17-152% for Mn and 70-500% for Zn, and correlated positively with the magnitude of diel increases of pH and temperature, indicating that geochemical processes involving reactive inorganic and organic surfaces on and in the streambed probably control these diel metal cycles. Diel increases of As concentrations (17-55%) were proportionally smaller and less variable among the seasonal sampling episodes than for Mn and Zn, and they correlated poorly with diel increases of pH and temperature. Streamflow among the seven sampling episodes ranged from 0.35-3.3 m$^{3}$/s. The timing of minimum and maximum values of diel streamflow cycles was inconsistent among sampling episodes and had little relation to the timing of metal concentration cycles, indicating that hydrological processes are not a primary control of diel metal cycles. Diel cycles of dissolved metal concentrations may occur at any time of year and during various hydrologic conditions in all streams with dissolved metals and neutral to alkaline pH.
H51G-02 INVITED 08:15h
The Fine Structure of Water-Quality Dynamics: the Wave of the Future in Catchment Hydrochemistry?
Until recently, most catchment hydrochemical studies have been based on hourly or sub-hourly measurements of water fluxes, and weekly or monthly samples of rainfall and streamflow chemistry. Trying to infer catchment hydrochemical dynamics from weekly or monthly grab samples, however, is like trying to understand a Beethoven symphony when one can only hear a single note every minute or two! New technological developments have now made it possible to monitor rainfall and streamflow chemistry at hourly or sub-hourly intervals -- similar to the timescales at which hydrometric data have long been available -- and to provide these measurements for long spans of time, not just for intensive field campaigns associated with individual storms. Here we explore how high-frequency chemical monitoring data are likely to transform catchment hydrology and biogeochemistry. We evaluate how the information content of a chemical time series depends on the sampling interval. As the sampling frequency increases, from weekly to daily to hourly, the information content of the chemical time series also increases, in some cases dramatically. However, above some sampling frequency (which varies from catchment to catchment) additional measurements will merely "connect the dots" between the existing observations, and will therefore add little additional information. We offer practical advice for estimating the frequency at which additional measurements become redundant. Experience to date shows that high-frequency chemical observations have led to important serendipitous discoveries. Continuous high-frequency monitoring of catchment hydrochemistry requires significant resources and tenacity, but what we stand to learn is well worth the effort.
H51G-03 08:30h
Effects of Photoirradiation on Natural Organic Matter: Importance for Studies of Stream Hydrobiogeochemistry
Natural organic matter (NOM) is a key component of lakes, streams, and wetlands, and it plays an important role in many biogeochemical processes. We investigated changes in molecular weight (by size exclusion chromatography, SEC), fluorescence, and UV/Vis absorbance of bulk filtered surface water NOM (SW) and XAD-8 and XAD-4 isolates upon irradiation in a solar simulator to 48 hours. Upon irradiation of SW and XAD-8 samples, the higher molecular weight NOM components were initially degraded to form more fluorescent intermediate molecular weight components, along with lower molecular weight components. After several hours of irradiation, the intermediate molecular weight fluorescent components degraded; little or no production of lower molecular weight components was detected by SEC. Hence, the overall sample fluorescence increased within the first few hours of irradiation and then decreased. The XAD-4 isolate did not contain high molecular weight components that could break down to form additional fluorescent components; therefore, irradiation resulted in a more simple and continuous decrease in molecular weight and fluorescence. Ongoing studies are focusing on the potential role of Fe in the observed changes to the NOM properties. These results demonstrate that photo-induced changes in NOM properties are a complex function of sample composition as well as amount of exposure to sunlight. This research further indicates that studies of NOM-related processes in surface waters need to consider the potential for a complex sequence of photo-induced changes to the NOM properties that are likely to affect reactivity.
H51G-04 08:45h
Photo-oxidation of Dissolved Organic Matter in River Water and its Effect on Trace Element Speciation
Photochemical effects on the chemistry of fresh surface waters are now well established. One interesting contrast in the studies of photochemical effects on dissolved organic carbon (DOC) versus trace metals is the difference in time scales generally considered. For DOC studies, timescales of days to weeks are most common in experiments whereas for trace metals, it is the diel cycle that tends to be studied. We conducted a three-week incubation in natural light (with dark controls) of filtered water from the lower Pearl River (Mississippi) examining both the changes in DOC and changes in physical-chemical speciation of a suite of trace metals. During the incubation, DOC decreased in the light by about 20% while UV light absorbance decreased by nearly 40%. This implies both the photo-oxidation of the DOC as well as a shift to a proportionately less aromatic nature for the DOC. For the trace elements, a variety of behaviors were observed. Some elements showed no change in speciation; for instance, the alkali and alkaline earth metals as well as some oxyanions (e.g., Mo) and some other elements (e.g., Mn). Other elements, however, did show significant changes in the light. Fe, for example, is a key trace element in this system with a colloidal concentration over 2 $\mu$M. There was a significant, continuous decrease in dissolved ($<$0.02 $\mu$m) Fe in the light samples during the experiment. This is best explained by release of organically-complexed Fe during photo-oxidation of the low-to-medium molecular weight fraction of the DOC followed by subsequent precipitation of the released Fe as additional colloidal iron(III) oxyhydroxide. A number of other elements (Ce, Cu, Cr, Pb, V, and U) also showed decreases in the dissolved ($<$0.02 $\mu$m) fraction with time, implying a release from low molecular weight complexes followed by sorption onto Fe colloids. All of these elements have been previously found to be associated with colloidal Fe in other systems. A number of these elements also showed decreases in their retention by an anion exchange column, likewise implying a decrease in the organically-complexed form. Our results suggest that as fluvial DOC becomes more photo-refractory upon exposure to sunlight, there will be a transfer of certain trace elements from dissolved, complexed forms to colloidal or particulate adsorbed forms. This should lead to fundamental differences in metal speciation, transport, and bioavailability between low-order streams with fresh allochthonous carbon inputs and floodplain rivers dominated by photo-refractory carbon.
H51G-05 09:00h
Tracing Nitrogen Sources in Forested Catchments Under Varying Flow Conditions: Seasonal and Event Scale Patterns
Our ability to assess how stream nutrient concentrations respond to biogeochemical transformations and stream flow dynamics is often limited by datasets that do not include all flow conditions that occur over event, monthly, seasonal, and yearly time scales. At the Sleepers River Research Watershed in northeastern Vermont, USA, nitrate, DOC (dissolved organic carbon), and major ion concentrations were measured on samples collected over a wide range of flow conditions from summer 2002 through summer 2004. Nutrient flushing occurred at the W-9 catchment and high-frequency sampling revealed critical insights into seasonal and event-scale controls on nutrient concentrations. In this seasonally snow-covered catchment, the earliest stage of snowmelt introduced nitrogen directly to the stream from the snowpack. As snowmelt progressed, the source of stream nitrate shifted to flushing of soil nitrate along shallow subsurface flow paths. In the growing season, nitrogen flushing to streams varied with antecedent moisture conditions. More nitrogen was available to flush to streams when antecedent moisture was lowest, and mobile nitrogen stores in the landscape regenerated under baseflow conditions on times scales as short as 7 days. Leaf fall was another critical time when coupled hydrological and biogeochemical processes controlled nutrient fluxes. With the input of labile organic carbon from freshly decomposing leaves, nitrate concentrations declined sharply in response to in-stream immobilization or denitrification. These high-resolution hydrochemical data from multiple flow regimes are identifying "hot spots" and "hot moments" of biogeochemical and hydrological processes that control nutrient fluxes in streams.
H51G-06 09:15h
Time Series Stream Temperature And Dissolved Oxygen Modeling In The Lower Flint River Basin
The tributaries of the Lower Flint River Basin (LFRB) are incised into the upper Floridan semi-confined limestone aquifer, and thus seepage of relatively old groundwater sustains baseflows and provides some control over temperature and dissolved oxygen fluctuations. This hydrologic and geologic setting creates aquatic habitat that is unique in the state of Georgia. Groundwater withdrawals and possible water supply reservoirs threaten to exacerbate low flow conditions during summer droughts, which may force negative impacts to stream temperature and dissolved oxygen (DO). To evaluate the possible effects of human modifications to stream habitat, summer time series (in 15 min interval) of stream temperature and DO were monitored over the last three years along these streams, and a Continuously Stirred Tank Reactor (CSTR) model was developed and calibrated with these data. The driving forces of the diel trends and the overall levels of stream temperature and DO were identified by this model. Simulations were conducted with assumed managed flow conditions to illustrate potential effects of various stream flow regimes on stream temperature and DO time series. The goal of this research is to provide an accurate simulation tool to guide management decisions.
H51G-07 09:30h
Diel Variation of Oxygen and Inorganic Carbon in a High Productivity, High Alkalinity Stream: Biological and Geochemical Controls
Dissolved oxygen (O$_{2}$) and inorganic carbon (DIC) measurements are frequently used for assessing the metabolic balance and ecological function of aquatic systems. In many cases, oxygen and DIC dynamics are related to each other with fixed stoichiometry (C:O$_{2}$ approximating 1:1) based upon photosynthesis and/or respiration. This stoichiometry has proven reproducible (with some variation) in systems where the oxygen and carbon balances are dominated by biological processes. However, high alkalinity aquatic systems may have an added geochemical control on daily DIC dynamics causing deviation of C:O$_{2}$ stoichiometry from that fixed by biological processes alone. As part of a study examining stream metabolism in agricultural watersheds, we measured diel variation of dissolved O$_{2}$, \delta $^{18}$O$_{2}$, DIC, and \delta $^{13}$C-DIC in a low-order, high-alkalinity stream, mid-continent USA to determine biological and geochemical controls on O$_{2}$ and DIC dynamics and stoichiometry. High rates of primary production generated large diel fluctuations in O$_{2}$ concentration (up to 50 percent above O$_{2}$ saturation during the day and 20 percent undersaturation at night). \delta $^{18}$O$_{2}$ isotopes varied inversely with O$_{2}$ concentration and were enriched by up to 12 permil at night. \delta $^{13}$C-DIC exhibited a 3.5 permil diel variation. Changes in DIC concentration were inversely correlated to changes in O$_{2}$ concentration at a C:O$_{2}$ ratio of nearly 3:1 indicating the influence of both biological and geochemical controls on the DIC pool. Relative contributions of stream metabolism and abiotic carbonate reactions were examined with a numerical model that included primary production, respiration, and carbonate reactions to simulate diel variation in O$_{2}$ and DIC concentration and isotopes.
H51G-08 09:45h
The Impact of Flow on Spatial Variability of Streamwater Chemistry in a Boreal Catchment
Spatial variability in water chemistry is one defining characteristic of stream ecosystems. An infrequently studied aspect of this variability is the impact of flow regime: for instance, does the variation in water chemistry within a watershed differ between times of baseflow and flood? We examined this issue in Krycklan, a largely forested till-underlain boreal catchment in Northern Sweden, with an emphasis on changes occurring during the annual spring flood. Utilizing 15 stream sites ranging in catchment size from 0.07-68 km$^{2}$ and in catchment wetland coverage from 0-40%, the between-site variance in several water parameters was compared between times of stable, relatively low flow and times of dramatically changing, high flow during spring flood. Distinctly different patterns were observed for different solutes. While dissolved organic carbon (DOC) varied widely between streams at baseflow (5-40 mg/l) and was positively correlated with catchment % wetland area, these differences were subdued during peak flood (DOC range 15-25 mg/l), and as such variability was lower. Conversely, differences in inorganic cation concentrations (electrical conductivity ranged from 15-30 $\mu$S/cm) were amplified during the spring flood, resulting in increased spatial variability. Much of the change in variability hinged on the contrasting behavior of streams draining catchments with high % wetland compared to forested catchments: wetland streams had lower concentrations of cations at baseflow, and were also diluted more dramatically than forest streams during the flood. Hydrological flowpaths were explored using isotopic and chemical signatures of streamwater and catchment source waters. Much of the observed patterns in spatial variability could be explained by divergent flowpaths within wetland and forest catchments during spring flood, and by differing source areas for DOC vs. cations.