H13I-01
Nitrate Temporal Dynamics and Controls on Spatial Variability in a Forested Headwater Catchment Following Disturbance
Nitrate (NO3-) leakage from forested watersheds due to disturbance is a well-documented but not well- understood process that can contribute to the degradation of receiving waters through eutrophication. Several studies have shown large-scale defoliation events in small forested watersheds in the Eastern U.S. cause immediate and dramatic increases in NO3- flux to steams with large differences in recovery time. Here, we analyze water-quality and discharge data collected from the time period 1992-2004 following a large-scale gypsy moth defoliation in Shenandoah National Park, Virginia. Following the defoliation, groundwater NO3- concentrations declined exponentially with a distinct seasonal pattern. Initial NO3- groundwater concentrations were related to the magnitude of defoliation within each watershed. Surprisingly, no long-term trend or seasonal pattern were found for soil water NO3- concentrations, as inferred from a mixing model applied to individual storm events. By comparing decay constants associated with groundwater discharge with constants for nitrate recovery to background concentrations, we find a hydrological imprint on the recovery time. This was confirmed by performing similar analysis on data from Hubbard Brook and Coweeta, where more rapid recovery times are attributed to the distinct biogeochemical processes associated with deforestation or crown damage. Synoptic measurements of NO3- concentrations collected on eight occasions within a stream network during the period of recovery are used to fit a model designed to capture the observed spatial variability. We find that upland terrestrial processes, rather than in-stream processes, account for the greatest proportion of this variability.
H13I-02
Isotopic Clues on Factors Controlling Geochemical Fluxes From Large Watersheds in Eastern Canada
A monitoring and monthly sampling program of the Nelson, Ottawa, St. Lawrence, La Grande and Great Whale rivers was started in September 2007. It provides information on the seasonality and sources of geochemical fluxes into the Hudson Bay and the North Atlantic from watersheds covering more than 2.6 106 km2 of the eastern Canadian boreal domain. Measurements of pH and alkalinity, analyses of major ions, strontium and dissolved silica, 2H and 18O of water, concentrations and isotopic properties of dissolved organic and inorganic carbon (13C) and uranium (234U/238U) were performed. Lithology more than latitudinal climatic gradients controls the river geochemistry. Rivers draining silicate terrains show lower dissolved U concentrations but greater 234U/238U disequilibria than rivers draining carbonates (average of 1.38 vs. 1.23). Groundwater supplies might exert some control on these U- isotope signatures. No clear seasonality is observed in 234U/238U ratios, but U concentrations are correlated to dissolved organic carbon (DOC) concentrations in most rivers. Rivers draining carbonates present higher total dissolved carbon concentrations and higher 13C-contents in dissolved inorganic carbon (DIC), in response to the dissolution of soil carbonates. DOC/DIC ratios above 2.4 are observed in rivers draining silicates; their lower 13C-DIC content directly reflects the organic matter oxidation in soils. Total dissolved solids are one order of magnitude or more greater in rivers draining carbonates, showing the strong difference in chemical weathering rates according to the geological setting. The stability in chemical fluxes and water isotopic compositions in the La Grande River, which hosts hydroelectric reservoirs covering more than 12 000 km2, indicates that it is the most buffered hydrological system among the investigated watersheds. Seasonal fluctuations are observed elsewhere, with maximum geochemical fluxes during the spring snowmelt. 2H-18O content of river water appears to be the only parameter presenting a strong latitudinal and climatic gradient (independent of lithology).
H13I-03 INVITED
Dissolved Organic Matter (DOM) Transport in Watersheds: Controls of Sources, Flowpaths and DOM Quality
We investigated the factors and processes influencing the transport of dissolved organic carbon (DOC) and
nitrogen (DON) in forested watersheds. Key questions that we addressed are: What are the sources and
flowpaths for DOC and DON in the watershed? Does DOM quality influence the concentrations and transport
of DOC and DON? How do DOC and DON concentrations change between baseflow and storm event
conditions and across the seasons? How do DOM concentrations and quality change with catchment scale?
These questions are being addressed in a 12 ha unglaciated, forested watershed located in the Piedmont
province of Maryland, USA. Watershed sources being sampled include – rainfall, throughfall, litter, soil water,
groundwater, hillslope seeps and hyporheic zone water. Watershed sources are sampled manually during
non-storm periods (every two weeks) while automated samplers are used for storm events. Water samples
are analyzed for cations, anions, DOC, DON, Silica and O18. DOM quality is characterized using specific
ultraviolet absorbance (SUVA), Fluorescence Index (FI), Excitation-Emission Matrices (EMMs), hydrophobic
and hydrophilic fractions, and concentrations of phenols, carboxylic acids, flavanoids, and amino acids.
Hydrologic monitoring includes discharge and groundwater elevations. Our initial results from endmember
mixing analysis (EMMA) indicate that stream chemistry is regulated by seep runoff, litter, and riparian soil
water with individual endmember contributions varying with size of the storm events and wetness conditions.
Highest DOC and DON concentrations were recorded for throughfall and litter layer. SUVA values were
highest and most variable for litter leachate followed by riparian water. Concentrations of DOC and DON in
streamflow increased dramatically during events and peaked at or after the discharge peak. While SUVA
values followed a similar trend, there were slight differences among events. DOM concentrations in stream
baseflow were influenced more by the percent wetland area in the catchment rather than the size of the
catchment. These observations will help us develop more accurate and reliable models of DOM transport in
the watershed.
http://udel.edu/~inamdar/
H13I-04
Solute Sourcing and Hydrologic Response to Monsoon Precipitation Along a Gradient of Urban Land Use
Urban storm runoff in arid and semiarid areas is used as a potential groundwater recharge resource, but knowledge gaps remain in our understanding on the underlying hydrologic and biogeochemical processes that control the water quality of urban runoff. This study addresses this gap by evaluating how hydrologic and biogeochemical processes interact to produce distinct storm runoff chemistry. We hypothesized that transport processes dominate the solute chemistry of highly urbanized watersheds with large impervious cover; whereas biogeochemical reactions dominate solute responses in less urbanized watersheds with potentially more vegetation and longer flow paths. Utilizing automatic water samplers, we collected urban storm runoff from five distinct urban land use watersheds: 1) low density residential (least urbanized), 2) old medium density residential, 3) new medium density residential, 4) mixed land use and 5) commercial (most urbanized). We coupled a conservative tracer (chloride, Cl-) with stable isotope data (δD and δ18O) to infer physical and biogeochemical processes contributing to the solute chemistry observed. Solute response was similar in the least and most urbanized watersheds, which had the highest mean seasonal concentrations of Cl-, DOC, fecal indicator bacteria (E. coli), Na, Hg and Cu among others, and had the lowest As, Ca and Ni concentrations. The low density site exhibited weak seasonal chloride flushing, contrasting with the commercial site's stronger flushing response. Coupling of Cl-, δD and δ18O data, and comparing it across sites demonstrates solute flushing and evapoconcentration in the commercial site as inferred by δ18O and δD values that plot along an evaporation trend (from -34 to -24 ‰ δD, and -5.3 to -3.5 ‰ δ18O) with increasing Cl- concentrations (from 1.8 to 7.4 mg L-1) during the runoff event. In contrast, high δD values (-27 to -22 ‰) of runoff and a simultaneous decrease in Cl- concentrations (from 11.5 to 3.7 mg L-1) at the low density site suggest watershed solute retention despite runoff evaporation which should concentrate Cl-. Lower δD values of runoff, closer to the meteoric water line, in the commercial site may indicate a shorter flow path when compared to the higher δD values (more evaporated signature) in the low density site. Our study demonstrates that the urban storm runoff quality can not be predicted by land use alone, and supports the study's hypothesis of transport controls on solutes at the most urbanized sites, and flow path and biogeochemical controls at least impervious sites.
H13I-05
Solute Dynamics in a Near-Surface Flowpath-Dominated Forest Catchment
Near-surface flowpaths are common in soilscapes that show a pronounced decrease in permeability and that
are subject to high rainfall amounts and intensities; Lutz Creek catchment in tropical Panama is no exception.
In this catchment, landscape position and antecedent wetness dictate whether overland flow is generated by
saturation excess (SOF) or by return flow (RF). In an ongoing study we explore the consequences of these
topography-controlled modes of overland flow generation for solute dynamics at different scales: catchment
scale and nested subcatchment scale. Total dissolved nitrogen, for example, is higher in RF-dominated than
in SOF-dominated subcatchments. Potassium, in contrast, shows higher concentrations in SOF than in RF,
and SOF resembles throughfall in this respect, whereas RF does not. Calcium is equally suitable to
distinguish between SOF and throughfall on the one hand, and RF on the other. Accordingly, RF-dominated
subcatchments show a dynamic within-event response whereas the calcium signal is nearly flat in SOF-
controlled subcatchments.
We conclude that the topography-controlled and flowpath-mediated hydrochemical signals persist over the
extent of our research catchment.
H13I-06
Modeling the Effects of Hydrological and Biogeochemical Processes on Denitrification and Stream Nitrogen Losses in River Networks
Nitrogen flux in streams is the cumulative result of biogeochemical and hydrological processes that control the supply and transport of nitrogen in terrestrial and aquatic ecosystems. These processes include the effects of denitrification on stream nitrogen removal, which influence the quantities of nitrogen delivered to downstream coastal waters, where increases in nitrogen flux have contributed to eutrophication and hypoxic conditions globally in recent decades. Despite progress in measuring and modeling stream denitrification, few studies have attempted to unravel the coupled effects of biogeochemical (nitrate loadings, concentration, temperature) and hydrological (streamflow, depth, velocity) factors on denitrification and stream nitrogen losses in river networks over space and time. We apply a dynamic nitrogen transport model to assess biogeochemical vs. hydrological effects on seasonal nitrate removal by denitrification in the river networks of two watersheds. The watershed streams have widely differing levels of nitrate concentrations, but similar flows. Unique to our model is the nonlinear dependence of stream denitrification on nitrate concentration, streamflow, and temperature, as determined by regression relations estimated from more than 300 published field measurements available for a variety of U.S. streams. We use these empirical relations to parameterize the nitrogen transport model, which was then applied to the first- through fourth-order stream reaches of the two watersheds. The model results indicate that in-stream nitrate removal by denitrification becomes less efficient as nitrate concentrations and flows increase. This is denoted by the appreciably low percentage of the in-stream nitrate flux (expressed per unit length of stream channel) that is removed in reaches during the highest nitrate concentration and flow months (Feb. to June). The importance of biogeochemical factors (which includes effects of anthropogenic nitrogen loadings, land use, and in-stream biogeochemical factors) is demonstrated as a major control on reach-scale denitrification and in-stream nitrate removal, as evidenced by the disproportionately lower nitrate removal efficiency in stream reaches of the highly nitrate-enriched watershed as compared with that in similarly sized reaches in the less nitrate-enriched watershed. Furthermore, results from model sensitivity analyses confirm the importance of biogeochemical factors (principally nitrate concentration), but suggest that hydrological factors contribute nearly equally to temporal and spatial variations in the percentage of the stream nitrate flux removed in the reaches of each watershed, with water depth indicated as a more important hydrological factor than water velocity.
H13I-07 INVITED
Inferences from a catchment-scale tracer circulation experiment
The coherent description of water flow and solute transport within heterogeneous hydrologic media typically involved at basin scales in response to external rainfall forcings represents a challenge in hydrological modeling. The mechanisms determining the mobilization and transport of solutes in soils through the paths of runoff formation are investigated by analyzing the results of a tracer experiment conducted within an instrumented hillslope draining into a tributary of the Dese river basin (North-eastern Italy). The response of the test catchment has been analyzed by employing two different chemical tracers: nitrates from diffuse agricultural sources (NO3-) and lithium from a point injection (Li+). Rainfall depths, streamflows and pressure heads within different soil horizons have been suitably measured. Inferences from the modeling exercise simulating the observed hydro-chemical response are discussed. Comparative analyses prove instructive, in particular concerning issues on the age of runoff water, an issue that is central to our current understanding of the mechanisms of transport at basin scale.