Nitrogen dynamics of storm runoff in the riparian zone of a forested watershed

The influence of storm runoff processes on stream nitrogen dynamics was investigated in a headwater riparian swamp on the Oak Ridges moraine in southern Ontario. Hydrologic data were combined with analysis of an isotopic tracer (¹⁸0) and nitrogen (NH ₄ ⁺ , NO ₃ ^– ) concentrations in saturation overland flow and stream discharge. Storm runoff was separated into its event and pre-event components using¹⁸O in order to examine the effect of water source on nitrogen chemistry. Laboratory experiments were also used to study nitrogen transformation associated with storm runoff-surface substrate interactions in the swamp. In most storms N0₃-N and NH₄-N concentrations in the initial 3–4 mm throughfall increment were 10–20x and 20–100x higher respectively than stream base flow concentrations. Maximum stream N0₃-N concentrations were < 2x to 6x higher than base flow concentrations and preceded or coincided with peak stream discharge. Storm-to-storm variations in stream N0₃-N behaviour also occurred during the hydrograph recession phase. NH₄-N concentrations attained an initial peak on the rising hydrograph limb, or at peak stream discharge. A second NH₄-N increase occurred during the late recession phase 3–5 h after maximum stream discharge. Inorganic-N concentrations in surface runoff were similar to peak streamflow.The close agreement between observed N0₃-N concentrations and values predicted from a chemical mixing model indicate that stream N0₃-N variations were controlled mainly by the mixture of throughfall and groundwater in surface stormflow from the swamp. Laboratory experiments also indicated that N0₃-N in surface runoff behaved conservatively when mixed with swamp substrates. With the exception of the late hydrograph recession phase, observed stream NH₄-N concentrations were much lower than concentrations predicted by the chemical mixing model. The rapid loss of NH₄-N from mixtures of surface stormflow and swamp substrates in laboratory experiments and the absence of uptake in sterilized substrates indicated that NH₄-N retention in surface storm runoff was due to biotic processes.     

Water chemistry and periphyton in an alpine wetland

Remote high elevation sites are thought to be good sites to monitor global change and anthropogenic effects on ecosystems. This study was conducted during 1987–1990 in a high elevation wetland (3593 m) located in the Green Lakes Valley, Front Range, Colorado (USA). Salix spp. was the dominant riparian species in this 2 ha. wetland. Small shallow pools (<0.5 m depth) constituted a water area of 236 m³. The major source of water during the study period was snowmelt. The wetland had a well defined outlet and inlet, although an undetermined amount of water entered as groundwater from the snow patch above. Outlet discharge was 424–460 m³ during the month of July and declined thereafter as water input from the snowpatch declined. Inlet discharge was 67% of outlet discharge. Water temperatures in the outlet were always less than 6.8°C, pH 6.0–6.3, and mean conductivity 30.8 µS cm^–1. Both NO _inf3 ^sup– and SO _inf4 ^sup–2 were higher in the inlet thanin the outlet. Dominant cations in the inlet and outlet waters were Ca⁺² Mg⁺² > K⁺ + Na⁺; dominant anions were SO _inf4 ^sup–2 HCO _inf3 ^sup– > NO _inf3 ^sup– Cl^–. Nutrient limitation by P was demonstrated once using nutrient diffusing substrata. No limitation could be shown for NO _inf3 ^sup– , HCO _inf3 ^sup– , or Fe+EDTA. Slow colonization rates of periphyton on tiles were attributed to low temperatures and/or ultraviolet radiation. However, interannual differences in biomass on tiles were as much as 300% after 35 days. A minimum of 16–54 samples would be needed to detect a significant interannual change in biomass on tiles after 35 days assuming that the extreme case for periphyton patchiness. Global climate change is likely to affect discharge and water temperature in this wetland which hill have direct and indirect affects on population dynamics and ecosystem function.     

Nitrate removal in a first-order stream: reconciling laboratory and field measurements

A combination of laboratory and field experiments were carried out to evaluate nitrate(NO ₃ ^t- ) removal during stream transport in a first-order agricultural drainage stream. Intact stream sediment cores overlain with stream and NO ₃ ^– -amended stream water indicated NO ₃ ^– losses averaging 93 — 353 mg m^–2 day^–1, with NO ₃ ^– concentration exerting a primary control on loss rate. Isotopic data indicated enrichment of NO ₃ ^– - ¹⁵N over time as NO ₃ ^– concentrations decreased, indicating a denitrification loss. Field experiments were designed to evaluate dilution of streamwater with low-NO ₃ ^– groundwater in addition to other NO ₃ ^– removal processes during transport. A series of bromide tracer and NO ₃ ^– - addition experiments were carried out in the field; groundwater dilution dominated the downstream NO ₃ ^– concentration trends, accounting for all observed decreases in NO ₃ ^– concentration. Isotopic data did not point to denitrification downstream as a major NO ₃ ^– removal process. This apparent disparity between simulated laboratory and in-situ stream removal rates appears to be a function of the hydrological processes controlling exchanges between stream bottom sediments and the overlying water. These results suggest that caution must be exercised in extrapolating potentials for NO ₃ ^– removal measured in laboratory experiments to the field, as these rates could be overestimated in some watersheds.     

Effect of whole catchment liming on the episodic acidification of two adirondack streams

During the fall of 1989 7.7Mg/ha of calcium carbonate was applied on two tributary catchments (40 ha and 60 ha) to Woods Lake, a small (25 ha) acidic headwater lake in the western Adirondack region of New York. Stream-water chemistry in both catchment tributaries responded immediately. Acid-neutralizing capacity (ANC) increased by more than 200 eq/L in one of the streams and more than 1000 eq/L in the other, from pre-liming values which ranged from –25 to +40 eq/L. The increase in ANC was primarily due to increases in dissolved Ca²⁺ concentrations. Most of the initial response of the streams was due to the dissolution of calcite that fell directly into the stream channels and adjacent wetlands. A small beaver impoundment and associated wetlands were probably responsible for the greater response observed in one of the streams.After the liming of subcatchmentIV (60 ha), Ca²⁺ concentrations increased with increasing stream discharge in the stream during fall rain events, suggesting a contribution from calcite dissolved within the soil and transported to the stream by surface runoff or shallow interflow. Concentrations of other ions not associated with the calcite (e.g. Na⁺) decreased during fall rain events, presumably due to mixing of solute-rich base flow with more dilute shallow interflow. The strong relation between changes in Ca²⁺ and changes in NO ₃ ^– concentrations during spring snowmelt, (r² = 0.93, slope = 0.96, on an equivalent basis) suggests that both solutes had a common source in the organic horizon of the soil. Increases in NO ₃ ^– concentrations during snowmelt were balanced by increases in Ca²⁺ that was released either directly from the calcite or from exchange sites, mitigating episodic acidification of the stream. However, high ambient NO ₃ ^– concentrations and relatively low ambient Ca²⁺ concentrations in the stream during the spring caused the stream to become acidic despite the CaCO₃ treatment.In stream WO2 (40ha), Ca²⁺ concentrations were much higher than in stream WO4 because of the dissolution of calcite which fell directly into the upstream beaver pond and its associated wetlands. Calcium concentrations decreased as both NO ₃ ^– concentrations and stream discharge increased, due to the dilution of Ca-enriched beaver pond water by shallow interflow. Despite this dilution, Ca²⁺ concentrations were high enough to more than balance strong acid anion (SO ₄ ^– , NO ₃ ^– , Cl^–) concentrations, resulting in a positive ANC in this stream throughout the year. These data indicate that liming of wetlands and beaver ponds is more effective than whole catchment liming in neutralizing acidic surface waters.     

Groundwater nitrogen dynamics at the terrestrial-lotic interface of a small catchment in the Central Amazon basin

Processes operating at the terrestrial-lotic interface may significantly alter dissolved nitrogen concentrations in groundwater as a result of shifting redox conditions and microbial communities. We monitored concentrations of total dissolved nitrogen, NO ₄ ^– , NH ₄ ^– , O₂ and Fe²⁺ for 10 months along two transects tracing groundwater flow from an upland (terra firme) forest, beneath the riparian forest, and into the stream channel of a small Central Amazonian catchment. Our aim was to examine the role of near-stream processes in regulating groundwater transfers of dissolved nitrogen from terrestrial to lotic ecosystems in the Central Amazon. We found pronounced compositional differences in inorganic nitrogen chemistry between upland, riparian, and stream hydrologic compartments. Nitrate dominated (average 89% of total inorganic nitrogen; TIN) the inorganic nitrogen chemistry of oxygenated upland groundwater but decreased markedly upon crossing the upland-riparian margin. Conversely, NH ₄ ^– dominated (average 93% of TIN) the inorganic chemistry of apparently anoxic riparian groundwater; NH ₄ ^– and TIN concentrations decreased markedly across the riparian-stream channel margin. In the oxygenated streamwater, NO ₃ ^– again dominated (average 82% of TIN) inorganic nitrogen chemistry. Denitrification followed by continued ammonification is hypothesized to effect the shift in speciation observed at the upland-riparian margin, while a combination of several processes may control the shift in speciation and loss of TIN observed at the riparian-stream margin. Dissolved organic nitrogen concentrations did not vary significantly between upland and riparian groundwater, but decreased across the riparian-stream margin. Our data suggest that extensive transformation reactions focused at the upland and stream margins of the riparian zone strongly regulate and diminish transfers of inorganic nitrogen from groundwater to streamwater in the catchment. This suggestion questions the veracity of attempts in the literature to link stream nitrogen chemistry with nutrient status in adjacent forests of similar catchments in the Central Amazon. It also complicates efforts to model nitrogen transfers across terrestrial-lotic interfaces in response to deforestation and changing climate.     

Dissolved organic nitrogen budgets for upland, forested ecosystems in New England

Relatively high deposition ofnitrogen (N) in the northeastern United States hascaused concern because sites could become N saturated.In the past, mass-balance studies have been used tomonitor the N status of sites and to investigate theimpact of increased N deposition. Typically, theseefforts have focused on dissolved inorganic forms ofN (DIN = NH₄-N + NO₃-N) and have largelyignored dissolved organic nitrogen (DON) due todifficulties in its analysis. Recent advances in themeasurement of total dissolved nitrogen (TDN) havefacilitated measurement of DON as the residual of TDN– DIN. We calculated DON and DIN budgets using data onprecipitation and streamwater chemistry collected from9 forested watersheds at 4 sites in New England. TDNin precipitation was composed primarily of DIN. Netretention of TDN ranged from 62 to 89% (4.7 to 10 kgha^{minus 1} yr^{minus 1}) of annual inputs. DON made up themajority of TDN in stream exports, suggesting thatinclusion of DON is critical to assessing N dynamicseven in areas with large anthropogenic inputs of DIN.Despite the dominance of DON in streamwater,precipitation inputs of DON were approximately equalto outputs. DON concentrations in streamwater did notappear significantly influenced by seasonal biologicalcontrols, but did increase with discharge on somewatersheds. Streamwater NO₃-N was the onlyfraction of N that exhibited a seasonal pattern, withconcentrations increasing during the winter months andpeaking during snowmelt runoff. Concentrations ofNO₃-N varied considerably among watersheds andare related to DOC:DON ratios in streamwater. AnnualDIN exports were negatively correlated withstreamwater DOC:DON ratios, indicating that theseratios might be a useful index of N status of uplandforests.     

The biogeochemistry of sulfur at Hubbard Brook

A synthesis of the biogeochemistry of S was done during 34 yr(1964–1965 to 1997–1998) in reference and human-manipulated forestecosystems of the Hubbard Brook Experimental Forest (HBEF), NH. There have beensignificant declines in concentration (–0.44µmol/liter-yr) and input (–5.44mol/ha-yr)of SO₄ ^2– in atmospheric bulk wet deposition, and inconcentration(–0.64 µmol/liter-yr) an d output (–3.74mol/ha-yr) of SO₄ ^2– in stream water ofthe HBEF since 1964. These changes arestrongly correlated with concurrent decreases in emissions of SO₂from the source area for the HBEF. The concentration and input ofSO₄ ^2– in bulk deposition ranged from a low of 13.1µmol/liter (1983–1984) and 211 mol/ha-yr(1997–1998) to a high of 34.7 µmol/liter(1965–1966) and 479 mol/ha-yr (1967–1968), with along-term mean of 23.9 µmol/liter and 336mol/ha-yr during 1964–1965 to 1997–1998. Despiterecentdeclines in concentrations, SO₄ ^2– is the dominantanion in both bulk deposition and streamwater at HBEF. Dry deposition is difficult to measure, especially inmountainousterrain, but was estimated at 21% of bulk deposition. Thus, average totalatmospheric deposition was 491 and 323 mol/ha-yr during1964–1969 and 1993–1998, respectively. Based on the long-term³⁴S pattern associated with anthropogenic emissions,SO₄ ^2– deposition at HBEF is influenced by numerousSO₂sources, but biogenic sources appear to be small. Annual throughfall plusstemflow in 1993–1994 was estimated at 346 molSO₄ ^2–/ha. Aboveground litterfall, for thewatershed-ecosystemaveraged about 180 mol S/ha-yr, with highest inputs (190 molS/ha-yr) in the lower elevation, more deciduous forest zone. Weatheringrelease was calculated at a maximum of 50 mol S/ha-yr. Theconcentration and output of SO₄ ^2– in stream waterranged from a low of 42.3µmol/liter (1996–1997) and 309 mol/ha-yr(1964–1965), to a high of 66.1 µmol/liter(1970–1971) and 849 mol/ha-yr (1973–1974), with along-term mean of 55.5 µmol/liter and 496mol/ha-yr during the 34 yrs of study. Gross outputs ofSO₄ ^2– in stream water consistently exceeded inputsin bulkdeposition and were positively and significantly related to annualprecipitationand streamflow. The relation between gross SO₄ ^2–output and annual streamflow changed with time asatmospheric inputs declined. In contrast to the pattern for bulk depositionconcentration, there was no seasonal pattern for streamSO₄ ^2– concentration. Nevertheless, stream outputs ofSO₄ ^2– were highly seasonal, peaking during springsnowmelt, andproducing a monthly cross-over pattern where net hydrologic flux (NHF) ispositive during summer and negative during the remainder of the year. Nosignificant elevational pattern in streamwaterSO₄ ^2– concentration was observed. Mean annual,volume-weightedsoil water SO₄ ^2– concentrations were relativelyuniform by soil horizon andacross landscape position. Based upon isotopic evidence, much of theSO₄ ^2– entering HBEF in atmospheric depositioncycles throughvegetation and microbial biomass before being released to the soil solution andstream water. Gaseous emissions of S from watershed-ecosystems at HBEF areunquantified, but estimated to be very small. Organic S (carbon bonded andestersulfates) represents some 89% of the total S in soil at HBEF. Some 6% exists asphosphate extractable SO₄ ^2– (PSO₄).About 73% of the total S in the soilprofile at HBEF occurs in the Bs2 horizon, and some 9% occurs in the forestfloor. The residence time for S in the soil was calculated to be 9 yr, butonly a small portion of the total organic soil pool turns over relativelyquickly. The S content of above- and belowground biomass is about 2885mol/ha, of which some 3–5% is in standing dead trees. Yellowbirch, American beech and sugar maple accounted for 89% of the S in trees, with31% in branches, 27% in roots and 25% in the lightwood of boles. The pool of Sin living biomass increased from 1965 to 1982 due to biomass accretion, andremained relatively constant thereafter. Of current inputs to the availablenutrient compartment of the forest ecosystem, 50% is from atmospheric bulkdeposition, 24% from net soil release, 11% from dry deposition, 11% from rootexudates and 4% is from canopy leaching. Comparing ecosystem processes for Sfrom 1964–1969 to 1993–1998, atmospheric bulk deposition decreasedby 34%, stream output decreased by 10%, net annual biomass storage decreased by92%, and net soil release increased by 184% compared to the 1964–1969values. These changes are correlated with decreased emissions of SO₂from the source area for the HBEF. Average, annual bulk deposition inputsexceeded streamwater outputs by 160.0 ± 75.3 SD molS/ha-yr,but average annual net ecosystem fluxes (NEF) were much smaller, mostlynegativeand highly variable during the 34 yr period (–54.3 ± 72.9 SDmol S/ha-yr; NEF range, +86.8 to –229.5). While severalmechanisms may explain this small discrepancy, the most likely are netdesorption of S and net mineralization of organic S largely associated with theforest floor. Our best estimates indicate that additional S from dry depositionand weathering release is probably small and that desorption accounts for about37% of the NEF imbalance and net mineralization probably accounts for theremainder (60%). Additional inputs from dry deposition would result fromunmeasured inputs of gaseous and particulate deposition directly to the forestfloor. The source of any unmeasured S input has important implications for therecovery of soils and streams in response to decreases in inputs of acidicdeposition. Sulfate is a dominant contributor to acid deposition at HBEF,seriously degrading aquatic and terrestrial ecosystems. Because of the strongrelation between SO₂ emissions and concentrations ofSO₄ ^2– in both atmospheric deposition and streamwater at HBEF,further reductions in SO₂ emissions will be required to allowsignificant ecosystem recovery from the effects of acidic deposition. Thedestruction or removal of vegetation on experimental watershed-ecosystems atHBEF resulted in increased rates of organic matter decomposition andnitrification, a lowering of soil and streamwater pH, enhancedSO₄ ^2– adsorption on mineral soil and smallerconcentrations andlosses of SO₄ ^2– in stream water. With vegetationregrowth, this adsorbedSO₄ ^2– is released from the soil, increasingconcentrations andfluxes of SO₄ ^2– in drainage water. Streamwaterconcentration ofSO₄ ^2– and gross annual output ofSO₄ ^2–/ha are essentially the same throughout theHubbard BrookValley in watersheds varying in size by about 4 orders of magnitude, from 3 to3000 ha.     

Controls of streamwater dissolved inorganic carbon dynamics in a forested watershed

Iinvestigated controls of stream dissolved inorganic carbon (DIC) sources andcycling along a stream size and productivity gradient in a temperate forestedwatershed in northern California. Dissolved CO₂ (CO_2(aq))dynamics in heavily shaded streams contrasted strongly with those of larger,open canopied sites. In streams with canopy cover > 97%, CO_{2 (aq)}was highest during baseflow periods (up to 540 M) and wasnegatively related to discharge. Effects of algal photosynthesis on CO_2(aq) were minimal and stream CO_{2 (aq)} was primarily controlledby groundwater CO_{2 (aq)} inputs and degassing losses to theatmosphere. In contrast to the small streams, CO_{2 (aq)} in larger,open-canopied streams was often below atmospheric levels at midday duringbaseflow and was positively related to discharge. Here, stream CO_2(aq) was strongly influenced by the balance between autotrophic andheterotrophic processes. Dynamics of HCO₃ ^– werelesscomplex. HCO₃ ^– and Ca²⁺ were positivelycorrelated, negatively related to discharge, and showed no pattern with streamsize. Stable carbon isotope ratios of DIC (i.e. ¹³C DIC)increased with stream size and discharge, indicating contrasting sources of DICto streams and rivers. During summer baseflows, ¹³C DIC were¹³C-depleted in the smallest streams (minimum of–17.7) due to the influence of CO_{2 (aq)} derived frommicrobialrespiration and HCO₃ ^– derived from carbonateweathering. ¹³C DIC were higher (up to –6.6)inthe larger streams and rivers due to invasion of atmospheric CO₂enhanced by algal CO_{2 (aq)} uptake. While small streams wereinfluenced by groundwater inputs, patterns in CO_{2 (aq)} and evidencefrom stable isotopes demonstrate the strong influence of stream metabolism andCO₂ exchange with the atmosphere on stream and river carbon cycles.     

Concentration and flux of solutes from snow and forest floor during snowmelt in the West-Central Adirondack region of New York

Decreases in pH and increases in the concentration of Al and NO ₃ ^– have been observed in surface waters draining acid-sensitive regions in the northeastern U.S. during spring snowmelt. To assess the source of this acidity, we evaluated solute concentrations in snowpack, and in meltwater collected from snow and forest floor lysimeters in the west-central Adirondack Mountains of New York during the spring snowmelt period, 29 March through 15 April 1984.During the initial phase of snowmelt, ions were preferentially leached from the snowpack resulting in elevated concentrations in snowmelt water (e.g. H⁺ = 140 eq.l^–1; NO ₄ ^2– = 123 eq.l^–1; SO ₃ ^– = 160 eq.l^–1). Solute concentrations decreased dramatically within a few days of the initial melt (< 50 eq.l^–1). The concentrations of SO ₄ ^2– and NO ₃ ^– in snowpack and snowmelt water were similar, whereas NO^– ₃ in the forest floor leachate was at least two times the concentration of SO ₄ ^2– .Study results suggest that the forest floor was a sink for snowmelt inputs of alkalinity, and a net source of H⁺, NO ₃ ^– , dissolved organic carbon, K⁺ and Al inputs to the mineral soil. The forest floor was relatively conservative with respect to snowmelt inputs of Ca²⁺, SO ₄ ^2– and Cl^–. These results indicate that mineralization of N, followed by nitrification in the forest floor may be an important process contributing to elevated concentrations of H⁺ and NO ₃ ^– in streams during the snowmelt period.     

15N evidence for the origin and cycling of inorganic nitrogen in a small Amazonian catchment

The ¹⁵N composition of the dominant form of dissolved inorganic nitrogen (DIN) was determined in upland groundwater, riparian groundwater, and stream water of the Barro Branco catchment, Amazônas, Brazil. The ¹⁵N composition of organic nitrogen in riparian and upland leaf litter was also determined. The data for these waters could be divided into three groups: upland groundwater DIN predominately composed of NO₃ ^– with ¹⁵N values averaging 6.25 ± 0.9 riparian groundwater DIN primarily composed of NH₄ ⁺ with ¹⁵N values averaging 9.17 ± 1.0 and stream water DIN predominately composed of NO₃ ^– with ¹⁵N values averaging 4.52 ± 0.8 Nitrate samples taken from the stream source and from the stream adjacent to the groundwater transects showed a downstream increase in ¹⁵N from 1.0to 4.5 Leaf litter samples averaged 3.5 ± 1.2The observed patterns in isotopic composition, together with previously observed inorganic nitrogen species and concentration shifts between upland, riparian and stream waters, suggest that groundwater DIN is not the primary source of DIN to the stream. Instead, the isotopic data suggest that remineralization of organic nitrogen within the stream itself may be a major source of stream DIN, and that the majority of DIN entering the stream via groundwater flowpaths is removed at the riparian-stream interface.     

Patterns in groundwater nitrogen concentration in the floodplain of a subtropical stream (Wollombi Brook,New South Wales)

The sources of groundwater and the patterns in groundwater dissolved N and DOC concentration in the floodplain of a subtropical stream (Wollombi Brook, New South Wales) were studied over a 2-year period using three piezometer transects. While the stream was generally a discharge area for regional groundwater, this source represented only a small contribution to either the water or N budget of the alluvial aquifer. Groundwater–surface water interactions appeared mostly driven by cycles of bank recharge and discharge between the stream and the alluvial aquifer. DON and NH₄⁺ were the principal forms of dissolved N in groundwater, consistent with the primarily suboxic to anoxic conditions in the alluvial aquifer. A plume of groundwater NO₃⁻ was found at one transect where oxic conditions persisted within the riparian zone. The origin of the NO₃⁻ plume was hypothesized to be soil NO₃⁻ from the riparian zone flushed to the water table during recharge events. When present, NO₃⁻ did not reach surface water because conditions in the alluvial aquifer in the vicinity of the stream were always reduced. The concentration of groundwater DOC was variable across the floodplain and may be related to the extent of the vegetation cover. Overall, transformation and recycling of N during lateral exchange processes, as opposed to discharge of new N inputs from regional groundwater, appears to primarily control N cycling during groundwater–surface water interactions in this subtropical floodplain.     

Dissolved organic carbon dynamics in the peat–streamwater interface

A series of experiments were conducted to address the fate of dissolved organic carbon (DOC) in the peat–stream interface zone linking a minerotrophic poor fen and an ombrotrophic mire with surrounding stream water in the drainage area of Lake Örträsket in northern Sweden. Transport and mineralisation of DOC were quantified in peat–stream interface cores in response to variations in pore water velocity, DOC concentration and the molecular size and source of DOC. Mineralisation and CH₄ production were positively correlated with pore water velocity at rates between 0.08 and 0.20cmh^–1 and negatively correlated at rates between 0.20 and 0.40cmh^–1. The DOC concentration of the effluent from the peat cores was independent of the pore water velocity but proportional to the DOC concentration of the source water. Higher concentrations of DOC were exported from than imported to the peat cores, and the cores exported DOC molecules of smaller average molecular size than received. Carbon mineralisation in the peat, assessed in a static system, was independent of the concentration of DOC. DOC with a nominal cutoff at 100Da was mineralised faster by streamwater bacteria than DOC dialysed with a cutoff at 3500Da, and their mineralisation rate was positively correlated with the DOC concentration. Streamwater bacteria mineralised streamwater DOC at a lower rate than the peat–stream interface zone pore water DOC. The pattern of velocity dependence of mineralisation was the same for both sources of peat DOC but the mineralisation rates, average molecular size, and bioavailability of DOC were different, emphasising the importance of the compositional heterogeneity of the peat–stream interface zone for the DOC budget of streamwater.     

Sulphate dynamics in a northern wetland catchment during snowmelt

Fluxes and stores of SO ₄ ^2– were measured in a small Canadian Shield basin during the 1989 snowmelt. Sulphate flux from the unsaturated zone (14.1 ± 7.3 kg ha^–1) was four times the amount supplied in meltwater and precipitation (3.5 ± 0.4 kg ha^–1). This reflects flushing of soluble S04- from organic and upper mineral soil horizons during melt, which counteracted potential dilution of groundwater SO ₄ ^2– levels by large water inputs to the basin. 35.6 ± 12.4 kg SO ₄ ^2– entered the saturated zone during melt, supplied equally by leaching from overlying soils and conversion of the capillary fringe to phreatic water due to rising water table levels. Streamflow conveyed 70% of the total SO ₄ ^2–1 export of 10.1 ± 2.3 kg ha^–1, and was largely supplied by groundwater discharge from a wotland in the lower portion of the basin. The remaining 30% of total export was via shallow subsurface flow. Results highlight the importance of unsaturated and saturated zone processes for SO ₄ ^2– dynamics and export during snowmelt.     

The importance of groundwater in the transportation of allochthonous dissolved organic matter to the streams draining a small mountain basin

The importance of groundwater in the dissolved organic matter (DOM) budget of small upland streams is not well understood. This paper is concerned with the amount of streamflow which can be attributed to groundwater, the organic chemistry of rainwater, streamwater, and groundwater, and the rate of transfer of DOM in groundwater to the streams of a small mountain catchment basin in Alberta. Using naturally occurring isotopes (¹⁸O and tritium) groundwater is concluded to be the largest contributor to stream discharge throughout the year. This means that most of the water which reaches the stream must pass through the soil column and be exposed to microbial attack. Groundwater in the Marmot Basin spends an average of about ten years in the ground before being discharged into streams. In this area it appears that the majority of DOM from forest productivity is consumed in the soil and only small amounts of refractory by-products reach the stream. This is in keeping with the finding of Fisher & Likens (1973) that 99% of forest productivity is consumed terrestrially. It is probable that bacteria in stream sediments are capable of taking up refractory compounds which deep soil bacteria can not. Increases in DOM concentration in streams are not usually observed during storm runoff because of the ability of bacteria to take up groundwater DOM and because most of stream discharge is groundwater low in DOM being flushed into the channel even during snowmelt and rainfall events.This work was supported by a Subvention from Environment Canada, Inland Waters Directorate, and partly by an operating grant from the National Research Council of Canada.This work was supported by a Subvention from Environment Canada, Inland Waters Directorate, and partly by an operating grant from the National Research Council of Canada.     

Spatial and temporal patterns of nitrogen concentrations in pristine and agriculturally-influenced prairie streams

Long-term data on nitrogen chemistry of streams draining Konza Prairie Biological Station (Konza), Kansas were analyzed to assess spatial and temporal patterns and examine the influence of agricultural activity on these patterns. Upland watersheds of Konza are predominantly tallgrass prairies, but agricultural fields and riparian forests border the lower reaches of the streams. We have up to 11 years of data in the relatively pristine upland reaches and 4 years of data on wells and downstream reaches influenced by fertilized croplands. Seasonal and spatial patterns in total nitrogen (TN) concentrations were driven largely by changes in the nitrate (NO₃ ^–) concentrations. A gradient of increasing NO₃ ^– concentrations occurred from pristine upland stream reaches to the more agriculturally-influenced lowland reaches. Nitrate concentrations varied seasonally and were negatively correlated with discharge in areas influenced by row-crop agriculture (p = 0.007). The NO₃ ^– concentrations of stream water in lowland reaches were lowest during times of high precipitation, when the relative influence of groundwater drainage is minimal and water in the channel is primarily derived from upland prairie reaches. The groundwater from cropland increased stream NO₃ ^– concentrations about four-fold during low-discharge periods, even though significant riparian forest corridors existed along most of the lower stream channel. The minimum NO₃ ^– concentrations in the agriculturally influenced reaches were greater than at any time in prairie reaches. Analysis of data before and after introduction of bison to four prairie watersheds revealed a 35% increase of TN concentrations (p < 0.05) in the stream water channels after the introduction of bison. These data suggest that natural processes such as bison grazing, variable discharge, and localized input of groundwater lead to variation in NO₃ ^– concentrations less than 100-fold in prairie streams. Row-crop agriculture can increase NO₃ ^– concentrations well over 100-fold relative to pristine systems, and the influence of this land use process over space and time overrides natural processes.     

Effect of Reduced Winter Precipitation and Increased Temperature on Watershed Solute Flux, 1988–2002, Northern Michigan

Since 1987 we have studied weekly change in winter (December–April) precipitation, snowpack, snowmelt, soil water, and stream water solute flux in a small (176-ha) Northern Michigan watershed vegetated by 65–85 year-old northern hardwoods. Our primary study objective was to quantify the effect of change in winter temperature and precipitation on watershed hydrology and solute flux. During the study winter runoff was correlated with precipitation, and forest soils beneath the snowpack remained unfrozen. Winter air temperature and soil temperature beneath the snowpack increased while precipitation and snowmelt declined. Atmospheric inputs declined for H⁺, NO₃⁻, NH₄⁺, dissolved inorganic nitrogen (DIN), and SO₄²⁻. Replicated plot-level results, which could not be directly extrapolated to the watershed scale, showed 90% of atmospheric DIN input was retained in surface shallow (<15 cm deep) soils while SO₄²⁻ flux increased 70% and dissolved organic carbon (DOC) 30-fold. Most stream water base cation (C_B), HCO₃⁻, and Cl⁻ concentrations declined with increased stream water discharge, K⁺, NO₃⁻, and SO₄²⁻ remained unchanged, and DOC and dissolved organic nitrogen (DON) increased. Winter stream water solute outputs declined or were unchanged with time except for NO₃⁻ and DOC which increased. DOC and DIN outputs were correlated with the percentage of winter runoff and stream discharge that occurred when subsurface flow at the plot-level was shallow (<25 cm beneath Oi). Study results suggest that the percentage of annual runoff occurring as shallow lateral subsurface flow may be a major factor regulating solute outputs and concentrations in snowmelt-dominated ecosystems.     

Solute dynamics in soil water and groundwater in a central Amazon catchment undergoing deforestation

Hydrochemical changes caused by slash-and-burnagricultural practices in a small upland catchment inthe central Amazon were measured. Soluteconcentrations were analyzed in wet deposition,overland flow, shallow throughflow, groundwater andbank seepage in a forested plot (about 5 ha) and anadjacent plot (about 2 ha) which had been deforestedin July 1989 and planted to manioc, and in streamwater in partially deforested and forested catchments. Measurements were made from November 1988 to June1990. The effects of slash-and-burn agriculturalpractices observed in the experimental plot includedincreased overland flow, erosion, and large losses ofsolutes from the rooted zone. Concentrations ofNO₃ ^-, Na⁺, K⁺, SO₄ ²-,Cl^- and Mn in throughflow of the experimentalplot were higher than those of the control plot bymore than a factor of 10. Extensive leaching occurredafter cutting and burning, but solute transfers werediminished along pathway stages of throughflow togroundwater, and particularly within the riparian zoneof the catchment. High concentrations of N and P inoverland flow indicate the importance of usingforested riparian buffers to mitigate solute inputs toreceiving waters in tropical catchments.     

Biogeochemistry of unpolluted forested watersheds in the Oregon Cascades: temporal patterns of precipitation and stream nitrogen fluxes

We analyzed long-term organic and inorganic nitrogen inputs and outputs in precipitation and streamwater in six watersheds at the H.J. Andrews Experimental Forest in the central Cascade Mountains of Oregon. Total bulk N deposition, averaging 1.6 to 2.0 kg N ha^–1 yr^–1, is low compared to other sites in the United States and little influenced by anthropogenic N sources. Streamwater N export is also low, averaging <1 kg ha^–1 yr^–1. DON is the predominant form of N exported from all watersheds, followed by PON, NH₄-N, and NO₃-N. Total annual stream discharge was a positive predictor of annual DON output in all six watersheds, suggesting that DON export is related to regional precipitation. In contrast, annual discharge was a positive predictor of annual NO₃-N output in one watershed, annual NH₄-N output in three watersheds, and annual PON output in three watersheds. Of the four forms of N, only DON had consistent seasonal concentration patterns in all watersheds. Peak streamwater DON concentrations occurred in November-December after the onset of fall rains but before the peak in the hydrograph, probably due to flushing of products of decomposition that had built up during the dry summer. Multiple biotic controls on the more labile nitrate and ammonium concentrations in streams may obscure temporal DIN flux patterns from the terrestrial environment. Results from this study underscore the value of using several watersheds from a single climatic zone to make inferences about controls on stream N chemistry; analysis of a single watershed may preclude identification of geographically extensive mechanisms controlling N dynamics.     

Runoff sources and land cover change in the Amazon: an end-member mixing analysis from small watersheds

The flowpaths by which water moves from watersheds to streams has important consequences for the runoff dynamics and biogeochemistry of surface waters in the Amazon Basin. The clearing of Amazon forest to cattle pasture has the potential to change runoff sources to streams by shifting runoff to more surficial flow pathways. We applied end-member mixing analysis (EMMA) to 10 small watersheds throughout the Amazon in which solute composition of streamwater and groundwater, overland flow, soil solution, throughfall and rainwater were measured, largely as part of the Large-Scale Biosphere-Atmosphere Experiment in Amazonia. We found a range in the extent to which streamwater samples fell within the mixing space determined by potential flowpath end-members, suggesting that some water sources to streams were not sampled. The contribution of overland flow as a source of stream flow was greater in pasture watersheds than in forest watersheds of comparable size. Increases in overland flow contribution to pasture streams ranged in some cases from 0% in forest to 27?C28% in pasture and were broadly consistent with results from hydrometric sampling of Amazon forest and pasture watersheds that indicate 17- to 18-fold increase in the overland flow contribution to stream flow in pastures. In forest, overland flow was an important contribution to stream flow (45?C57%) in ephemeral streams where flows were dominated by stormflow. Overland flow contribution to stream flow decreased in importance with increasing watershed area, from 21 to 57% in forest and 60?C89% in pasture watersheds of less than 10?ha to 0% in forest and 27?C28% in pastures in watersheds greater than 100?ha. Soil solution contributions to stream flow were similar across watershed area and groundwater inputs generally increased in proportion to decreases in overland flow. Application of EMMA across multiple watersheds indicated patterns across gradients of stream size and land cover that were consistent with patterns determined by detailed hydrometric sampling.