HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (13) Citing Articles via Google Scholar Google Scholar Articles by Hagedorn, F. Articles by Flühler, H. Search for Related Content PubMed Articles by Hagedorn, F. Articles by Flühler, H. GeoRef GeoRef Citation Agricola Articles by Hagedorn, F. Articles by Flühler, H.

DIVISION S-7-FOREST & RANGE SOILS

The Role of Rapid Flow Paths for Nitrogen Transformation in a Forest Soil

A Field Study with Micro Suction Cups

^a Swiss Federal Inst. for Forest, Snow and Landscape Research (WSL), Zürcherstr. 111, CH-8903 Birmensdorf, Switzerland
^b Inst. of Plant Biology/Microbiology, Univ. of Zürich, Zollikerstr. 107, CH-8008 Zürich, Switzerland
^c Inst. of Terrestrial Ecology, ETH Zürich, Grabenstr. 11, CH-8952 Schlieren, Switzerland

hagedorn{at}wsl.ch

	ABSTRACT

TOP ABSTRACT INTRODUCTION Materials and methods Results and discussion Conclusions REFERENCES

 
Preferential flow is a common phenomenon in soils. This study was conducted to investigate the significance of rapid flow paths for N transformation in a forested Humaquept in central Switzerland. Fifty micro suction cups, each with a surface area of 12 mm², were installed in a regular grid in the uppermost 5 cm. First, the location of each "microcup" relative to main flow paths was estimated based on the response to applications of a dye, SO^2-₄, and Cl^-. Then, a N-addition experiment was carried out to study the N transformation at locations along flow paths and within the soil matrix. Only 23 of 50 microcups responded to the application of the dye within the first 24 h, which indicates that a large portion of the soil volume is not in contact with the infiltrating rainwater. Those microcups which responded to the added dye were regarded to be located along flow paths. At depths below 2 cm, under temporarily reducing conditions, sampling locations in or near flow paths had higher NO^-₃ concentrations (20–25 µM) than those of the soil matrix (below 12 µM). Within 24 h after a simulated rainfall, the ratio decreased more in the flow paths (between -2.4 and -4.9 mol mol^-1) than in the soil matrix (-0.7 to -0.8 mol mol^-1), which indicates an enhanced denitrification at these locations. In the subsequent dry period, nitrification started 2 d earlier and was more pronounced along flow paths. The results of this study suggest that flow paths are microhabitats with an increased N transformation compared with the soil matrix.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION Materials and methods Results and discussion Conclusions REFERENCES

 
FORESTS IN LARGE areas of Europe and North America are subjected to elevated N deposition, which may impair the functioning of these ecosystems. Increased N inputs may cause N saturation when the retention capacity of the system is exceeded (Aber et al., 1989). Nitrogen saturation is associated with increased leaching of NO^-₃ from the soil compartment (Aber et al., 1989; Gundersen et al., 1998). Generally, N saturation and thus NO^-₃ leaching from forests is thought to occur when inorganic N is in excess of total plant and microbial demand. However, the ability of forest soils to retain increased N deposition also depends on the flow regime in the soil compartment. Field studies have shown that solutes bypass a large portion of the soil matrix (Flury et al., 1994; Stagnitti et al., 1995; Stamm et al., 1998). Travel times of solutes may be shorter than a priori expected (Stamm et al., 1998). Only a small part of the soil may come into contact with the infiltrating rainwater.

Soil biologists point out that, analogous to the heterogeneous nature of solute movement in the soil, microbial activity is heterogeneously distributed in the soil. This is attributed to the patchy distribution of organic matter (Parkin, 1987), to soil aggregation (Tiedje et al., 1984; Cambardella and Elliott, 1993; Chotte et al., 1998), and to the distribution of pore sizes (Juma, 1993), which lead to manifold microhabitats for soil biota. In this field study, we investigated how preferential flow paths are linked to locations of preferred N transformation in the soil. Habitats near or in flow paths are likely to provide a better supply of N, and since flow paths are more permeable, they drain more rapidly and are better aerated than the rest of the soil matrix.

Studies focusing on microscale variability of microbial processes are either visual (Morgan et al., 1991; Stamatiadis et al., 1990) or destructive by incubation of spatially separated soil material (e.g., Cambardella and Elliott, 1993; Nishio, 1994; Abbasi and Adams, 1998). Another approach is the in situ measurement of O₂ concentrations (Bakker and Bronswijk, 1993) and redox potentials (Cogger et al., 1992; Flessa and Beese, 1995) using microelectrodes. However, these two methods only provide information about the physicochemical conditions of microhabitats but not about the transformation of nutrients. Micro suction cups combined with a recently developed analytical technique, capillary electrophoresis, enables collecting and analysis of small samples of soil solution (Göttlein et al., 1996; Göttlein and Matzner, 1997). In contrast to destructive spatial separation of soil, micro suction cups allow the temporal dynamic of the in situ nutrient transformation to be estimated at a small scale. Using a dye, Cl^-, and SO^2-₄ as tracers, we located the micro suction cups relative to flow paths.

The objectives of this study were (i) to identify the spatial distribution of NO^-₃ concentrations and their relation to the main flow paths in the uppermost 5 cm of a Humaquept, (ii) to investigate the N transformations within the reach of flow paths and the soil matrix, and (iii) to explore the significance of microscale variability for the effects of increased N deposition on NO^-₃ leaching at a plot scale.

	Materials and methods

TOP ABSTRACT INTRODUCTION Materials and methods Results and discussion Conclusions REFERENCES

 
Study Site
The study area is located in the Alptal (47°03N, 8°43E) in the Prealps of central Switzerland. Mean annual air temperature is 6°C and precipitation averages 2300 mm yr^-1. The parent rock material is Flysch, from which heavy clay soils, mainly Humaquepts, have formed. The water table is close to the surface throughout the year.

Vegetation and soil types form a mosaic pattern that is closely related to the microtopography. Norway spruce (Picea abies L.) grows on the mounds, where the water table is below the 40-cm depth. Soils on the mounds have an oxidized Bw horizon and a forest floor. In the depressions, the water table frequently reaches the surface, leading to a muck topsoil with a high carbonate content. The mineral soil is a permanently reduced Bg horizon (Table 1) . The experiment with the micro suction cups was conducted in a depression with a dense ground floor vegetation, mainly marsh marigold (Caltha palustris L.), white butterbur [Petasites albus (L.) Gaertn.], and roughstalk bluegrass (Poa trivialis L.). Selected soil properties are given in Table 1.

The water discharge from the subcatchments was measured with "V"-notch weirs. Runoff proportional samples, bulk deposition, and throughfall were collected weekly (Schleppi et al., 1998).

Additionally, soil solution was sampled on five plots (20 m²) each with and without increased N deposition. Nitrogen was sprinkled simultaneously onto the NITREX subcatchments. Horizontally installed porous glass filter plates (diameter 90 mm, maximum pore diameter <16 µm) were used to sample at the 5-, 10-, and 20-cm depths, and suction cups were used at the 30-, 50-, and 100-cm depths. Samples were collected weekly and bulked every 2 wk.

Soil cores (5.5-cm diam., 10 cm) were taken monthly for measuring extractable N. Roots were removed, and the soil homogenized and shaken for 1 h with 0.5 M K₂SO₄ at a 1:4 soil/solution ratio. Analysis of NO^-₃ in the filtrates (filter model 589/3, Schleicher and Schuell, Dassel, Germany) was conducted photometrically at 210 nm according to Norman and Stucki (1981) with a Shimadzu UV-160 spectrophotometer (Shimadzu Scientific, Columbia, MD).

Denitrification was measured by the acetylene inhibition technique using a static core system (6.3-cm diam., 10 cm) (Ryden et al., 1987). Polypropylene tubes were inserted into the topsoil several weeks before sampling to allow equilibration of the disturbed soil system. The tubes were transferred into jars fitted with neoprene septa, and incubated at soil temperature in an atmosphere of 10-kPA acetylene. Gas samples were taken after 2, 3.5, and 5 h. Nitrous oxide concentrations were measured with a Perkin-Elmer gas chromatograph (model 8500, Perkin Elmer, Norwalk, CT) with a ⁶³Ni electron capture detector. The seasonal denitrification pattern was monitored from May 1996 to October 1997. The measurements were executed every 2 weeks on all 10 plots.

Redox potential measurements were performed with Pt electrodes (Pfisterer and Gribbohm, 1989; Cogger et al., 1992) inserted into the soil and an Ag/AgCl reference electrode placed into a piezometer within 50 cm of the redox electrodes. The Eh readings were conducted with a portable pH/mV meter (pH 320 Wissenschaftlich-Technische Werkstätten, Weilheim, Germany). Redox electrodes were installed at 1-, 2-, 3-, 4-, 5-, 10- and 20-cm depths (three per depth) 3 wk before the N-addition experiment. All electrodes were tested with a redox buffer K₄[Fe(CN)₆]/K₃[Fe(CN)₆] before and after the experiment.

Construction and Installation of Micro Suction Cups
The microcups were constructed by inserting and gluing a small ceramic plate (4-mm diameter, 1.5 mm thick) into a pipette tip (50–200 µL) (Fig. 1) . With a corer, the small ceramic plate was obtained from a large ceramic plate. The microcups were connected to 0.3-mm-i.d. teflon tubing (1/16'') and then glued onto plexiglass pipes (5-mm i.d.; 6.5-mm o.d.). The dead volume of the microcups and the teflon tubes was 40 µL. All microcups were individually tested for leaks. In the field, microcups were able to maintain a suction of 40 kPa for 2 wk. Prior to installation, all suction cups were first flushed with 1 M HCl and subsequently rinsed with distilled water.

View larger version (38K): [in this window] [in a new window]   Fig. 1 Micro suction cups: construction and instrumental array in the field

Fifty microcups were arranged in a systematic 5 by 20 cm grid to minimize the interaction between adjacent cups (Fig. 1). The vertical distance between the microcups was 1 cm; the lateral distance was 2 cm. The uppermost row of microcups at the 1-cm depth was installed at a horizontal distance of 10 cm from the plexiglass plate. The deeper rows of suction cups were installed 2 cm further into the soil.

Soil water samples were taken by connecting the microcups to an evacuation device, which allowed individual sampling from each microcup. The bottom of the evacuation device was filled with distilled water to produce water vapor–saturated conditions, thereby reducing evaporation losses from the samples. Sampling started in July 1997, 6 wk after the installation. In the period from installation to the first sampling, the microcups were flushed and allowed to equilibrate with the surrounding soil.

To estimate the microscale spatial variability, soil solution was collected for three consecutive weeks by applying a permanent suction of 30 kPa. The typical sample volume collected ranged from 200 µL to 10 mL and averaged 3.7 mL of soil water. Samples were frozen and stored until analysis.

In order to evaluate the relevance of microcup sampling of soil water on the larger-scale, additional suction plates were installed at 5-, 10-, and 20-cm depths (two per depth) within 50 cm of the microcups.

Identification of the Location of Microcups with Brilliant Blue
To identify the location of each microcup relative to the flow paths, we observed the response to infiltrating water, spiked with a Brilliant Blue FCF solution (1 g BB L^-1; Hoechst, Frankfurt, Germany), homogeneously sprinkled with a backpack sprayer onto an area of 50 by 50 cm above the microcups. The arrival time of the Brilliant Blue FCF solution at each cup was visually observed, and the solution was collected for 1 d. Microcups with arrival times of <24 h were considered to be within reach of rapid flow paths (fast response) and those >24 h to tap the soil matrix (slow response).

The homogeneity of tracer applications were tested with 100-cm² containers placed on the ground. Five containers were used during the Brilliant Blue addition and 10 during the simulated rainfalls. The coefficient of variation were always below 10%.

Microscale Nitrogen Transformation Experiment
In order to investigate the in situ nitrification and denitrification at the microscale, N-addition experiments were carried out during August 1997.

Ammonium Addition
A solution with a high NH⁺₄ concentration corresponding with the mean NH⁺₄ input of 2 wk (18 kg NH⁺₄–N ha^-1 yr^-1) was added with a backpack sprayer onto a surface of 2 m² around the microcups. First, the plot was irrigated with 8 mm of deionized water for 4 h. Then, 2.5 mmol (NH₄)₂SO₄ m^-2 dissolved in 2 mm water was applied within 10 min. During the next 2 h, an additional 4 mm of deionized water was added.

Nitrate Addition
During the second experiment, we applied Cl^-, as a conservative tracer, and NO^-₃ with an automated sprinkling device (described by Flury et al., 1994) onto a surface of 3.5 m² around the microcups. Since the experiment was designed to investigate the NO^-₃ transformation in the topsoil, rapid leaching was avoided by applying NO^-₃ at the end of the simulated rainfall. First, 15 mm of deionized water were applied within 5 h. Then, 5 mmol KNO₃ m^-2 and 5 mmol KCl m^-2 dissolved in 1.4 mm of rainwater were sprinkled within 10 min onto the plot. The NO^-₃ addition corresponded with the mean input of 2 wk (18 kg NO₃–N ha^-1 yr^-1). Directly after the tracer application, 2 mm of deionized water was sprinkled onto the plot to flush the sprinkling device. After these additions, the field plot was covered with a plastic sheet to prevent precipitation and additional N inputs.

A suction of 30 kPa was applied to the microcups and 5 kPa to the suction plates at 5-, 10-, and 20-cm depths during both the NH⁺₄ and the NO^-₃ additions. The solutions of the microcups from each depth were grouped into flow paths and soil matrix on the basis of the response of the microcups to the previous Brilliant Blue addition. To obtain sufficient volumes for the analysis, the solutions were bulked (50 µL per microcup) per group and depth, except for the first and the last sampling date, when the microcups yielded sufficient water for separate analysis. All samples were immediately frozen and stored until analysis.

Denitrification activity was measured with three pairs of cores during the NH⁺₄ addition and four pairs of cores during the NO^-₃ addition. The cores were inserted 3 wk before the experiment within a distance of 1.5 m to the microcups.

Laboratory Analysis
Microcup soil solutions were analyzed by capillary electrophoresis (BioFocus 3000, BioRad, Hercules, CA), which requires 5 to 10 nL per analytical run (Göttlein and Blasek, 1996). Major anions (Cl^-, NO^-₃, SO^2-₄) were measured using a pyromellitic acid buffer system (3 mM pyromellitic acid adjusted to pH 8 with TEMED [Fluka Chemical, Buchs, Switzerland]). Analysis of cations (Ca²⁺, Mg²⁺, K⁺, NH⁺₄) was performed with a metol buffer system (5 mM metol, 5 mM ascorbic acid, 2 mM 18-crown-6) (Göttlein and Blasek, 1996). Separation was conducted at 20°C with 20 kV voltage. Samples were injected by vacuum. Detection wavelength was 230 nm for anions and 220 nm for cations. Detection limits were 5 µM in the case of Cl^-, NO^-₃, and SO^2-₄, and 10 µM for NH⁺₄. Due to the minimal available volume, samples were not filtered prior to analysis.

Brilliant Blue FCF concentration was measured photometrically at a wavelength of 630 nm with a Shimadzu UV-160 spectrophotometer (Shimadzu Scientific).

Soil solutions sampled by suction plates and suction cups were filtered (0.45 µm; ME25, Schleicher and Schuell). Anions (Cl^-, NO^-₃, SO^2-₄) in the filtrates were measured by ion chromatography (DX-120, Dionex, Sunnyvale, CA).

Soil C contents were measured with a Carlo Erba C/N analyzer (Carlo Erba, Milan, Italy).

	Results and discussion

TOP ABSTRACT INTRODUCTION Materials and methods Results and discussion Conclusions REFERENCES

 
Spatial Distribution of Nitrate Concentrations and Redox Potentials
The depth and the frequency distribution of NO^-₃ in the soil solution sampled by the microcups and bulked during three consecutive weeks are shown in Fig. 2 . As expected, NO^-₃ concentrations decreased with soil depth. The main drop of NO^-₃ concentrations was found below the 3-cm depth, where redox potential measurements indicated NO^-₃-reducing conditions (Fig. 3) .

View larger version (19K): [in this window] [in a new window]   Fig. 2 Distribution of mean NO-3 concentrations and relative frequencies at individual depths. Means and standard errors of three sampling dates

View larger version (23K): [in this window] [in a new window]   Fig. 3 Soil redox potentials at different depths and water table position during the N-addition experiment. Lines represent individual redox electrodes

As observed for NO^-₃ concentrations, the measured redox potentials in the mineral soil below the 2-cm depth varied across a broad range, indicating the coexistence of anaerobic and aerobic microsites (Fig. 3). The temporal dynamic of individual electrodes shows the simultaneous occurrence of active and inactive microsites. At inactive locations (16 of 21 electrodes), the redox potentials remained stable, whereas the redox potential at active locations (5 of 21) exhibited a pronounced temporal dynamic after rainfall events.

Flow Regimes and Microcups Locations
In order to explain the microscale variability of NO^-₃ in the topsoil, the relative location of each microcup with respect to flow paths was estimated by the response to a pulse input of a dye tracer. We assumed that the fast responding microcups showing a high recovery of the Brilliant Blue FCF solution were located within the reach of flow paths. Microcups showing a slow response to the dye tracer input were regarded to be located within the soil matrix.

Twenty-seven of 50 microcups did not respond to the dye tracing, which indicates that a large portion of the soil volume did not come into contact with the dye solution within 24 h of application. The observed infiltration pattern was surprising, since preferential flow, studied in agricultural soils, has mainly been observed in subsoils (Flury et al., 1994; Heijs et al., 1996; Stamm et al., 1998). However, in these studies ground vegetation had been removed and the uppermost soil was homogenized. Our experiment shows that a large proportion of the topsoil was bypassed. One reason for this was probably the dense ground vegetation and the heterogeneity of the organic layer, which led to a redistribution of homogeneously applied rainfall. According to our observations, rainwater accumulated on foliage and dripped irregularly in space and time onto the soil surface. Additionally, water might have preferentially infiltrated along stems and roots (Dekker and Ritsema, 1996).

Response of Flow Paths and Soil Matrix to Tracer Applications
Based on the response of the microcups to the Brilliant Blue addition, the microcups were divided into two groups, flow paths and soil matrix.

Microcups from flow paths showed a faster SO^2-₄ breakthrough in the first and a faster Cl^- breakthrough in the second N-addition experiment than those from the soil matrix (Fig. 5 and 7) , which supports the grouping based on the Brilliant Blue response.

View larger version (21K): [in this window] [in a new window]   Fig. 5 Chloride concentrations in the soil solution of flow paths and soil matrix at the 3- and 5-cm depths. Due to the small sample volumes, the samples were bulked during the experiment. Standard errors represent the variability of collected solution at Day 20 (3-cm depth in soil matrix: n = 6; 3-cm depth in flow path: n = 4; 5-cm depth in soil matrix: n = 8). The 5 mM Cl- m-2 was applied at the end of the simulated rainfall of 18 mm

View larger version (23K): [in this window] [in a new window]   Fig. 7 Breakthrough curves of NH+4 and SO2-4 at various depths. The 5 mM NH+4 m-2 and 2.5 mM SO2-4 m-2 were applied after 8 mm of rain, followed by an additional 6 mm of rain. This resulted in a concentration of 360 µM NH+4 and 180 µM SO2-4 in the rainwater. The detection limit of NH+4 was 10 µM

Relation of Nitrate Concentrations to Flow Paths
The existence of flow paths affected the NO^-₃ concentrations in the topsoil (Fig. 4) . Nitrate concentrations of the microcups at the 3-, 4-, and 5-cm depths were positively correlated with the Brilliant Blue recovery. In contrast, NO^-₃ concentrations of the uppermost 2 cm were not related to the location of the microcups (expressed as Brilliant Blue recovery). This shows that the location of the microcups influenced the NO^-₃ concentration only under reducing conditions. Higher NO^-₃ concentrations along flow paths under reducing conditions were probably due either to a shorter contact time of NO^-₃ with the soil in the flow paths (i.e., too short for denitrification) and/or more favorable conditions for net nitrification. Our results suggest that both mechanisms hold true. A shorter residence time of the solution sampled along flow paths was shown by the shorter breakthrough time of Cl^-, SO^2-₄, and Brilliant Blue. The more favorable conditions for nitrification along flow paths compared with the soil matrix were indicated by the shorter tailing of Cl^- breakthrough curve in flow paths. This suggests that flow paths are better drained and warrant a better O₂ supply (Dziejowski et al., 1997).

View larger version (19K): [in this window] [in a new window]   Fig. 4 Relation of NO-3 concentrations to locations of microcups characterized based on Brilliant Blue FCF recovery (flow path = high recovery). Nitrate concentrations of each microcup are the mean of three sampling dates. Brilliant Blue recovery is the amount sampled by each microcup within 24 h after application

The NO^-₃ concentrations of the soil solution were related to the Cl^- concentrations to account for concentration effects. Assuming that Cl^- is inert in the soil system, an increase of the NO^-₃/Cl^- ratio suggests a net mobilization or production of NO^-₃, whereas a decrease of the NO^-₃/Cl^- ratio suggests a NO^-₃ consumption.

After the simulated rainfall of the NO^-₃ addition, the NO^-₃/Cl^- ratio decreased at all depths below 2 cm and remained constant in the upper 2 cm for the following 3 d (Fig. 6) . Since no water was applied within this period, this suggests that denitrification occurred below the 2-cm depth, but not within the uppermost 2 cm. This is supported by the corresponding redox potential measurements (Fig. 3). The measured denitrification showed a temporal pattern analogous to the NO^-₃/Cl^- ratio. It peaked immediately after the NO^-₃ application and then decreased drastically within the next 24 h (Fig. 8) .

View larger version (23K): [in this window] [in a new window]   Fig. 6 The ratio in the soil solution during the N-addition experiment. Due to the small sample volumes, the samples were bulked. Standard errors were estimated at the first and the last sampling date

View larger version (22K): [in this window] [in a new window]   Fig. 8 Denitrification rate and NO-3 concentrations at the 5- and 20-cm depths of the soil solution sampled by suction plates. Denitrifying activity was measured with three pairs of cores during the NH+4 addition and with four pairs of cores during the NO-3 addition. NO-3 concentrations are the mean of two suction plates, installed within 50 cm of the microcups

Retention of Ammonium and Consequences for Nitrification–Denitrification
In contrast to NO^-₃, NH⁺₄ concentrations showed no differences between flow paths and soil matrix. The added NH⁺₄ pulse corresponding with a 2-wk ambient input did not show up in the soil solution at all, in contrast to the simultaneously applied SO^2-₄ (Fig. 7). None of the samples from microcups nor suction plates indicated increased NH⁺₄ concentrations on any sampling date. At most locations, they remained below the detection limit of 10 µM NH⁺₄. Ammonium was immediately immobilized at its first soil contact. The extremely efficient retention of NH⁺₄ suggests that abiotic rather than biotic processes are responsible. Ammonium can be adsorbed to cation exchange sites or incorporated into organic matter through chemical reactions (Schimel and Firestone, 1989). Both processes are promoted by the high C (140 g kg^-1, 14%) and clay content (50%) of the topsoil, combined with a high pH. Strong retention of NH⁺₄ may also limit the direct nitrification of elevated atmospheric NH⁺₄ inputs, as suggested by the low increase of NO^-₃ in the soil solution after the NH⁺₄ application (Fig. 6 and 8).

This is consistent with the denitrification pattern; denitrification did not increase after the NH⁺₄ application (Fig. 8). However, the simulated rainfall after 10 d of drying caused increased denitrification, even before NO^-₃ was added.

Combining the denitrification pattern with the high NO^-₃/Cl^- ratio in flow paths during dry periods and its rapid drop after rainfalls suggests the following process. Nitrate is mainly produced in aerobic flow paths during rainless periods. The subsequent rainfall displaces this NO^-₃ into anoxic parts. The NO^-₃ enriched flow paths become anaerobic due to waterlogging. Both processes combined with an additional NO^-₃ input through infiltrating rainwater along flow paths stimulate denitrification. However, NO^-₃ was rapidly depleted and denitrification activity decreased drastically within 24h (Fig. 6 and 8). In this gleyic topsoil, nitrification and denitrification appear to occur temporally separated and to be not as closely coupled as reported in other studies (Abbasi and Adams, 1998; Nielsen and Revsbech, 1998). However, the quoted studies were carried out with NH⁺₄-rich aerobic soils in which local enrichments of organic materials cause an increased respiration, producing anaerobic microsites (Christensen et al., 1991; Flessa and Beese, 1995; Nielsen and Revsbech, 1998). Our results suggest that in the water-saturated topsoil, NH⁺₄ is very immobile and net nitrification can only start after a period of drying. The produced NO^-₃ is not denitrified until the next rain, which leads to NO^-₃ transport into anaerobic spots and to the formation of anaerobic conditions in formerly aerobic areas.

Significance of Microscale Variability for Processes at the Catchment Scale
The observed bypassing of a large portion of the soil volume by infiltrating rainwater and the increased NO^-₃ concentrations along flow paths may help to explain effects of N deposition on the catchment scale.

Three years of continuous addition of 30 kg of NH₄NO₃–N ha^-1 yr^-1 increased the NO^-₃ exported from the subcatchment (+4 kg N ha^-1 yr^-1), the NO^-₃ leaching at the 5-cm depth (P < 0.03), and N loss via denitrification (+1.6 kg N ha^-1 yr^-1, P < 0.01), but not the K₂SO₄-extractable NO^-₃ of the bulk soil (-0.3 kg N ha^-1 yr^-1, NS) (Fig. 9) . This appears to be a contradiction. Measurements of bulk soil samples integrate soil across large volumes, while, as indicated by the microcups, a large proportion of the soil is not in direct contact with the infiltrating rainwater. The bypassing probably affects both NO^-₃ leaching and denitrification. Only a small proportion of the soil solution collected by microcups appears to be relevant for NO^-₃ leaching and NO^-₃ export from the subcatchment, namely the soil solution from flow paths. This was suggested by the comparison of NO^-₃ and Cl^- concentrations of different soil waters (Fig. 10 , Table 2) . The concentrations of the runoff from the subcatchment and of the soil water collected by suction plates were more closely related to samples from flow paths than to those from the soil matrix.

View larger version (34K): [in this window] [in a new window]   Fig. 9 The K2SO4-extractable NO-3 concentrations of the bulk soil (0–5 cm) and fluxes within the soil compartment. Mean and mean spatial standard errors of five plots from May 1996 to November 1997. Sampling dates: n = 15 (K2SO4-extractable); n = 28 (denitrification); n = 30 (leached NO-3); NO-3 export from subcatchments (1500 m2), 1997

View larger version (18K): [in this window] [in a new window]   Fig. 10 Comparison of solute concentrations from flow paths and soil matrix locations to solutes at the 3-, 4-, and 5-cm depths sampled by suction plates and to solutes of the runoff from the subcatchment. Suction plates were installed within 50 cm of the microcups. Data points represent mean concentrations of 3 wk

	Conclusions

TOP ABSTRACT INTRODUCTION Materials and methods Results and discussion Conclusions REFERENCES

 
Both the soil solution sampled by microcups and the redox potential measurements reveal that only the uppermost 2 cm of the forest Humaquept were aerobic. Below the 2-cm soil depth, aerobic and anaerobic microsites coexisted.

Only 23 of 50 microcups responded to the application of the Brilliant Blue FCF dye tracer, which indicates that a large portion of the soil volume is not in direct contact with the infiltrating rainwater. The microcups that responded to the dye were regarded to be located within reach of flow paths. These microsites were more active with respect to the transformation of N than the nonresponding cups in the soil matrix. The active sites along flow paths had higher NO^-₃ concentrations at depths below 2 cm and showed increased denitrification activity after rainfalls and an earlier and enhanced net nitrification after drying. Locations within reach of flow paths are probably microhabitats that are characterized by a better nutrient as well as O₂ supply, and are more exposed to drying and wetting than the soil matrix. However, additional research is necessary to transfer our findings regarding the importance of flow paths for N transformation to other soils and conditions.

The observed heterogeneous distribution of N transformation at a microscale may explain the effects of increased N deposition at the catchment scale; a large proportion of the soil was bypassed by the infiltrating rainwater. This probably reduces the capacity of the soil to retain N deposition. As a consequence, elevated N inputs increase the N fluxes within the ecosystem (denitrification, NO^-₃ leaching, and NO^-₃ export from the subcatchment) without increasing the K₂SO₄-extractable NO^-₃ of the bulk soil.

	ACKNOWLEDGMENTS

 
A. Heim is thanked for the introduction and help with CE measurements. We are grateful to M. Bundt for discussions and A. Clark for the linguistic revision of the manuscript. Furthermore, we thank the central laboratory of the WSL for conducting the IC analyses and S. Kretschmar and M. Bundt for constructing redox electrodes. This study, as part of the European project NITREX, was partly financed by the Swiss Federal Office of Science and Education, Bern.

Received for publication December 1, 1998.

	REFERENCES

TOP ABSTRACT INTRODUCTION Materials and methods Results and discussion Conclusions REFERENCES

 

• Abbasi M.K., Adams W.A. Loss of nitrogen in compacted grassland by simultaneous nitrification and denitrification. Plant Soil 1998;200:265-277.
• Aber J.D., Nadelhoffer K.J., Steudler P., Melillo J.M. Nitrogen saturation in northern forest ecosystems. BioScience 1989;39:378-386.[ISI]
• Bakker D.M., Bronswijk J.J.B. Heterogeneous oxygen concentration in a structured clay soil. Soil Sci. 1993;155:309-315.
• Cambardella C.A., Elliott E.T. Carbon and nitrogen distribution in aggregates from cultivated and native grassland soils. Soil Sci. Soc. Am. J. 1993;57:1071-1076.[Abstract/Free Full Text]
• Chotte J.L., Ladd J.N., Amato M. Sites of microbial assimilation, and turnover of soluble and particulate ¹⁴C-labelled substrates decomposing in a clay soil. Soil Biol. Biochem. 1998;30:205-218.
• Christensen S., Simkens S., Tiedje J.M. Spatial variation in denitrification: Dependence of activity centers in the soil environment. Soil Sci. Soc. Am. J. 1991;54:1608-1613.
• Cogger C.G., Kennedy P.E., Carlson D. Seasonally saturated soils in the Puget Lowland II. Measuring and interpreting redox potentials. Soil Sci. 1992;154:50-58.
• Dekker L.W., Ritsema C.J. Uneven moisture patterns in water repellent soils. Geoderma 1996;70:87-99.
• Dziejowski J.E., Rimmer A., Steenhuis T.S. Preferential movement of oxygen in soils. Soil Sci. Soc. Am. J. 1997;61:1607-1610.[Abstract/Free Full Text]
• Flessa H., Beese F. Effects of sugarbeet residues on soil redox potential and nitrous oxide emission. Soil Sci. Soc. Am. J. 1995;59:1044-1051.[Abstract/Free Full Text]
• Flühler H., Stolzy L.H., Ardakani M.S. A statistical approach to define soil aeration in respect to denitrification. Soil Sci. 1976;122:115-123.
• Flury M., Flühler H., Jury W.A., Leuenberger J. Susceptibility of soils to preferential flow of water: A field study. Water Res. Res. 1994;30:1945-1954.
• Göttlein A., Blasek R. Analysis of small volumes of soil solution by capillary electrophoresis. Soil Sci. 1996;161:705-716.
• Göttlein A., Hell U., Blasek R. A system for microscale tensiometry and lysimetry. Geoderma 1996;69:147-156.
• Göttlein A., Matzner E. Microscale heterogeneity of acidity related stress-parameters in the soil solution of a forested cambic podzol. Plant Soil 1997;192:95-105.
• Gundersen P., Emmet P.A., Kjønnas O.J., Koopmans C.J., Tietema A. Impact of nitrogen deposition on nitrogen cycling in forests: A synthesis of NITREX data. For. Ecol. Manage. 1998;101:37-55.
• Heijs A.W.J., Ritsema C.J., Dekker L.W. Three-dimensional visualization of preferential flow patterns in two soils. Geoderma 1996;70:101-116.
• Juma N.G. Interrelationship between soil structure/texture, soil biota/soil organic matter and crop production. Geoderma 1993;57:3-30.
• Morgan P., Cooper N.S., Battersby N.S., Lee S.A., Lewis S.T., Machin T.M., Grahm S.C., Watkinson R.J. Automated image analysis method to determine fungal biomass in soils and on solid matrices. Soil Biol. Biochem. 1991;23:609-616.
• Nielsen T.H., Revsbech N.P. Nitrification, denitrification, and N-liberation associated with two types of organic hot-spots in soil. Soil Biol. Biochem. 1998;30:611-619.
• Nishio T. Estimating nitrogen transformation rates in a surface aerobic soil of a paddy field. Soil Biol. Biochem. 1994;26:1273-1280.
• Norman R.J., Stucki J.W. The determination of nitrate and nitrite in soil extracts by ultraviolet spectrophotometry. Soil Sci. Soc. Am. J. 1981;45:347-353.[Abstract/Free Full Text]
• Parkin T.B. Soil microsite as a source of denitrification variability. Soil Sci. Soc. Am. J. 1987;51:1194-1199.[Abstract/Free Full Text]
• Pfisterer U., Gribbohm S. Zur Herstellung von Platinelektroden für Redoxmessungen. (In German, with English abstract.) Z. Pflanzenernähr. Bodenk. 1989;152:455-456.
• Ryden J.C., Skinner J.H., Nixon D.J. Soil core incubation system for the field measurement of denitrification using acetylene-inhibition. Soil Biol. Biochem. 1987;19:753-757.
• Schimel J.P., Firestone M.K. Inorganic N incorporation by coniferous forest floor material. Soil Biol. Biochem. 1989;21:41-46.
• Schleppi P., Muller N., Feyen H., Papritz A., Bucher J.B., Flühler H. Nitrogen budgets of two small experimental forested catchments at Alptal, Switzerland. For. Ecol. Manage. 1998;101:177-185.
• Stagnitti F., Parlange J.-Y., Steenhuis T.S., J.Boll B., Pivetz, Barry D.A. Transport of moisture and solutes in the unsaturated zone by preferential flow. In: Singh V.P., ed. Environmental hydrology. Dordrecht, The Netherlands: Kluwer Academic Publishers, 1995:193-224.
• Stamatiadis S., Doran J.W., Ingham E.R. Use of staining and inhibitors to separate fungal and bacterial activity in soil. Soil Biol. Biochem. 1990;22:81-88.
• Stamm C., Flühler H., Gächter R., Leuenberger J., Wunderli H. Preferential transport of phosphorus in drained grassland soils. J. Environ. Qual. 1998;27:515-522.[Abstract/Free Full Text]
• Tiedje J.M., Sexstone A.J., Parkin T.B., Revsbech N.P. Anaerobic processes in soil. Plant Soil 1984;76:197-212.
• Wright R.F., Rasmussen L. Introduction to the NITREX and EXMAN projects. For. Ecol. Manage. 1998;101:1-7.

This article has been cited by other articles:

				A. R. Barzegar, S. J. Herbert, A. M. Hashemi, and C. S. Hu Passive Pan Sampler for Vadose Zone Leachate Collection Soil Sci. Soc. Am. J., May 1, 2004; 68(3): 744 - 749. [Abstract] [Full Text] [PDF]

				R. T. Venterea, G. M. Lovett, P. M. Groffman, and P. A. Schwarz Landscape Patterns of Net Nitrification in a Northern Hardwood-Conifer Forest Soil Sci. Soc. Am. J., March 1, 2003; 67(2): 527 - 539. [Abstract] [Full Text] [PDF]

				M. Bundt, M. Jaggi, P. Blaser, R. Siegwolf, and F. Hagedorn Carbon and Nitrogen Dynamics in Preferential Flow Paths and Matrix of a Forest Soil Soil Sci. Soc. Am. J., September 1, 2001; 65(5): 1529 - 1538. [Abstract] [Full Text] [PDF]

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in ISI Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (13) Citing Articles via Google Scholar Google Scholar Articles by Hagedorn, F. Articles by Flühler, H. Search for Related Content PubMed Articles by Hagedorn, F. Articles by Flühler, H. GeoRef GeoRef Citation Agricola Articles by Hagedorn, F. Articles by Flühler, H.