JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (49)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Clément, J.-C.
Right arrow Articles by Marmonier, P.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Clément, J.-C.
Right arrow Articles by Marmonier, P.
Agricola
Right arrow Articles by Clément, J.-C.
Right arrow Articles by Marmonier, P.
Related Collections
Right arrow Ecosystem Management
Right arrow Soil Microbiology
Right arrow Nutrient Cycling
Right arrow Water Pollution
Right arrow Wetland Soils
Journal of Environmental Quality 31:1025-1037 (2002)
© 2002 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America

TECHNICAL REPORTS
Wetlands and Aquatic Processes

Seasonal Dynamics of Denitrification along Topohydrosequences in Three Different Riparian Wetlands

Jean-Christophe Clément*, Gilles Pinay and Pierre Marmonier

U.M.R. 6553-ECOBIO, Université de Rennes I, Avenue du Général Leclerc, F-35042 Rennes cedex, France

* Corresponding author (clement{at}aesop.rutgers.edu)

Received for publication April 23, 2001.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSION
 REFERENCES
 
We investigated the seasonal patterns of denitrification rates and potentials in soil profiles along the topohydrosequence formed at the upland–wetland interface in three riparian wetlands with different vegetation cover (i.e., forest, understory vegetation, and grass). Denitrification was measured using the acetylene inhibition method on soil cores and slurries, which provided a means of comparing the relative activity of this process in different locations. We evaluated, on a seasonal basis, the respective importance of the vegetative cover and the hydromorphic gradient as factors limiting denitrification. Regardless of the season, vegetation type, or lateral position along each topohydrosequence in the riparian wetlands, strong significant gradients of both in situ and potential denitrification rates were measured within a soil profile. Results confirm that the upper organic soil horizon is the most active, when in contact with the ground water. In deeper soil horizons, denitrification activity was low (from 0.004 to 0.5 mg N kg-1 dry soil d-1), but contributed significantly to the reduction of ground water NO-3 load along the riparian ground water flowpath (from 9.32 to 0.98 mg NO3–N L-1). Along the soil topohydrosequence, the denitrifying community of the upper soil horizons did not vary significantly on a seasonal basis despite the large seasonal ground water fluctuations. Along each topohydrosequence, the denitrification-limiting factor gradually shifted from anaerobiosis to NO-3 supply. In situ denitrification rates in the forested, understory vegetation and grass sites were not significantly different. This result emphasizes the importance of the topography of the valley rather than the vegetation cover in controlling denitrification activity in riparian wetlands.

Abbreviations: ANA, denitrification activity of soil samples incubated with acetylene under N2 atmosphere alone • ANAC, denitrification activity of soil samples amended with carbon (4 g glucose kg-1 fresh soil) and incubated with acetylene under N2 atmosphere • ANACN, denitrification activity of soil samples amended with both nitrate and carbon (10 mg NO3–N kg-1 fresh soil + 4 g glucose kg-1 fresh soil) and incubated with acetylene under N2 atmosphere • ANAN, denitrification activity of soil samples amended with nitrate (10 mg NO3–N kg-1 fresh soil) and incubated with acetylene under N2 atmosphere • DNT, in situ denitrification activity incubated with acetylene alone • NITR, net nitrification • NMIN, net mineralization


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSION
 REFERENCES
 
ONCE CONSIDERED OF LITTLE VALUE, riparian wetlands are now recognized for their roles in maintaining water quality and sustaining biodiversity far in excess of their aerial extent in the landscape (Haycock et al., 1997; Naiman and Decamps, 1997). Indeed, since Peterjohn and Correll (1984), numerous studies have shown that ground water NO-3 concentrations may decrease substantially as ground water flows through riparian zones en route to the stream, and hence, improve water quality (Jordan et al., 1993; Pinay et al., 1993; Hanson et al., 1994b; McClain et al., 1994; Nelson et al., 1995; Hill, 1996; Hedin et al., 1998). Microbial transformations (i.e., denitrification or immobilization) and plant uptake are considered to be responsible for these buffering capabilities. Plant uptake and microbiological immobilization represent a temporary storage system since N will eventually return to the system via death and decomposition. In contrast, denitrification constitutes a real sink for NO-3 since it is transformed to gaseous products (N2, NO, N2O) that are emitted to the atmosphere (Knowles, 1982).

Environmental factors, which affect microbiological denitrification, have been identified (Tiedje, 1988; Gold et al., 1998; Luo et al., 1999) and investigated in several field studies (Schipper et al., 1993; Pinay et al., 1995; Schnabel et al., 1996; Schnabel, 1997). Denitrification is controlled by soil moisture and temperature (Grundmann and Rolston, 1987; Groffman et al., 1996a), nutrient supply as nitrate and soluble organic matter (Starr and Gillham, 1993; Paul and Clark, 1996), and soil texture (Davidsson and Sthal, 2000; Pinay et al., 2000). Spatial and temporal variations of these parameters and their interactions are often linked to the high variability of soil denitrification rates (Groffman and Tiedje, 1988; Hanson et al., 1994a; Montgomery et al., 1996; Hill et al., 2000). At the soil scale, this denitrification variability is under the control of pore size distribution and C availability (Sextone et al., 1985; Parkin, 1987). These two variables are controlled by land use (Balesdent, 1996). Yet comparative studies on denitrification rates under pasture and cropped soils have yielded contradictory results: some have found that denitrification rates were higher under pasture (Weier et al., 1993; Lensi et al., 1995), while others found that the cropped soils had higher rates (Bijay-Singh et al., 1989). These differences could be explained by the fact that these studies were based on potential rates and did not take into account the importance of the pore size distribution and soil clod distribution on in situ denitrification rates (Parry et al., 1999). At a larger scale, riparian ecosystems are often described as complex environments that are spatially heterogeneous, where denitrification is controlled by the interaction between flood hydraulics, floodplain topography, soil conditions, nitrogen concentration, and diffusion pattern (Groffman et al., 1996b; Stewart et al., 1998; Hill et al., 2000). Several recent studies have demonstrated the importance of the upland–wetland interface in controlling diffuse nitrate fluxes from upslope (see Hill, 1996 and Haycock et al., 1997 for reviews). Most of these studies have been focused on forested areas; the few that have reported on the role of grassed riparian sites are contradictory (Daniels and Gilliam, 1996). Some studies have demonstrated a low efficiency in NO-3 removal compared with the forest tree riparian zones (Osborne and Kovacic, 1993; Hubbard and Lowrance, 1997; Wenger, 1999; Lyons et al., 2000), while others observed substantial removal capacity of riparian grass (Haycock et al., 1993; Schnabel et al., 1996; Addy et al., 1999). At the riparian-zone scale, these apparently divergent results might have arisen from comparisons of sites with different hydrogeomorphic conditions.

The objective of this study was to evaluate the importance of different vegetation covers (i.e., forest trees, understory vegetation, and grass) on denitrification activity in adjacent riparian zones situated in the same hydrogeomorphic stream valley system. Although the three study sites were located along the same river, they were not hydrologically connected from one to another. Furthermore, nitrate fluxes coming from uplands and entering the wetland were in the same order of magnitude for the three study sites. The difference of vegetation type is explained by the fact that the abandonment of the sites has occurred at different periods. This typical gradual decrease of land use pressure in the riparian zones is due to the change of agricultural practices in the late 1960s when mechanization and intensification occurred. We investigated the seasonal pattern of in situ and potential denitrification rates by sampling soil profiles along the topohydrosequences formed at the upland–wetland interface.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSION
 REFERENCES
 
Study Area
The study was conducted in a riparian wetland located 10 km southwest of the Mont Saint-Michel Bay (Brittany, France: 48.3° N, 1.3° W). The wetland borders a fouth-order stream (the Hermitage River) that has its source 5 km upstream. Several other upstream perennial sources participate to the global flow. Mean annual discharge of the Hermitage River is approximately 90 L s-1. The region has a mild oceanic climate characterized by mild, humid weather all year round. Annual rainfall ranged between 850 and 900 mm during the study period. The annual mean temperature was 12.5°C, July and August being the warmest months with average temperatures ranging from 15 to 20°C. The lowest average temperatures (from -1.9 to 14°C) were measured in November and December. The landscape is predominantly agricultural, with maize (Zea mays L.) and wheat (Triticum sp.) as dominant crops. The upland is an arable field with a wheat–maize rotation. The agricultural field has a bank (3 m high) at its downslope end where it borders the wetland. The study site is about 80 m wide but our investigations are focused on the lower 15 m on the hillslope (Fig. 1a) .



View larger version (40K):
[in this window]
[in a new window]
  		Fig. 1. (a) Location of the study site and map of the studied wetland. (b) Average nitrate concentration (with standard deviation, n = 138) in ground water below the soil topohydrosequence. m.a.s.l., meters above sea level.

 
Within this riparian wetland, three sites were selected with different vegetation types: (i) a forested site with typical wetland trees (i.e., oak [Quercus sp.] and white willow [Salix alba L.]); (ii) an understory vegetation site characterized by abundant understory plants (white willow, blackberry bush [Rubus fructicosus L.], water persicaria [Polygonum amphibium L.], reed canary grass[Phalaris arundinacea L.], and bindweed [Calystegia sepium (L.) R. Br.]) without mature trees; and (iii) a grass site characterized by herbaceous species (soft meadow grass [Holcus lanatus L.], orchard grass [Dactylis glomerata L.], creeping bent grass [Agrostis stolonifera L.], soft rush [Juncus effusus L.], floating manna grass [Glyceria fluitans (L.) R. Br.], and marsh foxtail [Alopecurus geniculatus L.]). The three selected riparian sites were approximately 20 m in length (along the bank) and 15 m wide into the riparian area. Their differences in vegetation cover resulted from differential timing of pasture abandonment. Pasture abandonment in the forested site occurred earlier than in the understory vegetation site, while the wet grass site was still grazed during the summer period. The tree canopy in the forested site was approximately 20 yr old. Nine trees were present in the 100-m2 study area, with diameters ranging from 9 to 87 cm. Stem and leaf production were respectively 0.13 and 0.31 kg m-2 yr-1. The willows presented several main stems shooting from the same root system. Given the mean height of the tree canopy (approximately 13 m high) and its percentage of coverage (about 100%), the understory vegetation growth in this site was considerably limited by the tree shade. Thus, the density of the undergrowth vegetation in the forest tree site was reduced and its productivity (0.10 kg m-2 yr-1) was significantly smaller than the plant productivity measured in the grass site (0.60 kg m-2 yr-1) and in the understory vegetation site (1.94 kg m-2 yr-1). The understory vegetation site was mainly characterized by high standing plants, and was not coppiced. Each study site has been divided into three zones (Fig. 1b) corresponding to the hydromorphic gradient from hillslope to wetland. Each zone was characterized by different dominant plant species in the undergrowth vegetation. Zone III at the hillslope boundary was 3 m wide and mainly characterized by blackberry bush and stinging nettle (Urtica dioica L.) in both the forest trees and the understory vegetation sites, while soft meadow grass and orchard grass dominated in the grass site. Zone II was 6 m wide and dominated by meso-hygrophilous vegetation such as ivy (Hedera helix L.) and stinging nettle in the forested site, soft meadow grass and water persicaria in the understory vegetation site, and creeping bent grass and soft rush in the grass site. The most hydromorphic zone (Zone I) was 6 m wide and characterized by hygrophilous vegetation such as reed canary grass and bindweed in both the forest trees and the understory vegetation site, while floating manna grass and marsh foxtail dominated in the grass site.

The geological substratum of the study area is made of cornéennes and mica-schists (Brioverien Schist). The soils were fine silty-clay loam, mixed, mesic, Typic Haplaquoll. The three study sites presented a similar soil profile. Below a litter layer made of decomposed plant material, an organic O horizon from 0 to -5 cm rapidly evolved into a brown-grey transitional mineral hydromorphic A horizon with ferrous oxidation spots along the root system (from -5 to -25 cm, 9.6% OM, pH 4.8, 75% silt plus clay). Then, the soil profile (from -25 to -50 cm) evolved into a grey pseudogley with ferrous oxidation spots (3.9% OM, pH 5.2, 80% silt plus clay). Some roots could be found up to -50 cm deep. Beyond this depth, the gley horizon was temporary reoxidized until -200 cm deep (2.7% OM, pH 5.5, 76% silt plus clay). Then, a reduced blue-grey horizon was present and reached the geological substratum made of schist. Some rare coarse sand veins (2–10 cm in diameter) could be found between -100 and -200 cm depth, and illustrated the presence of residual marine sediments. The texture was loamy-clay in the first 100 cm, then loamy-sand in the deeper layers. The structure became more and more massive with depth, reducing the permeability, although the rare sandy veins could facilitate it. Bulk densities were measured in the first 25 cm and ranged between 0.45 and 0.70 g cm-3.

Sampling Strategy and Analytical Procedures
Within each site, three transects of four wells each were installed from the upland–riparian interface toward the stream. Wells transects were spaced about 10 m apart with an average distance of 3 m between wells. Wells were constructed from 2-cm-diameter PVC pipe, inserted to depths approximately 2 m below the ground surface, with the lowest 50 cm of each well perforated to allow water to be collected and water table elevation to be measured. Three topohydrosequences were selected along ground water flow paths identified by a bromide tracing experiment. Ground water was sampled and water table elevation measured on a biweekly basis. Water samples were analyzed colorimetrically for NO-2 and NO-3 using a Technicon (Tarrytown, NY) autoanalyzer including Cd reduction to NO-2 (Wood et al., 1967), and for NH+4 using the phenyl-hypochlorite method (Solorzano, 1969). Soils were sampled in the three sites at each season, between autumn 1998 and summer 1999 using a manual soil corer. Within each topohydrosequence, three soil replicates were sampled in each of the three zones and at three depths (0–25, 25–50, 50–75 cm) to determine lateral and vertical stratification of denitrification using the static core acetylene inhibition method (Yoshinari and Knowles, 1976). Since the penetration of acetylene within the soil cores can be disturbed by soil characteristics, we checked the efficacy of the inhibition by comparing N2O emissions of soil samples with and without acetylene addition (Table 1). The N2O emissions were between 31 and 58 times higher in the presence of acetylene than in its absence, and demonstrated that acetylene was actually penetrating the cores and acting as an effective inhibitor of N2O reduction. Acetylene also inhibits nitrification and therefore, if NO-3 is limiting, acetylene additions may decrease denitrification rates by preventing regeneration of consumed nitrate (Walter et al., 1979). To limit any problems in soils where nitrification and denitrification are "coupled," we decided to incubate soils in the presence of acetylene only for short time periods (up to 2 h). The upper 25 cm of soil were taken after the litter was discarded. All samples were stored at 4°C before analysis. In situ denitrification (DNT) was assayed immediately after sampling. Three intact cores sampled in each of the three depths in the three zones of the three sites were incubated in glass flasks five times larger than the soil core. The flasks were capped with rubber serum stoppers and then amended with acetone-free acetylene to bring the atmospheric concentration to 10 kPa (10% v/v) acetylene and 90 kPa air. Samples were incubated at field temperature and DNT rates were calculated as the rate of nitrous oxide (N2O) accumulation in the headspace after 2 h of incubation. Headspace samples were removed from all cores and stored in evacuated Venoject collection tubes (Terumo Scientific, Somerset, NJ). Gas samples were collected by syringe through a septum and analyzed by gas chromatography for N2O (Chrompack [Houten, the Netherlands] 9001-CP).


View this table:
[in this window]
[in a new window]
  		Table 1. Annual mean of N2O emissions from soil samples in each study site incubated with and without acetylene.

 
Denitrification enzyme activities (DEA) were measured from three replicates of each sample transformed into soil slurries using the procedure of Smith and Tiedje (Smith and Tiedje, 1979). Four sets of subsamples of each soil sample were made anoxic by flushing with N2 and incubated for 8 h with acetone-free acetylene to bring the soil atmosphere concentration to 10 kPa (10% v/v) acetylene and 90 kPa air at mean soil temperature. The first set was incubated with acetylene under an N2 atmosphere alone and referred to as ANA. One set of subsamples was amended with nitrate (10 mg NO3–N kg-1 fresh soil) and referred to as ANAN; a second set was amended with C (4 g glucose kg-1 fresh soil) referred to as ANAC. A last set was amended with both NO-3 and C (10 mg NO3–N kg-1 fresh soil + 4 g glucose kg-1 fresh soil) and referred to as ANACN. Denitrification rates are calculated as the rate of nitrous oxide (N2O) accumulation in the headspace between 4 and 8 h of incubation.

Soils samples were extracted with 100 mL of K2SO4 (35 g L-1). The extracts were filtered and analyzed for NH4–N, NO3–N, and dissolved organic N using a Technicon autoanalyzer (Technicon, 1977). Nitrate was analyzed colorimetrically by the Griess–Ilosvay method (Keeney and Nelson, 1982) after reduction by percolation on copperized cadmium column. The NH+4 was measured following the Indophenol Blue Method (Keeney and Nelson, 1982). Dissolved organic N was measured on the K2SO4 extract by oxidation to NO-3 with potassium persulfate at 120°C, and analyzed by the above-mentioned procedure. Soil water content was measured by drying fresh soil at 70°C for 48 h until a constant weight was reached. Soil total organic N and total organic C were measured using a CHN autoanalyzer (PerkinElmer [Wellesley, MA] Series II CHNS/O Analyzer 2400). Net nitrification (NITR) and net mineralization (NMIN) were seasonally evaluated on the 0- to 25-cm soil layer. Net nitrogen mineralization was calculated from measured changes in the mineral N content of largely undisturbed soil isolated inside polyethylene bags allowing air to pass through but preventing leaching (Pastor et al., 1987). After 30 d of incubation in the field, nitrogen content in the incubated bags was compared with the soil nitrogen content at the beginning of the incubation. Net nitrification was estimated from measured changes in NO3–N content during exposure in the same incubated soil samples.

Soil and air temperatures and rainfall were continuously measured every 15 min in the field using one probe, one temperature sensor, and one tipping bucket gauge, respectively, connected to a Campbell (Logan, UT) data logger.

Statistical Analyses
Differences between sites, zones, depths, and seasons for water content (% H2O), ammonium content (NH4–N), nitrate content (NO3–N), dissolved organic nitrogen content (DNO), in situ denitrification rate (DNT), potential denitrification rates (ANACN), and denitrification limiting factors (ANA, ANAC, ANAN) were tested using the Kruskal–Wallis nonparametric analysis of variance (ANOVA). Differences were considered statistically significant if p < 0.05. All statistical analyses were performed on Statistica for Windows (Stat Soft, 1999).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSION
 REFERENCES
 
Ground Water Fluctuations
Ground water level in the wetland followed the same pattern as the stream with high water levels in late winter and early spring and low water levels in summer and early autumn (Fig. 2) . Similar patterns were observed at the three sites, even though a ground water gradient exists from the forested wetland to the grass, the last being the wettest, due to its downstream position within the valley. At each site there was also a significant ground water level gradient. In Zone III (i.e., at the upland–wetland interface), the ground water level had the largest amplitude of variation within the year (i.e., almost 2 m), while in Zones II and I, the water level variation progressively decreased, maintaining the water level closer to the soil surface. In each site, the water table depth was significantly higher in Zone I than in Zone II or III (p < 0.001). On an annual basis, ground water NO-3 concentrations decreased significantly from the upper part of the topohydrosequence to the lowest part in all three sites (Fig. 1b). The slight increase in ammonium (data not presented here) measured in the lower part of the transect (from 0.45 to 0.48 mg N L-1) cannot explain the change in nitrate concentration by dissimilative reduction of NO-3 to NH+4.



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 2. Rainfall, air temperature, and water table elevation measurements in the three sampling zones of the three study sites between January 1998 and March 2000. Forest site data started later since this site was equipped with piezometers in August 1998.

 
Vertical Pattern
Strong gradients with depth were observed at all the sites for the nitrogen-related parameters. Kruskal–Wallis nonparametric analysis of variance of ranks confirmed the significant differences between depths for all sites. Soil parameters measured in the three sites presented a significant decreasing or increasing pattern with depth (p <= 0.001, r2 = 0.85). This strong vertical pattern was found in soil organic N and C along the three topohydrosequences (Table 2). Moreover, soil NH4–N concentrations (Table 3) were always significantly higher in the 0- to 25-cm upper organic horizon (between 1.8 and 8.1 g N kg-1 of dry soil) than in the deeper soil layers (between 0.9 and 3.2 g N kg-1 of dry soil). Similarly, soil NO3–N concentration (Table 3) presented the same pattern with higher values in the 0- to 25-cm layer (between 3.1 and 8.9 g N kg-1 of dry soil) than in the deeper layers (between 1.1 and 4.8 g N kg-1 of dry soil). This vertical pattern was also observed for in situ denitrification (DNT) and potential denitrification (ANACN) (Fig. 3 and Fig. 4) . Yet, despite the steep gradient of DNT rates measured along the soil profiles, some significant activities (from 0.004 to 0.5 mg N kg-1 dry soil d-1) could be detected down to 75 cm (Fig. 3).


View this table:
[in this window]
[in a new window]
  		Table 2. Soil carbon and nitrogen contents by zone and depth in the three study sites.

 

View this table:
[in this window]
[in a new window]
  		Table 3. Annual averages and standard errors of the mean of soil nutrient contents and nitrogen mineralization in the soil profile along the riparian catena. NITR, net nitrification; NMIN, net mineralization.

 


View larger version (46K):
[in this window]
[in a new window]
  		Fig. 3. Seasonal changes of in situ denitrification (DNT) (mean and standard errors, for n = 3) in soil profiles along the three topohydrosequences.

 


View larger version (44K):
[in this window]
[in a new window]
  		Fig. 4. Seasonal changes of potential denitrification (ANACN) (mean and standard errors, for n = 3) in soil profiles along the three topohydrosequences.

 
Importance of the Vegetation Cover versus Lateral Gradient
Soil characteristics and microbial processes involved in the nitrogen cycle of the upper soil layer of the three different sites were compared zone by zone for the four different seasons (Table 4). For most of the soil parameters, differences between sites were not significant, whatever the season or the zone considered. Even so, there are some significant differences between sites for in situ denitrification (DNT) and potential denitrification (ANACN). However, no consistent pattern could be found, either from the zone or the seasonal perspective. For instance, in situ denitrification (DNT) is significantly higher in Zone III of the forested site in autumn, Zone II of the grass site and Zone I of the understory vegetation site in winter, and Zones I and II of the forested site in spring, and not significantly different between sites in summer (Fig. 3). Similarly, no clear pattern could be found for the net mineralization (NMIN) and nitrification (NITR) processes, neither by site nor by season (Tables 3 and 4).


View this table:
[in this window]
[in a new window]
  		Table 4. Intersite differences for the 0- to 25-cm layer in the three zones and the four seasons. Nonparametric analysis of variance (ANOVA; Kruskal–Wallis test) equivalent to a one-way ANOVA based on median of ranks. H is the value of the Kruskal–Wallis test.

 
However, when the data sets of the three sites were pooled by zone, there were significant differences between zones (Table 5). Net mineralization (NMIN) was significantly higher in the upslope part (Zone III) than the lower ones (Table 3). The same pattern occurred for net nitrification (NITR). A significant lack of nitrification in Zones II and I was noticeable when compared with NMIN rates. It entailed an accumulation of ammonium in these lower zones except for the grass site (Table 3). Due to the large intersite variability (Table 4), no significant differences were found in the in situ denitrification rates (DNT) between zones when sites were pooled by zone (Table 5).


View this table:
[in this window]
[in a new window]
  		Table 5. Interzone differences for the 0- to 25-cm layer at the four seasons. The data for the three sites were pooled by zone. Nonparametric analysis of variance (ANOVA; Kruskal–Wallis test) equivalent to a one-way ANOVA based on median of ranks. H is the value of the Kruskal–Wallis test.

 
Factors Limiting Denitrification
The factors limiting denitrification were investigated. We tested the effects of various treatments (i.e., anaerobiosis, anaerobiosis + nitrate, anaerobiosis + carbon as glucose, anaerobiosis + carbon + nitrate) on the denitrifying activity of the bacterial communities sampled along soil profiles of the three zones of each site. The ANACN, which provides a potential rate of denitrification activity under nonlimiting factors, can be used as an index of denitrifying population density. If absolute values are not very informative per se, the relative comparison of ANACN rates within soil profiles, zones, and sites allow us to identify areas where denitrifying populations are the most abundant. Moreover, comparison rates of ANACN with other treatments (ANA, ANAN, and ANAC) reveal the denitrification-limiting factors.

The ANACN rates were significantly higher in the upper soil horizons than in the lower ones, whatever the site, the zone, or the season considered (Fig. 4). The ANACN rates in the lower zones (I and II) were generally higher than in Zone III, although not always significantly, than those measured in the upper zones. Moreover, Zone III in all sites generally had higher rates at depth. When comparing the annual average DNT and ANACN by site, zone, and profile, we found that ANACN was always significantly higher than DNT in the upper soils horizons, whatever the site and the zone considered (Fig. 5) . In the lower horizons, ANACN was not significantly different from DNT, except in the Zone III 25- to 50-cm horizon, where differences were still significant.



View larger version (38K):
[in this window]
[in a new window]
  		Fig. 5. Comparison of annual average (with standard errors, n = 12) of in situ (DNT) and potential (ANACN) denitrification rates in soil profiles along the three topohydrosequences. (Statistical differences between in situ and potential are given by asterisks: ***, p <= 0.001; **, p <= 0.01; *, p <= 0.05; ns, nonsignificant.)

 
Because we found the highest DNT and ANACN rates in the upper soil horizons, we then only investigated the seasonal pattern of the denitrification limiting factors in these upper horizons (sites pooled together) by comparing rates measured with different treatments (i.e., ANA, ANAC, and ANAN against ANACN) (Fig. 6) . During the spring high water season, ANACN is significantly lower than during the other seasons and denitrification rates are limited by anaerobiosis, except in the flooded lower part of the topohydrosequence that is limited by the nitrate availability (Fig. 6). In summer, when the water table is low in the upper parts of the topohydrosequence, ANACN rates are still low and limited by anaerobiosis, while in the lower zone, ANACN rates are significantly higher but cannot be triggered without the combination of anaerobiosis, nitrate, and organic carbon. In autumn, denitrification is limited in the upper and the lower parts of the topohydrosequence by anaerobiosis, while in the middle zone it requires the combination of all the factors. In winter, the importance of nitrate availability gradually increased from Zone III to Zone I.



View larger version (38K):
[in this window]
[in a new window]
  		Fig. 6. Seasonal variations of denitrification-limiting factors in the upper horizon (0–25 cm) of the three topohydrosequences. Average values (with standard errors) for all sites (n = 9).

 
Despite the fact that the limiting factors of denitrification varied seasonally as shown above, we analyzed the annual response of denitrification-limiting factors in each of the three zones by averaging the results obtained for the different modalities at the different seasons and in the different sites (Fig. 7) . Overall, ANACN rates increased from Zone III to Zone I, while the importance of anaerobiosis gradually disappeared, being replaced by nitrate availability.



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 7. Denitrification limiting factors in the upper soil horizon (0–25 cm) of the topohydrosequence. Seasons and sites have been pooled together (means and standard errors, n = 36). ANA, denitrification activity of soil samples incubated with acetylene under N2 atmosphere alone. ANAC, denitrification activity of soil samples amended with carbon (4 g glucose kg-1 fresh soil) and incubated with acetylene under N2 atmosphere. ANAN, denitrification activity of soil samples amended with nitrate (10 mg NO3–N kg-1 fresh soil) and incubated with acetylene under N2 atmosphere.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSION
 REFERENCES
 
Tri-Dimensional Structure of the Topohydrosequence
In each site, we found a significantly higher in situ denitrification activity in the upper horizon (Fig. 3), which corresponds to the rich organic layer. In this study the term in situ denitrification refers to measurements in unamended soil cores using the acetylene blocking method (as opposed to the potential denitrification rates measured by incubation of soil slurries after different amendment such as anaerobiosis, nitrate, and organic carbon). This strong vertical DNT gradient was maintained whatever the site or the zone. Moreover, we could not find any seasonal pattern in denitrification rates along the different topohydrosequences (Table 5). Control on denitrification activity is operated by the organic carbon availability, which decreases significantly with depth (Table 2).

Even so, soil-denitrifying activity was measured at all depths of the three riparian sites, whatever the season (Fig. 3). In a parallel study we could demonstrate that the nitrate decline along these riparian flow paths was due to denitrification, using the change of the 15N ratio of the remaining ground water nitrate (Clément et al., unpublished data). However, we could not determine whether or not deep riparian soil layers were involved in that reduction. The results presented here reveal that there exists enough available organic carbon to sustain some significant denitrification up to 75 cm deep, which contributes to the reduction of the ground water nitrate load. In these deeper soil horizons, measured rates of DNT were of the same range as ANACN, except in the 25- to 50-cm depth of Zone III (i.e., at the interface between the agricultural upland and the wetland; Fig. 5). Thus, in terms of nitrate removal, the active soil volume is potentially more important at the upland–wetland interface (Zone III) than in the other zones. The combination of high subsurface nitrate input from the upland fields (Fig. 1b) and the high frequency of soil saturation in this intermediary horizon (Fig. 2) maintains a higher denitrifying bacteria community than expected from the general model in which high density should occur mainly in the organic carbon-rich horizons. It confirms that in deeper soil profiles, despite the low denitrifying bacteria density, activity can be sustained by allochtonous nitrate input. This implies that, at the upland–wetland interface scale, even though there exists a strong vertical gradient, denitrification activity should be considered as a function of the entire soil profile and/or volume rather than as a surface-related process (Hill et al., 2000). This aspect clearly has some management implications in terms of ground water quality control. Indeed, considering a volume of soil, one might ease negotiation on a width basis because for a given width there is a longer surface of contact between soil matrix and water. Moreover, the ground water table does not have to be at the soil surface level. This is important in terms of surface water regime because the ground water level is often controlled by the adjacent stream water level.

Potential Denitrification
We measured the denitrifying enzyme activity (ANACN), which provides a tool to evaluate the pool of denitrifying bacteria by forcing their expression under nonlimiting conditions for a short period to avoid their multiplication. Hence, comparison of the results obtained allows us to determine the maximum denitrification capacity of a soil and provides an indirect evaluation of its denitrifying community. The steep vertical gradient measured for in situ denitrification was also found for ANACN (Fig. 4). This indicates that the upper soil horizon hosts the highest denitrifier density, whatever the season or the site, although the upper part of the topohydrosequence was rarely entirely flooded (Fig. 2). This can be explained by the presence of denitrifying microsites in soils, which are maintained by stable moisture content (Parkin, 1987; Lensi et al., 1995; Parry et al., 1999). Moreover, water movement by capillary rise to the upper soil horizons allows the maintenance of both aerobic and anaerobic environments in this organic horizon, a condition favorable for both nitrification and denitrification processes to occur (Groffman and Tiedje, 1988). Indeed, the upper soil horizons have the highest concentrations of both ammonium and nitrate whatever the season (Table 3). The apparent low net mineralization (NMIN) and nitrification (NITR) (Table 3) in the lower zones of the topohydrosequence can be explained by the high denitrification, which is not accounted for in the method used to evaluate these processes.

The potential denitrification activities in the upper horizons were always significantly higher than the measured DNT rates. This high potential denitrification at the boundary between aerobic and anaerobic conditions can lead to high emission of natural N2O both during the nitrification process (Firestone and Davidson, 1989; Burt et al., 1999; Mogge et al., 1999) and incomplete nitrate reduction through denitrification (Kliewer and Gilliam, 1995; Abbasi and Adams, 2000).

Denitrification Limiting Factors
Along the soil topohydrosequence, the denitrifying bacteria pool (as measured by ANACN) of the upper soil horizon did not vary significantly (Table 5) on a seasonal basis. Despite large ground water table variations (up to 1 m; Fig. 2), at least in the upper part of the topohydrosequence, the soil moisture content of this upper horizon did not vary significantly on a seasonal basis. It remained around 33% (w/v) because of the precipitation regime. Hence, this moderate moisture content coupled with high organic matter content (i.e., 6 to 20%) allowed the persistence of a high density of denitrifiers all year. Even so, the denitrification limiting factors vary seasonally (Fig. 6). This seasonal variation along the topohydrosequence of the type of factors limiting denitrification underpins the fact that this upland–wetland interface fluctuates seasonally as a function of the ground water table and most probably the allochtonous nitrate input from the upslope catchment. These two parameters (i.e., anaerobiosis via ground water level fluctuation and nitrate supply) represent the main triggers of denitrification in an environment rich in C.

On an annual basis, the denitrifier density of the upper soil horizon, estimated as their potential activity (ANACN), increased from the upper part of the topohydrosequence to the lower part (Fig. 4). The potential denitrification rates were within the range of those measured in the south of France along a tributary of the Garonne (Pinay et al., 1993), on the Garonne alluvial soils (Pinay et al., 2000), and on poorly drained soils of riparian forests in the northeastern USA. (Groffman et al., 1992). In the upper part of the topohydrosequence, close to the upland area, the most important limiting factor of denitrification was the lack of anaerobic conditions. Indeed, whatever the site considered, soil incubation under anaerobiosis (ANA) provoked denitrification rates as high as under the incubation with the combination of anaerobiosis, nitrate, and organic carbon (ANACN). The high seasonal ground water level variations hampered the maintenance of complete anaerobiosis conditions, at least from the surface to 50 cm deep, where ANACN was significantly higher than in situ denitrification rates (Fig. 5). In the lower zones, the denitrification-limiting factor gradually shifts from anaerobiosis to nitrate supply (Fig. 7). Due to its lower topographic position, the lower part of the topohydrosequence is for most of the year under anaerobic conditions caused by water table levels close to the ground surface. Nitrification is often blocked at the ammonification stage under anaerobiosis, limiting in situ nitrate supply from organic matter mineralization, except during the summer ground water table drawdown. This lower part of the topohydrosequence belongs to the so-called potential buffer zone where nitrate is removed from ground water (Haycock et al., 1993). Moreover, at this distance from the cropland (i.e., 12 m), ground water nitrate supply is very limited since a large part of it has been already reduced upslope (Fig. 1).

Does Vegetation Cover Matter?
In most cases, in situ denitrification rates did not differ significantly among the forested, understory vegetation, and wet grass sites (Table 4). This result is in accordance with Addy et al. (1999), who did not find any significant difference of nitrate removal between subsoils of forested and mown riparian buffer zones in a mesocosm experiment. One explanation is that the adjacent location of the three types of vegetation cover could result in a similar composition and distribution of roots in the different sites. However, the sites were wide enough apart to overcome such a problem. Indeed, our sampling points were surrounded at each site by at least 20 m of the same vegetation cover on both sides. Another reason could be that the highly variable responses of denitrification activity in soil cores, and the core incubation method used, could overwhelm intersite differences due to plant and bacterial interaction.

The main reason for this absence of significant differences among land covers could be that they all provide enough organic C for the heterotrophic denitrifying bacteria. The constant decrease of ground water nitrate concentration along the three different topohydrosequences supports this hypothesis. However, this does not mean that there could never be any shortage of available organic C under a higher carbon demand, which could be generated by either a seasonal or a chronic increase of allochtonous nitrate from the upland catchment (Haycock and Pinay, 1993). Nevertheless, this lack of a significant difference between land covers underscores the importance of the topography of the stream valley, rather than the vegetation cover, as a control of denitrification activity in a riparian wetland.


   CONCLUSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSION
REFERENCES
 
The strongly significant gradients of both in situ denitrification rates and denitrifier density (ANACN) confirm that the upper soil horizon of the riparian zone is the most active. Even so, this does not imply that it represents the only effective horizon to reduce subsurface allochtonous nitrate input. Indeed, during the low-water period, the ground water level is far below the soil surface, and in turn inaccessible to denitrifiers of the upper horizon. Hence, the observed disappearance of ground water nitrate is then due to denitrification in the deeper soil layers, this being confirmed by the measurement of significant denitrification rates up to 75 cm deep. It implies that there exists enough carbon availability for the denitrifying bacteria, at least up to 75 cm deep in these study sites, and whatever the land cover of the wetland (forest, understory vegetation, or grass). The vegetation covers tested did not exhibit any significant difference at the spatial and temporal scales analyzed. In terms of management, this implies that the perennial vegetation covers tested all provide enough organic C to sustain a high density of heterotrophic denitrifying bacteria. In fact, the most important factors were lateral zonation from the upland border to the riparian zone and the vertical gradient down the soil profile. At the upland–wetland interface scale, even though there exists a strong vertical gradient, denitrification activity should be considered as a function of the entire soil profile and/or volume rather than as a surface-related process. This volume varies within the aquifer according to the ground water table elevation.


   ACKNOWLEDGMENTS
 
This study is part of a European project NICOLAS (Nitrogen Control by Landscape Structures in Agricultural Environments), which was supported by a grant (ENV4-CT97-0395) from the Environment and Climate Directorate of the European Union.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSION
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2002 31: 705-710. [Full Text]  



This article has been cited by other articles:


Home page
 Soil Sci.Home page
J. H. Davis, S. M. Griffith, W. R. Horwath, J. J. Steiner, and D. D. Myrold
Denitrification and Nitrate Consumption in an Herbaceous Riparian Area and Perennial Ryegrass Seed Cropping System
Soil Sci. Soc. Am. J., September 1, 2008; 72(5): 1299 - 1310.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
E. O. Young and R. D. Briggs
Nitrogen Dynamics among Cropland and Riparian Buffers: Soil-Landscape Influences
J. Environ. Qual., May 7, 2007; 36(3): 801 - 814.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
M. E. Hernandez and W. J. Mitsch
Denitrification Potential and Organic Matter as Affected by Vegetation Community, Wetland Age, and Plant Introduction in Created Wetlands
J. Environ. Qual., January 9, 2007; 36(1): 333 - 342.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
O. Montreuil and P. Merot
Nitrogen Removal in Valley Bottom Wetlands: Assessment in Headwater Catchments Distributed throughout a Large Basin
J. Environ. Qual., October 27, 2006; 35(6): 2113 - 2122.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
P. J. A. Kleinman, M. S. Srinivasan, C. J. Dell, J. P. Schmidt, A. N. Sharpley, and R. B. Bryant
Role of Rainfall Intensity and Hydrology in Nutrient Transport via Surface Runoff.
J. Environ. Qual., July 1, 2006; 35(4): 1248 - 1259.
[Abstract] [Full Text] [PDF]

Home page
 Progress in Physical GeographyHome page
T. P. Burt and G. Pinay
Linking hydrology and biogeochemistry in complex landscapes
Progress in Physical Geography, September 1, 2005; 29(3): 297 - 316.
[Abstract] [PDF]

Home page
 J. Environ. Qual.Home page
D. Q. Kellogg, A. J. Gold, P. M. Groffman, K. Addy, M. H. Stolt, and G. Blazejewski
In Situ Ground Water Denitrification in Stratified, Permeable Soils Underlying Riparian Wetlands
J. Environ. Qual., March 1, 2005; 34(2): 524 - 533.