JEQ Grow Your Career With ASA
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 (15)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hill, A. R.
Right arrow Articles by Langat, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hill, A. R.
Right arrow Articles by Langat, J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Hill, A. R.
Right arrow Articles by Langat, J.
Related Collections
Right arrow Watershed and Landscape Processes
Right arrow Water Pollution
Right arrow Nitrogen
Right arrow Soil Analysis
Published in J. Environ. Qual. 33:911-919 (2004).
© ASA, CSSA, SSSA
677 S. Segoe Rd., Madison, WI 53711 USA

TECHNICAL REPORTS

Landscape and Watershed Processes

Denitrification Potential in Relation to Lithology in Five Headwater Riparian Zones

Alan R. Hill*, Philippe G. F. Vidon and Jackson Langat

Department of Geography, York University, Toronto, ON, Canada M3J 1P3

* Corresponding author (hill{at}yorku.ca).

Received for publication March 11, 2003.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 STUDY AREA
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The influence of riparian zone lithology on nitrate dynamics is poorly understood. We investigated vertical variations in potential denitrification activity in relation to the lithology and stratigraphy of five headwater riparian zones on glacial till and outwash landscapes in southern Ontario, Canada. Conductive coarse sand and gravel layers occurred in four of the five riparian areas. These layers were thin and did not extend to the field–riparian perimeter in some riparian zones, which limited their role as conduits for ground water flow. We found widespread organic-rich layers at depths ranging from 40 to 300 cm that resulted from natural floodplain processes and the burial of surface soils by rapid valley-bottom sedimentation after European settlement. The organic matter content of these layers varied considerably from 2 to 5% (relic channel deposit) to 5 to 21% (buried soils) and 30 to 62% (buried peat). Denitrification potential (DNP) was measured by the acetylene block method in sediment slurries amended with nitrate. The highest DNP rates were usually found in the top 0- to 15-cm surface soil layer in all riparian zones. However, a steep decline in DNP with depth was often absent and high DNP activity occurred in the deep organic-rich layers. Water table variations in 2000–2002 indicated that ground water only interacted frequently with riparian surface soils between late March and May, whereas subsurface organic layers that sustain considerable DNP were below the water table for most of the year. These results suggest that riparian zones with organic deposits at depth may effectively remove nitrate from ground water even when the water table does not interact with organic-rich surface soil horizons.

Abbreviations: DNP, denitrification potential • OM, organic matter


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
STUDY AREA
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
STREAM RIPARIAN ZONES have received much attention recently because of their role as critical landscape interfaces for the regulation of energy and material fluxes from terrestrial to aquatic ecosystems. Numerous studies have examined the effect of stream riparian areas on nitrate removal from subsurface flows contaminated by agriculture and other human activities (Peterjohn and Correll, 1984; Lowrance et al., 1984; Simmons et al., 1992; Gilliam, 1994; Hill, 1996). Despite this major research focus, there are still many gaps in our knowledge. One of these areas of uncertainty involves the effects of riparian zone lithology and stratigraphy on nitrogen removal. Variations in soil texture and the distribution of organic matter that supports microbial activity in riparian deposits can influence the effectiveness of riparian zones as pollutant buffers. However, most previous studies with the recent exception of a study by Devito et al. (2000) have not examined riparian zone lithology in detail.

The flow of water through riparian areas is strongly affected by the texture and the stratigraphy of the subsurface materials (Huggenberger et al., 1998). Highly conductive gravel layers at depth beneath low-permeability sediments can create a conduit for ground water flow that bypasses the riparian zone (Burt and Haycock, 1996; Burt, 1997). Recently, Devito et al. (2000) reported that coarse sand and gravel lenses were associated with the movement of nitrate-rich plumes of ground water across a southern Ontario floodplain.

Denitrification, the conversion of NO3– to N gases, has been identified as an important mechanism for nitrate removal in riparian areas (Schipper et al., 1993; Hanson et al., 1994; Verchot et al., 1997; Martin et al., 1999). Many studies have measured high rates of denitrification in the upper few centimeters of riparian soils, whereas a low or non-existent potential for denitrification in the subsurface has been linked to a lack of available carbon (Lowrance, 1992; Groffman et al., 1992, 1996; Schnabel et al., 1996; Burt et al., 1999). Recent evidence, however, indicates important interactions between denitrification and organic matter at depth in some riparian zones. Gold et al. (1998) and Jacinthe et al. (1998) found that small patches of organic matter in the C horizon of poorly drained riparian soils function as "hot spots" of denitrification activity. Hill et al. (2000) measured high rates of denitrification in narrow zones near interfaces between aquifer sands and either peat or buried river channel sediments at depths of several meters in a floodplain. Although denitrification rates were highest in the upper soil layer, significant denitrification activity was measured up to 75 cm deep along topohydrosequences in three riparian wetlands (Clement et al., 2002). These studies suggest that the occurrence and distribution of organic matter in the vertical dimension beneath riparian areas in relation to denitrification activity requires further study.

The goal of the present study was to analyze the lithology and stratigraphy of five headwater stream riparian zones. Our specific objectives were to (i) examine subsurface variations in sediment texture and organic matter distribution to identify the occurrence of coarse-grained and organically enriched layers and (ii) measure vertical variations in potential denitrification in relation to subsurface lithology.


   STUDY AREA
TOP
 ABSTRACT
 INTRODUCTION
 STUDY AREA
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The five study sites are located in headwater agricultural watersheds on sandy loam glacial till and glacial outwash landscapes near Toronto in southern Ontario, Canada (Fig. 1) . These riparian zones were selected to represent the range of geomorphic settings that is characteristic of this region. Three of the riparian areas are located on glacial tills and border a first-order tributary of the Nottawasaga River (Hwy. 27 site), a first-order tributary of Duffins Creek (Ganatsekiagon site), and a second-order stream that drains into Lake Simcoe (Maskinonge Creek site). The Hwy. 27 and Maskinonge riparian zones have a concave topography with moderately steep slopes near the riparian–field boundary and level terrain near the stream (Fig. 2) . The Ganatsekiagon riparian zone has a slightly convex topography with an average slope of 13.2° that extends to the stream (Fig. 3) .



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 1. Location of study sites.

 


View larger version (36K):
[in this window]
[in a new window]
  		Fig. 2. Vertical cross-sections of (top) Highway 27, (middle) Vivian Creek, and (bottom) Maskinonge riparian zones showing lithology and saturated hydraulic conductivity (cm d–1) for piezometer openings (dots). The < term indicates hydraulic conductivity of <0.2 cm d–1. Dashed lines indicate the maximum and minimum water table positions during 2000–2002. Vertical arrows indicate the position of the riparian–field perimeter and stream bank. Numbers above represent piezometer nests and soil sampling locations.

 


View larger version (33K):
[in this window]
[in a new window]
  		Fig. 3. Vertical cross-sections of (top) Ganatsekiagon and (bottom) Road 10 riparian zones showing lithology and saturated hydraulic conductivity (cm d–1) for piezometer openings (dots). The < term indicates hydraulic conductivity of <0.2 cm d–1. Dashed lines indicate the maximum and minimum water table during 2000–2002. Vertical arrows indicate the position of the riparian–field perimeter and stream bank. Numbers above represent piezometers nests and soil sampling locations.

 
The fourth riparian area borders Vivian Creek, a second-order stream that drains into Lake Simcoe. This riparian zone has very gently sloping topography and is underlain by an outwash silt deposit (Fig. 2). The final site is a 31-m-wide riparian zone that borders a first-order unnamed stream on the level Alliston sand plain (Fig. 3). This location, which is identified as the Road 10 riparian zone, is underlain by a 9- to 12-m thick unconfined sand aquifer (Devito et al., 2000). However, the riparian area only interacts with the upper part of the aquifer because the stream is incised to a shallow depth of 2 to 3 m below the sand plain surface. The vegetation of the riparian zones consists of an herbaceous community of grasses and forbs with scattered shrubs and small deciduous trees. Areas of fertilized cropland occur upslope from these riparian zones.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 STUDY AREA
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
A transect of piezometer nests and wells was installed extending from the field–riparian margin to the stream in each riparian zone. Piezometers constructed from 1.27-cm-i.d. PVC pipe with 20-cm slotted ends were installed at depths between 0.5 and 5.5 m depending on sites. Ground water wells (5.1-cm-i.d. ABS pipe) 1.5 to 2.0 m long, perforated throughout their length, were installed at most piezometer nests. Saturated hydraulic conductivities were measured in piezometers using the Hvorslev water recovery method (Freeze and Cherry, 1979). Ground water levels were measured at least monthly from March 2000 (Maskinonge, Ganatsekiagon, and Vivian sites), May 2000 (Hwy. 27), and July 2000 (Road 10 sites) to September 2002. The topography of the riparian zone and the adjacent upland were surveyed using a topographical station.

Riparian zone lithology was first determined from visual inspection of soils during piezometer and well installation. The confining layer defined as a horizon restricting ground water flow was determined in each riparian zone using a combination of the lithology survey and hydraulic conductivity data. Soil samples were later collected in 15- or 20-cm depth increments using a bucket augur at two or three additional locations in each riparian area. Gravel, different fractions of sand, and the silt + clay fraction were separated by dry-sieving to determine soil texture. Organic matter content was determined by loss on ignition at 430°C for 24 h (Davies, 1974).

Potential denitrification rates were determined on the soil samples at various depth increments using the acetylene block technique that inhibits the final conversion of N2O to N2 gas. Three soil replicates were sampled at two depths in the riparian zones to investigate variability in potential denitrification rates within sampling locations. Samples of homogenized fresh soil consisting of 40 g were placed in 250-mL serum bottles. The samples were slurried with 50 mL of solution treated with nitrate (100 mg NO3––N L–1 as KNO3–) and acetone-free acetylene gas was added to each bottle to achieve a final concentration of 10% (10 kPa) in the gas phase. Serum bottles were evacuated and flushed three times with argon to ensure anaerobic conditions. Abiotic control bottles in which bacteria were metabolically inhibited with chloroform received the same treatment as the experimental bottles. The slurries were incubated at 20°C and gas headspace samples were collected from each bottle at 2, 5, 24, 48, and 72 h of incubation. Incubations were extended to 144 h for soil samples collected from the Vivian, Hwy. 27, and Road 10 riparian zones. These incubations were limited to 6 d because acetylene can stimulate microbial activity after one week by acting as a C source (Terry and Duxbury, 1985). Before headspace sampling bottles were shaken vigorously. Nitrous oxide was determined by using a gas chromatograph equipped with an electron capture detector. The N2O produced was corrected for the amounts dissolved in the water. All values were also corrected for headspace concentrations in abiotic control bottles. Denitrification rates (mg kg–1 dry soil d–1) were estimated from the rate of nitrous oxide accumulation between 2 and 5 h, and after 24, 48, 72, and 144 h. At the conclusion of the incubations, gas phase volumes were measured and wet and dry (at 105°C) weights of sediment were determined. The nitrate concentrations remaining in the slurries were analyzed colorimetrically by the cadmium reduction method on a Technicon (Tarrytown, NY) AutoAnalyzer. Final concentrations of nitrate were always >30 mg NO3––N L–1, indicating that nitrate was nonlimiting during the incubations.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 STUDY AREA
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Considerable variations in texture with depth occurred in the five riparian areas. Soils at Hwy. 27 were sandy loams on the slope near the field boundary changing to sandy and coarse sandy soils downslope toward the stream (Fig. 2, top). Layers with a high gravel and coarse sand (0.5–2 mm) content occurred in the top 0 to 30 cm and at depths of >1 m near the stream (Fig. 4) . The deeper 10- to 20-cm-thick coarse-grained layer was found at many coring locations overlying the till aquitard, suggesting that it extends as a continuous layer from the stream bank to the hillslope base at Site 6 (Fig. 2, top). The Vivian Creek riparian area showed a similar pattern of finer-textured soils near the riparian–field boundary and an extensive coarse-grained horizon overlying the silt confining unit nearer the stream (Fig. 2, middle). This layer varied from 15 to 30 cm in thickness and had a higher gravel or coarse sand content than adjacent layers (Fig. 5) . The hydraulic conductivity (Ksat) of the soils overlying the confining unit near the riparian perimeter generally ranged from 5 to 22 cm d–1, whereas higher Ksat values of 44 to 186 cm d–1 were observed in the gravel layers at the Hwy. 27 and Vivian Creek riparian zones.



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 4. Vertical texture profiles of the Ganatsekiagon and Highway 27 riparian zones. Numbers indicate soil sampling locations shown in Fig. 2 and 3.

 


View larger version (35K):
[in this window]
[in a new window]
  		Fig. 5. Vertical texture profiles of the Vivian Creek and Road 10 riparian zones. Numbers indicate the soil sampling locations shown in Fig. 2 and 3.

 
Coarse-grained layers were not found at depth in the Maskinonge riparian zone (Fig. 2, bottom). Sediments overlying the till aquitard at the field–riparian margin were loamy sands and loamy fine sands (data not shown) with Ksat values of 3 to 18 cm d–1. A loamy sand horizon with low Ksat values of 0.1 to 2 cm d–1 separated an overlying peat deposit from the basal till (Ksat < 0.1 cm d–1) within the riparian area (Fig. 2, bottom). A coarse sandy ablation till with variable Ksat values (1.2 to >300 cm d–1) overlay a dense basal till (Ksat < 0.1 cm d–1) at depths of 1.4 to 1.6 m in the Ganatsekiagon riparian zone (Fig. 3, top). The texture in the upper 100 cm at the stream bank was loamy sand and sand. Upslope, many lenses with a high gravel or coarse sand content were present, particularly between 50- and 100-cm depths (Fig. 4).

Coring and textural analysis at the Road 10 riparian zone showed a high coarse sand content in the first 50 cm and at depths of >200 to 300 cm, whereas sediment layers at intervening depths often had higher silt plus clay and fine sand (0.05–0.25 mm) fractions (Fig. 5). Higher Ksat values (150 to >300 cm d–1) were observed in the coarser sediments at depth, whereas Ksat values were often low (1–19 cm d–1) in the sediments at depths of 100 to 250 cm across the riparian area (Fig. 3, bottom).

Organic matter (OM) content ranged from 5 to >15% in the 0- to 15-cm surface soil horizon in the five riparian areas (Fig. 6) . Although OM content declined below this surface horizon to depths of 40 to 50 cm, coring revealed extensive organic-rich layers at greater depths in four of the five riparian zones. A dark brown or black 10- to 15-cm-thick horizon containing small twigs and charred wood fragments overlay a gravel layer and extended from the stream bank to beyond Site 7 at Hwy. 27 and to between Sites 4 and 5 at Vivian Creek (Fig. 2). At Hwy. 27, a second organic-rich layer occurred beneath the surface gravel horizon at 45- to 55-cm depths from the stream bank to beyond Site 2. Organic matter content in these organic-rich layers ranged between 5 and 10% and 17 and 21% at Hwy. 27 and Vivian Creek, respectively (Fig. 6). Coring revealed that patches of organic matter and fine woody debris extended into the underlying coarse-grained horizon, particularly at Hwy. 27 (Sites 1 and 6) and Vivian Creek (Sites 1 and 2) (Fig. 2).



View larger version (28K):
[in this window]
[in a new window]
  		Fig. 6. Vertical profiles of organic matter content in the five riparian zones. Numbers on each profile indicate soil sampling locations shown in Fig. 2 and 3. Note that the horizontal scale differs among the riparian zones and the vertical scale for Road 10 differs from the scale of the other sites.

 
A thin organic-rich layer occurred at a depth of 45 to 55 cm from near the field edge at Maskinonge downslope to Site 8 where it began to thicken into a peat deposit with 30 to 62% OM that extended to the stream channel (Fig. 2 and 6). The peat was covered with a 30- to 50-cm-deep organic-rich sandy loam. We did not detect extensive organic-rich layers at depth in the Ganatsekiagon riparian area. Organic matter content on the upper and middle slope sections declined from 5 to 6% in the 0- to 15-cm horizons to <1% near the contact with the basal till (Fig. 6). However, near the stream bank OM increased to 5 to 6% at depths of 35 to 60 cm and was >2% to a depth of 80 cm.

Sediment cores from the Road 10 riparian zone showed an extensive organically enriched deposit at depths of 100 to 130 cm between Sites 3 and 6 that increased to a depth of 100 to 300 cm at Site 2 (Fig. 3, bottom). This deposit contained many thin layers of fine sand and organic matter as well as considerable amounts of wood fragments, many of which were charred by fire, and occasional thin bands of freshwater mollusc shells. The OM content of cores from this deposit ranged from 2 to 5.5% (Fig. 6).

Variations in denitrification potentials (DNP) with depth are presented for the five riparian areas in Fig. 7 through 10 . The standard deviation was 20 to 35% of the mean DNP for N = 3 replicates, suggesting that variability within sample locations is small compared with differences between sample locations. The change in nitrate concentration during the incubations was similar in magnitude to N2O production in the majority of the sediment samples. In general, N2O concentrations showed a rapid linear increase during the 72-h incubations of the surface 0- to 15-cm soil layer. In contrast, a lag time of 24 or 48 h followed by a rapid increase in N2O occurred in some of the deeper sediments, particularly at Hwy. 27, Vivian Creek, and Road 10. Rates of N2O production in these sediments were linear between 72 and 144 h (data not shown). These patterns resulted in similar daily rates of DNP in the surface layer, whether calculated from N2O accumulation between 2 and 5 h, or after 24 and 72 h (Fig. 7–10). In some of the deeper layers, daily rates of DNP after 72 h were 20 to 90 times higher than daily rates estimated from the 2- to 5-h incubation period. However, daily rates of DNP after 144 h were similar to the rates after 72 h, suggesting that these samples had achieved a steady state with respect to DNP rate.



View larger version (25K):
[in this window]
[in a new window]
  		Fig. 7. Vertical variations of potential denitrification (mg N kg–1 dry soil d–1) estimated from N2O accumulation between 2 and 5 h and after 24 and 72 h in the Highway 27 riparian zone (Sites 2 and 6). See Fig. 2 for site locations. The term T indicates <0.1 mg N kg–1 d–1, while N indicates no activity.

 


View larger version (22K):
[in this window]
[in a new window]
  		Fig. 10. Vertical variations of potential denitrification (mg N kg–1 dry soil d–1) estimated from N2O accumulation between 2 and 5 h and after 24 and 72 h in the Road 10 riparian zone (Sites 2, 3, and 5). See Fig. 3 for site locations. The term T indicates <0.1 mg N kg–1 d–1, while N indicates no activity.

 


View larger version (36K):
[in this window]
[in a new window]
  		Fig. 8. Vertical variations of potential denitrification (mg N kg–1 dry soil d–1) estimated from N2O accumulation between 2 and 5 h and after 24 and 72 h in the Vivian Creek riparian zone (Site 2 and 4) and Ganatsekiagon riparian zone (Site 1). See Fig. 2 and 3 for site locations. The term T indicates <0.1 mg N kg–1 d–1, while N indicates no activity.

 


View larger version (35K):
[in this window]
[in a new window]
  		Fig. 9. Vertical variations of potential denitrification (mg N kg–1 dry soil d–1) estimated from N2O accumulation between 2 and 5 h and after 24 and 72 h in the Maskinonge riparian zone (Sites 4 and 7). See Fig. 2 for site locations. The term T indicates <0.1 mg N kg–1 d–1, while N indicates no activity.

 
The highest DNP rates were usually observed in the top 0- to 15-cm soil horizon; however, a steep gradient of DNP decline with depth was absent in sediments from the Hwy. 27, Vivian Creek, and Maskinonge riparian zones (Fig. 7–9). We measured considerable denitrification activity in organic-rich layers at depths of 45 to 85 cm (Vivian Creek), 40 to 55, 90 to 100, and 130 to 145 cm (Hwy. 27), and 40 to 100 cm (Maskinonge). Daily DNP rates estimated from the 72-h incubations at these depths were often similar to rates in the 0- to 15-cm soil layer. Despite a steep decline in DNP activity with depth at the Road 10 riparian zone, rates of 0.02 to 0.07 mg N kg–1 dry soil d–1 were detected to depths of 210 cm at Site 2 (Fig. 10).

We found a positive correlation between DNP rates in subsurface sediment horizons after 72-h incubations and organic matter content (r = 0.76, N = 42, p < 0.05) when all the riparian zones with the exception of the peat deposit at Maskinonge were considered together (Fig. 11) . In contrast, correlation analysis did not show a significant relationship between subsurface DNP and silt + clay content (r = –0.10, N = 42, p > 0.05).



View larger version (11K):
[in this window]
[in a new window]
  		Fig. 11. Relationship between denitrification potential (DNP) and (top) subsurface sediment organic matter content and (bottom) subsurface sediment silt + clay content in the five riparian zones.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 STUDY AREA
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Permeable sediments underlain at depths of 1 to 3 m by a confining layer that restricts ground water flow were found in four of the five riparian zones in this study. Most previous research on nitrate removal in riparian zones has also been conducted at sites with a shallow aquitard (Hill, 1996). Despite the shallow depth of permeable sediments in many riparian zones, our results suggest that these deposits may have a heterogeneous lithology. The data obtained from the five study sites indicate the widespread occurrence of conductive coarse sediment layers. These deposits are a legacy of glacial and fluvio-glacial processes in some riparian zones, for example the ablation till layer at Ganatsekiagon and the coarse sands at depths of ≥3 m at Road 10. In other riparian zones (Hwy. 27 and Vivian Creek), the depth and lateral extent of coarse sand and gravel layers suggest that stream processes produced them. Reineck and Singh (1980) indicate that gravel layers are often the former bed sediments of laterally migrating stream channels and are common features of floodplain lithology. Coarse-grained layers can also be deposited in overbank alluvium by extreme high-energy floods (Knox, 2001).

Highly conductive coarse sediments can increase ground water flow rates and limit nitrate removal in some riparian areas (Correll et al., 1997). However, in our study ground water nitrate concentrations were low in coarse sediment layers, except in the ablation till at the Ganatsekiagon site (Vidon and Hill, 2004). The coarse sand and gravel layers at the Hwy. 27 and Vivian Creek sites were thin and did not occur within the first 10 to 15 m from the field–riparian perimeter. These layers also contained patches of organic matter and were overlain by an extensive organic-rich horizon. This lithostratigraphy limits the potential for ground water to bypass the riparian area without nitrate removal.

We found widespread organic layers at depth in four of the five riparian zones. Recently, Gold et al. (2001) have also reported layers of organic deposits and buried surface soils to depths of 3 m below the water table in five riparian areas with hydric soils in Rhode Island, USA. The presence of woody debris in the organic layer at several of our field sites and the extension of this layer upslope at Hwy. 27 and Maskinonge indicate a buried surface soil that may represent the time of initial European settlement in the early to mid-19th century. Widespread deforestation and agricultural land use during this period led to accelerated soil erosion and valley-bottom sedimentation in many areas of eastern North America (Costa, 1975; Fitzpatrick and Knox, 2000). Buried soils that were surface soils before European settlement have been found at depths of 0.3 to 4 m beneath alluvial deposits in agricultural watersheds in southern Ontario and throughout the eastern USA (Happ et al., 1940; Costa, 1975; Hill, 1976; Knox, 1977; Trimble, 1981; Jackson and Coleman, 1986; Lowrance et al., 1986; Phillips, 1993; Faulkner, 1998; Fitzpatrick and Knox, 2000). Older buried soils that pre-date European settlement also occur frequently in floodplains as a consequence of episodic patterns of Holocene sedimentation (Brakenridge, 1988; Brown, 1996).

The nature and configuration of the organic-rich deposits in the Road 10 riparian area suggests the presence of a buried stream channel that was infilled by sedimentation during overbank flood events. Several other studies have reported subsurface organic-rich sediments in relic channels and channel bar deposits preserved in floodplains (Fustec et al., 1991; Haycock and Burt, 1993; Hill et al., 2000).

Our analysis of sediment samples amended with nitrate indicated the potential for considerable denitrification activity in organic-rich layers that are widespread at depth in four of the five riparian sites investigated. These estimated denitrification rates are potential rates under conditions where nitrate supply is a non-limiting factor and do not indicate in situ rates. Denitrification activity in some portions of many riparian zones may be limited by low ground water nitrate concentrations because of rapid nitrate depletion upslope near the riparian perimeter. Some of the deeper sediment samples showed a lag phase of 24 or 48 h before rapid N2O accumulation was measured. Similar time lags in N2O production have been previously reported, particularly for sites where ambient nitrate levels were low (Aelion and Shaw, 2000). These data suggest that short-term incubations may seriously underestimate the DNP of deeper riparian sediments that require longer time periods of nitrate amendment to provide an opportunity for the microbial community to respond.

Potential denitrification rates in the surface soils of our riparian sites were similar to those measured in the 0- to 25-cm horizon of riparian wetland soils in Brittany, France (Clement et al., 2002). In contrast, DNP rates in deeper horizons at the Maskinonge and Vivian Creek sites were considerably higher than DNP rates measured in 25- to 50- and 50- to 75-cm horizons at the sites in Brittany. The significant positive correlation between subsurface DNP rates and sediment organic matter content indicates the importance of organically enriched layers as an energy source for denitrification at depth in riparian sediments. Sediment silt + clay content that provides a greater surface area for microbial habitats did not influence subsurface DNP. However, the silt + clay range was 2 to 25% in the samples analyzed and texture may exert a greater effect on denitrification when silt + clay content is >65% (Pinay et al., 2000).

Although the highest DNP rates were generally found in the top 0- to 15-cm riparian soil layer, this layer rarely interacted with ground water in the five southern Ontario riparian sites. Maximum water table positions were usually only near the ground surface for a few weeks in late March to May after spring snow melt, whereas subsurface organic layers that sustain significant DNP rates were below the water table for much of the year (Fig. 2 and 3). Minimum water table positions, even during the summer of 2001, which was the driest recorded in 45 yr, remained at depths that maintained contact between ground water and organic-rich sediments in several of the riparian zones.

The extensive literature documenting the widespread occurrence of subsurface organic layers and buried soils in near-stream areas together with our results indicating high DNP rates at depth in these organic-rich layers suggest a need to reassess several current perspectives of riparian zone nitrate dynamics. Researchers have suggested that nitrate removal occurs mainly in riparian zones with hydric soils where the water table lies near the ground surface during much of the year (Burt et al., 1999; Gold et al., 2001; Rosenblatt et al., 2001). In Rhode Island, USA, minimal nitrate removal was found in riparian sites with non-hydric soils, whereas high nitrate removal occurred in hydric riparian sites (Gold et al., 2001). These patterns imply that effective nitrate removal requires an interaction of nitrate-rich ground water with the surface soil horizon where denitrification potential is considered to be greatest. In contrast, our data suggest that riparian zones with non-hydric upper soil horizons and water tables that are usually deeper than 40 cm below the surface can remove nitrates from ground water, if organic-rich layers with significant denitrification potentials are present at depth. This conclusion is supported by data from a parallel study that found that ground water nitrate inputs from adjacent cropland were rapidly removed in the riparian sites except at Ganatsekiagon Creek (Vidon and Hill, 2004).

The influence of different types of riparian vegetation on ground water nitrate removal by plant uptake and denitrification, stimulated by litter inputs to the upper soil horizons, has received considerable attention. However, these studies have produced contradictory results, with some reporting that forested riparian zones had higher nitrate removal rates than grassed riparian areas (Osborne and Kovacic, 1993; Verchot et al., 1997; Lyons et al., 2000), whereas others have found substantial removal in grassed riparian sites (Haycock and Pinay, 1993; Lowrance et al., 1995; Schnabel et al., 1996; Correll et al., 1997; Clement et al., 2002).

The results of our study indicate that the link between riparian vegetation and ground water nitrate removal may be more complex than is usually assumed. At present, the vegetation on the five riparian sites is an herbaceous plant community. However, the wood content of the buried organic-rich layers that displayed significant DNP suggests a former forest cover on these sites. In many riparian areas, it may be necessary to consider how ground water removal is affected by links between past vegetation communities and the availability of carbon in organic-rich deposits at depth, rather than the influence of the contemporary vegetation community.

Researchers have also suggested that denitrification is the major mechanism of nitrate removal from riparian ground water in winter when vegetation is dormant and a high water table increases interaction with organic-rich upper soil horizons, while vegetation uptake is predominant in the growing season when the water table is low (Groffman et al., 1992; Simmons et al., 1992; Haycock et al., 1993; Correll, 1997). Our evidence that organic-rich layers at depth below the water table in riparian areas can sustain significant DNP rates suggests that denitrification can also be a dominant mechanism of nitrate removal from ground water during the growing season. Consequently, for many riparian zones it may be unnecessary to invoke explanations for nitrate removal that involve seasonal changes in removal mechanisms.

Natural floodplain processes and human activities that result in accelerated erosion and valley-bottom sedimentation create a complex stratigraphy in riparian zones. These heterogeneous deposits can include highly conductive coarse sands and gravel layers as well as buried surface soils, organic-rich sediments, and woody debris. Our study indicates that a greater focus on riparian lithology and stratigraphy is needed to improve our understanding of riparian zone nitrate dynamics. In particular, the occurrence of organic-rich deposits that provide energy for microbial transformations at depth beneath the water table has important implications for the role of riparian areas as pollution buffers.


   ACKNOWLEDGMENTS
 
The authors are grateful to Alan Michalsky for preparing the piezometers and wells, Robert McDonald and Tim Duval for assistance in the field and the analysis of soil texture and organic matter content, and the York Geography Department cartographic office for the figures. We also thank the landowners for access to riparian sites. The constructive comments of three anonymous reviewers on the original manuscript are appreciated. The research was supported by grants from the Natural Sciences and Engineering Council of Canada to A.R. Hill.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 STUDY AREA
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2004 33: 799-804. [Full Text]  



This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
W. D. Robertson and S. L. Schiff
Persistent elevated nitrate in a riparian zone aquifer.
J. Environ. Qual., March 1, 2008; 37(2): 669 - 679.
[Abstract] [Full Text] [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.
[Abstract] [Full Text] [PDF]

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 (15)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hill, A. R.
Right arrow Articles by Langat, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hill, A. R.
Right arrow Articles by Langat, J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Hill, A. R.
Right arrow Articles by Langat, J.
Related Collections
Right arrow Watershed and Landscape Processes
Right arrow Water Pollution
Right arrow Nitrogen
Right arrow Soil Analysis

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Vadose Zone Journal Journal of Plant Registrations
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal