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 Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (10) Citing Articles via Google Scholar Google Scholar Articles by Bruland, G. L. Articles by Richardson, C. J. Search for Related Content PubMed Articles by Bruland, G. L. Articles by Richardson, C. J. GeoRef GeoRef Citation Agricola Articles by Bruland, G. L. Articles by Richardson, C. J. Related Collections Wetlands and Aquatic Processes Ecosystem Restoration Wetland Soils

Division S-10—Wetland Soils

Hydrologic Gradients and Topsoil Additions Affect Soil Properties of Virginia Created Wetlands

^a Soil and Water Science Dep., Univ. of Florida, Institute of Food and Agricultural Sciences, 2169 McCarty Hall, P.O. Box 110290, Gainesville, FL 32611-0290
^b Duke Univ. Wetland Center, Nicholas School of the Environment and Earth Sciences, Box 90333, Durham, NC 27708-0333

^* Corresponding author (GBruland{at}ifas.ufl.edu)

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
As the role of soil properties in the development of created wetlands (CWs) has not received adequate attention in regulatory or scientific communities, this study was conducted to evaluate the development of soil properties in 11 CWs in Virginia ranging from 4 to 16 yr since creation. Six of the 11 sites received at least 15 cm of topsoil (TS) while the other five sites received no topsoil (No TS). Cores collected from wet, intermediate, and dry positions at each site were analyzed for moisture, bulk density (D_b), soil organic matter (SOM), texture, water-holding capacity (WHC), P sorption index (PSI), and microbial biomass C (MBC). Both positions along the hydrologic gradient and topsoil status were hypothesized to be significant factors in explaining the variability of the measured soil properties. Soil moisture decreased significantly while D_b increased significantly from wet to dry zones. Moisture, WHC, and PSI were all significantly elevated in certain zones of TS compared with No TS sites. Soil organic matter had significant Spearman correlations with all other measured soil properties, revealing that this parameter was an important indictor of soil quality. In addition, sites with high mean moisture, SOM, and PSI values all received TS; conversely, the site with the lowest mean moisture and SOM content did not receive TS. Thus, amending CW soils with TS appeared to be an effective strategy for increasing soil moisture, WHC, and PSI. Whenever possible, practices such as TS or organic amendments should be employed, especially if wetland creation involves excavation into subsoils with low SOM and high D_b.

Abbreviations: CWs, created wetlands • CV, coefficient of variation • D_b, bulk density • GLM, generalized linear model • MBC, microbial biomass carbon • No TS, did not receive topsoil • PSI, phosphorus sorption index • SOM, soil organic matter • TS, received topsoil • VDOT, Virginia Department of Transportation • WHC, water-holding capacity

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
SECTION 404 of the Clean Water Act mandates compensatory mitigation whenever a wetland is impacted or destroyed by human land-use activities. The ability of these CWs to functionally replace natural wetlands is a topic of considerable debate (Zedler and Calloway, 1999; Kentula, 2000; Hunter and Faulkner, 2001). As current monitoring of CWs is based on meeting hydrologic and vegetative success criteria, it does not require any monitoring of soil properties or processes (Rheinhardt and Brinson, 2000). This is a cause for concern for the following reasons: (i) soil forms the foundation of these developing ecosystems (Stolt et al., 2000); (ii) inadequate soil properties can be detrimental to vegetative survival and the establishment of wetland hydrology (Shaffer and Ernst, 1999); and (iii) soil is the medium for biogeochemical processes that transform and retain nutrients (Mitsch and Gosselink, 2000). Without suitable soil properties, CWs may never functionally replace the natural wetlands that were destroyed.

As soils of natural wetlands have been exposed to wetland processes for periods of decades, centuries, or even millennia, it is unfair to expect CWs that have been exposed to wetland processes for only a few years to immediately exhibit soil properties that are characteristic of natural wetlands. Thus, for this study, comparisons of CWs to natural wetlands were omitted, as this research was conducted to evaluate the development of soil properties in 11 Virginia Department of Transportation (VDOT) CWs ranging in age from 4 to 16 yr since construction. In the period from 1980 to 1996, VDOT constructed more than 105 non-tidal CWs (Whittecar and Daniels, 1999). The focus of this study was on CWs in the Coastal Plain, as this is the location of the majority of wetland impacts in Virginia (Whittecar and Daniels, 1999). The investigation was limited to non-tidal forested CWs for two main reasons: (i) non-tidal CWs have been studied much less than their tidal counterparts; and (ii) natural non-tidal forested wetlands have been shown to perform important nutrient transformation and retention functions (Walbridge and Struthers, 1993; Axt and Walbridge, 1999; Verhoeven et al., 2001). As few CW sites are assessed beyond what is needed to satisfy success criteria mandated by the Army Corps of Engineers, the objective of this study was to investigate the development of soil properties both within and among the 11 CWs. It was hypothesized that position along the hydrologic gradient and topsoil status would be significant factors in explaining the variability of soil properties in the CW sites.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Study Sites
The CWs sampled in this study were located in the Coastal Plain physiographic region of Virginia (Fig. 1 , Table 1). Each site was property of the VDOT and part of its compensatory mitigation program. Sites that met the following criteria were selected for sampling: (1) the site was designed as a created palustrine forested wetland; (2) no jurisdictional wetlands previously existed at the site; and (3) adequate documentation was available from the VDOT central office to locate the site. The 11 sites selected for sampling were located in the counties of Greenville, Prince George, Southhampton, and Sussex as well as the cities of Chesapeake and Suffolk (Table 1). Sites ranged in size from 0.2 to 13.8 ha. Site age, which ranged from 4 to 16 yr, was calculated as the number of growing seasons a site had experienced since planting. Seven of the sites were located in riverine hydrogeomorphic settings and were associated with streams of various orders (Western Freeway [1st order], Courtland Bypass [Nottaway River], Emporia Bypass [4th], Stony Creek [4th], Rowanty Creek [4th], 2nd Swamp [2nd], and Fort Lee [2nd]). The other four sites were located in nonriverine hydrogeomorphic settings (Bower's Hill, Franklin Bypass, Otterdam Bank, and 664 Site 7).

View larger version (30K): [in this window] [in a new window]   Fig. 1. Map of study site locations across the Coastal Plain of Virginia.

After reviewing USDA County Soil Surveys and speaking to Natural Resource Conservation Service personnel in Virginia, it was determined that 14 different soil series from two soil orders (Ultisols and Entisols) were present at the 11 sites sampled in this study (Table 1). All but one of the series (Pactolus, thermic, coated Aquic Quartzipsamments) were Endoaquults, Hapludults, or Paleudults. The Altavista (fine-loamy, mixed, semiactive, thermic Aquic Hapludults), Craven (fine, mixed, subactive, thermic, Aquic Hapludults), Dragston (coarse-loamy, mixed, semiactive, thermic Aeric Endoaquults), Sassafras (fine-loamy, siliceous, semiactive, mesic Typic Hapludults), and Woodstown (fine-loamy, mixed, active, mesic Aquic Hapludults) series were present at both TS and No TS sites, while the State (fine-loamy, mixed, semiactive, thermic Typic Hapludults) series occurred in two TS sites and the Slagle (fine-loamy, siliceous, thermic, Aquic Hapludults) series occurred at four of the No TS sites.

Vegetation planted at these sites during construction included Betula nigra L. (river birch), Cephalanthus occidentalus L. (buttonbrush), Fraxinus pennsylvanica Marsh. (green ash), Magnolia virginiana L. (sweetbay), Myrica cerifera [L.] Small (wax myrtle), Nyssa sylvatica Marsh. (black gum), Quercus bicolor Willd. (swamp white oak), Quercus michauxii Nutt. (swamp chestnut oak), and Taxodium distichum [L.] Rich. (bald cypress). A variety of other woody and emergent species have naturally colonized these sites such as Acer rubrum L. (red maple), B. nigra, Salix nigra Marsh. (black willow), and Typha spp. (cattail).

Sampling Design
During site visits, 20- to 40-m transects were established that ran perpendicular to existing topographic and hydrologic gradients at each site. Each transect was divided into three zones that were representative of the dry, intermediate, and wet sections of the site. In the few cases where microtopography was present, care was taken to locate transects in areas where microtopographic features did not confound the changes that occurred across the hydrologic gradient. The three zones were delineated visually based on observations of the vegetation, hydrology, and soils. The dry zones were characterized by: (i) a variety of upland, old-field, and wetland plant species; (ii) water tables that were below the soil surface; and (iii) coarse-textured soils. The intermediate zones were characterized by: (i) fewer upland species, more facultative and obligate wetland species; (ii) water tables that were near the soil surface; and (iii) finer-texture soils. The wet zones were characterized by: (i) more flood-tolerant, obligate wetland species; (ii) water tables that were near or at the soil surface; and (iii) finer-texture soils. Areas that appeared to permanently inundate were not included in the wet zones. One soil core from each zone was collected at each site for a total of three cores per site.

The cores were sampled from the upper 10 cm of the soil profile in plastic sleeves of 5-cm diam. with a piston corer. This type of sampling allowed for the determination of the bulk density of each individual core. Cores were collected from the upper 10 cm of the soil profile as this corresponds to the zone that is most biologically active and subject to physical processes of erosion and deposition (Pinay et al., 2000). As logistical constraints associated with sampling 11 sites made it impossible to sample all sites on the same day, cores were collected between 18 and 22 Feb. 2002. During the sampling period, climatic conditions were generally consistent across the Coastal Plain. Once collected, cores were stored in a cooler with ice until being transported back to the Duke Wetland Center Laboratory.

Laboratory Analyses
Soil moisture, D_b, SOM, and texture were measured for each core as these soil properties have been shown to affect hydrology, vegetation, and biogeochecmical cycling in wetlands (Axt and Walbridge, 1999; Bridgham et al., 2001; Hunter and Faulkner, 2001; Verhoeven et al., 2001). The WHC, the PSI, and MBC were also measured as indicators of moisture and nutrient retention potential. No other study has quantified the WHC or the PSI of CWs. Furthermore, only one other study has measured MBC in CWs (Duncan and Groffman, 1994). Additional research is needed to determine if CWs in various regions of the country are developing viable microbial communities.

Upon arrival at the laboratory, cores were extruded from the plastic sleeves and split in half vertically with a sharp knife. Half of the core was oven dried at 105°C for 24 h to determine the moisture content and D_b. The dry half was then ground and passed through a 2-mm sieve to remove pebbles and macro-organic matter. The sieved soil was then used to determine percentage of SOM by loss on ignition (Campbell et al., 2002) and soil texture by the pipette method (Sheldrick and Wang, 1993).

The other half of the core was also passed through a 2-mm sieve while being kept at field moisture. The field-moist sieved soil was analyzed for WHC by the percolation method (Forster, 1995), for P sorption capacity by the PSI (Richardson, 1985), and for MBC by the chloroform fumigation extraction (CFE) method (Brookes et al., 1985; Allen, 1999). Previous studies have established that the PSI is a reliable gauge of a wetland soil's P sorption potential and less time-consuming to measure than multiple-point P sorption isotherms (Richardson, 1985; Axt and Walbridge, 1999; Bridgham et al., 2001). The PSI was determined by shaking 2.0 g of sterilized soil with a solution of 130 mg PO₄–P L^–1 for 24 h. Soils were sterilized with two drops of toluene to prevent microbial uptake of phosphate (Richardson and Vaithiyanathan, 1995). The difference in concentration of inorganic P between the initial (130 mg PO₄–P L^–1) and final concentration represents the amount of P sorbed. The index was then calculated as X(LogC)^–1 where X equals the amount of P sorbed (Mg P 100 g soil^–1) and C equals the final inorganic P concentration (Mg P L^–1) in solution.

To determine the MBC, two subsamples of 10 g of oven dry equivalent (ODE) of wet soil were weighed into centrifuge tubes. The first subsample was extracted with 0.5 M K₂SO₄ by shaking for 1 h. Extracts were then filtered with Whatman 42 filter paper. The second subsample was fumigated with ethanol-free chloroform that was pipetted onto two large cotton balls placed in the headspace of the centrifuge tube (Allen, 1999). Each tube was then capped tightly and stored in a dark container for 7 d. Following incubation, chloroform was removed by placing the tubes in a vacuum dessicator. The dessicator was evacuated multiple times and flushed with room air after each evacuation. Incubated soils were then extracted, shaken, and filtered. Control and fumigation extracts were analyzed for total organic C content using a Shimadzu TOC 5000 solution C analyzer (Shimadzu, Inc., Columbia, MD). Microbial biomass C was calculated for each sample as the difference between control and fumigated values of TOC (Brookes et al., 1985).

Statistical Analyses
Due to the unbalanced study design (six sites receiving topsoil or organic amendments and five sites receiving no topsoil or organic amendments) the data were analyzed with a generalized linear model (GLM) procedure in SAS for Windows Version 8.2 (SAS Institute, Cary, NC). The differences among soil properties along the hydrologic gradient and between topsoil (TS) and no topsoil (No TS) sites as well as any interactions between the two factors were examined with this model. As D_b, SOM, Clay, WHC, PSI, and MBC were non-normally distributed, these soil properties were log-transformed before running the GLM analysis. The other three soil properties met the GLM assumptions of normality and homogeneous variances. When main effects were significant, differences in TS and No TS sites across the gradient were determined with a least-squared differences (LSD) procedure (SAS Institute, Cary, NC). A significance level of p < 0.10 was used for the GLM and the LSD tests.

A correlation analysis was also conducted to investigate the relationships among the nine different soil properties. Due to the non-normally distributed data, the soil properties were rank transformed and a Spearman correlation analysis was conducted with Statistica version 5.5 (StatSoft, Tulsa, OK). A significance level of p < 0.10 was used to assess the significance of the Spearman correlations. As there were no significant differences across the hydrologic gradient and between TS and No TS sites for most of the soil properties, the data were not stratified by either one of these factors for the correlation analysis.

Site mean values were also determined for each of the soil properties measured in this study. As the sample size at each site was small (n = 3) and the soils were taken along transects designed to capture hydrologic gradients, it would have been inappropriate to compare these site mean values with t tests or analysis of variance. Despite these caveats, the site mean values were presented as qualitative descriptors of the soils of each site. The coefficient of variation (CV) was also determined for the site means, the means of all TS and all No TS sites, and the mean of all sites. The CV (CV = standard deviation/mean) provided a standardized method for comparing the variability among soil properties with mean values that spanned over two orders of magnitude.

To determine whether there were any temporal trends in soil properties, site mean soil property values were regressed versus site age. No significant trends were discovered. Sample sizes became too small to further stratify regressions by position along the gradient or by topsoil status. The lack of change with site age was attributed to variety of factors including: (i) the fact that pedogenesis occurs over much longer time scales than that captured in this study; (ii) the greater use of TS additions in the younger sites than in the older sites; and (iii) the uneven distribution of site ages in the TS and No TS groups. Furthermore, as there were no significant differences between soil properties of riverine CW sites and nonriverine CW sites, no stratification of the data by hydrogeomorphic setting was employed.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Hydrologic Gradient and Topsoil Status
Statistical analysis of soil moisture indicated that both position along the hydrologic gradient and topsoil status were significant, although their interaction was not (Table 2). Moisture tracked the wetness gradient at both types of sites, and according to the LSD test, TS sites had significantly higher moisture at the wet position than No TS sites at the intermediate and dry positions (Fig. 2a) . The analysis of D_b indicated that position was significant while topsoil status and their interaction were not (Table 2). According to the LSD test, D_b values in the wet zones of both types of sites were significantly lower than D_b values in the dry zones of TS sites (Fig. 2b).

View larger version (40K): [in this window] [in a new window]   Fig. 2. Effect of position along the hydrologic gradient and topsoil status on (a) soil moisture, (b) bulk density, (c) soil organic matter, (d) clay, (e) silt, (f) sand, (h) water holding capacity, (h) the P sorption index, and (i) microbial biomass C. Data presented are means and error bars represent ±1 standard error. Data points with different letters are significantly different at the p < 0.10 level according to a least-squared differences test.

Correlations among Soil Properties
The Spearman correlation analyses revealed that soil moisture had significant negative correlations with D_b and percentage of sand, as well as significant positive correlations with SOM and WHC (Table 3). Bulk density had significant negative associations with SOM and WHC. Soil organic matter was significantly correlated to every other soil property measured in this study. Percentage of silt also had significant positive associations with WHC, and MBC and percentage of sand had significant negative associations with WHC and MBC. Water holding capacity was also significantly related to MBC.

View this table: [in this window] [in a new window]   Table 3. Spearman rank correlations among the soil properties measured at the 11 created wetland sites in Virginia. Only those correlations significant at the p < 0.10 level were reported in the table. (ns = not significant.)

View this table: [in this window] [in a new window]   Table 4. Site mean values (coefficient of variation [CV] in parenthesis) for the soil properties of sites that received topsoil (TS) and sites that did not receive topsoil (No TS). Immediately below each section are rows of mean values (CV) for all TS and No TS sites. The final row in the table indicates the overall mean (CV) for all sites.

The mean WHC for all sites was 55.3%. Although there was some within-site variability, the CVs for WHC of the two types of sites (TS sites = 0.22, No TS sites = 0.23) and across all sites (0.25) were quite similar. Bower's Hill had the highest mean WHC at 85.6%, while Rowanty Creek had the lowest mean WHC at 44.3%. The mean PSI for all sites was 22.8, and within- and among-site variability was similar to that of WHC. The Stony Creek site had the highest mean PSI of 36.6 while the Second Swamp sites had the lowest PSI of 12.2. The mean MBC for all sites was 138.6 Mg C kg soil^–1. After SOM, MBC displayed the next highest within and among site variability of all the soil properties measured in this study. The CVs for MBC were 0.29 for all TS sites, 0.61 for all No TS sites, and 0.43 for all sites. Second Swamp had the highest mean MBC at 239.6 Mg C kg soil^–1, while the 10-yr-old No TS site, Otterdam Bank had the lowest mean MBC at 53.5 Mg C kg soil^–1.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Hydrologic Gradient and Topsoil Status
The small number of sites and replicate cores collected at each site may have limited our ability to detect significant differences; however, differences across the gradient were significant for soil moisture and D_b, and differences between topsoil status were significant for moisture, WHC, and PSI. Soil moisture decreased significantly while D_b increased significantly from wet to dry zones in both types of sites. While not statistically significant, SOM also decreased from wet to dry zones in both types of sites. The lack of significant differences in soil texture in the TS and No TS sites indicated that differences in moisture, WHC, and PSI were related to topsoil additions or hydrologic gradients instead of textural differences in the two site types. Soil moisture, WHC, and PSI were all significantly higher in certain zones of TS sites than they were in No TS sites. These results were similar to those of a study of organic matter amendments at a CW in Pennsylvania, where amended areas exhibited higher soil moisture than unamended areas (Stauffer and Brooks, 1997).

If wetland creation involves excavation of surface horizons to expose nutrient-poor subsoils with high bulk densities, as is often the case in the Southeastern Coastal Plain, it may be necessary to amend these types of sites with topsoil or organic matter to improve retention of moisture, development of plant and microbial communities, and storage of P. As deficiencies in the soil properties of CWs may remain problematic for long periods of time, it is imperative for initial soil conditions to be considered carefully. Spending more money on topsoil or organic amendments during construction may actually pay for itself in the long-term, as sites with improved soil conditions will be more likely to meet the hydrologic and vegetative success criteria mandated by the Army Corps of Engineers. These sites will also be much less likely to require costly remedial work such as regrading of the soil surface or replanting of additional vegetation.

While it may be too costly to amend large CW sites in their entirety, there would be value in amending certain sections or subplots within the site. It appears that funding may be best spent by amending sections of CWs of intermediate elevation and hydrology. Unlike wet areas that may accumulate SOM, or dry areas in which added SOM may be subject to decomposition due to aerobic conditions, topsoil, or amendments added to intermediate zones may actually persist in these zones long enough to stimulate vegetative and microbial development. Sites amended with topsoil or organic matter would provide valuable opportunities for collaboration with university researchers and desperately needed quantitative data about management options for CWs. Designers, engineers, and site managers should consider innovative approaches to wetland creation such as salvaging soils of impacted sites, amending with topsoil or organic matter, deep ripping, and reestablishing microtopography. All of these practices have the potential to create better initial soil conditions that will, in turn, encourage more rapid development of wetland hydrology, vegetation, and biogeochemical cycling.

Correlations Among Soil Properties
Soil organic matter had significant positive correlations with moisture, clay, silt, WHC, PSI and MBC, and significant negative correlations with D_b and the percentage of sand, revealing that this parameter was an important indictor of soil quality at these young CWs. Furthermore, in studies with budget or time constraints, SOM may be the best single variable to measure at these types of sites because it provides information about a number of other soil properties and processes. Cores with a high percentage of sand content had low moisture, SOM, and WHC. In contrast, cores with a high percentage of silt content had high moisture, SOM, WHC, and MBC. Previous research has shown that soils dominated by coarse-textured sands to be unable to hold sufficient water for plant survival (Sopher and Baird, 1978) and much less effective in retaining nutrients than soils dominated by fine-textured silts and clays (Poach and Faulkner, 1998). Cores with high SOM content also had high PSI values. Contrary to the assumption that SOM inhibits P sorption, in this study SOM and PSI had a significant positive association. A recent study of P sorption in natural wetlands in Virginia also found a strong positive correlation between SOM content and PSI (Axt and Walbridge, 1999). Soil organic matter, percentage of silt, and WHC were positively correlated to MBC. Thus, it may be safe to assume that future CW sites in the Virginia Coastal Plain with low SOM and percentage of silt content will have low MBC in their initial years of development.

To recap the significant differences in the data across the hydrologic gradient, between TS and No TS sites, and significant relationships among soil properties, a summary table was created (Table 5). From the table it is interesting to note position along the gradient accounted for two significant differences (only one if the p-value for the effects of position on moisture would have been rounded up to 0.10), while topsoil status accounted for three significant differences. Thus, topsoil status may be slightly more important than position along the hydrologic gradient in explaining the variability of soil properties in the CWs.

View this table: [in this window] [in a new window]   Table 5. Table columns represent soil properties while rows represent different comparisons including from wet to dry along the hydrologic gradient, between sites receiving topsoil and sites that did not receive topsoil, and correlation among the measured soil properties. Statistically significant results are listed in the table.

It was interesting to note that the youngest site, Stony Creek, had the highest mean SOM, and the oldest site, Rowanty Creek, had the lowest mean SOM. This was attributed to the fact that Stony Creek received TS and organic amendments while Rowanty Creek did not. The different mitigation practices, positions along the hydrologic gradient, and site ages all contributed to the high standard deviations and coefficients of variation reported in this study for SOM. Despite this variation, however, our results for SOM were similar to those reported in other studies of CWs. For example, the mean SOM content of the 11 CW sites sampled in this study was 7.0%, which was very close to the mean SOM content of 6.2% reported in a study of 44 CWs in Pennsylvania (Bishel-Machung et al., 1996), and the mean SOM content of 6.7% reported in a study of nine CWs in Oregon (Shaffer and Ernst, 1999). A study of three natural wetlands in the Virginia Coastal Plain reported mean SOM contents of 23.4, 9.8, and 4.9% (Axt and Walbridge, 1999). Thus, compared with natural wetlands in the region, the mean SOM contents reported for CWs in this study were comparable, albeit somewhat lower.

The CWs in this study displayed a mean clay content of 8.8%, a mean silt content of 20.5%, and a mean sand content of 70.7%. The site mean sand contents of the CWs appeared to be comparable, though slightly higher, than those reported in a recent study of natural wetlands of the Virginia Coastal Plain (Axt and Walbridge, 1999). The values obtained in this study at the Western Freeway site for mean clay, silt, and sand (10, 17, and 73%) content were similar to those reported for a previous study of this site (6, 14, and 79%) (Stolt et al., 2000).

The site mean PSI values ranged from 12.2 to 36.6. In comparison, a seminal study of P sorption in natural wetlands reported PSI values for a Maryland swamp ranging from 21 to 163, a Michigan swamp ranging from 30 to 45, a Michigan fen ranging from 35 to 41, and a North Carolina pocosin ranging from 8 to 30 (Richardson, 1985). Thus, the range of PSI values from the CWs in Virginia was lower than the ranges of PSI values observed for other types of natural wetlands with mineral soils. There are a variety of factors that may have explained the low PSI values in the CWs, such as differences in soil type, parent material, weathering, type of vegetation, and organic matter compared with that of the natural wetlands.

Microbial biomass C mirrored SOM in terms of variability, as it was highly inconsistent within and among sites. The range of MBC values reported in this study (53.5–239.6 mg C Kg soil^–1) was lower than those reported in a study of constructed wetlands in Massachusetts (768–1088 mg C Kg soil^–1) (Duncan and Groffman, 1994), possibly due to the following factors: (i) different methods were used to quantify MBC in each study; and (ii) the soils of the constructed wetlands in Massachusetts consisted of substrate from a nearby pond-dredging project that would be expected to have a higher SOM and MBC than the upland subsurface horizons that were the typical substrates for the CWs in Virginia.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 
Soil properties of CWs in Virginia were affected both by position along the hydrologic gradient and by topsoil status. Soil moisture and D_b decreased significantly from wet to dry zones. Perhaps even more important than the variability explained by the hydrologic gradient, was that explained by topsoil status. Moisture, WHC, and PSI were all significantly elevated in certain positions of TS sites in comparison with No TS sites. Soil organic matter was significantly correlated with the other eight soil properties measured in this study, suggesting that SOM was an important indicator of soil quality and should be measured in future studies. Furthermore, sites with high moisture, SOM, and PSI all received topsoil or topsoil and organic amendments; conversely, the site with the lowest moisture, SOM, and highest sand content, did not receive any topsoil or organic amendments. Thus, in the Virginia Coastal Plain, an effective strategy for increasing soil moisture, WHC, and PSI in CWs appears to be amending sites with topsoil or organic matter. Whenever possible, and particularly when wetland creation involves excavation into highly compacted or low SOM subsoils, practices such as topsoil and organic matter amendments should be encouraged. Without suitable soil properties, CWs may never functionally replace the natural wetlands that were destroyed.

	ACKNOWLEDGMENTS

 
We thank L. Snead of the VDOT Central Office in Richmond, VA for help with site selection, S. Russell and P. Constanzer of VDOT for assistance with site location and soil coring, and W. Willis for aiding with laboratory analysis at the Duke University Wetland Center. H. Bruland's editing of this manuscript is gratefully acknowledged. Comments by two anonymous reviewers also improved this manuscript. Funding was provided by the Duke Wetland Center Case Studies Program and by a Graduate Research Fellowship to the first author from the Center for Transportation and the Environment in Raleigh, NC. Although this paper has gone through the VDOT publications review process, it may not necessarily reflect the view of VDOT and no official endorsement should be inferred.

Received for publication August 11, 2003.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION CONCLUSIONS REFERENCES

 

• Allen, A.S. 1999. Effects of elevated atmospheric CO₂ on soil nitrogen availability. Ph.D. Diss. Duke University, Durham, NC.
• Axt, J.R., and M.R. Walbridge. 1999. Phosphate removal capacity of palustrine forested wetlands and adjacent uplands in Virginia. Soil Sci. Soc. Am. J. 63:1019–1031.[Abstract/Free Full Text]
• Bishel-Machung, L., R.P. Brooks, S.S. Yates, and K.L. Hoover. 1996. Soil properties of reference wetlands and wetland creation projects in Pennsylvania. Wetlands 16:532–541.
• Bridgham, S.D., C.A. Johnston, and J.P. Schubaurer-Berigan. 2001. P sorption dynamics in soils and coupling with surface and pore water in riverine wetlands. Soil Sci. Soc. Am. J. 65:577–588.[Abstract/Free Full Text]
• Brookes, P.C., A. Landman, G. Pruden, and D.S. Jenkinson. 1985. Chloroform fumigation and the release of soil nitrogen: A rapid and direct extraction method to measure microbial biomass nitrogen in soil. Soil Biol. Biochem. 17:837–842.
• Campbell, D.A., C.A. Cole, and R.P. Brooks. 2002. A comparison of created and natural wetlands in Pennsylvania, USA. Wetlands Ecol. Manage. 10:41–49.
• Duncan, C.P., and P.M. Groffman. 1994. Comparing microbial parameters in natural and constructed wetlands. J. Environ. Qual. 23:298–305.[Abstract/Free Full Text]
• Forster, J.C. 1995. Soil physical analysis. p. 106. In K. Alef and P. Nannipieri (ed.) Methods in applied soil microbiology and biochemistry. Academic Press, London.
• Hunter, R.G., and S.P. Faulkner. 2001. Denitrification potentials in restored and natural bottomland hardwood forests. Soil Sci. Soc. Am. J. 65:1865–1872.[Abstract/Free Full Text]
• Kentula, M.E. 2000. Perspectives on setting success criteria for wetland restoration. Ecol. Eng. 15:199–209.
• Mitsch, W.J., and J.G. Gosselink. 2000. Wetlands. 3rd ed. John Wiley & Sons, New York.
• Pinay, G., V.J. Black, A.M. Planty-Tabacchi, B. Gumiero, and H. Decamps. 2000. Geomorphic control of denitrification in large river floodplain soils. Biogeochemistry 50:163–182.
• Poach, M.E., and S.P. Faulkner. 1998. Soil phosphorus characteristics of created and natural wetlands in the Atchafalya Delta, LA. Est. Coast. Shelf Sci. 46:195–203.
• Rheinhardt, R.D., and M.M. Brinson. 2000. An evaluation of the effectiveness of existing NCDOT wetland mitigation sites. CTE/NCDOT Joint Environmental Research Program, Raleigh, NC.
• Richardson, C.J. 1985. Mechanisms controlling phosphorus retention capacity in freshwater wetlands. Science 228:1424–1427.[Abstract/Free Full Text]
• Richardson, C.J., and P. Vaithiyanathan. 1995. Phosphorus sorption characteristics of Everglades soil along an eutrophication gradient. Soil Sci. Soc. Am. J. 59:1782–1788.[Abstract/Free Full Text]
• Shaffer, P.W., and T.L. Ernst. 1999. Distribution of SOM in freshwater wetlands in the Portland, Oregon area. Wetlands 19:505–516.
• Sheldrick, B.H., and H.L. Wang. 1993. Particle size distribution. p. 499–511. In M.R. Carter (ed.) Soil sampling and methods of analysis. Lewis Publishers, Boca Raton FL.
• Sopher, C.D., and J.D. Baird. 1978. Soils and soil management. Reston Publishing Co., Reston, VA.
• Stauffer, A.L., and R.P. Brooks. 1997. Plant and soil responses to salvaged marsh surface and organic matter amendments at a created wetland in central Pennsylvania. Wetlands 17:90–105.
• Stolt, M.H., M.H. Genthner, W.L. Daniels, V.A. Groover, S. Nagle, and K.C. Haering. 2000. Comparison of soil and other environmental conditions in constructed and adjacent palustrine reference wetlands. Wetlands 20:671–683.
• Verhoeven, J.T.A., D.F. Whigham, R. van Logtestijn, and J. O‘Neill. 2001. A comparative study of nitrogen and phosphorus cycling in tidal and non-tidal riverine wetlands. Wetlands 21:210–222.
• Walbridge, M.R., and J.P. Struthers. 1993. Phosphorus retention in non-tidal palustrine forested wetlands of the Mid-Atlantic Region. Wetlands 13:84–94.
• Whittecar, R.G., and W.L. Daniels. 1999. Use of hydrogeomorphic concepts to design created wetlands in southeastern Virginia. Geomorphology 31:355–371.
• Zedler, J.B., and J.C. Calloway. 1999. Tracking wetland restoration: Do mitigation sites follow desired trajectories. Rest. Ecol. 7:69–73.

This article has been cited by other articles:

				S. Yu and J. G. Ehrenfeld Relationships among plants, soils and microbial communities along a hydrological gradient in the New Jersey Pinelands, USA Ann. Bot., July 30, 2009; (2009) mcp183v1. [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 Web of Science Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (10) Citing Articles via Google Scholar Google Scholar Articles by Bruland, G. L. Articles by Richardson, C. J. Search for Related Content PubMed Articles by Bruland, G. L. Articles by Richardson, C. J. GeoRef GeoRef Citation Agricola Articles by Bruland, G. L. Articles by Richardson, C. J. Related Collections Wetlands and Aquatic Processes Ecosystem Restoration Wetland Soils