JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 23 June 2008
Published in J Environ Qual 37:1634-1643 (2008)
DOI: 10.2134/jeq2007.0116
© 2008 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
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 Similar articles in this journal
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 Google Scholar
Google Scholar
Right arrow Articles by Nahlik, A. M.
Right arrow Articles by Mitsch, W. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Nahlik, A. M.
Right arrow Articles by Mitsch, W. J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Nahlik, A. M.
Right arrow Articles by Mitsch, W. J.
Related Collections
Right arrow Wetlands and Aquatic Processes
Right arrow Surface Water Quality
Right arrow Colloids

TECHNICAL REPORTS

Wetlands and Aquatic Processes

The Effect of River Pulsing on Sedimentation and Nutrients in Created Riparian Wetlands

Amanda M. Nahlik* and William J. Mitsch

Wilma H. Schiermeier Olentangy River Wetland Research Park, Environmental Science Graduate Program and School of Environment and Natural Resources, The Ohio State Univ., 352 W. Dodridge Street, Columbus, OH 43202

* Corresponding author (nahlik.1{at}osu.edu).

Received for publication March 5, 2007.

   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 Materials and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
Sedimentation under pulsed and steady-flow conditions was investigated in two created flow-through riparian wetlands in central Ohio over 2 yr. Hydrologic pulses of river water lasting for 6 to 8 d were imposed on each wetland from January through June during 2004. Mean inflow rates during pulses averaged 52 and 7 cm d–1 between pulses. In 2005, the wetlands received a steady-flow regime of 11 cm d–1 with no major hydrologic fluctuations. Thirty-two sediment traps were deployed and sampled once per month in April, May, June, and July for two consecutive years in each wetland. January through March were not sampled in either year due to frozen water surfaces in the wetlands. Gross sedimentation (sedimentation without normalizing for differences between years) was significantly greater in the pulsing study period (90 kg m–2) than in the steady-flow study period (64 kg m–2). When normalized for different hydrologic and total suspended solid inputs between years, sedimentation for April through July was not significantly different between pulsing and steady-flow study periods. Sedimentation for the 3 mo that received hydrologic pulses (April, May, and June) was significantly lower during pulsing months than in the corresponding steady-flow months. Large fractions of inorganic matter in collected sediments indicated that allochthonous inputs were the main contributor to sedimentation in these wetlands. Organic matter fractions of collected sediments were consistently greater in the steady-flow study period (1.8 g kg–1) than in the pulsed study period (1.5 g kg–1), consistent with greater primary productivity in the water column during steady-flow conditions.

Abbreviations: FPC, flood pulse concept • Is, sedimentation index • TSS, total suspended solids


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
Materials and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
THE landscape position of riparian wetland ecosystems, located between terrestrial and riverine systems, makes them receptive to runoff from terrestrial systems and floodwaters from aquatic systems (Brown, 1985; Mitsch, 1995; Mitsch and Gosselink, 2007), both of which can be sources of large sediment loads. Reduced water velocity of wetlands promotes sedimentation, one of the fundamental biogeochemical processes in wetlands (Harter and Mitsch, 2003; Mitsch and Gosselink, 2007). Sedimentation is the process of organic and inorganic particulate matter gravimetrically accumulating as substrate within a system. Sediment sources include the flooding rivers (allochthonous inputs) and the wetlands themselves (autochthonous inputs). Specific autochthonous inputs may include precipitation of calcium carbonates, dead and/or decaying algal mats, phytoplankton, vegetation from the wetland, and resuspension of wetland sediments.

The process of sedimentation contributes significantly to several wetland functions. Because sediments sorb nutrients, other minerals, and contaminants, the process of sedimentation can contribute to nutrient and contaminant removal from the water column. The negative charge of some ions, such as phosphates, makes them particularly susceptible to binding to positively charged ions, such as calcium and iron, and precipitating as sediments. Negative ions can also bind directly to positively charged sites on mineral sediments. Sedimentation also contributes to improved water conditions by increasing water clarity via reduced turbidity. As a result of increased water clarity, more sunlight is able to penetrate the water column, thus potentially increasing aquatic productivity and dissolved oxygen in the water column.

Once sediments have settled as substrate, nutrients and contaminants are stored within the wetland, transformed through nutrient cycling, and retained or made bioavailable to plants. Some wetlands are nutrient sinks and have especially large storage capacities for nitrogen and phosphorus, nutrients that have been attributed to the eutrophication of inland lakes and reservoirs and hypoxia in coastal systems, such as the Gulf of Mexico (Rabalais et al., 1999; Mitsch et al., 2001; Scavia et al., 2003).

The flood pulse concept (FPC) encompasses the idea that connectivity of rivers and floodplains through the lateral movement of water (i.e., floods) directly affects the integrity of the river and riparian systems (Junk et al., 1989). Although the FPC was originally developed using tropical Amazonian systems as a model, the underlying premise of the FPC has been widely applied to temperate riverine systems (Galat et al., 1998; Sparks et al., 1998; Tockner et al., 2000; Delong et al., 2001; Tockner and Stanford, 2002; Toth et al., 2002). Riparian wetlands connected by floodplain to adjacent rivers or streams receive hydrologic pulses during over-bankflow conditions, which generally occur during winter and spring in the Midwestern USA. River overflow into riparian wetlands subsidizes the wetland by supplying fresh sediments, providing an influx of nutrients, and introducing seeds and organisms (Mitsch et al., 1979a,b; Loucks, 1990; Odum et al., 1995; Heimann and Roell, 2000). Likewise, an export of wetland sediments may accompany large flood pulses, providing the river with sediments rich in organic matter and detritus, a critical input to the benthic riverine food chain (Vannote et al., 1980).

Fluctuations in water level and allochthonous inputs provided by a pulsing regime are also essential for several wetland functions, including production, succession, and decomposition (Middleton, 2002). In the Midwestern USA, a natural pulsing regime consisting of flooding in winter and spring and a drier period in summer and autumn allow for effective dispersal, introduction, and germination of seeds into riparian wetlands (Middleton, 2000). Robertson et al. (2001) and Darke and Megonigal (2003) reported that the spring timing of flood pulses significantly increased biomass production and species richness in forested riparian wetlands. Disturbances caused by rapid changes in water levels and velocities that accompany pulses help drive succession by creating open niches and introducing new species into the system (Parsons et al., 2005). The amount of moisture in a wetland can affect decomposition rates such that the wettest areas have the greatest decomposition rates (Brinson, 1977), although some studies have shown that continuous anaerobic conditions do not support microbes responsible for decomposition (Reddy and Patrick, 1975). Brinson et al. (1981) suggest that the most effective hydrologic regime for rapid decomposition rates is a pulsed system with dry periods in between wet periods.

The ability of scientists to quantify sedimentation rates and patterns, particularly during pulsing events, is important for predicting wetland succession and viability, land accretion, and nutrient removal by wetlands. There have been relatively few studies on sedimentation conducted in freshwater wetlands compared with saltwater marshes, with the exception of Mitsch et al. (1979a), Hupp and Morris (1990), Brueske and Barrett (1994), Fennessy et al. (1994), Kleiss (1996), Wardrop and Brooks (1998), Craft and Casey (2000), Liptak (2000), Mann and Wetzel (2000), Sánchez-Carrillo et al. (2001), and Harter and Mitsch (2003); however, none of these studies addresses the effects of hydrologic pulsing on sedimentation. Understanding sedimentation and being able to better predict its effect on short-term and long-term nutrient dynamics will allow for better wetland design and placement so that goals for water quality and habitat creation can be met.

The goal of this project was to determine how hydrologic pulsing in riparian wetlands affects rates and patterns of sedimentation and accompanying nutrient dynamics. It was hypothesized that sedimentation would be greater in pulsed rather than steady-flow conditions due to more rapid sedimentation and subsidized autochthonous inputs (i.e., aquatic productivity). This objective was investigated by the following specific studies in two full-scale (1-ha) experimental wetlands:


   Materials and Methods
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
Results
 Discussion
 Conclusions
 REFERENCES
 
Study Area
Two 1-ha experimental wetlands located at the Wilma H. Schiermeier Olentangy River Wetland Research Park in Columbus, Ohio (40'021°N, 83'017°E) (Fig. 1 ) were used in this study. The experimental wetlands were created in 1993 on floodplain previously farmed by The Ohio State University. The parent soil type at the site is Ross soil, ranging from silt loam to silty clay loam to loam (McLoda et al., 1980). Water is pumped into the wetlands from the adjacent Olentangy River, a fourth-order stream that drains primarily agricultural watershed with a mean annual discharge of 400 to 700 ft3 s–1. The water intake system allows hydrologic control of the system (i.e., water input can be manipulated) and parity between the two wetlands (i.e., the two wetlands always receive the same amount of water). One wetland was planted with 13 species of hydrophytes in 1994, and the other was allowed to colonize naturally (Mitsch et al., 1998, 2005a, 2005b). After 13 yr, the two wetlands have remarkably similar vegetative communities that occur in the same locations. Dominant vegetation in both wetlands includes soft-stem bulrush (Schoenoplectus tabernaemontani), narrow-leaved and wide-leaved cattail (Typha angustifolia and Typha latifolia, respectively), and rice cut grass (Leersia oryzoides).


Figure 1
View larger version (29K):
[in this window]
[in a new window]

  		Fig. 1. Location of 16 sediment trap stations in each experimental wetland at the Olentangy River Wetland Research Park. Deep marsh areas (generally >30 cm deep) are approximated by the circles within the basins. Lines represent boardwalks.

 
Experimental Design
In 2004, the two experimental wetlands (Fig. 1) were subjected to pumped high-flow pulses for the first week of the first 6 mo (January through June), during the typically wet season in the Midwestern USA. Mean inflow rates from the river to the wetland were increased to an average of 52 ± 3 cm d–1 for 6 to 8 d, after which flow rates were reduced to a mean low-flow rate of 7 ± 0.2 cm d–1 (Fig. 2 ). During the last 6 mo of 2004 (July through December), during the typically dry season, a mean moderate-flow rate of 10 ± 0.01 cm d–1 was maintained. In 2005, the wetlands received a steady-flow regime for the duration of the year, where water was pumped into the wetlands at a constant rate of 11 ± 0.01 cm d–1 with no major hydrologic fluctuations (i.e., no pulses). To reduce experimental variability, the total hydrologic input into the wetlands in 2005 was designed to be equivalent to that in 2004. Hydrologic loading rates take into account the area of the wetland and are reported as m3 m–2 d–1 or m d–1. Sediment influxes originated from allochthonous and autochthonous sources and were not controlled. Neither hydrology (Fig. 2) nor sedimentation patterns were significantly different between the two experimental wetlands; therefore, the wetlands were treated as replicates throughout this study.


Figure 2
View larger version (32K):
[in this window]
[in a new window]

  		Fig. 2. Total surface water inflow (cm d–1) in experimental Wetland 1 and Wetland 2 for 2004 pulsed and 2005 steady-flow years. Shaded boxes indicate periods for this study. Monthly pulses in 2004 are designated with arrows.

 
Sediment Traps
To sample deposition of suspended sediments (and coarser material traveling in suspension) as wetland substrate, sediment traps were constructed using a modified design from Fennessy et al. (1994). Two replicate 500-mL wide-mouth Nalgene bottles attached to each other by 0.5-cm rubber bands and wire were secured to a metal stake driven deep into wetland sediments. The two bottles used for each trap had a height/diameter ratio >3 to minimize resuspension of trapped sediments (Hakanson and Jansson, 1983). Bottles were sunk vertically into the substrate so that the opening was approximately 5 cm above the soil surface; this biased for new or resuspended sediments falling from the water column, as opposed to older sediments that rolled or slid across the soil surface via bed-material transport (Gardner, 1980) and allowed traps to function during low flows. To minimize trapping sediments disturbed during placement, sediment traps were pre-filled with water and capped before deployment. After fixing the traps to the stakes, a minimum of 30 min was allowed for any disturbed sediments to settle before lids were removed from the bottles.

Sediment traps were deployed and retrieved on the same schedule—once per month in April, May, June, and July of 2004 and 2005 (our pulsed and steady-flow study periods, respectively)—at 16 sites in each of the two wetlands (Fig. 1). January, February, and March were not sampled in either year due to frozen water surfaces in the wetlands. Despite the fact that July did not experience a pulse in 2004, we sampled July in both years to compare sedimentation rates in steady-flow conditions after flood pulses (2004) with sedimentation rates during long-term, continuous, steady-flow conditions (2005). Because previous studies have shown that most sediments settle near the inflow (Hakanson and Jansson, 1983; Brueske and Barrett, 1994; Fennessy et al., 1994; Kleiss, 1996; Sánchez-Carrillo et al., 2001), 10 sites were clustered near the inflow of the wetland, and the remaining six were spaced at wider intervals toward the outflow. Traps were deployed in deep and shallow areas of each wetland (stations 1–7, 11 and 12, and 15 and 16 were placed in deep areas, whereas stations 5, 8, 9 and 10, and 13 and 14 were placed in shallow areas of the wetlands). Vegetation communities were similar in both wetlands and remained the same in sediment trap locations between years. Traps were capped to minimize sediment loss immediately before retrieval, labeled, and stored at 4°C until analysis.

Laboratory Analyses
After sample collection, sediments were allowed to re-settle for 2 to 14 d in bottles after transport from the field to the lab. Supernatant was poured with caution to avoid disturbing the sediments, and sediments were poured into pre-weighed aluminum drying pans. Any remaining sediments in the bottles were rinsed into the drying pans using deionized water. Sediments were dried in a drying oven at 60 to 65°C until sediments reached constant weight, cooled in a desiccator for a minimum of 12 h, and weighed with a top-loading balance (Mettler AE200; Metteler, Columbus, OH) with 0.0001 g sensitivity. Because standard methods for total suspended solids call for drying temperature of 103 to 105°C (APHA, AWWA, and WEF, 1998) and collected sediments were dried at 60 to 65°C to preserve the integrity of nutrients (Comin et al., 1997; Verhoeven et al., 2001), subsamples of collected sediments were dried at 103 to 105°C. Sediment weights dried at 60 to 65°C were normalized for the difference and reported as dried at 103 to 105°C.

Dried samples were ground to pass through a 2-mm sieve and stored in airtight plastic bags for organic and nutrient analyses. Loss on ignition was used to estimate organic content. Subsamples of approximately 5 g were placed into porcelain crucibles, weighed, combusted at 550°C for approximately 3 h in a Fisher Scientific Isotemp Programmable muffle furnace (650–750 Series) (Davies, 1974; Forster, 1995), cooled to room temperature in a desiccator, and weighed to determine combusted weight.

Nutrient analyses were performed for sediment samples from selected sites (Sites 6, 7, 11, 12, 15, and 16) of each wetland for each sampling month in 2004 and 2005 by the Ohio Agricultural Research and Development Center Service Testing and Research Laboratory at The Ohio State University Wooster Campus. Soils were analyzed for total N using the Dumas Method combustion technique (AOAC International, 2002) and for P and Ca by inductively coupled plasma emission spectrometry after 3051 Microwave Digestion (USEPA, 1994).

To determine the flux of suspended sediments in and out of the wetland, water samples were collected from the inflow and outflows of the wetlands in the morning and evening throughout this study and from the river weekly, stored at 4°C, and analyzed for turbidity with a Hach 2100N Turbidimeter (Hach, Loveland, CO) using nephelmetric methods (APHA, AWWA, and WEF, 1998). Total suspended solids (TSS), used to determine sediment retention within the study wetlands, were estimated using the turbidity measurements and a regression equation reported in Harter and Mitsch (2003) for the same river and wetlands.

Data Analysis
The two bottles in each sediment trap were analyzed separately in the lab, but means for the two bottles were used for statistical analyses because the two bottles functioned as replicates. Sediment accumulation data were determined by applying the sum of total captured sediment at each site by the total area of the bottles to the entire wetland area for each month. Minitab 14 was used to conduct ANOVA, Fishers pair-wise comparisons, t tests, and Pearson correlations. ANOVA tests were used to compare temporal and spatial sedimentation. Fishers pair-wise comparisons were conducted at 95% confidence interval to determine differences between months, and t tests were used to compare July 2004 (pulsed study period) and July 2005 (steady-flow study period) with other monthly data. Significant differences are reported at p ≤ 0.05, and sample variance is expressed as mean ± SE.


   Results
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
Discussion
 Conclusions
 REFERENCES
 
Sedimentation Rates
Sediments were collected monthly over a period of 126 d in both 2004 and 2005. Total gross sedimentation (sedimentation without accounting for differences in hydrologic and TSS inputs between years) for April, May, June, July was significantly greater during the pulsing study period than during the steady-flow study period (p = 0.032; 90 kg m–2 and 64 kg m–2 in pulsing and steady-flow study periods, respectively). When taking into account only the months in which pulsing occurred (April, May, June), total gross sedimentation was not different between the pulsed months in 2004 (45 kg m–2) and corresponding steady-flow months in 2005 (39 kg m–2). There was greater total gross sedimentation during the pulsed study period than the steady-flow study period because of significantly greater sedimentation rates in July 2004 (1 mo after the flood pulses) than in July 2005 (p = 0.011) (Fig. 3a ).


Figure 3
View larger version (29K):
[in this window]
[in a new window]

  		Fig. 3. Sedimentation rates (a) separated by organic matter sedimentation (white striped bars, bottom) and inorganic material sedimentation (gray striped bars, top) and (b) reported as monthly sedimentation index for inorganic and organic matter combined in pulsed and steady-flow years. Total sedimentation is reported as g dry wt m–2 d–1, and the sedimentation index is unitless. Means are reported with SE. Number of samples for the pulsed study period are 32, 32, 31, and 29 for each month, respectively. Number of samples for the steady-flow study period are 32 for each month. Different letters signify significant differences at p ≤ 0.05.

 
Sedimentation was dominated by inorganic (mineral) material throughout the study (Fig. 3a). Inorganic gross sedimentation did not statistically change between April and May of the pulsed and steady-flow years. It significantly increased from 350 to 712 g dry wt m–2 d–1 in May to June in the pulsed year (ppulsed < 0.001) and from 271 to 614 g dry wt m–2 d–1 in May to June in the steady-flow year (psteady flow < 0.001). Inorganic gross sedimentation increased significantly again by 420 g dry wt m–2 d–1 from June to July in the pulsing year (p = 0.020) but did not change significantly in the steady-flow year. Inorganic gross sedimentation in July was significantly different between pulsing and steady-flow years (p = 0.006) despite similar hydrology in both years for that month. Organic matter gross sedimentation comprised a small fraction of the total gross sedimentation throughout the study and significantly increased in both years in June and July (Fig. 3a).

Normalizing Sedimentation
Total surface inflow (averaged between the wetlands) for the study period (April through July) was 20 m in the pulsed year and 13 m in the steady-flow year, with a significant difference between the 2 yr (p = 0.011). Statistical analyses also revealed significant differences between years for TSS in the river (14 ± 1 and 10 ± 1 mg L–1 for 2004 and 2005, respectively; p = 0.049) (Table 1 ) and the wetland inflows (11 ± 1 and 7 ± 1 mg L–1 for 2004 and 2005, respectively; p = 0.001). To correct for the differences in TSS and hydrologic input between the 2004 and 2005 yr at the inflow, a normalized sedimentation index (Is) was calculated for each month (Fig. 3b) as

Formula
where s = uncorrected sedimentation rate (g m–2 mo–1), q = total water input into the wetland (m mo–1), and Ci = the mean concentration of TSS imported into the wetland (mg L–1 = g m–3). The sedimentation index is unitless and provides a comparison between years that normalizes for different flows and river sediment conditions.


View this table:
[in this window]
[in a new window]

  		Table 1. Mean total suspended solids (mg L–1) estimated from turbidity analyses in the Olentangy River and inflow to the experimental wetlands (always measured at Wetland 1) during both years of experimental months. Monthly averages are reported with SE. Number of samples are in parentheses.

 
Normalized Is showed different patterns than did the uncorrected sedimentation rates. Total Is was not significantly different between study periods (April, May, June, July) in the pulsed year and the steady-flow year; however, Is for each of the first 3 mo—April, May, June—was significantly lower in pulsing months than in the corresponding steady-flow months (pApril = 0.002; pMay = 0.008; pJune = 0.010). There was an eightfold increase in Is from June to July in the pulsing year (p < 0.001); however, Is decreased in the steady-flow year for the same period after steadily increasing over the previous 3 mo (Fig. 3b).

Spatial Patterns
Sedimentation patterns varied with distance from the inflow, with similar patterns in the pulsed and steady-flow years (Fig. 4 ). Significant differences between the pulsed and steady-flow study periods occurred at 110 and 140 m from the inflow (p110 = 0.047; p140 = 0.008). Sedimentation rates above 600 g dry wt m–2 d–1 occurred at 5, 45, and 140 m from the inflow during the pulsed study period, whereas peaks during the steady-flow study period were present only at 45 and 110 m from the inflow. The wetland cross-section shown in Fig. 4 suggests that sedimentation peaks at 45 and 110 m may coincide with shallow, normally vegetated areas within the wetlands, whereas lower sedimentation rates generally correspond to the deep, open water basins.


Figure 4
View larger version (27K):
[in this window]
[in a new window]

  		Fig. 4. Sedimentation rates (g dry wt m–2 d–1) as a function of distance from the inflow (m) in pulsed and steady-flow years. Means are reported with SE. *Significant differences between pulsed and steady-flow years at p ≤ 0.05. Cross-section of the wetlands approximates the basin morphology and ecology, including basin depth and presence of vegetation, at corresponding distances.

 
Organic and Nutrient Concentrations
Organic matter associated with collected sediments was consistently lower in the pulsed year than in the steady-flow year (1.5 and 1.8 g kg–1, respectively; p < 0.001). June was the only month that was not significantly different between years. This pattern is consistent with results described by Tuttle et al. (2008), who found greater aquatic productivity in the steady-flow than the pulsed year of the same wetlands used in this study.

Total nitrogen concentrations of sediments varied from month to month between hydrologic treatments and showed no consistent temporal pattern between pulsed and steady-flow years. Significant differences in total nitrogen content of sediments were not present between years in April; however, differences in May, June, and July were evident. Sediment nitrogen concentrations were greater in pulsed May than in steady-flow May (p = 0.055), whereas sediments exhibited greater nitrogen concentrations in steady-flow months of June and July than the same months in the pulsed year (pJune = 0.045; pJuly = 0.007).

Neither hydrologic regime nor month had a significant effect on phosphorus content of sediments. Monthly phosphorus concentrations were consistent within the range of 0.98 to 1.07 g kg–1. Although there were no significant differences in phosphorus contents, they tended to follow the same temporal pattern as total nitrogen content of sediments.

Calcium content of captured sediment was greater overall in the steady-flow year than in the pulsed year, and significant differences between years occurred in June and July (pJune < 0.001; pJuly = 0.002) (Fig. 5 ). Calcium content increased between the pulsed year and the steady-flow year from 29.7 to 103.1 g kg–1 in June and 43.2 to 80.2 g kg–1 in July. The pulsed year showed little variation in calcium content from month to month; only June had significantly lower calcium content than the other months with a mean of 29.7 g kg–1 (p = 0.009). There were no significant differences between calcium content of sediments from April and May of the steady-flow year. June and July of the steady-flow year had significantly greater calcium concentrations than earlier months (p < 0.001). Calcium concentrations in the steady-flow year nearly doubled from May to June and remained high in July.


Figure 5
View larger version (13K):
[in this window]
[in a new window]

  		Fig. 5. Mean concentrations of calcium (g kg–1) of trapped sediments in pulsed and steady-flow years. Means are reported with SE. Number of samples are 9, 11, 11, and 9 in the pulsed study period and 12, 12, 10, and 12 in the steady-flow study period for April, May, June, and July, respectively. Different letters signify significant differences at p ≤ 0.05.

 

   Discussion
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion
Conclusions
 REFERENCES
 
Effects of Pulsing on Sedimentation
Using gross accumulation data, an estimated 900 ± 15 t ha–1 of sediments were deposited during the pulsed study period, and an estimated 640 ± 10 t ha–1 were deposited during the steady-flow study period in the experimental wetlands. The difference in sedimentation rates and estimated sediment deposition between study periods may be partially attributed to differences in hydraulic loading and sediment import between years. When gross sedimentation rates were normalized for differences in flow and total suspended solids, the same amounts of total sediments were deposited during pulsed and steady-flow study periods. Normalized sedimentation was greater in 3 mo (April, May, June) in the steady-flow year compared with the pulsed year. However, normalized sedimentation rates in July were greater in the pulsed year than the steady-flow year, offsetting the differences in the previous 3 mo (Fig. 3b). It seems that sedimentation as a result of flood pulsing into wetlands can lag far after the actual hydrology pulse has disappeared.

Allochthonous versus Autochthonous Sources
Inorganic (mineral) sediments were 82 to 85% of the collected sediments on a dry-weight basis in this study. Mean inorganic gross sedimentation rates were greater in the pulsed year than those in the steady-flow year, with some of this difference due to the estimated 40% greater river sediment load during the pulsed year. Months that experienced long periods of steady flow after pulsing, such as 2004 July, had significantly greater inorganic sedimentation rates than the steady-flow year according to gross and normalized sedimentation data. Although not statistically significant, organic gross sedimentation rates tended to be lower in the pulsed year than in the steady-flow year. The current study showed that sedimentation in riparian wetlands is dominated by inorganic material that likely originates from allochthonous inputs, regardless of hydrologic regime.

Organic sedimentation was greater in the steady-flow year than the pulsed year, with significant differences in June and July. The absence of river pulses and lower total suspended solid content of inflow water during the steady-flow year may have resulted in lower amounts of introduced inorganic sediments, thus increasing the percentage of organic/inorganic sediments. In addition, Tuttle et al. (2008) reported greater water column productivity (combination of phytoplankton, submersed aquatic plants, and benthic algae) in the same wetlands during the steady-flow year, suggesting that greater organic matter in the sediments may be linked to autochthonous sources. Significantly greater organic matter content during the warm months of June and July in the steady-flow year than in the pulsed year supports the premise that hydrologic-pulsed wetlands may accumulate less organic carbon, even though pulsing may lead to greater overall productivity (Odum et al., 1995).

Effect of Macrophytes and Water Depth
Emergent macrophytes seemed to positively affect sedimentation rates in this study, with the greatest sedimentation rates occurring in shallow, vegetated areas. These results are consistent with sedimentation studies done several years prior in the same experimental wetland basins by Harter and Mitsch (2003), who found greatest sedimentation rates in shallow areas even though sedimentation was measured by horizon markers rather than sedimentation traps. Darke and Megonigal (2003) found that sedimentation increased as the growing season progressed in some tidal fresh water wetlands, and Peterson and Teal (1996) reported similar trends in the marsh component of a septage treatment system, most likely due to more trapping of sediments by emergent vegetation. An indirect contribution of vegetation to sedimentation appears as increased organic (autochthonous) sedimentation rates during the growing season as a result of resuspension and biodegradation of the previous year's vegetation as new vegetation begins to emerge. The findings of greater sedimentation rates in shallow vegetated water have significance for the Louisiana Delta where vegetated wetlands are being lost at an alarming rate because of land subsidence (Day et al., 2007). Sedimentation, minimized by river management, does not compensate, and as the water gets deeper, the situation spirals out of control as sedimentation becomes less in deeper water.

Sedimentation and Nutrient Dynamics
The Redfield (1958) ratio, which describes a carbon/nitrogen/phosphorus (C/N/P) ratio in a balanced system at 41C/7.2N/1P (by weight), is often used to calculate excess or limiting nutrients within ecosystems. Carbon is generally not limiting in wetland systems; this is evident in the ratios calculated from nutrient data: Mean ratios for sediments were 77C/5N/1P for the steady-flow study period and 93C/5N/1P in the pulsed study period. The N/P ratio is more important. Mean N/P ratios calculated from nutrient data on collected sediments were 5:1, with no significant differences between the pulsed and steady-flow years (with a range of 3:1 to 7:1 in individual samples). These N/P ratios suggest that nitrogen, rather than phosphorus, may be the limiting nutrient in the experimental wetlands. In an experiment conducted in 1996, Svengsouk and Mitsch (2001) reported nitrogen-limited conditions in the same experimental wetlands used in this study.

Organic matter plays an important part in nitrogen cycling because many forms of nitrogen (e.g., ammonium) are bound to organic matter until decomposition or uptake by vegetation (Mitsch and Gosselink, 2007); thus, nitrogen concentrations are often correlated with organic matter concentrations in sediments. In this study, there is a weak positive relationship between the concentrations of organic matter and nitrogen in collected sediments (r = 0.354; p = 0.001) (Fig. 6a ). Because rates of organic accumulation are low (total sediment accumulation was 14 ± 2 kg m–2 in the pulsed year and 11 ± 2 kg m–2 in the steady-flow year), nitrogen associated with these types of sediments are accumulating in the wetland substrate at low rates. It is also possible that nitrogen in sediments is absorbed by vegetation or quickly transformed, resulting in very short residence times as sediment-bound nitrogen.


Figure 6
View larger version (19K):
[in this window]
[in a new window]

  		Fig. 6. (a) Total nitrogen (g kg–1) as a function of organic matter (g kg–1) and (b) phosphorus concentrations (g kg–1) as a function of inorganic material (g kg–1) in collected sediments. p values were determined using a Pearson correlation.

 
Phosphorus concentrations were expected to increase with the concentrations of inorganic material that comprised sediments because positively charged clay particles, such as inorganic material, have a high binding potential to phosphorus (Mitsch and Gosselink, 2007). Regression analyses did not support this hypothesis, showing that there was a significant inverse relationship between inorganic material and phosphorus concentrations (r = –0.297; p = 0.006) (Fig. 6b). This may be because many of the trapped sediments were from allochthonous sources, which contributed sediment particles that were relatively devoid of nutrients.

The amount of nutrient sedimentation measured is considerably large in the experimental wetlands, with means of 834 ± 190 and 441 ± 130 g N m–2 yr–1 of nitrogen deposited in the pulsed and steady-flow study periods, respectively. Although these large numbers may be in part due to resuspension, nitrogen loading into wetlands at this rate surpasses the sustainable range of 10 to 40 g N m–2 yr–1 suggested by Mitsch et al. (2000) for freshwater wetlands. Phosphorus sedimentation rates were 200 ± 40 and 100 ± 40 g P m–2 yr–1 in the pulsed and steady-flow study periods, respectively. These sedimentation rates for phosphorus also exceed the range of 0.5 to 5 g P m–2 yr–1 of designated sustainable levels (Mitsch et al., 2000). For the nitrogen and phosphorus sedimentation rates given here, much of this nutrient sedimentation is due to resuspension, and the net rate of nutrient sedimentation is much less.

Calcium Precipitation
Under high temperatures and high pH conditions that are typical in many standing-water wetlands, calcium carbonate precipitation occurs (Liptak, 2000). Algae (benthic and planktonic) play a large role in calcium precipitation directly by increasing pH through removal of CO2 during photosynthesis and indirectly by incorporating calcium carbonate into their structure, which eventually (i.e., in death) contribute to sedimentation rates (Liptak, 2000). Positive correlation of gross primary productivity of the water column (data from Tuttle, 2005) and rates of calcium sedimentation measured in our study during the same study periods supports the argument that calcium precipitation is a biologically driven and important function in these wetlands (r = 0.760; p = 0.047) (Fig. 7 ). Mean calcium sedimentation rates in this study ranged from 12 to 39 g Ca m–2 d–1 during the pulsed year to 6 to 29 g Ca m–2 d–1 during the steady-flow year. Liptak (2000) reported that mean net calcium carbonate deposition in the experimental wetlands at the Olentangy River Wetland Research Park occurred at a rate of 0.48 g Ca m–2 d–1, a rate far lower than the gross rate of calcium sedimentation reported here. This difference is due to the resuspension that is measured by our sedimentation traps.


Figure 7
View larger version (13K):
[in this window]
[in a new window]

  		Fig. 7. Calcium sedimentation (g Ca m–2 d–1) as a function of gross primary productivity in the water column (kcal m–2 d–1). Error bars represent SE. p values were determined using a Pearson correlation. Gross primary productivity data provided by Tuttle (2005).

 
Comparison to Other Wetland Sedimentation Studies
Sedimentation is affected by many outside factors, such as flow rates, water source, vegetative cover, wind speeds, and animal activity. For this reason, comparing sedimentation rates between wetlands is difficult, and reported sedimentation rates vary greatly in the literature. In a comparison with other sedimentation studies, it was expected that sedimentation rates would increase logarithmically with increased hydrologic loading rates into the wetland. Sedimentation rates determined in this study were greater than other freshwater studies examined; the greater hydrologic loading rates in this study probably account for some, but not all, of these greater rates (Fig. 8 ). Also, although the bottle trap method is a good repeatable indicator of relative sedimentation rates, it has been reported to overestimate sediment retention (Kleiss, 1996; Steiger et al., 2003). Sedimentation rates measured by our bottle trap method were influenced by resuspension. In a previous sedimentation study conducted in the same wetlands, Harter and Mitsch (2003) estimated that approximately 36 kg m–2 yr–1 of sediments were resuspended due to wind, ice, and animal activity, whereas net sediment input and output were an order of magnitude lower. Substantial amounts of resuspension most likely contributed to the large sedimentation rates calculated from this study.


Figure 8
View larger version (15K):
[in this window]
[in a new window]

  		Fig. 8. Sedimentation rates (g dry wt m–2 d–1) as a function of loading rate (cm d–1) for several studies. Means are reported with SE when possible. Regression line is for wetland studies (Fennessy et al., 1994; this study) using the bottle trap method. Unfilled diamonds are used for studies (Mitsch, 1979; Sánchez-Carrillo et al., 2001; Harter and Mitsch, 2003) that used other methods of estimating sedimentation rates or studies focused on wetlands other than riparian wetlands.

 
Study Limitations
The bottle trap method in this study was necessary to capture a sufficient amount of sediment for lab analysis and to provide an easily repeatable technique. It was equally as important for the traps to set for the full month because, even after a month, some traps did not catch enough sediment for full characterization analysis. In the future, the combination of bottle traps to catch sediments for analysis, horizon markers to more accurately estimate net sedimentation rates, particle size analyses to determine spatial distribution, and stable isotope characterization to determine the source of sediments would help describe sedimentation pathways in wetlands.

Simulated flood pulses in this study did not necessarily exemplify sediment load characteristics of actual flood pulses because the simulated flood pulses did not occur during natural flooding in the river. Flooding events generally result in bank erosion and greater suspended material loads (Knighton, 1998); therefore, river floodwaters naturally deposit more sediments into the riparian wetlands to which they are connected (Mitsch et al., 1979a). Due to the experimental design of this study, simulated flooding often occurred outside natural flood conditions in the river; therefore, sedimentation rates, especially those for allochthonous inputs, could have been lower than they would be during a natural pulse.


   Conclusions
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion
 Conclusions
REFERENCES
 
When gross sedimentation rates were normalized for differences between years, sedimentation rates were similar in pulsing and steady-flow years; however, short-term (monthly) patterns of sedimentation were different between pulsing and steady-flow years, regardless of whether rates were normalized. We found evidence that allochthonous inputs accounted for the majority of introduced sediments in these riparian wetlands, with autochthonous inputs becoming more important in the growing season and in steady-flow conditions. Pulsing did not have an effect on nitrogen and phosphorus sedimentation in this study; however, calcium sedimentation was greater during the steady-flow year, perhaps due to greater primary productivity in the wetlands. Overall, the pulsing hydrology influenced sedimentation directly because of its physics and indirectly because of its effects on other ecosystem functions.


   ACKNOWLEDGMENTS
 
We acknowledge Dr. M. Siobhan Fennessy and Dr. Virigine Bouchard for their insight and contributions to the manuscript. We thank Anne Altor, Chris Anderson, Jake Elder, Dan Fink, Jim Hamski, Maria Hernandez, Sarah Hunt, Jeremiah Miller, and Cassie Tuttle for their invaluable help in the field and lab. Dr. Li Zhang assisted with associated data gathering throughout this study. This project was supported by an OARDC Payne Grant and a USDA NRI CSREES Award (2003-35102-13518). Olentangy River Wetland Research Park publication number 2008-003.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 Materials and Methods
 Results
 Discussion
 Conclusions
 REFERENCES
 





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 Similar articles in this journal
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 Google Scholar
Google Scholar
Right arrow Articles by Nahlik, A. M.
Right arrow Articles by Mitsch, W. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Nahlik, A. M.
Right arrow Articles by Mitsch, W. J.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by Nahlik, A. M.
Right arrow Articles by Mitsch, W. J.
Related Collections
Right arrow Wetlands and Aquatic Processes
Right arrow Surface Water Quality
Right arrow Colloids

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