Soil Science Society of America Journal 67:961-969 (2003)
© 2003 Soil Science Society of America
DIVISION S-10—WETLAND SOILS
A Method to Predict Soil Saturation Frequency and Duration from Soil Color
X. Hea,
M. J. Vepraskas*,a,
D. L. Lindboa and
R. W. Skaggsb
a Dep. of Soil Science, Box 7619, North Carolina State University, Raleigh, NC 2795
b Dep. of Biological and Agricultural Engineering, Box 7625, North Carolina State University, Raleigh, NC 27695
* Corresponding author (Michael_Vepraskas{at}NCSU.edu)
 |
ABSTRACT
|
|---|
Saturation frequency and duration must be estimated to determine if a site is a jurisdictional wetland, and such data also aid in assessing sites for on-site waste disposal. This study developed a method to estimate saturation frequency and duration by calibrating redoximorphic features to a 40-yr record of water table simulations in a catena of Atlantic Coastal Plain soils in North Carolina. Thirteen plots were established along a toposequence with moderately well-drained (Aquic Paleudults) and very poorly drained soils (Umbric Paleaquults) as end members. A hydrologic model (DRAINMOD) was calibrated for each plot. Redox potential measurements showed that an average of 21 consecutive days of continuous saturation was sufficient for Fe reduction to occur in the soils. Historic rainfall data were used in the DRAINMOD model to estimate the number of times each plot was saturated for 21 consecutive days or longer in each year of a 40-yr period. Redoximorphic features were significantly correlated with average number of saturation events computed to have occurred at depths of 45, 60, 75, and 90 cm across all soils. Relationships were linear and varied by depth when all soils were analyzed as a single population. The r2 values for relationships between redox depletions and saturation events were >0.85 for saturation occurring during the growing season, and were >0.75 for saturation events occurring at any time during the year. These relationships allow prediction of the likelihood that a soil will saturate for 
21 d by simply estimating the percentage of redoximorphic features at a given depth.
Abbreviations: NSE, number of saturation events
|
INTRODUCTION
|
|---|
SOIL COLOR PATTERNS that include low chroma or gray colors are commonly used to predict where seasonal saturation occurs in soils (Daniels et al., 1971; Bouma, 1983; Pickering and Veneman, 1984; Buol and Rebertus, 1988; Veneman et al., 1998). Low chroma colors increase in abundance the longer a soil is saturated and chemically reduced (Vepraskas, 1999). Daniels et al. (1971) used a statistical model to compute water table levels in a toposequence of Paleudults and Paleaquults and then estimated water table levels during the course of an average year. Data were collected for a 2-yr period. They noted that as soils became saturated for longer periods the chroma of the Bt horizons decreased. Horizons containing colors with chromas of 2 or less and values of 4 or more were saturated from 10 to 50% of the year. Simonson and Boersma (1972) related faint and distinct mottling to average durations of saturation that were estimated for a 29-yr period. The studies of Daniels et al. (1971) and Simonson and Boersma (1972) were original, but they did not define a simple relationship between saturation frequency and redoximorphic feature abundance. As a result, it has been difficult to extrapolate their saturation data to other sites to make specific interpretations regarding saturation frequency or duration. Later work has attempted to define the relationship between saturation duration and occurrence of redoximorphic features, but because these studies were short-term (<3yr), the effects of saturation frequency on morphology could not be addressed (Franzmeier et al., 1983; Megonigal et al., 1993; and Jacobs et al., 2002).
Some land-use regulations require that frequency and duration of saturated conditions be determined to assess a soil's suitability for a given use. For example, jurisdictional wetlands must meet a standard for wetland hydrology, which requires an area be saturated to the surface for at least 5% of the growing season with a frequency of at least 5 yr out of 10 (Environmental Laboratory, 1987). On-site wetland inspections to determine whether a site meets the wetland hydrology requirement must be completed quickly to keep up with the demand. There is currently no technology known to the authors that would enable a field scientist to determine whether a site affected by shallow ground water actually meets the saturation frequency and duration requirements for wetland hydrology during a single site visit.
We hypothesized that rapid assessment of the frequency of soil saturation for specific durations may be accomplished if the soil color patterns (redoximorphic features) were correlated with these hydrologic parameters. Testing this hypothesis requires that a hydrologic model first be calibrated on-site to simulate water table levels using rainfall as the major input variable. Once the model is calibrated for a specific soil, long-term (e.g., 40 yr) rainfall data could then be read into the model to compute saturation duration and frequency for each depth in the soil. By correlating the historic data on saturation with the percentages of redoximorphic features in the soil, it should be possible to calibrate the percentages of redoximorphic features to specific saturation durations and frequencies. If the statistical analyses are successful, then we might be able to predict, for example, that a soil having 40% redox depletions at a depth of 60 cm is saturated for at least 21 d in 8 out of 10 yr. Once the redoximorphic features at one soil have been calibrated in this way, they then can be used to predict saturation frequency and duration in similar soils where water table levels have not been measured. Vepraskas (2000) reported on a pilot study where this approach was successfully applied to three soil plots.
This study investigated the relationship between redoximorphic feature percentage and saturation frequency and duration for a toposequence of soils. The specific objectives were: (i) to compute a 40-yr record of water table fluctuations for the soils using a hydrologic model; (ii) to compute saturation-event index values that summarize the saturation frequency and duration data for each soil in the toposequence; and (iii) to correlate percentages of redox depletions and concentrations with the saturation event values.
|
MATERIALS AND METHODS
|
|---|
Experimental Sites
Research was conducted in the North Carolina Coastal Plain in a toposequence of soils that ranged from moderately well drained to very poorly drained. The site was located in Pitt County, NC, approximately 5.1 km southwest of Greenville at N 35°34'10'' lat. and W 77°26'26'' long.. A schematic map showing the experimental area is given in Fig. 1
. A ditch was present along the western and northern perimeter of the site. The ditch ranged from 1 to 2 m wide and was 0.6 to 1.5 m deep. The soil toposequence at the site consisted of soils in the following series: Goldsboro (G) (fine-loamy, siliceous, subactive, thermic Aquic Paleudults), Lynchburg (L) (fine-loamy, siliceous, semiactive, thermic Aeric Paleaquults), Rains (R) (fine-loamy, siliceous, semiactive, thermic Typic Paleaquults), and Pantego (P) (fine-loamy, siliceous, semiactive, thermic Umbric Paleaquult). Experiments were conducted along four transects which contained a total of 13 plots. These transects were established at increasing distances from the ditch. Each transect consisted of three or four plots along the soil toposequence that were placed to include Goldsboro (moderately well drained), Lynchburg (somewhat poorly drained), and Rains (poorly drained) soils in each transect. Transect 4 contained an additional plot in the (very poorly drained) Pantego soil. These plots were labeled as: 1G, 1L, 1R ... 4G, 4L, 4R, and 4P to represent transect and series name. The slope at the site was 2%. Vegetation consisted of loblolly pine (Pinus taeda L.), red maple (Acer rubrum L.), and white oak (Quercus alba L.) etc. Most of trees were between 10 and 50 yr old. Hayes and Vepraskas (2000) previously reported additional data on soil morphology at the site, and also discussed how the ditch affected soil morphology at this site.
Water table levels were monitored daily to depths of 2 m at 2400 h (midnight) in each of the 13 plots using RDS automatic monitoring wells (Remote Data Systems, Inc., Wilmington, NC). The water table data were collected from November 1996 until March 1999 and were used to calibrate a hydrologic model. Wells were installed by boring a 10-cm diam. hole to a depth of 2.25 m, inserting the well, and filling in the space between the well screen and soil with sand. A 3-cm thick layer of dry bentonite pellets was then placed on the top of the sand to seal the well from surface water inflow. A conical mound of soil was placed over the bentonite to direct surface water away from the well. To ensure that the recording wells were monitoring daily water levels accurately, a manual check well was installed at each plot to a depth of approximately 127 cm below the mineral soil surface. Every 2 to 3 wk the check wells were measured to compare with the water table data from the recording wells. Rainfall was also measured daily at each site using recording gauges (Onset Computer Corp., Bourne, MA). Additional details regarding hydrologic monitoring were reported by Hayes and Vepraskas (2000) and He et al. (2002).
Soil profiles were described from pits (1.5 by 1.5 by 1.0 m) placed in each plot at the site (Fig. 1). Each pit was approximately 10 m from the water monitoring instruments to avoid interference with a plot's hydrology. The soils were described to approximately 100 cm below the soil surface with some data having been reported earlier by Hayes and Vepraskas (2000). Soil morphological features were estimated using standard field methods (Schoeneberger et al., 1998). The abundance and size of redoximorphic features were estimated visually by comparing the features with proportion charts. These visual estimates were made independently by the same two individuals in all pits. The results from both individuals were averaged for each depth in each profile.
Redox Potentials and Soil pH
Five Pt electrodes were inserted at depths of 15, 30, and 60 cm in each plot to measure oxidation-reduction (redox) potential. Hayes (1998) described construction of these electrodes. Field voltage was measured weekly at each depth in each plot using an Accumet 1002 pH/mV meter from the Fisher Scientific Co. (Norcross, GA) and a Ag/AgCl, saturated KCl reference electrode from Jensen Instruments (Tacoma, WA). Salt bridges were used to connect the reference electrode to the soil (Pickering and Veneman, 1983). All the field readings were converted to EH values (true redox potentials) by adding a temperature-dependent conversion factor to the voltages measured in the field. In the summer of every year when the redox potentials were very high all the electrodes were pulled from the field. The Pt tips were cleaned with steel wool, and the electrodes were checked for accuracy with a redox buffer and water, and working electrodes were reinstalled. New electrodes were inserted into plots when necessary to maintain five replicated electrodes at each depth. The EH measurements were made weekly for approximately 2.5 yr.
Soil pH values at depths of 15, 30, and 60 cm were also measured on soil slurries (1:1 soil/water ratio) in each plot several times from 1998 to 1999. Soil pH was determined using an Accumet 1002 pH/mV meter and a glass pH electrode that were calibrated on-site.
Soil Temperature
Soil temperature was measured weekly with thermocouples at depths of 15, 30, and 60 cm in each plot. A separate thermocouple was used for each depth. One end of the thermocouple was placed at the desired depth and the other end was extended above the soil surface for the convenience of reading temperature. The thermocouples were made using ANSI Type TX Extension Grade Copper/Constantan 20 gauge thermocouple wire with polyvinyl insulation (Omega Engineering, Inc., Stamford, CT). The soil temperatures were read with an Omega Microcomputer Thermometer Model HH-71 T (Omega Engineering, Inc., Stamford, CT).
Model Calibration and Simulation
DRAINMOD is a hydrologic model that simulates water table levels in a soil plot over time (Skaggs, 1978). Input data for the simulations included precipitation, evapotranspiration, infiltration, runoff, and subsurface drainage. DRAINMOD computed a water balance on a soil pedon of unit cross-sectional area on a day-by-day, hour-by-hour basis. From this water balance the depth to the water table was computed and duration of saturation for any depth was determined.
The DRAINMOD model was calibrated separately for each experimental plot using a short-term (approximately 3 yr) record of observed weather data and water table measurements collected on-site. Soil water characteristics and saturated hydraulic conductivity were also determined for each plot (He et al., 2002). Details of the model calibration procedure were reported by He et al. (2002). Predicted and monitored water table fluctuations were compared and then selected model parameters were adjusted to bring predicted values in line with measured ones. Measured and predicted daily water table depths differed by an average absolute deviation of 20 cm or less for most plots.
Historic water table levels were predicted using the calibrated DRAINMOD models for each plot and long-term rainfall data. A 40-yr record of daily rainfall data were available for the period from 1 Jan. 1959 through 31 Dec. 1998 from a weather station located 9.2 km from the site. The daily rainfall data were then disaggregated into hourly data and converted to a format compatible with DRAINMOD using a computer program developed by Robbins (1988). Daily maximum and minimum temperature data and geographic location were required for the potential evapotranspiration (PET) calculation in DRAINMOD. These data were also obtained from the weather station that supplied rainfall data.
Daily water table levels were computed for each of the 13 plots for the previous 40 yr using the calibrated DRAINMOD models. Required inputs for the hydrology analysis included the starting day and ending day of the simulation, continuous days of saturation, and maximum depth to saturation. Hydrology "within the growing season" and "outside of the growing season" was of interest. The growing season used here began on 16 March and ended on 13 November in Pitt County, and these dates were based on an air temperature of -2°C (28°F) and a probability of 5 yr in 10 (Karnowski et al., 1974). To determine whether soil temperature influenced relationships between saturation frequency and soil color, water table levels were simulated over four time periods each year (1 January to 15 March, 16 March to 15 July, 16 July to 13 November, and 14 November to 31 December). Saturation frequency was determined for durations 21 consecutive days, in depth increments of 15 cm to a depth of 90 cm.
Percentages of redoximorphic features were determined by soil horizon, and horizon boundaries did not coincide with the depth increments used for the model simulation. Therefore, percentages of redoximorphic features were estimated in depth increments of 15 cm to a depth of 90 cm to perform statistical correlations. Percentages of redox depletions (with chromas of 2 or less and values of 4 or more) and concentrations (or red mottles) were estimated for each depth increment using a weighted average approach based on the soil profile description data. For example, we assumed that the percentage of redox depletions determined for a horizon applied to all depths that the horizon spanned. When two horizons were included in a single 15-cm depth increment, then the percentage of redox depletions used for that depth increment was a weighted average based on the proportion of each horizon's thickness in the depth increment. Statistical analyses were conducted to correlate the percentages of redox depletions and concentrations, and saturation events using SAS statistical software (Version 7.0; SAS Institute, 1998). Saturation events within the growing season, out of the growing season and during a year were used in the correlation.
|
RESULTS AND DISCUSSION
|
|---|
Matrix Color and Redoximorphic Features
Soil profile descriptions, including matrix color, percentages of redoximorphic features, and organic C concentrations for each horizon were reported previously (Hayes and Vepraskas, 2000). The estimated percentages of redox depletions and redox concentrations that were used for statistical correlations are summarized for each depth range to illustrate the range in values encountered (Table 1). When redox depletions occurred in abundances >50% they formed the matrix color of the horizon, while smaller percentages occurred as mottles. The E and B horizons of some Rains and Pantego soils had dark organic stains or organic materials with Munsell values 
4 and chromas 1. These features occupied between 10 to 30% of E or B horizons in four plots and were included with the percentage of redox depletions for this study. We assumed that the organic accumulations would only be preserved in soils that were saturated and reduced, so we expected they would be related to saturation frequency and duration. The abundance of redox depletions increased in a consistent manner in the sequence from moderately well-drained Goldsboro soils to very poorly drained Pantego soils (Table 1). Percentages of redox concentrations increased from moderately well-drained Goldsboro soils to poorly drained Rains soils and then decreased in very poorly drained Pantego soils.
View this table:
[in this window]
[in a new window]
|
Table 1. Means (with minimum and maximum values) of redoximorphic features found in the four soils of the Greenville site. Data were obtained from profile descriptions and estimated for the preselected depths. Data for the Pantego soil were obtained from one pit. Ranges in values were influenced by distance from the ditch as discussed by Hayes and Vepraskas (2000).
|
|
Iron Reduction and Redox Potentials
Iron reduction must occur for low chroma colors or redox depletions to develop in a soil (Vepraskas, 2000). Below a temperature of 5°C microbial activity within soils is assumed to be insignificant (Soil Survey Staff, 1999). Soil temperatures for different soils were averaged across the transects for depths of 60 cm and are shown in Fig. 2
. Soil temperatures measured at a depth of 60 cm approximate the average daily soil temperature (Soil Survey Staff, 1999). The average daily soil temperatures were above 5°C throughout the year indicating that Fe reduction could occur year round in this landscape.
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 2. Average daily soil temperatures in the Goldsboro (G), Lynchburg (L), Rains (R), and Pantego (P) plots.
|
|
The occurrence of Fe reduction was determined by redox potential measurements. The threshold EH value for the beginning of Fe reduction was estimated from an EH–pH diagram developed for the mineral FeOOH that can be computed from (Vepraskas, 2000):
 | [1] |
The pH values determined for each plot were used for the estimation of EH(Fe2+). Average soil pH values ranged from 3.8 to 4.9 across all soil plots for depths between 15 to 60 cm. Hayes (1998) verified that Fe2+ was present in the soil water when field redox potentials fell below the EH(Fe2+) value computed by Eq. [1] for the soils of this study.
Mean values for redox potentials in the Rains plots were examined from 1997 to 1998 at depths of 15, 30, and 60 cm to determine the amount of time required for Fe reduction to occur after an aerated soil became saturated. An example of the redox data is shown in Fig. 3
for soil at a depth of 30 cm in Plot 2R to illustrate the data of interest. The line showing when Fe reduction occurred was computed using Eq. [1]. For one wetting period in 1997 to 1998, the lag period was 21 d between the onset of saturation (which began on 25 Dec. 1997) and the beginning of Fe reduction (which occurred on 15 Jan. 1998). Table 2 lists the lag periods for all Rains plots determined for 1997 and 1998 when the water table rose and Fe reduction began. A longer duration of saturation was required for Fe reduction to occur in deeper horizons, probably because of the lower organic matter contents found there (Hayes and Vepraskas, 2000). The average lag period between the onset of saturation and onset of reduction for the 15-, 30-, and 60-cm depths was found to be 21 d for this landscape.
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 3. Variation in redox potential (EH) over time for Plot 2R at a depth of 30 cm. Mean and range in EH is shown as determined from five electrodes at this depth. The data were used to determine the time required for the mean EH value to fall to where Fe reduction would occur following a saturation event. In this example, the soil had a high mean EH before December 25, and it required 21 d of saturation before the mean EH value fell into the level where Fe reduction was expected.
|
|
View this table:
[in this window]
[in a new window]
|
Table 2. Lag time between the onset of saturation and onset of Fe reduction based on average EH values measured in Rains plots during the fall to spring of 1997 and 1998 (Hayes unpublished data, 1998).
|
|
This 21-d value was the average amount of time that a horizon had to be saturated before the Fe-reducing reactions that produce redoximorphic features would occur for all depths between 15 and 60 cm. This value pertains to this landscape and does not necessarily represent conditions found in other areas. We expect that longer lag periods will be found on soils with less soluble organic C, higher pH's, and steeper slopes where water will move faster and take longer to become reduced than in flatter landscapes (Grieve, 1990). The 21-d value was used to approximate a minimum period of saturation to use for comparison with soil colors, although shorter periods are clearly required near the surface where greater amounts of soluble C typically occur.
Saturation Events
DRAINMOD computed the number of times a soil plot was saturated for 21 consecutive days or longer each year above specific depths. The model also computed the longest continuous period of saturation found during the year. An example of the output from DRAINMOD is shown in Table 3 for Plots 2R and 2G at a depth of 75 cm. The period of simulation extended from Day 1 (1 January) through Day 74 (15 March) of each year. It can be seen from the table that Plot 2R was predicted to have a water table within 75 cm of the surface only once per year. The water table remained above 75 cm for either 55 or 74 d for the period shown.
View this table:
[in this window]
[in a new window]
|
Table 3. Calculation of saturation events occurring at a depth of 75 cm for periods lasting a minimum of 21 consecutive days in Plots 2R and 2G during the first simulated period from 1 January to 15 March. Total and average values were computed for the full record of data which lasted from 1959 to 1998, while only a partial record is reproduced below. Abbreviations represent variables used in Eq. [2].
|
|
We initially used the value for "number of periods of saturation" (Table 3) for correlation with percentages of redoximorphic features. This was found to be unsuitable because two soil plots could both have been saturated once during the period while their durations of saturation ranged widely. For example, Plot 2G was saturated above a depth of 75 cm for as little as 25 d in some years, while Plot 2R was saturated for nearly three times longer (Table 3). Soil horizons that were saturated for the longer period had more redoximorphic features. To overcome this problem we developed another measure of saturation, termed saturation event, that incorporated both the number of periods of saturation and the longest duration of saturation as computed by DRAINMOD. This variable gave more weight to longer saturation events than to shorter ones. The number of saturation events was calculated from the DRAINMOD output as follows:
 | [2] |
where NSE is the number of saturation events in a given time interval, LDS is the integer of the longest duration of a single period of saturation (in days) during the period of interest, and NPS is the number of times the soil was saturated for periods lasting 21 d or longer during the period of interest.
We computed NSE values for each depth simulated in the 13 plots of the toposequence. Examples of the computations are shown in Table 3 for plots in the Rains and Goldsboro soils. The Rains Plot 2R experienced three saturation events in most years. On average, 2.9 saturation events occurred per year over the period from 1958 through 1998. On the other hand, the water table rose more frequently in the Goldsboro Plot 2G, but the longest periods of saturation were shorter than those of the Rains Plot 2R. The computed saturation events in Plot 2G ranged from 1 to 3 for the 40-yr simulation period, and the average number of saturation events per year was 2.2.
The computation of NSE allowed us to summarize saturation frequency data for durations 21 d. Table 4 shows the average NSEs that were computed for each plot. The average NSEs is the probability that the water table will rise above a specific depth in a year and remain above that depth for 21 consecutive days. The average number of saturation events could be met with different frequencies and durations of saturation as shown in Table 3.
View this table:
[in this window]
[in a new window]
|
Table 4. Saturation events computed for the time the water table was within the indicated depth consecutively for at least 21 d during the given period on the 13 plots at the site.
|
|
Correlation of Soil Redoximorphic Features with Saturation Events
Once the average numbers of saturation events were determined for the depth ranges of interest, these values were then correlated with the percentage of redox concentrations, percentage of redox depletions, and a combination of both. Periods of interest included within growing season, out of growing season, and an entire year. The average NSEs for an entire year was computed by simply adding the saturation event values in the four simulated periods in a year. Average number of saturation events within the growing season and out of growing season was calculated in the same way by adding the average values in the two simulated periods within or out of the growing season. Because there were few redoximorphic features in the top 30 cm of most soils, the correlations were completed for the depths below 30 cm from the surface. All correlations between percentage of redoximorphic features and saturation events within the growing season, out of growing season and throughout a year were determined using the simple regression models:
 | [3] |
 | [4] |
Data plots of relationships between redox depletion percentage and average number of saturation events were generally linear but varied with depth (Fig. 4)
, so regression analyses were completed separately for individual depths. The r2 values (Table 5) were generally >0.78 for relationships with both types of features indicating a linear relationship existed between saturation events and redoximorphic feature percentage. Correlations between percentage of redox depletions and saturation events in deeper horizons (i.e., 75 and 90 cm) were lower than for shallower horizons. One reason for this is that only saturation events lasting 21 d were considered. Longer durations of saturation are probably needed for Fe reduction to occur in horizons below 75 cm in these soils because less soluble organic C is available at depth to promote rapid Fe reduction, and hence a greater production of redox depletions (Hayes and Vepraskas, 2000). Nevertheless, the results show that once saturation duration exceeds 21 d then the percentage of redox depletions increases with increasing frequency or duration of saturation.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 4. Regression lines and data points for the relationship between the average number of saturation events (NSEs) during the growing season and the percentage of redox depletions. The data show that linear relationships are justified for all depths.
|
|
View this table:
[in this window]
[in a new window]
|
Table 5. Summary of parameters for relating average number of saturation events to percentage of redox depletions and redox concentrations in single variable linear regression models.
|
|
Looking at results for redox depletions (Table 5), the strongest correlations were found within the growing season where r2 values were >0.9 at depths of 45 and 60 cm, and >0.85 at 75 and 90 cm. Simulated water tables stayed above 75 cm almost the entire period outside of the growing season in most soils. This limited the range of saturation events and contributed to lower r2 values. Regressions computed for saturation events throughout a year improved the r2 values over what was obtained for the period outside the growing season. Correlations between redox concentration percentage and average NSE were not as strong as those obtained for redox depletions.
Number of saturation events was related to both the percentage of redox depletions and percentage of redox concentrations using the following regression model:
 | [5] |
Model parameters A and B for percentage of redox depletions and percentage of redox concentrations are tabulated in Table 6. Most R2 values were above 0.9 for regression at each depth and during each period of interest. Regression results for depths of 75 and 90 cm were improved over those found when only percentage of redox depletions was put in the model. Parameters A and B both increased with increasing depth showing that longer saturation durations were required to form comparable percentages of redoximorphic features in deeper horizons as compared with shallower horizons.
View this table:
[in this window]
[in a new window]
|
Table 6. Summary of regression parameters for the relationship between percentages of two kinds of redoximorphic features and saturation index. Linear regression models were determined for three periods, within growing season, out of growing season, and throughout a year, using data for the Greenville site.
|
|
|
APPLICATIONS
|
|---|
The occurrence of soil saturation is commonly assessed by looking for the depth that redox depletions occur. For example, if redox depletions are found at a depth of 45 cm in a soil, then it is assumed that the seasonal high water table occurs at a depth of 45 cm. While it is usually not known whether the seasonal high water table rises annually to the shallowest depth where redox depletions are found, soil scientists have generally believed that the water table rises to that level, and stays there for a prolonged period. Estimation of the duration of saturation at the depth where redox depletions are found requires long-term hydrologic data. Tables 5 and 6 show that percentages of redoximorphic features are related to saturation data predicted from hydrologic models based on historic rainfall and soil hydraulic properties. Consequently, we can estimate saturation frequency for fixed durations from the percentages of redoximorphic features (Fig. 5) . We caution, however, that these results are site specific and can only be applied to the Goldsboro, Lynchburg, Rains, and Pantego soils at this site.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 5. Relationships between average number of saturation events (NSEs) and percentage of redox depletions for the periods covering the entire year and outside the growing season. Data were obtained from all plots at the site.
|
|
Figure 5A shows that rather than estimate a vague concept such as seasonal high water table, the depth and percentage of redox depletions can be used to estimate both a frequency and duration of saturation. For example, if a soil were found to have 15% redox depletions at a depth of 45 cm, then the soil would have an average NSE of 1 for the entire year. We predict from this that the soil at a depth of 45 cm saturates for 21 to 41 d on average about once per year. In cases where soils have 5% depletions at 45 cm, the soils have an average NSE of 0.3. These soils saturate for 21 d or more approximately 3 yr in 10. The redox depletions in this case are not relict, they simply form during periodic events that don't occur every year.
If time of year is important for an interpretation then graphs can also be developed to show the likelihood of saturation occurring inside or outside of the growing season (Fig. 5B). Graphs such as these will allow evaluations of seasonal wetness in soils with greater precision than present estimation techniques. With additional work that determines NSE values within 30 cm of the surface, we may even be able to estimate jurisdictional wetland hydrology from soil color patterns.
Statistical models using only one variable, such as percentage of redox depletions, are easily displayed graphically and are simple to use and apply in the field. Considering both the percentages of redox depletions and redox concentrations provide better estimates of saturation conditions, but multiple variables are cannot easily be displayed graphically.
The methods used here can be applied to virtually any soil or toposequence, although the results of this study should be considered site-specific at this point. Different results might be expected for soils containing more soluble organic C, those found in cooler temperatures, or on different landscape positions. These results are based on a minimum saturation duration of 21 consecutive days. This value clearly differs with depth but was selected because it was the average length of time a soil needed to be saturated before Fe reduction occurred. Iron reduction is the process that begins the formation of redox depletions or gray soil colors. As discussed earlier, the 21-d value is site-specific and was intended to be applied to all depths between 0 and 90 cm. Should hydrologic information be needed solely for shallow depths then a shorter duration of saturation might be appropriate.
|
ACKNOWLEDGMENTS
|
|---|
Funding for the research was obtained from the U.S. Environmental Protection Agency (contract no. CR 824735-01-0) and the water Resources Research Institute of the University of North Carolina (WRRI Project no. 70175). Their assistance was greatly appreciated.
Received for publication April 30, 2002.
|
REFERENCES
|
|---|
- Bouma, J. 1983. Hydrology and soil genesis of soils with aquic moisture regimes. p. 253–281. In L. P. Wilding et al. (ed.) Pedogenesis and soil taxonomy. I. Concepts and interactions. Elsevier, Amsterdam, the Netherlands.
- Buol, S.W., and R.A. Rebertus. 1988. Soil formation under hydromorphic conditions. p. 253–262. In D. D. Hook et al. (ed.) The ecology and management of wetlands. Vol. 1: Ecology of wetlands. Timber Press, Portland, OR.
- Daniels, R.B., E.E. Gamble, and L.A. Nelson. 1971. Relations between soil morphology and water-table levels on a dissected North Carolina Coastal Plain Surface. Soil Sci. Soc. Am. Proc. 35:781–784.
- Environmental Laboratory. 1987. Corps of Engineers Wetland Delineation Manual. Technical Report Y-87–1. U.S. Army Engineers Waterways Experiment Station, Vicksburg, MS.
- Franzmeier, J., E. Yahner, G.C. Steinhardt, and H.R. Sinclair, Jr. 1983. Color patterns and water table levels in some Indiana soils. Soil Sci. Soc. Am. J. 47:1196–1202.[Abstract/Free Full Text]
- Grieve, I.C. 1990. Variation in chemical composition of the soil solution over a four-year period at an upland soil site in southwest Scotland. Geoderma 46:351–362.
- Hayes, W.A., Jr. 1998. Effect of ditching on soil morphology, saturation, and reduction in a catena of Coastal Plain soil. M.S. thesis, North Carolina State University, Raleigh, NC.
- Hayes, W.A., Jr., and M.J. Vepraskas. 2000. Morphological changes in soils produced when hydrology is altered by ditching. Soil Sci. Soc. Am. J. 64:1893–1904.[Abstract/Free Full Text]
- He, X., M.J. Vepraskas, R.W. Skaggs, and D.L. Lindbo. 2002. Adapting a drainage model to simulate water table levels in Coastal Plain soils. Soil Sci. Soc. Am. J. 66:1722–1731.[Abstract/Free Full Text]
- Jacobs, P.M., L.T. West, and J.N. Shaw. 2002. Redoximorphic features as indicators of seasonal saturation, Lowndes County, Georgia. Soil Sci. Soc. Am. J. 66:315–323.[Abstract/Free Full Text]
- Karnowski, E.H., J.B. Newman, J. Dunn, and J.A. Meadows. 1974. p. 71. Soil Conservation Service. U.S. Gov. Print. Office. Washington, DC.
- Megonigal, J.P., W.H. Patrick, Jr., and S.P. Faulkner. 1993. Wetland identification in seasonally flooded forest soils: Soil morphology and redox dynamics. Soil Sci. Soc. Am. J. 57:140–149.[ISI]
- Pickering, E.W., and P.L.M. Veneman. 1983. Salt bridges for field redox potential measurements. Commun. Soil Sci. Plant Anal. 14:669–677.
- Pickering, E.W., and P.L.M. Veneman. 1984. Moisture regimes and morphological characteristics in a hydrosequence in Central Massachusetts. Soil Sci. Soc. Am. J. 48:113–118.[Abstract/Free Full Text]
- Robbins, K.D. 1988. Simulated climate inputs for DRAINMOD. Ph.D. dissertation. North Carolina State University. Raleigh, NC.
- SAS Institute. 1998. SAS user's guide. Version seven. SAS, Inc., Cary, NC.
- Schoeneberger, P.J., D.A. Wysocki, E.C. Benham, and W.D. Broderson. 1998. Field book for describing and sampling soils. Version 1.1. Natural Resources Conservation Center, USDA, National Soil Survey Center, Lincoln, NE.
- Simonson, G.H., and L. Boersma. 1972. Soil morphology and water table relations: II. Correlation between annual water table fluctuations and profile features. Soil Sci. Soc. Am. Proc. 36:649–653.
- Skaggs, R.W. 1978. A water management model for shallow water table soils. Technical Report 134. Water Res. Research Inst., Raleigh, NC.
- Soil Survey Staff. 1999. Soil taxonomy. A basic system of soil classification for making and interpreting soil surveys. US Dep. of Agriculture Handbook No. 436, 2nd ed. U.S. Gov. Print. Off., Washington, DC.
- Veneman, P.L.M., L.A. Spokas, and D.L. Lindbo. 1998. Soil moisture and redoximorphic features: a historical persepective. P. 1–23. In M.C. Rabenhorst et al. (ed.) Quantifying soil hydromorphology. SSSA Special Publ. No. 54. SSSA, Madison, WI.
- Vepraskas, M.J. 1999. Redoximorphic features for identifying aquic conditions. Tech. Bull. 301. NC Agric. Exp. Stn., Raleigh, NC.
- Vepraskas, M.J. 2000. Morphological features of seasonally reduced soils. P. 163–182. In J.L. Richardson and M.J. Vepraskas (ed.) Wetland soils: Genesis, hydrology, landscapes, and classification. Lewis Publ., Boca Raton, FL.
This article has been cited by other articles:
|
|
|
|
 
M. J. Vepraskas, X. He, D. L. Lindbo, and R. W. Skaggs
Calibrating Hydric Soil Field Indicators to Long-Term Wetland Hydrology
Soil Sci. Soc. Am. J.,
July 1, 2004;
68(4):
1461 - 1469.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
