| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Environmental Science and Technology, Univ. of Maryland, 1109 H.J. Patterson Hall, College Park, MD 20742
b USDA-ARS, Pasture Systems and Watershed Management Research Unit, University Park, PA 16802
c Dep. of Agriculture, Univ. of Maryland Eastern Shore, Princess Anne, MD, 21853
* Corresponding author (bneed{at}umd.edu)
Received for publication March 7, 2006.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONREFERENCES |
|---|
Abbreviations: Alox, acid ammonium oxalate-extractable Al • DPS, percent degree of P saturation • Feox, acid ammonium oxalate-extractable Fe • Pox, acid ammonium oxalate-extractable P • UMES, University of Maryland Eastern Shore
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONREFERENCES |
|---|
The southern Delmarva Peninsula has a relatively flat relief and is dominated by poorly drained soils. The water table in this region is close to or at the surface for extended periods of time during the year. To permit cultivation, the region relies on open-air ditches to lower the water table and quickly remove overland flow during periods of intense rainfall. Agricultural ditches are connected hydrologically to local streams and rivers and are a pathway for sediment and nutrients from agricultural ecosystems (Vadas and Sims, 1998). Ditches are widespread in the USA either as the primary means of land drainage or as collection ditches for tile-drained and irrigated lands (Fausey et al., 1995; Thomas et al., 1995; Evans et al., 1996).
The southern Delmarva Peninsula contains an intense poultry industry, which produced more than 560 million broiler birds and more than 1.3 million Mg of chicken in 2004 (Delmarva Poultry Industry, 2005). Large quantities of poultry litter (poultry manure combined with woodchips, shavings, or other bedding material) are produced each year on the Delmarva Peninsula, much of which is land applied as fertilizer for crops. Poultry litter has a low N/P ratio (Kleinman et al., 2005) such that application of litter at a rate suited to meet crop N requirements generally results in application of P above that required by the crop. The continual application of poultry litter in excess of crop needs leads to the accumulation of P in soils and increased potential for P loss in runoff (Sharpley, 1999).
Ditch soils may act as both a sink and source of P (Sallade and Sims, 1997a, 1997b; Nguyen and Sukias, 2002). Mechanisms that control this relationship include both sedimentation and resuspension of organic matter and P-enriched soil particles, the sorption and desorption reactions of P in solution with mineral and organic compounds, and the uptake and release of P by plants and microorganisms (Johnston et al., 1997).
In acidic soils, such as those found in the Atlantic Coastal Plain, controls of P can be attributed to Fe and Al hydroxides and organic matter cycling (Vadas and Sims, 1998). Low redox potentials develop in ditch soils during periods of warm weather and slow overlying water movement when the decomposition of organic matter is occurring. In the absence of oxygen, ferric Fe may be used as an electron acceptor causing the dissolution of ferric Fe and the subsequent release of Fe-bound P (Reddy et al., 1995; Vadas and Sims, 1998). Aluminum-bound P is not affected by anoxic conditions (Darke and Walbridge, 2000).
Phosphorus in agricultural field soils can exhibit high variation with important implications for agronomic (Gupta et al., 1997) and environmental P management (Larson et al., 1997; Needelman et al., 2001). In wetlands, the distribution of P has been correlated with distance to surface inflows (DeBusk et al., 1994) and sorption capacity (Bruland and Richardson, 2004). The spatial variation of P in ditch soils has not previously been investigated. An understanding of the degree of variation (variance) and spatial patterns (autocorrelation) of P in ditch soils would assist in the understanding of transport and deposition processes (i.e., zones of sink, source, stability), the development of ditch sampling designs, and the improvement of ditch management, modeling, and mass-balance estimation.
The objectives of this study were to (i) assess the spatial variation of oxalate-extractable P, Al, Fe, and pH in a soil depth of 0 to 5 cm within a vegetated agricultural ditch network and (ii) examine the relationships between ditch soil P variation and both ditch soil properties and farm structure.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONREFERENCES |
|---|
|
We categorized the ditches within the study area as either primary, shallow collection, or deep collection. Primary ditches are shallow (<1.5 m), drain surface runoff and shallow subsurface flow, often contain stagnant water, and are intermittently inundated. Collection ditches transport outflow from primary ditches. Two types of collection ditches can be discerned: shallow collection (approximately 1.5–2.0 m in depth) and deep collection (>2 m in depth). Shallow-collection ditches are seasonally connected to shallow ground water and intermittently inundated. Deep collection ditches are connected to deep, regional ground water and are permanently inundated. Ditches were labeled using the letter D appended with either an X for primary field ditches or an XX for collection ditches. The letters S and D were used to distinguish between shallow (i.e., S) and deep (i.e., D) collection ditches and a numerical identifier (e.g., 1, 2, 3...) was used to distinguish between ditches of the same type. Thus, DX1 and DX2 distinguish between two primary ditches, whereas DXXS1 and DXXD1 identify shallow and deep collection ditches, respectively.
Ditch Soils
Previous investigations have indicated that soils have formed in the sediments in the ditches at the study area. These sediments are able to support rooted vegetation and horizons have formed through pedogenesis, including organic horizons at the soil surface and gleyed horizons in the profile (Vaughan, 2005). We, therefore, use the term "ditch soils" although the parent material of the sola of these soils is sediment, just as one would discuss a floodplain soil though it is formed from sediment.
Ditch soil profiles were generally A horizons formed in alluvium in-filled since the most recent ditch clean-out over C horizons formed in the original Coastal Plain sediment-derived soils. Many soils had thin Oi horizons overlying the A horizons. Ditch A horizons were dark in color (mean value = 3.3; chroma = 1.8); ditch C horizons were lighter in color (mean value = 5; chroma = 2). Ditch A horizons were loamy in texture while ditch C horizons were coarser in texture and dominated by very gravelly sands, gravelly sands, and sands. Small segments of the ditch network had sandier A horizons. Subangular and granular soil structure was generally observed in the A horizons of primary and shallow collection ditches while the A horizons of deep collection ditches were structureless (high n-value). Ditch C horizons were primarily structureless. Redoximorphic features such as depleted matrices and Fe depletions and concentrations were common in both ditch A and C horizons. Ditch soils were generally acidic, with pH ranging from 2.6 to 6.1, and a mean of 4.7. Ditch A horizons were enriched with organic carbon from 0 to 124 mg kg–1 with a mean of 24 mg kg–1. Iron monosulfides (FeS) were observed on the surface of ditches DX1, DX2, and DX3 when they were submerged for extended periods of time. These surficial iron-monosulfides are referred to as monosulfidic black oozes (Smith, 2004). Additionally, geologic sulfidic materials that contain pyrite (FeS2) were found at a depth below most ditches located on the farm. It is thought that these sulfidic materials were deposited during a marine transgression, which is believed to have occurred either 82000 yr B.P. or 125000 yr B.P. (Toscano and York, 1992; Groot and Jordan, 1999; Wah, 2003).
All ditch soil profiles described were classified to the suborder level as Aquents. Endoaquents accounted for 61% of all profiles, with the subgroups being Sulfic, Aeric, and Hummaqueptic. Particle-size family classes were coarse-loamy or coarse-loamy over sandy or sandy-skeletal.
Field Methods
Soil samples were collected in the spring (March–April) of 2004. At the time of sampling, all ditches contained >8 cm of water. Sampling sites within each ditch were located at 10-m intervals starting from the intersection of a primary ditch and collection ditch using a wheeled measuring device in the field adjacent to each ditch, and then marking the sampling site within the ditch with either a flag or spray paint on the side of the ditch. Samples were composites of three cores (0–5 cm) extracted using a 7.6-cm open-face gouge auger in three evenly spaced (approximately 5 cm) distances perpendicular to the flow direction in the ditch. At every third sampling location (30-m intervals) an additional composite sample was collected within 5 cm of the original sample. The cores were placed into plastic sampling bags and composited by hand. Samples were transported back to the laboratory at air temperature and air-dried within 1 d of sampling. Upon returning to the laboratory and before air-drying, concentrations of what was presumably ferric Fe were observed on the inside of the sealed sample bags. Coarse organic debris was removed and the sample was ground to pass a 2-mm sieve. All analyses were performed on crushed, air-dried samples. A total of 405 samples were collected.
Laboratory Methods
Soil samples were extracted with acid ammonium oxalate elements by shaking 0.5-g air-dried soil with 20 mL 0.1 M (NH4)2C2O4·H2O + 0.1 M H2C2O4·2H2O (pH adjusted to 3.0) in the dark for 4 h. Extractable Al, Fe, and P (Alox, Feox, Pox) were determined on the supernatant by inductively coupled plasma atomic emission spectroscopy (ICP–AES) (Thermo Jarrel Ash 61 E, Franklin, MA) (Ross and Wang, 1993). Oxalate-extractable P was used rather than the commonly used Mehlich III P as a measure of soil P because of its ability to extract a greater proportion of P that is occluded by or tightly sorbed to Fe oxides that can form in soils with frequent oxidation–reduction cycles. Phosphorus that is occluded by Fe may become soluble under prolonged reducing conditions, therefore making acid ammonium oxalate-extractable P a better measure of P that can be potentially released to overlying drainage waters (Rhue and Harris, 1999). The degree of P saturation (DPS) was estimated as:
 | [1] |
, to estimate the proportion of Feox and Alox dedicated to P sorption. The use of in the literature has been varied (e.g., Lookman et al., 1996; Leinweber et al., 1997; Kleinman and Sharpley, 2002), and a growing number of researchers now do not use an in calculating DPS. Samples were analyzed for pH at a soil/water ratio of 1:1. These dry soil pH measurements may be lower than the pH of the soils at their initially moist condition because some ditch samples likely contained oxidizable sulfides (Vaughan, 2005). Previous investigations in these ditches have revealed the presence of monosulfides at or near the surface of ditch soils at concentrations up to 870 mg kg–1 (acid-volatile S) (Vaughan, 2005).
Statistical Analyses
Statistical analyses were performed using S-Plus and S+ Spatial Stats (Insightful Corporation, 2001). Statistical analyses were performed in two parts: (i) all ditches combined and (ii) individual ditches analyzed independently. Kolmogorov-Smirnov tests and descriptive statistics were used to assess normality. Only soil pH was found to be normally distributed; all other variables were found to be normal after log-transformation and were, therefore, log-transformed before statistical analyses (Press et al., 1989). Spatial autocorrelation was described using semivariance analysis (McBratney and Webster, 1986). A pooled variogram was generated for each variable by normalizing semivariance values for each ditch by the variance of that ditch (Goovaerts, 1997, p. 187). This was necessary due to the linear correlation between the mean and the variance (proportional effect). Semivariogram bins, or classes between point pairs, were set at 10-m increments to a maximum of 200 m. A minimum of 103 point pairs were present in all bins.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONREFERENCES |
|---|
|
|
|
|
|
|
|
|
While area A was located in a primary ditch, area B comprises roughly 100 m of a deep collection ditch (DXXD2). Area B contained the highest Pox concentration (6919 mg kg–1) found within the study area and ditch DXXD2 had the largest coefficient of variation (CV) of any ditch (123%). The mean Pox of area B was 2617 mg kg–1. Oxalate-extractable Fe was moderate while Alox was very high relative to other areas (Table 1). The DPS values observed in area B were also quite high, although soils were not as saturated with P as in area A due to their greater Feox and Alox concentrations (Table 2).
As with area A, DPS in area B was likely affected by the presence of an adjacent farm structure, or point source of P, and significant relationships between Feox and Pox (r = 0.66***) and between Alox and Pox (r = 0.98***) were observed. Near area B, a poultry broiler house (operational until spring of 2004) lay to the north in close proximity (Fig. 1). An access door where poultry litter and broiler chickens were routinely removed was located approximately 15 m to the north of Area B, separated by a compacted gravel road. During heavy rain events, we observed poultry manure solids floating in runoff water that flowed into area B from the direction of the barn. Thus, direct inputs of poultry manure into area B through runoff likely account for the extremely high concentrations of Pox and relatively high DPS measured in area B soils.
For both areas A and B, the transitions from high Pox and DPS to downstream zones of low Pox and DPS are abrupt. In the case of area A, there is an abrupt transition from ditch DX8, a shallow primary ditch, to deep collection ditch DXXD3. We can assume that the flow in DXXD3 is far greater than DX8 because DXXD3 is deeper, has continual baseflow, and receives flow from several ditches. This difference in flow may result in rapid dilution of P transfers from DX8 into DXXD3. In the case of area B, there is a rapid transition within the deep collection ditch DXXD2 from area B (high Pox and Alox) to an area of relatively low Pox and Alox (Fig. 3, 4, and 5). Although it is possible that some Alox found in area B was anthropogenic, it is possible that the change in Alox values reflects differing alluvial parent materials (i.e., those with contrasting levels of Alox enrichment or particle size distribution). Indeed, Feox also differed substantially between the two areas.
Ditch Areas C, D, and E (Relatively High Acid Ammonium Oxalate-Extractable Phosphorus from Nonpoint Sources)
The relatively high Pox concentrations in ditch soils sampled from ditch areas C, D, and E are likely caused by nonpoint sources of P. Ditch areas C, D, and E are all located at the downstream end of primary ditches that, superficially, have similar characteristics. Ditch soil samples from ditch areas C, D, and E possessed relatively high Pox and Feox concentrations (though substantially lower than areas A and B) but moderate Alox concentrations and DPS values. The three ditch areas were characterized by pH levels that were considerably lower than other areas in the study area (mean = 4.6) (Fig. 7
).
|
All three areas are located at potential depositional areas within the primary ditches, where P may accumulate. Following periods of storm flow, flow in these ditches diminishes rapidly, particularly near their juncture with the collection ditch DXXS1. As the primary ditches broaden, the deposition of sediment is more likely to occur in these areas. Sharpley et al. (1985) described the enrichment of P in eroded sediments relative to the soils from which they were derived, which would explain the higher Pox of areas C, D, and E. Although erosion rates in the region are relatively low and observed ditch flow appeared to be of low turbidity, the increase in Pox of areas C, D, and E is consistent with the deposition of sediments containing higher concentrations of Pox relative to adjacent fields.
Another hypothesis is that the high Pox concentrations in ditch areas C, D, and E result from pedogenic mechanisms. These areas are high in Feox relative to other ditch areas (Fig. 4). Phosphorus is most commonly found associated with Fe in acidic Coastal Plain ecosystems; therefore, the high concentration of poorly crystalline Fe (Feox) may be acting as a source of binding sites to retain Pox. The cause of such high concentrations of Feox in areas C, D, and E may be due to release of ferrous Fe during the oxidation of Fe monosulfides found at these soil surfaces and of sulfidic materials containing pyrite found at depth in this area of the farm. The oxidation of Fe monosulfides and pyrite can produce significant quantities of Fe in a ferrous (Fe2+) form, which can be converted to a ferric (Fe3+) form on exposure to oxygen (Fanning et al., 2002). This process also releases appreciable amounts of sulfuric acid (Fanning et al., 2002). The production of sulfuric acid through sulfide oxidation could explain the low soil pH values in these ditches (Table 2, Fig. 7).
The additional Feox produced by the oxidation of sulfidic materials may have a favorable effect on controlling P losses to ditch DXXS1, which receives water from DX1, DX2, and DX3 (Fig. 1). Ditch DXXS1 is enriched with Feox due to underlying sulfidic materials (Fig. 4), but is not enriched with Pox relative to DX1, DX2, and DX3 (Fig. 3). Thus, the enrichment of Feox in DX1, DX2, and DX3 at the outlets may be buffering losses of P to DXXS1.
Implications of Phosphorus Variation in Ditch Soils
If the spatial variation of soil P found at the UMES Research Farm is representative of ditches in agroecosystems with a history of manure application and intensive animal agriculture, the findings of this study have important implications for the management and understanding of P fate and transport. Traditionally, ditch management has focused on clean-outs (dredging) and woody vegetation control to maintain hydrologic function. Best management practices are being developed and assessed to maintain and improve hydrologic function while increasing nutrient retention and denitrification (Evans et al., 1996). Ditch management practices are currently applied without sampling or characterization of ditch soils and their geomorphic environment. Knowledge of the spatial variation of P within ditches would allow for targeting of best management practices. For instance, areas within a ditch network identified as having high DPS values could be selectively targeted for clean-outs to reduce P desorption from ditch soils to ditch waters.
If sampling strategies for ditch soils were to be developed, sampling design decisions would include point sampling versus compositing across a ditch, sample number, sampling depth, sampling method, P and other analyses such as Fe and Al, and full-ditch vs. zonal sampling. Sampling strategies and data interpretation developed to understand P loss potential from field soils may not apply to ditches due to redox fluctuations, high organic matter content, and the different hydrology of ditches. Sampling method should allow for the calculation of bulk density to quantify P storage in ditch networks. Measuring soil volume in ditch soils can be difficult due to saturated soils with organic horizons, loamy layers, and sandy layers; the push auger used in the present study caused some densification and therefore we were not able to determine bulk density.
At UMES ditches DXXD2, DX1, and DX3 exhibited substantial within-ditch zonation that would not have been ascertained with composite sampling across a ditch; in each case division of the ditch into thirds would have adequately captured this variation. The zonal nature of the variation of these ditches seems to have been caused by either local farm structure (DXXD2) or geomorphic setting (DX1 and DX3). It may not be necessary to sample ditches in zones that are unaffected by locally variable factors, but this would require methodology to consistently predict which ditches have substantial within-ditch variation.
There are currently no models available to estimate P transport processes within and over ditch soils. To manage and model P losses from ditches, an improved understanding is needed of P transport pathways from the landscape to and through ditch soils, P retention processes in ditch soils, and direct P losses from ditch soils to overlying waters. Areas of very high P and DPS (e.g., areas A and B) may be ditch critical source areas of P to downstream water bodies. Targeting of ditch critical source areas with best management practices may yield large improvements in water quality. Areas of lower P and DPS in the ditch network (e.g., ditch DXXD3) may sorb P from these areas, providing a natural mitigation mechanism. However, if stable and accreting, ditch soils may serve primarily as P sinks over the long term, in which case management should be performed to maximize this role.
|
SUMMARY AND CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  D.E. Radcliffe, J. Freer, and O. Schoumans Diffuse Phosphorus Models in the United States and Europe: Their Usages, Scales, and Uncertainties J. Environ. Qual., August 24, 2009; 38(5): 1956 - 1967. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. E. Vaughan, B. A. Needelman, P. J.A. Kleinman, and A. L. Allen Vertical Distribution of Phosphorus in Agricultural Drainage Ditch Soils J. Environ. Qual., October 24, 2007; 36(6): 1895 - 1903. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
