| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a USDA-ARS, Pasture Systems and Watershed Management Research Unit, Curtin Road, University Park, PA 16802-3702
b Dep. Natural Resource Sciences and Landscape Architecture, Univ. Maryland, College Park, MD 20742
* Corresponding author (pjk9{at}psu.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: Alox, oxalate-extractable Al • DRP, dissolved reactive P • Feox, oxalate-extractable Fe • Pox, oxalate- extractable P • Psat, P sorption saturation • PVC, polyvinyl Cl • TP, total P
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Direct observation of P leaching can be difficult, since leaching losses often occur in discrete events, or pulses, that can be easily missed; continuous flow-monitoring equipment is expensive. Traditional measures of leaching potential used for N, such as saturated hydraulic conductivity, may not be applicable to P leaching potential, as they emphasize matrix rather than macropore flow. For instance, preferential flow through macropores may be aggravated by low saturated hydraulic conductivity, as water is locally forced through macropores (Chardon and van Faassen, 1999).
While surface runoff has long been seen as the dominant pathway by which P is transported from land to water, subsurface flow pathways can be important to P transport in certain landscapes (Sharpley and Halvorson, 1994). For instance, subsurface P transport has been well documented through sandy soils (Sims et al., 1998), well-structured, fine-textured soils with tortuous macropores (Chardon and van Faassen, 1999; Djodjic et al., 1999) and porous organic soils with low P sorption capacities (Miller, 1979; Cogger and Duxbury, 1984). Subsurface transport is undoubtedly dominated by preferential flow through soil macropores (Simard et al., 2000), and environmentally important losses to surface waters are facilitated by the introduction of artificial drainage, which provides lateral connectivity between subsurface macropores and surface water (Dils and Heathwaite, 1999). In addition, factors such as soil P sorption saturation (Psat) and oxidation-reduction cycles can greatly increase P mobility through soils (Behrendt and Boekhold, 1993; Heckrath et al., 1995).
Where no direct P leaching data exist, investigators have turned to evidence of long-term P translocation within soil profiles as an indicator of P leaching potential (Kuo and Baker, 1982; Mozaffari and Sims, 1994; Eghball et al., 1996). Often, P profile data from reference soils with little to no history of P additions are compared with P profile data from similar soils with histories of P amendment. For instance, Eghball et al. (1996) concluded that significant leaching of P had occurred in a sandy loam soil (a Typic Haplustoll) because of elevated P concentrations in the subsoil horizons of soils receiving manure and mineral P applications. Elsewhere, Mozaffari and Sims (1994) and Kuo and Baker (1982) derived similar conclusions after comparing P profile data from unmanured and manured members of various mineral and organic soils (Aquic Hapludults, Typic Umbraquults, Typic Fluvaquents, Terric Medisaprists).
In contrast to those inferential studies reporting accumulations of subsurface P because of leaching, Thomas et al. (1997) observed significant leaching of P through a silty clay loam soil but little accumulation of P with depth. Similarly, Haygarth et al. (1998) observed total P (TP) concentrations in leachate of up to 0.89 mg L-1 from a heavily manured Typic Haplaquept with little evidence of P translocation below the Ap horizon. One reason that P profile data may not reveal evidence of P translocation, and therefore may provide an incomplete picture of P leaching potential, is that translocated P may be concentrated along preferential flow pathways (e.g., Hansen et al., 1998).
Data obtained from bulk soil samples favor detection of P translocated under matrix flow. As bulk sampling of soils homogenizes ped interiors with the walls of macropores, P sorbed to macropore walls may be sufficiently diluted by mixing that it is not detected. Sharpley et al. (1993), for instance, found little evidence of subsurface P accumulation in any of twelve heavily fertilized Oklahoma soils with 12 to 35 yr of poultry litter application. Possibly, bulk soil sampling diluted areas of P concentration thereby masking differences between manured and unmanured sites.
The objective of this study was to determine whether soil P profile data could be reliably used to assess P leaching potential in heavily manured soils. Bulk and clay film samples collected from soil horizons were analyzed for evidence of P translocation. To validate conclusions drawn from P profile data, tile-drain monitoring data and column leaching experiments were assessed.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
|
To assess subsurface losses from the hillslope, the tile-drain outlet was equipped with an automatic sampler in May 2000 (Sigma 900 Max Portable Sampler, HACH/American Sigma Company, Loveland, CO1) and programmed to collect flow-weighted composite samples of tile drainage. In addition, in December 2000, intact soil columns were collected adjacent to each of the profile description pits following the method of McDowell and Sharpley (2001a). A 30-cm diameter polyvinyl Cl (PVC) cylinder was pushed into the soil by constant pressure from a 2-Mg drop weight. To prevent surface compaction, the drop weight was not allowed to contact the soil surface and no visual evidence of compaction was observed following cylinder insertion into the soil. Columns were removed by excavating the soil adjacent to the submerged cylinder and then tilting the cylinder to cleanly break contact between the soil column and the underlying subsoil. Two 30-cm deep columns and two 50-cm deep columns were collected from each of the three locations.
Ped Dissection
Intact peds from each horizon were dissected with stainless steel scalpels to physically separate components (Buol and Hole, 1959). Subsoil horizons were carefully dissected, or scraped, to separate clay films, from the intact peds. Following dissection, samples were air dried prior to laboratory analysis. In all horizons, except for the Hartleton 2Bt4 horizon, which could not be sampled, the clay films of the argillic horizons are thick and distinctly colored. We do recognize that there was some contamination of soil components during the scraping process. However, because of the distinct morphological features, we believe that these soils lend themselves to effective ped dissection.
Leaching Experiments
Preparation of the columns for leaching experiments followed the method of McDowell and Sharpley (2001b). Figure 3
illustrates the design of the intact cores. To prevent by-pass flow between the soil and the PVC cylinder, soils were allowed to dry for several weeks to induce shrinkage. Once dry (at approximately the permanent wilting point), a void of approximately 0.5-cm width appeared between the soil and the cylinder. The lower end of the void was sealed with nonreactive silicone. Paraffin wax was then poured into the void, effectively sealing the interface between soil and cylinder (Fig. 3). Following this procedure, the bottom of the column was supported by a layer of cheese cloth and a 30-cm diam. PVC disk, perforated with roughly 60, 0.2-cm perforations. The disk, in turn, was held in place by a PVC cap sealed to the cylinder with silicone. To allow drainage, a hole was drilled into the cap and fitted with a 1-cm PVC nipple that could be inserted into a plastic collection container (Fig. 3). This design ensured that drainage did not become reductive during the leaching experiments.
|
|
Laboratory Analysis
Tile drainage and leachate from the soil columns were analyzed for dissolved reactive P (DRP) by filtering water samples (0.45 µm) and analyzing the filtrate colorimetrically by the method of Murphy and Riley (1962). Total P was measured on unfiltered leachate by modified semimicro-Kjeldahl procedure following Bremner (1996).
Soil samples were subjected to extraction with distilled water (Kuo, 1996), Mehlich-3 solution (Mehlich, 1984) and acid ammonium oxalate (Ross and Wang, 1993). Distilled water extraction was conducted by shaking 0.5 g of soil in 5-mL of distilled water for 1 h. Mehlich-3 and water soluble P concentrations were determined colormetrically (Murphy and Riley, 1962). Acid ammonium oxalate P, Fe, and Al concentrations were determined by inductively coupled plasma-atomic emission spectroscopy. For each soil extraction, four bulk horizon samples were analyzed to estimate variance within the horizons. Because of limited sample size, clay films, Fe concentrations, and Fe depletions were only analyzed in duplicate.
Soil P sorption saturation was determined from the acid ammonium oxalate data, as detailed in Kleinman and Sharpley (2002):
 |
Particle-size analysis was conducted on bulk samples by the hydrometer method (Gee and Bauder, 1986).
Data Analysis
Only bulk samples laboratory analysis was replicated to allow for statistical analysis. To statistically evaluate profile trends in soil P (i.e., P concentrations related to bulk and clay film samples), differences in bulk sample concentrations between individual horizons were assessed by ANOVA (Neter et al., 1996). Because clay film samples were collected in very small quantities, only single replicates were obtained. Therefore, to quantify trends we observed in clay film data, we made the assumption that the variance within the clay film data was equal to the variance observed with the bulk samples. Pair-wise comparisons were conducted by Student's t test (Snedecor and Cochran, 1991).
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
In the county-level soil survey, the permeability of Buchanan soils is listed as 1.5 to 5.1 cm h-1 in the Ap and Bt horizons and 0.2 to 0.5 cm h-1 in the fragipan (Eckenrode, 1985). The permeability of Hartleton soils is listed as 1.5 to 15 cm h-1 throughout the solum. Despite the contrasting permeabilities of Buchanan and Hartleton soils, macropore distribution in these soils is likely more important to P transport, given the processes described above. Macropores were abundant within the subsurface horizons of all three soils, as reflected by visually observable pores in the argillic horizons (quantified as "common" in all three soils), well-expressed structural properties (subangular blocky to prismatic), clay films, and root distributions (Table 1).
The Lower and Upper Buchanan soils have well developed fragipans. Fragic properties are well expressed including brittleness, excavation difficulty, and lack of roots in the matrix. The absence of a fragipan in the Hartleton soil is most likely because of the high rock fragment content (extremely channery). The fragipans cause seasonally perched water tables, as reflected by the somewhat poorly drained classification, and are key to the variable source area hydrology of this location (Gburek and Sharpley, 1998). In addition to serving as a cause of saturation-related overland flow, fragipan-perched water tables can possess substantial lateral flow that may be an important P transport mechanism (Reuter et al., 1998). Therefore, in these soils, tile drains enhance the transport of P to the stream channel.
Finally, the lithologic discontinuity within the solum of all three soils is of particular note to this investigation (Table 1). The discontinuity is demarcated by a rapid change in rock fragment content, and rock fragment characteristics (e.g., type, color, and grain size). In the Lower Buchanan, horizons beneath the discontinuity exhibit a redder hue. This layering of materials within the solum is likely a result of spatially variable processes operating during colluviation. Trends in P distribution related to the lithologic discontinuity are discussed in the following section.
Interpreting Phosphorus Distribution with Depth
Chemical analyses of samples taken from individual horizons reveal similar gross trends in P distribution within the three soil profiles. Table 3 presents chemical and physical properties for each profile, differentiating bulk samples from clay-film samples.
|
Soil Pox data do provide some evidence of P leaching via soil macropores. Specifically, Pox concentrations in the clay films of four of the subsoil horizons (Bt1 of Lower Buchanan, Bt2 of Upper Buchanan, Bt1 and 2Bt3 of Hartleton) are significantly higher than Pox concentrations in corresponding bulk samples (P < 0.05) (Table 3). This suggests that P has been translocated along soil macropores, possibly as colloidal P during lessivage or as dissolved P that is now sorbed to macropore walls. The elevated Pox concentrations (37 mg kg-1) in the clay films of the 2Bt3 horizon of the Hartleton soils occur at the approximate depth of artificial drainage, indicating the potential for environmental losses of leached P. However, such observations and interpretations must be tempered with the fact that an equal number of horizons exhibited the opposite trend, with Pox significantly lower in the clay films than in the bulk samples. Here it is important to reiterate that not all clay films are necessarily associated with active macropores. Rather, some clay films may belong to macropores that are now discontinuous with surface horizons, while others may represent plugged macropores that are no longer functional as conduits for bypass flow.
The enhanced mobilization of P under reducing conditions because of saturation has been well documented (e.g., Jensen et al., 1998). Reduction of Fe from the ferric (Fe3+) to ferrous (Fe2+) state results in the dissolution of Fe phosphates. Following dissolution, P may leach lower within the soil profile where it is resorbed following oxidation. One striking trend observed in all three soils is that Pox concentrations in the lower solum are significantly higher than in the upper argillic horizons (P < 0.001). Such a disparity in Pox concentrations could result from the dissolution and mobilization of P from saturated and reduced overlying argillic horizons (Table 3). Trends in bulk sample Feox with depth reveal significantly greater Feox concentrations (P < 0.05), and hence greater P sorption capacity, below the lithologic discontinuity where Pox concentrations are greater than in the overlying argillic horizons (Fig. 4) .
|
Water-Soluble Phosphorus
Water-soluble P is an indicator of P readily available to the soil solution, and is strongly correlated with DRP losses in runoff (Pote et al., 1999). Elevated water soluble P at depth would suggest that these soils have a potential to desorb P to drain water. Water-soluble P declines from the surface (1.3–13.5 mg kg-1) to the subsoil (0.1–1.9 mg kg-1), with the highest concentrations occurring in the surface and subsurface horizons of the Hartleton soil (Table 3). In the subsoil, water-soluble P is significantly higher in the Bt1 horizons of all soils than in lower horizons (P < 0.05). The changes in water-soluble P with depth highlight the role of manure additions as a source of this highly labile fraction, as well as the high P buffering capacities of the subsoil, which result in low absolute concentrations. The high water-soluble P concentrations in the bulk samples of the Bt1 horizons, relative to the lower argillic horizons, indicate limited translocation of soluble P from the soil surface horizons, possibly by matrix flow. However, the absence of elevated concentrations of water-soluble P lower in the profile, where artificial drainage is installed, indicates that leaching of water-soluble P through the soil matrix may not be important to subsurface P transport in these soils.
Unlike Pox and Psat, we anticipated significantly lower water-soluble P in the clay films than in the bulk samples because of increased sorption of solution P in the clay films related to their greater reactive surface area and elevated sorption capacities (Alox + Feox). In five of the eleven clay film samples (Lower Buchanan 2Btx1, Upper Buchanan Bt1 and Bt3 and Hartleton Bt1 and 2Bt3) water-extractable P is significantly lower than in the corresponding bulk samples (Table 3). Notably, in one horizon (the Lower Buchanan Bt1 horizon), water-extractable P was significantly higher in the clay film than in the bulk sample. In this case, the clay film sample had a considerably high Psat (3%) relative to the bulk sample (1%), which may explain the elevated water-extractable P concentration, as P desorption to water is strongly correlated with Psat (Sharpley and Rekolainen, 1997).
Mehlich-3 Phosphorus
Mehlich-3 P, an indicator of plant available P, ranges from 9 to 177 mg kg-1 at the soil surface to 0.3 to 3.0 mg kg-1 in the argillic horizons and 1.4 to 8.6 mg kg-1 below the lithologic discontinuity (Table 3). As a point of reference, in Pennsylvania, a crop response threshold has been estimated at a Mehlich-3 P concentration of 65 mg kg-1, above which the addition of fertilizer P is not expected to increase crop yield (Beegle, 1999). As such, Mehlich-3 P concentrations in the surface horizons of the Upper Buchanan and Hartleton soils are well in excess of crop requirements. Mehlich-3 P concentrations in the surface horizon of the Lower Buchanan soil are below crop requirements. Trends with Mehlich-3 P are similar to those with Pox, and both are highly correlated (r = 0.90, Mehlich-3 P = 0.3 Pox - 4.8). As with Pox, Mehlich-3 P concentrations were significantly greater in six of the eleven clay-film samples relative to corresponding bulk samples. Similarly, the Mehlich-3 P concentration of the C horizons (1.4–7.5 mg kg-1) was not significantly different from those below the lithologic discontinuity, again indicating an authigenic cause of elevated P below the lithologic discontinuity.
Origin of Phosphorus in the Lower Solum
To further deduce whether elevated P in the lower solum resulted from long-term leaching or from varied parent materials within the regolith, we collected samples beneath the solum of each of the three soils (approximately 2-m depth). Concentrations of Pox (12–20 mg kg-1) in the C horizon cores are not significantly different from the lower solum concentrations (P > 0.1). Similarly, Psat in the C horizons (2–3%) is not significantly different from Psat in the lower solum (P > 0.1). Thus, the elevated P concentrations in the lower solum appear to be authigenic (i.e., because of a discontinuity in soil parent materials). It is unlikely that leaching of surface-applied manure P would uniformly enrich the matrices of the lower solum and C horizons and that sufficient manure P has been translocated to account for the mass of Pox found in the regolith.
Leaching Observations
Table 4 presents leachate DRP, TP, and volume data from the soil columns. Prior to manure addition, leachate DRP concentrations ranged from 0.004 to 0.011 mg L-1 and TP ranged from 0.047 to 0.308 mg L-1, while leachate DRP loads ranged from 0.009 to 0.042 mg and leachate TP loads ranged from 0.107 to 1.026 mg. On average, DRP accounted for only 5% of TP concentration. No statistically significant (P > 0.05) differences in leachate P concentrations and loads were observed between the 30-cm deep cores, representing leaching potential to the argillic horizon, and the 50-cm deep cores, representing leaching potential to the lower solum and tile drainage (Table 4).
|
Results from the leaching experiment indicate the importance of manure addition to P leaching, and confirm the potential to leach P to the lower solum, approximately the depth of the tile drain, in all three soils. As manure was added at a rate corresponding with the annual average, and manure is typically applied to these soils in the spring, when soils are wet, it is likely that this experiment accurately simulates field conditions.
Because there was no systematic difference in leachate P losses between the 30- and 50-cm deep cores, results suggest that P is transported via preferential flow pathways. If matrix flow was the dominant mechanism of subsurface P transport, then one would expect P concentrations in leachate to be greater from the 30-cm cores than from the 50-cm cores. In fact, the greatest leaching losses (concentrations and loads) among all replicates were consistently observed from a 50-cm core (Lower Buchanan). Furthermore, the ranking of TP losses from individual cores after manure application was identical to the ranking of TP losses prior to manure application, such that the Lower Buchanan 50-cm core consistently had the highest losses while the Lower Buchanan 30-cm core had lowest losses. Such a consistent response illustrates that P transport from the soil surface, particularly from manure P, is responsible for subsurface P concentrations and that an active macropore flow pathway is necessary for P transport.
Monitoring of tile drainage from May to October 2000 further supports the importance of subsurface flow as a pathway for P loss from these soils (Table 5). Mean (flow weighted) drainage DRP concentrations ranged from 0.010 mg L-1 during base flow to 0.102 mg L-1 during storm event flow. Mean TP concentrations ranged from 0.019 mg L-1 during base flow to 0.283 mg L-1 during storm event flow. Although base flow volume (458 895 L) was greater than storm flow volume (58 935 L), loads of DRP and TP were substantially higher during storm flow than during base flow. Increases in drainage P concentrations during storm flow more than offset the greater flow volumes observed during base flow. The elevated losses of P in tile drainage during storm events support the conclusion that rapid, subsurface P transport can be significant within this system. In fact, storm flow concentrations are frequently above eutrophic criteria (0.1 mg L-1 as TP) established for streams or other flowing waters not discharging directly into lakes or impoundments (Dodds et al., 1998; U.S. Environmental Protection Agency, 1994).
|
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Increases in Pox, Psat, and Mehlich-3 P concentrations in bulk samples of the lower solum may be explained by two competing theories: (i) discontinuity in parent material (at the fragipan boundary in Buchanan soils); and, (ii) preferential flow of P from the soil surface to the lower horizons (i.e., bypass leaching). Based upon P profile data alone, we cannot discount the first theory, as rock fragment quantity and type differences within the subsoil indicate that two stratified parent materials are present in all three soils.
This study shows that soil P profile data must be carefully interpreted to properly assess P leaching potential. Bulk horizon samples appeared to provide evidence of P translocation, but the elevated P in the subsoil was ultimately attributed to lithological differences. The need for scrutiny of P profile data is because of the dominant mechanism of subsurface P transport: preferential flow. Leachate P bypasses the matrix of subsoil horizons via select macropores that are not taken into account by bulk sampling of horizons. Indeed, leaching experiments confirmed that the potential to leach P to the depth of artificial drainage was equivalent to the potential to leach P to shallow depths. Monitoring data revealed an increase in drainage P concentrations with flow, providing clear indication of the importance of bypass flow via soil macropores.
The transport of P by subsurface pathways can be an important mechanism of P transfer from land to water in heavily manured soils, especially those that are artificially drained or have preferential flow pathways connected to stream channel discharge. This study reveals that detailed description and interpretation of soil P profile data provides limited insight into P leaching potential: correlation of soil P fractions with leachate P from lysimeters was poor.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSNOTESRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. O. Young and R. D. Briggs Phosphorus Concentrations in Soil and Subsurface Water: A Field Study among Cropland and Riparian Buffers J. Environ. Qual., January 4, 2008; 37(1): 69 - 78. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Chrysostome, V. D. Nair, W. G. Harris, and R. D. Rhue Laboratory Validation of Soil Phosphorus Storage Capacity Predictions for Use in Risk Assessment Soil Sci. Soc. Am. J., August 9, 2007; 71(5): 1564 - 1569. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Page, P. M. Haygarth, K. J. Beven, A. Joynes, T. Butler, C. Keeler, J. Freer, P. N. Owens, and G. A. Wood Spatial Variability of Soil Phosphorus in Relation to the Topographic Index and Critical Source Areas: Sampling for Assessing Risk to Water Quality J. Environ. Qual., November 7, 2005; 34(6): 2263 - 2277. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Bechmann, P. J. A. Kleinman, A. N. Sharpley, and L. S. Saporito Freeze-Thaw Effects on Phosphorus Loss in Runoff from Manured and Catch-Cropped Soils J. Environ. Qual., November 7, 2005; 34(6): 2301 - 2309. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. S. Butler and F. J. Coale Phosphorus Leaching in Manure-Amended Atlantic Coastal Plain Soils J. Environ. Qual., January 1, 2005; 34(1): 370 - 381. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Sharpley and P. Kleinman Effect of Rainfall Simulator and Plot Scale on Overland Flow and Phosphorus Transport J. Environ. Qual., November 1, 2003; 32(6): 2172 - 2179. [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 | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
