| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Graduate Program in Water Resources Science, 500 Pillsbury Dr. SE, Univ. of Minnesota, Minneapolis, MN 55455
b Dep. of Civil Engineering and Water Resources Center, Univ. of Minnesota, St. Paul, MN 55108
c Water Resources Center, 173 McNeal Hall, Univ. of Minnesota, St. Paul, MN 55108
d Dep. of Soil, Water and Climate, Univ. of Minnesota, MN 55108
* Corresponding author (brezo001{at}umn.edu)
Received for publication July 15, 2001.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: BAP, bioavailable phosphorus • EPC, equilibrium phosphorus concentration • PP, particulate phosphorus • PSIm, phosphorus saturation index based on sorption maximum • PSIs, phosphorus saturation index based on sorptivity • SRP, soluble reactive phosphorus • STP, soil test phosphorus • TP, total phosphorus • TSS, total suspended sediment
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Correlations between soil test phosphorus (STP) levels and P concentrations in agricultural runoff have been reported in several field and laboratory studies (Sharpley, 1995a; Yli-Halla et al., 1995; Pote et al., 1996, 1999; Hooda et al., 2000). Results showed that STP extraction methods such as Mehlich-III P, Bray P, and Olsen P give soil P levels significantly correlated with dissolved P concentrations in surface runoff.
The P sorption saturation approach is based on the principle that soil P loss is a function of both the soil P level and the soil P sorption capacity. The degree of P saturation was defined by Breeuwsma et al. (1995) as the ratio of sorbed P to P sorption capacity of a soil. Sharpley (1995a) and Pote et al. (1996)( 1999) applied this concept to their studies and found that soil P sorption saturation also gave results that correlated well with surface runoff dissolved P. However, soils used in these studies mostly had pH levels less than 7, and many of the soils were acidic (pH < 6.0). No similar work has been done on calcareous soils (pH > 7.6; Thomas, 1996) that are common in some areas of the upper Midwest, including the Minnesota River basin in southern and western Minnesota. Most of the soils in these areas are of calcareous and glacial origin. These areas are intensively farmed, and P pollution to surface waters via surface runoff resulting from application of P fertilization and manure is a major concern (Minnesota Pollution Control Agency, 1997).
Here we report results from a laboratory runoff simulation experiment designed to study relationships between the dissolved P and BAP levels in runoff, and soil physical and chemical properties, including soil STP levels and P sorption saturation, of mostly calcareous soils from the Minnesota River basin. The four main objectives of the study were to (i) evaluate the effectiveness of P sorption saturation and STP methods in predicting dissolved P levels in runoff from calcareous soils in the Minnesota River basin; (ii) compare the effectiveness of P sorption saturation and routine STP methods in predicting runoff BAP levels (estimated with iron oxide–impregnated paper); (iii) determine relationships between runoff total P (TP), runoff PP, and runoff particulate BAP; and (iv) estimate critical soil test P levels to manage stream eutrophication.
From the Minnesota–South Dakota border, the Minnesota River flows 539 km through southern Minnesota, joining the Mississippi River at Fort Snelling in the Minneapolis–St. Paul metro area (Fig. 1) . The Minnesota River is polluted with sediment and nutrients (N and P) originating primarily from urban and agricultural activities, the latter of which accounts for 92% of the land use in the river basin (Mulla and Mallawatantri, 1997). As the state's largest tributary of the Mississippi River, the Minnesota River on average increases the Mississippi River's flow by 47% and adds disproportionately to the latter river's pollutant load (Minnesota Pollution Control Agency, 1997). The Minnesota River basin is composed of 12 major and 1208 minor watersheds in four physiographic regions (glacial moraines and bedrock highlands; glacial till plains and drumlins; lacustrine sediments of lowland river valleys and lake plains; and the Minnesota River valley outwash [Minnesota Pollution Control Agency, 1997]). Mean annual precipitation ranges from 787 mm in the eastern part of the basin to 584 mm in the west.
|
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Soil Phosphorus Sorption and Desorption
Soil P sorption was measured by a standardized procedure (Nair et al., 1984) with small modifications. For each sample location, 0.52 g of soil was placed into 50 mL Nalgene (Rochester, NY) Oak Ridge centrifuge tubes (n = 6). The total weight of each tube with the soil was measured (wo). A 13-mL aliquot of one of the following P solutions (0, 1000, 3000, 5000, 7500, and 10 000 µg P L-1, respectively as KH2PO4) was added to each tube along with 0.18 mL of chloroform to suppress microbial activity during shaking. The centrifuge tubes were tumbled in an end-over-end shaker for 24 h to reach equilibrium (Detenbeck and Brezonik, 1991) and then centrifuged at 12 000 rpm for 30 min. Supernatant was carefully transferred to 30-mL acid-washed Nalgene bottles and stored at 4°C in the dark.
After centrifugation, the weight of each centrifuge tube with remaining wet soil was measured (wt). The difference between wt and the initial dry weight wo was the weight of residue solution after centrifugation, which was translated into volume (density of the residue solution assumed to be 1.0 g mL-1) and used to calculate the initial P concentration of the desorption step. For desorption, 13 mL of 0.01 M CaCl2 and 0.18 mL of chloroform was added to the remaining wet soil in each centrifuge tube. After shaking for 24 h, tubes were centrifuged at 12 000 rpm for 30 min and supernatant was poured into 30-mL Nalgene bottles. The SRP in the supernatant from both the adsorption and desorption parts of the experiment then was measured after appropriate dilution. The amount of P adsorbed onto and desorbed from soil particles was calculated from these results, and the Langmuir sorption model was used to interpret the data.
Langmuir Sorption Model Fitting
An expanded Langmuir sorption model (Pollman, 1983) was used to plot the experimental data:
 | [1] |
x = mass of P released from soil (mg); v = volume of solution (L); m = mass of soil (mg); 
o = mass of P initially present on soil (mg kg-1 soil); 
= maximum sorption (mg kg-1 soil); b = constant related to the sorption energy (L µg-1); and Co + x/v: = C* = P concentration (µg L-1) at equilibrium.
When x = 0 (i.e., no net adsorption or desorption in the system), Eq. [1] becomes:
 | [2] |
The term Co, commonly known as the equilibrium phosphorus concentration (EPC), defines an important property of a soil: the ambient P concentration at which no net sorption or desorption occurs when a soil is suspended in the solution (Brezonik and Pollman, 1999). As Eq. [2] shows, EPC is a function of the P maximum sorption and binding energy.
Phosphorus Sorption Saturation
A generalized definition of P sorption saturation can be written as (Sharpley, 1995a):
 | [3] |
To apply Eq. [3] and calculate P sorption saturation, we used sodium-hydroxide P as the measure of extractable P. This was a natural choice because sodium-hydroxide P also was used in this study as the initial sorbed P, o, in fitting the Langmuir sorption model (Detenbeck and Brezonik, 1991; Sallade and Sims, 1997). Two measures of sorption capacity were used: (i) sorption maximum, ; and (ii) sorptivity = x b (He et al., 1999). Results are denoted as either PSIm (phosphorus saturation index based on sorption maximum; values in percent):
 | [4] |
 | [5] |
Equation [5] has units of concentration (mg L-1) for PSIs, and values were in the order of 10-1. For easy of interpretation, calculated values of PSIs were multiplied by 100 and the actual units ignored in data interpretation. Desorbability (the percentage of P desorbed from the soil after 24 h of equilibration with 0.01 M CaCl2 following the 24 h sorption equilibrium phase) and EPC also were used as indicators of soil P sorption capacity.
Runoff Simulation Experiment
Runoff boxes were constructed from Type-II PVC sheets (0.635 cm in thickness; Fig. 2)
. The boxes had an effective volume of 0.61 m (length) x 0.15 m (width) x 0.10 m (height). A space of 2.5 cm was created between the soil and the bottom of the boxes with a perforated PVC sheet to facilitate soil wetting and rainwater drainage. Mesh window screen, followed by a double-layer cheese cloth, was laid on top of the perforated PVC sheet to prevent loss of soil particles. Then, a V notch was cut on the downslope end wall of the box to aid runoff discharge, and a piece of mesh screen was attached to the notch to prevent sloughing of soil particles from the notch (Fig. 2).
|
Runoff was collected and measured in 1000-mL graduated cylinders. Runoff from each soil for an entire rain event was composited in an acid-washed 3-L plastic container. The measured runoff volumes from the rainfall simulations had a coefficient of variation of only 5%. A 30-mL aliquot was immediately filtered (0.45 µm) with Millex (Bedford, MA) HA syringe-driven filter units and stored at 4°C for SRP analysis. Unfiltered runoff was stored at 4°C for BAP, TP, and total suspended sediment (TSS) analysis. All analyses were completed within 5 d of sample collection.
Chemical Analyses
Total suspended sediment was determined by oven-drying (105°C) 30 mL of filtered (Whatman [Maidstone, UK] GF/F glass fiber filter) runoff samples (American Public Health Association, 1995). The SRP was determined by the ascorbic acid method (Murphy and Riley, 1962; American Public Health Association, 1995). Total P was measured by the same method following persulfate digestion (American Public Health Association, 1995).
Soil pH, organic matter, calcium carbonate equivalent, and TP, Mehlich-III P, Bray P, and Olsen P were measured by the Research Analytical Laboratory, University of Minnesota, St. Paul. Sodium-hydroxide P was extracted with a combined solution of 0.1 M NaOH and 1.0 M NaCl at a soil to solution ratio of 1/500 (w/v) in 50-mL Nalgene Oak Ridge centrifuge tubes. This method differs from that of Sharpley et al. (1991) in that 1.0 M NaCl was added to minimize P re-adsorption (Barbanti et al., 1994). After mixing in an end-over-end shaker for 17 h, tubes were centrifuged at 12 000 rpm for 30 min. Supernatant was removed from the centrifuge tubes and neutralized to pH approximately 8, with phenolphthalein as indicator. Phosphorus in neutralized extract was determined as SRP. Water-extractable P, which is effectively 0.01 M CaCl2–extractable P, was obtained as the result from the zero P addition of the P adsorption experiment.
Bioavailable P in runoff was measured with the iron oxide–impregnated paper method modified from Sharpley et al. (1994b) and Chardon et al. (1996). Iron oxide–impregnated paper was prepared by immersing Whatman 541 filter paper disks (55 mm in diameter, pore size 20–25 µm) into a solution of 0.37 M FeCl3·6H2O for 1 h. After air-drying, filter paper disks were dropped into a 2.7 M NH4OH solution for exactly 1 min, washed twice in distilled water, and air-dried.
To measure BAP, an iron oxide–impregnated paper disk was shaken with 1 g of soil (with 40 mL 0.01 M CaCl2) or 40 mL of unfiltered runoff sample in a 250-mL Erlenmeyer on a reciprocal mixer for 16 h at 150 rpm. The disk then was removed from the beaker, rinsed thoroughly with distilled water, and air-dried on a mesh sheet. Phosphorus absorbed on the disk was extracted with 40 mL of 0.1 M H2SO4 on a reciprocal mixer (150 rpm) for 1 h and measured as SRP. The SRP standards (0, 100, 250, 500, 1000 µg L-1) were run with the samples.
Algal Bioassay
The bioavailability of P from the 10 soils also was measured by algal bioassay procedures patterned after the method of Sharpley et al. (1991). We used Stephanodiscus hantzschii as the culture organism because it is the most common and abundant diatom in the Minnesota River. A stock culture of S. hantzschii was acquired from the diatom herbarium at Loras College (Dubuque, IA); this culture was originally isolated from the Minnesota River. Batch cultures were grown in full-nutrient COMBO medium for 5 wk (Kilham et al., 1998), filtered onto 1.2-µm glass fiber filters, and transferred to P-free COMBO medium for 21 d to induce a P-starved state. Soil samples (2-mm sieved, autoclaved, 50 mg per sample) were added to 25-mL of P-starved culture in duplicate 50 mL culture tubes to achieve a 500:1 solution to soil ratio. Four control replicates were performed (i.e., no sediment additions). A series of nine orthophosphate-addition treatments (1–100 µg P L-1) was incubated in addition to the sediment-amended treatments. The 10-d incubation consisted of a 12:12 light–dark cycle at 20°C; tubes were mixed daily. Contents of each culture tube were filtered (1.2-µm glass fiber filter), frozen (-70°C), and extracted overnight in 90% acetone. Centrifuged supernatant was analyzed for chlorophyll a (Welschmeyer, 1994) on a precalibrated Turner (Norwood, MA) 10-AU fluorometer. Results for chlorophyll a response are expressed as the ratio of mean treatment chlorophyll a divided by mean control chlorophyll a.
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Among the eight STP methods (Table 2), sodium-hydroxide P always had the largest values. Bray P, Olsen P, and P from iron oxide–impregnated paper yielded lower values, reflecting the relatively mild extractants used in these tests. All STP measurements were significantly correlated with one another (Table 3; r2 = 0.43–0.96, p < 0.05), indicating there is an overlap in the portion of P that can be extracted by various chemical reagents or used by algae. Among the STP methods, Olsen P, Mehlich-III P, sodium-hydroxide P, and P from iron oxide–impregnated paper were closely correlated with one another (r2 = 0.78–0.96, p < 0.001). The four measures all were correlated with chlorophyll a response, with P from iron oxide–impregnated paper showing the best linearity (r2 = 0.88, p < 0.001). This indicates that a good portion of the BAP was captured in these four tests.
|
|
0.97; Fig. 3b, Table 4). We recognize that the Langmuir model does not fully capture the complicated nature of soil P sorption, but it is a valuable tool in providing sorption parameters, such as and b, for comparing a set of soils tested under uniform conditions.
|
|
|
|
|
Sodium-hydroxide P, which has been used to estimate BAP by Sharpley et al. (1991) and others, explained 81% of the variation in SRP release to runoff. A threshold effect was observed; SRP in runoff increased only after the soil sodium-hydroxide P exceeded approximately 40 mg kg-1 (Fig. 4c). For discussion purposes, we define the soil P level that, when exceeded, results in detectable SRP in runoff, as the "threshold level." The threshold level was much smaller for Mehlich-III P and Olsen P (about 12 and 8 mg kg-1, respectively), apparently due to the generally smaller values for these two soil tests. Although NaOH, which is effective in removing Fe- and Al-bound P, extracted more P from the soils, it did not predict runoff SRP levels as well as Mehlich-III P and Olsen P.
The increase in scatter from Olsen P to sodium-hydroxide P (cf. Fig. 4b and 4c) is due primarily to two data points with sodium-hydroxide P values of 81 and 82 mg kg-1 (St. Peter Low and St. Peter Mid, respectively). These two samples had low pH and calcium carbonate contents compared with the other soils (Table 1). The acidic nature of these two soils, which may reduce the effectiveness of NaOH by partially neutralizing it, probably led to the deviation of these data from the trend line for runoff SRP versus soil sodium-hydroxide P.
Bray P was given special consideration in this study because it is the primary extractant used by farmers and consultants to measure soil P for fertilizer P recommendations in the Minnesota River basin. Figure 4d and Table 5 show that compared with the other STP methods, Bray P was least effective in explaining the variability in runoff SRP. Kuo (1996) suggested that the Bray P extractant, 0.03 M NH4F and 0.025 M HCl, is not suitable for highly calcareous soils because they neutralize the acidic extractant and form CaF2, which rapidly immobilizes P. The two soils contributing most to the spread of data points about the linear trend line in Fig. 4d (Morris West Mid and Morris East Up; Bray P of 1 and 15 mg kg-1, respectively) had the highest CCE contents among the 10 soils, 10.5 and 9.1%, respectively (Table 1). The ineffectiveness of the Bray test to measure available P in highly calcareous soils apparently is the reason that a close-fitting relationship was not obtained between runoff SRP and Bray P.
Among the soil P sorption capacity indicators, EPC best explained the variation in SRP, accounting for 89% of the variance (Fig. 5). As Eq. [2] shows, EPC is a function of the intrinsic P sorption capacity (), sorption energy (b), and actual P sorbed by a soil sample (o). Equilibrium P concentration values calculated from the Langmuir model thus reflect the key factors (, b, and o) that determine P sorption and desorption. Because the EPC corresponds to the P concentration of a soil solution in equilibrium with the soil, it is reasonable that the EPC should influence runoff SRP.
Desorbability also provided fairly good estimates of runoff SRP (r2 = 0.75; Fig. 5a). The lower fit compared with Mehlich-III P, Olsen P, and other STP methods (except Bray P) may reflect a different nature between the P extracted by these STP methods and the P targeted by the desorbability test. Mehlich-III P, Olsen P, and other STP methods are typically used to extract P sorbed by soil particles before the collection of field samples. In our study, "desorbability" measured the release of P added to the soils as free phosphate in the laboratory in a time span of 48 h (24 h for sorption and 24 h for desorption). The SRP in runoff from the rainfall simulations came from the same type of P as the STP methods measured. Consequently, Mehlich-III P, Olsen P, water P, and sodium-hydroxide P were more closely related to runoff SRP than desorbability. Although the parameters for EPC were obtained under the same experimental conditions as desorbability, EPC captures more information on the intrinsic sorption–desorption capacity of a soil (because it is a function of o and the parameters of the Langmuir model) and thus is a better predictor of runoff SRP than is desorbability.
Although both P sorption saturation indices, PSIm and PSIs, were correlated with runoff SRP (Fig. 5c,d), PSIs had a much higher r2 (0.88) and significance level (p < 0.001); PSIm accounted for only 50% of the variance. Soils with high sorption maxima often have low sorption energy; that is, some soils are capable of holding more P (have more sorption sites), but the binding can be relatively weak, which means adsorbed P can be lost more readily. Inclusion of the sorption energy parameter b in PSIs (see Eq. [2–4]) provided useful information to explain the variation in SRP released during rain events. The PSIm is similar to the P sorption saturation in Sharpley's (1995a) runoff simulations, except that Sharpley used Mehlich-III P for the extractable P (Eq. [3]). His P sorption saturation accounted for 86% of the variance in SRP release.
Combining the sorption energy term and sorption maximum yields a more reliable indicator for the sorption capacity of a soil. For example, compare Soils A and B in Fig. 5c and 5d, which represent the (A) St. Peter Mid and (B) St. Peter Low soil samples. Soil A has a PSIm greater than that for Soil B (41 versus 36%), but runoff from Soil A had a lower SRP (96 µg L-1) than that from Soil B (126 µg L-1). Soil A, however, has a much higher sorption energy (b = 3.00 x 10-3 L µg-1) than Soil B (b = 2.44 x 10-3 L µg-1), leading to a lower PSIs for Soil A (13.5) than for Soil B (14.7), and this resulted in a much-improved fit for these samples. As Fig. 5d also shows, there was little release of SRP from soils with PSIs values < 10%. As mentioned previously, a soil sodium-hydroxide P of 40 mg kg-1 was the threshold level for measurable levels of SRP to appear in runoff (Fig. 4c). These results imply that the soil P sorption binding energy is different (lower) for P sorbed beyond a PSIs of 10% and a sodium-hydroxide P of 40 mg kg-1, which results in the release of desorbable P. This release is presumably related to saturation of the high-energy binding sites on soil particle surfaces (Hooda et al., 2000).
Runoff Bioavailable Phosphorus and Particulate Phosphorus
Most of the predictor variables explained less of the variance in runoff BAP levels (measured as P from iron oxide–impregnated paper) than they did for runoff SRP (Table 5). In runoff simulations, large quantities of soil particles detached by raindrop impact and running water were washed into collected runoff. Runoff BAP includes the part of PP that can be extracted from soil particles in runoff by the iron oxide–impregnated filter paper. The lower r2 values for regressions of runoff BAP versus the predictor variables imply that this PP is only partially accounted for by the STP measurements and sorption capacity indicators of the soils. Nevertheless, the r2 values from Mehlich-III P, water P, EPC, and PSIs are still greater than 80% (Table 5). This indicates that in spite of the addition of a small portion of PP into the BAP pool, these STP methods and soil P sorption capacity indicators still provide reasonable estimates of runoff BAP levels. The good fit (r2 = 0.83, p < 0.001) that PSIs provided shows again that it is a suitable indicator to estimate runoff P loss.
Two exceptions to the general trend of poorer prediction of runoff BAP than SRP are P from iron oxide–impregnated paper (r2 = 0.78) and the chlorophyll a response of the soils (r2 = 0.77) (Table 5). Because runoff BAP was measured by the same method as the P from iron oxide–impregnated paper of the soils, it is not surprising that runoff BAP was well correlated with soil P from iron oxide–impregnated paper. Algae, including the diatom S. hantzschii used in this study, acquire P from both the solution and soil particles. Because iron oxide–impregnated paper absorbs P from both the solution and the soil particles, and because of the excellent agreement between runoff P from iron oxide–impregnated paper (BAP) and runoff algal bioavailable P (Sharpley, 1993), chlorophyll a response predicted runoff P from iron oxide–impregnated paper better than runoff SRP alone. The large and unanticipated difference in correlation coefficients for Olsen P (0.95 for SRP versus 0.77 for BAP; Table 5) probably reflects the inadequacy of the Olsen method in extracting adsorbed BAP from calcareous soil particles.
Runoff PP was approximated by subtracting runoff SRP from runoff TP. The nonreactive soluble portion of TP, which generally is small (Sharpley et al., 1991; Sonzogni et al., 1982), thus was assigned to runoff PP. Results indicated that PP contributed the major portion of runoff TP (59–98%, all but one value > 85%). This range is wider than that (75–95%) reported by Sharpley et al. (1994a). Figure 6 shows that runoff PP was determined largely by runoff TSS, as one would expect.
|
Implications
Results from this study can be used to identify a critical range for STP levels or soil sorption capacity indicators to lower the risk of excessive P losses from calcareous soils to surface waters. The major pathway for transport of P into streams and lakes is by overland flow in runoff and eroded sediment. Losses of P by leaching through the soil and entry into subsurface tile-drainage systems generally are considered much smaller than losses by runoff and erosion because sorption by soil limits P transport by the former pathway. Exceptions can occur on soils with very shallow ground water and sandy or high organic matter soils having a history of excessive P fertilizer and manure applications (Sims et al., 1998). Because of the generally high clay content of soils in the Minnesota River basin (Table 1), we assume here that P losses via leaching and subsurface tile drainage are negligible for the purpose of identifying critical ranges of STP and soil sorption capacity indicators. A study by Randall and coworkers (cited by Sims et al. [1998]) of tile-drain effluent from a flat Webster clay loam soil (fine-loamy, mixed, superactive, mesic Typic Endoaquoll) fertilized with dairy manure and urea in Minnesota also supports this assumption: SRP was detected in only one of 35 samples from manured plots, and TP was detected in only 7% of the samples from manured and urea-fertilized plots.
Eutrophic conditions due to excessive input of P can occur in streams at TP levels of about 75 to 100 µg L-1 (Brezonik et al., 1999; Dodds et al., 1998). Under conditions common in the Minnesota River basin, overland runoff is typically diluted by a factor of six by baseflow and subsurface tile drain flow (Davis et al., 2000). The critical TP concentration in runoff thus is approximately 450 to 600 µg L-1. If we assume that all this P is SRP, the corresponding STP levels of Mehlich-III P, Olsen P, and sodium-hydroxide P at which runoff SRP of 450 to 600 µg L-1 would occur are approximately 90 to 115, 55 to 70, and 120 to 145, respectively, based on results in Fig. 4a–c and Table 5. (Sodium-hydroxide P is included here because it has gained increasing acceptance as a surrogate for BAP.) In addition, a runoff SRP of 450 to 600 µg L-1 corresponds to soil EPC and PSIs values of 500 to 630 µg L-1 and 27 to 34, respectively. These values are at the high end of values for both STP levels and sorption capacity indicators in Minnesota River basin soils (Fang, unpublished data, 2002).
For river eutrophication management, it would be more realistic to modify the above analysis by accounting for BAP (particulate bioavailable P plus SRP), rather than just SRP. Bioavailable P values are more applicable because they reflect the risk of a soil to produce runoff and erosion containing the type of P that directly contributes to eutrophication of receiving waters. For a BAP level of 450 to 600 µg L-1, the corresponding critical STP levels for Mehlich-III P and Olsen P are 65 to 85 and 40 to 55 mg kg-1, respectively (Table 5). Despite the fact that Bray P is a widely used STP extractant in the Minnesota River basin, Bray P was not nearly as good as other STP methods in predicting runoff BAP (Table 5). Nevertheless, consultants are unlikely to stop using the Bray P extractant in Minnesota in the short-term future. In addition, extensive amounts of Bray P data were collected in the past, and these data could be useful in estimating soil P loss potential. Therefore, we report here that the critical Bray P level is 35 to 50 mg kg-1 (Table 5). For sorption capacity indicators, corresponding critical values of soil EPC and PSIs are 360 to 480 µg L-1 and 20 to 26, respectively.
Relationships established between runoff P and STPs using soil runoff boxes may not closely reflect edge-of-field losses of BAP due to the many transport factors and field characteristics that cannot be captured in laboratory studies. Our analysis of critical STP levels given above implicitly assumes dilution of BAP with tile drain effluent containing no phosphorus, but does not account for a sediment delivery ratio or for long-term changes in bioavailability of the particulate phosphorus. Typical sediment delivery ratios for major watersheds in the Minnesota River basin are from 5 to 15% (Roehl, 1962). In the long term, about 66% of the PP in the Minnesota River basin becomes bioavailable (James et al., 2001). If we adjusted for sediment delivery ratios, the critical BAP level in our analysis above would increase. If we adjusted for long-term bioavailability of PP, the critical BAP level would decrease. The two factors dampen the effects of each other. Our results thus provide a simple and rapid reference tool to estimate P pollution potential in agricultural watersheds. This simple approach could be incorporated into more sophisticated tools such as the phosphorus index being adopted by many states (Lemunyon and Gilbert, 1993). The critical levels of STP derived from this study could be used to establish ranges of STP levels corresponding to very low, low, medium, high, or very high phosphorus loss potential ratings in the phosphorus index.
Under the rainfall intensity and soil slope conditions of our runoff simulations, the effective depth of the interaction between soil and rainfall runoff is estimated to be around 5 mm according to Sharpley (1985). Consequently, to replicate field conditions soil samples should be collected and STP levels or P sorption capacity indicators should be measured for only the top soil (<1 cm depth). Better correlations between these parameters and runoff SRP or BAP thus can be established in a runoff simulation study using only the top centimeter of soil. Because samples in this study were collected from a depth of 15 to 20 cm, and because STP levels typically decrease with depth, our critical levels of STP probably represent an upper limit. Eroding surface soils would have higher STP levels than the samples we studied and would produce a greater risk for river eutrophication.
Some states have used critical Mehlich-III P levels ranging from 75 to 200 mg kg-1 and Bray P levels from 100 to 200 mg kg-1 to limit manure applications on fields with high STP levels (Lory and Scharf, 1999). The fact that these values are much higher than the critical Mehlich-III P levels (65–85 mg kg-1) and Bray P levels (35–50 mg kg-1) in our study suggests that implementation of their policies may not result in effective environmental protection against phosphorus losses. More in line with our analysis, some states are adopting a critical Mehlich-III P level of 65 mg kg-1 (Sharpley, 1995b) as a cutoff value in their P indexing procedures for phosphorus loss potential assessment. This value is more consistent with environmental levels of phosphorus that produce eutrophication than are higher values used by many other states.
|
CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Soil P from iron oxide–impregnated paper and chlorophyll a response, both of which account for the bioavailable portion of soil PP, had an equally good or better correlation with runoff BAP than runoff SRP, reflecting the importance of PP in the runoff BAP pool. As a sorption capacity indicator, EPC was correlated well with both runoff SRP and BAP. The PSIm (sodium-hydroxide P divided by sorption maximum) did a poor job of estimating runoff SRP and BAP. On the other hand, PSIs (sodium-hydroxide P divided by sorptivity), by including sorption energy term b, was highly correlated with both runoff SRP and BAP. Runoff PP was strongly associated with runoff TSS.
In spite of having a better theoretical interpretation, PSIs was not a better predictor of runoff P than routine STP methods such as Mehlich-III P. However, our study did not include soil samples from sites other than the Minnesota River basin and soil samples with very high STP values. As a result, the ability of PSIs to estimate runoff P over a broad range of soil types and STP levels was not completely tested here. Sharpley (1995a) and Pote et al. (1996), who used much wider ranges of Mehlich-III P (7–360 mg kg-1 and 54–490 mg kg-1, respectively), showed that Mehlich-III P was not a good predictor of runoff SRP over such broad ranges. To fully test the reliability of PSIs, a sampling scheme with soils from different geographic regions and with a wider range of STP levels should be employed.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. R. Guidry, F. V. Schindler, D. R. German, R. H. Gelderman, and J. R. Gerwing Using Simulated Rainfall to Evaluate Field and Indoor Surface Runoff Phosphorus Relationships J. Environ. Qual., October 27, 2006; 35(6): 2236 - 2243. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J. Penn, G. L. Mullins, L. W. Zelazny, and A. N. Sharpley Estimating Dissolved Phosphorus Concentrations in Runoff from Three Physiographic Regions of Virginia Soil Sci. Soc. Am. J., September 20, 2006; 70(6): 1967 - 1974. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. Miller, E. C. S. Olson, D. S. Chanasyk, B. W. Beasley, F. J. Larney, and B. M. Olson Phosphorus and nitrogen in rainfall simulation runoff after fresh and composted beef cattle manure application. J. Environ. Qual., July 1, 2006; 35(4): 1279 - 1290. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. R. Wright, M. Amrani, M. A. Akbar, D. J. Heaney, and D. S. Vanderwel Determining Phosphorus Release Rates to Runoff from Selected Alberta Soils Using Laboratory Rainfall Simulation J. Environ. Qual., April 3, 2006; 35(3): 806 - 814. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Little, D. R. Bennett, and J. J. Miller Nutrient and Sediment Losses Under Simulated Rainfall Following Manure Incorporation by Different Methods J. Environ. Qual., September 8, 2005; 34(5): 1883 - 1895. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Davis, H. Zhang, J. L. Schroder, J. J. Wang, M. E. Payton, and A. Zazulak Soil Characteristics and Phosphorus Level Effect on Phosphorus Loss in Runoff J. Environ. Qual., August 9, 2005; 34(5): 1640 - 1650. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Fang, P. L. Brezonik, D. J. Mulla, and L. K. Hatch Characterization of Soil Algal Bioavailable Phosphorus in the Minnesota River Basin Soil Sci. Soc. Am. J., June 2, 2005; 69(4): 1016 - 1025. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. A. Vadas, P. J. A. Kleinman, A. N. Sharpley, and B. L. Turner Relating Soil Phosphorus to Dissolved Phosphorus in Runoff: A Single Extraction Coefficient for Water Quality Modeling J. Environ. Qual., March 1, 2005; 34(2): 572 - 580. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. N. Sharpley, R. W. McDowell, and P. J. A. Kleinman Amounts, Forms, and Solubility of Phosphorus in Soils Receiving Manure Soil Sci. Soc. Am. J., November 1, 2004; 68(6): 2048 - 2057. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. J. A. Kleinman, A. N. Sharpley, T. L. Veith, R. O. Maguire, and P. A. Vadas Evaluation of Phosphorus Transport in Surface Runoff from Packed Soil Boxes J. Environ. Qual., July 1, 2004; 33(4): 1413 - 1423. [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 | |||
