| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a USDA-ARS, Dale Bumpers Small Farms Research Center, 6883 South State Highway 23, Booneville, AR 72927-9214 USA
b Dep. of Agronomy, 115 Plant Science, Univ. of Arkansas, Fayetteville, AR 72701 USA
c Jr., USDA-ARS, 115 Plant Science, Fayetteville, AR 72701 USA
d Biosystems and Agricultural Engineering Dep., 128 Agric. Engineering Building, Univ. of Kentucky, Lexington, KY 40546 USA
dpote{at}ag.gov
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMaterials and methodsResults and discussionConclusionsREFERENCES |
|---|
and Mehlich III STP
were not affected by season, but DRP concentration in August runoff
was almost double that in May
, so the resulting correlations were affected. Water-extractable STP from field-moist soil was higher in August
than in May
, and P saturation levels showed a similar trend. Runoff volumes were smaller in August, so season had little effect on mean DRP–mass loss.
Abbreviations: CV, coefficient of variation • DRP, dissolved reactive P • ICP, inductively coupled plasma spectrometer • M3, Mehlich III extraction method for soil P • PMAX, P-sorption maximum • PSI, P sorption index • STP, soil test P
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionConclusionsREFERENCES |
|---|
Yli-Halla et al. (1995) report a study in which they monitored DRP levels in runoff from cultivated field plots for more than two years. They attributed DRP losses in runoff primarily to desorption of STP from surface soil, and they observed that mean DRP concentrations from all plots were consistently higher in autumn runoff than in spring runoff. During the course of our research on runoff from grassland, we also observed that DRP concentrations generally seemed to be higher in runoff samples collected during late summer and early autumn than in runoff samples collected during the spring. As a result, we hypothesized that seasonal changes in field conditions affect the correlation between STP levels and DRP concentrations in runoff. Also, since laboratory research by Miller et al. (1993) had shown that air-drying of surface soil increased the amount of P that was extracted by distilled water, we believed that soil moisture content was probably the primary driving factor for any such seasonal effects. Therefore, we tested our hypothesis by comparing results from a runoff event in the wet part of the growing season (May) to results from a runoff event in the dry season (August). Because the level of water-extractable STP is affected by soil moisture content (Miller et al., 1993), we also felt that it was important to identify any effect that air-drying of soil samples prior to analysis might have on the correlation between STP levels and DRP concentrations in runoff. Thus, our objectives were to test the hypotheses that (i) DRP concentrations in grassland runoff increase as seasonal soil conditions become dry, (ii) that the usual method of air-drying soil samples prior to analyses has a significant impact on the amount of extractable soil P, and (iii) that relationships between STP and runoff DRP concentrations are, therefore, affected by season and soil drying.
|
Materials and methods |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionConclusionsREFERENCES |
|---|
1.3 g cm-1, pH of 5.0, organic matter content of 11 g kg-1, and particle-size distribution of 23% sand, 69% silt, and 8% clay (Pote et al., 1996). Fescue (Festuca arundinacea Schreb.) was established in the plots immediately after their construction in the fall of 1990, and it has been continuously maintained. For this study (1996), all of the plots had a dense stand of fescue that provided virtually uniform 100% ground cover as measured by the line–transect method (Laflen et al., 1981). These plots provided a wide range of STP levels because they were previously used to investigate water-quality effects of pasture fertilization with swine manure, commercial fertilizer, and poultry manure. They had received surface applications of swine, broiler, or caged-layer manure to furnish 0, 220, 440, or 880 kg N ha-1 in 1991, with treatments assigned to plots in a randomized complete block design. These manure application rates supplied 0, 19, 38, 54, 76, 108, 215, or 304 kg P ha-1, depending on the type of manure and the application rate. Some had also received manure (swine, broiler, or caged layer), or commercial inorganic fertilizer to supply 220 kg N ha-1 and 87 kg P ha-1 in 1992. However, at the time of this study, the plots had received no nutrient applications during the previous four years. Thus, the amendments had plenty of time to decompose and equilibrate in the soil prior to our simulated rainfall event.
In order to compare results from different seasons, the 36 plots used in this study were randomly assigned to two groups of 18 plots each. The two groups were treated identically, using the procedures described in the following paragraphs, except that the sampling and runoff event took place in May (wet season) for the first group, but was delayed until August (dry season) for the second group.
Sampling Methods
A representative soil sample was collected from each plot in the group and consisted of a composite of 10 cores (2.54-cm diameter) taken randomly from the top layer of soil (0–2 cm depth), based on work by Sharpley et al. (1978). This provided adequate soil for analysis with minimal damage to the plot surface. Immediately following the soil sampling, a simulator described by Edwards et al. (1992) was used to apply simulated rainfall at an intensity of 75 mm h-1 and to produce 30 min of runoff from each plot. Runoff was sampled manually at 5-min intervals throughout the runoff event beginning 2.5 min after initiation of continuous-flow runoff. For each discrete runoff sample, the volume and time required to collect it were recorded and used to calculate the mean flow rate and total volume of runoff for the 5-min interval. Using these runoff data, the six discrete runoff samples from a given plot were used to construct a composite sample in a flow-weighted manner to represent all of the runoff from that plot. An aliquot of each composite runoff sample was filtered (0.45-µm pore diam.) and stored in the dark at 4°C until analyzed by the molybdenum-blue method for DRP in water samples (Murphy and Riley, 1962). Total DRP–mass losses from each plot were calculated as the plot's total runoff volume multiplied by DRP concentration in the flow-weighted composite runoff sample from that plot.
Each soil sample was placed in a sealed plastic bag and stored in the dark at 4°C until a subsample could be gravimetrically analyzed for water content and the remaining soil sieved (2 mm) to remove larger rock particles and most of the grass thatch material. Each sample was then mixed thoroughly and divided into two subsamples. The first soil subsample from each plot was kept in the dark in a sealed plastic container at 4°C until it could be analyzed, while the remaining subsample was air dried. The amount of each subsample weighed out for analysis was based on its water content in order to assure that equal amounts of soil were being analyzed by each method and for every plot.
Soil Analyses
Each soil subsample was analyzed for extractable P by both Mehlich III (Mehlich, 1984) and distilled water (Pote et al., 1996) methods. The Mehlich III chemical extractant was selected because it is commonly used for STP analysis in soil testing laboratories and was originally developed to assess the amount of P available in the soil for growing plants. The distilled water method extracts less soil P than Mehlich III (Pote et al., 1996), but the former is intended primarily to extract P forms that would be soluble in actual runoff solution. Mehlich III extract was analyzed for P by an inductively coupled plasma spectrometer (ICP) (Model D, Spectro Analytical Instruments, Fitchburg, MA), while distilled water extracts were analyzed colorimetrically by the molybdenum-blue method (Murphy and Riley, 1962).
The P saturation (%) of each soil sample was calculated by using the P sorption index (PSI) method to estimate the maximum amount of P (PMAX) that a given soil can adsorb (Mozaffari and Sims, 1994). The initial STP content (mg kg-1), obtained by either the Mehlich III (M3–PSI method) or distilled water (H2O–PSI method) extractions, was then divided by the PMAX (mg kg-1) and multiplied by 100. In each case, the P saturation was correlated to DRP runoff levels. Details of the methods for determining the PSI and PMAX for each soil sample are given in the following paragraph.
A single-point isotherm was used to obtain the PSI as described by Mozaffari and Sims (1994) by mixing each air-dried soil subsample with a P-sorption solution containing 300 mg P L-1 (1.318 g of KH2PO4 dissolved in distilled, deionized H2O to make 1 L of solution). The PSI was determined by weighing 1.00 g of soil into a 50-mL centrifuge tube, adding 20 mL of 0.0125 M CaCl2 · 2H2O, and adding 5 mL of P-sorption solution to make a combined solution containing 0.01 M CaCl2 and 60 mg P L-1. After two drops of toluene were added to minimize microbial activity, the tubes were capped, shaken for 18 h on a reciprocating shaker (3 oscillations s-1), and centrifuged for 10 min at 27000 g. Each sample was then membrane filtered (0.45-µm pore diam.) and the filtrate analyzed for P by ICP. The PSI can be calculated as X/log PF where:
The PMAX can be estimated by the equation
, given that PMAX < 1400 mg kg-1 (Mozaffari and Sims, 1994).
Statistical Methods
For each STP method, the results were correlated to DRP concentrations in runoff from the plots, a linear regression was developed for each season (May and August), and the sample correlation coefficient (r value) was determined for each correlation. To test for statistical differences between regression lines due to season, a classification variable for time and its interaction with STP were included in the regression model to test for homogeneity of the linear relationships between STP and DRP for May and August.
For the two methods of handling soil samples (field-moist vs. air-dried), STP results from the moist subsample were correlated to STP results from the air-dried subsample of the same plot, and the r value, slope, and y-intercept of the resulting regression line were determined. Results from each sample-handling method were also correlated to DRP concentrations in plot runoff, and regression analysis was used to identify any statistical differences between slopes of the resulting regression lines. To simplify the discussion of results, the term wet is used to refer to field-moist soil samples and the term dry is used to refer to air-dried soil samples.
|
Results and discussion |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionConclusionsREFERENCES |
|---|
,
, and y-intercept = 0.015 in May, and a regression line with
,
, and y-intercept = -0.403 in August, so the two lines were almost identical. This result and its possible cause are included in the discussion of DRP in runoff below. Since Mehlich III results were not significantly affected by air-drying the soil, Mehlich III results from wet soil samples are not discussed further and are not included in correlations of Mehlich III STP to DRP concentrations in runoff (Table 2
and Fig. 2)
.
|
|
|
|
We are not sure what caused the seasonal difference in soil P adsorption, but one possibility is that rapid drying of soil samples taken in May could result in relatively small crystal size and more total surface area for soil compounds that adsorb P. The importance of surface area in P-sorption capacity is well documented (e.g., Kuo and Lotse, 1972; Pena and Torrent, 1984; Goldberg and Sposito, 1984), so it is conceivable that if slow drying of soil during the summer months allows the formation of larger crystals with less surface area, P-sorption capacity might be reduced.
Alternatively, the explanation may be related to the fact that higher microbial populations tend to flourish in the soil during May moreso than under the dry soil conditions of August (Paul and Clark, 1989). For example, microbial oxidation of Fe usually produces ferrihydrite (Chukhrov et al., 1973), a compound that often covers bacterial or algal cells as spherical aggregates (Schwertmann and Taylor, 1989) and is known to have a large surface area for P sorption. If such compounds break down as the soil dries during the summer months, then P-sorption capacity of the soil may be reduced. We hope that future investigations may shed more light on the cause(s) of this seasonal influence on P-sorption capacity of soil.
Runoff
For the May runoff event, total surface runoff from each plot ranged from a low of 6.0 mm to a high of 25.2 mm, with a mean of
. Total runoff was much lower during the August runoff event, ranging from 2.1 mm up to 11.7 mm, with a mean of
. These lower runoff totals in August were not surprising, given that it was a much drier month than May. In fact, just prior to the runoff event, the mean water content of the surface soil was only 4.6% in August compared with 26.9% in May.
Dissolved Reactive Phosphorus in Runoff
In May, the mean DRP concentration in all of the runoff from each plot ranged from 0.28 to 0.86 mg L-1, with a mean value of 0.57 mg L-1 for all of the plots combined. The mean DRP concentrations were much higher in August runoff, ranging from 0.72 to 1.36 mg L-1, and averaging 1.05 mg L-1 for all of the plots. It may seem logical to attribute the lower P concentrations in May runoff primarily to dilution effects that are caused by the higher volume of runoff in May; however, some of our previous research (Pote et al., 1999) has indicated that when other factors are controlled, soils with higher runoff volumes have higher P concentrations in the runoff. Also, Miller et al. (1993) found that equal amounts of distilled water extracted significantly larger amounts of P from air-dried soil samples than from field-moist soil samples. The same effect was observed in this study (Table 1), providing further evidence that runoff extracts smaller quantities of P from moist soil than from dry soil and implying that dilution was not the only process causing P runoff concentrations to be lower in May than in August.
In addition, at least two possible biological reasons seem likely to cause higher concentrations of DRP to be released in August than in May. First, microbial populations generally flourish in the soil during the warm, rainy season, but tend to die off somewhat as the soil becomes dry (Paul and Clark, 1989). Therefore, it seems likely that much of the DRP that is immobilized in microbial tissue during the spring season may become available again as the microorganisms die and decompose during the hot, dry season. This possibility is supported by the fact that Mehlich III extractions of P were not affected by soil drying, while distilled water extractions of STP were increased by soil drying (Table 1). Theoretically, the strongly acidic Mehlich III solution is able to extract P that has been immobilized by living microbial tissue, while distilled water can only extract P that is loosely held on soil surfaces or is readily available from decomposing microorganisms. Thus, the only condition under which distilled water extracted approximately the same amount of STP from both the May and August soil samples was when the samples had been thoroughly air dried (Table 1), thus reducing microbial populations to similar levels for both months. Yet, since the May runoff was extracting P from soil that was far from air-dry (26.9% water content), DRP concentrations in May runoff were much lower than DRP concentrations in August runoff that came from soils that were almost air-dry (4.6% water content).
Secondly, the fescue crop was growing rapidly during the warm, rainy season (May), but it tended to go dormant as conditions became hot and dry (August), leaving large quantities of wilted or dried plant tissue on the soil surface (Sleper and West, 1996). Most of this material did not get included in the soil sample, but Timmons et al. (1970) showed that such crop residues release significant quantities of DRP to runoff. Sharpley (1981) found that the concentration of DRP in plant leachate increased with soil water stress and plant age, and it eventually contributed >50% of the DRP loss in runoff from some crops. Thus, STP analysis of soil samples did not completely account for a major source of DRP that was not as available for May runoff but may have become more available in late summer as the fescue cover crop aged and dried.
Correlations of Soil Test Phosphorus to Dissolved Reactive Phosphorus
The r value, slope, and y-intercept for each STP correlation to DRP in runoff are given in Table 2. Regardless of the season, STP values obtained by each of the methods in this study were significantly correlated (P < 0.01) to DRP concentrations in plot runoff. Statistical comparison of the regression lines for Mehlich III STP correlations to DRP (Fig. 2) showed the lines were parallel. However, the two lines were significantly different (P < 0.05) for May and August because Mehlich III-extractable STP levels remained constant (Table 1), while DRP concentrations were much lower in May runoff than in August runoff.
Since distilled water extracted less STP from field-moist samples than from air-dried samples, regression lines for the correlation of DRP concentrations in runoff to distilled water extractable STP (Fig. 3) are shown for both sample-handling methods in both seasons. All of the regression lines in Fig. 3 have statistically equal slopes (P < 0.05). From air-dried soil samples, distilled water extracted approximately the same amount of STP in May as in August (Table 1). Thus, statistical comparison showed this sample-handling method (like the Mehlich III method) produced parallel regression lines but significantly different correlations (P < 0.05) for May and August, because DRP concentrations were much lower in May runoff than in August runoff. However, from field-moist soil samples, water-extractable STP levels were lower in May than in August, thus following the same trend as DRP concentrations in runoff. Because of these findings, we investigated the possibility that for water-extractable STP from field-moist soil correlated to DRP concentrations in runoff, the May and August data might have the same regression line, thus showing some potential for estimating DRP concentrations in runoff during either season. However, regression analysis showed that the intercepts of the regression lines were statistically different, so the data sets could not be combined into a single regression for both seasons.
|
|
for wet soil; regression line
for dry soil. We had observed a similar trend in our previous work, where STP showed much better correlations to DRP loads in spring runoff studies (Pote et al., 1999) than to DRP loads in late-summer runoff studies (Pote et al., 1996). |
Conclusions |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionConclusionsREFERENCES |
|---|
The practice of drying soil samples prior to analysis increased the amount of soil P extracted by distilled water, but it also gave a stable measurement of soil P that was not influenced by season or changes in soil water content. Although water-extractable STP from air-dry soil and Mehlich III STP did not change significantly from May to August, DRP concentrations in August runoff were almost twice as high as those in May runoff. Thus, a given level of STP produced higher DRP concentrations in August runoff than in May; however, P saturation of the soil (calculated by the PSI method) and water-extractable STP from field-moist soil were higher in August than in May, following the same trend as DRP concentrations in runoff. As a result, the P-saturation (H2O–PSI method) data correlated to DRP in runoff gave statistically the same regression line for May and August, thus allowing the data for May and August to be combined into a single linear regression.
Any of the STP methods used in this study may be potentially useful for estimating DRP concentrations in runoff, but each has seasonal characteristics that should be considered, and some seasonal adjustments may be necessary to obtain the most accurate estimates. For example, the water extract (from air-dry soil) and Mehlich III methods gave consistent STP results regardless of the season, but users need to remember that DRP concentrations associated with these STP results may double during the growing season. For the P-saturation (H2O–PSI) method, it may be possible to establish a single correlation that gives reasonable estimates throughout the growing season, but this method may also require frequent sampling to accomodate changing STP levels during the season.
At this time no conclusions can be drawn concerning the reason that air-dried soil samples collected in May showed a higher P-sorption capacity than those collected in August. However, this fact may be important information for understanding the factors that affect soil P loss in runoff. Accordingly, we hope that future investigations will be conducted to identify the cause(s) of this seasonal influence on P-sorption capacity of soil.
Received for publication December 12, 1997.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and methodsResults and discussionConclusionsREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. B. Owens and M. J. Shipitalo Surface and Subsurface Phosphorus Losses from Fertilized Pasture Systems in Ohio J. Environ. Qual., May 31, 2006; 35(4): 1101 - 1109. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Brauer and G. Aiken Effects of a waste paper product on soil phosphorus, carbon, and bulk density. J. Environ. Qual., May 1, 2006; 35(3): 898 - 902. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. L. Allen, A. P. Mallarino, J. G. Klatt, J. L. Baker, and M. Camara Soil and surface runoff phosphorus relationships for five typical USA midwest soils. J. Environ. Qual., March 1, 2006; 35(2): 599 - 610. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Brauer, G. E. Aiken, D. H. Pote, S. J. Livingston, L. D. Norton, T. R. Way, and J. H. Edwards Amendment Effects on Soil Test Phosphorus J. Environ. Qual., August 9, 2005; 34(5): 1682 - 1686. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. C. Schwartz and T. H. Dao Phosphorus Extractability of Soils Amended with Stockpiled and Composted Cattle Manure J. Environ. Qual., April 20, 2005; 34(3): 970 - 978. [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]
|
|
|
|
|
|
P. D. Schroeder, D. E. Radcliffe, M. L. Cabrera, and C. D. Belew Relationship between Soil Test Phosphorus and Phosphorus in Runoff: Effects of Soil Series Variability J. Environ. Qual., July 1, 2004; 33(4): 1452 - 1463. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. S. Griffin, C. W. Honeycutt, and Z. He Changes in Soil Phosphorus from Manure Application Soil Sci. Soc. Am. J., March 1, 2003; 67(2): 645 - 653. [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 | |||
