| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a East Dakota Water Development District, 132 B Airport Drive, Brookings, SD 57006
b Chemistry Department, Southwest Minnesota State University, Marshall, MN 56258
c Water Resources Institute, Agricultural Engineering, Box 2120, Room 211, South Dakota State University, Brookings, SD 57007
d Plant Science Department, Agricultural Hall, Box 2207A, South Dakota State University, Brookings, SD 57007
* Corresponding author (schindlerfr{at}southwestmsu.edu)
Received for publication April 19, 2006.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: RP(<0.45), dissolved reactive phosphorus or orthophosphate • STP, soil test phosphorus • TP, total phosphorus • TP(unf), total phosphorus on a raw unfiltered sample • TP(<0.45), total dissolved phosphorus
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The management of P loss from agricultural land and the use of field rainfall simulation to develop soil test P (STP) and runoff P concentration relationships are well documented (Sharpley, 1995; Pote et al., 1996; Daverede et al., 2003; Kleinman et al., 2004; Tarkalson and Mikkelsen, 2004). Microplot field rainfall simulations, although useful in obtaining P loss information and in providing greater control of landscape variables relative to watershed-scale research, are time consuming, labor intensive, and are often limited by seasonal timeframes. Consequently, assessment of alternative soil P loss evaluation measures is warranted. Indoor rainfall simulations using runoff boxes may prove to be a timelier means of attaining soil P loss information, since they are less labor intensive, provide improved control of confounding variables, and provide researchers the opportunity to evaluate a greater number of soil types without the seasonal limitations associated with in situ field rainfall simulations.
Sharpley (1995) used indoor rainfall simulations and runoff boxes to evaluate the relationship between STP and dissolved P (DP) in runoff of poultry-manure-amended soils collected from southeast Oklahoma. He found that Mehlich-3 P was strongly correlated to DP concentrations in surface runoff (r2 range of 0.90–0.96), but found that a single relationship could not describe the release of P from surface soil, since the slope of the relationship varied among soil types. Although useful in describing the effect of poultry manure on DP loss, and in predicting DP loss as a function of STP, his study did not establish a direct link between field and indoor DP loss to surface runoff. Vadas et al. (2005) was able to show that a single P extraction coefficient relating soil P to dissolved P in runoff could be used across a range of soils with different hydrologic and intrinsic properties and under varying management situations. This proves to be very valuable in terms of model development for evaluating and minimizing P transport from agricultural soils. Kleinman et al. (2004) evaluated surface runoff from grassed field plots (1 m wide by 2 m long) and bare soil boxes (0.2 m wide by 2 m long) at similar slopes and found that packed boxes had greater total P (TP) loss compared with field plots due to decreased water infiltration and greater surface runoff and erosion. They concluded that, despite erosion and hydrological differences, either field plots or packed boxes would provide reasonable estimates of the rate of P loss to surface runoff.
The work of Kleinman et al. (2004) provided valuable information regarding the use of indoor runoff boxes to describe P loss and devise improved P management strategies; however, their work was done on the more weathered Ultisol soils of the northeastern USA and compared surface-protected field plots with bare-surface runoff boxes. Our study supports the National Phosphorus Research Project by providing additional information regarding the use of indoor runoff boxes to describe P loss for Mollisol soils. The objective of the study was to evaluate the efficacy of indoor rainfall simulation to predict total dissolved P concentrations in field runoff for four intermediately weathered Mollisol soils of South Dakota under conventionally tilled, low-residue environments. This study contributes to the field objectives of the National Phosphorus Research Project, which is "to characterize soil test P–runoff P relationships for a representative cross-section of important agricultural soils across all Major Land Resource Areas in the U.S." (National Phosphorus Research Project, 2001).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
30 field sites for each soil series were collected and analyzed for their Olsen-P concentrations. Ten of the 30 sites were selected for rainfall simulation based on their agronomic Olsen-P concentration (i.e., sites ranged from low to very high Olsen-P). Within each soil series, all soil sites had similar slopes (2–3%), topographic characters, and constituted similar cropping systems, i.e., a conventionally tilled corn (Zea mays L.) and soybean [Glycine max (L.) Merr.] rotation. No site received manure or fertilizer P within 9 mo of rainfall simulation.
Rainfall Simulations
Field Plots
In situ rainfall simulations were conducted according to the protocol of the National Phosphorus Research Project (National Phosphorus Research Project, 2001). Metal plot boundaries with dimensions of 2 m long by 1 m wide were installed (in duplicate) at the selected runoff sites to a height of 5 cm above the ground to isolate surface runoff. Each boundary contained runoff collection troughs at the downslope side of the plot area. Vacuum was used to remove runoff from the collection troughs. Deflection shields were installed to prevent source water from entering the collected runoff. All residual plant material was removed from the soil surface after boundary installation, and plots were tilled to a depth of 5 cm with a hand rake to mimic seedbed conditions and because this depth is considered the zone of maximum interaction between surface runoff and soil P. A 2 to 3% slope was maintained for all plot areas.
A rainfall simulator similar to that used by the National Phosphorus Research Project (National Phosphorus Research Project, 2001) was constructed by Joern Inc., (Purdue Univ., West Lafayette, IN), and used to conduct in situ field plot rainfall simulations. The use of this type of rainfall simulator to produce runoff has been evaluated by Sharpley and Kleinman (2003) and has been proven to be an effective measure of dissolved reactive P loss (Sharpley and Kleinman, 2003). A TeeJet 1/2 HH SS 50 WSQ nozzle (Spraying Systems Co., Wheaton, IL) was centered over the plot area and located 3.3 m above the plot surface to approximate terminal raindrop velocity (Kleinman et al., 2004). Rainfall distribution at this height had a coefficient of uniformity >85% within the 4-m2 plot area. The rainfall simulator was fitted with tarpaulins to minimize the effects of wind on raindrop distribution. Source water was applied to plot areas at a rate of 65 mm h–1 (National Phosphorus Research Project, 2001), which, based on a 1-yr 24-h rainfall, approximates a 1- to 1.5-yr return period (South Dakota State Climatologist, personal communication, 2006). Natural rainwater was collected and used as source water due to its low P content (mean total P = 0.027 mg P L–1, n = 51) and low soil flocculative character (as determined by the recommended source water test described in the national protocol [National Phosphorus Research Project, 2001]) relative to South Dakota's ground water.
The source water was periodically analyzed (i.e., when a different lot of collected rainwater was used) for total P on a raw, unfiltered sample [TP(unf)] and ranged from 0.004 to 0.054 mg P L–1 [mean TP(unf) = 0.027 mg P L–1, n = 51]. The very low P values were the result of algal P immobilization during times of water storage. Algae grew on the interior walls of the indoor storage tanks, thus depleting the dissolved P of the source water. To adjust for this variability, initial source water TP(unf) was subtracted from all runoff TP(unf) and TP(<0.45) concentrations before data analyses (Schroeder et al., 2004).
Rainfall simulations were conducted at 1-d intervals for three consecutive days as per the national protocol: Day 1 at antecedent soil moisture, and Days 2 and 3 at field capacity. Rainfall simulations were conducted at antecedent moisture contents to minimize the variability in hydrology among field plots. Average results from only the second and third runoff events were used to assess the in situ soil test–runoff P relationships. Soil moisture contents were monitored and field capacity verified with a Theta Probe type mL2x soil moisture sensor (Dynamax, Houston, TX) coupled with a Type HH2 moisture meter (Delta-T Devices, Cambridge, UK). Runoff collection began at 2.5 min after the start of continuous runoff and was collected in toto for 30 min. The preliminary 2.5-min runoff events were used as a means of rinsing and conditioning the runoff troughs and the composite collection containers. Runoff weights were recorded every 5 min for the 30-min runoff duration.
Following each simulation, composite runoff water was thoroughly mixed to ensure complete sediment suspension and then immediately sampled. Unfiltered and filtered (0.45-µm) samples were kept at –20°C until analysis for their total P content by H2SO4–S2O8 digestion and ascorbic acid reduction, as described in American Public Health Association (1998), by South Dakota State University (SDSU) Analytical Services. The P forms of the unfiltered and filtered samples in this study were classified as particulate plus dissolved P [TP(unf)] and dissolved forms of P only [TP(<0.45)] (Haygarth and Sharpley, 2000). Since we observed no significant differences between the TP(<0.45) and dissolved reactive P or orthophosphate [RP(<0.45)] concentrations of the studied soils, only the TP(<0.45) concentrations are reported here. In addition, since TP(<0.45) is considered the most immediate P fraction contributing to lake eutrophication, and little correlation existed between runoff TP(unf) and STP (attributed to differences in management practices and other edaphic factors), only correlations with the TP(<0.45) fractions are reported.
Following rain simulations, 12 soil core samples (0–5 and 0–15 cm) were collected per duplicate plot area using a 2.0-cm-diameter, stainless steel probe. Composite soil samples were air dried, ground, and sieved to 2 mm. Samples were analyzed by the SDSU Soil and Plant Testing Laboratory according to the recommended chemical soil test procedures for the North Central Region (Brown, 1998). Soil P contents were determined in triplicate using the Olsen (NaHCO3) method (Frank et al., 1998) and pH by 1:1 soil/water slurry (Watson and Brown, 1998). The pipette method was used for particle size analyses and was performed by the SDSU Soil Pedology Laboratory using standard procedures (NRCS, 1996).
Bulk soil surface samples (0–5 cm) were collected from each plot after field simulation, dried in forced air at 35°C, crushed, and sieved to 19 mm for use in indoor rainfall simulations. A laboratory sample was collected, ground, and sieved to 2 mm. Samples were analyzed by the SDSU Soil and Plant Testing Laboratory for Olsen P and pH using the methods as described above for the probe samples. In addition, bulk samples were analyzed for organic matter by loss of weight on ignition (Combs and Nathan, 1998), and pH by the method described by Watson and Brown (1998). Since surface soil–runoff interactions are greatest at soil depths <5 cm (Sharpley, 1995) and the greatest amount of variability in dissolved P is usually explained by STP levels at these depths (Schroeder et al., 2004), all soil test and runoff P relationships reported here were derived from the 0–5-cm soil depth only.
Packed Runoff Boxes
Bulk soils collected from each of the 10 field plots and for each soil series were packed into runoff boxes constructed according to the national protocol (National Phosphorus Research Project, 2001). Runoff boxes were constructed of high-density polyvinyl chloride sheets with dimensions of 1 m long by 20 cm wide by 7.5 cm deep. Each box contained nine drainage holes (5-mm) at the bottom, which were covered with cheesecloth at the time of packing to prevent soil loss. The drainage holes were established according to the national protocol, and were evenly spaced along the rear, middle, and downslope side of the box. For ease and uniformity of packing, source water was added to enough dry soil to achieve a gravimetric moisture content of 0.12 kg water kg–1 dry soil. The premoistened soil was packed into the runoff boxes in four separate 1.25-cm layers for a final soil packing depth of 5 cm. Each layer was levelled to a uniform depth of 6 mm higher than the desired depth and then packed with a wooden tamper to achieve the final depth and hence a uniform bulk density of 1.2 g soil cm–3. The final soil depth of each layer needed to achieve the desired bulk density was monitored with fabricated depth gauges. Depth gauges were made from pieces of 1 by 6 (2.5 by 15 cm) pine board, cut to fit the inside width of the packed boxes, and painted. Each soil was replicated three times and packed boxes were placed on a runoff table with adjustable slope. The table containing nine packed boxes, i.e., three boxes per soil, was centered under the rainfall simulator at a slope corresponding to the field plots. Plexiglas dividers were inserted between boxes to prevent soil splash between boxes. A V-shaped trough was constructed for each box to collect surface runoff, and a deflection shield attached to prevent source water from entering the sample runoff. Runoff from each box was collected in 10-L polycarbonate carboys.
All rainfall simulation setup parameters and implementations including runoff sampling, storage, and analysis were as described above for the field plots. The packed boxes were covered with cut-to-fit fiberglass base furnace filters to protect the soil surface from excessive raindrop impact, saturated by low-intensity rainfall, and allowed to drain overnight before the simulated rainfall. Filters were removed for all simulation events to more closely mimic the low-residue field conditions. Furthermore, since the antecedent runoff of the field plots was subject to inherent soil moisture and hydrologic variably and thus not reported in this study, and in an effort to draw apposite comparisons with field plots, only the average runoff results for the second and third simulation events for the packed boxes are reported here. Surface runoff data for both the packed boxes and field plots in Table 2 are reported in terms of P concentration (mg P L–1) and total mass of P loss per standardized unit area (kg P ha–1) during the 30-min rainfall event.
|
|
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Statistical differences in water quality variables existed between rain simulation method and among soil series (Table 2). Field runoff produced from the Vienna, Kranzburg, and Poinsett soils were similar, whereas the Barnes soil produced more field runoff than the Vienna soil. Runoff amounts from field plots ranged from 2.3 to 2.7 cm and were similar to the field runoff data reported by Pote et al. (1999) on tall fescue (Festuca arundinacea Schreb.)-covered Nella (fine-loamy, siliceous, semiactive, thermic Typic Paleudult) and Linker (fine-loamy, siliceous, semiactive, thermic Typic Hapludult) soils in northeast Arkansas, but slightly more than those reported by Kleinman et al. (2004) on the mixed-grass-covered Hartleton (loamy-skeletal, mixed, active, mesic Typic Hapludult) and Honeoye (fine-loamy, mixed, active, mesic Glossic Hapludalf) soils of the northeastern USA. As with our study, both the Pote et al. (1999) and Kleinman et al. (2004) studies were based on 30-min runoff events with similar rainfall intensities, i.e., 75 mm h–1.
Runoff produced from packed boxes was statistically greater than from field plots among all soil types except Poinsett, which showed no significance at P = 0.06. When packed in boxes, significantly more runoff was produced from the Vienna soil than the other three soils. Runoff produced from packed boxes was greater with both the Barnes and Vienna soils, although the Barnes was not statistically different from the Kranzburg soil. The greater runoff from the packed boxes was presumably not a consequence of impeded infiltration. Kleinman et al. (2004) showed how the drainage design of the packed boxes used in this study, i.e., nine 5-mm drainage holes, does not significantly impede infiltration into sieved soil. Our study shows that the infiltration effects of packed boxes depend on soil type.
The runoff results of our study were similar to those of Kleinman et al. (2004), who reported greater runoff production from packed boxes than field plots. Runoff amounts from the packed boxes in our study ranged from 2.8 to 3.4 cm and were also the same order of magnitude as those reported by Kleinman et al. (2004). Unlike our study, however, which compared residue-free field plots and packed boxes, the results of Kleinman et al. (2004) were based on comparisons between bare and grass-covered soil surfaces for packed boxes and field plots, respectively. Greater surface runoff from low-residue surfaces is expected due to aggregate breakdown from raindrop impact and subsequent surface sealing, and from reduced infiltration resultant from greater runoff velocities (Jones et al., 1994; Potter et al., 1995; Chambers et al., 2000; Thompson et al., 2001; Findeling et al., 2003). Even with residue-free field plots, our study generated greater surface runoff from packed boxes than field plots.
The most likely reason for greater surface runoff from packed boxes than field plots is that, despite the extent of residue cover, soil from field plots maintains a greater amount of macroporosity than soil packed in boxes, and thus is expected to have greater infiltration. We also noted significant earthworm activity among many of the field plots, which presumably impacted the amount of water infiltrating the soil surface. Although the most significant earthworm activity was noted with the Kranzburg soil, field surface runoff amounts for Kranzburg were similar to the Vienna and Poinsett, and were lower than the Barnes soil (Table 2). Consequently, although earthworms play a significant role in soil porosity and water infiltration (Brady and Weil, 2002), their effect is clearly modified by other soil properties (e.g., soil texture and organic matter content).
Since residue cover was not a factor in our study, the greater runoff amounts produced from the packed boxes compared with field plots for the Vienna, Kranzburg, and Barnes (P < 0.01) soils may be explained, in part, by the their lower organic matter contents. There appeared to be an inverse relationship between organic matter content with the amount of and variability in runoff produced from the packed boxes (r2 = 0.75; Tables 1 and 2), and the soils with the lowest organic matter content, i.e., the Vienna, Kranzburg, and Barnes, resulted in the most significant difference in runoff between packed boxes and field plots. Organic matter enhances aggregate stability (Brady and Weil, 2002; Lado et al., 2004; Sasal et al., 2006); therefore, soils that are lower in organic matter would be subject to greater aggregate breakdown during the soil preparatory and packing phases of indoor rainfall simulation. The lower aggregate stability of the Vienna, Kranzburg, and Barnes soils resulting from their lower organic matter contents would cause decreased infiltration due to greater macropore destruction during the crushing, sieving, and packing sequence. Moreover, the lower organic matter contents could have lowered the soils' water-holding capacities, thus resulting in greater surface runoff from the packed boxes. The less discernable difference in runoff between rain simulation methods for the Poinsett soil was probably a combination of its high clay and organic matter and low sand contents.
No statistical differences existed between rain simulation method or among soil series for either TP(<0.45) concentrations or TP(<0.45) loss (Table 2). The lack of significance may be attributed to method and soil series similarities and not to a wide range of values within each soil series since standard deviations were similar. The TP(<0.45) concentrations and loads to surface runoff from field plots and packed boxes ranged from 0.27 to 0.50 and 0.25 to 0.59 mg P L–1 and from 0.06 to 0.13 and 0.08 to 0.16 kg P ha–1, respectively. These values agree closely with RP(<0.45) concentrations and losses reported in other studies. For instance, Kleinman et al. (2004) observed a range of RP(<0.45) concentrations and losses to surface runoff of 0.07 to 0.35 and 0.08 to 0.22 mg P L–1 and 0.01 to 0.06 and 0.03 to 0.05 kg P ha–1 for field plots and packed boxes, respectively. Daverede et al. (2003) reported an average loss of 0.05 and 0.02 kg RP(<0.45) ha–1 under no-till and chisel-plow plot conditions, respectively, when plots were established on a Typic Argiudoll soil and under a corn and soybean rotation. Schroeder et al. (2004) reported a RP(<0.45) concentration range in surface runoff of 0.15 to 0.80 mg P L–1 from pastured soils in the Piedmont region of Georgia. Vadas et al. (2005) reviewed published data representing 31 different soils and reported filterable reactive P concentrations in surface runoff ranging from 0 to 2.0 mg P L–1.
The most significant differences between rain simulation methods appeared in the TP(unf) losses. All soils except Poinsett showed significance in TP(unf) loss between rain simulation methods (P < 0.05). The TP(unf) losses from field plots were similar among the Vienna, Kranzburg, and Barnes soils, but greater for the Poinsett soil. The TP(unf) losses from packed boxes showed no significant differences among soil series. Generally, the packed boxes resulted in greater TP(unf) losses among soils than the field plots (Table 2), but the differences between them were considerably smaller than the differences in TP(unf) losses between residue-protected field plots and bare-surfaced packed boxes as reported in other studies. For example, the data from Kleinman et al. (2004) showed that the average amount of TP(unf) lost from packed boxes was 4.8 times greater than that of the mixed-grass-covered field plots, whereas in our study, the packed boxes resulted in only 1.6 times more TP(unf) lost to surface runoff compared with the field plots. This was expected since our study removed the majority of the surface residue from the field plots before rainfall simulation, thus exposing the soil surface to greater raindrop energy disaggregation processes. The usually high TP(unf) losses of the packed boxes was probably the result of the sieving and packing process, which destroyed the larger aggregates and increased the amount of smaller P-containing separates exposed to runoff (Kleinman et al., 2004).
Total Dissolved Phosphorus and Soil Test Phosphorus Relationships
The 0- to 5-cm bulk and probe soil samples were used in evaluating the surface runoff TP(<0.45) concentration and STP relationships for the indoor and field rain simulation methods, respectively. Figure 1 depicts the linear regressions relating STP (i.e., Olsen-P) with the surface runoff TP(<0.45) concentration for field and packed-box rainfall simulations among the four soil series. For all soils and rain simulation methods, STP was significantly correlated (P < 0.002) to the TP(<0.45) concentration in the surface runoff. Positive, linear relationships between runoff RP(<0.45) and STP have been well documented (Sharpley, 1995; Pote et al., 1996, 1999; McDowell and Sharpley, 2001; Fang et al., 2002; Daverede et al., 2003; Sharpley and Kleinman, 2003; Kleinman et al., 2004; Schroeder et al., 2004; Tarkalson and Mikkelsen, 2004). A greater amount of variation in TP(<0.45) concentration was explained by STP using packed boxes compared with field plot runoff among the Kranzburg, Poinsett, and Barnes soil series. The greater r2 of the packed boxes was probably the result of greater control over the confounding variables typical of field plots (Kleinman et al., 2004).
|
The slope or the extraction coefficient of regression equations estimate the rate of P released to surface runoff (Sharpley, 1995). Slopes ranged from 0.0027 to 0.0040 and from 0.0025 to 0.0059 for the field plots and packed boxes, respectively, when Olsen-P was used as the predictor variable (Fig. 1). The slopes of our study agree closely with other studies. For instance, Pote et al. (1996) studied fescue-covered Captina silt loam soil and reported a regression slope of 0.0088 when Olsen-P was correlated to RP(<0.45). Pote et al. (1999) reported model regression slopes of 0.004 to 0.009 mg P L–1 (r > 0.86) for grassed plots on Ultisols using Olsen-P to predict dissolved reactive P in surface runoff. Schroeder et al. (2004) reported model regression slopes of 0.002 and 0.014 when Mehlich-III-P and deionized-water-extracted P, respectively, were used to regress surface runoff RP(<0.45) from pastured soils in the Piedmont region of Georgia. Vadas et al. (2005) concluded that "a single value for an extraction coefficient relating soil P to dissolved P in runoff can be used across a wide range of soil, hydrology, or management scenarios." Kleinman et al. (2004) concluded that, despite marked variability in hydrologic and erosion properties among soil types, both field plots and packed boxes can be used to estimate P extraction coefficients.
Sharpley (1995) discussed how regression equations that provide similar extraction coefficients for multiple soils could be combined into a single equation to describe the rate of P release to surface runoff. No significant differences in slopes between the field plots or packed boxes were observed in our study (0.03 
P 0.83). Consequently, a regression equation estimating TP(<0.45) concentration in field surface runoff was developed for each soil using TP(<0.45) concentrations from packed boxes as the predictor variable (Table 3). Table 3 shows that TP(<0.45) concentration in surface runoff from field plots and packed boxes were significantly correlated (P 0.003) and that the regression slopes (0.36 < P < 0.98) and intercepts (0.10 < P < 0.77) were not significantly different among soil series. Given the lack of significance of the regression slopes in Table 3, a single relationship was derived to predict field TP(<0.45) concentration in surface runoff using the surface runoff TP(<0.45) concentrations from packed boxes as the predictor variable. The derived equation is
 | [1] |
|
|
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
No statistical differences in either TP(<0.45) concentration or TP(<0.45) loss existed between packed boxes and field plots among soil series; however, TP(<0.45) concentrations and losses were numerically greater for the packed boxes. An exception, however, was the field-plot Barnes soil, which exhibited a greater average TP(<0.45) concentration than its packed-box counterpart. The most significant differences between rain simulation methods appeared in the TP(unf) losses. All soils except Poinsett showed significance in TP(unf) loss between rain simulation methods. The TP(unf) lost to surface runoff from packed boxes was consistently greater for all soil types. This was expected, since the sieving and packing process destroys many of the larger aggregates, increasing the availability of fine particles to runoff.
The TP(<0.45) concentration in surface runoff from field plots can be predicted from the TP(<0.45) concentration in surface runoff from the packed boxes (0.68 < r2 < 0.94) for the studied soils. Since regression slopes (0.36 < P < 0.98) were not significantly different among the soil series, a single relationship was derived to predict the field TP(<0.45) concentration in surface runoff using the surface runoff TP(<0.45) concentration from packed boxes as the predictor variable. The derived equation is: TP(<0.45)Field = 0.7030 TP(<0.45)Indoor + 0.0579.
This study provides evidence of the efficacy of the indoor runoff protocol (National Phosphorus Research Project, 2001) to predict the TP(<0.45) concentration in field surface runoff, and its capacity to serve as a screening tool to identify select soils considered at high risk for P loss. This study contributes to the field objectives of the National Phosphorus Research Project (National Phosphorus Research Project, 2001) by characterizing the STP–runoff P relationships for the dominant agricultural soils of South Dakota representing the Calcic Hapludolls subgroup.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Vadose Zone Journal | Journal of Plant Registrations | ||||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal |
Regression significant at 0.001 < P < 0.002 probability level.