SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 2 February 2006
Published in Soil Sci Soc Am J 70:454-463 (2006)
DOI: 10.2136/sssaj2005.0031
© 2006 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Allen, B. L.
Right arrow Articles by Mallarino, A. P.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Allen, B. L.
Right arrow Articles by Mallarino, A. P.
Agricola
Right arrow Articles by Allen, B. L.
Right arrow Articles by Mallarino, A. P.
Related Collections
Right arrow Nutrient Management
Right arrow Phosphorus
Right arrow Soil Chemistry

Nutrient Management & Soil & Plant Analysis

Relationships between Extractable Soil Phosphorus and Phosphorus Saturation after Long-Term Fertilizer or Manure Application

Brett L. Allen and Antonio P. Mallarino*

Dep. of Agronomy, Iowa State Univ., Ames, IA 50011

* Corresponding author (apmallar{at}iastate.edu)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Total soil P (TP), soil-test P (STP), and the degree of soil P saturation are affected by long-term P application but relationships between these measurements need to be established for grain production cropping systems to improve P management guidelines. This research studied these relationships from samples collected from 11 long-term (4–23 yr) Iowa P trials. Mean soil clay content and pH (0- to 15-cm depth) ranged from 171 to 375 g kg–1 and 6.1 to 6.8, respectively, and maximum cumulative P application was 192 to 1098 kg P ha–1. Soil was analyzed for Bray-P1 (BP), Mehlich-3 P (M3P), Olsen P (OP), TP, P sorption index (PSI), and P saturation by STP/PSI and Mehlich-3 extractable P, Al, and Fe (M3sat) indices. Soil-test P increased as P applied increased and declined when P was not applied. Total P increased linearly with increasing BP, M3P, and OP (r = 0.52–0.55), and increases were 1.8, 1.7, and 3.5 mg TP kg–1 per mg STP kg–1 for BP, M3P, and OP, respectively. Usually STP was linearly correlated to M3sat and STP/PSI (r = 0.80–0.94), and M3sat was linearly correlated to STP/PSI (r = 0.86–0.92). Results indicate that STP can approximately estimate long-term effects of P application on TP, and soil P saturation for conditions similar to those in this study, but TP estimates are improved by grouping similar soil series. Further research for a wider range of soils and STP would be useful to better describe relationships between these measurements.

Abbreviations: BP, Bray-P1 • DPSox, soil P saturation index based on ammonium-oxalate extractable P, Fe, and Al molar ratio • M3, Mehlich-3 • M3P, Mehlich-3 P • M3sat, soil P saturation index based on Mehlich-3 extractable P, Fe, and Al molar ratio • OM, organic matter • OP, Olsen P • PSI, P sorption index • STP, soil-test P • TP, total P


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
PHOSPHORUS is an essential nutrient for crop growth, and fertilization sometimes is necessary to achieve optimal levels for crop production. However, P fertilization seldom is needed for crop production in high-testing soils (Jokela, 1992; Webb et al., 1992) and excessive P application can cause eutrophication of surface freshwater resources (Correll, 1998; Sharpley and Rekolainen, 1997). Extensive animal production in Iowa and the Corn Belt results in significant manure P applications to many fields. Applications of fertilizer or manure P in excess of P removal with harvested products have resulted in sharp STP increases in many areas of the USA Corn Belt (Potash and Phosphate Institute PPI, 2001).

Several STP methods, such as BP, M3P, and OP, are routinely used to predict P sufficiency for crops. These tests were designed to extract a fraction of P that correlates well with plant uptake, but not necessarily to correlate well with TP or P loss from fields through surface runoff or subsurface drainage. Environmental P assessment tools, such as the P index, include STP from routine soil test methods among several other field characteristics to rank fields for risk of P loss (Lemunyon and Gilbert, 1993; Mallarino et al., 2002). Furthermore, these tools sometimes use STP directly or implicitly to estimate TP and degree of soil P saturation (Kleinman and Sharpley, 2002; Maguire and Sims, 2002b; Mallarino et al., 2002; Sims et al., 2002). Alternative soil tests have been proposed to predict dissolved P loss from fields. Some tests measure or estimate soil P saturation or P sorption potential. The result of excessive P loading rates is that the P sorption capacity of many soils becomes increasingly saturated, the capacity of the soil to "fix" P is reduced, which increases the rate of P release or desorption. Study of P quantity/intensity relationships based on P sorption/desorption isotherms can be time-consuming, so more practical indices of soil P saturation have been proposed. Two indices commonly used in recent years estimate degree of P saturation from P, Fe, and Al molar ratios after extraction with ammonium oxalate (DPSox) or Mehlich-3 (M3sat) extractants (van der Zee et al., 1987; Khiari et al., 2000). The M3sat index is appealing because the M3 extractant is widely used for routine analysis of P, Al, Fe, and other elements. This index is based on empirical data showing that P saturation is determined mainly by P binding to Fe and Al oxides in noncalcareous soils (Parfitt, 1978). Research has shown that M3sat often is highly correlated to DPSox in acidic or near neutral soils (Maguire and Sims, 2002a; Sims et al., 2002) and slightly less but still highly correlated in alkaline soils (Kleinman and Sharpley, 2002). Another index is derived from STP and PSI calculated from a single-point isotherm (Bache and Williams, 1971). Researchers have used different STP extractants to estimate the P saturation index STP/PSI in different soils (Pote et al., 1999; Westermann et al., 2001; Börling et al., 2004).

Several studies have assessed the effect of long-term P application on STP and crop yield in the U.S. Corn Belt (Webb et al., 1992; Randall et al., 1997a, 1997b; Dodd and Mallarino, 2005). Fewer studies have reported long-term changes of TP. For example, Barber (1979) sampled plots from an Indiana study where up to 54 kg P ha–1 was applied annually for 25 yr and reported that TP (15-cm depth) increased from 400 mg P kg–1 with no P applied to 632 mg P kg–1. Whalen and Chang (2001) summarized results from a 16-yr study in Alberta, Canada, showing TP concentrations (15-cm depth) ranging from 920 mg P kg–1 without P application to 3750 mg P kg–1 with cattle manure applications totaling 5.1 Mg P ha–1. Long-term effects of P application on TP and soil P saturation have not been studied in the Corn Belt. Cattle manure additions to Texas soils over a period of 8 yr increased TP and reduced P sorption compared with soils that were not manured (Sharpley et al., 1984). Börling et al. (2004) found increases in soil P saturation and decreases in PSI (decreased P sorption) in a Sweden soil with more than 30 yr of P application history.

Concerns related to water quality and excess P inputs to soils warrant further study of soil P fractions and their relationships in typical Iowa soils. Relationships between STP, TP, and indices of soil P saturation have not been studied in Iowa. An unpublished review of available Iowa TP data since the 1940s (A.P. Mallarino and B.L. Allen, presentation to Iowa P Index Task Force, 2001) showed a TP range from 400 to near 1000 mg P kg–1 and that the higher TP values usually were associated with higher BP values. Relationships between STP, TP, and indices of soil P saturation are of interest because they can be used to rank fields for risk of particulate and dissolved P loss to water resources. Therefore, the objective of this research was to study relationships between P application rates, STP, TP, and two indices of soil P saturation based on soil samples collected from long-term experiments.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Field Trial Descriptions
Soil samples (0- to 15-cm depth) were collected from selected treatments and plots of 11 Iowa long-term P field experiments. Table 1 shows selected site characteristics, soil series, soil analyses, and the range of cumulative P application rates and estimated P removal with harvest. Soil textures were loam, clay loam, or silty clay loam (clay content ranged from 171 to 355 g kg–1). Most trials, except Trial 11, were managed with corn (Zea mays L.)–soybean [Glycine max (L.) Merr.] rotations, and the soils represented typical agricultural soils of Iowa and neighboring states. Soil P across trials and treatments ranged from 3 to 120 mg kg–1 BP, 4 to 140 mg kg–1 M3P, 1 to 74 mg kg–1 OP, and 305 to 678 mg kg–1 TP. Cumulative P applications across trials and treatments ranged from 0 to 1098 kg ha–1 over periods of 4 to 23 yr. The P concentrations of harvested grain were not measured in these trials. Therefore, P removal estimates in Table 1 assumed average P concentration values recommended by Iowa State University for maintaining STP (Sawyer et al., 2002), which are 5.8 g P kg–1 soybean grain and 2.9 g P kg–1 corn grain.


View this table:
[in this window]
[in a new window]
  		Table 1. Soils, selected soil properties, and summarized information of treatments.

 
Trials 1 through 5 were established in 1994 near the towns of Nashua, Kanawha, Sutherland, Crawfordsville, and Lewis, respectively. Soil samples were collected in October 2001 (after crop harvest) from the three replications of four treatments that were a control that received no P and a rate of 27.4 kg P ha–1 yr–1 (as broadcast triple superphosphate) for both no-till and chisel-plow/disk tillage treatments. Details of crop and soil management practices and partial yield results were published by Bordoli and Mallarino (1998) and Mallarino et al. (1999). Trials 6 and 7 were identical fertilizer P rate experiments established near Boone in 1975 (Trial 6) and near Kanawha in 1976 (Trial 7) for corn–soybean rotations managed with chisel-plow/disk tillage. Soil samples were collected in October 1996 from all replications (three at Trial 6 and four at Trial 7) of eight P treatments (as broadcast triple superphosphate) that were the combinations of two initial treatments (0 and 291.2 kg P ha–1) and four annual rates (0, 11.2, 22.4, or 33.6 kg P ha–1 yr–1). Details of crop and soil management practices as well as partial yield and BP soil-test results were published by Webb et al. (1992) and Dodd and Mallarino (2005). Trial 8 was a long-term P-K fertilization experiment established in 1979 near Nashua for corn–soybean rotations managed with chisel-plow/disk tillage. Soil samples collected in October 2001 were from three replications of three P treatments (0, 22.4, or 44.8 kg P ha–1 yr–1 as broadcast triple superphosphate). Crop yield and BP results were published by Dodd and Mallarino (2005). Plot size was uniform within each trial but ranged from 55 to 110 m2 across Trials 1 to 8.

Trials 9, 10, and 11 evaluated several poultry or liquid swine manure application rates for corn and (or) soybean managed with chisel-plow/disk tillage. The manure rates were based on the total N content of manure from analyses of preliminary samples collected from each storage site. To accurately determine the amounts of nutrients applied, manure samples were collected for analysis while the manure was being applied (several loads from the manure applicator were needed for each experiment). Therefore, the actual amount of P applied with each treatment varied within a site and over time due to variability in nutrient content of manure, and the range of cumulative P applied over time is shown in Table 1. For all trials the manure was applied before chisel-plowing and disking by broadcasting poultry manure or injecting (10- to 15-cm depth) liquid swine manure. Trial 9 was established near Boone in 1996 and treatments were a control and two rates of liquid swine manure applied before corn of corn–soybean rotations. Trial 10 was also established near Boone in 1998 and treatments were a control and two rates of poultry manure (from egg layers) applied before corn of corn–soybean rotations. Crop and soil management practices for these two trials and partial results for BP and P concentrations of subsurface tile drainage were summarized by Klatt et al. (2002). At both trials, manure was applied only for corn at 0, 168, and 336 kg N ha–1 of total manure N, respectively, and three replications were used. Soil samples were collected in 2001 from all replications of the treatments. Plot size was uniform within each trial but was 87 m2 in Trial 9 and 0.2 ha in Trail 10. Trial 11 was established near Nashua in 1993 to evaluate fertilizer and manure management systems for continuous corn or corn–soybean rotations. Details of site characteristics, management practices, and crop yields were published by Bakhsh et al. (2001). Soil samples were collected in June 2001 from three replications of three liquid swine manure treatments applied to plots measuring 0.4 ha managed with chisel-plow/disk tillage. Two treatments involved application of a similar manure rate before corn of corn–soybean rotations in the fall, and the third treatment involved manure application in the fall of each year for continuous corn. The manure rates were based on the N needs of corn or estimated N removal with soybean grain and varied over time from 112 to 168 kg N ha–1.

Soil Sampling and Analyses
A composite soil sample was collected from a 0- to 15-cm depth from each plot, and included 12 to 15 cores using a 2-cm diam. probe. Samples were dried at 40°C, ground to pass through a 2-mm sieve, and stored in plastic-lined paper bags before analyses. All tests were conducted in duplicate. To reduce costs, soil organic matter (OM) and particle-size distribution for each trial were measured on a composite sample collected from all replications of the lowest P rate. Organic matter was determined by a combustion method based on the procedure of Wang and Anderson (1998). Soil particle-size distribution was measured as suggested by Walter et al. (1978). Soil pH of all plots was determined with a pH meter in a 1:1 soil/water ratio. Soil-test P was analyzed for BP, M3P, and OP following procedures recommended for the North-Central Region of the USA (Frank et al., 1998) using a colorimetric determination of extracted P based on the method developed by Murphy and Riley (1962). Total P was determined by an alkaline oxidation digestion procedure (Dick and Tabatabai, 1977) adapted to an Al digestion block (Cihacek and Lizotte, 1990) and by measuring P in digests colorimetrically (Murphy and Riley, 1962).

Soil P saturation was estimated by STP/PSI and M3sat indices. The STP/PSI index was calculated following procedures described by Pote et al. (1999) for M3P and by Westermann et al. (2001) for OP. A first step involved calculating PSI for soil from each plot using the single-point P sorption procedure suggested by Bache and Williams (1971) and Sims (2000). Briefly, 20 mL of a 75 mg P L–1 solution was added to 1 g soil, equilibrated on a horizontal shaker for 18 h, filtered through a 0.45-µm pore-size filter, and orthophosphate P was measured colorimetrically (Murphy and Riley, 1962). The PSI was calculated using Eq. [1], where q is the amount of P sorbed (mg kg–1) and C is the equilibrium solution P concentration (mg L–1).

Formula 1[1]

A second step involved dividing STP measured with BP, M3P, or OP by PSI and multiplying by 100. The M3sat index was calculated by dividing M3P (mmol kg–1) by M3 extractable Fe plus Al (mmol kg–1) and multiplying by 100 (Khiari et al., 2000). Mehlich-3 extracted Fe and Al were measured by atomic absorption spectroscopy.

Statistical Analyses
This study derived relationships between P accumulation, TP, STP, and P saturation in soils. As expected because of the long-term P application at several rates, analysis of variance of treatment effects on soil P measurements for randomized, complete-block designs showed significant treatment effects for all trials and results are not shown. Correlation analysis and linear or quadratic regression analyses (SAS Institute, 2000) were used to study relationships between the variables. Quadratic equations are shown only when the quadratic term was significant (P ≤ 0.05) after the linear term. Data used in these analyses were means of two duplicate laboratory analyses for each sample collected from soil of each field treatment and replication. The linear coefficients of relationships between measurements were compared across trials using the GLM procedure of SAS combined with a Bonferroni test (SAS Institute, 2000).


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Relationships between Cumulative Phosphorus Application and Soil Phosphorus
Soil-Test Phosphorus
Soil P measured with the three soil P test methods and TP increased with increasing P application rate in all trials, although the nature of the relationship depended on the P treatment and trial. Comparisons of initial STP results from samples collected from the trial areas before P applications started (using BP, not shown) with results in Fig. 1 indicate that BP declined in plots that received no P and increased in plots that received the largest P rate. Furthermore, comparisons of P applied and estimates of P removal with harvest in Table 1 indicate that P applied with the largest rates always was larger than P removal. The positive balance was small for Trials 1 through 5 (10–53 kg P ha–1) and larger in other trials (154–571 kg P ha–1). The actual positive balance probably was even larger because recent grain P concentration data from other Iowa trials (A. P. Mallarino and M. M. Barbazan, Iowa State University, unpublished, 2004) indicate that recommended P concentrations for STP maintenance in Iowa (Sawyer et al., 2002) used for P removal estimates in Table 1 are within the upper 20% range of measured concentrations.


Figure 1
View larger version (29K):
[in this window]
[in a new window]
  		Fig. 1. Relationship between cumulative P application and soil-test P measured with three routine tests for trials with more than 20-yr histories. Curvilinear trends are shown only when the quadratic term was significant (P < 0.05) after a linear term. Different letters by the regression lines indicate significantly different linear coefficients between trials for each soil P test.

 
Figure 1 shows results for trials with long P application histories (>20 yr) and for which curvilinear relationships (P < 0.05) were observed (Trials 6, 7, and 8). In these trials the STP rate of increase was greater as the P rate increased. The departure from linearity was small for all trials and P tests, however, and R2 values increased only by <0.07 over those for linear trends. The STP change per unit P applied for M3P and OP was highest for Trial 8 (Kenyon and Readlyn soils), lowest for Trial 6 (Nicollet and Webster soils), and intermediate for Trial 7 (Webster soil). The BP changes per unit P applied ranked similarly to those for M3P and OP across trials, but the differences between trials were not statistically different at P < 0.05. Table 2 shows results for trials with short P application histories (≤8 yr) and for which linear relationships were observed, even at trials with more than two P application rates. The linear coefficients of these relationships did not differ between trials.


View this table:
[in this window]
[in a new window]
  		Table 2. Rates of increase of soil-test P, total P, and P saturation indices (0- to 15-cm depth) per unit of cumulative P applied for trials with a P application period shorter than 20 yr.{dagger}

 
A greater rate of STP increase with increasing P application rates for Kenyon and Readlyn soils at Trial 8 might be explained by significantly lower clay and extractable Ca concentrations than for Nicollet and Webster soils (Table 1). Dodd and Mallarino (2005) found that less P fertilizer was needed to maintain a high crop yield level over time at Trial 8 than at Trials 6 or 7. However, the results cannot be explained with certainty because OM was lower and extractable Al or Fe was higher for the Kenyon and Readlyn soils. Phosphorus sorption can increase as clay and extractable Ca, Fe, and Al increase and OM decreases but net effects depend on the relative levels of these properties and other factors (Parfitt, 1978; Huang and Schnitzer, 1986; Mozaffari and Sims, 1994; McDowell and Condron, 2001).

The range of STP increase per unit P applied observed in this study encompassed values reported for other states of the central and western Corn Belt. In a 25-yr experiment in Indiana managed with a corn–soybean rotation, Barber (1979) reported 0.05 mg BP kg–1 per kg P ha–1. Randall et al. (1997b) reported 0.04 mg BP kg–1 per kg P ha–1 for a Minnesota Webster soil where continuous corn was grown for 7 yr followed by 11 yr of a corn–soybean rotation. In a 4- yr study on Illinois soils, Peck et al. (1971) found an increase of 0.11 mg BP kg–1 per kg P ha–1.

Total Soil Phosphorus
Total soil P increased linearly (P < 0.05) with P accumulation at both the long-term trials (Fig. 2 ) and short-term trials (Table 2). The linear coefficients of the relationships did not differ (P < 0.05) between trials and ranged from 0.1 to 0.6 mg TP kg–1 per kg P ha–1 applied. Similar or larger ranges of linear coefficients for this relationship were reported by others (Barber, 1979; Sharpley et al., 1984; Whalen and Chang, 2001; Börling et al., 2004). In an Indiana soil (Barber, 1979), TP in the top 15 cm increased 0.17 mg P kg–1 per kg ha–1 P applied during 25 yr. Different rates of P removal with harvest and P loss from the soil through erosion could explain differences in these relationships across studies.


Figure 2
View larger version (30K):
[in this window]
[in a new window]
  		Fig. 2. Relationship between cumulative P application and Mehlich-3 P saturation index, Mehlich-3 P/PSI (P sorption index), and total P for trials with more than 20-yr histories. Curvilinear trends are shown only when the quadratic term was significant (P < 0.05) after a linear term. The linear coefficients of the relationships for each test did not differ.

 
Soil Phosphorus Saturation Indices
Both P saturation indices (M3sat and M3P/PSI) increased with P accumulation, however, like STP, relationships were curvilinear or linear depending on the extent of P accumulation (Table 2 and Fig. 2). The relationship between M3sat and cumulative P input from long-term trials (Fig. 2) was curvilinear for Trial 6 and linear for Trials 7 and 8. Although not significant at P < 0.05, a clear curvilinear trend is also observed for Trial 7 that suggests a higher rate of soil P saturation as the cumulative P application rate increased. The M3P/PSI index increased with cumulative P input, and the relationship was curvilinear for Trials 6 and 7 and linear for Trial 8. The results for Trials 6 and 7 revealed an increasing rate of soil P saturation change at high P application rates, which coincides with results for M3sat. For other trials (Table 2) both saturation indices increased linearly with applied P but linear coefficients did not differ between trials, and M3sat ranged from 0.01 to 0.03% and M3P/PSI ranged from 0.03 to 0.07% per kg P ha–1 applied. Other research has also shown that long-term P applications increase soil P saturation (Sharpley, 1996, Börling et al., 2004) and decrease soil P sorption (Reddy et al., 1980; Sharpley et al., 1984). As expected, however, rates of saturation change varied greatly likely due to variation in both soil properties and ranges of P application rates used in the studies. For example, Börling et al. (2004) reported that DPSox ranged from 7.7 to 16.8% in soil that received no P and from 9.9 to 33.7% in soil that received annual rates of 24 to 49 kg P ha–1 for up to 31 yr.

Relationships between Soil Phosphorus Measurements
Total Soil Phosphorus and Soil-Test Phosphorus
Total soil P and STP were linearly related to each other (TP increased 1.7 to 3.4, 1.8 to 2.6, and 3.1 to 6.4 mg P kg–1 per mg P kg–1 of BP, M3P, and OP, respectively), and relationships were not significantly different between trials (Fig. 3 and Table 3). However, data in Fig. 3 show a much higher scattering of data points for BP in Trial 7 compared with M3P and OP. Although all r2 values were lower for Trial 7 than for other trials, the relationship for BP was the lowest (r2 = 0.28). This scattering, and mainly some high TP values for low BP values, may be explained by relatively lower P extraction by BP from alkaline, CaCO3 affected soils than for other soils (Sen Tran et al., 1990; Mallarino, 1997). Although average pH for Trial 7 was 6.7 (Table 1), a few plots (9 of 32) tested from pH 7.3 to 7.9. The M3P and OP tests are comparatively much less affected by pH than BP in Iowa soils (Mallarino, 1997). Regression analyses across all trials (Table 3) showed that TP increased 1.8, 1.7, and 3.5 mg P kg–1 per mg kg–1 BP, M3P, and OP, respectively. Others have observed significant correlation between TP and routine soil P tests, although as expected the relationships sometimes differed across soils (Sharpley, 1996; Pautler and Sims, 2000).


Figure 3
View larger version (26K):
[in this window]
[in a new window]
  		Fig. 3. Relationship between soil-test P measured with three routine tests and soil total P for trials with more than 20-yr histories. Statistical tests for differences between trends for each soil test are shown in Table 3.

 

View this table:
[in this window]
[in a new window]
  		Table 3. Relationship between total soil P and three routine P tests for all plots at each trial and across all trials.

 
Calculations from averages across the 11 trials showed that OP was 1.3% of TP for plots with no P application to 5.8% (a 4.5-fold increase) for plots that received the highest P rate. Similar calculations for the other tests were 2.5 to 11.1% for BP and 2.8 to 12.4% for M3P (a 4.4-fold increase for both tests). Sharpley et al. (1984) reported the proportion of TP as BP in the surface 30 cm of a clay-loam Texas soil increased from 4% with no P application to 29% when 2.1 Mg P ha–1 as cattle manure was applied over a 5-yr period. Whalen and Chang (2001) reported that the proportion of TP as OP in the surface 15 cm of a clay-loam Alberta (Canada) soil increased from 13% with no P application to 27% with 5.1 Mg P ha–1 as cattle manure applied over a 16-yr period. The P application rates in these studies were higher than in our study. These results indicate that application of a given P rate would likely result in a greater increase of plant-available P in high-testing soils than in low-testing soils. Also, this implies a higher risk of dissolved P loss from fertilization of high-testing soils compared with low-testing soils because STP often correlates well with dissolved P loss from fields (Pote et al., 1999; Schoumans and Groenendijk, 2000; Sims et al., 2002). Study of relationships between STP and soil TP is also important because TP is directly related to particulate P delivery to surface water supplies (Correll, 1998) and some P indices estimate TP from STP (Jarrell and Bundy, 2002; Mallarino et al., 2002). While updated records for STP are prevalent, those for current measures of soil TP are not, and laboratory tests for TP are more costly than tests for STP. This study showed that the relationship between TP and STP is weak across soils with contrasting properties.

Soil Phosphorus Saturation and Soil-Test Phosphorus
The M3sat index increased linearly (P < 0.05) with increasing STP in all trials (Fig. 4 ; Table 4). For trials with more than 20-yr histories (Fig. 4), M3sat increased 0.10 to 0.16, 0.11 to 0.20, and 0.22 to 0.47% per mg P kg–1 for BP, M3P, and OP, respectively. Across all trials (Table 4), the M3sat rate of increase was similar for BP and M3P (0.12% per mg P kg–1). The M3sat increase per unit BP did not differ between trials, but M3sat increase per unit M3P and OP sometimes differed. For example, the M3sat increase with increasing OP and M3P was greater for Trials 6 and 7 compared with Trials 8, 10, and 11 (Table 4). This difference cannot be explained with certainty. Soils at Trials 6 and 7 had higher clay, OM, and M3 extractable Ca concentrations but significantly lower M3-extractable Fe and Al concentrations than soils at Trials 8, 10, and 11 (Table 1). Therefore, probably OM, Fe, and Al oxide concentrations had a greater impact on P sorption as estimated by M3sat than clay and Ca species in these soils. Organic matter can coat P sorption sites limiting P sorption (Daly et al., 2001; Huang and Schnitzer, 1986; McDowell and Condron, 2001) and P sorption decreases with decreasing Fe and Al oxides (Parfitt, 1978; McDowell and Condron, 2001). However, data in Fig. 4 suggest that the main reason for the difference between P tests across sites is much higher scattering of data points for BP in Trials 6 and 7 compared with M3P and OP. This scattering, and mainly some high M3sat values for low BP values for Trial 7, may be explained by lower P extraction by BP from alkaline, CaCO3 affected soils referred to above. Although M3P extraction is not affected by high pH in Iowa soils (Mallarino, 1997) it is also possible that a P saturation estimate based on extractable Al and Fe may be unreliable in these soils. We did not observe a statistically significant relationship between soil pH and other measurements across or within soils (not shown) probably because pH of the vast majority of the samples was within a narrow range.


Figure 4
View larger version (25K):
[in this window]
[in a new window]
  		Fig. 4. Relationship between soil-test P measured with three routine tests and soil P saturation for trials with more than 20-yr histories. Statistical tests of differences between trends for each soil test are shown in Table 4.

 

View this table:
[in this window]
[in a new window]
  		Table 4. Relationship between P saturation estimated with M3sat{dagger} and soil-test P measured with three routine P tests across all plots of each trial and across all trials.

 
High correlation between M3sat and STP observed across most soils (poor for BP only in Trial 7 having some high-pH plots) and small differences between soils indicate that STP could provide reasonable estimates of M3sat when conditions are similar to those in this study (mainly soil properties and STP ranges). Maguire and Sims (2002a) also reported high linear correlations between M3sat and STP measured with M3P (r = 0.91) for silt-loam and sandy-loam soils. A better knowledge of relationships between STP and M3sat could play an important role for improved P management because some authors have suggested that M3sat could be a better predictor of runoff or leached P than STP alone (Maguire and Sims, 2002a). For example, an environmental threshold of 25% soil P saturation (DPSox) has been suggested for the Netherlands (Schoumans and Groenendijk, 2000) and studies (Kleinman and Sharpley, 2002; Maguire and Sims, 2002a; Sims et al., 2002) have shown linear and high correlation between DPSox and M3sat across many soils. Sims et al. (2002) reported that a M3sat value of 14 reflected a change point after which P concentration in surface runoff increased at a greater linear rate for coarse-textured Delaware soils. Nair et al. (2004) estimated such a change point at 16% M3sat [M3P/0.5(M3 Fe + Al)] for Florida sandy soils based on water-extractable soil P (1:10 soil to water ratio). An M3sat change point for Iowa soils included in our study probably would differ from the more acidic and coarse-textured (sandy to silty-loam) Delaware or Florida soils. Nevertheless, calculations based on means across all our trials suggest that BP, M3P, and OP values of 96, 106, and 49 mg P kg–1, respectively, would reflect a M3sat value of 14%.

The STP/PSI saturation index increased with increasing BP, M3P, and OP (Table 5). Across all trials, STP/PSI increased linearly and correlations were high when STP/PSI was calculated with the three soil tests (r = 0.93–0.94). Results by trial and soil test also showed linear increasing trends, except in Trial 6 for M3P and OP and in Trial 8 for M3P where STP/PSI increased at slightly greater rates at high soil P levels. However, the R2 increase over the linear model was only 0.01 in the three instances and these curvilinear trends were disregarded. Statistical differences between linear coefficients across trials (Table 5) indicated a highest rate of STP/PSI increase for Trials 2 and 7 and lowest for Trial 3 for the three routine P tests. These differences could not be fully explained, although soils at Trials 2 and 7 had higher OM and lower clay and M3-extractable Fe and Al concentrations than at Trial 3 (Table 1). These soil property differences could indicate lower P sorption for soils at Trials 2 and 7 than at Trial 3 and may explain observed greater rates of P saturation with increasing STP for Trials 2 and 7. Although other research reported PSI was highly correlated (r = 0.90) with soil clay content (Mozaffari and Sims, 1994), this correlation was very weak (r = 0.35) in our study probably because of a small range in soil clay content. We also found a weak and barely significant (P < 0.05) negative correlation between soil pH and PSI (r = –0.21). Others (Duffera and Robarge, 1999) have suggested that soil pH should not be expected to have a large or clear effect on P sorption over a narrow soil pH range as that in this study.


View this table:
[in this window]
[in a new window]
  		Table 5. Relationship between the soil-test P/P sorption index (PSI) soil P saturation index calculated using three routine soil P tests and soil-test P across all plots at each trial and across all trials.

 
The two P saturation indices were highly correlated and relationships were generally linear for each trial or departures from linearity were small (Fig. 5 ; Table 6). There were only a few instances in which the two indices differed. For example, greater rates of STP/PSI with increasing M3sat at Trials 8 and 11 compared with Trials 6 and 7 suggest that STP/PSI measured either a smaller portion of P sorbed to Fe and Al oxides or a larger portion of P sorbed to other soil constituents compared with M3sat. The soils at Trials 8 and 11 had higher concentration of M3 extractable Fe and Al and lower clay, OM, and M3 extractable Ca concentrations than soils at Trials 6 and 7 (Table 1). One could expect that Nicollet and Webster soil at Trials 6 and 7, for example, would have significant P sorption on Ca species as suggested by higher pH and M3-extractable Ca than for other soils. Across all trials, correlation coefficients between STP/PSI and M3sat were 0.86, 0.92, and 0.86 for BP, M3P, and OP, respectively. Pote et al. (1999) reported significant (P < 0.01) correlations between DRP in runoff and soil P saturation measured with M3P/PSI (r = 0.92–0.94) and DPSox (r = 0.90–0.93) for three Arkansas Ultisols. However, they doubted the usefulness of these indices as universal predictors of DRP because slopes of regression lines were quite different for the three soils. Westermann et al. (2001) reported a significant correlation (r = 0.55) between OP/PSI and DRP in irrigated calcareous soils. Theoretically, an OP/PSI index could provide better estimates of P saturation in alkaline (CaCO3–affected) soils than M3sat, because P sorption by Ca species would be more significant than in acid or neutral soils, the PSI does not distinguish between sorption mechanisms, and OP usually is recommended as the most appropriate routine P test for these soils. However, results from our study cannot confirm this expectation, probably because only a few plots had alkaline soil.


Figure 5
View larger version (26K):
[in this window]
[in a new window]
  		Fig. 5. Relationship between Mehlich-3 saturation index and P saturation estimated with three indices of soil-test P/PSI (P sorption index) for trials with more than 20-yr histories. Curvilinear trends are shown only when the quadratic term was significant (P < 0.05) after a linear term. Statistical tests of differences between trends for each saturation index are shown in Table 6.

 

View this table:
[in this window]
[in a new window]
  		Table 6. Relationship between soil-test P/P sorption index (PSI) soil P saturation index calculated using three routine soil P tests and M3sat{dagger} for all plots at each trial and across all trials.

 

   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Long-term P application at various rates increased STP measured with three routine tests, TP, and soil P saturation. In trials receiving various P rates for more than 20 yr, STP increased at higher rates as the P applied increased, but trends departed little from linearity. Total soil P always increased linearly as STP increased and correlation coefficients for the three tests within trials ranged from 0.53 to 0.98. Across all soils, however, correlation coefficients were approximately similar for the three tests (r = 0.52–0.55) and TP increased 1.8, 1.7, and 3.5 mg P kg–1 per mg P kg–1 of BP, M3P, and OP, respectively. These results suggest that STP databases could provide approximate estimates of TP but estimates are poorer across all soils.

Soil P saturation estimated by M3sat or STP/PSI indices was positively and linearly correlated with STP across trials (r = 0.80 to 0.94 depending on the P test used for STP/PSI). Increasing rates of soil P saturation as STP increases should be expected at high STP values, but this was not observed probably because STP was not sufficiently high. The M3sat and STP/PSI saturation indices usually were related similarly to other measurements, although they differed in a few instances. For example, STP/PSI increased at a faster rate with increasing STP than M3sat for two soils with higher concentration of M3 extractable Fe and Al and relatively lower clay, OM, and M3 extractable Ca compared with two other soils with opposite properties. The highest M3sat levels measured in this study were near an environmental threshold suggested for the Atlantic region of the USA.

Overall, the results suggest that routine soil P tests can approximately estimate long-term effects of P application on TP, and soil P saturation for the soils and STP ranges included in the study. However, some estimates (such as TP from STP) can be improved by developing relationships for groups of soils with similar chemical and mineralogical properties. Although P application rates, STP, and TP values were representative of large areas of Iowa and neighboring states, further investigation is needed for a wider range of soil chemical properties and P levels.

Received for publication January 24, 2005.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Allen, B. L.
Right arrow Articles by Mallarino, A. P.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Allen, B. L.
Right arrow Articles by Mallarino, A. P.
Agricola
Right arrow Articles by Allen, B. L.
Right arrow Articles by Mallarino, A. P.
Related Collections
Right arrow Nutrient Management
Right arrow Phosphorus
Right arrow Soil Chemistry

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