JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 27 October 2006
Published in J Environ Qual 35:2212-2221 (2006)
DOI: 10.2134/jeq2006.0085
© 2006 American Society of Agronomy, Crop Science Society of America, and 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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Casson, J. P.
Right arrow Articles by Ontkean, G. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Casson, J. P.
Right arrow Articles by Ontkean, G. R.
Agricola
Right arrow Articles by Casson, J. P.
Right arrow Articles by Ontkean, G. R.
Related Collections
Right arrow Surface Water Quality
Right arrow Runoff
Right arrow Phosphorus

TECHNICAL REPORTS

Surface Water Quality

Degree of Phosphorus Saturation Thresholds in Manure-Amended Soils of Alberta

Janna P. Cassona,*, D. Rodney Bennetta, Sheilah C. Nolanb, Barry M. Olsona and Gerald R. Ontkeana

a Alberta Agriculture, Food and Rural Development, 100, 5401 1st Avenue South, Lethbridge, AB, Canada T1J 4V6
b Alberta Agriculture, Food and Rural Development, 206, 7000 113 Street, Edmonton, AB, Canada T6H 4Z9

* Corresponding author (janna.casson{at}gov.ab.ca)

Received for publication February 28, 2006.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
The risk of P losses from agricultural land to surface and ground water generally increases as the degree of soil P saturation increases. A single-point soil P sorption index (PSI) was validated with adsorption isotherm data for determination of the P sorption status of Alberta soils. Soil P thresholds (change points) were then examined for two agricultural soils after eight annual applications of different rates of cattle manure and for three agricultural soils after one application of different rates of cattle manure. Linear relationships were found between soil-test P (STP) levels up to 1000 mg kg–1 and desorbed P in the five Alberta soils. Weak linear relationships were also found between STP and runoff dissolved reactive phosphorus (DRP) in three of these soils. Change points for the degree of P saturation (DPS) were detected in four of the five soils at 3 to 44% for water-extractable P (WEP) and at 11 to 51% for CaCl2–extractable P (CaCl2–P). Change points were not found for DPS or runoff DRP. Overall DPS thresholds for the five soils combined were 27% for WEP and 44% for CaCl2–P at a critical desorbable-P value of 1 mg L–1. The corresponding STP levels (44 mg kg–1 for WEP and 71 mg kg–1 for CaCl2–P) are similar to agronomic thresholds for crops grown on Alberta soils. Soluble P losses in overland flow and leaching may be greater in soils with DPS values that exceed these thresholds than in soils with lower DPS values.

Abbreviations: CaCl2–P, calcium chloride extractable phosphorus • DPS, degree of phosphorus saturation • DRP, dissolved reactive phosphorus • PSC, phosphorus sorption capacity • PSI, phosphorus sorption index • STP, soil test phosphorus • WEP, water-extractable phosphorus


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
REPEATED APPLICATIONS of high rates of livestock manure result in P accumulation in soils (Whalen and Chang, 2001; Qian et al., 2004). This accumulation may partially or completely saturate soil P sorption sites, leading to an increase in P leaching into subsurface soil layers (Eghball et al., 1996) and an increase in P transported into surface waters via runoff (Schroeder et al., 2004; Vadas et al., 2005). Losses of P from agricultural land may result in accelerated eutrophication of surface waters (Campbell and Edwards, 2001).

The P sorption capacity (PSC) of soils is a finite characteristic that varies widely according to clay content, clay mineralogy, organic matter content, exchangeable Al, Fe, and Ca concentrations, and soil pH (Tisdale et al., 1993). Soil processes that affect adsorption of applied P to soil are influenced by the current sorption status and the P sorption capacity of a given soil and determine how much P is available for crop uptake or susceptible to loss in runoff or leachate (Hansen et al., 2002). The DPS of soils has been identified as a potential P loss risk indicator because it has a strong relationship with runoff P concentrations (Sharpley, 1995; Sims et al., 2002). Transport of P from soil to water at a given level of STP or DPS is also influenced by the cation status and ionic strength of the aqueous phase (Beauchemin et al., 1996). The DPS is a function of the portion of soil exchange sites that are bound with P (sorbed P) in relation to the number of sites available for P binding (PSC). The DPS is generally defined as

Formula 1[1]
The PSC of soils may be determined from the sorption maxima of adsorption isotherms using different concentrations of P solutions and measurements of the quantity of P sorbed from solution to solid phase at equilibrium (Sharpley, 1995). A single-point PSI has also been developed to replace the time-consuming adsorption isotherms (Bache and Williams, 1971; Mozaffari and Sims, 1994).

As a soil becomes enriched or depleted in P, changes in the concentration in solution (I) and the supporting pool of labile P (Q) occur (Bache and Williams, 1971). The manner in which Q and I change depends on the slope of the sorption isotherm, which expresses the buffering capacity of the soil with respect to P (Mattingly, 1965). Kleinman et al. (2000) proposed a soil chemical approach for determination of soil P sorption thresholds (i.e., change points) related to soil P transfer to waterways. These thresholds are based on P sorption saturation levels in the soil and they delineate a critical soil P loading level above which any added P may be lost more readily via surface runoff or leaching. Extractions of surface soils with water and 0.01 M CaCl2 have been used for estimation of P losses in overland flow or leaching (McDowell and Sharpley, 2001). McDowell et al. (2001a) found that WEP had the strongest relationship with P lost via overland flow and CaCl2–P had the strongest relationship with P lost via leaching. Split-line models have been used to determine thresholds where the relationship between STP or DPS and DRP in runoff or drainage are split into two sections, one with greater P loss per unit soil P than the other (Hesketh and Brookes, 2000; McDowell and Trudgill, 2000). Quantity–intensity (Q–I) relationships such as these have been used to identify change points in several recent studies (McDowell and Sharpley, 2001; Maguire and Sims, 2002a, 2002b; Indiati and Sequi, 2004; Nair et al., 2004).

Soil DPS values from 25 to 40% are generally associated with a greater risk of P losses in leaching or overland flow (Pautler and Sims, 2000). In the Netherlands, the DPS is determined from the content of oxalate-extractable P, Al, and Fe in the soil (Breeuwsma et al., 1995). A DPS of 25% or more has been established as a critical value, above which the potential for P losses through runoff and leaching become unacceptable (Breeuwsma et al., 1995). Hooda et al. (2000) studied the relationship between DPS and P release to solution on a number of soils in the UK. They discovered that little P desorption occurred below a DPS of 20% (by the Dutch method) and P leachate losses increased linearly above this value. In Quebec, DPS is determined from the content of Mehlich-3-extractable P and Al in the soil (Giroux and Tran, 1996). The surface water quality objective for P in Quebec (0.03 mg total P L–1) is lower than the objective in the Netherlands (0.15 mg total P L–1), thus a DPS value of 9% has been proposed as a standard to represent the DPS limit in the A horizon of agricultural soils in Quebec (Sims et al., 1998). In Delaware, the DPS is determined from the content of Mehlich-3-extractable P, Al, and Fe in the soil, and soils above 11% are considered to have above-optimal P levels, while soils above 15% DPS may require remedial action to minimize the risk of nonpoint-source P pollution by runoff and leaching (Sims et al., 2002).

Determination of the P sorption status of soils is dependant on the methods used to measure STP and PSC, and on the equation used to calculate DPS. Acid oxalate extraction is not considered suitable for alkaline or calcareous soils, which occur widely in the Canadian prairie region, because oxalic acid precipitates Ca during the oxalate extraction and changes the pH of the acid buffer when it reacts with CO3 (Loeppert and Inskeep, 1996). Mehlich-3 extractions may also not be appropriate for calcareous soils because NH4F reacts with CaCO3 and forms CaF2, which may precipitate soluble P (Kleinman and Sharpley, 2002), and the amount of dissolved P that might potentially be lost in overland flow from heavily manured soils may be overestimated (Sharpley et al., 2004).

The objectives of this study were to validate and adapt a single-point PSI method for Alberta soils, to examine the P sorption characteristics of five Alberta agricultural soils, and to determine the site-specific soil P thresholds (change points) in these soils.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Phosphorus Sorption Index Validation and Adaptation
Forty-seven air-dried soil samples from a previous study (Wright et al., 2003) were used to compare single-point PSI methods to PSC values determined using adsorption isotherms. The soils were collected from manured and unmanured farmers' fields and from an experiment where variable rates of hog and beef manure were applied. The STP values for these surface soils (0–10 cm) collected throughout Alberta ranged from 15 to 680 mg kg–1. Wright et al. (2003) fitted a number of isotherm equations to their data, including Langmuir and Freundlich equations, but the best fit was an equation referred to as the Alberta Model:

Formula 2[2]
where PSC is the maximum amount of P sorbed by soil (mg kg–1), Peq is the P concentration at equilibrium (mg L–1); and a and b are constants.

A single-point PSI was determined by two methods: a KCl PSI method and a CaCl2 PSI method. The Bache and Williams (1971) method was used to determine the KCl PSI of the soils. Two grams of each soil were shaken with 40 mL of 0.01 M KCl containing 75 mg L–1 inorganic P (KH2PO4) and two drops of chloroform (to inhibit microbial activity) for 18 h on an end-over-end shaker at 20°C. Two grams of each soil were also shaken with 40 mL of 0.01 M CaCl2 containing 75 mg L–1 inorganic P (KH2PO4) and two drops of chloroform for 16 h on an end-over-end shaker at 20°C. The CaCl2 PSI method extraction ratio (1:20 soil/solution) and time of shaking (16 h) were the same as used in the P sorption capacity determinations by Wright et al. (2003). All samples were centrifuged at 27 000g for 10 min at 4°C, filtered (0.45 µm), and aliquots of the filtrate were analyzed for orthophosphate-P by the ammonium molybdate–ascorbic acid method (Murphy and Riley, 1962). The PSI was then defined as

Formula 3[3]
where PSI is the P sorption index (mg kg–1), V is the solution volume (L), S is the soil weight (kg), and X = initial – final solution P (mg L–1).

Degree of Soil Phosphorus Saturation
The DPS of five Alberta soils was assessed using archived soil samples obtained from two small-plot field studies: a manure rate study where solid cattle manure had been applied annually to irrigated soils at rates of 0 (control), 20, 40, 60, and 120 Mg ha–1 (wet weight) for 8 yr (1993–2000) at two sites near Picture Butte, ~25 km northeast of Lethbridge, Alberta (Olson et al., 2003); and a manure rate and tillage rainfall simulation study where solid cattle manure was applied once at various rates of total P at three sites (Ontkean et al., 2006). The manure rate and tillage study was conducted at Agriculture and Agri-Food Canada research centers at Beaverlodge and Lacombe, Alberta, and on privately owned land near Wilson Siding, ~16 km southeast of Lethbridge. Rainfall simulation tests were completed at these three sites immediately after manure application and tillage and again ~1 yr later. The three study sites were cropped with an annual cereal crop under rainfed conditions between the rainfall simulation tests. Soils at all five sites were characterized before manure was applied by measuring soil reaction (pH), the electrical conductivity of a saturated paste extract (Rhoades, 1982a), CaCO3 content (Rhoades, 1982b), and particle-size distribution (Gee and Bauder, 1986). Both experiments were conducted with a randomized block design with five replicates of each treatment for the manure rate study and four replicates of each treatment for the manure rate and tillage study.

Soils at the 8-yr manure rate study sites consisted of neutral to slightly alkaline Typic Haplustolls, with one site predominantly coarse-textured soils and the other site mainly medium-textured soils (Table 1). Fresh cattle manure supplied by three local feedlot operations was weighed in the field and applied to 8 by 16 m plots using a tractor-pulled, rear-delivery manure spreader. Manure was incorporated on the day of application with two perpendicular tillage passes with a double disk. Five samples of manure collected annually during field applications were analyzed for moisture content, total P, and total N, NH4–N, and NO3–N (Table 2). Nitrate-N and available P were determined using a 5:1 solution/soil Miller–Axley extractant (0.03 M NH4F–0.015 M H2SO4 extractant) followed by colorimetric analysis (Miller and Axley, 1956; Technicon Industrial Systems, 1978), and NH4–N was determined using a 5:1 2 M KCl extractant and colorimetric analysis (Maynard and Kalra, 1993). Total P was measured using a HNO3–HClO4 digest and inductively coupled plasma spectrometric analysis (Jones et al., 1991) and total N was determined by the Kjeldahl digestion method (McGill and Fiqueiredo, 1993). Soils at each site were cropped to silage barley (Hordeum vulgare L.) from 1994 to 1998 and in 2000, and to silage triticale (xTriticosecale spp.) in 1999 and 2001 (Olson et al., 2003). Each site was irrigated as required using a solid-set irrigation system. Soil samples were collected annually from each plot in September or October before manure was applied. A composite sample was obtained from two 75-mm-diameter soil cores collected from each plot at increments of 0 to 0.15, 0.15 to 0.3, 0.3 to 0.6, 0.6 to 0.9, 0.9 to 1.2, and 1.2 to 1.5 m. Samples were placed in coolers with ice packs for transportation to the laboratory.


View this table:
[in this window]
[in a new window]
  		Table 1. Soil classification and mean surface soil characteristics before manure application.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Mean characteristics of the solid cattle manure applied.

 
Soils used for the manure rate and tillage study included: acidic, fine-textured Mollic Haplocryalfs; slightly acidic, coarse-textured Typic Argicryolls and Typic Haplocryolls; and neutral, medium-textured Typic and Entic Haplustolls (Table 1). Solid cattle manure, supplied by local feedlot operations, was weighed in the field in garbage cans and applied to 7 by 10 m plots (Ontkean et al., 2006). Manure was spread as uniformly as possible on the plots using garden rakes and was either incorporated on the day of application with one pass of a double disk or was not incorporated. Samples of manure collected from each plot during field applications were analyzed for moisture content, total P (AOAC, 2003a), total N (AOAC, 2003b), WEP (Kleinman et al., 2002), NH4–N, and NO3–N. Moisture and nutrient content were variable in the three different manure sources (Table 2). Treatments for all manure-amended plots were based on the total P content of each manure source ~1 wk before land application. Plots at the Lacombe site received approximately 0 (control), 50, 100, and 200 kg total P ha–1. Manure rates at Beaverlodge were reduced to about half these values due to the low P content of the manure source. The manure total P actually applied at the Wilson Siding site was approximately twice the target rates due to a reduction in the moisture content of the manure between the time of analysis of the manure source and land application of the manure. A composite soil sample was collected with a 19-cm-wide by 50-cm-long by 2.5-cm-deep metal frame and a 2.5-cm-deep scoop from two sampling sites in each plot before each rainfall simulation test (Nolan et al., 2005). Soil samples collected before the second rainfall simulation tests were screened to remove aboveground plant material.

Soil samples used for DPS analysis were from the 0- to 15-cm layer of selected treatments sampled in 2001 in the 8-yr manure rate study and from the 0- to 2.5-cm layer 1 yr after manure had been applied for all the treatments in the manure rate and tillage study. Samples were air dried, sieved (2 mm), and analyzed for STP using the modified Kelowna method (Qian et al., 1991). The modified Kelowna extractant is composed of 0.25 M HOAc, 0.25 M NH4OAc, and 0.015 M NH4F with a measured pH of 4.9 (Qian et al., 1991). Forty milliliters of modified Kelowna extractant was added to 4 g of each soil and the soil suspensions were shaken on a reciprocating shaker for 5 min. The soil extracts were then filtered using Whatman no. 42 filter paper and the extractable P contents were determined colorimetrically (Qian et al., 1991). Measured STP values were also converted to Mehlich-3-equivalent values using Eq. [4] (Wright et al., 2003) for comparison with other studies:

Formula 4[4]
where STPM3P is the Mehlich-3-equivalent STP (mg kg–1) and STPMK is the modified Kelowna STP (mg kg–1). Phosphorus sorption indices were measured using the CaCl2 PSI method described above. The DPS was determined from the ratio of STP to PSI plus STP using Eq. [5] (adapted from Pautler and Sims, 2000; Indiati and Sequi, 2004):

Formula 5[5]

Phosphorus Desorption and Runoff Phosphorus
Calcium chloride extractable P in soils from the five sites was determined by shaking 4 g of each soil sample with 20 mL of 0.01 M CaCl2 for 1 h on an end-over-end shaker at 20°C. Water-extractable P in soils from the five sites was determined by shaking 2 g of each soil with 20 mL of deionized water for 1 h on an end-over-end shaker at 20°C. All samples were centrifuged at 27 000 x g for 10 min at 4°C, filtered (0.45 µm), and aliquots of the filtrate were analyzed for orthophosphate P by the ammonium molybdate–ascorbic acid method (Murphy and Riley, 1962). The CaCl2–P and WEP values were used as proxies for P that could be lost in runoff from these five soils.

Rainfall simulation tests were conducted on each plot of the manure rate and tillage study after harvest of an annual cereal crop in the fall of 2003 at the Beaverlodge and Lacombe sites, and before seeding in the spring of 2004 at the Wilson Siding site. Results from rainfall simulation tests conducted again 1 yr after manure application were used for this study. Four rainfall simulators (Wright et al., 2003) were operated simultaneously on four adjacent plots. Each simulator was fitted with a single Fulljet 1/2 SS HH WSQ nozzle (Spraying Systems Co., Wheaton, IL) centered ~3.05 m above the soil surface above a 1.5 by 2.0 m runoff frame. The simulators were operated at a pressure of ~28 kPa and generated continuous flow at an intensity of 70 mm h–1. Frame borders were constructed of galvanized or painted steel, with top and side plates driven into the soil to a depth of ~10 cm and the front plate was level with the soil surface. Runoff water was collected with a triangular metal tray attached to the front plate and samples were collected within a 0.3-m-deep hole excavated at the lower end of the tray. The collection tray was covered with a 1.2 by 1.8 m sheet of clear Plexiglas, which prevented water from spraying directly onto the tray. Source water used from treated municipal supplies near each site contained negligible concentrations of nutrients. Composite runoff water samples were collected during consecutive intervals at 5, 10, 20, and 30 min after commencement of continuous runoff. The total volume of water collected during each time interval was also recorded and a 1-L sample from each time interval was transported to the laboratory in a cooler with ice packs. A 200-mL subsample was obtained within 24 h with a 0.45-µm membrane filter unit or a 0.45-µm high-capacity filter, and the filtered sample was analyzed for DRP by the ammonium molybdate–ascorbic acid method (Murphy and Riley, 1962). The DRP mass load was calculated by multiplying the DRP concentration by the runoff volume for each interval and by summing the mass loads for the four intervals. A DRP flow-weighted mean concentration was computed by dividing the total DRP mass load for the 30-min interval by the total volume of flow for the same period.

Statistical Analysis
Linear regression analysis was used to evaluate the relationships between PSC values determined using an Alberta PSC model developed with adsorption isotherms (Wright et al., 2003) and the single-point KCl PSI and CaCl2 PSI values measured in our study. Significant differences among STP, PSI, and DPS values for the different manure rate treatments at each site were determined using the Tukey adjustment at the P < 0.05 level. Slope differences between STP and runoff DRP, CaCl2–P, or WEP relationships were determined using PROC MIXED estimate statements at the P < 0.05 level. Quantity–intensity relationships were also examined between STP or DPS and CaCl2–P or WEP for the 8-yr manure rate study, and between STP or DPS and CaCl2–P, WEP, or DRP flow-weighted mean concentration from the 30-min runoff event for the manure rate and tillage study. A PROC NLIN split-line model was used to determine the environmental soil P thresholds (change points). The NLIN procedure required estimation of linear (d + ex) and quadratic (a + bx + cx2) estimation parameters, and solved for the threshold between the linear and quadratic regressions by iterative reevaluation of the equation. All statistical analyses were performed using SAS 8.1 (SAS Institute, 2000). The STP values that corresponded to change points were subsequently determined from the STP–desorbed P relationships.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Phosphorus Sorption Index Validation
A significant relationship was found between the Alberta Model PSC data (Wright et al., 2003) and the CaCl2 PSI and KCl PSI measurements for the same soils (Fig. 1a and 1b). Negative PSI values were detected for some soils with the KCl PSI method (i.e., the KCl extractant may have removed P from sorption sites), whereas negative values were not observed for any of the soils with the CaCl2 PSI method. A stronger relationship was also found between the Alberta Model PSC and the CaCl2 PSI method than the KCl PSI method; therefore, the CaCl2 PSI method was selected for determination of the single-point PSI and the DPS of the soils subsequently examined in the two manure rate studies.


Figure 1
View larger version (22K):
[in this window]
[in a new window]
  		Fig. 1. Relationship between the Alberta Model P sorption capacity (PSC) and (a) the CaCl2–P sorption index (CaCl2 PSI) and (b) the KCl P sorption index (KCl PSI).

 
Eight-Year Manure Rate Experiment
In the coarse- and medium-textured soils at the Picture Butte sites, the PSI decreased significantly and STP increased with increasing rates of manure application (Table 3). Mean PSI values for the highest manure rate decreased significantly by 95.1 mg kg–1 (80%) in the coarse-textured soil and by 254 mg kg–1 (56%) in the medium-textured soil compared with the control. The DPS in these soils also increased significantly as the rate of manure application increased, with DPS values for the highest manure rates >90% in the coarse-textured soil and >70% in the medium-textured soil.


View this table:
[in this window]
[in a new window]
  		Table 3. Treatment means and standard errors (SE) of modified Kelowna soil-test P (STP), P sorption index (PSI), and degree of soil P saturation (DPS) of soils from the 8-yr manure rate study.

 
Change points were not detected using the STP–WEP and STP–CaCl2–P relationships because they were linear (Table 4). When DPS was expressed as a function of STP and PSI, however, the resulting curvilinear relationships with WEP and CaCl2–P enabled determination of change points where slopes increased with increasing DPS. The DPS and WEP were positively related and change points were identified at a DPS of 44% for the coarse-textured soil and at a DPS of 3% for the medium-textured soil. Change points for WEP corresponded with an STP of 57 mg kg–1 in the coarse-textured soil and an STP of 44 mg kg–1 in the medium-textured soil. The DPS and CaCl2–P were also positively related and change points were identified at a DPS of 48% for the coarse-textured soil and a DPS of 11% for the medium-textured soil. These change points corresponded with STP values of 46 mg kg–1 in the coarse-textured soil and 61 mg kg–1 in the medium-textured soil.


View this table:
[in this window]
[in a new window]
  		Table 4. Relationships between modified Kelowna soil-test P (STP) or the degree of P saturation (DPS) with water-extractable P (WEP), CaCl2-extractable P (CaCl2–P), and runoff dissolved reactive P (DRP).

 
Manure Rate and Tillage Rainfall Simulation Experiment
Soil-test P and DPS increased significantly with increasing manure application rates in all three soils and the PSI decreased significantly with higher manure rates in soils at the Lacombe and Wilson Siding sites (Table 5). Mean PSI values for the highest manure rate decreased significantly by 60.4 mg kg–1 (52%) at Lacombe and by 49.3 mg kg–1 (33%) at Wilson Siding compared with the control. The DPS in these three soils increased significantly as the rate of manure application increased and the mean DPS values for the highest manure rates were ~25% in soils at Beaverlodge, whereas mean DPS values for the highest manure rates were >80% in soils at Lacombe and Wilson Siding (Table 5).


View this table:
[in this window]
[in a new window]
  		Table 5. Treatment means and standard errors (SE) of modified Kelowna soil-test P (STP), P sorption index (PSI), and degree of soil P saturation (DPS) of soils from the manure rate and tillage study.

 
Relationships between STP and WEP and between STP and CaCl2–P were linear for all three soils in this study (Table 4). Change-point analysis of soils at the Beaverlodge site was not possible due to the relatively low DPS of these soils. Curvilinear relationships were found between DPS and WEP, and between DPS and CaCl2–P, in soils at Lacombe and Wilson Siding (Table 4). Change points for DPS and WEP were detected at a DPS value of 44% at Lacombe and at 40% at Wilson Siding. These change points correspond to STP values of 96 mg kg–1 at Lacombe and 68 mg kg–1 at Wilson Siding. Change-point analysis of DPS and CaCl2–P yielded change points of 49% at Lacombe and 51% at Wilson Siding, which corresponds to STP values of 103 mg kg–1 at Lacombe and 57 mg kg–1 at Wilson Siding.

Weak, but positive, linear relationships were also detected between STP and runoff DRP for soils at these three sites (Ontkean et al., 2006; Table 4). The slope of the regression equation between STP and runoff DRP was significantly greater than the slope for STP and CaCl2–P, but significantly less than the slope for the relationship between STP and WEP, for soils at Beaverlodge. Slopes for the relationships between STP and runoff DRP were significantly less than the slopes for STP and WEP or CaCl2–P for soils at Lacombe and Wilson Siding. Stronger, curvilinear relationships were found between the DPS and runoff DRP for soils at Lacombe and Wilson Siding; however, change points between DPS and runoff DRP were not found in these soils. The relationship between DPS and runoff DRP was not significant for soils at Beaverlodge.

Overall Change-Point Analysis
Strong linear relationships were found between STP and WEP and between STP and CaCl2–P (Fig. 2a and 2c) and strong curvilinear relationships were found between DPS and WEP and between DPS and CaCl2–P when data from all five sites were combined (Fig. 2b and 2d). The change point for DPS and WEP was observed at 16%, which is equivalent to an STP level of 34 mg kg–1. A change point was also detected at 26% for DPS and CaCl2–P, which equates to an STP level of 27 mg kg–1.


Figure 2
View larger version (24K):
[in this window]
[in a new window]
  		Fig. 2. Relationships between (a) modified Kelowna soil-test P (STP) and water-extractable P (WEP); (b) modified Kelowna-based degree of P saturation (DPS) and WEP; (c) STP and CaCl2-extractable P (CaCl2–P); and (d) DPS and CaCl2–P in five Alberta soils.

 
Change-point analysis following conversion of all STP data to Mehlich-3-equivalent values increased the change point for DPS and WEP to 29%, a Mehlich-3-equivalent STP value of 66 mg kg–1, and increased the change point for DPS and CaCl2–P to 38%, a Mehlich-3-equivalent STP value of 35 mg kg–1.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Soil Phosphorus Saturation
The increase in STP and DPS and the decrease in the PSI with increasing rates of manure application in the Alberta soils investigated are consistent with similar studies on alkaline, calcareous soils in Alberta (Whalen and Chang, 2002), on neutral soils in the UK (Siddique and Robinson, 2003), and on acidic soils in the USA (Reddy et al., 1980; Sharpley et al., 1984; Mozaffari and Sims, 1994; Iyamuremye et al., 1996). These changes have been attributed to organic acids, produced during the mineralization of added organic matter, that compete for P sorption sites, modify surface charge characteristics, or form complexes with Fe, Al, or Ca (Reddy et al., 1980; Iyamuremye and Dick, 1996; Whalen and Chang, 2002). This results in a reduction in the P binding energy in manure-amended soils, and this increases the susceptibility of these soils to P losses (Holford et al., 1997; Whalen and Chang, 2002). Annual applications of high rates of cattle manure for 20 yr have increased soil labile P to as much as 61% of total P in a Lethbridge clay loam (Typic Haploboroll) soil (Dormaar and Chang, 1995). Sharpley et al. (2004) found that P was increasingly precipitated in more soluble Ca-P forms (tricalcium-P and octocalcium-P) in manured soils compared with less soluble Ca-P forms (hydroxapatite) in untreated soils. Runoff DRP concentrations at all levels of STP have been found to be lower in calcareous than in noncalcareous soils (Torbert et al., 2002). They postulated that the presence of free CaCO3 resulted in precipitation of the insoluble minerals hydroxyapatite and fluorapatite, which reduced the soluble P concentration in calcareous soils. Organic C has been shown to inhibit Ca-P precipitation and amorphous and organically complexed Fe and Mn have been shown to contribute to P retention in calcareous soils (Leytem and Westermann, 2003).

Comparison of our DPS results with other studies must be made with caution since the method of STP analysis and the equation for calculation of DPS are different from methods used in most other studies. The modified Kelowna extraction method (Qian et al., 1991) results in slightly lower STP values than the Mehlich-3 method (Eq. [4]; Wright et al., 2003). Calculation of the DPS by the Pautler and Sims (2000) method ensures that DPS values do not exceed 100% since the current STP status of soils is included in the numerator and denominator of Eq. [5]. Calculation of the DPS by this method results in an apparent increase in the PSC of our soils (STP + PSI, the denominator in Eq. [5]) with manure rate (Tables 3 and 4); however, this apparent increase in PSC is related to increased levels of STP with manure rate and not to a dramatic increase in sorption sites. Soluble P in the soil solution and desorbable P from the labile fraction of soil P generally increase with manure rate, even though adsorption sites on the mineral and organic surfaces are satisfied to a greater degree, as indicated by the decreased PSI for the higher manure rates. Relationships between single-point PSI estimates and PSC maxima determined using adsorption isotherms have been improved by taking previously sorbed P into account (Burkitt et al., 2002; Bolland and Allen, 2003). The PSC estimates for the control treatments of each soil in our study were similar to PSC values determined for alkaline soils in Alberta (Whalen and Chang, 2002) and in Manitoba (Ige et al., 2005).

Soil Phosphorus Thresholds
In our study, soil P thresholds (change points) for WEP, CaCl2–P, or runoff DRP with increases in STP were not evident for STP values less than ~1000 mg kg–1 in any of the soils examined because the relationships were linear (Fig. 2a and 2c); however, change points were detected at DPS values of 3 to 44% for WEP and at 11 to 51% for CaCl2–P in four of the five soils investigated. A change point was not detected in the Beaverlodge soil. The low STP and high PSI values associated with this soil suggest there was a large quantity of vacant P sorption sites available for P binding. Manure applied to this soil also had a lower P content; therefore, less manure P was available for binding to P sorption sites compared with the other soils. Howard (2003) defined agronomic thresholds as being a soil P level, as determined by STP analysis, beyond which there is no practical economic or crop response to added P from inorganic or organic fertilizer sources. Howard (2003) reported modified Kelowna STP agronomic thresholds for long-term crop production on a wide variety of soils in Alberta were 50 mg kg–1 for wheat, barley, and pea (Pisum sativum L.), 60 mg kg–1 for canola (Brassica napus L. var. napus), and 70 mg kg–1 of for potato (Solanum tuberosum L.). The STP values that corresponded to the change points detected in the other four soils were similar to the agronomic threshold of 50 to 70 mg STP kg–1 for optimum crop growth in Alberta agricultural soils (Howard, 2003), except for WEP and CaCl2–P at the Lacombe site. Change-point analysis of data from the five soils combined resulted in change points that were approximately half the agronomic threshold for Alberta soils (Fig. 2b and 2d).

Pote et al. (1999) used the ammonium oxalate extraction method and determined DPS values of 20 to 30% in three acidic Arkansas soils. They also described the relationship between DPS and runoff DRP using a quadratic equation [Runoff DRP(mg L–1) = 0.3083 – 0.0353(DPS, %) + 0.0014(DPS, %)2, r2 = 0.87]. Sims et al. (2002) reported a change point of 14% for Mehlich-3 DPS and runoff DRP in northeastern U.S. soils, which is equivalent to 28% as determined by the Dutch calculation method (Breeuwsma et al., 1995). Vadas et al. (2005) recently compared 10 noncalcareous soils from a number of studies, including the work by Pote et al. (1999), and found that runoff DRP increased rapidly at DPS values >12.5%, as determined by an acid ammonium oxalate extraction. This threshold was equivalent to 25% when DPS was computed by the Dutch method. Other studies have found change points between DPS and WEP of 10% (Hooda et al., 2000), equivalent to 20% using the Dutch calculation method, and 16 to 20% (Nair et al., 2004). Change points between DPS and CaCl2–P were observed at 18% in an Italian soil (Indiati and Sequi, 2004); from 26 to 34% in a wide range of soils from the UK, New Zealand, and the northeastern USA (McDowell et al., 2001b); and from 26 to 38% in Florida soils (Nair et al., 2004). Borling et al. (2004) found strong linear relationships between the ratio of Olsen STP to PSI and CaCl2–P and between the ratio of ammonium lactate extractable STP to PSI and CaCl2–P in several Swedish soils. Change points were not observed within the range of STP values of the soils examined (up to 80 mg Olsen P kg–1 or about 190 mg ammonium lactate extractable P kg–1).

Laboski and Lamb (2004) recently determined that a DPS value of about 22% in Minnesota soils corresponded with a critical WEP concentration of 1 mg P L–1. Using a desorbed P target value for our soils of 1 mg P L–1, critical DPS values were 27% for WEP and 44% for CaCl2–P. These critical DPS values correspond to modified Kelowna STP levels of ~44 mg kg–1 for WEP and 71 mg kg–1 for CaCl2–P.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
A single-point CaCl2–PSI method was validated using adsorption isotherm data from a wide range of Alberta soils. This method was subsequently used to determine the P sorption status of five Alberta soils following land application of cattle manure at different rates.

Strong linear relationships were determined between STP and desorbed P (WEP and CaCl2–P) in five Alberta soils. The WEP method generally resulted in higher concentrations of desorbed P than the CaCl2–P method. Change points for DPS were identified in four of the five soils at 3 to 44% for WEP and at 11 to 51% for CaCl2–P using DPS–desorbed P relationships determined in the laboratory. Weak linear relationships were also found between STP and runoff DRP for the three soils used for field rainfall simulation tests. Change points for DPS and runoff DRP were not evident in these three soils. Desorbed P concentrations (WEP or CaCl2–P) determined in the laboratory were generally substantially higher in relation to STP or DPS than runoff DRP concentrations measured in the field.

Analysis of all five soils combined resulted in DPS change points that were equivalent to approximately half the agronomic threshold of 50 to 60 mg kg–1 for crops grown on Alberta soils. Overall DPS thresholds for these five soils, at a desorbable-P target of 1 mg L–1, were 27% for WEP and 44% for CaCl2–P. These critical DPS thresholds correspond to modified Kelowna STP levels of ~44 mg kg–1 for WEP and 71 mg kg–1 for CaCl2–P, which are similar to agronomic thresholds. Soils with DPS values that exceed these thresholds may be more susceptible to soluble P losses in overland flow and leaching.


   ACKNOWLEDGMENTS
 
We wish to thank Toby Entz and Dennis Mikalson for assistance with data analysis; Ki Au, Mark Kadijk, Gyan Mankee, Linda Broderson, and Fawzi Bichai from the Irrigation Branch Soil and Water Assessment Unit in Lethbridge for soil analyses; Bonnie Hofer and Bob Winter for help with presentation graphics; Tony Brierley and Gerry Coen, Agriculture and Agri-Food Canada, for assistance with soil classification; and Mohamed Amrani and Atta Atia for use of soil samples and adsorption isotherm data from the Soil Phosphorus Mobility study.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
J. L. Little, S. C. Nolan, J. P. Casson, and B. M. Olson
Relationships between Soil and Runoff Phosphorus in Small Alberta Watersheds
J. Environ. Qual., July 17, 2007; 36(5): 1289 - 1300.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
C. A. Volf, G. R. Ontkean, D. R. Bennett, D. S. Chanasyk, and J. J. Miller
Phosphorus Losses in Simulated Rainfall Runoff from Manured Soils of Alberta
J. Environ. Qual., April 5, 2007; 36(3): 730 - 741.
[Abstract] [Full Text] [PDF]

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 PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Casson, J. P.
Right arrow Articles by Ontkean, G. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Casson, J. P.
Right arrow Articles by Ontkean, G. R.
Agricola
Right arrow Articles by Casson, J. P.
Right arrow Articles by Ontkean, G. R.
Related Collections
Right arrow Surface Water Quality
Right arrow Runoff
Right arrow Phosphorus

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS