soil temperature regulates phosphorus loss from ... - naldc - USDA

SOIL TEMPERATURE REGULATES PHOSPHORUS LOSS FROM LYSIMETERS FOLLOWING FALL AND WINTER MANURE APPLICATION M. R. Williams, G. W. Feyereisen, D. B. Beegle, R. D. Shannon

ABSTRACT. Applying manure in the fall and winter increases the potential for nutrient loss prior to crop uptake in the spring. In order to minimize the risk of nutrient loss, recommendations are often based on soil temperature, since biological activity has been shown to decrease substantially at temperatures less than 10°C. These recommendations are often targeted toward reducing nitrogen (N) losses; thus, a smaller body of information is available on the fate of phosphorus (P) from fall and winter applied manure. The objective of this research was to determine how soil temperature affects P loss in runoff and leachate, and assess overwinter P losses based on application date and soil temperature. Nitrogen losses are discussed in a separate article. Dairy manure was surface applied to a channery silt loam soil contained in lysimeters at soil temperatures of 15.7°C, 4.8°C, and -1.1°C, which represented early fall, late fall, and winter applications, respectively. Phosphorus losses were determined during a series of rainfall simulations and natural precipitation events from October 2009 through March 2010. Phosphorus losses were significantly influenced by the soil temperature at the time of manure application and first rainfall-runoff event. As the soil temperature decreased, losses of DRP, TDP, and total P increased. Overwinter losses were also significantly impacted by soil temperature. The winter treatment had two times higher total P losses compared to the manure applied during the early fall. Results of this research show that soil temperature is important for determining P losses and that incorporating quantitative tools, such as soil temperature, into manure management plans could enhance P retention and help reduce the risk of overwinter P losses. Keywords. Fall-applied manure, Leachate, Lysimeter, Phosphorus, Runoff, Soil temperature.

P

hosphorus loss to surface and ground waters is a widespread environmental concern, and agricultural practices are often a key contributor of this nutrient loss in watersheds across the northeastern U.S. (e.g., Carpenter et al., 1998; Daniel et al., 1998). In response to impaired water resources, many states have developed and adopted nutrient management guidelines, which include manure handling, storage, and land application, and integrate other conservation measures to reduce

Submitted for review in September 2011 as manuscript number SW 9423; approved for publication by the Soil & Water Division of ASABE in February 2012. Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the USDA. The USDA is an equal opportunity provider and employer. The authors are Mark R. Williams, ASABE Member, Graduate Research Assistant, Department of Agricultural and Biological Engineering, The Pennsylvania State University, University Park, Pennsylvania; Gary W. Feyereisen, ASABE Member, Research Agricultural Engineer, USDA-ARS Soil and Water Management Unit, St. Paul, Minnesota; Douglas B. Beegle, Distinguished Professor, Department of Agronomy, The Pennsylvania State University, University Park, Pennsylvania; and Robert D. Shannon, ASABE Member, Associate Professor, Department of Agricultural and Biological Engineering, The Pennsylvania State University, University Park, Pennsylvania. Corresponding author: Gary W. Feyereisen, USDA-ARS Soil and Water Management Unit, 439 Borlaug Hall, 1991 Upper Buford Circle, St. Paul, MN 55108; phone: 612-625-0968; fax: 651-649-5175; e-mail: [email protected].

phosphorus (P) loss (Sharpley and Moyer, 2000). These guidelines often recommend that producers apply manure in the spring, coinciding with crop nutrient needs. Many producers, however, practice fall and winter manure spreading. Applying manure in the fall and winter is often done for economic and practical reasons (e.g., fewer field operations take place in the fall and winter compared to the spring) (Flemming and Fraser, 2000). Alternatively, applying manure in the fall and winter increases the potential for loss of some portion of the nutrients prior to crop uptake. The majority of research on overwinter nutrient loss from fall and winter applied manure has focused on nitrogen (N) losses. Denitrification (Bremner and Zantua, 1975; Dorland and Beauchamp, 1991; Chantigny et al., 2002) and nitrification (Malhi and Nyborg, 1979; Malhi and McGill, 1982; Cookson et al., 2002) have been reported in agricultural soils at sub-zero temperatures; thus, in some instances, significant overwinter N losses have been reported (Bole and Gould, 1986; Nyborg and Malhi, 1986; Nyborg et al., 1990). In order to minimize the risk of N loss between application in the fall and crop uptake in the spring, university extension publications and industry professionals often recommend applications when daily soil temperatures are cooler, typically below 10°C (Snyder et al., 2000). While numerous studies have examined overwinter N losses, few have investigated the effect of soil temperature and application date on P losses in surface runoff and subsurface leachate. Soil hydraulic properties are affected by

Transactions of the ASABE Vol. 55(3): 871-880

2012 American Society of Agricultural and Biological Engineers ISSN 2151-0032

871

soil temperature (Zuzel and Pikul, 1987; Bachmann et al., 2002) and have been shown to be a primary factor in determining P losses in runoff (e.g., Sharpley, 1995). Additionally, seasonal applications may also affect P leaching losses, but no consensus exists on this issue. Sims et al. (1998) reported that fall applications of swine manure resulted in higher P losses compared to winter applications, but Geohring et al. (2001) suggested that late summer and early fall applications of dairy manure result in lower P losses. The objective of this research, therefore, was to (1) determine how soil temperature affects P losses in surface runoff and subsurface leachate and (2) assess overwinter P losses based on manure application date and soil temperature.

MATERIALS AND METHODS LYSIMETER DESIGN AND OPERATION An experimental system for simulating a natural soil temperature profile, described in detail by Williams et al. (2010, 2011), was used in this study (fig. 1). In September 2009, twenty-four intact soil cores from Klingerstown, Pennsylvania (40° 39′ 39″ N, 76° 41′ 37″ W) were collected. The soil at the site was Leck Kill channery silt loam (fine-loamy, mixed, semiactive, mesic Typic Hapludult), and the surface was covered with corn residue at the time of collection. Each core was collected by driving a schedule 80 PVC pipe (15 cm diameter by 50 cm long) into the soil between two rows of corn stubble with a 1.1 Mg drop hammer. An end cap (schedule 40 PVC) was placed on the bottom of each lysimeter following collection and contained a tapped 1.25 cm diameter hole to which a leachate collection system was attached. In order to suppress sidewall bypass flow during subsequent rainfall simulations, the lysimeters were built after the design of Feyereisen and Folmar (2009). PVC spacers, cut from SDR 35 pipe (0.5 cm thick), were designed to provide a gap between the soil core and the lysimeter wall. The spacers were in place during insertion of the lysimeter into the soil and subsequently removed. The resultant space was backfilled with liquefied petroleum jelly, which creat-

Figure 1. Cross-section of a single lysimeter (left) and the soil thermal cycling system (right): (1) PVC lysimeter, (2) petroleum jelly, (3) perforated PVC bottom, (4) runoff collection, (5) undisturbed soil core, (6) PVC end cap, (7) extruded polystyrene insulation, (8) leachate collection, (9) plywood bin, (10) thermocouple, (11) sand, (12) thermistor, (13) heating cable, and (14) plywood bottom. Only two of the four lysimeters are shown in the plywood bin.

872

ed a watertight seal between the soil column and the lysimeter wall. A 1.6 cm diameter hole was then drilled into the lysimeter wall at or slightly below the soil surface, and a runoff collection system was installed. The lysimeters were randomly divided into six groups of four, and each group was placed in a 61 × 61 × 61 cm plywood bin (fig. 1). Each bin had a 2 cm thick perforated plywood bottom with extruded polystyrene insulation (7.5 cm thick, R-value = 15) on top of the plywood. The walls of the bin were also encased with extruded polystyrene insulation (5 cm thick, R-value = 10). A commercially available electric resistance heating cable (Orbit Radiant Heating, Perkasie, Pa.) was placed in two layers, 2.5 and 7.6 cm, above the insulation on the bottom of each bin in order to create an upward heat flux representative of heat flow under field conditions (fig. 1). Masonry sand was carefully added and packed over the heating cable and around the lysimeters until the bins were filled up to the soil surface. One lysimeter from each bin was fitted with four type-T thermocouples (Culik et al., 1982) at 5, 10, 20, and 30 cm depths below the soil surface. The thermocouples along with two thermistors and an electronic relay that controlled the heating cable were connected to a datalogger, which was programmed to maintain a set temperature at the 40 cm depth. Each lysimeter-bin assembly was placed on a custombuilt cart and pushed outside a USDA-ARS building on the Pennsylvania State University campus from October 2009 through March 2010. The site was level and grass-covered. A tarp canopy was erected over the site to prevent unwanted precipitation from coming in contact with the lysimeters. The temperature of the heating cable in each bin was manually decreased weekly or biweekly from 16°C to 2°C to reflect the 30-year average soil temperature cycle at a depth of 40 cm in State College, Pennsylvania (40° 47′ 29″ N, 77° 51′ 31″ W). Water was added weekly (0.5 cm depth) in order to prevent the soil surface from drying out. MANURE APPLICATION AND RAINFALL SIMULATION The dairy manure used in this study was collected from the Pennsylvania State University Research Dairy. Feces and urine were collected and stored separately in a refrigerator no more than three days prior to application. The feces and urine were mixed at a 1.7:1 ratio one day prior to application and were analyzed for nutrient content (table 1). The lysimeters inside each of the bins were randomly designated as either a control (C) or one of three manure application treatments: early fall (EF), late fall (LF), and winter (W). On October 21, 2009, all of the lysimeter-bin assemblies were pulled inside a temperature-controlled room and the soil temperature was allowed to equilibrate to the change in air temperature in the room (Tair = 16°C) for a one-day period (table 1). On October 22, manure was surface-applied to the EF treatment lysimeter in each bin at a rate of 0.39 g cm-2, which was equivalent to an application rate of 37,400 L ha-1. The application rate was predetermined based on a manure analysis and a 225 kg total N ha-1 recommendation. All of the bins were then individually subjected to a rainfall simulation on October 25 using a

TRANSACTIONS OF THE ASABE

[a] [b]

[c]

Table 1. Treatment, manure application date, air and soil temperatures, and selected characteristics of the dairy manure. Application Tair[b] Organic N Total N Total P2O5 Tsoil[c] Solids NH4-N Treatment[a] Date (kg N kL-1) (kg N kL-1) (°C) (°C) (%) (kg N kL-1) (kg P kL-1) EF October 22 16.0 15.7 9.0 3.78 2.45 6.21 1.21 LF November 17 4.5 4.8 9.9 3.73 2.37 6.11 1.16 W December 15 -6.5 -1.1 9.1 3.18 2.77 5.93 1.13 EF = early fall, LF = late fall, and W = winter. Tair = air temperature that was maintained for a four-day period (one day prior to manure application through the rainfall simulation) inside the temperature-controlled room/freezer. Tsoil = soil temperature at 5 cm depth following one-day equilibration period prior to manure application.

modified protocol of Sharpley et al. (2001). A single nozzle (FullJet 1/2 HH SS 14WSQ, Spraying Systems Co., Wheaton, Ill.) was attached to a frame at a height of 3.05 m above the top of the bin. Each rainfall simulation was one hour at an average rainfall intensity of 3.8 cm h-1. Following the same procedure as the EF treatment, manure was surface-applied to the LF and W treatment lysimeters in each bin on November 17 and December 15, respectively, with all bins receiving rainfall simulations three days after each application (table 1). In order to achieve the desired air temperatures for the late fall and winter applications (Tair = 4.5°C and -6.5°C, respectively), all of the lysimeter-bin assemblies were pushed into a large walk-in freezer (Leer ICE Merchandisers, New Lisbon, Wisc.) one day prior to manure application and maintained there until the rainfall simulation. A final, fourth rainfall simulation was conducted on January 15 (Tair = -6.5°C). The uniformity coefficient among all (6 bins × 4 simulations = 24) of the rainfall simulations was 0.93. The lysimeter-bin assemblies were all stored under the tarp canopy for the three-week periods between the rainfall simulations, where they were subjected to ambient temperature fluctuations. NATURAL PRECIPITATION EVENTS Upon completion of the final set of rainfall simulations in January 2010, the tarp canopy covering the lysimeter-bin assemblies was removed. This exposed the lysimeters to both ambient air temperature and precipitation. From January through March 2010, a total of five events (two snowmelt, three rainfall) produced runoff, leachate, or both. Snowmelt events occurred on February 22 and March 5, 2010, and were the result of 15 cm of snow that fell on February 12. Rainfall events occurred on January 25, March 15, and March 29, 2010, with rainfall depths of 3.5, 3.6, and 1.5 cm, respectively. The soil temperature from mid-January to mid-March at the 40 cm depth was held constant at 2°C in all of the bins. As air temperatures began to increase, soil temperatures at the 40 cm depth were increased biweekly to reflect the change in air temperature. SAMPLE ANALYSIS During both the rainfall simulations and natural precipitation events, runoff and leachate were collected. Water samples were refrigerated at 4°C immediately after collection until analysis. Dissolved reactive phosphorus (DRP) was determined colorimetrically on filtered (0.45 μm) runoff and leachate water. Total dissolved phosphorus (TDP) was determined using ICP spectrometry analysis on filtered runoff and leachate water. Total P was determined by alkaline persulfate digestion (Patton and Kryskalla, 2003) and

55(3): 871-880

ICP spectrometry. Nutrient losses were then calculated by multiplying the concentration of the nutrient by the volume of the sample. Particulate phosphorus (PP) losses were calculated as the difference between total P and TDP. Dissolved organic phosphorus (DOP) loss was calculated as the difference between TDP and DRP. When the lysimeters were excavated, soil samples were collected from 0-15, 15-30, and 30-45 cm depths adjacent to the outside of the lysimeter wall. These samples were used to establish a baseline for the soil P content prior to manure application. At the conclusion of the study, the manure remaining on the soil surface of the EF, LF, and W treatments was removed. Soil samples were then collected from 0-15, 15-30, and 30-45 cm depths within all of the lysimeters. All soil samples were air-dried and sieved (2 mm). Extraction of soil samples was performed using 0.01 M CaCl2 as a measure of plant-available P and P that is more easily transported by water (Self-Davis et al., 2009). Extractions were also performed by shaking 2.5 g of soil with 25 mL of Mehlich-3 solution (0.2 M CH3COOH + 0.25 M NH4NO3 + 0.013 M HNO3 + 0.001 M EDTA) for 5 min (Mehlich, 1984) as a soil P measure including more recalcitrant forms. The final and initial soil samples were then compared to calculate the nutrient level change, thus determining the nutrients that would remain in the soil for potential crop use in the spring. STATISTICAL ANALYSIS The effect of soil temperature on hydrology was determined by calculating the percentage of total water loss (runoff + leachate) that was partitioned to runoff (or leachate). The percentages were then evaluated by an ANOVA using the general linear models procedure (PROC GLM). Soil temperature effects on P loss during the first rainfall simulation after manure application for each treatment were evaluated by an ANOVA using PROC GLM, with results shown as bolded values in table 2. Similarly, soil temperature effects on overwinter P loss from all rainfall simulations and natural precipitation events were evaluated by ANOVA using PROC GLM. Water quality data that were below the analytical detection limit for P (0.01 mg L-1) were assigned a value that was equal to one-half of the detection limit for statistical analysis purposes (Hornung and Reed, 1990). Manure application treatment effects on soil P concentrations were evaluated by an ANOVA using the mixed model procedure (PROC MIXED) with manure treatment as the main effect and depth as the split-plot effect. Pairwise comparisons for all analyses were made using Tukey’s Studentized range (HSD) test in order to separate treatment means. A probability level of 0.05 was used to evaluate statistical significance of treatment effects in all

873

(a)

(b)

Figure 2. Mean soil and air temperatures for all treatments from October 2009 through March 2010: (a) air temperature, and (b) soil temperature at the 5 and 40 cm depths. Solid arrows indicate rainfall simulations. Dashed arrows indicate natural precipitation events that produced runoff, leachate, or both. Asterisks (*) above the dashed arrows indicate snowmelt.

analyses. All statistics were completed in SAS version 9.1 (SAS, 2002).

RESULTS AIR TEMPERATURE, SOIL TEMPERATURE, AND HYDROLOGY The soil thermal cycling system was able to simulate a natural soil temperature profile by controlling the upper and lower boundary conditions (air temperature and soil temperature at the 40 cm depth, respectively) of the intact soil core (Williams et al., 2010). During the study, ambient air temperatures decreased throughout the fall and winter and began to increase at the onset of spring (fig. 2a). Air temperatures ranged from 27°C (Oct. 23) to -21°C (Jan. 27) over this period, and mean monthly temperatures were similar to the 30-year temperature record (1978-2008). The heating cable has been shown to control the soil temperature at the 40 cm depth and produce a natural vertical soil temperature gradient within the soil core (Williams et al., 2010). In this study, the manual adjustment of the heating cable temperature over a seven-month period was also able to provide an accurate simulation of seasonal changes in soil temperature. However, the manual adjustment resulted in step changes rather than the gradual change observed naturally (Johnsson et al., 1995; Hu et al., 2006) (fig. 2b). Soil temperature at the 5 cm depth was approximately equal to the air temperature following the one-day equilibration prior to the October and November rainfall simulations (table 1). Conversely, when the air temperature was maintained at -6.5°C prior to the December and January simulations, the soil temperature at the 5 cm depth reached equilibrium at -1.1 and -0.5°C, respectively. These values correspond to the temperature of the frozen water contained in the soil pores. The partitioning of rain water during the simulations was significantly influenced by the soil temperature at the 5 cm depth (fig. 3). Both the manured and control treatments followed similar trends, although adding manure to the soil surface increased the volume of runoff compared to the control when the soil temperature was greater than 0°C (fig. 3). A decrease in the soil temperature at the 5 cm depth from 15.7°C to 4.8°C resulted in an average increase

874

Figure 3. Runoff and leachate water volume based on soil temperature at 5 cm depth. Bars with the same color and letter are not statistically different at α = 0.05. Error bars represent one standard deviation.

in runoff volume of 12% for all treatments; however, this relationship was not significant due to high variability among the soil cores. Alternatively, significant differences were observed in runoff volume when the soil temperature at the 5 cm depth decreased from 4.8°C to -1.1°C. This resulted in an average increase of 56% in runoff volume for all treatments. The frozen soil (-1.1°C), therefore, resulted in 79% of the water volume becoming runoff for all treatments, which was significantly more than the 23% average for non-frozen soil (15.7°C and 4.8°C). During the five natural rainfall and snowmelt events following the final set of rainfall simulations on January 15, only two events produced runoff. The first event occurred on January 25 in which 3.5 cm of rain fell over the course of the day on frozen soil. The second event was a snowmelt on February 22. The other three events produced exclusively leachate due to their low intensity and unfrozen soil conditions. PHOSPHORUS LOSSES DURING FIRST RAINFALL SIMULATION AFTER MANURE APPLICATION Soil temperature significantly affected the loss of P in both runoff and leachate (table 2). Phosphorus losses in this section are reported in terms of the first rainfall simulation


after manure application for each treatment (table 2, bolded values) in order to assess trends in P loss at the different soil temperatures. Results from the other rainfall simulations as well as the natural precipitation events are presented in the subsequent section on overwinter P losses. Runoff Losses Dissolved P losses for all manured treatments were significantly greater than the control treatment during the first rainfall simulation after manure application (table 2). Mean DRP losses did not differ significantly between the EF and LF treatments, which represented manure application at soil temperatures of 15.7°C and 4.8°C, respectively. Significantly more DRP was lost when the soil temperature was 1.1°C. Observed DRP losses for the W treatment were nearly three-fold greater than the losses for the EF and LF treatments (table 2, bolded values). Dissolved organic P losses were relatively low compared to DRP losses for the manured treatments; however, significantly more DOP was lost in runoff for the LF and W treatments (soil temperatures of 4.8°C and -1.1°C, respectively) compared to the EF treatment (15.7°C). Total dissolved P losses in runoff significantly increased as the temperature of the soil decreased (table 2, bolded values), with DRP comprising 86%, 50%, and 78% of TDP losses at soil temperatures of 15.7°C, 4.8°C, and -1.1°C, respectively. Particulate P losses in runoff during the first rainfall after manure application were also significantly different among manured treatments (table 2). Particulate P losses accounted for 62%, 65%, and 41% of the total P losses for the EF, LF, and W treatments, respectively, which represented manure application at 15.7°C, 4.8°C, and -1.1°C (table 2). Total P losses for all manured treatments were 8 to 20 times greater than the control treatment during the first rainfall-runoff event after manure application (table 2, bolded values). There were significant differences among manured treatments as well. Similar to the trend found for

[a]

[b]

DRP, the W treatment had significantly more total P losses compared to the EF and LF treatments. Leachate Losses Leachate losses of P were significantly greater for the EF and LF treatments compared to the control during the first rainfall simulation after manure application. However, there were no observed differences between the EF and LF treatments (table 2, bolded values). Dissolved reactive P losses were less than the analytical detection limit (0.01 mg) for both the EF and LF treatments. Losses of DOP in leachate were similar between EF and LF treatments and comprised 99% of the mean losses of TDP (table 2). Both the EF and LF treatments also had significantly more losses of PP and total P in leachate compared to the control; however, there were no differences observed between the manured treatments (table 2, bolded values). Particulate P losses accounted for 82% and 78% of the total P losses for the EF and LF treatments, respectively. The W treatment did not have any leachate losses due to the frozen soil conditions (-1.1°C) at the time of the first rainfall simulation after manure application. OVERWINTER PHOSPHORUS LOSSES Losses of P in runoff and leachate from the rainfall simulations and natural precipitation events are shown in tables 2 and 3, respectively, on an event-by-event basis. Cumulative overwinter losses are presented in table 4. Runoff Losses Dissolved reactive P losses in runoff during the rainfall simulations following manure application and natural precipitation events did not decrease significantly for the EF and LF treatments compared to the first rainfall simulation with manure (tables 2 and 3). During each event, DRP losses ranged from 0.21 to 0.53 mg and from 0.16 to 0.58 mg for the EF and LF treatments, respectively. Losses of DRP

Table 2. Mean P losses in runoff and leachate during the rainfall simulations.[a] Mean Runoff Loss (mg) Mean Leachate Loss (mg)[b] C EF LF W C EF LF W Analyte Date DRP Oct. 25 0.07 a 0.53 b 0.04 a 0.06 a BD BD BD BD Nov. 20 0.02 a 0.21 b 0.58 b 0.02 a BD 0.06 a BD BD Dec. 18 0.06 a 0.34 b 0.54 b 1.87 c n/a n/a n/a n/a Jan. 15 0.05 a 0.30 b 0.44 b 1.60 c n/a n/a n/a n/a DOP Oct. 25 0.00 a 0.08 b 0.00 a 0.00 a 0.01 a 0.13 b 0.01 a 0.01 a Nov. 20 0.01 a 0.10 b 0.58 c 0.01 a 0.02 a 0.10 b 0.13 b 0.02 a Dec. 18 0.01 a 0.08 b 0.30 bc 0.53 c n/a n/a n/a n/a Jan. 15 0.01 a 0.08 b 0.20 bc 0.68 c n/a n/a n/a n/a TDP Oct. 25 0.07 a 0.61 b 0.04 a 0.06 a 0.01 a 0.13 b 0.01 a 0.01 a Nov. 20 0.03 a 0.31 b 1.16 c 0.03 a 0.02 a 0.16 b 0.13 b 0.02 a Dec. 18 0.07 a 0.42 b 0.84 bc 2.40 d n/a n/a n/a n/a Jan. 15 0.06 a 0.38 b 0.64 b 2.28 d n/a n/a n/a n/a PP Oct. 25 0.12 a 0.99 b 0.10 a 0.07 a 0.00 a 0.60 b 0.02 a 0.00 a Nov. 20 0.04 a 0.06 a 2.11 d 0.10 a 0.06 a 0.04 a 0.47 b 0.05 a Dec. 18 0.08 a 0.12 a 0.20 a 1.68 c n/a n/a n/a n/a Jan. 15 0.17 a 0.12 a 0.11 a 0.10 a n/a n/a n/a n/a Total P Oct. 25 0.19 a 1.60 bc 0.14 a 0.13 a 0.01 a 0.73 b 0.03 a 0.01 a Nov. 20 0.07 a 0.37 ab 3.27 c 0.13 a 0.08 a 0.20 a 0.60 b 0.07 a Dec. 18 0.15 a 0.58 b 1.04 b 4.08 d n/a n/a n/a n/a Jan. 15 0.23 a 0.50 b 0.75 b 2.38 c n/a n/a n/a n/a C = control, EF = early fall, LF = late fall, and W = winter. Means for all dates within an analyte section and loss pathway (Runoff and Leachate) and followed by the same letter are not significantly different at α = 0.05. Bold values indicate the first rainfall simulation after manure application. BD denotes below detection, and n/a denotes that there was not any leachate collected during the rainfall simulation.

55(3): 871-880

875

[a]

Table 3. Mean P losses in runoff and leachate during the natural precipitation events.[a] Mean Runoff Loss (mg) Mean Leachate Loss (mg) C EF LF W C EF LF W Analyte Date DRP Jan. 24 0.03 a 0.28 b 0.16 b 0.97 c 0.02 a 0.04 a 0.05 a 0.06 a Feb. 22 0.03 a 0.34 b 0.33 b 0.34 b 0.01 a 0.01 a 0.01 a 0.01 a Mar. 5 n/a n/a n/a n/a 0.01 a BD BD 0.01 a Mar. 15 n/a n/a n/a n/a BD BD BD BD Mar. 29 n/a n/a n/a n/a BD BD BD BD DOP Jan. 24 0.00 a 0.02 a 0.03 a 0.06 a 0.01 a 0.04 a 0.04 a 0.09 a Feb. 22 0.00 a 0.04 a 0.05 a 0.07 a 0.00 a 0.01 a 0.00 a 0.01 a Mar. 5 n/a n/a n/a n/a 0.00 a 0.01 a 0.01 a 0.01 a Mar. 15 n/a n/a n/a n/a 0.01 a 0.01 a 0.01 a 0.02 a Mar. 29 n/a n/a n/a n/a 0.01 a 0.01 a 0.01 a 0.01 a TDP Jan. 24 0.03 a 0.30 b 0.19 b 1.03 c 0.03 a 0.08 b 0.09 b 0.15 c Feb. 22 0.03 a 0.38 b 0.38 b 0.41 b 0.01 a 0.02 a 0.01 a 0.02 a Mar. 5 n/a n/a n/a n/a 0.01 a 0.01 a 0.01 a 0.02 a Mar. 15 n/a n/a n/a n/a 0.01 a 0.01 a 0.02 a 0.02 a Mar. 29 n/a n/a n/a n/a 0.01 a 0.01 a 0.01 a 0.01 a PP Jan. 24 0.09 a 0.06 a 0.03 a 0.17 b 0.07 a 0.07 a 0.06 a 0.10 a Feb. 22 0.01 a 0.05 a 0.03 a 0.13 b 0.02 a 0.02 a 0.02 a 0.02 a Mar. 5 n/a n/a n/a n/a 0.01 a 0.02 a 0.01 a 0.01 a Mar. 15 n/a n/a n/a n/a 0.03 a 0.03 a 0.02 a 0.03 a Mar. 29 n/a n/a n/a n/a 0.01 a 0.02 a 0.02 a 0.01 a Total P Jan. 24 0.12 a 0.36 b 0.22 b 1.20 c 0.10 ab 0.15 b 0.15 b 0.25 c Feb. 22 0.04 a 0.43 b 0.41 b 0.54 c 0.03 a 0.04 a 0.03 a 0.04 a Mar. 5 n/a n/a n/a n/a 0.02 a 0.02 a 0.02 a 0.03 a Mar. 15 n/a n/a n/a n/a 0.04 a 0.05 a 0.04 a 0.05 a Mar. 29 n/a n/a n/a n/a 0.02 a 0.03 a 0.03 a 0.02 a C = control, EF = early fall, LF = late fall, and W = winter. Means for all dates within an analyte section and loss pathway (Runoff and Leachate) and followed by the same letter are not significantly different at α = 0.05. BD denotes below detection limit, and n/a denotes that there was not any runoff collected during the event.

Table 4. Total cumulative overwinter mean P losses (mg) in runoff and leachate. Treatment[b] C EF LF W Nutrient Loss[a] DRP RO 0.26 a 2.00 b 2.09 b 4.86 c L 0.04 a 0.11 a 0.06 a 0.08 a Total 0.30 a 2.11 b 2.15 b 4.94 c DOP RO 0.03 a 0.40 b 1.16 c 1.35 c L 0.06 a 0.31 b 0.22 b 0.17 b Total 0.09 a 0.71 b 1.38 c 1.52 c TDP RO 0.29 a 2.40 b 3.25 c 6.21 d L 0.10 a 0.42 c 0.28 b 0.25 b Total 0.39 a 2.82 b 3.53 c 6.46 d PP RO 0.51 a 2.04 b 2.48 b 2.18 b L 0.20 a 0.80 a 0.60 b 0.22 a Total 0.71 a 2.84 bc 3.08 c 2.40 b Total P RO 0.80 a 3.84 b 5.73 c 8.39 d L 0.30 a 1.22 b 0.88 b 0.47 a Total 1.10 a 5.06 b 6.61 c 8.86 d [a] RO denotes runoff losses, and L denotes leachate losses. [b] C = control, EF = early fall, LF = late fall, and W = winter. Means within the same row and followed by the same letter are not significantly different at α = 0.05.

for the W treatment significantly decreased during the fourth event after manure application compared to the previous three events (table 3). All manured treatments had significantly greater DRP loss in runoff compared to the control for all events after manure application. Cumulative overwinter losses of DRP in runoff were significantly different among treatments, with W > LF = EF > C (table 4). Total overwinter losses of TDP were primarily comprised of DRP. Dissolved reactive P accounted for 83%, 64%, and 78% of TDP for the EF, LF, and W treatments, respectively. Similar to DRP, TDP losses for the EF treatment did not decrease significantly in subsequent events compared to the first rainfall simulation after manure application. For the EF treatment, individual event losses ranged from 0.30 to

876

0.61 mg (tables 2 and 3). Total dissolved P decreased significantly during the third rainfall event after manure application compared to the first two events after application for both the LF and W treatments (tables 2 and 3). Cumulative overwinter TDP losses in runoff were significantly different among treatments, with W > LF > EF > C (table 4). Cumulative overwinter PP losses in runoff for the manured treatments were significantly greater than the control; however, there were no significant differences among treatments, with W = LF = EF > C (table 4). Particulate P losses for all manured treatments decreased significantly following the first rainfall simulation after manure application and were not significantly different from the control treatment in subsequent events (table 2). Since PP accounted for 41% to 64% of the total P lost during the first rainfall simulation after manure application, the significant decreases observed in PP during the second event resulted in significant decreases in total P for all treatments as well. Cumulative overwinter total P losses in runoff were significantly different among treatments, with W > LF > EF > C (table 4). For all P forms, the majority of P was lost in runoff, with an average of 95%, 65%, 87%, 83%, and 95% of DRP, DOP, TDP, PP, and total P, respectively, for all treatments. Leachate Losses Dissolved reactive P losses in leachate were low for all treatments compared to runoff and often below the analytical detection limit (tables 2 and 3). Overwinter losses of DRP for the manured treatments were not significantly different from the control. Losses of DOP in leachate were significantly greater than the control for the EF treatment during the first and second rainfall simulations after manure application (table 2). While cumulative losses of DOP for


all manured treatments were significantly greater than the control, there were no differences observed among manured treatments, with EF = LF = W > C (table 4). Alternatively, cumulative overwinter losses of TDP in leachate were significantly different among manured treatments, with EF > LF = W > C (table 4). The January 24 rainfall event produced the largest losses of TDP in leachate for all manured treatments, with the W treatment having significantly higher losses than the EF and LF treatments (table 3). This was the first event that produced leachate for the W treatment. The majority of total P losses occurred during the first three rainfall simulations after manure application for the EF, LF, and W treatments and then significantly decreased in subsequent events. Total P losses in leachate were found to be significantly greater for the EF and LF treatments compared to the W and C treatments over the winter. EXTRACTABLE SOIL PHOSPHORUS LEVELS Concentrations of both CaCl2 and Mehlich-3 extractable P in the soil collected adjacent to the lysimeter wall during excavation at the beginning of the study were not significantly different among treatments (table 5). Mean CaCl2 extractable P concentrations for all treatments decreased with depth and averaged 3.83, 0.75, and 0.08 mg kg-1 for the 0-15, 15-30, and 30-45 cm layers, respectively (table 5). Similarly, mean Mehlich-3 extractable P decreased significantly with depth and averaged 114.8, 72.1, and 11.2 mg kg-1 for the 0-15, 15-30, and 30-45 cm layers, respectively. Calcium chloride extractable P concentrations after completion of the study were significantly different among treatments (table 5). The W treatment had significantly higher CaCl2 concentrations in the 0-15 cm layer compared to all of the other treatments, with W > LF > C = EF. Compared to the initial conditions, CaCl2 extractable P concentrations for both the EF and C treatments decreased in the 0-15 cm layer by 29% and 34%, respectively. Alternatively, CaCl2 extractable P concentrations for both the LF and W treatments increased in the 0-15 cm layer by 16% and 29%, respectively, compared to the initial conditions. In both the 15-30 and 30-45 cm layers, CaCl2 extractable P concentrations increased significantly for the EF and LF treatments, while no differences were observed for the C and W treatments (table 5). Mehlich-3 extractable P concentrations in the 0-15 cm layer increased significantly for all manured treatments at the completion of the study compared to the control (table 5). The W treatment had a significantly greater increase

[a]

[b]

compared to the EF and C treatments, while the LF treatment was not significantly different from either the W or EF treatment (table 5). There were no significant differences in Mehlich-3 extractable P concentrations in the 1530 and 30-45 cm layers, with average increases for all treatments of 5% and 20%, respectively (table 5).

DISCUSSION The results of this study show that sub-zero soil temperatures and subsequent frost formation in the soil can have a pronounced influence on the infiltration of meltwater and precipitation. A 56% increase in runoff volume was observed for frozen soils compared to non-frozen soils. Soils can often turn into massive, dense, concrete-like structures that are nearly impermeable to water when they are frozen at high soil moisture contents (Zuzel and Pikul, 1987). The rapid cooling of the soil surface in the fall season often induces temperature gradients within the soil profile, which causes a migration of moisture to the surface from deeper soils (Bullock et al., 1988). Soils that are frozen at high moisture contents, therefore, have been shown to produce increases in runoff volume (Zuzel and Pikul, 1987; Gray et al., 2001). Zuzel et al. (1982) found that when the soil was frozen, 87% of the precipitation became runoff on field plots, whereas the majority of the water infiltrated when the soil temperature was above freezing. Soil temperature also markedly influenced the infiltration of rain water at temperatures above freezing. A 12% increase in runoff volume was observed when the soil temperature decreased from 15.7°C to 4.8°C. The difference in runoff between temperatures was possibly due to changes in the viscosity of the water contained in the soil pores as well as the rain water. Within the range of environmental temperatures, water viscosity changes by approximately 2% per °C (Lin et al., 2003). Thus, changes in water viscosity at the different temperatures would lead to an estimated decrease in infiltration of 10% to 20%. Not only does this have implications for the seasonal temperature changes in the fall, but also in the spring season when temperatures are increasing. Numerous studies have investigated the processes and mechanisms related to P loss from surface-applied manure (e.g., Shapley, 1995). Few have specifically documented P losses at different soil temperatures; however, several studies have shown that nutrient losses can be significantly greater when the soil is frozen compared to non-frozen soil. Hensler et al. (1970) reported that up to 20% of N, 12% of

Table 5. Extractable soil P levels prior to and after completion of the study.[a] Mehlich-3 Extractable P (mg kg-1) CaCl2 Extractable P (mg kg-1) Depth Sampling C EF LF W C EF LF W (cm) Date[b] 0-15 Initial 4.26 a 4.09 a 3.39 a 3.59 a 120.5 a 119.8 a 111.4 a 107.4 a Final 3.04 a 2.72 a 4.01 b 5.02 c 107.3 a 124.1 b 127.4 b 139.6 b % Diff. -29 a -34 a 16 b 29 c -11 a 4b 13 bc 23 c 15-30 Initial 0.87 a 0.77 a 0.54 a 0.80 a 75.9 a 73.3 a 73.1 a 66.1 a Final 1.27 a 1.50 a 1.40 a 1.36 a 82.0 a 76.9 a 75.3 a 70.4 a % Diff. 32 a 49 b 61 b 41 ab 7a 5a 3a 6a 30-45 Initial 0.12 a BD BD 0.17 a 12.6 a 11.1 a 11.4 a 9.5 a Final 0.26 a 0.34 a 0.25 a 0.40 a 15.5 a 14.5 a 13.8 a 12.3 a % Diff. 54 a 98 b 98 b 58 a 19 a 21 a 17 a 23 a C = control, EF = early fall, LF = late fall, and W = winter. Means within the same row and followed by the same letter are not significantly different at α = 0.05. BD denotes below detection limit. % Diff. = percentage change in soil P concentrations between the initial and final sampling dates.

55(3): 871-880

877

P, and 14% of potassium were lost in surface runoff from manure applied to frozen soil. Similarly, Steenhuis et al. (1981) found that when infiltration was limited, approximately 50% of the original manurial N was lost in surface runoff. In these studies as well as ours, P losses are strongly correlated to differences in runoff volumes between frozen and non-frozen soils. Since soil temperature influences the infiltration of precipitation, it also plays a significant role in determining P losses. A decrease in soil temperature of 10.9°C between the early and late fall application dates resulted in a doubling of dissolved P lost in runoff. Furthermore, a 16.8°C change in soil temperature from the early fall to winter increased runoff dissolved P concentrations four-fold. Therefore, if the majority of the precipitation infiltrates into the soil during the first rainfall event (early fall), then the P losses will be relatively small. If, however, the precipitation becomes runoff due to decreased soil temperatures (winter), the losses will be substantially greater. In addition to hydrology, other factors, such as the length of time between manure application and the first rainfall-runoff event, likely influenced P loss during the first precipitation event after manure application. In a previous study, an irrigation event that occurred immediately following a fertilizer application doubled the P concentrations in runoff compared to when the first irrigation occurred at a water deficit of 50 mm (Bush and Austin, 2001). Schroeder et al. (2004) also observed that as the time between poultry litter application and the first rainfall event increased, P concentrations in runoff decreased. Phosphorus losses potentially could be greater than those reported in this study if the first rainfall-runoff event occurred as soon as one day following manure application or significantly less if the length of time reached one to two weeks after application. Information on the interaction between soil temperature and time (application to first rainfallrunoff event) on nutrient loss from fall and winter applied manure is lacking and should be investigated in future research. Phosphorus losses during the rainfall simulation three days after manure application accounted for 46% to 58% of the total overwinter loss of P regardless of application date. Losses in runoff generally decreased for all manured treatments following the first rainfall-runoff event but were still greater than P losses from the control after four to six rainfall events over a five-month period. The decreases in P losses over the winter season, however, were not consistent among manured treatments. Total P losses in the second rainfall after manure application were 75% less than the losses observed during the first rainfall for the early and late fall treatments, while only a 40% reduction was seen from the winter-applied manure. This resulted in the winter-applied manure having significantly greater overwinter P losses than both the early and late fall treatments. Phosphorus losses in sequential rainfall simulations (Kleinman et al., 2006) and changes in the relative distribution of manure P within inorganic and organic fractions over several simulations (Sharpley and Moyer, 2000) have been examined. However, all of these studies were conducted at warmer temperatures. No studies have document-

878

ed changes in manurial P form or losses of P from cold or frozen manure. Thus, we can only hypothesize why there were differences observed among manure application treatments. Freezing of soil that is high in organic matter content has been shown to increase P solubility because of the dissolution of organic compounds and disruption of plant cells (Ron Vaz et al., 1994). Additionally, research aimed at sample storage and preservation has shown that freezing is the safest way to store soil samples and maintain sorptive properties of the soil because the reaction between soil and P proceeds even in air-dry soil but ceases at subzero temperatures (Bramley et al., 1992). When manure, which is high in organic matter content, is applied in the winter at temperatures less than 0°C, P solubility may increase due to freezing and result in significant losses of P in runoff during the first precipitation event. Additionally, frozen soil following the first rainfall-runoff event may preserve P remaining on the soil surface from interacting with the soil and stop the reaction of P within the manure, which could lead to larger losses in subsequent events. The findings from the extractable soil P concentrations in the top 15 cm of soil support the hypothesis that freezing increases the solubility of manurial P and slows the interaction between the P and the soil. Calcium chloride extractable P is generally thought of as P that is readily available or plant available, whereas Mehlich-3 extractable P accounts for both the labile P and a portion of P that is bound more tightly to clay particles. For the early-fall-applied manure, the CaCl2 extractable P in the top 15 cm decreased by 34%, while the Mehlich-3 extractable P increased by 4%. Conversely, both CaCl2 and Mehlich-3 extractable P increased for the winter-applied manure by 29% and 23%, respectively. Therefore, manure application during both the fall and winter seasons increased the P concentrations in the soil (Mehlich-3 extractable P), but the form and availability of the P remaining in the spring depended on the application date and overwinter soil temperatures. The freezing soil temperatures and differences in leachate losses of P likely influenced the amount of P remaining in the soil as well. The early fall and late fall applications of manure resulted in significantly more P in leachate compared to both the winter-applied manure and the control. The majority (59% to 66%) of the losses for the fall applications occurred during the first rainfall simulation after manure application. Higher infiltration volumes during the first rainfall transported the P through the soil, whereas there was no leachate collected for the winter treatment until the third rainfall event after application. Similar results were observed by Sims et al. (1998), in which fall application of swine manure was observed to have significantly higher P concentrations in agricultural drainage water compared to winter application of manure.

CONCLUSION University extension publications and industry professionals often make recommendations for fall and winter season manure application based on soil temperature. The focus of these recommendations is typically N manage-


ment; however, the results of this study show that soil temperature is also important for overwinter P dynamics. Phosphorus losses from fall and winter applied manure were primarily a function of the soil temperature at the time of the first rainfall-runoff event. The amount of infiltration significantly decreased as soil temperatures decreased, which resulted in higher P losses in runoff. A decrease in soil temperature of 16.8°C between the early fall and winter application dates resulted in a 56% increase in runoff volume and a four-fold increase in dissolved P losses. The losses from this first rainfall-runoff event accounted for 46% to 58% of the total overwinter P loss. Frozen soil with little to no infiltration capacity at the time of the winter application of manure resulted in significantly higher overwinter P losses compared to both early and late fall applications. Despite the highest overwinter P losses, however, the winter manure application treatment had the highest concentrations of P remaining in the top 15 cm of soil at the conclusion of the study. When manure is applied at sub-zero temperatures, P solubility may increase due to freezing. Freezing of the soil and manure may also maintain sorptive properties of the soil because the reaction between soil and P essentially ceases at temperatures below 0°C. Therefore, the freezing of the soil and manure during the winter significantly increases the risk of P losses in runoff due to decreased infiltration capacity of the soil and increased P solubility. Conversely, the cold temperature inhibits manurial P from reacting with the soil and allows the P to remain in more available forms. In order to minimize P losses from fall and winter manure applications, producers should spread their manure in the early fall when soil temperatures are greater than 10°C and infiltration potential is the greatest. Manure application on frozen soil is highly discouraged from an environmental loss perspective; however, our results indicated that winter application resulted in significantly higher soil P concentrations and thus more crop-available P in the spring growing season. More research is needed on changes in manurial P form and losses of P from cold and frozen manure in order to better assess the trade-off risks of applying manure at different dates and soil temperatures.

REFERENCES Bachmann, J., R. Horton, S. A. Grant, and R. R. van der Ploeg. 2002. Temperature dependence of water retention curves for wettable and water repellant soils. SSSA J. 66(1): 44-52. Bole, J. B., and W. D. Gould. 1986. Overwinter losses of nitrogen-15 labeled urea fertilizer. Canadian J. Soil Sci. 66(3): 513-520. Bramley, R. G. V., N. J. Barrow, and T. C. Shaw. 1992. The reaction between phosphate and dry soil: I. The effect of time, temperature, and dryness. J. Soil Sci. 43(4): 749-758. Bremner, J. M., and M. I. Zantua. 1975. Enzyme activity in soils at subzero temperatures. Soil Biol. Biochem. 7(6): 383-387. Bullock, M. S., W. D. Kemper, and S. D. Nelson. 1988. Soil cohesion as affected by freezing, water content, time, and tillage. SSSA J. 52(3): 770-776. Bush, B. J., and N. R. Austin. 2001. Timing of phosphorus fertilizer application within an irrigation cycle for perennial pasture. J. Environ. Qual. 30(3): 939-946. Carpenter, S. R., N. F. Caraco, D. L. Correll, R. W. Howarth, A. N.

55(3): 871-880

Sharpley, and V. H. Smith. 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecol. Appl. 8(3): 559-568. Chantigny, M. H., P. Rochette, and D. A. Angers. 2002. Fate of carbon and nitrogen from animal manure and crop residues in wet and cold soils. Soil Biol. Biochem. 34(4): 509-517. Cookson, W. R., I. S. Cornforth, and J. S. Rowarth. 2002. Winter soil temperature (2-15°C) effects on nitrogen transformations in clover green manure amended or unamended soils: A laboratory and field study. Soil Biol. Biochem. 34(10): 1401-1415. Culik, M. N., J. W. Doran, and K. A. Richards. 1982. Construction of soil thermocouples for the novice. SSSA J. 46(4): 882-884. Daniel, T. C., A. N. Sharpley, and J. L. Lemunyon. 1998. Agricultural phosphorus and eutrophication: A symposium overview. J. Environ. Qual. 27(2): 251-257. Dorland, S., and E. G. Beauchamp. 1991. Denitrification and ammonification at low soil temperatures. Canadian J. Soil Sci. 71(3): 293-303. Feyereisen, G. W., and G. J. Folmar. 2009. Development of a laboratory-scale lysimeter system to simultaneously study runoff and leaching dynamics. Trans. ASABE 52(5): 1585-1591. Fleming, R., and H. Fraser. 2000. Impacts of winter spreading of manure on water quality: Literature review. Ridgetown, Ontario, Canada: Ridgetown College, University of Guelph. Geohring, L. D., O. V. McHugh, M. T. Walter, T. S. Steenhuis, M. S. Akhtar, and M. F. Walter. 2001. Phosphorus transport into subsurface drains by macropores after manure applications: Implication for best manure management practices. Soil Sci. 166(12): 896-909 Gray, D. M., B. Toth, L. Zhao, J. W. Pomeroy, and J. Granger. 2001. Estimating areal snowmelt infiltration into frozen soils. Hydrol. Proc. 15(16): 3095-3111. Hensler, R. F., R. J. Olsen, S. A. Witzel, O. J. Attoe, W. H. Paulson, and R. F. Johannes. 1970. Effect of method of manure handling on crop yields, nutrient recovery, and runoff losses. Trans. ASAE 13(6): 726-731. Hornung, R. W., and L. D. Reed. 1990. Estimation of average concentration in the presence of nondectectable values. Appl. Occup. Environ. Hyg. 5(1): 6-51. Hu, Q. C., E. van Bochove, J. Warland, B. Kay, and C. WagnerRiddle. 2006. New method to simulate soil freezing and thawing cycles for studying nitrous oxide flux. SSSA J. 70(6): 2106-2113. Johnsson, H., B. Thunholm, and L. C. Lundin. 1995. Experimental system for one-dimensional freezing of undisturbed soil profiles. Soil Tech. 7(4): 319-325. Kleinman, P. J. A., M. S. Srinivasan, C. J. Dell, J. P. Schmidt, A. N. Sharpley, and R. B. Bryant. 2006. Role of rainfall intensity and hydrology in nutrient transport via surface runoff. J. Environ. Qual. 35(4): 1248-1259. Lin, C., D. Greenwald, and A. Banin. 2003. Temperature dependence of infiltration rate during large-scale water recharge into soils. SSSA J. 67(2): 487-493. Malhi, S. S., and W. B. McGill. 1982. Nitrification in three Alberta soils: Effects of temperature, moisture, and substrate concentration. Soil Biol. Biochem. 14(4): 393-399. Malhi, S. S., and M. Nyborg, 1979. Nitrate formation during winter from fall-applied urea. Soil Biol. Biochem. 11(4): 439-441. Mehlich, A. 1984. Mehlich No. 3 soil test extractant: A modification of Mehlich No. 2 extractant. Comm. Soil. Sci. Plant Anal. 15(12): 1409-1416 Nyborg, M., and S. S. Malhi. 1986. Comparison of fall and spring application of nitrogen fertilizers in northern and central Alberta. Canadian J. Soil Sci. 66(2): 225-236. Nyborg, M., S. S. Malhi, and E. D. Solberg. 1990. Effect of date of application of the fate of 15N-labeled urea and potassium nitrate. Canadian J. Soil Sci. 70(1): 21-31. Patton, C. J., and J. R. Kryskalla. 2003. Methods of analysis by the

879

U.S. Geological Survey National Water Quality Laboratory: Evaluation of alkaline persulfate digestion as an alternative to Kjeldahl digestion for determination of total dissolved nitrogen and phosphorus in water. USGS Water Resources Investigations Report 03-4174. Reston, Va.: U.S. Geological Survey. Ron Vaz, M. D., A. C. Edwards, C. A. Shand, and M. S. Cresser. 1994. Changes in the chemistry of soil solution and acetic-acid extractable P following different types of freeze/thaw episodes. European J. Soil Sci. 45(3): 353-359. SAS. 2002. SAS User’s Guide: Statistics. Cary, N.C.: SAS Institute, Inc. Schroeder, P. D., D. E. Radcliffe, and M. L. Cabrera. 2004. Rainfall timing and poultry litter application rate effects on phosphorus loss in surface runoff. J. Environ. Qual. 33(6): 2201-2209. Self-Davis, M. L., P. A. Moore Jr., and B. C. Joern. 2009. Determination of water and/or dilute salt-extractable phosphorus. In Methods of Phosphorus Analysis for Soils, Sediments, Residuals, and Waters, 24-26. 2nd ed. J. L. Kovar and G. M. Pierzynski, eds. Southern Cooperative Series Bulletin No. 596. Raleigh, N.C.: North Carolina State University and Southern Extension/Research Activity - Information Exchange Group 17 (SERA 17). Sharpley, A. N. 1995. Identifying sites vulnerable to phosphorus loss on agricultural runoff. J. Environ. Qual. 24(5): 947-951. Sharpley, A. N., and B. Moyer. 2000. Phosphorus forms in manure and compost and their release during simulated rainfall. J. Environ. Qual. 29(5): 1462-1469. Sharpley A. N., R. W. McDowell, and P. J. A Kleinman. 2001.

880

Phosphorus loss from land to water: Integrating agricultural and environmental management. Plant Soil 237(2): 287-307. Sims, J. T., R. R. Simard, and B. C. Joern. 1998. Phosphorus loss in agricultural drainage: Historical perspective and current research. J. Environ. Qual. 27(2): 277-293. Snyder, C. S., G. W. Randall, R. E. Lamond, and R. G. Hoeft. 2000. Fall nitrogen management for agronomic response and environmental protection. Norcross, Ga.: Potash and Phosphate Institute. Available at: https://www.ipni.net/ppiweb/ppibase.nsf/ b369c6dbe705dd13852568e3000de93d/bfa77b79e6c8028c85256 94e002d096d/$FILE/Fall%20N%20Mgmt.pdf. Steenhuis, T. S., G. D. Bubenzer, J. C. Converse, and M. F. Walter. 1981. Winter-spread manure nitrogen loss. Trans. ASAE 24(2): 436-449. Williams, M. R., G. W. Feyereisen, G. J. Folmar, and H. S. Lin. 2010. Experimental system for simulating a natural soil temperature profile during freeze-thaw cycles. Applied Eng. in Agric. 26(5): 843-848. Williams, M. R., G. W. Feyereisen, D. B. Beegle, R. D. Shannon, G. J. Folmar, and R. B. Bryant. 2011. Manure application under winter conditions: Nutrient runoff and leachate losses. Trans. ASABE 54(3): 891-899. Zuzel J. F., and J. L. Pikul. 1987. Infiltration into a seasonally frozen agricultural soil. J. Soil Water Cons. 42(6): 447-450. Zuzel, J. F., R. R. Allmaras, and R. Greenwalt. 1982. Runoff and soil erosion on frozen soils in northeastern Oregon. J. Soil Water Cons. 37(6): 351-354.
