Development of a Phosphorus Index for Pastures ... - PubAg - USDA

Development of a Phosphorus Index for Pastures Fertilized with Poultry Litter—Factors Affecting Phosphorus Runoff

Reproduced from Journal of Environmental Quality. Published by ASA, CSSA, and SSSA. All copyrights reserved.

Paul B. DeLaune,* Philip A. Moore, Jr., Dennis K. Carman, Andrew N. Sharpley, Brian E. Haggard, and Tommy C. Daniel ABSTRACT

P.B. DeLaune and T.C. Daniel, Department of Crop, Soil, and Environmental Sciences, University of Arkansas, Fayetteville, AR 72701. P.A. Moore, Jr. and B.E. Haggard, USDA-ARS, Poultry Production and Product Safety Research Unit, Fayetteville, AR 72701. D.K. Carman, USDA-NRCS, National Water Management Center, Little Rock, AR 72203. A.N. Sharpley, USDA-ARS, Pasture Systems and Watershed Management Research Unit, University Park, PA 16802. Mention of a trade name, proprietary product, or specific equipment does not constitute a guarantee by the USDA and does not imply its approval to the exclusion of other products that may be suitable. *Corresponding author ([email protected]).

rather than the practices for nitrogen, especially for soils susceptible to surface runoff (Sharpley et al., 1993). Annual applications of animal manures to land for several years that exceed plant P removal increase soil test P levels, which leads to greater P losses in surface runoff water (Pote et al., 1996, 1999; Sharpley, 1995; Schreiber, 1988). Most of these researchers have found a positive linear relationship between soil test P and the soluble reactive phosphorus (SRP) concentration in surface runoff water. As a result of these studies, threshold soil test P levels have been established in some regions of the United States above which P application is not recommended or allowed. However, manure applications were generally not made within 1 to 2 yr of the research and runoff concentrations were typically lower than what would be observed from pastures receiving recent or annual manure applications. Although soil test P is being used to guide P management in some states, factors such as manure application, fertilizer application, crop type, erosion, surface runoff, and residue management may override soil P in determining potential P losses (Sharpley and Tunney, 2000). Sauer et al. (2000) showed that poultry litter applications overwhelmed the soil P influence on SRP concentrations in surface runoff water. Kuykendall et al. (1999) suggested that for pastures receiving routine applications of poultry litter, soil test P should not be the sole criterion for determining the P runoff potential. If soil test P is not a dominant factor controlling P runoff from fertilized pastures, then P management guidelines should reflect the other factors that may contribute to P losses in runoff. Gburek et al. (2000) suggested that both source and transport factors should be incorporated in agricultural management practices for P. Several management practices have been studied that reduce the risk of P runoff. Precipitation of SRP using alum has been shown to reduce soluble P concentrations in both manure and surface runoff water (Moore and Miller, 1994; Shreve et al., 1995; Moore et al., 2000). Manipulation of animal diets using phytase enzymes and low phytic acid corn also reduces total P in manure (Nelson et al., 1971; Beers and Jongbloed, 1992; Raboy et al., 1994). Transport factors that consider the timing and occurrence of rainfall events have shown that P runoff decreases with an increase in length of time between a manure application and a runoff event (Westerman and Overcash, 1980; Sharpley, 1997). Phosphorus transport has also been shown to occur predominantly from hydrologically ac-

Published in J. Environ. Qual. 33:2183–2191 (2004). © ASA, CSSA, SSSA 677 S. Segoe Rd., Madison, WI 53711 USA

Abbreviations: HAP, high available phosphorus; SRP, soluble reactive phosphorus; TP, total phosphorus; TSP, triple superphosphate.

Currently, several state and federal agencies are proposing upper limits on soil test phosphorus (P), above which animal manures cannot be applied, based on the assumption that high P concentrations in runoff are due to high soil test P. Recent studies show that other factors are more indicative of P concentrations in runoff from areas where manure is being applied. The original P index was developed as an alternative P management tool incorporating factors affecting both the source and transport of P. The objective of this research was to evaluate the effects of multiple variables on P concentrations in runoff water and to construct a P source component of a P index for pastures that incorporates these effects. The evaluated variables were: (i) soil test P, (ii) soluble P in poultry litter, (iii) P in poultry diets, (iv) fertilizer type, and (v) poultry litter application rate. Field studies with simulated rainfall showed that P runoff was affected by the amount of soluble P applied in the fertilizer source. Before manure applications, soil test P was directly related to soluble P concentrations in runoff water. However, soil test P had little effect on P runoff after animal manure was applied. Unlike most other P indices, weighting factors of the P source components in the P index for pastures are based on results from runoff studies conducted under various management scenarios. As a result, weighting factors for the P source potential variables are well justified. A modification of the P index using scientific data should strengthen the ability of the P index concept to evaluate locations and management alternatives for P losses.

P

hosphorus pollution from nonpoint sources has received increasing attention due to accelerated eutrophication of nearby water bodies. When manure is applied at recommended nitrogen (N) rates, concentrations of P in runoff can be relatively high (Edwards and Daniel, 1993a). More than 90% of surface runoff P from pastures fertilized with animal manures has been shown to be of the soluble form (Edwards and Daniel, 1993b). Soluble P readily affects water quality since it is directly bioavailable to macrophytes and algae, whereas particulate P is bioavailable only when converted to inorganic phosphate (Sonzogni et al., 1982). Thus, nutrient management practices for P have recently been emphasized

2183


2184

J. ENVIRON. QUAL., VOL. 33, NOVEMBER–DECEMBER 2004

tive areas of a watershed (Gburek et al., 1996; Gburek and Sharpley, 1998). A few simulation programs have attempted to calculate edge-of-field P losses. However, none of the existing models provide users with a simple tool to measure the relative potential of P movement from sites based on readily available field parameter values (Lemunyon and Gilbert, 1993). To account for both P transport factors and P sources, the P index was developed to provide conservation planners, watershed planners, and farmers with a simple tool to assess the various landforms and management practices for the potential risk of P movement (Lemunyon and Gilbert, 1993). However, weighting factors for each of the risk variables in the original P index were based on professional judgement. Because the P index was not based on scientific data, a modification of the P index using scientific data should strengthen the ability of the P index concept to evaluate locations and management alternatives for P losses. Several states have modified the original P index to better assess the risk of P movement in localized regions. Most of the modified P indices contain weighting factors that were assigned by regional and state specialists in consultation with local field conservationists. Information from actual runoff studies were not incorporated into many of the currently used P indices. The objective of this series of studies was to determine the effect of (i) soil test P, (ii) soluble P in poultry litter, (iii) P in poultry diets, (iv) fertilizer type, and (v) poultry litter application rates on P concentrations in surface runoff water. The results from these runoff studies were then used to develop a P source component for a P index for pastures fertilized with poultry litter. MATERIALS AND METHODS Rainfall Simulation Studies Seventy-two runoff plots (1.52 ⫻ 6.10 m) were established in 1998 on a Captina silt loam soil (fine-silty, siliceous, mesic Typic Fragiudults) at the University of Arkansas Agricultural Research Station in Fayetteville, AR. The plots were built with a 5% slope and were hydrologically isolated from surrounding land with 15-cm metal borders inserted vertically so that 5 cm of the strips were exposed above the soil surface on three sides. An aluminum collection trough was located at the downslope edge. After the initial construction of the plots was completed, the concentration of P in the soil was augmented on 24 plots with commercial P fertilizer. Triple superphosphate (TSP, 200 g P kg⫺1) was incorporated to about 15 cm at rates of 0, 168, 336, 672, 1008, and 1344 kg P ha⫺1. Triple superphosphate was used to increase soil test P since it would not result in the input of other nutrients or contaminants, such as arsenic, copper, zinc, and 17␤-estradiol, that may have been added using organic fertilizers. The addition of organic fertilizer could have also increased infiltration rates of the plots.

six levels of P augmentation. Two rainfall simulations were conducted 1 wk apart to determine the relationship between soil P and P runoff. A third rainfall simulation was conducted to determine the relationship between soil P and P runoff after poultry litter application. One day before the third rainfall simulation, poultry litter was applied to each plot at a rate equivalent to 5.60 Mg ha⫺1.

Effect of Diet The effect of dietary reductions in P was evaluated in a second runoff study. There were four treatments replicated three times in a randomized complete block design. Treatments included litter from poultry that were fed diets consisting of a (i) normal diet, (ii) phytase diet, (iii) high available phosphorus (HAP) corn diet, and (iv) HAP ⫹ phytase diet. Litter samples from poultry that were fed these diets were collected from a study conducted in Delaware and complete details of formulated diets were provided by Saylor et al. (2001). The litter had been deep-stacked for 6 to 8 mo before collection and shipping. Litter was applied at rates equivalent to 4.48 Mg ha⫺1 (dry wt.) 1 d before the first rainfall simulation. Rainfall simulations were conducted 1, 7, and 14 d after litter application.

Effect of Alum Amendment A third study was conducted to evaluate the effect of soluble P in manure on P runoff. The experimental design consisted of a randomized complete block design with six replications of four treatments consisting of (i) normal poultry litter, (ii) poultry litter treated with 5% alum, (iii) poultry litter treated with 10% alum, and (iv) poultry litter treated with 20% alum. Each treatment was applied at rates equivalent to 5.60 Mg ha⫺1. These plots were the same as those used in the first study (see Effect of Soil Test Phosphorus, above). Each treatment was applied to each of the six levels of soil test P; therefore, there were six replications of each treatment. Rainfall simulations occurred 1, 16, and 21 d after litter application. Litter used for the study was collected from a poultry (broiler) farm in northwestern Arkansas that had six broiler houses, three of which were treated with alum at 10% rates after each flock was removed. The 5% alum treatment was made by mixing one-half alum-treated litter and one-half untreated litter. The 20% alum treatment was made by adding the appropriate amount of alum to the 10% alum-treated litter and mixing.

Effect of Fertilizer Type A fourth study was conducted to determine the effect of fertilizer type on P runoff. Treatments included (i) an unfertilized control, (ii) normal poultry litter, (iii) alum-treated poultry litter, and (iv) triple superphosphate. All fertilizer types were applied at a similar total P rate (78 kg P ha⫺1). Poultry litter was applied at rates equivalent to 5.60 Mg ha⫺1. Treatments were replicated three times in a randomized complete block design. Rainfall simulations were conducted 1, 7, and 14 d after fertilizer application. Both alum-treated and untreated poultry litter were collected from a poultry farm in northwestern Arkansas. Alum was applied at a 10% rate in the broiler house.

Effect of Soil Test Phosphorus The first study was conducted on the P-augmented plots to determine the effect of soil P on P concentrations in runoff. The experimental design was a randomized complete block design with four replications of six treatments consisting of

Effect of Litter Application Rate The effect of litter application rates on P runoff was determined in a fifth study. Treatments included poultry litter application rates equivalent to (i) 2.24, (ii) 4.48, (iii) 6.72, and (iv)

DELAUNE ET AL.: POULTRY LITTER FERTILIZATION AND P RUNOFF

8.96 Mg ha⫺1. Simulated rainfall was applied 1, 7, and 14 d after litter application. Litter was collected from a poultry farm in northwestern Arkansas. Treatments were replicated three times in a randomized complete block design.


Chemical Analysis A subsample of manure applied to each plot for each study described above was taken for analysis of SRP and total phosphorus (TP). Upon return to the laboratory, 20 g of poultry litter from each sample was placed into a 250-mL polycarbonate centrifuge tube and extracted with 200 mL of double deionized water for 2 h on a mechanical shaker (Self-Davis and Moore, 2000). After each sample was centrifuged, aliquots were filtered through a 0.45-␮m membrane and acidified to pH 2 with HCl for SRP analysis. Subsamples of TSP were also taken and extracted with double deionized water. Soluble reactive P was determined colorimetrically using the automated ascorbic reduction method (American Public Health Association, 1998). Total P was determined by digesting ovendried litter (60⬚C) with HNO3 and 30% H2O2 and analyzing the digested sample using an inductively coupled plasma spectrometer (ICP) (Zarcinas et al., 1987). After fertilizer application to each plot, simulated rainfall was applied to each plot using a large rainfall simulator with eight TeeJet 1/2HH-SS30WSQ nozzles (Spraying Systems Co., Wheaton, IL) approximately 3 m above the soil (Miller, 1987). To provide the best coefficient of uniformity and to produce drops with size, velocity, and impact energies characterizing natural rainfall, a water pressure of 28 kPa was regulated at the nozzle to establish a water flow rate of 126 mL s⫺1 (Shelton et al., 1985). Using solenoid valves, a rainfall intensity of 50 mm h⫺1 was used. Runoff samples were collected at 2.5, 7.5, 12.5, 17.5, 22.5, and 27.5 min after continuous runoff was achieved from each plot and composited based on flow rates at the time of sampling. Composited runoff samples were filtered through a 0.45-␮m membrane and acidified to pH 2 with concentrated HCl. Soluble reactive P concentrations in the runoff water were determined colorimetrically on filtered, acidified samples using the automated ascorbic acid reduction method (American Public Health Association, 1998). Soil cores were taken from each plot immediately before each rainfall simulation. Composite soil samples consisting of five random cores were taken each for Mehlich-III (0–15 cm) and water-soluble P (0–5 cm). Soil was collected immediately adjacent to each plot to replace the soil within each plot removed via sampling. Cores were dried in an oven for 48 h at 60⬚C and ground to pass a 2-mm sieve. Mehlich-III P extracts were analyzed using ICP after extracting 2 g of soil with 14 mL of Mehlich-III solution for 5 min and filtering through a Whatman (Maidstone, UK) no. 1 filter (Mehlich, 1984). A 1:7 ratio was used since this is the method used by the University of Arkansas Soil Test Laboratory. Water-soluble P was determined by measuring 2.5 g soil into a 40-mL polycarbonate centrifuge tube and extracting with 25 mL of double deionized water for 2 h on a mechanical shaker (Self-Davis et al., 2000). After each sample was centrifuged, aliquots were filtered through a 0.45-␮m membrane and acidified to pH 2 with HCl. Water-soluble P was determined colorimetrically using the automated ascorbic acid reduction method (American Public Health Association, 1998).

Statistical Analysis and Phosphorus Index Development Analysis of variance was used to determine significant treatment effects on P loss from each runoff event (SAS Institute,

2185

1990). When significance was indicated, means were separated using Fisher’s protected LSD (P ⬍ 0.05). For linear regression data, an F test was performed using JMPIN to test if the slope was significant (SAS Institute, 1996). To develop an index based on the data collected from runoff studies, multiple regression was used (SAS Proc Reg, stepwise option). Annual P loads were calculated for each of the 72 plots used in the runoff study. Calculations were made using the following assumptions: (i) six runoff events per year, (ii) 1 cm of runoff per event, and (iii) three runoff events occur before litter application and three occur after litter application. An assumption of six runoff events per year and 1 cm runoff per event was used based on the observation of several studies in Arkansas. Bartholomew et al. (1954) measured 0.75 cm runoff yr⫺1 for 19 yr from a Newtonia silt loam on ungrazed bermudagrass plots in the Ozark Highlands. Hood and Bartholomew (1956) measured 3.5 cm runoff yr⫺1 from bermudagrass on a Baxter silt loam over a 10-yr period in central Arkansas. These studies have reported the amount of runoff to be 0.7 and 3.9% of the total precipitation, respectively. Ma et al. (1998) measured 1.35 cm runoff yr⫺1 from tall fescue on a Captina silt loam in northwestern Arkansas. Sauer et al. (2002) recently reported that runoff from one-acre watersheds occurred in response to three precipitation events and totaled 3.6 cm of water or 2.6% of the total precipitation that fell at the site during the 14-mo study period (3.1 cm yr⫺1). Edwards et al. (1994) measured 19.3 cm runoff yr⫺1 from a Captina silt loam soil cropped to tall fescue. The average annual runoff volume from published reports on silt loam soils in Arkansas is 5.6 cm yr⫺1, which is similar to the 6 cm runoff yr⫺1 assumed in the loading calculations. An assumption of 6 cm of runoff per year in the Ozark Highlands means that it is assumed that approximately 5% of the annual precipitation contributes to runoff. Load calculations were made using the runoff concentrations from the three runoff events that occurred from each of the fertilized plots after litter application. The load for each runoff event from each plot was calculated and summed. Load calculations for runoff events that occurred before litter application were calculated from unfertilized control plots. The P load resulting from three runoff events from control plots was added to the calculated load from fertilized plots. The calculated P load was set as the dependent variable whereas water-soluble soil P, Mehlich-III P, amount of total P applied, and amount of soluble P applied were set as the independent variables. Parameter estimates were used as the weighting factor for respective variables.

RESULTS AND DISCUSSION Runoff Studies Effect of Soil Test Phosphorus Average soil test P (Mehlich III) levels were 117, 159, 220, 305, 369, and 473 mg P kg⫺1 for additions of 0, 168, 336, 672, 1008, and 1344 kg P ha⫺1, respectively. Although these high additions of P typically do not occur, the resulting soil test P levels can be found in soils that have received excessive, long-term annual P applications. A strong positive linear relationship existed between the amount of P incorporated and the resulting Mehlich-III P. The slope of the line indicated that TSP increased Mehlich III–extractable P by 0.26 mg P kg⫺1 per kg P ha⫺1 added. Initial rainfall simulations showed a positive linear


2186


uble P concentrations in runoff water (Burch et al., 2001; Lory et al., 2001; Humphry, 2000). Since poultry litter is a liming agent, the pH of manured soils increases. As calcium (Ca) is added and the pH increases, P appears to be partitioned into less-soluble Ca-bound phases. Thus, precipitation of Ca-phosphate minerals results in a diminishing nonlinear relationship between soil test P and P concentrations in runoff water. Burch et al. (2001) noted this trend once the soil pH increased to greater than 6.4. A significant linear relationship (r2 ⫽ 0.52) was observed between Mehlich-III P and P concentrations in runoff water from the second rainfall simulation, which occurred 1 wk after the first simulation (data not shown). However, once manure was applied to these plots at equivalent rates, the relationship between SRP concentrations in runoff water and Mehlich-III P was not significant (Fig. 1b). Sauer et al. (2000) also showed that a good relationship between Mehlich-III soil P and SRP concentrations in runoff water existed before litter application; however, soil test P was overwhelmed by soluble P application rates once litter was applied. These results add to the evidence that soil test P would be useful in estimating SRP in runoff only in areas where manure or other sources high in SRP are not applied. Effect of Soluble Phosphorus on Runoff

Fig. 1. Relationship between soluble reactive phosphorus (SRP) in runoff water and (a) Mehlich-III soil test P before litter application; (b) Mehlich-III soil test P after litter application; and (c) the amount of soluble P in litter applied.

relationship (r ⫽ 0.86) between SRP concentrations in runoff water and Mehlich-III P from the P-augmented plots (Fig. 1a). A strong positive linear relationship between SRP in runoff and soil P has also been shown in past studies (Pote et al., 1996, 1999; Sharpley, 1995). Concentrations of P in runoff ranged from 2.32 to 3.2 mg P L⫺1 from plots receiving 1344 kg P ha⫺1. This may be attributed to the addition of a highly soluble P source such as TSP to augment soil P levels. Most studies relating soil test P to soluble P concentrations in runoff report SRP concentrations less than 2.0 mg P L⫺1. Burch et al. (2001) reported soluble P concentrations in runoff of 1.6 mg P L⫺1 at a Mehlich III–extractable P concentration of 1500 mg P kg⫺1 in the top 5 cm of soil. On heavily manured soils, a nonlinear diminishing relationship, rather than a strong positive linear relationship, has been shown between Mehlich III–extractable P and sol2

In contrast, a significant linear relationship was found between soluble P in the litter and SRP concentrations in runoff water (r2 ⫽ 0.75; Fig. 1c). Although SRP concentrations in runoff water were positively correlated to TP in the litter, they were more strongly correlated to the soluble P in the litter (data not shown). A poor relationship also existed between TP in the manure and SRP concentrations in runoff water. Sharpley et al. (2001) also showed that Mehlich III–extractable soil P had little or no effect on P concentration in runoff from soils fertilized with dairy manure. Therefore, it may be concluded that soluble P in manure is one of the most important factors contributing to SRP concentrations in runoff water following litter application. Effect of Diet on Phosphorus Runoff Total P levels in the manure from broiler chickens fed low-P diets were significantly lower than that from birds fed normal diets (Table 1). These data support the findings of past studies (Nelson et al., 1971; Jongbloed and Kemme, 1990; Raboy et al., 1994). Poultry and swine lack the phytase enzyme necessary to utilize the P from phytic acid, thereby making the P unavailable to these animals. As a result, only 10 to 20% of the P in corn is available to swine and poultry (Cromwell et al., 1993; Jongbloed and Kemme, 1990). Because nonruminant animals, such as poultry and swine, lack the phytase enzyme, animal producers must add inorganic phosphate, such as defluorinated phosphate, to animal diets to provide sufficient levels of P. Addition of phytase to feed results in increased availability of phytate P by converting unavailable phytic acid to a more readily available form of P (Cromwell et al., 1993). An alterna-

2187


Table 1. Soluble reactive phosphorus (SRP) concentrations, total phosphorus (TP) concentrations, SRP applied, TP applied, and Mehlich-III soil test P from various fertilizer sources in the studies used for the development of weighting factors in the phosphorus index. Concentration† Study


Effect of alum

Fertilizer type

Diet type

Application rate

Applied

Fertilizer source

SRP

TP

SRP

normal 5% alum 10% alum 20% alum normal alum TSP§ unfertilized control normal HAP¶ phytase HAP/phytase 1 ton 2 ton 3 ton 4 ton

mg kg⫺1 556 226 159 68.2 629 149 171 014 0 2 069 3 434 4 762 1 158 583 590 608 626

g kg⫺1 20.8 21.2 20.1 16.6 19.0 18.9 17.9 0 27.7 20.6 17.3 14.1 20.8 21.0 22.6 22.3

2.45 0.96 0.67 0.29 2.76 0.63 74.9 0 8.98 15.4 20.1 5.02 0.97 1.94 2.91 3.88

TP

Mehlich-III soil P‡

93.3 92.3 87.1 72.2 85.3 82.1 78.5 0 124 94.0 74.5 63.1 35.4 71.1 111 142

mg kg⫺1 218 324 275 319 102 118 122 100 125 114 103 110 99.1 107 111 108

kg ha⫺1

† Dry-weight basis. ‡ Mehlich-III soil P immediately before the second rainfall simulation after fertilizer application. § Triple superphosphate. ¶ High available phosphorus.

tive approach to adding phytase enzyme is the use of a low-phytate corn, thus increasing the availability of P in corn grain by genetically manipulating the form of P. Raboy et al. (1990) first noted a concomitant increase in kernel inorganic P with a reduction in phytic acid in corn. Two nonlethal corn mutants, lpa-1 and lpa-2, showed reductions in phytic acid P ranging from 50 to 60% with no decrease in total P in the seed (Raboy and Gerbasi, 1996). As a result of phytase additions, as well as the use of low-phytate corn, inorganic phosphate additions to feed can be dramatically reduced. Therefore, total P concentrations in manure are reduced. It is commonly believed that P concentrations in runoff will be higher from litter containing the highest total P concentrations. However, our analysis showed that SRP concentrations in runoff water were highest from the manure from birds fed HAP and phytase diets, which resulted in lower total P concentrations in manure (Table 1). While total P concentrations in manures are often reduced by dietary modification, it has not been clearly established if soluble P concentrations in manures are also reduced. Phytase and HAP diets actually increased soluble P concentrations in the litter in this study (Table 1). High levels of soluble P in the litter resulted in higher SRP concentrations in runoff water (Fig. 2). Litter from the phytase diet had significantly lower total P concentrations than litter from the normal diet, but had significantly higher soluble P concentrations than all other diet-modified litter. These data indicate that P runoff is more closely related to soluble P in manure than total P. Litter from phytase diets resulted in significantly higher SRP concentrations in runoff water than normal litter, with concentrations as high as 85 mg P L⫺1 (117% increase over normal litter). Again, this shows that the amount of SRP being applied is a very important factor in regulating P runoff. The litter from the diet modification study had been deep-stacked for 6 to 8 mo before shipping. Apparently, deep stacking affected the P solubility in the manure.

Codling et al. (2000) reported a 34% increase in waterextractable P on poultry litter over a 7-wk incubation period. Vadas et al. (2004) found that water-extractable P in poultry litter samples, which had gone through a few freeze–thaw cycles and a room temperature mixing and compositing, increased by an average of 52%. Storage and handling of poultry litter can affect the litter soluble P contents. Although not significant, Moore et al. (1998) found that SRP concentrations in runoff from plots amended with litter from broilers fed a phytase diet were numerically higher (12%) than a standard diet. Smith et al. (2001) found a 25% increase in SRP concentrations in runoff water from plots amended with manure from swine fed a phytase diet compared with a normal diet. While studies have shown that total P concentrations are decreased in litter from modified diets, more research is needed to define how diet modification and litter storage affect the SRP of these manures. Phosphorus is more readily available using phytase as it releases inor-

Fig. 2. The effect of diet type on soluble reactive phosphorus (SRP) concentrations in runoff water.


2188


litter, 15.1 mg P L⫺1 for 5% alum, 13.4 mg P L⫺1 for 10% alum, and 0.88 mg P L⫺1 for 20% alum. A significant linear relationship was found between the amount of alum in the litter and SRP concentrations in the runoff water (Fig. 3a). Soluble reactive P concentrations significantly decreased as the alum rate increased. Soluble reactive P concentrations in runoff were reduced by 42% with 5% alum, 49% with 10% alum, and 97% with 20% alum. These data support the findings of Shreve et al. (1995), who found that alum additions to poultry litter reduced P runoff by 87% from small plots of tall fescue. Moore et al. (2000) showed a 73% reduction in SRP concentrations in runoff from field-scale watersheds fertilized with alum-treated litter receiving natural rainfall compared with those amended with normal litter for a 3-yr period. Alum additions to litter in broiler houses also improve bird performance parameters, such as weight gain and feed conversion and lower energy use, enhancing economic returns for both integrators and growers (Moore et al., 1999). Moore and Miller (1994) showed that alum additions to poultry litter greatly decreased water-soluble P, with virtually 100% of the soluble P being removed from solution in some instances. The exact mechanism in which P solubility is reduced by alum is still unclear. Effect of Fertilizer Type on Phosphorus Runoff

Fig. 3. The effect of (a) alum application rate; (b) fertilizer type; and (c) poultry litter application rate on soluble reactive phosphorus (SRP) concentrations in runoff water. TSP, triple superphosphate.

ganic orthophosphate from phytate. The activity of the phytase enzyme could potentially continue in stored manure and increase the soluble P content of the manure. Controlling SRP concentrations in litter is particularly important when litter is surface-applied to pastures, since P runoff from pastures is predominantly of the soluble form. As phytase decreases the total P content of manures, the addition of chemical amendments to reduce soluble P levels in the manure could result in reduced P concentrations in runoff water. Smith et al. (2001) found a 41% reduction in P concentrations in runoff when adding AlCl3 to manure from swine fed a phytase diet compared with standard diets with an equal AlCl3 rate. Reducing total P concentrations in manure could also reduce increases in soil P with continuing litter application. Effect of Alum Amendment on Phosphorus Runoff Mean runoff concentrations for the first runoff event after litter application were 26.0 mg P L⫺1 for normal

Soluble reactive P concentrations in runoff water 1 d after application were 103, 15.7, and 10.8 mg P L⫺1 for triple superphosphate, normal poultry litter, and alumtreated poultry litter, respectively (Fig. 3b). All fertilizers were applied at rates equivalent to 78 kg P ha⫺1. Plots fertilized with alum-treated litter resulted in the lowest concentrations of P in runoff among fertilized plots. Although all of the fertilizer treatments were applied at equivalent total P rates, alum-treated litter and normal litter resulted in much lower soluble P application rates. Higher P runoff from plots fertilized with TSP can be expected since the water solubility of TSP is about 94% compared with 2 to 4% in poultry litter. Regardless of fertilizer type, SRP concentrations in runoff were dependent on the water solubility of the P source. Nichols et al. (1994) demonstrated higher P concentrations in runoff water from plots receiving inorganic fertilizer for runoff events occurring 7 d after application. These data also demonstrate the importance of water solubility of P in the fertilizer source. Effect of Poultry Litter Application Rate on Phosphorus Runoff Phosphorus concentrations in runoff increased linearly as the P application rate increased for each rainfall event. Phosphorus concentrations in runoff were 8.8, 16.6, 27.7, and 33.0 mg P L⫺1 for applications of 2.24, 4.48, 6.72, and 8.96 Mg ha⫺1, respectively, for the first runoff event (Fig. 3c). A positive linear relationship was seen for each simulated rainfall event, although the magnitude of SRP concentration in runoff decreased with each rainfall event (Fig. 3c). This would be expected if the soluble P concentration in poultry litter is a major contributor to SRP concentrations in runoff.

2189



The mean SRP concentration in runoff water from plots fertilized with litter at rates equivalent to 2.24 Mg ha⫺1 was 4.75 mg P L⫺1 after the third runoff event. The Mehlich-III P concentration in the soil for the fertilized plots was not significantly different from the unfertilized plots. Mean Mehlich-III P concentrations (0–15 cm) were 103 and 102 mg P kg⫺1 for unfertilized control plots and plots receiving 2.24 Mg ha⫺1 litter application, respectively. These data suggest that even after three runoff events, soluble P applied in manure was still overwhelming the effect of soil P with respect to P concentrations in runoff water.

Determination of Weighting Factors for the Phosphorus Source Components Both water-soluble soil P and Mehlich-III P have been shown to be positively correlated to P concentrations in runoff water when no manure has been applied (Pote et al., 1996, 1999). However, a weak correlation was found between soil test P and P concentrations in runoff water in this study (Fig. 1b) as well as previous studies soon after manure has been applied (Pierson et al., 2001; Sauer et al., 2000; Sharpley and Moyer, 2000). Both water-soluble soil P (r ⫽ 0.24) and Mehlich-III soil P (r ⫽ 0.24) were poorly correlated with annual P loading in runoff water from plots receiving P amendments (Table 2). A highly significant relationship was found between TP application rates and annual P loading (r ⫽ 0.67; Table 2). However, soluble P application rates accounted for more of the variability in annual P loading (r ⫽ 0.87; Table 2). This provides evidence that P application rates (both total and soluble) are more important than soil test P in regulating P runoff when manure applications have been made. To determine the weighting factors of the P source component, P loads in runoff were initially modeled using Mehlich-III soil P, water-soluble soil P, TP application rates, and soluble P application rates. Results showed that both water-soluble soil P and Mehlich-III P concentrations did not make a significant contribution to the model, whereas both soluble P and TP application rates were highly significant (P ⬍ 0.0001). Water-soluble soil P and Mehlich-III soil P were highly correlated (y ⫽ 4.54x ⫹ 115; r ⫽ 0.93). The Mehlich-III extraction procedure is used by many laboratories that perform routine soil analysis. Therefore, Mehlich-III rather than water-soluble soil P was used in the model. Although total P applications were highly correlated to resulting runoff P loads, SRP in the litter is more highly correlated to runoff P loads. The dominant form of P found in runoff from pastures is SRP. Thus, SRP was used in the model. When the multiple regression analysis was performed, with water-soluble P in the soil and TP in litter omitted, results showed that the weighting factors for the source factors were 0.404 for soluble P application rate (P ⬍ 0.0001) and 0.000666 for soil test P (P ⬎ 0.2; Mehlich III) (r2 ⫽ 0.74, RMSE ⫽ 1.11) (Eq. [1]). Although Mehlich-III soil P was not significant, it was included in the index to account for background levels of P in runoff or for fields in which poultry litter has not been applied,

Table 2. Relationship between calculated phosphorus loading in runoff and water-soluble soil P, Mehlich-III soil P, the amount of total P applied, and the total amount of soluble P applied. Variable Water-soluble soil P Mehlich-III soil P Total P applied Soluble P applied

Equation y y y y

⫽ ⫽ ⫽ ⫽

⫺0.0094x ⫹ 3.32 ⫺0.0021x ⫹ 3.56 0.036x ⫹ 0.85 0.44x ⫹ 1.67

r

P⬎F

0.24 0.24 0.67 0.87

0.0294 0.0203 ⬍0.0001 ⬍0.0001

thereby providing an estimate of P loss for both fertilized and unfertilized fields. The vast difference in weighting factors is attributed to the much larger values of Mehlich-III P in soil compared with soluble P applied in the manure. Whereas Mehlich-III values for soils are often on the order of 100 to 1000 lb P acre⫺1 (112 to 1120 kg ha⫺1), soluble P applied in manure is typically 2 to 3 lb P acre⫺1 (2.24 to 3.36 kg ha⫺1). These weighting factors were assigned to P source factors in the P index for pastures (Eq. [1]): P source potential ⫽ (0.404 ⫻ lb acre⫺1 SRP applied) ⫹ (0.000666 ⫻ lb acre⫺1 Mehlich-III P)

[1] This P source potential in Eq. [1] is one of the first to use actual runoff data to derive weighting factors. As the P index is an assessment tool implemented by USDA Natural Resources Conservation Service (NRCS) personnel, units of lb acre⫺1 rather than kg ha⫺1 are used throughout the P index for convenience and simplicity to nutrient management planners.

CONCLUSIONS Results from this study show that manure P solubility is the most important source factor affecting P runoff from small runoff plots under simulated rainfall. When no manure was applied, Mehlich-III soil P concentrations were highly correlated with P concentrations in runoff water. However, once manure was applied, SRP concentrations in runoff were not correlated to MehlichIII P, but were highly correlated to the SRP concentrations in the applied manure. Throughout the series of studies, it was evident that P runoff increased with increasing soluble P concentration in the manure. The lowest P runoff was observed from alum-treated litter, which had the lowest SRP levels of all the litter samples. Treatments containing the highest P solubility, such as commercial P fertilizer and HAP or phytase litter, resulted in the highest P runoff. Phosphorus concentrations in runoff also increased with increasing application rates for all three rainfall events, but were highest in the first rainfall event after application. For each of these studies, P concentrations in runoff water remained higher from plots receiving P inputs than plots without any P additions. This shows that P sources are very important in controlling the amount of P lost from runoff plots several weeks after application as well as after multiple runoff events. Unlike other P indices, the weighting factors for the P source variables in this index were developed using results from runoff studies conducted under various management scenarios. As a re-

2190


sult, weighting factors for the P source potential variables are well justified. Validation studies on a larger scale are warranted to better extrapolate results from simulated runoff studies to the field scale.


REFERENCES American Public Health Association. 1998. Standard methods for the examination of water and wastewater. 19th ed. APHA, Washington, DC. Bartholomew, R.P., D.A. Hinkle, and K. Engler. 1954. Effect of rainfall characteristics and soil management practices on soil and water losses in northwest Arkansas. Bull. 548. Univ. of Arkansas Agric. Exp. Stn., Fayetteville. Beers, S., and A.W. Jongbloed. 1992. Effect of supplementary Aspergillus niger phytase in diets for piglets on their performance and digestibility of phosphorus. Br. Soc. Anim. Sci. 55:425–430. Burch, J.A., T.C. Daniel, and J.A. Lory. 2001. Soil phosphorus thresholds to protect water quality: Implications for manure management. In 2001 Annual meetings abstracts [CD-ROM]. ASA, CSSA, and SSSA, Madison, WI. Codling, E.E., R.L. Chaney, and C.L. Mulchi. 2000. Use of aluminum and iron-rich residues to immobilize phosphorus in poultry litter and litter-amended soils. J. Environ. Qual. 29:1924–1931. Cromwell, G.L., T.S. Stahly, R.D. Coffey, H.J. Moneque, and J.H. Randolph. 1993. Efficacy of phytase in improving the bioavailability of phosphorus in soybean meal and corn-soybean meal diets for pigs. J. Anim. Sci. 71:1831–1840. Edwards, D.R., and T.C. Daniel. 1993a. Runoff quality impacts of swine manure applied to fescue plots. Trans. ASAE 36:81–86. Edwards, D.R., and T.C. Daniel. 1993b. Effects of poultry litter application rate and rainfall intensity on quality of runoff from fescuegrass plots. J. Environ. Qual. 22:361–365. Edwards, D.R., T.C. Daniel, J.F. Murdoch, and P.F. Vendrell. 1994. The Moore’s Creek monitoring project. Publ. MSC-162. Arkansas Water Resour. Center, Fayetteville. Gburek, W.J., and A.N. Sharpley. 1998. Hydrologic controls on phosphorus loss from upland agricultural watersheds. J. Environ. Qual. 27:267–277. Gburek, W.J., A.N. Sharpley, L. Heathwaite, and G.J. Folmar. 2000. Phosphorus management at the watershed scale: A modification of the phosphorus index. J. Environ. Qual. 29:130–144. Gburek, W.L., A.N. Sharpley, and H.B. Pionke. 1996. Identification of critical sources for phosphorus export from agricultural catchments. p. 263–282. In M.G. Anderson and S.M. Brooks (ed.) Advances in hillslope processes. John Wiley & Sons, Chichester, UK. Hood, G.W., and R.P. Bartholomew. 1956. Soil and water conservation studies in the Ozark Highlands area of Arkansas. Bull. 563. Univ. of Arkansas, Agric. Exp. Stn., Fayetteville. Humphry, J.B. 2000. A portable rainfall simulator for comparing phosphorus in soil and runoff. M.S. thesis. Univ. of Arkansas, Fayetteville. Jongbloed, A.W., and P.A. Kemme. 1990. Effect of pelleting mixed feeds on phytase activity and apparent absorbability of phosphorus and calcium in pigs. Trans. ASAE 36:1673–1678. Kuykendall, H.A., M.L. Cabrera, C.S. Hoveland, M.A. McCann, and L.T. West. 1999. Stocking effects on nutrient runoff from pastures fertilized with broiler litter. J. Environ. Qual. 28:1886–1890. Lemunyon, J.L., and R.G. Gilbert. 1993. Concept and need for a phosphorus assessment tool. J. Prod. Agric. 6:483–486. Lory, J.A., T.C. Daniel, J. Burch, P.J.A. Kleinman, and P.C. Scharf. 2001. Nonlinear effects of soil test P extraction method and sampling depth on runoff [P] predictions. In 2001 Annual meetings abstracts [CD-ROM]. ASA, CSSA, and SSSA, Madison, WI. Ma, L., H.D. Scott, M.J. Shaffer, and L.R. Ahuja. 1998. RZWQM simulations of water and nitrate movement in manured tall fescue. Soil Sci. 163:259–270. Mehlich, A. 1984. Mehlich 3 soil test extractant: A modification of Mehlich 2 extractant. Commun. Soil Sci. Plant Anal. 15:1409–1416. Miller, W.P. 1987. A solenoid-operated, variable intensity rainfall simulator. Soil Sci. Soc. Am. J. 51:832–834. Moore, P.A., Jr., T.C. Daniel, and D.R. Edwards. 1999. Reducing

phosphorus runoff and improving poultry production with alum. Poult. Sci. 78:692–698. Moore, P.A., Jr., T.C. Daniel, and D.R. Edwards. 2000. Reducing phosphorus runoff and inhibiting ammonia loss from poultry manure with aluminum sulfate. J. Environ. Qual. 29:37–49. Moore, P.A., Jr., and D.M. Miller. 1994. Decreasing phosphorus solubility in poultry litter with aluminum, calcium, and iron amendments. J. Environ. Qual. 23:325–330. Moore, P.A., Jr., M.L. Self-Davis, T.C. Daniel, W.E. Huff, D.R. Edwards, D.J. Nichols, W.F. Jaynes, G.R. Huff, J.M. Balog, N.C. Rath, P.W. Waldroup, and V. Raboy. 1998. Use of high available phosphorus corn and phytase enzyme additions to broiler litter diets to lower phosphorus levels in poultry litter. p. 346–352. In J.P. Blake and P.H. Patterson (ed.) Proc. of the Natl. Poultry Waste Manage. Symp. Auburn Univ. Print. Serv., Auburn, AL. Nelson, T.S., T.R. Shieh, R.J. Wodzinski, and J.H. Ware. 1971. Effect of supplemental phytase on the utilization of phytase phosphorus by chicks. J. Nutr. 101:1289–1294. Nichols, D.J., T.C. Daniel, and D.R. Edwards. 1994. Nutrient runoff from pasture after incorporation of poultry litter or inorganic fertilizer. Soil Sci. Soc. Am. J. 58:1224–1228. Pierson, S.T., M.L. Cabrera, G.K. Evanylo, H.A. Kuykendall, C.S. Hoveland, M.A. McCann, and L.T. West. 2001. Phosphorus and ammonium concentrations in surface runoff from grasslands fertilized with broiler litter. J. Environ. Qual. 30:1784–1789. Pote, D.H., T.C. Daniel, A.N. Sharpley, P.A. Moore, Jr., D.R. Edwards, and D.J. Nichols. 1996. Relating extractable soil phosphorus to phosphorus losses in runoff. Soil Sci. Soc. Am. J. 60:855–859. Pote, D.H., T.C. Daniel, A.N. Sharpley, P.A. Moore, Jr., D.M. Miller, and D.R. Edwards. 1999. Relationship between phosphorus levels in three ultisols and phosphorus concentrations in runoff. J. Environ. Qual. 28:170–175. Raboy, V., D.B. Dickinson, and M.G. Neuffer. 1990. A survey of maize kernel mutants for variation in phytic acid. Maydica 35:383–390. Raboy, V., and P. Gerbasi. 1996. Genetics of myo-inositol phosphate synthesis and accumulation. p. 257–285. In B.B. Biswas and S. Biswas (ed.) Subcellular biochemistry. Vol. 26. Myo-inositol phosphates, phosphoinositides, and signal transduction. Plenum Press, New York. Raboy, V., K. Young, and P. Gerbasi. 1994. Maize low phytic acid (lpa) mutants. No. 1827. In 4th Int. Congr. of Plant Molecular Biol. SAS Institute. 1990. SAS/STAT user’s guide. Version 6. 4th ed. SAS Inst., Cary, NC. SAS Institute. 1996. JMPIN user’s guide. Version 3. Student ed. SAS Inst., Cary, NC. Sauer, T.J., T.C. Daniel, D.J. Nichols, C.P. West, P.A. Moore, Jr., and G.L. Wheeler. 2000. Runoff water quality from poultry littertreated pasture and forest sites. J. Environ. Qual. 29:515–521. Sauer, T.J., P.A. Moore, Jr., J.M. Ham, W.L. Bland, J.H. Prueger, and C.P. West. 2002. Seasonal water balance of an Ozark hillslope. Agric. Water Manage. 55:71–82. Saylor, W.W., J.T. Sims, G.W. Malone, and M.F. Lavahum. 2001. Use of phytase and high-available phosphorus corn in broiler diets: Impact on litter phosphorus levels and solubility. p. 43–57. In Proc. 49th Maryland Nutr. Conf. for Feed Manufacturers, Baltimore. College of Agric. and Life Sci., Univ. of Maryland, College Park. Schreiber, J.D. 1988. Estimating soluble phosphorus in agricultural runoff. J. Miss. Acad. Sci. 33:1–15. Self-Davis, M.L., and P.A. Moore, Jr. 2000. Determining water-soluble phosphorus in animal manure. p. 74–76. In G.M. Pierzynski (ed.) Methods of phosphorus analysis for soils, sediments, residuals, and waters. Southern Coop. Ser. Bull. 396 [Online]. Available at www.soil.ncsu.edu/sera17/publications/sera17-2/pm_cover.htm (verified 12 May 2004). North Carolina State Univ., Raleigh. Self-Davis, M.L., P.A. Moore, Jr., and B.C. Joern. 2000. Determination of water- and/or dilute salt-extractable phosphorus. p. 24–26. In G.M. Pierzynski (ed.) Southern Coop. Ser. Bull. 396 [Online]. Available at www.soil.ncsu.edu/sera17/publications/sera17-2/pm_cover.htm (verified 12 May 2004). North Carolina State Univ., Raleigh. Sharpley, A.N. 1995. Dependence of runoff phosphorus on extractable soil phosphorus. J. Environ. Qual. 24:920–926. Sharpley, A.N. 1997. Rainfall frequency and nitrogen and phosphorus runoff from soil amended with poultry litter. J. Environ. Qual. 26: 1127–1132.



Sharpley, A.N., P.J.A. Kleinman, R.W. McDowell, and J.L. Weld. 2001. Assessing site vulnerability to phosphorus loss in an agricultural watershed. J. Environ. Qual. 30:2026–2036. Sharpley, A.N., and B. Moyer. 2000. Phosphorus forms in manure and compost and their release during simulated rainfall. J. Environ. Qual. 29:1462–1469. Sharpley, A.N., S.J. Smith, and W.R. Bain. 1993. Nitrogen and phosphorus fate from long-term poultry litter application to Oklahoma soils. Soil Sci. Soc. Am. J. 57:1131–1137. Sharpley, A.N., and H. Tunney. 2000. Phosphorus research strategies to meet agricultural and environmental challenges of the 21st century. J. Environ. Qual. 29:176–181. Shelton, C.H., R.D. von Bernuth, and S.P. Rajbhandari. 1985. A continuous-application rainfall simulator. Trans. ASAE 28:1115–1119. Shreve, B.R., P.A. Moore, Jr., T.C. Daniel, D.R. Edwards, and D.M. Miller. 1995. Reduction of phosphorus in runoff from field-applied poultry litter using chemical amendments. J. Environ. Qual. 24:106– 111. Smith, D.R., P.A. Moore, Jr., C.V. Maxwell, and T.C. Daniel. 2001. Dietary phytase and aluminum chloride manure amendments to

2191

reduce phosphorus and ammonia volatilization from swine manure. p. 502–507. In G.B. Havenstein (ed.) Proc. Int. Symp. Addressing Animal Production and Environ. Issues, Research Triangle Park, NC. 3–5 Oct. 2001. College of Agric. and Life Sci., North Carolina State Univ., Raleigh. Sonzogni, W.C., S.C. Chapra, D.E. Armstrong, and T.J. Logan. 1982. Bioavailability of phosphorus inputs to lakes. J. Environ. Qual. 11: 555–563. Vadas, P.A., J.J. Meisinger, L.J. Sikora, J.P. McMurtry, and A.E. Sefton. 2004. Effect of poultry diet on phosphorus in runoff from soils amended with poultry manure and compost. J. Environ. Qual. 33:1845–1854. Westerman, P.W., and M.R. Overcash. 1980. Short-term attenuation of runoff pollution potential for land-applied swine and poultry manure. p. 289–292. In Livestock wastes—A renewable resource. Proc. of the 4th Int. Symp. on Livestock Wastes, Amarillo, TX. Apr. 1980. Am. Soc. Agric. Eng., St. Joseph, MI. Zarcinas, B.A., B. Cartwright, and L.R. Spouncer. 1987. Nitric acid digestion and multi-element analysis of plant material by inductively coupled argon plasma spectrometry. Commun. Soil Sci. Plant Anal. 18:131–146.