Department of Agronomy, Kansas State University, Manhattan, KS 66506
* Corresponding author (firstname.lastname@example.org ).
Received for publication March 3, 2004.
Interest in plant nutrient issues for sustainable land application of residuals is increasingly driven by environmental concerns. The indicators of concern are P and N in surface waters, nitrate leaching, and emissions of ammonia and greenhouse gases. Federal regulations require residual application rates to be on a N basis at most, and on a P basis when risk of P loss in surface runoff is high. Modeling of mineralization offers the potential for more accurate determinations of potentially available nitrogen (PAN) and quick tests could allow the determination of PAN on residuals immediately before land application. Methods for reducing ammonia emissions from livestock operations and new techniques for quantifying emissions under field conditions are being developed. Calibration and validation of P loss assessment tools is an ongoing concern and the interpretation of edge of field P losses warrants further attention. The solubility of P in residuals and soils can be influenced by various amendments or treatment processes. High available P grains or phytase enzyme supplementation can reduce total and soluble P in animal manures by reducing the need for diet supplementation with inorganic P. The use of synchrotron-based X-ray absorption spectroscopy has identified chemical forms of inorganic P. Considerable progress has been made addressing plant nutrient issues for sustainable land application and interest in this topic will remain strong into the foreseeable future.
Abbreviations: CAFO, concentrated animal feeding operation • HAP, high available phosphorus • PAN, potentially available nitrogen • PMN, potentially mineralizable nitrogen • STP, soil test phosphorus • TSP, triple superphosphate
Source: J. Environ. Qual. 34:18-28 (2005).
ONE OF THE OBJECTIVES of the Sustainable Land Application Conference was to provide an update on the body of literature related to the land application of biosolids and wastewaters since the previous major conference on this topic in 1993. An additional objective was to expand the scope to include other organic by-products such as animal manures. The purpose of this paper is to provide a broad overview of research and relevant issues for soil reactions of nutrients as constituents in land-applied agricultural and municipal by-products. Additional details on select topics are provided in subsequent papers in this section.
Earlier reports from conferences on plant nutrients in biosolids and wastewaters discussed nitrogen (N), phosphorus (P), potassium (K), sulfur (S), calcium (Ca), and magnesium (Mg) (Linden et al., 1983; Pierzynski, 1994). This report focuses on only N and P, as they have significant environmental implications that often determine residual application rates.
Previous interest in land application aspects of N and P focused on nutrient content, mineralization of organic N, variability in characteristics within and between treatment plants and processes, and nitrate movement. Ammonia volatilization and P chemistry and availability have only recently garnered attention. Plant nutrient issues remain a significant topic for land application due to existing and emerging environmental concerns.
Recent U.S. federal legislation has addressed various aspects of nutrient management, including land application of by-products. These include the promulgation of 40 CFR Part 503 regulations for land-applied biosolids (USEPA, 1993), the USDA-NRCS Conservation Practice Standards for Nutrient Management Code 590 (USDA, 1999), the USEPA concentrated animal feeding operation (CAFO) regulations (USEPA, 2003), and the USDA National Organic Standards (USDA, 2000). The influence of these regulations on scientific literature has been considerable as they have stimulated a great deal of research.
The plant nutrient aspects of the Part 503 regulations were based on concerns regarding nitrate leaching, and required estimates of PAN, including potentially mineralizable nitrogen (PMN), to determine appropriate application rates. The 503 regulations raised the issue of N need versus N removal as a basis for determining application rates. For example, biosolids could be applied to legume crops, whose relatively high N removal rates permit greater application rates than those made to some nonleguminous crops. Because the 503 regulations did not address P, the greater application rates were viewed positively in the overall efforts of a land application program to manage N.
At the national level, the Code 590 standards provide guidelines for nutrient management to be developed further and followed by each state. The Code 590 standards require that animal wastes be applied on a N basis if risk of P loss is low and on a P basis if P loss risk is high. The Code 590 standards offer three possible means for identifying areas with high risk of P loss. There are two threshold approaches, one based on environmental criteria and the other on agronomic criteria, and a P index rating. Combinations of these approaches are permitted. Soil test P levels are used as criteria to limit or prohibit P applications when the P threshold is reached. Agronomic thresholds, which are based on potential crop response to P applications, are typically the most restrictive approach. Environmental thresholds are designed to prevent unacceptable P losses via surface or subsurface flow, and are based on the relationship between soil test phosphorus (STP) and runoff or tile drainage composition.
Adoption of a P index rating system by a state agency necessitates development based on local climatic conditions, soils, and management practices. The complexity of systems varies widely among states (Sharpley et al., 2003). Factors typically used in P index systems include STP, P additions from inorganic and organic sources, method of application for P additions, soil erosion, irrigation erosion, and proximity to bodies of water. These source and transport factors interact to influence the potential for P impairment of surface waters. Many P index systems are constructed with a great deal of professional judgment as to how the site characteristics will collectively influence P loss and, therefore, require validation and/or calibration. However, relatively few states have correlated measured versus predicted P losses in some fashion.
Threshold approaches have the advantages of being simple and related to water quality or agronomic P needs. Disadvantages include the lack of consideration of P transport, the implication that widespread distribution of P is better than limited distribution, and lack of inclusion of best management practices. The P index systems provide a more comprehensive view of P fate and transport, which implies the necessity of best management practices. Use of a P index system introduces flexibility to allow greater STP levels when P transport potential is low. This can be a critical factor in areas where conservative thresholds may severely restrict land application of by-products. One significant disadvantage to the P index system is the requirement of complex inputs and the need of a trained professional to complete the assessment.
The CAFO regulations, officially published in the Federal Register early in 2003, are in the early stages of implementation. Animal feeding operations are defined as systems that confine animals for 45 d or more per year where little or no vegetation exists. An animal feeding operation can be designated a CAFO if the number of animals exceeds certain limits or if the operation is deemed to have significant negative effects on the environment. A permit that includes a nutrient management plan must be obtained to begin or continue a CAFO. Each state will determine how it handles the nutrient management plan requirements. Most states will likely simply adopt the Code 590 standards to satisfy this need.
The USDA national organic standards stipulate that foods that are labeled "organic" must be produced without the use of synthetic substances, with some exceptions, and without certain nonsynthetic substances. The standards also stipulate that application of plant or animal materials must be managed ... in a manner that does not contribute to the contamination of crops, soil, or water by plant nutrients, pathogenic organisms, heavy metals, or residues of prohibited substances. Raw or composted animal manures are allowed, whereas biosolids are prohibited. Harvest restrictions are required following the use of raw manures. One of the key challenges to organic production is providing adequate N while minimizing the accumulation of P and heavy metals in soils (Mikkelsen, 2000).
Potentially Available Nitrogen and Nitrogen Mineralization
The USEPA 40 CFR Part 503 regulations require that sewage sludge land applications not exceed the agronomic rate for a given crop. The agronomic rate provides the amount of N required by the crop and minimizes the amount of N available for leaching below the root zone to ground water supplies. Similarly, the Code 590 guidelines state that animal manures can be applied on a N basis provided the risk of P loss is low. A generalized approach for determining PAN is illustrated in the following equation:
Researchers have attempted to calculate PMN using laboratory methods, field studies, decay series, and computer models. Field studies by Cogger et al. (1999) found apparent biosolids N recovery of 28 to 40% for forage grasses and 11 to 44% for dryland winter wheat. Results indicated a greater availability of biosolids N in the second year of application than predicted from commonly used biosolids decay series (USEPA, 1983). Carryover N mineralized from biosolids applied to forage grasses can be significant; cumulative apparent N recovery increased by an average of 9% in the year following biosolids application (Sullivan et al., 1997). In a similar field study, Barbarick et al. (1996) reported net N mineralization from 25 to 57% for five to six applications of 6.7 Mg biosolids ha–1 on dryland winter wheat. More recent research from Barbarick and Ippolito (2000) found that 1 Mg of biosolids was equal to approximately 8 kg N fertilizer, whereas estimates calculated using the USEPA (1983) approach for the same material were only 6 to 7 kg N fertilizer equivalent per 1 Mg biosolids. They also estimated first-year mineralization rates that ranged from 25 to 32%.
Computer models and constant decay series have also been used to predict the amount of available N in municipal waste products. Gilmour and Skinner (1999) estimated PAN from six biosolids using decay-series modeling. This included average biosolids decomposition kinetic data from laboratory incubations, average field site weather data, and biosolids analytical data. Potentially available N released from biosolids during the field study was linearly related to the biosolids C to N ratio, organic N, or total N. Biosolids C to N ratio was the best predictor of biosolids PAN in the field, followed by organic N and total N. Slopes from the relationships of PAN to organic N and total N suggested that about 45 and 40% of the biosolids N was made plant available during the growing season, respectively.
Gilmour and Skinner (1999) used two approaches to calculate annual estimates of PAN from biosolids. For each approach, the computer model Decomposition employed analytical data for organic C, organic N, and inorganic N. One method used actual decomposition kinetics and weather data, and the other method used mean biosolids kinetics and mean weather data to estimate biosolids decomposition (Gilmour et al., 1996). Both methods produced similar estimates of PAN for the same biosolids source, but variability among biosolids was great. Mean estimates for PAN during the first year for Methods 1 and 2 were almost identical at 32.4 and 32.0 kg N Mg–1 biosolids, respectively. Overall, similar PAN predictions were obtained with decay series using the computer model with average or actual decomposition kinetics and weather data. This approach was verified with field studies located across a wide range of climatic conditions. Gilmour et al. (2003) reported that observed and predicted PAN were strongly correlated (r2 = 0.72), with the slope and intercept not significantly different than unity and zero, respectively (Fig. 1) . The application of this approach to other by-products has yet to be demonstrated.Accurate estimates of PMN are also needed for animal waste products. Land application of livestock manure can lead to surface and ground water contamination if rates exceed crop needs and reduce crop yields if rates limit N availability. Nitrogen mineralization in manure is influenced by a host of factors, both product-specific and environmental. Variability in N mineralization is due partly to soil type and field environmental conditions; however, the composition of the manure can significantly influence the rate and amount of organic N mineralization (Van Kessel et al., 2000). Net N mineralization in an incubation study ranged from –29.2 to 54.9% for several different dairy manure samples (Van Kessel and Reeves, 2002). While manure sample analysis found compositional differences in the samples, there was no strong correlation between various composition parameters and organic N mineralization (Van Kessel and Reeves, 2002). Van Kessel and Reeves (2002) identified no relationship between near-infrared spectral characteristics and N mineralization, but Qafoku et al. (2001) found a strong correlation (r2 = 0.82) between PMN and predicted N in poultry litter using near-infrared reflectance spectroscopy (Fig. 2) .
Haney et al. (2001) evaluated the relationships between the flush of CO2 during 1 d after rewetting of dried soil and potential soil N mineralization in the laboratory and forage N uptake in the field following dairy cattle manure application. The flush of CO2, or mineralized C, during 1 d was highly correlated with potential N mineralized in soil after a 24-d incubation and forage N uptake in the field. There was poor correlation between residual inorganic soil N and N mineralized during the same incubation. The C mineralized in 1 d after soil rewetting represents a rapid and reliable laboratory test to estimate potentially mineralizable N in manure-amended soils.
The potential for NH3 volatilization from many land-applied by-products, particularly poultry litter and anaerobically digested and alkaline biosolids, poses environmental and agronomic dilemmas. Concentrated animal feeding operations and wastewater treatment facilities can also be significant sources of NH3 emissions. Animal agriculture contributes approximately 50% of the total U.S. anthropogenic ammonia emissions (van Aardenne et al., 2001). Such N loss reduces the agronomic fertilizer value of the by-products and contributes to environmental problems such as acid deposition, eutrophication, reductions in biodiversity, and human and animal health concerns. In addition, NH3 exacerbates odor problems and reacts with other atmospheric constituents to produce haze. Conversely, NH3 volatilization could be viewed as advantageous by allowing greater residual application rates through reductions in the concentration of total N.
Quantification of NH3 emissions from various sources including CAFOs, wastewater treatment facilities, and land application of by-products is critical because of increased interest in the environmental effects of atmospheric NH3 and the lack of research on this topic. Unfortunately, the methodologies for measuring of NH3 emissions in the field limit our capabilities to predict N loss. Laboratory measurements are more useful for determining relative differences in NH3 volatilization potential than absolute field balances. Methods to accurately quantify NH3 losses in the field have not yet been developed.
Chemical amendments can reduce NH3 loss from poultry manure. Moore et al. (1995) tested several Al, Fe, and Ca amendments for reducing N loss from poultry litter via volatilization. Calcium hydroxide did not significantly affect NH3 loss in litter. Ammonia volatilization was reduced by 11 and 58% with high and low concentrations of FeSO4·7H2O, respectively, and by 24 to nearly 100% with alum + CaCO3 or alum alone. The higher alum rate resulted in a twofold increase of the N concentration in the litter, which increased the value of the material as an N fertilizer. Moore et al. (2000) also demonstrated that poultry health and production could be increased by alum reduction of NH3 emissions (Table 1).
Methods of application of by-product influence NH3 losses as well. Hansen et al. (2003) compared NH3 losses from six injection techniques with surface broadcast of liquid cattle manure and found subsurface placement reduced NH3 losses 20 to 75%.
Composting, a commonly employed organic waste processing method, can increase NH3 volatilization from manure depending primarily on the C to N ratio. Organic and inorganic amendments can also affect NH3 volatilization during the composting process. Kithome et al. (1999) evaluated the addition of adsorbents and chemical amendments to reduce NH3 emission during the composting of poultry manure. Adsorbents, such as zeolites, clay, and coir (fibrous material that makes up the thick mesocarp of the coconut fruit), are employed to bind NH3 and NH+4. Inorganic chemicals, such as CaCl2, CaSO4, MgCl2, MgSO4, and Al2(SO4)3, will inhibit the production of NH3. Ammonia loss from the unamended manure ranged from 14 to 29% of total N during the first 3 d and increased to 47 to 62% after 25 d. Manure amended with 38% zeolite and 33% coir reduced NH3 losses by 44 and 48%, respectively. The decreased volatilization resulted in greater extractable N in amended composts, thereby increasing the N fertilizer value. Alum also reduced NH3 volatilization, and increased the total N concentration in the resulting compost material.
Soil NO–3 accumulation and leaching to ground water are increased on land that routinely receives animal wastes. Nitrate from the Mississippi River basin has been implicated as one of the primary causes for the hypoxic zone in the Gulf of Mexico (USGS, 2000) (Fig. 3) . Efficient N management is vital for reducing N loss through both surface and subsurface transport. Recent interest in hypoxia and in N-based land application systems has generated interest in improving N fertilizer recommendation models to minimize nitrate losses from the landscape.
Andraski et al. (2000) evaluated four cropping–manure management system effects on soil water NO–3 concentrations and leaching below the root zone in corn. Early-season soil profile NO3–N contents varied among systems. Soil profile NO3–N contents increased by 80 to 210 kg N ha–1 from April to June in systems that had received past manure applications. The continuous corn system, with no manure history, had an increase of only 20 kg N ha–1. Management systems with a manure history had the greatest soil water NO3–N concentrations. Total NO3–N leaching amounts were greater for management systems with a history of manure application than for cropping systems with no manure history, and the economically optimum N rate was lower for sites with a history of manure applications (Table 2).
Several processes associated with the production and land application of by-products can lead to greenhouse gas emissions. The primary gases of concern are CO2, CH4, and N2O. Carbon dioxide released from digestive or waste handling processes has no net effect on global warming potential because the C was taken from the atmosphere via photosynthesis. If C is released as CH4, however, there is an effect on global warming potential because CH4 is a more potent greenhouse gas than CO2. In addition, N2O is a greenhouse gas and plays a role in the destruction of stratospheric ozone and waste management practices can influence the amount released to the atmosphere.
Hao et al. (2001) measured C and N2O losses from passive (no turning) or active (turned six times) composting methods for cattle feedlot manure. Most C loss was as CO2; however, a substantial amount of C was lost by CH4 emission. Carbon losses as CO2 and CH4 were 73.8 and 6.3 kg C Mg–1 manure, respectively, for the passive method and 168.0 and 8.1 kg C Mg–1 manure, respectively, for the active method. Nitrogen loss through N2O emission was 0.11 and 0.19 kg N Mg–1 manure for the passive and active methods, respectively. Consumption of fuel to aerate and maintain the manure windrows added an additional C loss of 4.4 kg C Mg–1 manure for the active method. Total greenhouse gas emission, expressed as CO2–C equivalent, was 240 and 401 kg C Mg–1 manure for the passive and active composting methods, respectively. Aerating the manure increased gas emissions due to greater biological activity, increased N cycling, and increased gas diffusion. A smaller gas diffusion rate and incomplete decomposition reduced gas emissions in the passive manure treatment.
Most estimates of N2O emission have been based on laboratory tests or from short-term field studies. Lessard et al. (1996) found that about 1 kg N ha–1 was lost due to N2O emission during 185 d from soil that received two applications of cattle manure for 2 yr. The effect of long-term manure application on N2O emission has been largely ignored. Chang et al. (1998) studied the effect of long-term manure application (21 yr) on the annual emission of N2O and whether emission rate was related to various environmental factors. Emission rates ranged from 2 to 4% of total N applied manure. These rates are much greater than results from the Lessard et al. (1996) short-term study. Greater emission rates from the long-term study may be the cumulative effect of repeated manure applications over several years and/or the mineralization of organic N reserves. The relationship of different combinations of environmental factors only accounted for 30% or less of the variability in N2O flux. The rate of N2O emission was greatest in the spring, but flux rates were significant throughout the winter months.
Similar to NH3 emissions, application method can greatly influence N2O emissions. Flessa and Beese (2000) compared N2O flux from surface and injected liquid cattle manure and injection greatly increased flux (Table 3). The researchers did not measure NH3 emissions, but injection for the purpose of reducing NH3 emissions may increase N2O emissions.
Relative Availability of Phosphorus in By-Products and By-Product Amended Soils
Interest in the relative availability of P in by-products and in by-product amended soils has increased for two reasons. The first is the goal of reducing P availability to reduce its potential environmental impact, and the second is accounting for relative differences in by-products that might be useful in determining relative risk of P loss for assessment tools such as the P index.
Wastewater treatment processes can influence the availability of P in biosolids. Of particular interest is the addition of metal salts, specifically Fe and Al, and lime used in some wastewater treatment processes. A decrease in the P solubility of biosolids resulting from certain chemical additions could reduce the risk of P loss to surface water.
Maguire et al. (2001) tested the effect of lime and metal salt addition and digestion on the extractability of P in eight biosolids and one poultry litter product in a 51-d incubation study using two soils. Six of the eight biosolids were treated with metal salts (usually Fe), two received addition of lime and metal salts, and two were not treated with either. Five of the biosolids underwent anaerobic digestion. Each of the biosolids and the poultry litter were incorporated at a rate of 8 Mg ha–1. The addition of biosolids consistently increased the water-soluble phosphorus (WSP) concentrations compared with the unamended soil. A rapid decrease in WSP was observed after the first week of incubation, followed by a more gradual decrease until week three, and minimal change thereafter. Biosolids not treated with a metal salt showed the greatest increases and sustained the greatest concentrations of WSP, Fe oxide strip–extractable P, and Mehlich-1 P throughout the incubation. Increased soil pH associated with the application of biosolids treated with lime may have resulted in the greater P extracted from those soils. Results suggested the possibility of predicting changes in soil P following additions of biosolids using the biosolids properties. Both the Mehlich-3 soil test and the molar ratio of P to (Al + Fe) measured with acid-peroxide digestion or oxalate extraction resulted in "reasonably accurate" predictions of soil P changes after biosolids application.
O'Connor et al. (2004) assessed phytoavailability of P in 12 biosolids relative to triple superphosphate (TSP) fertilizer in a greenhouse study using two P-deficient, sandy soils. Relative phytoavailability ranged from very low (not detectable) to greater than 100%. The materials were placed into groups of high (>75% of TSP), moderate (25–75% of TSP), or low ( TSP) phytoavailability. The high group contained two biosolids produced by a biological P removal process and without any metal salt additions. The moderate group had biosolids produced by conventional treatment processes, including composting. Within the low group were materials with high total Fe and Al concentrations and that were developed into products with low moisture content.
Reducing the solubility of P in manure before land application with chemical amendments (e.g., alum) can reduce P loss in runoff. Moore et al. (2000) (Table 4) and Shreve et al. (1995) reported increased dissolved reactive and total P concentrations in runoff from pastures amended with untreated poultry litter compared with alum-amended litter applications. Dou et al. (2003) tested coal combustion by-products and alum as amendments for reducing soluble P in dairy, swine, and broiler litter. Alum additions reduced readily soluble P (inorganic P) by 99, 86, and 80% in the dairy, broiler, and swine samples, respectively, as compared with the control. Fluidized bed combustion fly ash and flue gas desulfurization by-product decreased readily soluble P by 50 to 80% at the highest application rates. Sequential extraction results from both the alum and by-product treatments suggest that manure P was shifted from the water-soluble fraction into more stable fractions. These shifts reduce the potential for P transport offsite as soluble P in runoff, but have minimal influence on the long-term P availability to the crop.
Robinson and Sharpley (1996) compared the short-term extractability, sorption, and fractionation of P in soils treated with poultry litter leachate or KH2PO4. The short-term extractability of P was lower in the soils treated with litter leachate than with KH2PO4, as measured by extraction with Fe-oxide strips. Soils treated with KH2PO4 had an average of 50% more Fe-oxide strip P than soils treated with litter leachate. Following removal of solution P by water extraction, leachate-treated soils had greater amounts of NaHCO3–P than KH2PO4–treated soils. Phosphorus sorption was greater from the litter leachate than from the KH2PO4 solution for all soils. Averaged across all soils, P sorption maximum was 548 and 304 mg kg–1 for the leachate-treated and KH2PO4–treated soils, respectively. Results suggest that poultry litter may maintain more residual P in soil than KH2PO4 due to lower solution P and strip P, but greater NaHCO3–P.
Siddique and Robinson (2003) investigated P sorption and extractability in soils receiving poultry litter, poultry manure, cattle slurry, and urban sewage sludge. All of the organic sources of P and KH2PO4 decreased the P sorption maximum and binding energy and increased CaCl2–P and resin P in all the soils. The changes increased in the order: poultry manure sewage sludge 2PO4suggested that soil solution P concentrations in the poultry manure, litter, and sewage sludge treatments were controlled by large amounts of available Ca. Organic acids released in the cattle slurry treatments were hypothesized to increase solubility compared with the KH2PO4–treated soils by blocking P sorption sites, thus reducing the P sorption capacity. Findings from Robinson and Sharpley (1996) and Siddique and Robinson (2003) suggest that P from inorganic sources that contained organic by-products does not undergo identical reactions in soils and that these differences should be taken into account when addressing agronomic or environmental issues from P. Few states or regulatory agencies address the potential differences, though a few states are moving to incorporate phosphorus availability coefficients (PACs) for various organic sources of P in P index values (Shober and Sims, 2003).
Phosphorus in Runoff and Erosion
Application of animal manures and biosolids based on the N requirement of the crop typically results in overapplication and accumulation of P in the soil. High concentrations of soil P increase the potential of P loss and surface water quality degradation. Phosphorus loss in surface runoff is mainly controlled by transport (i.e., runoff and erosion) and source (e.g., manure, biosolids, or fertilizer applications, soil test P level) factors (Pote et al., 1996).
Eghball and Gilley (1999) evaluated the effects of N or P-based applications of beef cattle manure or compost on runoff concentrations of dissolved P, FeO sink P, particulate P, and total P from no-till and disked fields using rainfall simulations. Runoff concentrations of dissolved and FeO sink P were significantly greater from the no-till than the disked treatment. Tillage did not affect total P concentration in the runoff. Nitrogen-based manure and compost applications resulted in runoff dissolved P concentrations significantly greater than 1 mg L–1; however, dissolved P concentrations for the P-based treatments were not significantly different than 1 mg L–1 during both simulated rainfall events. This indicates that N-based manure or compost application to no-till systems will result in P enrichment of surface water bodies. The researchers concluded that P-based manure and compost application appears to achieve beneficial agronomic and environmental goals with respect to P.
Waste source and treatment processes may explain differences in P availability and their potential loss in surface runoff (Frossard et al., 1996). Withers et al. (2001) compared P transfer in surface runoff from plots amended with inorganic fertilizer, cattle manure, and sewage sludge. The specific amendments were TSP, liquid cattle manure (LCS), liquid anaerobically digested sewage sludge (LDS), and dewatered cake sludge (DSC). Laboratory analysis of the waste materials found substantially more P extracted from the cattle manure compared with the sludges using water and NaHCO3. The liquid amendments (LCS and LDS) had higher HCl-extractable P than the DSC sample. High P solubility for both the NaHCO3– and HCl-extractable fractions in the cattle manure indicated that most P was inorganic. Based on laboratory analyses, the relative differences in the potential P release to water can be ranked in the order TSP > LCS > LDS > DSC. Application of either inorganic P fertilizer or liquid cattle manure transferred more P to surface water than application of liquid or dewatered biosolids. This is probably due to lower P solubility in water and/or NaHCO3 associated with the biosolids amendments.
Many studies have measured edge of field P losses or concentrations in runoff; however, there is a strong need for interpretation of such data relative to potential environmental impact. Such an interpretation would probably be watershed specific, but should be more instructive than merely making relative comparisons or comparing dissolved P concentrations in runoff with an arbitrary criterion (e.g., 1 mg P L–1).
Phosphorus that accumulates in the soil can be susceptible to loss via leaching, particularly in coarse-textured soils where large amounts of animal manure or biosolids have been applied. Leaching of soil P can lead to ground water contamination in areas with shallow water tables or contribute to the degradation of surface water through subsurface flow, particularly when tile drainage is used. Phosphorus can leach in sandy soils because these soils quickly become P saturated due their low contents of P-binding constituents (e.g., clays and Fe- and Al-oxides). The saturation of the P sorption capacity in the surface soil could lead to further saturation of the deeper soil horizons (Sims et al., 1998).
Novak et al. (2000) evaluated the effects of long-term (10 yr), intensive swine manure application on soil P accumulation and movement in a Coastal Plain soil. Three soil cores were taken within the spray field in 1991 (4 yr of manure application) and 1997 (10 yr of manure application). In 1991, the mean surface soil (0–15 cm) value of Mehlich-3 P was 376 mg P kg–1 soil, substantially greater than the maximum soil test value for crop production (100 mg P kg–1). High values (168 mg P kg–1) were also observed down to the 30- to 45-cm depth, suggesting that 4 yr of manure applications was sufficient to cause P accumulation in the topsoil and some leaching. In 1997, soil Mehlich-3 P concentrations were significantly greater for the spray field compared with the control area down to 90 to 135 cm (Fig. 4) . In a similar study using cattle feedlot manure, Eghball et al. (1996) found greater STP concentrations to 1.8 m in soils amended with manure compared with inorganic fertilizer. Minimal amounts of fertilizer P had moved deeper in the soil than 1.1 m. At similar P loading rates, P movement was deeper in the manure treatment than the fertilizer treatment. The researchers found no correlation between the P adsorption maximum or P index and soil P movement.
Phosphorus leaching was low in Coastal Plain soils amended with biosolids whose high concentrations of Al and Fe increased P sorption and reduce P solubility (Elliot et al., 2002). Similar results were found by Siddique et al. (2000) in a study on soils whose surface horizon texture was loam. Significantly greater amounts of P leached from inorganic fertilizer–amended soil than sludge-amended soil. The difference in P mobility was explained by the lower P solubility in the sludge compared with the fertilizer.
New Tools for Understanding and Minimizing the Environmental Impacts of Phosphorus
Phosphorus Loss Assessment Tools
In 1999, the USDA and the USEPA issued a national strategy outlining P-based nutrient management approaches for the land application of animal manure in response to increasing concern over nonpoint-source P pollution of surface waters associated with the agronomic use of animal wastes. The three approaches base P management on either (i) STP and crop response, (ii) an environmental soil P threshold limit, or (iii) a P index that identifies fields with the greatest risk for P loss. Relative advantages and disadvantages of each approach were discussed previously. The P index identifies areas where the risk of P transport is probably greater than other areas by accounting for source and transport factors that control P loss in runoff. Several investigators have demonstrated that a P index can accurately rank sites based on their potential for P loss in runoff (Birr and Mulla, 2001; Gburek et al., 2000; Sharpley, 1995).
An evaluation of the three P-based nutrient management strategies was conducted on 10 Pennsylvania farms using site management and transport factors from the Pennsylvania P index (Weld et al., 2002). The revised universal soil loss equation (RUSLE) and soil surveys were used to calculate soil loss for each site. A surface runoff class was assigned based on the soil permeability class and slope. Also included in transport factor category was contributing distance, generally defined as the distance between a given field and the edge of a body of water. Source factors included soil test P (Mehlich-3 P), fertilizer rate, fertilizer application method, manure rate, manure application method, and P availability coefficient. A P index value was calculated by multiplying the summed transport, source, and management factors. Results from this study found that the P index was less restrictive, more flexible, and allowed for more management options compared with the other two strategies. In addition, the P index strategy had the lowest first year total cost estimate across all 10 farms.
Modifications of Animal Diets to Reduce Phosphorus Losses in Runoff
Livestock diet modifications can change the digestibility, forms, and amounts of P in feed, and can subsequently alter the quantity and forms of P excreted in manure. Dairy, swine, and poultry are the animal species for which these modifications are most appropriate. Three strategies can be used: (i) reductions in P supplementation to the diet without reducing animal performance, (ii) supplementation of phytase enzyme to increase the digestibility of phytic acid P in the diet, and (iii) use of high-available-phosphorus (HAP) corn in the diet. These changes can reduce P losses in runoff following manure land applications.
Swine manure has some of the greatest P contents compared with other livestock (Barnett, 1994). Typical swine diets include corn and/or soybean meal, whose P is largely in the form of myo-inositol hexakisphosphate (phytate). Swine are unable to use phytate P due to insufficient amounts of the phytase enzyme in their upper digestive tract. For this reason, supplements of inorganic P are often used in swine diets, thus further increasing the amount of P excreted in the manure.
Microbial phytase and HAP are two products that can eliminate the need for inorganic P supplement in swine diets. Microbial phytase catalyzes the hydrolysis of phytate in the upper digestive tract and makes orthophosphate available for use. Several studies have shown that microbial phytase is effective in improving P adsorption and reducing P excretion. High-available-P corn contains substantially less phytate than normal corn, but does not differ in total P content. This product is not commercially available at present but may be a promising substitute for normal corn in swine diets. High-available-P corn diets for swine and young poultry have resulted in improved P utilization and decreased P excretion compared with diets using conventional corn (Spencer et al., 2000). Phosphorus digestibility was further increased when phytase was added to HAP corn diets (Sands et al., 2001).
Baxter et al. (2003) evaluated the effects of swine diets formulated with normal and HAP corn, with and without phytase additions, on the relative amounts of various P fractions in the manure. The researchers also studied the changes in the relative contents of these P fractions when manure was stored as slurry. Fecal total P concentration was reduced 17, 19, and 40% in diets formulated with phytase, HAP corn, and phytase plus HAP corn, respectively (Table 5). Phytase supplements had no effect on water-soluble dissolved molybdate-reactive phosphorus (DRP) when added to the normal corn diet, but significantly reduced DRP by 27% when in the HAP corn diet (Table 5). Fecal phytate P was significantly reduced in all the treatments compared with the control diet, with the HAP plus phytase diet having the lowest phytate P concentration (Table 5). Reductions were observed across all diets in the relative pools of DRP, phytate P, dissolved organic P, and acid-soluble organic P over a storage time of 150 d. Significant effects of diet on the relative amounts of different P pools, averaged across storage time, indicated that diet may affect the availability and types of P in slurry that contains different-aged material.
Peak et al. (2002) used XANES to study P forms in alum-amended poultry litter. Alum amendment converts P from weakly adsorbed and Ca phosphate forms to forms associated with amorphous Al(OH)3.
SUMMARY AND RESEARCH NEEDS
Nitrogen mineralization and nitrate leaching remain important today for land application programs. In addition, new regulations and existing and emerging environmental issues warrant additional attention and research. Nutrient management legislation, eutrophication from both N and P, and NH3 and greenhouse gas emissions are relevant examples. Advances in research methods will allow us to further our knowledge in a number of areas and will increasingly influence our research approaches and productivity.
Research needs are either general in nature and apply to all plant nutrient issues, or are specific for N or P. A framework for determining plant nutrient issues that can be used for all land-applied materials is needed. Formal guidelines exist for manures and biosolids, but not for other materials or for mixtures of materials. This forces managers to conduct a complete evaluation for new materials or mixtures when this may not be necessary. Similarly, modeling exercises should supplement field or laboratory evaluations and investigators are encouraged to perform extensive characterizations of materials and conditions to support model calibration and validation. Best management practices for nutrients need to be integrative across other issues related to sustainable land application such as metals, pathogens, and aesthetics. The effects of BMPs at the watershed scale warrant additional investigation. Ancillary benefits from land application, such as C sequestration, plant disease suppression, and soil quality benefits, need to be further explored. Overall, a systems approach should be adopted in more studies. Relevant examples include the effect of animal diet modifications ranging from animal productivity to water quality benefits, or the influence of injection of manures and biosolids for decreasing NH3 emissions while increasing N2O emissions.
Quick tests for PMN offer the promise of improving our estimates of PAN, and possibly reducing environmental issues associated with N from land application of residuals, and warrant further investigation. It is possible, however, that variability in residuals and issues with precision and accuracy during application would overshadow small improvements in PAN estimates, and this should be considered as well. Basic research is needed on the mineralization process. This will likely involve the study of microbial and macrofaunal ecology of the process as affected by temperature, water content, and soil characteristics. Similarly, further research is needed to identify pools of mineralizable N and their relationship to easily measurable parameters in residuals. Interest in NH3 as an atmospheric pollutant will likely increase, and studies on amendments for reducing NH3 emissions is needed along with field-based techniques for estimating NH3 losses. Feedlots and waste lagoon remediation after facility closure will become an important topic. These sites have high NH4–N concentrations and represent significant and long-term sources of NO3–N for ground water when nitrification occurs.
Numerous research needs are evident for P. Work should continue on environmentally available P in residuals with some consideration of a national P index. In a related issue, much needs to be learned about interpretation of edge of field P losses and the fate of P between the edge of the field and the point of impact. Strategies for removing or reducing the solubility of P in residuals warrant further investigation. Regulations that prohibit residuals application, or restrict them to less than N-based applications, may discourage land application in the long term and encourage disposal strategies that are not based on beneficial reuse. In addition, the long-term efficacy of alum and other amendments used to reduce P bioavailability in soils and residuals is unknown. Finally, the continued use of new and improved spectroscopic techniques will move us toward a more mechanistic understanding of environmental P chemistry.
Contribution no. 04-308-J of the Kansas Agricultural Experiment Station, Manhattan.
Fig. 1. Predicted versus measured potentially mineralizable N in biosolids. Data symbols correspond to field studies conducted in Arkansas, Michigan, Virginia, and Washington in 1998 and 1999. From Gilmour et al. (2003).
Fig. 2. Measured potentially mineralizable N for broiler litter versus that predicted using near infrared spectroscopy (NIRS). From Qafoku et al. (2001).
Fig. 3. Estimated N source inputs over time from the Mississippi River watershed. From USGS (2000).
Fig. 4. Mehlich 3–extractable P concentrations with depth after 10 yr of intensive swine manure applications. Data from Novak et al. (2000).
Fig. 5. X-ray absorption near edge structure (XANES) spectroscopy of a high P soil and three P standards (P adsorbed to ferrihydrite, octacalcium phosphate, and hydroxyapatite). A linear combination of the spectra from the three standards matches the spectra from the soil and suggests P may exist in similar forms in the soil. From Beauchemin et al. (2003).
Table 1. Alum effects on ammonia flux and poultry production.
All treatment effects are significant at the 0.05 probability level. Data from Moore et al. (2000).
Table 2. The effect of crop rotation and manure application history on nitrate leaching and the economically optimum N rate.
Values represent the average over two years and two N rates. Data from Andraski et al. (2000).
The amount of N leached was calculated from estimated soil drainage based on water balance and the mean monthly soil water NO3–N concentrations.
Table 3. The influence of injection on N2O fluxes (±SD) from land-applied cattle slurry.
Data from Flessa and Beese (2000).
Table 4. The influence of alum amendments to poultry litter on dissolved reactive and total P concentrations in runoff from surface applications of litter to pasture.
All treatment effects are significant at the 0.05 probability level. Data from Moore et al. (2000).
Table 5. The effect of diet modification on swine manure composition.
Data from Baxter et al. (2003).
All diets contained approximately the same total P content. PHY, phytase-amended; HAP, high-available-P corn.
Means with the same letter within a column are not significantly different at the 0.05 probability level.