LA hosphorus Accumulation and Loss from Alabama Soils Receiving Poultry Litter Builletin No. (-31 F[ebruary 1-9 Alabama Agricultural Experiment Station Dr. James E. Marion, Director Auburn LUniv'erst, Alabama CONTENTS INTRODUCTION .............................................. ............. 1 Phosphorus Retention and Buildup ....................... ........... 2 Soil Properties ............................................. .............. 3 Phosphorus Adsorption and Desorption................. .......... 4 Models Used to Evaluate P Adsorption.......... ........... 5 The Erosion Productivity Impact Calculator (EPIC) ............... 5 Solubility and Speciation of Solution and Solid P.................. 6 Phosphorus and Suspended Eroded Soil Material ................. 6 SUMMARY ........................................ ................... 8 METHODS ................................................... ............... 11 Soil Characterization ...................................... ................. 11 Paired Fields ............................................. ............. 11 Incubation .................................................. .............. 12 MINTEQA2/PRODEFA2 Ion Speciation......... .............. 13 Erosion Productivity Impact Calculator Simulation...... ......... 14 RESULTS AND DISCUSSION ......................... .................. 14 Soil Characteristics ........................................ ............. 14 Phosphorus Adsorption...................................15 Adsorption Isotherms................................... ................... 28 Soluble Phosphorus Determinations Following Incubation .......... 35 Predicting Long-term Phosphorus Accumulation, Leaching, and Runoff..... .................... 44 RELATED PHOSPHORUS RESEARCH IN ALABAMA ............ 54 Phosphorus Mobility Associated with Suspended Eroded Soil Material ....................................... 54 Watershed Study of Conventional- and Conservation-tillage Cotton Practices ....................................... 57 Long-term Phosphorus Levels Determined by Soil Test Analysis ............................................. 57 Soil Test Phosphorus Levels from Long-term Broiler litter Application .................................. 58 P Accumulation and Runoff from Litter and Conventional Fertilizer Applications ................. 58 LITERATURE CITED .......................................... 60 Information contained herein is available to all without regard to race, color, sex, or national origin. Phosphorus Accumulation and Loss from Alabama Soils Receiving Poultry Litter GREGORY L. MULLINS AND BENJAMIN F. HAJEK INTRODUCTION Each year, Alabama's broiler industry produces an estimated 1.5 million tons of litter, containing approximately 14,645 tons of phosphorus (P). Poultry litter is an excellent source of plant-available P for pasture and row crop lands if properly managed, but excess application rates can pose an environmental threat. Excess P can move from the soil into natural water bodies and lead to eutrophication. Broiler litter application to cropped and grassed agricultural land has increased the level of extractable P in soils (40). Increasing levels of litter application can increase the amount of P in runoff water (19, 20). Chemical fertilizer or manures come into direct contact with the soil; however, contact between litter P and the soil is reduced by the bedding portion of poultry litter and the practice of applying litter without incorporation. As a result, high percentages of P can be found in dissolved forms. In areas of concentrated broiler production, such as North Alabama, poultry litter may be more of a disposal problem than a resource management issue. Under these conditions, one may be more interested in the maximum amount of P that can be safely added to a given soil without harming the environment. An Alabama Agricultural Experiment Station (AAES) study was conducted to refine the best management practices for applying broiler litter to farmland. Objectives of the study included: (1) Determining the P adsorption characteristics and the maximum P adsorption capacity of selected soils in North Alabama as affected by the addition of poultry litter and fertilizer P; (2) Evaluating P desorption characteristics of North Alabama soils as affected by the addition of poultry litter and fertilizer P; (3) Evaluating and predicting P uptake by plants and P concentrations in soil water leaching through poultry litter incorporated and spread on soil surfaces, in runoff waters, through the soil profile, and below the root zone; (4) Evaluating environmentally acceptable P loading limits for North Alabama soils; and (5) Examining previous related research in relation to results of this study and in relation to P runoff losses. Mullins is a Professor and Hajek is a Professor Emeritus in the Auburn University Department of Agronomy and Soils. 2 Phosphorus Accumulation and Loss from Alabama Soils Phosphorus Retention and Buildup In soils with high phosphorus retention capacities, surface water eutrophication can occur from the erosion of P-rich soil material and from runoff water with high phosphorus concentrations. There is no reason for long-term commercial fertilizer use to cause problems of accumulation of excess soil P either in surface water or in ground water. What is needed is the use of P rates consistent with plant needs and with conservation-based management to minimize erosion and runoff. Management should include a regular soil testing program to prevent over application and buildup. Most inorganic P forms in soils are so insoluble that only a small fraction (less than 10%) is available at any one time for plant uptake (15). There is a large body of literature on the loss of P from fields due to erosion and runoff (70). However, literature on losses from fields receiving animal wastes is limited (40). Phosphorus loss from pastures can be significant for periods soon after application since it is not incorporated by plowing. Phosphorus loss can be significant even if erosion is slight. Best management practices for application of poultry litter should focus on controlling and reducing runoff and erosion. In addition to controlling overland transport, management practices that limit soil test P buildup to excessive levels should be developed (70). Estimating buildup of soil test P levels is a simple test available from all agronomic soil testing laboratories. Several studies have shown that soil test P and soluble P in runoff are related (68). This has lead to state and federal water quality management agencies attempting to identify "cut-off-levels" based on soil test P. Proposed "phosphorus indexing" methods have been developed to plan management strategies when fields receiving P fertilizers approach the cut-off-level (41). In the Netherlands, animal-based agriculture has resulted in high soil P levels such that regulations have been set to limit the additional application of P. A critical degree of P accumulation in a soil has been set and defined as 25% of the P adsorption capacity of the soil, calculated as follows: DPS = (Pact/PSCa)xl00 DPS is the degree of P saturation, Pact is the oxalate extractable P, and PSCa is the P sorption capacity. All values are expressed on the basis of soil surface area (70). In Alabama, the Auburn University Soil Testing Laboratory uses an "extremely high" soil test P rating as an environmentally critical level. The level is based on Mehlich-I extractable P. It is defined as four times the amount of a nutrient considered to be adequate for optimum plant growth (2). This level is considered excessive and further additions could be detrimental to the crop and may contribute to pollution of ground and surface waters. Alabama Agricultural Experiment Station 3 Soil Properties Goethite and gibbsite sorb phosphates through ligand exchange on exchange sites located on the crystalline edges of gibbsite and the surface of goethite (44). The adsorption of phosphate onto kaolinite, usually the dominant mineral in clay fractions of southeastern U.S. soils, has been proposed to consist of three types of binding sites which correspond to three regions of kaolinite-phosphate adsorption isotherms: Region I involves P adsorption at low P concentration (less than 10-4 M P) onto high affinity sites; Region II begins at P concentrations at approximately 10 4 M P; and Region III is the linear portion of the isotherm from medium to high P concentrations (10 3 to 10-' MP). Phosphorus is first sorbed onto sites in regions I and II. These are proposed to represent two-OH sites associated with Al at the edge of a kaolinite crystal octahedral sheet and the Al associated with exchange sites (47). Region III may be a poorly crystalline region of the clay surface, or may represent a site that sorbs a small amount of P compared to the number of sites (17). Several soil chemical and physical properties have been shown to affect P sorption. Jones et al. (37) separated soils into four classes for predicting the P sorption coefficient that is used in the Erosion Productivity Impact Calculator (EPIC). The first separation is between calcareous and non-calcareous soils. In calcareous soils sorption is a function of CaCO 3 content. Non-calcareous soils are classed into slightly, moderately, and highly weathered. In slightly weathered soils the P sorption coefficient (PSP) is considered a function of active and labile P. In moderately weathered soils, PSP is estimated using percent base saturation and pH, and in highly weathered soils clay content is used. Mullins (48) evaluated the P adsorption characteristics of Benndale, Hartsells, Lucedale, and Dewey soils as affected by long-term fertility treatments. Sorption of P within a soil was affected by the rate of added P and past fertility treatments. P adsorption capacity decreased with decreasing clay content. For example, the Dewey (27% clay) soil had the highest P sorption capacity, and the Benndale (4% clay) soil had the lowest P sorption capacity. In a literature review on the chemistry of soil P, Sanyal and DeDatta (57) summarized P sorption data where correlations had been made between P sorption parameters and soil properties. They showed that several workers have reported a significant correlation of P sorption parameters with clay content. Sanyal and DeDatta (57) speculated that correlation between P sorption and clay content may reflect primarily the effect of surface area on P adsorption. For example, Sanyal et al. (56) reported a high degree of correlation between the Langmuir adsorption maximum and Freundlich K and the clay content of several acid and acid sulfate soils of South and Southeast Asia. 4 Phosphorus Accumulation and Loss from Alabama Soils Recently, Schunost and Schwertmann (60) developed pedotransfer functions proposed by Bouma (10) for modeling P availability. They related the P sorption capacity to clay content, citrate-bicarbonate-dithionite extractable iron (CBD iron), and pH. The iron (Fe) content could be replaced by a yellowness component (goethite). In addition, a one-point partition function was required. This relationship was developed for soils that have smectite: goethite soil mineral assemblages. Alabama is dominated by noncalcareous, highly weathered soils in the Ultisol soil order. Soils with smectite and goethite mineral assemblages are found in Alabama. They would be considered non-calcareous moderately weathered soils. Calcareous soils in Alabama are found in the Blackland Prairie Soil Provence (28). Phosphorus Adsorption and Desorption When water soluble P fertilizers are applied to soils, most of the added P is converted very quickly to one of many possible low-soluble forms of P. Soil pH is one of the most important factors that affects the forms of P in soil and the solubility and availability of soil P. In acid soils, P is precipitated as insoluble Fe- or aluminum (Al)-phosphates or is adsorbed on the surface of hydrous oxides (75). In calcareous soils, P is precipitated as calcium (Ca)phosphates of low solubility. When considering adsorption and the solubility of both Fe- and Al-phosphates and Ca-phosphate minerals, a pH of 6.5 generally corresponds to a maximum solubility and availability of soil P. Adsorption by ligand exchange between phosphate and H20 and/or OH is believed to be the primary mechanism for P adsorption on hydrous oxide surfaces. Phosphate retained through a single coordinate linkage is considered to be labile (i.e. weakly adsorbed) and readily available to plants. In contrast, P retained through a double coordinate linkage is difficultly available (75). As summarized by Stevenson (75), Riley and Syers (55) proposed three mechanisms for P adsorption in soils and on the surfaces of Fe-oxide gels. The first mechanism is by chemisorption at protonated surface sites with the surface release of one H20 for each phosphate ion. A second mechanism is chemisorption by replacement of surface OH groups by adsorbed phosphate ions. The third mechanism is believed to operate at high concentration and represents a physical sorption of P as a potential determining ion. Phosphate may also interact with the surface of clay minerals (75). At exposed crystal edges, phosphate can be bound by replacement of a OH group from exposed Al atoms. On clay surfaces, phosphate may be held by forming a linkage with exchangeable Ca. This extent of adsorption is considered to be more prevalent for 1:1 type clays The high affinity of oxide and clay surfaces for soluble P results in a low rate of release of added P. Thus, the release or desorption of adsorbed P is very slow to almost being irreversible, resulting in a large hysteresis effect Alabama Agricultural Experiment Station 5 (57). Hysteresis effects make the use of P adsorption isotherms unsuitable as a means of estimating the solubility and release of added P. Models Used to Evaluate P Adsorption Phosphorus retention by soils has been typically evaluated using adsorption isotherms. Adsorption isotherms are constructed by plotting the amount of P adsorbed per unit weight of soil as a function of the equilibrium P concentration in solution. It is assumed in the use of adsorption isotherms that the P removed from solution is sorbed, that precipitation does not occur or is accounted for and that equilibrium conditions have been reached (9). The two adsorption models most commonly used to study P adsorption include the Langmuir and the Fruendlich equations. The Langmuir equation was developed on the basis of kinetic theory and was initially derived to study the adsorption of gasses on solids (9, 30). It is an attractive model since it produces a theoretical adsorption maximum and a coefficient that is theoretically related to a binding energy. The adsorption maximum is assumed to be a monomolecular layer on all reactive sites. The equation is commonly presented as: x/m = (kXb)/(1 +kX) where x/m is the weight of P per unit weight of soil, X is the equilibrium concentration of P in solution, b is a constant related to the binding strength, and k is the maximum adsorption capacity. If adsorption data do not conform to the Langmuir equation, the Freundlich equation is often used. The Freundlich equation was originally empirical, and is less demanding since it implies that the energy of adsorption decreases logarithmically as the fraction of covered surface increases. The equation is given as: x/m = kX M where k and n are empirical constants and the other terms are defined as for the Langmuir equation. The Erosion Productivity Impact Calculator (EPIC) Several soil and plant P models have been developed for specific objectives (36). However, none provide the long-term capability required for evaluating buildup and depletion simulations associated with litter application. A soil and plant model with long-term capability was developed as a component of EPIC, which is a comprehensive model developed for application to erosion-productivity problems (79). EPIC can be used to predict current-year crop yields using measured variables such as climate and soils, or it can be used to predict long-term yields using simulated weather based on nearest station input and various management strategies. The P part of the model simulates plant uptake, attenuation in soil, runoff P losses both in solution and adsorbed to soil particles, mineral P, organic P, labile P, active P, and 6 Phosphorus Accumulation and Loss from Alabama Soils fixed P (37). This is the most comprehensive model available to predict the buildup of P based on fertilizer and other management practices for long time periods. In addition, EPIC is an excellent tool to evaluate soil erosion using USLE, MUSLE, and other erosion models. EPIC has an extensive soil and climate data base for all regions of the U.S. Solubility and Speciation of Solution and Solid P Measurement and prediction of P concentrations in runoff waters is a key factor in determining mobility and transport of solution P and suspended solid P from fields, pastures, and woodland. MINTEQA2 is a geochemical equilibrium speciation model capable of computing equilibria among the dissolved, adsorbed, solid, and gas phases in an environmental setting (5). Several problems are often encountered in measuring total elemental concentrations in soils and water (43). Often concentrations are below detection limits of the methods used, organic complexes can chelate elements, and fine suspended clays may adsorb some ions that are determined as being soluble. MINTEQA2 can be used to address these and a wide variety of phosphorus solubility problems. Phosphorus and Suspended Eroded Soil Material Phosphorus is readily adsorbed onto amorphous oxides and hydroxides of Al and Fe (76) and onto short-range order aluminosilicates (47). The sorption of inorganic P is rapid. Because of the ubiquitousness of Al- and Feoxides and hydroxides in soils and their rapid sorption rate of P (32), they have been reported to be associated with up to 50% of the total soil P content (58). Soil P can be carried off-site by water erosion. Suspended solids, particularly the water-dispersible clay, can transport P to receiving waters. The forms of P entering waterways can be divided into dissolved and particulate P (74). The dissolved forms are those that pass through a 45-micrometer (gm) pore size filter. These include inorganic, condensed, and organic forms. Dissolved forms are more readily available to algae. The particulate forms are less readily available to algae. They also include inorganic, condensed, and organic forms. Of the particulate forms, the inorganic form contributes the most bioavailable P. This form includes P associated with Al and Fe hydrous oxides and non-apatite minerals. Most of the P (75-90%) in runoff from agricultural land is associated with suspended mineral and organic material (45, 63). Phosphorus in runoff from pasture, which contains little suspended material, may be mostly in the soluble form (11, 71). Excessive P in waterways can lead to plankton blooms and eutrophication. The quantity of P entering waterways is dependent upon quantity of P in the soil, topography, vegetative cover, quantity and duration of runoff, and land use (39). Phosphorus loss from agricultural lands occurs from its leaching from plant material during rainfall, dissolution of fertilizer material, desorption of soil P by runoff, and the erosion of P-laden soil (59, 62). Alabama Agricultural Experiment Station 7 Scarseth and Chandler (58) reported that 60% of superphosphate applied over a 26-year period was lost by erosion from a nearly level Norfolk loamy sand soil planted to a cotton, corn, oat, and legume rotation. A 63% loss of applied P was attributed to erosion in a Hartsell fine sandy loam of 2-4% slope (21). Schuman et al. (59) applied P at the recommended rate of 39 kilograms per hectare (kg/ha) in Treynor, Iowa, and at 2.5 times this rate to contour-planted corn and pasture. The average P loss for the normally and heavily fertilized plots, over a three-year period, were 0.691 and 1.221 kg/ha per year. They divided P loss into solution P and aggregate-bound P. The solution P for the normally and heavily fertilized plots averaged 0.110 and 0.171 kg/ha per year and the aggregate-bound P averaged 0.581 and 1.050 kg/ha per year. The concentration of soil-bound P in runoff water tends to be considerably higher than its concentration in the original soil. Enrichment ratios are used to compare the P adsorbed onto an eroded particle or aggregate compared to the amount of P in an equal mass of soil. Enrichment ratios greater than one indicate that the eroded soil material contains more P than the soil from which it eroded (8, 80). Enrichment ratios ranging from 0.95 to 1.96 have been r'eported for various-size aggregates eroded from plots subjected to different tillage methods in experiments conducted by Alberts and Moldenhauer (4). They observed that the 0.21-0.05-millimeter (mm) and the less than 0.05-mm fractions had the highest enrichment ratios. Sharpley (61) studied the effect of the amount of P applied to a soil and the enrichment ratio of eroded soil material. Enrichment increased with increasing P addition. Enrichment ratios were greatest in the 5-2-gm and less than 2-gm fractions. Barrows and Kilmer (8) hypothesized that low amounts of soluble ions in runoff water may be due to the reabsorption of the ions in solution by the colloidal material in runoff. Sharpley et al. (62) observed that the soluble P concentrations in surface runoff from several cropped and grassed watersheds decreased as the amount of eroded soil material increased. In experiments using soil from these watersheds, soluble P added in rainfall was adsorbed by the eroded soil material transported in the runoff. Their results indicated that suspended eroded soil material may act as a P sink rather than as a P source. They also stated that the magnitude of the sorption was more dependent on the sorptive capacity of the eroded material than that of the surface soil. Section 319 of the 1990 Water Quality Act gives states the responsibility of developing management plans to limit P pollution from agricultural lands and aquacultural operations. Farmers must have a best management plan developed by the Alabama Department of Environmental Management (2). Currently, the amount of broiler litter that is applied to crop land and pasture is based primarily on nitrogen (N) content of the litter and crop N requirements. However, recent Alabama Cooperative Extension Service publications discourage additional P applications, especially in broiler litter and other animal manures, when soil test P levels reach the "Extremely High" rating. 8 Phosphorus Accumulation and Loss from Alabama Soils SUMMARY In the fall of 1990, soil samples were collected from the Ap and Bt horizon of three major soil series found within the poultry-producing region of North Alabama. The soils sampled included a Hartsells (fine loamy, thermic, siliceous Typic Hapludults), Linker, and Dewey (clayey, kaolinitic, thermic Typic Paleudults) series. According to available records, none of the sites had been treated with poultry litter. The samples were used for laboratory soil characterization and to determine P adsorption-desorption of soils treated with both mineral fertilizer P and broiler litter P. In 1991, a paired pasture and a paired cotton field were sampled to one meter. One field in each pair had received litter for 18-20 years, and the other was under conventional N and P fertilization. Resident mineral (total and resin extractable) and organic P were determined to sampled depths. EPIC was used to simulate P uptake, leaching and runoff losses under fescue and cotton receiving litter and under conventional fertilization. Data obtained by adsorption isotherm analysis, incubation of litter, and monocalciPum-P-treated samples, analysis of field soil samples, EPIC simulation of P loading, and fitting equilibrium distribution data to various models show these North Alabama soils have high P adsorption capacities. The P adsorption capacity is essentially the same when based on clay content. Following a three-month incubation period of monocalcium-phosphate- (MCP) and broiler-litter-treated soil samples, significantly more P was soluble and leached from MCP-treated soil than from broiler-litter-treated soil with comparable amounts of P added. Measured and predicted results show that large amounts of P are retained on clay (5,000 mng/kg) before dissolved P concentrations reach 1,000 gg per liter, a concentration considered as a limit for disposal of treated water from sewage treatment plants. In watershed studies in North Alabama, dissolved P in runoff water exceeded this limit even when conventional application rates recommend by soil test were used, regardless of P source (litter P or mineral fertilizer P). The dissolved P concentrations are related to date of application, degree of incorporation, management schedules, and time after application when runoff events occurred (29, 72). Reviews of available literature for research conducted outside of Alabama (66, 65) show that dissolved P in surface runoff can exceed this limit under a variety of cropping systems, with and without the addition of fertilizer P. Runoff P losses are more related to management practices than they are to the source of P or, to a point, the amount of resident soil-P.A 20-year EPIC simulation of conventional P application and of annual 7,200 kg/ha litter applications indicated that the annual average dissolved P concentration of runoff does not change appreciably over a 20-year period, while total soil P levels increased from 500 mg/kg to 1,500 mg/kg. Alabama Agricultural Experiment Station 9 The amount of P that soil can adsorb depends on its P adsorption capacity and its soil mineralogy. Results of particle-size and mineralogy characterization from this study indicate that the P-adsorbing capacity of these soils should be governed by their kaolinite, HIV, gibbsite, and goethite contents. The clay percentage should be a more relevant factor in describing P movement since active P-sorbing minerals are in the clay fraction. Approximately 40% to 60% of the water-dispersible clay fractions in runoff from highly weathered Ultisols in North Alabama should be kaolinitic. Hydroxy-interlayered vermiculite (HIV) would comprise about 15-20% of the water-dispersible clay of suspended solids. Mica, goethite, and gibbsite would make up the rest. More accurate descriptions of P adsorption mechanisms and movement possibly could be attained by utilizing mineral percentages in adsorption models. However, the analysis procedures are complex and not suitable for routine soil tests such as used in the soil testing laboratory. An ion speciation model MINTEQA2 (43), which uses mineralogy input, did not adequately predict soluble P in equilibrium with these soil clays. Mineralogy and claycontent is somewhat accounted for in the EPIC model by use of a P sorption parameter (PSP) obtained by considering classification of the soil and clay content in highly weathered soils and CaCO 3 in relatively unweathered soils (37). The need to consider clay content is evident from results obtained by isotherms and solubility of P from 91-day incubated soil samples with litter P and MCP applied and with no P applied. Hartsells Ap sample has 5.7% clay and Dewey has 20.2%. When P adsorption was put on a clay percent basis, adsorption and desorption was essentially the same per unit of clay. This should hold on a regional basis for soils with kaolinitic mineralogy in Ultisol and Alfisol soil orders. The complex nature of soil:solution P equilibria is well documented (42). Isotherm results in the AAES study show that a single adsorption mechanism does not explain or model adsorption and solution concentration from low to high loading. Several phenomena were found to occur: adsorption on multiple surfaces; higher soil pH at high litter loading rates, and slightly higher pH as MCP rates increased; and precipitation of Al-, Fe-, and Ca-phosphates. These isotherm adsorption curves can be represented by combining an initial Langmuir adsorbing surface followed by a linear representation of equilibrium solid:solution P concentrations (Figures 13-18). The linear nature of this equilibrium concentration distribution also is evident when solid:solution equilibria were evaluated on a clay suspension basis. In all cases, the linear phase is reached when solution concentrations are greater than 1,000 gg/L (67). Alabama does not have a discharge limit for municipal treatment facilities; however, one facility discharging into a stream flowing into a reservoir for a city water supply uses the 1,000 gg/L limit. At this concentration, MCP data (Figure 23) in the AAES study show that the total mineral P in the solid phase is 5,884 mg/kg of clay, of which 2,878 mg/kg is stable fixed and 128 10 Phosphorus Accumulation and Loss from Alabama Soils mg/kg is exchangeable or active P. In the same clays incubated with litter, 9,096 mg/kg of P are in the solid phase, 3,163 are fixed, and 980 are active. An approximation of exchangeable or active P can also be made by considering all MCP used in developing adsorption isotherms as being active. The tests were made after only 16 hours of shaking. Considering that the linear adsorption region is reached at 1,000 gg/L, a linear equation can be developed, and the active or exchangeable solid phase P at 1,000 gg/L can be calculated. This and other investigations have shown that soil test P (Mehlich-I extractable) gives an indication of dissolved P. At 1,000 gg/L dissolved P, the soil test level is 300 mg/kg. Kingery et al. (40) found that the average soil test P from 11 sites were near 300 mg/kg for surface soil layers of fescue pastures on the Hartsells series receiving litter for about 20 years. A 20-year simulation of soil test P levels using the EPIC model compared favorable to Kingery' s measured results. Many researchers have expressed concern about the level of P buildup in soils receiving high rates of P from poultry litter applications. Critical values for soil test P have been defined on the basis of crop requirements. The rationale has been that the limits proposed are far in excess of crop P requirements, and further applications enhance the risk for P losses in runoff. There is considerable evidence that on soils with moderate levels of soil test P, there are P losses in runoff from P applied at soil test recommended levels. In fact, both Scarseth (58) and Ensminger (21) reported substantial amounts of P losses from applied P fertilizer on Alabama soils. Losses in excess of biological algae limits are not the exception. These losses are management- and rainfallrelated and are not totally subject to P inputs from long-term P application. Isotherms show that resident soil test P levels have to exceed 300 mg/kg before equilibrium dissolved P levels greater than 1,000 gg/L are supported. This is a reasonable maximum P limit level based not on crop needs, but on both soil and solution levels obtained from controlled experimentation. Lower soil test critical values ranging from 75 to 200 mg/kg have been proposed (26, 67). Results from this study show that as the 300 mg/kg critical soil test value is approached or exceeded, best management plans (BMPs) should be evaluated to guide operators in selecting practices to limit further P buildup, minimize erosion and runoff losses, and maintain soil productivity. For fields that currently exceed the 300 mg/kg limit, remediation strategies that will accelerate P depletion should be made a part of management. Evaluating proposed practices or developing recommended practices through the use of models such as EPIC can help provide a comprehensive approach based on real onsite input of climate, soil, agronomic factors, management, and economic returns. Recommendations can be flexible, site specific, and environmentally sound. Alabama Agricultural Experiment Station 11 METHODS Soil Characterization In the fall of 1990, soil samples were collected from the Ap and Bt horizon of three major soil series found within the poultry-producing region of North Alabama. The soils sampled were in the Hartsells, Linker and Dewey series. According to available records, none of the sites had been treated with poultry litter. Approximately 150 kg of soil was collected from each horizon, air-dried and ground to pass a 0.25-inch sieve. Subsamples were obtained for incubation and adsorption isotherm determinations. In addition subsamples were used for soil physical, chemical, and mineralogical characterization. Physical: Particle-size analysis was performed by the sieve and pipette method (73). Percentages of the various fractions were calculated on an ovendry weight basis. The samples were treated for removal of organic matter and free Fe oxides. Chemical: Soil pH was determined on 1:1 soil-water suspensions. Cation exchange capacity (CEC ) was determined by N NH 4OAc saturation at pH 7.0 and KC1 extraction (73). Exchangeable Ca, magnesium (Mg), potassium (K), and sodium (Na) were extracted with N NH4OAc and determined by atomic absorption spectrophotometry. Exchangeable Al and hydrogen (H) were extracted with a N KC1 solution and determined by titration (73). Free Fe oxides, extracted by the Na-citrate-bicarbonate-dithionite method (34) were determined in solution extracts by X-ray emission spectroscopy (22). Mineralogical: Clay fractions (less than 2 im) were obtained for mineralogical analysis by ultrasonic bath dispersion in a pH 9.5 Na2CO3 suspension (38). Quantitative interpretations were based on procedures given by Karathanasis and Hajek (38). X-ray diffraction patterns were obtained for magnesium- and potassiumsaturated clay fractions placed on glass slides by a modified filter peel technique (18). A Siemens D5000 diffractometer equipped with a monochrometer and Cu tube was used. Magnesium-saturated clays were analyzed by Thermogravimetry and Differential Scanning Calorimetry. A Thermal Analysis "TA" 1200 controller unit was used for all TG and DSC analyses. Paired Fields Paired fescue pasture and paired cotton fields were sampled to one meter. One field in each pair had received litter for 18-20 years, and the other was under conventional N and P fertilization. The collected samples were airdried and ground to pass a 2-mm sieve. They were analyzed for Mehlich-1 (46) extractable P, K, Ca, Mg, and pH by the Auburn University Soil Testing Laboratory. Organic P was determined using an ignition procedure (52). A 10-gram subsample of soil from each treatment was ground to pass a 60-mesh 12 Phosphorus Accumulation and Loss from Alabama Soils sieve. This ground subsample was used in the determination of resin-extractable P (52) and total P. Total P was determined using a nitric-perchloric acid, wet ash digestion procedure (33). Incubation Subsamples of soil from each horizon were ground to pass a 2-mm sieve and treated with poultry litter at rates equivalent to 0, 25, and 50 tons per acre (0, 25, and 50 g of litter per kilogram of soil). A rate of 50 tons per acre would represent the approximate amount of litter what would be applied to supply the N needs of a corn crop for a period of 10 years. Litter used in the study was obtained from North Alabama and contained 32.14% carbon (C), 3.5% N, and 1.37% P. The litter was air-dried and ground to pass a 1-mm sieve prior to application. A second series of treatments received P as inorganic, reagent-grade monocalcium phosphate, MCP, [Ca(H 2PO4) 2 H 20] at rates equivalent to the amount of P applied in the poultry litter treatments (0, 342, and 684 mg/kg). Treatments consisted of 200 g of air-dried soil that was thoroughly mixed (by hand) with the respective rate of litter or MCP. Treated soil was placed into treatment vessels which consisted of one-pint, wide-mouth mason jars. Deionized water was added to bring the soil to 85% of field capacity. Treatments were incubated at 30°C for 91 days. In order to measure the amount of CO 2 produced during the incubation, a vial containing 10 mL of 2 N NaOH was placed into each treatment vessel. Two treatment vessels without soil were included as checks. Residual NaOH in the vials was titrated (in the presence of BaCl 2) periodically throughout the incubation study using 1 N HCl. Data from the titrations were used to determine the amount of CO 2 that was produced during the incubation study. Treatment vessels were aerated 10 minutes each time the NaOH vials were removed for titration and once a week when the time between titration was extended to 14 days. At the end of 91 days, the samples were air-dried and ground to pass a 2mm sieve. Organic P was determined using an ignition procedure (52) and pH using a 1:1 soil to water (V:V) ratio (33). A 10-gram subsample of soil from each treatment was ground to pass a 60-mesh sieve. This ground subsample was used in the determination of resin-extractable P (52) and total P. Total P was determined in soil from selected treatments using a nitric-perchloric, wet ash digestion procedure (33). Phosphorus desorption in each treatment was evaluated using a batch equilibrium method where soil was extracted with 0.01 M CaCl 2. Subsamples of soil (5 g) from each treatment were weighed into 50 mL centrifuge tubes and equilibrated with 20 mL of 0.01 M CaCl 2. One drop of toluene was added to reduce microbial activity. Each tube was weighed before and after the addition of the CaC12 solution. The tubes were placed in a 30°C shaking water bath for 24 hours at approximately 25 rpm; tubes were then centrifuged, and Alabama Agricultural Experiment Station 13 10 to 15 mL of solution was removed. The tubes were reweighed and enough CaCl 2 solution was added to bring the solution back to 20 mL. A vortex mixer was used to resuspend the soil, and the samples were shaken for an additional 24 hours. This process was repeated for six days. Leachates were analyzed using the Murphy and Riley (49) procedure. The amount of P released was calculated by knowing the concentration of P in the leachate, the amount of leachate removed each day, and the amount of leachate that was carried over to the next sampling. Adsorption Isotherms: Phosphorus adsorption isotherms were constructed for soil from each repetition of each treatment. Subsamples (1 g) of soil from each treatment were weighed into 50-mL centrifuge tubes and equilibrated with 30 mL of 0.01 M CaC12 containing various levels of P. Phosphorus was supplied as KH 2PO 4 at rates of 0, 5, 10, 25, 75, 150, 250, and 500 mg/ kg soil. One drop of toluene was added to each tube to reduce microbial activity. Treated samples were shaken for 18 hours in a reciprocating, shaking water bath (30°C), centrifuged, and filtered through Gelman SUPOR-200 membrane filters (0.2 gm pores). Phosphorus in the filtrates was determined by the Murphy and Riley (49) procedure. Adsorbed P was calculated as the difference between the amount of P added and the resulting equilibrium solution concentration. The Langmuir, cumulative, Fruendlich, and linear equations were applied to isotherm data. Adsorption capacities were needed as input into a soil P simulation model. An additional subset was equilibrated in 0.01 M CaC12 for 30 days, with no P added to the background solution. Following equilibration the suspensions were centrifuged and filtered to obtain clear solutions for P analysis. This set was used to determine the relation of P sorption to clay content and the solution P concentrations were used to evaluate an ion speciation model. MINTEQA2/PRODEFA2 Ion Speciation The concentration of P in solution was evaluated relative to soil mineral and soluble ion composition by submitting equilibrium solution P concentrations to the MINTEQA2 ion speciation model (5). Equilibrium P concentrations were obtained by suspending enough soil in solution to achieve clay suspensions of 500 and 1,000 mg/L. The Hartsells and Dewey soils were obtained from an incubated set with the equivalent of 25 and 50 tons of litter added and with equivalent amounts of MCP added and incubated for 91 days at 30°C. Check samples were also included (no litter or MCP added). The suspensions were allowed to equilibrate for 30 days with occasional shaking. A microbial inhibitor was added. Following equilibration the suspensions were centrifuged, filtered, and analyzed for soluble P and other major soluble ions. The data required to predict the equilibrium P in these suspensions include: (1) dissolved concentrations of P and other relevant ions, such as Ca, Mg, Fe, Al, pH, SO 4, and anions, to nearly maintain neutrality and to simulate the ionic strength of the solution; (2) dissolved CO 2; and (3) minerals, espe- 14 Phosphorus Accumulation and Loss from Alabama Soils cially kaolinite, gibbsite, quartz, and iron oxyhydroxides. Initial soil mineral and total P composition was available from analysis done in other parts of this study. MINTEQA2/PRODEFA2 was installed on a 486/33 PC, and the equilibrium constant data furnished with the program was replaced by a database sent on request by Lindsay (personal communication). Both the Langmuir and the Kd adsorption equations were included in the simulation. Erosion Productivity Impact Calculator Simulation In this study, a paired fescue pasture (Hartsells fine sandy loam) and a paired cotton field (Dewey silt loam) were sampled to one meter. The pasture sites were in DeKalb County, Ala., and the cotton fields were in Cherokee County, Ala. One field in each pair had received litter for 18-20 years, and the other was under conventional N and P fertilization. Isotherms developed from similar soils were used to determine P sorption ratios of surface and subsoils which were input into the model. Resident mineral (total and resin extractable), Mehlich-1 extractable P, and organic P were determined to sampled depths for the four fields. Soil input variables were obtained from available soil data sampled in support of other research and the soil survey program. Management was obtained from farm operator records and from practices recommended by professional agricultural workers in North Alabama. Average fresh broiler litter composition was obtained from an Extension Service survey of litter composition needed to make fertilizer recommendations for litter used on pasture or other crops. Before the model was used, it was evaluated by comparing predicted forage and cotton yields to average yields expected in North Alabama on these or similar soils. Predicted yields were well within the expected range when evaluated for a 10-year period. EPIC was used to simulate P uptake, leaching, buildup, and runoff losses under fescue and cotton receiving litter and under conventional fertilization for 20 years. RESULTS AND DISCUSSION Soil Characteristics The surface and subsoil samples collected from three pedons were characterized to confirm placement into Soil Taxonomy (73). In addition, selected data were obtained that were useful in interpreting differences in P loading and retention. The data showed that all three soils should be classified as Ultisols, which are considered to be highly weathered (Table 1). Particle-size distribution properties are given in Table 2, and chemical properties are in Table 3. Since clay is the active fraction, particle size data indicate that the clay content of the Ap horizon in Dewey and Linker soils (about 20%) will adsorb more P per unit weight of soil than Hartsells (5.7% clay) (37). This is also evident for subsoil Bt horizons. Other particle-size differences are reflected in higher silt Alabama Agricultural Experiment Station 15 TABLE 1. CLASSIFICATION OF NORTH ALABAMA SOILS SAMPLED FOR P ADSORPTION AS AFFECTED BY THE APPLICATION OF POULTRY LITTER Soil Dewey ............................. ......... Linker ......................... ............ Hartsells .................................... Classification Clayey, kaolinitic, thermic, Typic Paleudults Fine-loamy, siliceous, thermic, Typic Hapludults Fine-loamy, siliceous, thermic, Typic Hapludults in the Dewey and corresponding lower sand. Chemical properties, OM, CEC, and pH are also similar for Dewey and Linker, but the higher gibbsite and Fe 20 3 of the Linker clay fraction (Table 4) should enhance P adsorption and fixation. Although all soils were classified in the Ultisol soil order, differences in clay content and mineralogy should be reflected by substantially higher P fixation and sorption in Dewey and Linker. This is supported by total P content and organic P (Table 3) but not by soil test P. The soil test P was higher in Hartsells and Dewey because they were collected in cultivated fields. The Linker was collected in mixed pasture-woodland that had received no fertilizer P. Phosphorus Adsorption Although P was the primary interest of this study, CO 2 production was measured in order to evaluate the relative degree of litter decomposition that occurred during the 91-day incubation study (Figures 1-3). As expected, C released as CO 2 increased with litter rate (Figures 1-3 and Table 5). A power function was used to describe the cumulative release of C with time from the litter treated soils (Figures 1-3). The amount of added litter-C released ranged from 47-64% of the added C in the Ap horizon samples and from 37-47% of the added C in the Bt horizon samples (Table 5). These levels of decomposition are consistent with other reports evaluating litter decomposition in soils. Sims (69) reported that from 30-60% of the organic N in poultry manure was TABLE 2. PARTICLE SIZE DISTRIBUTION OF THE THREE SOILS USED TO EVALUATE P Soil & Horizon ADSORPTION AS AFFECTED BY THE APPLICATION OF LITTER Particle size distribution Sand Silt Clay pct. pct. pct. 2-1 pct. Sand size distribution (mm) 1-.5 .5-.25 .25-.1 .1-.05 pct. pct. pct. pct. Hartsells-Ap' ....... Hartsells-Bt' ........ Linker-Ap 2 ........... Linker-Bt 3 ............ Dewey-Ap 4 ........... Dewey-Bt 5 ............ 'Sandy loam; 2 61.0 56.8 48.8 40.1 29.9 19.0 3 33.28 31.20 31.44 27.22 49.86 46.16 Clay loam; 4 5.72 12.00 19.76 32.68 20.24 34.84 Silt loam; 0.66 1.06 5.57 3.50 6.10 3.72 5 3.78 3.89 2.27 2.25 14.92 11.17 32.84 31.63 17.32 13.75 27.46 25.00 51.56 52.30 42.89 39.25 32.54 35.11 11.17 11.13 31.96 41.25 18.98 25.00 Loam; Silty Clay loam. 16 TABLE Phosphorus Accumulation and Loss from Alabama Soils 3. CHEMICAL PROPERTIES OF THREE SOIL SERIES USED TO EVALUATE P ADSORPTION CHARACTERISTICS AS AFFECTED BY POULTRY LITTER Soil & horizon Organic CEC matter pct. cmol/kg pH Total P mg/kg Organic Resin P ext. P mg/kg mg/kg P lb./a. Mehlich I extractable Mg Ca K lb./a. lb./a. lb./a. Hartsells-Ap ... 1.5 Hartsells-Bt..... 0.3 Linker-Ap ....... 1.8 Linker-Bt ........ 0.7 Dewey-Ap ....... 2.1 Dewey-Bt ........ 0.7 4.86 4.68 6.34 6.61 6.73 7.35 6.6 5.1 5.8 4.9 5.3 5.2 311 223 479 497 656 488 86 40 149 113 138 77 27.0 1.5 0.9 1.6 5.5 1.0 96 3 5 2 53 3 70 44 74 39 413 250 52 11 112 70 127 157 1,470 390 1,290 780 830 1,240 mineralized within 90 to 150 days after incorporation. Castellanos and Pratt (13) reported that approximately 45% of the total C in chicken manure was decomposed during a four-week incubation study. Gale and Gilmour (25) reported that C release from litter-treated soil consisted of three phases (rapid, intermediate, slow). In their study, decomposition shifted from a rapid to an intermediate and ultimately to a slow phase. The shift in decomposition phase occurred when approximately 15% and 30% of the added C was mineralized, respectively. Under field conditions, Flynn (24) evaluated the decomposition of peanut-hull- and wood-shaving-based poultry litter. Both sources of litter were applied at a rate of four tons per acre. At 112 days after application, he reported an average release of 52% and 49% of the litter-C on a Wynnville and a Norfolk soil, respectively. Monocalcium phosphate resulted in higher levels of resin-extractable P, as compared to the litter treatments, since part of the litter P is organic (Table 6). The litter treatments resulted in large increases in organic P in all soils tested. For some unexplained reason, the low-litter treatment to the Ap horizon of the Linker soil resulted in very little change in organic P. If the Ap horizon of the Linker soil is excluded, an average of 37% and 33% of the TABLE 4. MINERALOGICAL CHARACTERISTICS OF SOILS SAMPLED FOR P ADSORPTION AS AFFECTED BY THE APPLICATION OF POULTRY LITTER Soil & horizion Surface Fe203 equivalent area, clay pct. 2.3 3.5 0.5 1.0 2.6 4.6 2 Surface area, soil 2 HIV pct. 27 32 63 53 58 55 Kaolinite pct. 48 45 34 31 39 44 Mineralogy Mica Gibbsite pct. 5 4 --- Quartz pct. 6 2 3 10 7 7 Dewey-Ap ........... Dewey-Bt ............ Hartsells-Ap ....... Hartsells-Bt......... Linker-Ap ........... Linker-Bt ............ 'Trace amount. m /g clay m /g soil 133 27 147 51 8 144 137 16 36 183 54 166 tr tr pct. tr' tr 2 5 9 14 Alabama Agricultural Experiment Station 17 Alabama Agricultural Experiment Station 17 CO2 1 mlO/kg 8001- Hartsells Ap Low litter: -1 .20E-05 High litter: -7.47E-05 + 74.95X.5 00 , r2 = 0.991 +,41 .86X.546, r2 = 0.996 o io 20 30 40 50 60 70 80 90 Days of incubation AControl * Low MCP Low Litter p High Litter 0 High MCP CO 1mmol/kg 2 800 600 400 200 0 0 - Hartsells Bt Low litter: -1.29E-05 + 33.09X"53 5 , r2 = 0.993 High litter: -3.14E-06 +,43.32X. 603, r2 = 0.995 10 20 30 40 50 60 70 80 90 Days of incubationI Figure 1. Cumulative production of CO 2 from the Ap and Bt horizon of a Hartsells soil as affected by the application of poultry litter and inorganic fertilizer. Low litter = 25 tons! acre, High litter = 50 tons/acre, Low MCP = monocalcium phosphate applied to equal P in Low litter treatment (342 mg P/kg), High MCP = monocalcium phosphate applied to equal P in High litter treatment (684 mg P/kg). 18 Phosphorus Accumulation and Loss from Alabama Soils 18 Phosphorus Accumulation and Loss from Alabama Soils CO 1mmol/kg 2 800 600 400 200 Linker Ap Low litter: -2.48E-05 High litter: -1.71 E-06 + 41 .31 X 63 8, r2 = 0.992 +,46.88X. 523, r2 = 0.998 ,A o io 20 30 Days of incubation 40 50 60 70 80 90 A Control " Low Litter o High Litter Low MCP Q High MCP CO mmol/kg 21 800 hLow IHigh LinkerBt litter: -1.94E-05 + 35.21 X522 , r2 = 0.998 litter: -2.45E-05 + 55.22X"53 1 , r2 = 0.987 0 10 20 30 Days of incubation] 40 50 60 70 80 90 Figure 2. Cumulative production of CO 2 from the Ap and Bt horizon of a Linker soil as affected by the application of poultry litter and inorganic fertilizer. Low litter = 25 tons! acre, High litter = 50 tons/acre, Low MCP = monocalcium phosphate applied to equal P in Low litter treatment (342 mg P/kg), High MCP = monocalcium phosphate applied to equal P in High litter treatment (684 mg P/kg). Alabama Agricultural Experiment Station 19 Alabama Agricultural Experiment Station 19 CO2 1mmol/kg Dewey Ap 800 (600 Low litter: -2.42E-05 + 43.57X 5 2 2, r2 = 0.996 High litter: -2.76E-05 + 65.45X"53 2, r2 = 0.989 00 0 0 400 400 0 0 10 20 30 40 50 60 Days of incubation " 70 80 90 AControl Low Litter o High Litter Low MCP CO mmol/kg 21 800 - Q High MCP Dewey Bt Low litter: -1.33E-05 + 32.0X"53 2, r2 = 0.997 High litter: -7.17E-05 + 60.29X"4 94 , r2 = 0.989 [ 0 10 20 30 40 50 60 70 80 90 Days of incubation Figure 3. Cumulative production of CO2 from the Ap and Bt horizon of a Dewey soil as affected by the application of poultry litter and inorganic fertilizer. Low litter = 25 tons! acre, High litter = 50 tons/acre, Low MCP = monocalcium phosphate applied to equal P in Low litter treatment (342 mg P/kg), High MCP = monocalcium phosphate applied to equal P in High litter treatment (684 mg P/kg). 20 Phosphorus Accumulation and Loss from Alabama Soils TABLE 5. AMOUNT OF CARBON RELEASED AS CO 2 FROM SOILS DURING 91-DAY INCUBATION AS AFFECTED BY LITTER AND MONOCALCIUM PHOSPATE Soil & horizon Source P rate Added C Carbon release Released C Pct. added C mmol/kg 31 457 650 35 36 14 352 614 6 22 43 472 667 66 68 -352 538 19 34 433 650 64 74 336 511 19 25 pct. -64 46 -- Hartsells-Ap ............... Check Hartsells-Ap ............... Litter Hartsells-Ap ............... Litter Hartsells-Ap .................. MCP Hartsells-Ap .................. MCP Hartsells-Bt................. Check Hartsells-Bt ................. Litter Hartsells-B t .................. Litter Hartsells-Bt ............ MCP Hartsells-Bt.................. MCP Linker-Ap .................. Check Linker-Ap ................... Litter Linker-Ap ............. Litter Linker-Ap ................... MCP Linker-Ap ................... MCP Linker-Bt ................... Check Linker-Bt .................... Litter Linker-Bt .................... Litter Linker-Bt ............ MCP Linker-Bt .................... MCP Dewey-Ap .................. Check Dewey-Ap ................... Litter Dewey-Ap ................... Litter Dewey-Ap ................... MCP Dewey-Ap ................ MCP Dewey-Bt ................... Check Dewey-Bt .................... Litter Dewey-Bt .................... Litter Dewey-Bt ....................... MCP Dewey-Bt ....................... MCP mg/kg 0 342 684 342 684 0 342 684 342 684 0 342 684 342 684 0 342 684 342 684 0 342 684 342 684 0 342 684 342 684 mmol/kg 0 669 1,338 0 0 0 669 1,338 0 0 0 669 1,338 0 0 16 669 1,338 0 0 54 669 1,338 0 0 22 669 1,338 0 0 --51 45 -- 64 47 -50 39 -- 57 45 -- 47 37 --- litter P from the low and high litter treatments, respectively, could be accounted for as organic P in the Ap horizon samples. In the Bt horizon samples averages of 40% and 39% of the litter P could be accounted for as organic P in the low- and high-litter treatments. Litter also resulted in an increase in soil pH, as compared to the check treatment. Monocalcium phosphate also resulted in the greatest release of added P during a continuous desorption study that was conducted over a period of six days (Figures 4-6). However, when expressed as a fraction of the added P, relatively small amounts of P were released from the treated soils during the batch desorption study (Table 6). The greatest release on a percentage basis was from the low-MCP treatment (342 mg/kg) on the Hartsells soil, where 20% of the added P was released. Data from the desorption study suggest that the leaching potential for P from the Ap horizon would be greatest for the Hartsells soil and lowest for the Alabama Agricultural Experiment Station 21 Dewey soil. Leaching potential for P applied as both sources would be very low in the Bt horizon of all three soils (Table 6, Figures 4-6). Solution P concentrations (PO4-P) maintained during the batch desorption study are shown in Figures 7-9. Consistently, the MCP treatments maintained the highest solution P concentrations. In the Ap horizon samples, the high-MCP treatment had solution P concentrations that were always higher as compared to the high-litter treatments. Higher solution concentrations in the MCP treatments were due, in part, to a lower soil pH in the MCP versus the litter treatments (Table 6). For the MCP and high-litter treatments, the highest concentrations were always observed in the first extract followed with a decrease in concentration with each successive extract. In the Ap horizon TABLE 6. SOIL PH, RESIN EXTRACTABLE P, ORGANIC P, 1 AND THE AMOUNT OF P RELEASED DURING A SIX-DAY DESORPTION STUDY Soil Source P rate Soil pH Resin P Organic P Desorbed P 2 Total Pct. added P Hartsells-Ap ....... Check Hartsells-Ap ....... Litter Hartsells-Ap ....... Litter Hartsells-Ap ....... MCP Hartsells-Ap ....... MCP Hartsells-Bt......... Check Hartsells-Bt......... Litter Hartsells-Bt......... Litter Hartsells-Bt......... MCP Hartsells-Bt......... MCP Linker-Ap ........... Check Linker-Ap ........... Litter Linker-Ap ........... Litter Linker-Ap ........... MCP Linker-Ap ........... MCP Linker-Bt ............ Check Linker-Bt ............ Litter Linker-Bt............Litter Linker-Bt ............ MCP Linker-Bt ............ MCP Dewey-Ap ........... Check Dewey-Ap ........... Litter Dewey-Ap ........... Litter Dewey-Ap ........... MCP Dewey-Ap ........... MCP Dewey-Bt ............ Check Dewey-Bt ............ Litter Dewey-Bt ............ Litter Dewey-Bt ............ MCP Dewey-Bt ............ MCP pct. mg/kg mg/kg mg/kg 1.04 0 6.2 76 27.00 2.802 342 7.3 110.00 189 11.00 166.00 684 7.5 3.20 303 23.00 69.00 20.00 342 123.00 6.0 5.8 92 132.00 684 216.00 19.20 0 5.1 40 0.24 1.50 342 0.91 21.00 173 0.20 6.8 7.3 386 5.30 108.00 0.70 684 1.80 342 4.9 37 6.40 35.00 56.00 8.20 684 5.1 54 110.00 149 0.36 0 5.2 0.89 0.30 161 1.50 5.5 342 16.00 68.00 684 325 3.90 0.50 7.3 147 19.00 5.50 5.3 57.00 342 192 65.00 118.00 9.50 5.4 684 0.48 113 0.21 0 5.0 0.34 4.5 2.40 254 0.04 342 281 1.40 0.20 684 6.9 31.00 8.80 4.9 92 1.30 0.30 342 138 9.30 5.5 1.30 684 66.00 138 0.33 5.0 5.50 0 279 1.30 29.00 0.30 342 6.6 684 6.9 46.00 361 2.50 0.30 52.00 177 12.00 3.40 342 5.0 196 47.00 6.80 684 5.1 92.00 0.74 0.14 5.2 77 0 2.10 215 0.19 0.01 342 5.3 0.94 7.9 31.00 394 0.10 684 1.44 0.40 342 5.1 84 8.30 2.40 684 III n 60.00 70 16.00 r-v 5.1 'Measurements were taken on soils that had been incubated for 91 days after being treated with P as either poultry litter or monocalcium phosphate (MCP). 2 Percent added P = (desorbed P - desorbed P from check)/added P)x100 22 Phosphorus Accumulation and Loss from Alabama Soils P release, mg/kg 140 140 Hartsells Ap 120 100 80 6040 20 1 2 3 4 5 6 7 Time, days A Control " Low Litter a High Litter LowMCP O High MCP P release, mg/kg 60 50 40 30 20 10 1 2 3 4 5 6 7 Time, days Figure 4. Cumulative P release during a batch desorption study using 0.01 M CaCl 2. Soil samples were collected from the Ap and Bt horizon of a Hartsells soil and treated with poultry litter and inorganic fertilizer. Solution phase was collected daily and soil resuspended in 0.01 M CaCI2 . Low litter = 25 tons/acre, High litter = 50 tons/acre, Low MCP = monocalcium phosphate applied to equal P in Low litter treatment (342 mg P/kg), High MCP = monocalcium phosphate applied to equal P in High litter treatment (684 mg P/kg). Alabama Agricultural Experiment Station 23 P release, mg/kg 1 -2 3 4 Time, days 5 6 7 A Control " Low Litter o High Litter Low MCP 0 High MCP P release, mg/kg 10 Linker Bt 2 1 2 3 4 5 6 7 Time, days Figure 5. Cumulative P release during a batch desorption study using 0.01 M CaCl 2. Soil samples were collected from the Ap and Bt horizon of a Linker soil and treated with poultry litter and inorganic fertilizer. Solution phase was collected daily and soil resuspended in 0.01 M CaCl 2. Low litter = 25 tons/acre, High litter = 50 tons/acre, Low MCP = monocalcium phosphate applied to equal P in Low litter treatment (342 mg P/kg), High MCP = monocalcium phosphate applied to equal P in High litter treatment (684 mg P/kg). 24 Phosphorus Accumulation and Loss from Alabama Soils P release, mg/kg 50- Dewey Ap 40 30 20 10 0 3 4 Time, days 1 .2 5 6 7 A Control " Low Litter o High Litter Low MCP O High MCP P release, mg/kg 20 rDewey Bt 15 10 0 1 2 3 4 5 6 7 Time, days Figure 6. Cumulative P release during a batch desorption study using 0.01 M CaCl 2. Soil samples were collected from the Ap and Bt horizon of a Dewey soil and treated with poultry litter and inorganic fertilizer. Solution phase was collected daily and soil resuspended in 0.01 M CaCl 2. Low litter = 25 tons/acre, High litter = 50 tons/acre, Low MCP = monocalcium phosphate applied to equal P in Low litter treatment (342 mg P/kg), High MCP = monocalcium phosphate applied to equal P in High litter treatment (684 mg P/kg). Alabama Agricultural Experiment Station 25 SoAlabama Agricultural Experiment Station Harisells Ap 15 10 1 .2 3 4 5 6 7 Time, days A Control " Low Litter o High Lifter Low MCP 0 High MCP Solution P, mg/L 6 1. .11. . 1 2 3 4 Time, days 5 6 7 Figure 7. Phosphorus concentration in the equilibrium solution during a batch desorption study using 0.01 M CaCl 2. Soil samples were collected from the Ap and Bt horizon of a Hartsells soil and treated with poultry litter and inorganic fertilizer. Solution phase was collected daily and soil resuspended in 0.01 M CaCl 2. Low litter = 25 tons/acre, High litter = 50 tons/acre, Low MCP = monocalcium phosphate applied to equal P in Low litter treatment (342 mg P/kg), High MCP = monocalcium phosphate applied to equal P in High litter treatment (684 mg P/kg). 26 Phosphorus Accumulation and Loss from Alabama Soils 26 Phosphorus Accumulatio Linker Ap 1 2 4 Time, days " Low Litter o High Litter 0 A Control Solution P,mg/I 0.7 Low MCP High MCP Linker Bt 0.5 0.4 0.3 0.2- 0.1 r Jviuiviirl rlya 1 2 3 4 Time, days 5 6 7 Figure 8. Phosphorus concentration in the equilibrium solution during a batch desorption study using 0.01 Mi CaCl2 . Soil samples were collected from the Ap and Bt horizon of a Linker soil and treated with poultry litter and inorganic fertilizer. Solution phase was collected daily and soil resuspended in 0.01 M CaCl 2. Low litter = 25 tons/acre, High litter = 50 tons/acre, Low MCP = monocalcium phosphate applied to equal P in Low litter treatment (342 mg P/kg), High MCP = monocalcium phosphate applied to equal P in High litter treatment (684 mg P/kg). Alabama Agricultural Experiment Station 27 Solution P, mg/L 5 Dewev Ao 1 2 3 4 5 6 Time, days A Control " Low Litter o High Litter Low MCP 0 High MCP Solution P, mg/L 1.4A.,- 1 2 3 4 Time, days 5 6 7 Figure 9. Phosphorus concentration in the equilibrium solution during a batch desorption study using 0.01 M CaCl2. Soil samples were collected from the Ap and Bt horizon of a Dewey soil and treated with poultry litter and inorganic fertilizer. Solution phase was collected daily and soil resuspended in 0.01 M CaCl2. Low litter = 25 tons/acre, High litter = 50 tons/acre, Low MCP = monocalcium phosphate applied to equal P in Low litter treatment (342 mg P/kg), High MCP = monocalcium phosphate applied to equal Pin High litter treatment (684 mg P/kg). 28 Phosphorus Accumulation and Loss from Alabama Soils samples, the Hartsells soil had the highest while the Dewey soil had the lowest concentrations. Surprisingly, the first extract (day 1) from the Ap horizon of the Hartsells soil had a solution P concentration of approximately 20 mg/ L. Even the Dewey soil had a solution P concentration of 4.1 mg/L in the first extract from the high-MCP treatment. Even after the final extract (day 7), solution P concentrations in the MCP and litter treatments would be within or above the minimum concentration that is needed for algae growth (27). These solution values which would be representative of a soil receiving litter for several years are much higher than soluble P concentrations observed in runoff collected from fields treated with annual applications of poultry litter (29, 31). A higher solution P concentration in the MCP treatments is in agreement with the results of Nichols et al. (50). They looked at runoff P concentrations (rainfall simulator) from fescue pasture at seven days after being treated with two tons of poultry litter per acre and an equivalent rate of fertilizer P. Concentrations of PO4-P from the litter treatment averaged 10.5 mg/L, while runoff from the fertilizer treatment averaged 26.1 mg/L. Adsorption Isotherms Five point isotherms were developed for the three soils following incubation for 91 days. An isotherm in which P concentration is plotted vs. concentration in the solid phase was developed for all treatments of all soils. Representative isotherms for surface and subsoil horizons of each soil are shown in Figures 10-12. Several P isotherm models have been proposed for predicting the P adsorption capacity of soils. Most P adsorption data are compared to a fit to the Langmuir type isotherm (9) which was initially derived for the adsorption of gases and vapors on surfaces. Two constants are obtained from a best fit of the adsorption data to a linear form of the Langmuir equation. yl= kbX -1+kX The constant b is related to energy of adsorption, the constant k is the adsorption capacity, yl is the solid phase P concentration, and X is the solution P concentration. The fit generally is good at lower adsorption levels and under-predicts at very high solution-sorbed concentrations. Figure 13 shows an example in which the Langmuir fit is good to approximately 15 mg/L in solution and about 400 mg/kg on the solid phase. However, the distribution becomes essentially linear from 200 to the last point at 600 mg/kg adsorbed P. Other models can be used and in some cases can be useful in predicting adsorption capacities. A cumulative model was considered because the equation has an adsorption maximum and a nonlinear distribution at lower adsorption amounts. In the cumulative equation: Alabama Agricultural Experiment Station 29. Alabama Agricultural Experiment Station Surface soil Adsorbed P, ppm 1,000 29 Subsurface soil Adsorbed P, ppm [O - 800 600 400 200 20 0 1,000 800 600 400 200 0 -200 0 1,000 800 600 50T equivalent MCP added Control, no litter added 5 10 15 20 25 0 1000 r 800 5 Control, no litter added 10 15 20 25 6001 400 50T litter added 200 v 50T litter added -LVUU I I I I I 5 10 15 20 25 ,000 0v 5 10 15 20 25 8001 600 50T equivalent MCP added 400 200 0, -200- 400 200 VI I 0 5 10 15 20 25 0 Equilibrium solution P, ppm 5 10 15 20 25 Equilibrium solution P, ppm Figure. 10. Phosphorus adsorption isotherms of Dewey surface and subsurface soil following addition of poultry litter and monocalcium phosphate and incubation for 91 days. 30 Phosphorus Accumulation and Loss from Alabama Soils 30 Phosphorus Accumulation and Loss from Alabama Soils Surface soil Subsurface soil Adsorbed P, ppm 1,000 800 600 400 Adsorbed P,ppm 1,000 800 600 400 200 0 -2001 I I-2001 Control, no litter added 200 0L Control, no litter added 0 1,000 5 10 15 20 25 0 , 5 10 15 ,I 20 25 I1,000800 8001 600 400 200 50T litter added 600 400 200 501 litter added 0 -200 0 5 10 15 I 0 20 I 25 i -200 0 5 10 15 20 25 1,000 800 600 5OT equivalent MCP added 1,000 800 600 50T equivalent MCP added 400 200 0 -200 400 200 0 5 10 15 20 25 Equilibrium solution P,ppm I I i 0 -200 0 5 10 15 20 25 Equilibrium solution P,ppm Figure. 11. Phosphorus adsorption isotherms of Linker surface and subsurface soil following addition of poultry litter and monocalcium phosphate and incubation for 91 days. Alabama Agricultural Experiment Station 31 Surface soil Adsorbed P, ppm 1,000 800 Control, no litter added Subsurface soil Adsorbed P, ppm 1,000 800 600 400 200 0 -200 - 600 400 200 Control, no litter added 0 0 I 5 I I i 20 25 -200 I I 10 15 0 1,000 800 600 400 5 10 15 20 25 1,000 800 600 400 200 50T litter added 200 50T litter added 0 -200 I 0 I -200 I I I 0 1,000 800 600 5 10 15 20 25 0 1,000 - 5 10 15 20 25 · 50T equivalent MCP added 800 600 50T equivalent MCP added 400 200 400200 - 0 -200 0 i i I I I 0 15 20 25 5 10 Equilibrium solution P, ppm -200 0 I I 5 10 15 20 25 Equilibrium solution P, ppm Figure. 12. Phosphorus adsorption isotherms of Hartsells surface and subsurface soil following addition of poultry litter and monocalcium phosphate and incubation for 91 days. 32 Phosphorus Accumulation and Loss from Alabama Soils X-al y2=O.5a0[1+erf(X-al a2\2 y2 is the amount adsorbed on soil, aO is the adsorption capacity, al is the cumulative midpoint, and a2 is the width from no adsorption to capacity. Figure 14 shows that a cumulative curve can be drawn that passes near most points. It is obvious that a single linear equation would not improve the fit of a line that would include all data points. However, a linear equation gives a good fit to adsorption data points at high concentrations. In the linear expression: y 3 =a+KdX y3 is the amount of P in the solid phase, a is the intercept, and K1 (equilibrium distribution coefficient) is the slope. If only the solution concentration points greater than 5 mg/L are used, a reasonable fit is obtained (Figure 15) in which the intercept is the amount of P in the solid phase that is "fixed" and/or that is tightly bound. Soil P, mg/kg YHartsells A 0 0 - Langmuir S Experimental 0 5 10 15 20 25 Solution P,mg/L Figure 13. An example of the Langmuir equation applied to equilibrium phosphorus adsorption data for a Hartsells surface soil. Alabama Agricultural Experiment Station 33 To include all data points, the Langmuir and cumulative equations can be summed to create the following equation: aOalX X-a3 yl= +(0.5a2[l1+erf(X-a3 ) 2a4 1+alX in which aO and al are Langmuir constants, a2 corresponds to the cumulative maximum, and a3 and a4 set width and cumulative curve midpoint. Figure 16 shows the resulting fit using a spreadsheet to generate the curve (Hartsells Ap) and Peakfit software (Figure 17; Linker Ap). A summation of the Langmuir and linear equations gives a comparable data fit (Figure 18). Using a model developed by combining equations to simultaneously model multiple P sorption reactions allows interpretation of the entire isotherm in terms of different adsorbing surfaces, precipitation, and fixation (47); or in terms of initially tightly bound P during initial loading and linear distribution between solid and solution phases after initial tightly bond sites are filled. Soil P, mg/kg Hartsells 800 - - - Cumulative 600 Experimental 600 400 .*' 200 * ' 0 0 5 10 15 20 25 Solution P, mg/L Figure 14. An example of the cumulative equation applied to equilibrium phosphorus adsorption data for a Hartsells surface soil. 34 Phosphorus Accumulation and Loss from Alabama Soils In Figures 19-21 (showing P adsorption isotherms following incubation), these isotherms show that resident mineral P following incubation was added to that retained from solution to give the total solid-phase mineral P. The relatively linear soil:solution concentration distribution is evident for Ap horizons of all soils. Solution concentrations of Ap horizons also indicate that after 91 days of incubation, more P was in solution when MCP was applied than from the addition of litter. In both P treatments on surface soil, the Langmuir capacity was exceeded, and adsorption distribution between solid and solution phases was essentially a linear function at high-solution concentrations. Phosphorus adsorption in soils is not a simple process, and it is not surprising that no one equation can be successfully used to model a complete adsorption isotherm from none adsorbed or in solution, to complete capacity loading. Excellent Langmuir representation is possible at initial loading to about 5 mg/L in solution or less. These points were used from all check samples represented by examples in Figures 10-12..The constants in Table 7 were calculated from this set of four repetitions for each surface and subsoil. The R2 indicated the "goodness of fit." The Langmuir capacities from this set of control samples were com- Soil P, mg/kg Hartsells 800 600 - 400 200 - Linear " " Experimental 0 5 10 15 20 25 Solution P, mg/L Figure 15. An example of the linear equation applied to equilibrium P adsorption data for a Hartsells surface soil. Alabama Agricultural Experiment Station 35 pared to soil clay content of the sample. The Langmuir P sorption capacity relationship to clay can be considered good (Figure 22), despite mineralogy differences between soils, and lends support to the use of clay content in the EPIC model (37). Soluble Phosphorus Determinations Following Incubation The samples in which the equivalent of 0, 25, and 50 tons of litter were applied and an equivalent sample set that were treated with monocalcium phosphate and equilibrated for 91 days were selected for solubility determinations. Only the Ap horizon samples of the Hartsells and Dewey soils were selected since litter and fertilizer P are applied to the surface layers only and these are the most extensive soils being used for pasture and cropland in North Alabama. Two soil:solution ratios were used for each soil. The two suspensions of both soils and all treatments contained approximately 500 and 1,000 mg of clay per L. This is within the range of suspended solids in runoff studies from small watersheds on the Tennessee Valley Substation (29) and from a cotton field in Colbert County, Ala. (72). The suspensions were allowed to equilibrate for 30 days with daily Solution P, mg/kg 800 - Hartsells Langmuir -600 - Sum of 1+2 line Cumulative 600- Experimental 400 - 200 - " 0 0 5 I 10 I 15 I 20 25 Soil P, mg/L Figure 16. An example of the Langmuir and cumulative equations summed and applied to equilibrium phosphorus adsorption data for a Hartsells surface soil. 36 Phosphorus Accumulation and Loss from Alabama Soils shaking, centrifuged, filtered, and soluble P determined. The suspension clay concentrations, pH, initial and final adsorbed phosphorus, measured and simulated solution phosphorus are shown in Table 9. The initial resident (total) P in the soils is given in Table 8. MINTEQA2 soluble ion speciation model did not satisfactorily predict solution P concentrations. In most cases, but not all, the model over-predicted solution P concentration. The model included both an adsorption distribution coefficient and a Langmuir exchange component. Input could be adjusted to make any one sample fit the data. However, the adjustments could not be applied across all soils and treatments. The simulated concentration values reported were calculated using Langmuir and Kd constants calculated from isotherm data, final suspension pH, and mineralogy of each soil clay. The differences from measured were not consistent enough to suggest possible modification of model parameters. Soluble P was also evaluated with respect to clay suspension concentration, equilibrium adsorbed P concentrations, and litter-applied and monocalcium-P treat- Soil, ppm 700 600 500 Linker, P-Isotherm Langmuir (354.911, 0.88916) Cum 2 (*233.747, *14.1246, *3.13271) No Background X2 = 2509.6077 r2 = 0.9873677 400 300 200 100 " 0 0 i I I I 5 10 15 20 25 Solution, ppm Figure 17. An example of the Langmuir and cumulative equations summed and applied to equilibrium P adsorption data for a Linker surface soil developed with Peakfit software. Alabama Agricultural Experiment Station 37 ments. The data were grouped into litter-treated and monocalcium-P treated samples. Each group was plotted by individual soil and both soils combined into one graph. The monocalcium-P treatments are shown in graphs in Figure 23. The figures show the linear relation between P adsorbed on clay and "in solution." The slope, approximately 3,000 mL/g, corresponds to an equilibrium distribution coefficient that is essentially a constant throughout the concentration range of the experiment. Extrapolating to zero concentration in solution should be an approximation of "fixed" or very tightly held P. The extrapolated constant 2,884 mg/kg clay corresponds to 583 mg/kg for fixed P in the Dewey soil and 164 mg/kg in Hartsells. Similar linear relationships resulted from litter-applied equilibrium experiments (Figure 24). The slope in this group is approximately 6,000 mL/g, twice that of the MCP test. This indicates more litter P in the suspended solid phase and also in the "fixed" phase as indicated by higher extrapolated values. Since samples receiving litter had higher pH following incubation, this difference could reflect the effect of pH on solution complexes of Ca, Al, and Fe phosphates (42). Soil P, mg/kg Hartsells - Languir 800 --- Sum of 1+2 line -. -- 60 - * Linear Experimental data pts. 400 00 s f f It. 100 00 00 00 . 100 0 _._ __ 200 oL 0 10 15 Solution P, mg/L 20 25 Figure 18. An example of the linear and Langmuir equations applied to equilibrium phosphorus data for a Hartsells surface soil. 38 Phosphorus Accumulation and Loss from Alabama Soils 38 PopouAcuuainadLsfrmAaaaSI Soil P,mg/kg 1,400 1,200 1,000 800 - "r No phosphorus applied I 600 400 200 - () 1,400 1,200 1,000 800 600 400 200 0( 1,400 1,200 1,000 800 600 400 200 0( - U 2 4 6 8 56 Mg/ha litter applied I I I I 4 6 -,0 Mineral phosphorus applied 4 Solution P, mg 6 Figure 19. Dewey surface soil equilibrium P adsorption from solution following poultry litter and monocalcium phosphate addition and incubation for 91 days. Alabama Agricultural Experiment Station 39 Soil P,mg/kg 1,400 1,200 1,000 800 - 600 400 200 0 I I I I No phosphorus applied I I I ! 0 1,400 1,200 1,000 800 600 - 2 4 6 8 10 12 14 16 400 200 - 56 Mg/ha litter applied 0 1,400 1,200 1,000 800 - 2 4 6 8 10 12 14 16 Mineral phosphorus applied 600 400 200 - 0 2 4 10 8 6 Solution P, mg 12 14 16 Figure 20. Hartsells surface soil equilibrium P adsorption from solution following poultry litter and monocalcium phosphate addition and incubation for 91 days. 40 Phosphorus Accumulation and Loss from Alabama Soils 40 Phosphorus Accumulation and Loss from Alabama Soils Soil P, mg/kg 1,400 1,200 1,000 800 600 - DC No phosphorus applied I I 400 200 - 0 1,400 ) 1,200 1,000 1,000 0 2 4 6 5 8 10 12 - 800 600 400 200 56 Mg/ha litter applied I I I I I I oc )= _ 4 6 8 10 12 1,400 1,200 1,000 800 600 400 200 Mineral phosphorus applied oC I I I I I I 2 4 6 Solution P, mg 8 10 12 Figure 21. Linker surface soil equilibrium P adsorption from solution following poultry litter and monocalcium phosphate addition and incubation for 91 days. Alabama Agricultural Experiment Station 41 Alabama Agricultural Experiment Station 4 The solution:soil P distribution is also affected by the suspended clay concentration (Figure 25). More clay in suspension resulted in more P remaining on the clay and yielding more P into solution. Several studies have documented a relationship between soil test P concentration and dissolved, runoff, and soil solution P (67, 1). All show that there is a potential for enrichment in dissolved P as soil test values increase. This approximately linear relationship was also observed in this study (Table 10). Figure 26 shows dissolved P and soil test values from surface samples collected from paired cotton fields, paired pastures, and incubated samples of Dewey and Hartsells soils following addition of litter. Dissolved P was obtained by suspending an equivalent amount of clay from the different soils and shaking overnight. According to these data 300 mg/kg Mehlich-1 extractable P would predict a dissolved P concentration of 1,000 gg/L. A solution concentration of 1,000 g/L has been accepted as a limit for treated municipal water discharged into streams and suggested as a possible limit for runoff P concentration (67). TABLE 7. LANGMUIR CAPACITY CONSTANTS CALCULATED FROM EQUILIBRIUM ISOTHERM DETERMINATIONS USING SOIL CHECK SAMPLES Soil P capacity R-squared' 0.90 0.83 0.93 0.92 0.97 0.98 0.98 0.96 0.91 0.94 0.94 0.94 0.98 0.98 0.98 0.98 0.95 0.95 0.95 0.95 Avg. capacity mg/kg 300 --- mg/kg 292 Hartsells-Ap .............. 334 Hartsells-Ap .............. 291 Hartsells-Ap .............. 281 Hartsells-Ap .............. Hartsells-Bt ................... 612 Hartsells-Bt ................... 606 Hartsells-Bt ................... 607 Hartsells-Bt ................... 612 Linker-Ap ......................... 631 Linker-Ap ......................... 595 Linker-Ap ......................... 600 Linker-Ap ......................... 604 Linker-Bt .......................... 850 Linker-Bt.........844 Linker-Bt.........865 Linker-Bt .......................... 865 Dewey-Ap ......................... 722 Dewey-Ap ......................... 729 DeweyAp........741 DeweyAp........738 610 --- 608 --- 856 -- 733 -- - Dewey-Bt .......................... Dewey-Bt .......................... Dewey-Bt .......................... Dewey-Bt .......................... 855 864 859 865 0.98 0.98 0.98 0.98 861 --- 'R2 for fit of data to the linear form of the langmuir equation, four reps. 42 Phosphorus Accumulation and Loss from Alabama Soils 42 Phosphorus AccumulationadLs rm lbm ol mg/kg 800 Languir Pcapacities Constant = 306.1455 2 = 0.862001 X2 -=16.95165 400 -' 400 " .'' " 200 " 0 0 5 10 - - -- Reg. line 0 Experimental data pts. 15 Clay, % 20 25 Figure 22. Surface and subsurface Langmuir P adsorption capacities obtained from Hartsells, Dewey and Linker soils. Each observation is the average of four replicate isotherms. . INITIAL AMOUNTS OF MINERAL P IN SOILS BEFORE BATCH EQUILIBRIUM TEST FOR ION SPECIATION MODEL VALIDATION (MINTEQ) TABLE 8 Soil & horizon Treatment Total P mg/kg 656 311 969 702 1,938 1,404 Dewey, Ap ............................. Check, 0 added Hartsells, Ap ..................... Check, 0 added Dewey, Ap ............................. 25 tons/acre litter' Hartsells, Ap ..................... 25 tons/acre litter' Dewey, Ap ............................. 50 tons/acre litter' Hartsells, Ap ..................... 50 tons/acre litter' 'Soil samples were also incubated with equivalent amounts of mineral fertilizer P added. Alabama Agricultural Experiment Station 43 Alabama Agricultural Experiment Station TABLE 9. MEASURED AND PREDICTED EQUILIBRIUM SOLUTION 4 P pH CONCENTRATIONS IN DILUTE CLAY SUSPENSIONS FOLLOWING INCUBATION Soil & treatment Soil Clay g/L 0.51 1.01 0.51 1.01 0.51 1.01 0.51 1.01 0.51 1.01 0.57 1.14 0.57 Initial P Final P Solution P 1 MINTEQ mg/L 0.04 0.04 0.33 0.42 0.36 2 g/L Decatur-Ap, check ........... 2.50 Decatur-Ap, check ........... 5.00 Decatur-Ap, 25 tons litter............ 2.50 Decatur-Ap, 25 tons litter............ 5.00 Decatur-Ap, 50 tons litter............ 2.50 Decatur-Ap, 50 tons litter............ 5.00 Decatur-Ap, MCP, 25T equv...... 2.50 Decatur-Ap, MCP, 25T equv...... 5.00 Decatur-Ap, MCP, 50T equv...... 2.50 Decatur-Ap, MCP, 50T equv...... 5.00 Hartsells-Ap, check check mg/kg 2,559 mg/kg 2,520 mg/L 0.02 0.02 0.07 0.08 0.13 0.18 0.24 0.35 0.55 0.90 0.11 0.15 0.56 0.87 1.23 1.88 1.25 2.11 2.27 3.84 5.15 5.09 2,559 3,409 3,409 4,258 4,258 4,249 4,249 5,939 5,939 3,934 3,934 8,969 2,539 3,271 3,330 4,001 4,080 3,775 3,903 4,852 5,050 3,742 3,803 7,990 6.62 6.68 6.74 6.47 0.39 0.75 1.20 2.19 0.56 0.04 0.31 0.31 0.51 0.46 3.08 4.72 2.51 3.35 5.21 5.12 5.22 5.19 5.93 6.22 6.97 6.95 ............. 10.00 Hartsells-Ap, ............. 20.00 Hartsells-Ap, 25 tons litter .... 10.00 20.00 Hartsells-Ap, 25 tons litter .... Hartsells-Ap, 50 tons litter 1.14 0.57 8,969 14,003 8,209 11,853 .... 10.00 7.07 7.28 6.29 6.10 6.42 6.31 Hartsells-Ap, 20.00 50 tons litter .... Hartsells-Ap, MCP, 25T equv. ... 10.00 Hartsells-Ap, MCP, 25T equv. ..20.00 Hartsells-Ap, MCP,S50T equv. ..10.00 Hartsells-Ap, MCP,S50T equv. ..20.00 1.14 0.57 1.14 0.57 1.14 14,003 9,913 9,913 15,892 15,892 12,360 7,728 8,069 11,924 12,535 'Predicted 'Measured. equilibrium solution P. 44 Phosphorus Accumulation and Loss from Alabama Soils PredictingLong-term Phosphorus Accumulation, Leaching, and Runoff A paired pasture and a paired cotton field were sampled to one meter. One field in each pair received litter for 18-20 year and the other was under conventional N and P fertilization. Phosphorus adsorption isotherms from the previous studies, and the constants in Table 7 were used to determine P sorption ratios for the Hartsells and Dewey surface and subsoils. Resident mineral (total and resin extractable) P and organic P were determined to sampled depths. EPIC (Erosion/Productivity Impact Calculator) was used to simulate P uptake, leaching and runoff losses under fescue and cotton receiving litter and under conventional mineral fertilization. The EPIC model has a P component that simulates uptake, leaching, and runoff losses both soluble and on susClay phosphorus mg/kg, thousands 6 5Average clay: Solution P distribution Dewey AP Horizon Monocal-P Clay phosphorus mg/kg, thousands clay 14 Average Pdisttribution Solution Hartsells AP Horizon 14 - 12 10 Monocal-P 9 * 8 6 - 0 - * 2L 0 I I I I 4 2 ( 0.2 0.4 0.6 0.8 0 14 -12 10 - 14 _ Solution P distribution 12 10 8 6 4 2 ff. Average clay: - Dewey and Hartsells AP Horizon Monocal-P with controls 0 I I I 1 2 3 4 Average clay: Solution Pdistribution Dewey and Hartsells AP Horizon Monocal-P " without controls - . 8 -* 1 9. 6 4 0 I I I I 2 I I I I 0 1 2 3 4 0 1 2 3 I 4 Solution phosphorus, mg/L Solution phosphorus, mg/L Figure 23. Equilibrium solution:clay P distribution following high levels of monocalcium phosphate addition, 91 days incubation, and 30 days in a 0.01M CaCl 2 suspension. Alabama Agricultural Experiment Station 45 pended solids. At equilibrium, P is considered to exist in three pools: labile, active, and stable. The stable form is four times the active and labile is about one-tenth of the active but has a temperature and time relationship. Additions or losses from any phase causes the equilibrium to adjust back to these limits (79). Figure 27 shows a flow chart of P partitioning as simulated by EPIC. Soil, climate, management, and crop input was obtained from various sources. EPIC's climate generator was driven by weather data from the station nearest to the pastures and fields selected for study (Cullman and Birmingham, Ala.; and Chattanooga, Tenn.). Soil layer information was obtained from EPIC's soil database and modified with on-site data when available (Hartsells #309, and Dewey #186). Management was obtained from land managers, the farmers managing the selected fields. Tables 11 and 12 give the Clay phosphorus, mg/lkg 4,500 - Solution Pdistribution Dewey AP Horizon Average clay: Clay phosphorus mg/kg, thousands 14 * 12 10 0 8 6 0 Average clay: Solution P distribution Hartsells AP Horizon 0 4,000 3,500 3,000 - 2,500 · I - "g/k4 2,000 L0 14,000 12,000 10,000 8,000 6,000 4,0002,000 0 I I I 0.05 0.1 0.15 0.2 Average clay: Solution P distribution Dewey and Hartsells AP Horizon Litter with controls * 0 14 12 1 2 3 4 Average clay: Solution P distribution Dewey and Hartsells AP Horizon Lifter without controls0 10 0 8s 1 2 3 4 I I I I I 2 . 1 2 3 4 0 Solution phosphorus, mg/L Solution phosphorus, mg/L Figure 24. Equilibrium solution:clay P distribution following high levels of poultry litter addition, 91 days incubation, and 30 days in a 0.01M CaCl 2 suspension. 46 Phosphorus Accumulation and Loss from Alabama Soils EPIC operation schedules for grazed fescue pastures and cotton. Tables 13 and 14 give background soil test and P partitioning data for input and comparison with EPIC predictions. The comparisons of EPIC-simulated, litter-applied, conventional-fertilizer-applied, and sampled-soil-layer P concentrations are presented as total mineral P distributions with depth in Figures 28 and 29. The distributions after 20 years of litter application compare favorable with actual concentra- Clay phosphorus, mg/kg 6150T 5r Dewey Ap B A 1 4 3 Check I A B 25T I I 0 14- 0.2 0.4 50T 0.6 0.8 B 12 10 A A B 25T Hartsells Ap U Check -I, 0 2 3 Solution P, mg/L Figure 25. Effect of clay suspension concentration on equilibrium P in solution; A = 0.5, B = 1.0 g/L. Alabama Agricultural Experiment Station 47 TABLE 10. SOIL TEST AND SOLUTION P FROM PAIRED-FIELD, FESCUEPASTURE, AND INCUBATED LITTER-TREATED SURFACE SOIL SAMPLES Series Dewey ............ Dewey ................ Dewey ............ Dewey ............ Dewey ............ Hartsells ............ Hartsells ............ Hartsells ............ Hartsells ............ Hartsells ............ Treatment cotton, no litter cotton, litter check 25 tons litter 50 tons litter fescue, no litter fescue, litter check 25 tons litter 50 tons litter Clay pct. 23.20 13.92 20.41 20.41 20.41 6.99 11.23 5.71 5.71 5.71 pH 4.99 6.56 6.38 6.56 6.88 5.48 6.74 6.47 6.89 7.18 Soil-test P mg/kg 13 120 27 61 91 59 163 48 205 305 Dissolved P mg/L 0.14 0.48 0.03 0.15 0.20 0.14 0.59 0.16 0.69 0.93 tions for both fescue pasture and cotton. In both cropping systems, measured and simulated P accumulated to over 1,000 g per ton of soil in the littered treatments. Under conventional cotton, P buildup was approximately 500 g per ton at the surface and in the 1-15 cm layer. EPIC simulates high concentrations at the surface when conventional P is used because it is not incorporated. Considering the many assumptions required to run the model, simulated results predict actual measured concentrations well and support the use Soil test, mg/kg 400 300 r 0 200 100 -0 0 I I 0 0.2 0.4 0.6 Solution P, mg/L 0.8 1 Figure 26. Relationship between soil test P and dissolved P of Hartsells and Dewey surface soil samples from paired cotton fields and fescue pastures and incubated samples. co I.r~lCmineral and0 I Fresh Sal organic I. I0 C) L ie Activehumus 5 0 a) 0 0* Figure 27. Phosphorus partioning as simulated in the EPIC model. 2. N. Alabama Agricultural Experiment Station 49 TABLE 11. FESCUE PASTURE MANAGEMENT INPUT USED TO DRIVE EPIC SIMULATION' Month Day EPIC ID# Name Tillage Operation Graze Drill plant 2 Graze Weight kg/ha Grazing days 3 ................ 10 .............. 10 .............. 2 1 1 67 3 67 --- 90 -- 90 3 ................ 5 9 .......... 20 9 .............. 20 3 ................ 1 6 .............. 15 63 63 63 21 21 Fertilizer Applied 34-0-0 34-0-0 11-52-0 Litter Applied Poultry-fresh broiler 3 Poultry-fresh broiler 150-100 75 3,750 3,750 -- 'Operation schedule for first year, one-year rotation, 20-year simulation. Hartsells soil, EPIC Soil data ID no. 309, modified for North Alabama soils. Land slope - 2% and slope length was 30 meters. 2 First year of simulation. 3 Fraction of broiler litter weight: mineral N = 0.012; NH3 = 0.99; organic N = 0.027; mineral P = 0.012; and organic P = 0.004. TABLE 12. COTTON MANAGEMENT INPUT USED TO DRIVE EPIC SIMULATION' Month Day EPIC ID# Name Weight kg/ha Tillage Operation Disk bed 16 3 ............................ 30 Row bed-17 4 .................................. 1 2 Row plant cotton-15 4 ............................... Row cultivate-30 19 4 ............................... Row cultivate-19 10 5 ............................... Row cultivate 19 1 6 .................................. 50 Harvest (95% eff.)-10 10 ............................. Kill 41 11 10 .............................. 28 Mold board plow-25 10 ............................. Fertilizer Applied 63 34-0-0 14 4 ............................... 53 11-52-0 14 4 ............................... Litter Applied 2 Poultry-fresh broiler 28 21 3 ............................... 21 Poultry-fresh broiler 15 6 ............................... --- 300 75 7,200 3,750 'Operation schedule for first year, one-year rotation, 20-year simulation. Decatur soil, EPIC Soil data ID no. 186, modified for Dewey soils in North Alabama. Land slope = 2% and slope length was 30 meters. 2 Fraction of broiler litter weight: mineral N = 0.012; NH 3= 0.99; organic N = 0.027; mineral P = 0.012; and organic P = 0.004. 50 TABLE Phosphorus Accumulation and Loss from Alabama Soils 13. SOIL TEST DATA FOR LITTERED AND NON-LITTERED SOILS IN TO EVALUATE THE ABILITY OF NORTH ALABAMA-USED EPIC TO PREDICT THE FATE AND TRANSFORMATIONS OF Soil Soil pH Litter Non Soil test P Litter Non P APPLIED AS LITTER Soil test Mg Litter Non Soil test Ca Litter Non Soil test K Litter Non Decatur-Ap...... Decatur-BA .... Decatur-Btf .... Decatur-Bt....... Decatur-CB .... Hartsells-Ap... Hartsells-BA .. Hartsells-Btf... Hartsells-Bt .... Hartsells-CB ... 7.1 7.0 7.2 7.0 5.3 6.5 6.0 5.8 4.8 4.6 5.0 5.4 5.8 5.3 4.9 5.2 5.2 5.0 4.8 4.8 mg/kg 120.0 126.0 128.0 2.5 1.3 163.0 100.0 66.0 1.3 0.8 mg/kg 13.0 5.5 3.3 0.9 0.8 59.0 19.0 5.3 0.8 0.8 mg/kg 204 153 151 118 38 235 104 84 65 56 mg/kg 122 68 47 22 24 79 43 42 44 34 mg/kg 93 82 85 72 56 197 67 56 48 35 mg/kg 58 56 63 108 123 31 10 8 13 12 mg/kg mg/kg 1,985 397 2,024 512 1,365 645 529 512 425 387 2,505 237 810 250 559 273 344 113 254 80 of EPIC in planning BMPs for litter application. In addition, EPIC is an excellent tool to use in predicting soil loss by water erosion under many crops and management schedules. Kingery et al. (40) reported P depth distributions determined by soil testing methods (Mehlich-1 extractable P). They sampled 11 paired fescue pastures on Hartsells soils in the poultry-producing region of North Alabama. Direct comparisons of total P cannot be made. However, Jones et al, (37) used soil test data to develop their active P pool concept. Soil test P depth distribution data were averaged for all pastures, and the same EPIC pasture management schedule used for the single site to predict total P after 10 and 20 years of litter application and conventional P fertilization was used for this averaged set. In this case active P was predicted with depth. The litter soil test and EPIC active P correspond well between simulated and measured (Figure 30). In this graph the weighted mean of the 0-15 cm layer predicted P was used since this is what was sampled by Kingery et al. (40). Only one EPIC-predicted point, 15-30 cm, under conventional fertilization does not follow the measured trend. There is a considerable amount of soil test P data available from many regions and there is considerable support for using soil test values as evidence of P accumulation in surface soil layers (67). EPIC offers a tool to predict P buildup to critical levels and the magnitude of soluble and sediment P in runoff when critical levels are reached. Figures 31 and 32 show 20year EPIC simulations of P loss on sediment and as soluble P concentration in runoff from fescue pasture and cotton. Most notable is the high soluble P concentration predicted for runoff from pastures receiving litter. It is also notable that concentrations in runoff are not increasing appreciably. This shows that P losses are controlled, to a greater extent, by levels of annual applications rates and not by levels of P accumulation as fixed and tightly bound P. Alabama Agricultural Experiment Station 51 Alabama Agricultural Experiment Station 51 Total P,G/T O 20 40Q 60 80 Depth, cm 500 Hartsells fsl Fescue pasture Conventional fertilizer applied o 0Oyears Sampled *l0 years 20 years Q Last sampled depth Total P, G/T A 00500 20k 1,000 A 40 60 80 Q Hartsells fsl Fescue pasture Broiler litter applied Figure 28. Sampled and EPIC-predicted P distribution with depth after 10 and 20 years of poultry litter and conventional mineral fertilizer P application on a fescue pasture. 52 Phosphorus Accumulation and Loss from Alabama Soils 52 Phosphorus Accumulation and Loss from Alabama Soils Total P, G/T 0r 0 1,000 1,500 20 40 60 80 Decatur Cotton Conventional fertilizer applied 100 Depth, cm o 0 years * 10 years A 20 years a Last sampled depth Sampled Total P, G/T 00 20 40 60 80 100 Depth, cm Figure 29. Sampled and EPIC-predicted P distribution with depth after 10 and 20 years of poultry litter and conventional mineral fertilizer P application on a cotton field. Decatur Cotton Broiler litter applied Alabama Agricultural Experiment Station 53 Alabama Agricultural Experiment Station TABLE 14. BACKGROUND INFORMATION FOR LITTERED AND NON-LITTERED SOILS IN NORTH ALABAMA USED TO EVALUATE THE ABILITY OF EPIC TO PREDICT THE FATE AND TRANSFORMATIONS OF P APPLIED AS LITTER Soil & horizon Total P mg/kg Decatur-Ap ............... 1,048 Decatur-BA .............. 1,115 338 Decatur-Bt ......... Hartsells-Ap ............. 1,231 Hartsells-BA ....... 658 Hartsells-Bt ........... 345 Decatur-Ap .......... 506' Decatur-BA ........... 5101 Decatur-Bt ............. 3401 Hartsells-Ap .......... 4731 Hartsells-BA.........4331 Hartsells-Bt ........... 266 Organic P mg/kg 181 283 53 416 123 70 125 104 96 269 117 40 Resin P Soil test P mg/kg 120.0 126.0 2.5 163.0 100.0 1.3 13.02 5.5' 0.93 59.0' 19.0 0.8 RatioSTP mg/kg 6.6 5.7 121.0 4.0 4.4 307.0 28.0 74.0 828.0 2.6 8.8 173.0 RatioACT mg/kg 25.0 14.0 138.0 6.9 6.8 407.0 37.0 110.0 1,351.0 2.7 9.5 264.0 5 mg/kg Littered 75.00 78.00 1.27 68.00 36.00 0.91 Non-littered 2.941 1.41' 1.09 7.622 4.423 1.43 'Indicates significant difference at the 0.01 level of probability. 2 lndicates significant difference at the 0.05 level of probability. 3 lndicates significant difference at the 0.10 level of probability. 0 Active P, mg/kg 200 100 0 300 o No litter-soil test " No litter-EPIC Litter-soil test 0 Liter-EPIC Depth, cm Figure 30. Soil test P and EPIC-predicted active P after 20 years of poultry litter application to fescue pastures on Hartsells soils in North Alabama. 54 Phosphorus Accumulation and Loss from Alabama Soils RELATED PHOSPHORUS RESEARCH IN ALABAMA Phosphorus Mobility Associated with Suspended Eroded Soil Material' Caldwell (12) conducted a study to characterize the mineralogy of water dispersible clays (WDC) in relation to soils and cropping practices. The objectives of this study were to identify the mineralogy of eroded soil material in runoff water from cropped and pasture land, half of which had received annual applications of poultry litter, and to suggest how the suspended material would behave with respect to P adsorption. 1 Caldwell, T. 1996. Mineralogical comparisons of surface soils and their water dispersible clay fractions with the suspended fractions in surface water of selected Alabama agricultural lands. M.S. Thesis, Auburn Univ. AL. 91p. P, P, kg/ha/yr o Litter applied kg/ha/yr Phosphorus lost with sediment * Convetional fertilizer 4- 2 0 1970 P, ppm 100 - 1975 1980 1985 1990 Phosphorus concentration in runoff water 10 1- 0.1- 0.01 1970 1975 1980 1985 1990 Figure 31. EPIC-predicted average annual P concentrations in runoff water from fescue pasture receiving poultry litter and under conventional fertilization for 20 years. Alabama Agricultural Experiment 55 Alabama Agricultural Experiment Station Gibbsite and goethite are noted for their phosphate adsorption capabilities (16, 76, 77, 81), but they comprise a small percentage of the mineralogy of the Hartsells and Dewey soils studied from Cullman and Cherokee counties. Kaolinite and hydroxy-interlayered-vermiculite (HIV) do not have the phosphate adsorbing capabilities of Fe and Al oxides, but they comprise the majority of the mineralogy of these soils. In general, the P adsorption capacities are as follows: amorphous hydrated oxides > goethite-gibbsite > kaolinite > 2:1 clays (16). Phosphates bind to edge sites of HIV as well as to the sites located on the hydroxy interlayer.The Hartsells and Dewey soils also contained muscovite. Phosphate adsorption onto micas is minor and occurs only at the edge surfaces. Any reported significant phosphate adsorption onto mica surfaces is probably because of Fe oxide coatings (54). Station 55 P, kg/ha/yr 7141 1r Phosphorus lost with sediment o Litter applied " Convetional fertilizer 65432 1 LI1970 I I I I I I I I I I I I I I I I I I I I I I 1975 1980 1985 1990 Average P, ppm/yr 0.4 I- Phosphorus concentration in runoff water 0.3 I0.2 0.1 0L Li I I I 1970 I I I I II I I I I I I I I I I I I I I I I I I I I I I I I 1 1975 1980 1985 1990 Figure 32. EPIC-predicted average annual P concentrations in runoff water from cotton fields receiving poultry litter and under conventional fertilization for 20 years. 56 Phosphorus Accumulation and Loss from Alabama Soils The amount of phosphate desorbed is related to the amount of Bray Pl-extractable P contained in the soil or eroded material (51, 77). Phosphate desorption is determined by: (1) P-desorption capacity of the soil or eroded material; (2) equilibrium chemical potential; and (3) rate of release (51). Oloya and Logan (51) observed that the soil material which contained low amounts of extractable P appeared to contain a pool of desorbable P that is greater than the amount desorbed. This pool of P was not depleted with sequential extraction. They suggested that this pool of P could be released over long periods of time. On the other hand, soil material that was high in extractable P contained a pool of easily desorbable P, which was depleted at about the fifth extraction. Riley and Syers suggested that the easily desorbed P is physically sorbed, while the more strongly held P is chemisorbed (55). Large amounts of applied P can be eroded from cultivated land (21, 58). However, loss of applied P is generally less than 1% (64). Even at such low levels, the total and soluble P can exceed the critical values of 0.02 and 0.01 mg/L, above which biological growth can be stimulated (64). Particulate P levels ranging from 0.06-9.61 mg/L have been reported for runoff from cultivated land, pasture, and forest. Alberts and Moldenhauer (4) measured available P levels of 68182 mg/L for the less than 0.05-mm fraction in runoff from a Sidell silt loam under various tillage treatments. The contribution of the water-dispersible clay (WDC) fraction to the particulate P content is difficult to assess. The primary clay, or WDC fraction, has been reported to be 5.2% or less of the total eroded soil material for a Miami silt loam (3), a Sidell silt loam (4), and Monona and Ida soils (3). The dissolved P concentrations from a recent study (0.170-3.408 mg/L for a Hartsells site in North Alabama and 0.0116-2.606 mg/L for an Orangeburg site in South Alabama) were within the ranges reported in the aforementionedmentioned studies (12). Though most P in runoff water is lost in a particulate form (45, 63), loss of soluble P may increase as land-application rates of animal wastes increases. Annual animal waste applications to cultivated land and pasture in excess of crop requirements and P-adsorption capacity of soils can result in downward P movement (7, 23, 40, 53). If there is downward movement of P then it also can be desorbed into runoff water. Depending on the P-adsorption capacity of the suspended material and of the soil over which the runoff flows, the soluble P may be readsorbed (62) or remain in solution. Stream and river banks, which usually have high P-adsorption capacities, can reduce the P load entering waterways by adsorbing soluble P (78). Concern over the pollution potential of P has led to an increased awareness of its transport from cultivated land, pasture, and aquacultural facilities. The Clean Water Act gave states the responsibility to develop strategies for controlling agricultural pollution. In Alabama, no laws exist which require monitoring of P movement. Currently, recommended limits set on the amount of broiler litter that can be applied to crop land and pasture are based on N content of the litter and crop N requirements. Alabama Agricultural Experiment Station 57 Litter was observed to have no effect on suspended solid mineralogy. Pasture reduced the relative percentage of quartz in runoff water. Quartz was the dominant mineral in runoff from Hartsells and Orangeburg soils. Hydroxy-interlayered vermiculite and kaolinite, the major P adsorbing minerals in these soils, comprised over 50% of the suspended solid mineralogy in farm ponds and streams. Hydroxy-interlayered vermiculite was dominant in receiving waters in Hartsells watersheds, while kaolinite was dominant in Orangeburg watersheds. The waterdispersible clay of the Hartsells Ap horizon was a better indicator of suspended solid mineral distribution in receiving waters than was the water-dispersible clay of the Orangeburg Ap horizon. For the Orangeburg soil watersheds, subsoil mineralogy was the best indicator of suspended solid mineralogy in receiving waters because of contributions from exposed B and C horizons. 2 Watershed Study of Conventional- and Conservation-tillage Cotton Practices Soileau et. al. (72) conducted a watershed study in Colbert County, Ala., from 1984-89. Runoff sediment N and P were determined for both conventional and conservation tillage on cotton for a Decatur soil on 1-6% slopes. Surface runoff of P was markedly increased for storm events immediately after fertilizer P was surface applied to cotton under conservation tillage. In both tillage systems, most of the runoff losses of P were associated with the solution phase rather than the suspended solid phase. Under conservation tillage, average loss in the solution phase was 2.62 kg/ha per year; in the solid phase, 0.30 kg/ha per year. Under conventional tillage, average loss in the solution phase was 0.75 kg/ha per year; in the solid, 0.24 kg/ha per year. The study concluded that because of elevated losses of both N and P after surface application, it is important that NP applications be timely and at rates not exceeding crop needs. Total annual sediment losses for this watershed were much less than the tolerance limit of five tons per year and also less than predicted by the USLE equation. Long-term Phosphorus Levels Determined by Soil Test Analysis 3 Cope (14) conducted a study of the effects of 50 years of fertilization with P and K on soil test levels at six locations in Alabama. He found that on soils that received 14-18 kg/ha of P from 1929 through 1957, soil-test P increased from an average of 19 mg/kg in 1929 to 33 mg/kg by 1957. When applications were discontinued, the level dropped to that of untreated plots by 1973. Application of 27 kg/ha or more of P rapidly increased soil test P to "high" or "very high" levels, and these have been maintained by annual applications of 30 kg/ha during 20 years of high crop production. The highest rate, 54 kg/ha, raised average levels of soil-test P (Mehlich-1 extractable) from 43 kg/ha in 1929 to 217 kg/ha in 1957. This was one-ninth of the 1,566 kg/ha of P applied during this period. 2 Soileau, J.M., J.T. Touchton, B.F. Hajek, and K.H. Yoo. 1994. Sediment, nitrogen and phosphorus runoff with conventional- and conservation-tillage cotton in a small watershed. J. Soil and Water Cons. 49 (1):82-89.(14) 3 Cope, J.T., Jr. 1981. Effects of 50 years of fertilization with phosphorus and potassium on soil test levels and yields at six locations. Soil Sci. Soc. Am. J. 45:342-347. 58 Phosphorus Accumulation and Loss from Alabama Soils Soil Test Phosphorus Levels from Long-term Broiler Litter Application 4 Kingery et. al. (40) conducted an investigation of the impacts of longterm land application of broiler litter on soil-chemical properties in the Sand Mountain region of North Alabama. The sampling sites were located in Cullman, Blount, Marshall, and Dekalb counties. From each county, three pairs of sites were chosen that consisted of long-term (15-25 years) litter and nonlittered fescue pastures on matching soil series. The study showed P accumulation and downward movement to approximately 60 cm in littered pastures. Long-term land application of broiler litter increased soil test P (Mehlich-1 extractable) levels an average of 530% in the 0-60 cm depth, as compared to no-litter pastures. P concentrations measured in the 0-15 cm depth (greater than 250 mg/kg) in litter sites have a rating of "extremely high" (greater than 100 mg/kg), according to the Auburn University Soil Testing Laboratory. P concentrations of 25 mg/kg in the 0-15 cm depth are considered adequate for most crops on these soils. PAccumulation and Runofffrom Litterand Conventional Fertilizer Applications s Hall (29) conducted a study during 1991-1993 to determine the impact of land-applied broiler litter on water quality, soil quality, and biomass production. The soil was a Decatur silt loam on 2% and 4% slopes on the Tennessee Valley Substation in Limestone County, Ala. Treatments on the two slopes included commercial fertilizer applied at soil test recommendation levels, and broiler litter at four and eight tons per acre. The cropping system was conventionally tilled corn and a winter cover of rye. The soil was sampled before, during, and after the study. Runoff water was collected with a Coshocton type runoff sampler. Soil percolate was collected with a wick lysimeter. The analysis of runoff water samples included total and dissolved P. The P content of the broiler litter was 14.8 g/kg in 1991 and 34 g/kg in 1992. Total P concentrations in runoff waters from both slopes, and all P application rates exceeded the limit that will support algae growth in surface lakes and streams (0.002-0.1 mg/L) in the 1991-1992 crop year. The highest TP concentration was 2.66 mg/L from 4% slopes occurring in runoff in late December. In the 1992-1993 crop year, TP was significantly related to treatment. Total P concentrations reached a maximum of 8.1 mg/L in a November runoff event from plots receiving eight tons of litter per acre. The maximum concentration from 4% slopes this crop year was 6.6 mg/L. Dissolved P concentrations in runoff during the 1991-1992 crop year from both slopes were not significant between treatments. Concentrations on all sampling occasions exceeded algae growth limits. The highest concentration, 4 Kingery, W.L., C.W. Wood, D.P. Delaney, J.C. Williams, and G.L. Mullins. 1994. Impact of long-term land application of broiler litter on environmentally related soil properties. J. Env. Qual. Vol 23, no. 1, pp. 139-147. 5 Hall, B.M. 1994. Broiler litter effects on crop production, soil properties, and water quality. M.S. Thesis. Auburn University, Auburn, AL. Alabama Agricultural Experiment Station 59 Alabama Agricultural Experiment Station 5 1.95 mg/L, occurred from a 2% slope in late September. During 1992-1993, dissolved P concentrations were highest on most runoff events where eight tons of litter were applied, reaching a maximum of 6.4 mg/L in a November runoff event. Litter applications promoted greater concentrations of sediment P than fertilization. During the 1991-1992 crop year, the highest sediment P concentration occurred under four-ton application rates (0.8 mg/L), and all concentrations for all events exceeded algal limits for P effects on eutrophication. In the 1992-1993, sediment P concentrations were usually highest from the eightton treatment, reaching a peak of 3 mg/L in a November runoff event. This is three times the concentration observed in the preceding year. Mehlich-1 extractable P in the 0-15 cm and 15-30 cm layers at the end of the study was significantly affected by litter application on both slopes. Residual soil P concentrations were increased from about 25 mg/kg in commercial fertilizer treatments to 115 mg/kg at eight-ton-per-acre litter application rates. In the 0-5 cm layer, soil-test P had a rating of "extremely high," according to the Auburn University Soil Testing Laboratory. The results of runoff water collection and P determinations indicated that all treatments, slope, two litter application levels, and commercial fertilizer application, have potential for degrading surface waters via loses of P. However, impacts of runoff losses of P are usually observed in receiving waters, such as rivers and lakes, rather than at the edge of the contributing field. 60 Phosphorus Accumulation and Loss from Alabama Soils LITERATURE CITED (1) Adams, J.F., Fred Adams, and J. Odom. 1982. Interaction of phosphorus rates and soil pH on soybean yield ans soil solution composition of two phosphorus-sufficient ultisols. Soil Sci. Soc. Am. J. 46:323-328. (2) Adams, J.F., C.C. Mitchell, and H.H. Bryant. 1994. Soil Test Fertilizer Recommendations for Alabama Crops. Alabama Agric. Exp. Stn. Cir. 178. (3) Alberts, E.E., R.C. Wendt, and R.F. Piest. 1983. Physical and chemical properties of eroded soil aggregates. Trans. ASAE 26:465-471. (4) Alberts, E.E., and W.C. Moldenhauer. 1981. Nitrogen and phosphorous transported by eroded soil aggregates. Soil Sci. Soc. Am. J. 45:391-396. (5) Allison, J.D., D.S. Brown, and K.J. Novo-Gradac. 1991. MINTEQA2/PRODEFA2, a geochemical assessment model for environmental systems: Version 3.0 users manual. EPA/ 600/3-91/021. Environmental Research Laboratory, Office of Research and Development, U.S. EnvironmentalProtection Agency. Athens, GA 30613. (6) American Public Health Association, American Water Works Association and Water Pollution Control Federation. 1989. Standard methods for the examination of water and wastewater. 1989. 17th ed. APHA. Washington , DC. (7) Barrow, N.J. 1980. Evaluation and utilization of residual phosphorus in soils. In R.E. Khasawneh et al. (ed.) The role of phosphorus in agriculture. ASA, CSSA, and SSSA, Madison, WI. (8) Barrows, H.L., and V.J. Kilmer. 1963. Plant nutrient losses from soils by water erosion. Adv. Agron. 15:303-316. (9) Bohn, H.L., B.L. McNeal, G.A. O'Connor. 1985. Soil Chemistry, 2nd Ed. John Wiley and Sons, New York. (10) Bouma, J. 1989. Using soil survey data for quantitative land evaluation. Adv. Soil Sci. 9:177-213. (11) Burwell, R.E., D.R. Timmons, and R.F. Holt. 1975. Nutrient transport in surface runoff as influenced by soil cover and seasonal periods. Soil Sci. Soc. Am. Proc. 39:523-529. (12) Caldwell T. 1996. Mineralogical comparisons of surface soils and their water dispersible clay fractions with the suspended fractions in surface water of selected Alabama agricultural lands. M.S. Thesis, Auburn Univ. AL. 91p. (13) Castellanos, J.Z., and P.F. Pratt. 1981. Mineralization of manure nitrogen-correlation with laboratory indexes. Soil Sci. Soc. Am. J. 45:354-357. (14) Cope, J.T., Jr. 1981. Effects of 50 years of fertilization with phosphorus and potassium on soil test levels and yields at six locations. Soil Sci. Soc. Am. J. 45:342-347. (15) Daniel, T.C., A.N. Sharpley, D.R. Edwards, R. Wedepohl, and J.L. Lemunyon. 1994. Minimizing surface water eutrophication from agricultue by phosphorus management. J. Soil and Water Con. 49:30-38. (16)DeDatta, S.K. 1964. Availability of phosphorus and utilization of phosphate in some great soil groups of Hawaii. Diss. Abstr. 25:716. (17) Dixon, J.B. 1989. Kaolin and serpentine group minerals. In J. B. Dixon and S.B. Weed. (ed.) Minerals in soil environments. 2nd ed. Soil Sci. Soc. Amer. Madison, WI. (18) Drever, J.I. 1973. The preparation of oriented clay mineral specimens for x-ray diffraction by a filter membrane peel technique. Am. Miner. 58:553-554. (19) Edwards, D.R., and T.C. Daniel. 1993b. Runoff quality impacts of swine manure applied to fescue plots. Trans. ASAE 36:81-86. Alabama Agricultural Experiment Station 61 (20) Edwards, D.R., and T.C. Daniel. 1993a. Effects of poultry litter application rate and rainfall intensity on quality of runoff from fescue grass plots. J. Environ. Qual. 22:361-365. (21) Ensminger, L.E. 1952. Loss of phosphorus by erosion. Soil Sci Soc. Am. Proc. 16:338-342. (22) Fanning, D.S., R.F. Korcak, and C.B. Coffman. 1970. Free ion oxides: rapid determination utilizing x-ray spectroscopy to determine iron in solution. Soil Sci. Soc. Am. Proc. 34:941946. (23) Field, J.A., R.B. Reneau, and W. Kroontje. 1985. Effects of anaerobically digested poultry manure on soil phosphorus adsorption and extractability. J. Environ. Qual. 14:105-107. (24) Flynn, R.P. (1995). Comparative evaluation of composted broiler litter for crop production. Ph.D. Dissertation. Auburn University, Auburn, AL. (25) Gale, P.M., and J.T. Gilmour. 1986. Carbon and nitrogen mineralization kinetics for poultry litter. J. Environ. Qual. 15:423-426. (26) Gartley , K.L., and J.T. Sims. 1994. Phosphorus soil testing : Environmental uses and implications. Commun. Soil Sci. Plant Anal. 25:1565-1582. (27) Greeson, P.E. 1969. Lake eutrophication-a natural process. Water Resources Bull. No. 5(4): 1630. (28) Hajek, B.F., F.L. Gilbert, and C.A. Steers. 1975. Soil associations of Alabama. Agronomy and Soils Departmental Series No. 24. Alabama Agric. Exp. Station/Auburn Univ. Auburn, AL. 30p. (29) Hall, B.M. 1994. Broiler litter effects on crop production, soil properties, and water quality. M.S. Thesis. Auburn University, Auburn, AL. (30) Harter, R.D., and G. Smith. 1981. Langmuir equation and alternate methods of studying "adsorption"reactions in soils. In R.H. Dowdy et al. (Eds.) Chemistry in The Soil Environment, ASA Special Publication Number 40, Soil Sci. Soc. Am., Madison, Wisconsin. pp. 167182. (31) Heathman, G.C., A.N. Sharpley, S.J. Smith, and J.S. Robinson. 1995. Land application of poultry litter and water quality in Oklahoma, U.S.A. Fertilizer Research. 40:165-173. (32) Hsu, P.H. 1964. Adsorption of phosphate by aluminum and iron in soils. Soil Sci. Soc. Am. Proc. 28:474-478. (33) Hue, N.V., and C.E. Evans. 1986. Procedures used for soil and plant analysis by the Auburn University Soil Testing Laboratory. Dep. Ser. 106. Alabama Agric. Exp. Stn. (34) Jackson, M.L. 1956. Soil Chemical Analysis-Advanced Course. Published by author, Dept. of Soils, Univ. of Wisconsin, Madison, WI. (35) John, M.K. 1970. Colorimetric determination of Phosphorus in soil and plant materials with ascorbic acid. Soil Sci. 109:214-220. (36) Jones, C.A., A.N. Sharpley, and J.R. Williams. 1991. Modeling phosphorus dynamics in the soil-plant system. In Modeling Plant and Soil Systems, Hanks, John and J.T. Ritchie editors. Agronomy Series no. 31. Am. Soc. Agron. Madison, WI. (37) Jones, C.A., A.N. Sharpley, and J.R. Williams. 1984. A simplified soil and plant phosphorus model. Soil Sci. Soc. Am. J. 48:800-805. (38) Karathanasis, A.D., and B.F. Hajek. 1982. Revised methods for rapid quantitative determination of minerals in soil clays. Soil Sci. Soc. Am. J. 46:419-425. (39) Keup, L.E. 1968. Phosphorus in flowing waters. Water Res. 2:373-386. (40) Kingery, W.L., C.W. Wood, D.P. Delaney, J.C. Williams, and G.L. Mullins. 1994. Impact of long-term land application of broiler litter on environmentally related soil properties. J. Env. Qual. Vol 23, no. 1, pp. 139-147. 62 Phosphorus Accumulation and Loss from Alabama Soils (41) Lemunyon, J.L. and R.G. Gilbert. 1993. Concept and need for a phosphorus assessment tool. J. Prod. Agric. 6:483-486. (42) Lindsay, W.L., P.L.G. Vlek, and S.H. Chen. 1989. Phosphate minerals. In Dixon, J.B. and S.B. Weed. Minerals in Soil Environments, 2nd ed. SSSA Book Series no. 1. Soil Sci. Soc. Am. Madison, Wisconsin. pp. 1089-1130. (43) Lindsay, W.L. 1992. MINTEQA2 as a chemical speciation model for use in soil and water investigations. Proceedings of water resources and environment: Education, training and research. Colorado Water Research Institute, Information series 69. (44) McBride, M.B. 1989. Surface chemistry of soil minerals. In J. B. Dixon and S.B. Weed. (ed.) Minerals in soil environments. 2nd ed. Soil Sci. Soc. Amer. Madison, WI. (45) McLeod, R.V., and R.O. Hegg. 1984. Pasture runoff water quality from application of inorganic and organic nitrogen sources. J. Environ. Qual. 13:122-126. (46) Mehlich, A. 1953. Determinations of P, Ca, Mg K, Na and NH 4 by North Carolina soil testing laboratories. Memeo. North Carolina State University, Raleigh. (47) Muljadi, D., A.M. Posner, and J.P. Quirk. 1966. The mechanisms of phosphate adsorption by kaolinite, gibbsite, and pseudoboehmite:I. J. Soil Sci. 17:212-247. (48) Mullins, G.L. 1991. Phosphorus sorption by four soils receiving long-term applications of fertilizer. Commun. Soil Sci. Plant Anal. 22(7&8):667-681. (49) Murphy, J., and J.P. Riley. 1962. A modified single solution method for the determination of phosphate in natural waters. Anal. Chim. Acta. 27:31-36. (50) 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. (51) Oloya, T.O., and T.J. Logan. 1980. Phosphate desorption from soils and sediments with varying levels of extractable phosphate. J. Environ. Qual. 9:526-531. (52) Olson, S.R., and L.E. Sommers. 1982. Phosphorus. pp. 403-430. In A.L. Page et al. (Eds.) Methods of Soil Analysis, Part 2, 2nd ed. American Society of Agronomy, Madison, WI. (53) Reddy, K.R., M.R. Overcash, R. Khaleel, and P.W. Werserman. 1980. Phosphorus adsorption-desorption characteristics of two soils utilized for disposal of animal wastes. J. Environ. Qual. 9:86-92. (54) Reichenback, R.G.V., and C.I. Rich. 1975. Finer-grained micas in soils. p. 59-95. In J.E. Gieseking (ed.) Soil components. Vol. 2. Inorganic components. Springer Verlag, New York. (55) Riley, J.C., and J.K. Syers. 1977. Desorption and isotopic exchange relationships of phosphate sorbed by soils and hydrous ferric oxide gel. J. Soil Sci. 28:596-609. (56) Sanyal, S.K., P.Y. Chan, and S.K. De Datta. 1990. Phosphate sorption-desorption behavior of some acidic soils in South and Southeast Asia. Paper presented at the 6th Philippine Chemistry Congress, Cebu City, Philippines, 24-26 May, 1990. (57) Sanyal, S.K., and S.D. De Datta. 1991. Chemistry of phosphorus transformations in soil. In. B.A. Stewart (ed.)Advances in Soil Science, Volume 16, Springer-Verlag Inc., New York. pp. 1-120. (58) Scarseth, G.D., and W.V. Chandler. 1938. Losses of phosphate from a light textured soil in Alabama and its relation to some aspects of soil conservation. J. Am. Soc. Agron. 30:361374. (59) Schuman, G.E., R.G. Spomer, and R.F. Piest. 1973. Phosphorus losses from four agricultural watersheds on Missouri Valley loess. Soil Sci. Soc. Am. Proc. 37:424-427. (60) Schunost, A.C. and U. Schwertmann. 1995. Predicting phosphate adsorption-desorption in a soilscape. Soil Sci. Soc. Am. J. 59:1575-1580. Alabama Agricultural Experiment Station 63 (61) Sharpley, A.N. 1980. The enrichment of soil phosphorus in runoff sediments. J. Environ. Qual. 9:521-526. (62) Sharpley, A.N., R.G. Menzel, S.J. Smith, E.D. Rhoades, and A.E. Oldness. 1981. The sorption of soluble phosphorus by soil material during transport in runoff from cropped and grassed watersheds. J. Environ. Qual. 10:211-215. (63) Sharpley, A.N., S.J. Smith, and J.W. Nancy. 1987. Environmental impact of agricultural nitrogenand phosphorus use. J. Agric. Food Chem. 35:812-817. (64) Sharpley, A.N., and R.G. Menzel. 1987. The impact of soil and fertilizer phosphorus on the environment. Adv. Agron. 41:297-324. (65) Sharpley, A.N., and A.D. Halvorson. 1994. The management of soil phosphorus availability and its impact on surface water quality. P. 7-90. In R. Lal and B.A. Stewart (Eds.), Advances in Soil Science: Soil Processes and Water Quality. Lewis Publishers, Ann Arbor, MI. (66) Sharpley, A.N, T.C. Daniel, and D.R. Edwards. 1993. Phosphorus movement in the landscape. J. Prod. Agric. 6:492-500. (67) Sharpley, A.N., T.C. Daniel, J.T. Sims, and D.H. Pote. 1996. Determining environmentally sound soil phosphorus levels. J. Soil and Water Con. 51(2) 160-166. (68) Sharpley, A.N., S.J. Smith, and R.G. Menzel. 1986. Phosphorus criteria and water quality management for agricultural watersheds. Lake Reserv. Mgmt. 2:177-182. (69) Sims, J.T. 1986. Nitrogen transformations in a poultry manure amended soil: temperature and moisture effects. J. Environ. Qual. 15:59-63. (70) Sims, J.T. and D.C. Wolf. 1994. Poultry waste management: agricultural and environmental issues. In Adv. in Agron. Vol. 52. Aca. Press, Inc. pp. 2-72. (71) Singer, M.J., and R.H. Rust. 1975. Phosphorus in surface runoff from a deciduous forest. J. Environ. Qual. 4:307-311. (72) Soileau, J.M., J.T. Touchton, B.F. Hajek, and K.H. Yoo. 1994. Sediment, nitrogen and phosphorus runoff with conventional- and conservation-tillage cotton in a small watershed. J. Soil and Water Cons. 49 (1):82-89. (73) Soil Survey Staff. 1994. Keys to Soil Taxonomy, USDA-SCS, U.S. Govt. Printing Office. 300-124/00122. 306p. (74) 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. (75) Stevenson, F.J. 1986. Cycles of Soil: Carbon, Nitrogen, Phosphorus, Sulfur, Micronutrients. John Wiley and Sons, New York. (76) Syers, J.K., T.D. Evans, J.D. Williams, and J.T. Murdock. 1971. Phosphate sorption parameters of representative soils from Rio Grande Do Sul, Brazil. Soil Sci. 112:267-275. (77) Syers, J.K., J.T. Murdock, and J.D.H. Williams. 1970. Adsorption and desorption of phosphate by soils. Soil Sci. Plant Analysis. 1:57-62. (78) Taylor, A.W., and H.M. Kunishi. 1971. Phosphate equilibria on stream sediment and soil in a watershed draining an agricultural region. J. Agr. Food Chem. 19:827-831. (79) Williams, J.R., C.A. Jones, and P.T. Dyke. 1984. A modeling approach to determining the relationship between erosion and productivity. Trans. ASAE 27(1):129-144. (80) Young, R.A., A.E. Oldness, C.K. Mutchler, W.C. Moldenhauer. 1986. Chemical and physical enrichments of sediment from cropland. Trans. ASAE 29:165-169. (81) Younge, OR., and D.L. Plucknett. 1966. Quenching the high phosphorus fixation of Hawaiian latosols. Soil Sci. Soc. Am. Proc. 30:653-655.