7
Phosphorus in the Soil-Crop System

Phosphorus is an essential plant nutrient and a necessary input for acceptable crop yields. The beneficial effect of phosphorus on crop yields has been known for well over a century (Kamprath and Watson, 1980). Viets (1975) estimated that one-third to one-half of modern yields are attributable to fertilizer additions and that maintenance of present production levels without fertilizer would require a 20 to 29 percent increase in the area of cultivated land.

However, when phosphorus enters surface waters in substantial amounts it becomes a pollutant, contributing to the excessive growth of algae and other aquatic vegetation and, thus, to the accelerated eutrophication of lakes and reservoirs. (Eutrophication is the process by which a body of water becomes, either naturally or by pollution, rich in dissolved nutrients and, often, seasonally deficient in dissolved oxygen.) Development of strategies to reduce phosphorus loadings to surface water requires an understanding of phosphorus inputs and outputs and the transport mechanisms that deliver phosphorus to surface water. Simplistic solutions may exacerbate trade-offs. Simply eliminating phosphorus additions might bring marginal lands into production, increasing the amount of erosion on such lands (Sharpley and Menzel, 1987). Practices to reduce phosphorus loadings must be based on an understanding of phosphorus sources, a balance between inputs and outputs, and transport processes.

THE PROBLEM OF PHOSPHORUS DELIVERY TO SURFACE WATERS

Excessive nutrient loads in surface water bodies lead to accelerated eutrophication. Algal blooms are one result of accelerated eutrophication



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture 7 Phosphorus in the Soil-Crop System Phosphorus is an essential plant nutrient and a necessary input for acceptable crop yields. The beneficial effect of phosphorus on crop yields has been known for well over a century (Kamprath and Watson, 1980). Viets (1975) estimated that one-third to one-half of modern yields are attributable to fertilizer additions and that maintenance of present production levels without fertilizer would require a 20 to 29 percent increase in the area of cultivated land. However, when phosphorus enters surface waters in substantial amounts it becomes a pollutant, contributing to the excessive growth of algae and other aquatic vegetation and, thus, to the accelerated eutrophication of lakes and reservoirs. (Eutrophication is the process by which a body of water becomes, either naturally or by pollution, rich in dissolved nutrients and, often, seasonally deficient in dissolved oxygen.) Development of strategies to reduce phosphorus loadings to surface water requires an understanding of phosphorus inputs and outputs and the transport mechanisms that deliver phosphorus to surface water. Simplistic solutions may exacerbate trade-offs. Simply eliminating phosphorus additions might bring marginal lands into production, increasing the amount of erosion on such lands (Sharpley and Menzel, 1987). Practices to reduce phosphorus loadings must be based on an understanding of phosphorus sources, a balance between inputs and outputs, and transport processes. THE PROBLEM OF PHOSPHORUS DELIVERY TO SURFACE WATERS Excessive nutrient loads in surface water bodies lead to accelerated eutrophication. Algal blooms are one result of accelerated eutrophication

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture and can result in oxygen depletion, fish kills, and other water quality problems. Phosphorus is most often the limiting nutrient in freshwater aquatic systems and was thought to be the major contributor to nuisance algal blooms in Wisconsin lakes in the late 1940s (Sawyer, 1947). In a variety of Japanese and U.S. lakes, Dillon and Rigler (1974) found a consistent relationship between the phosphorus concentration in the water and the size of the algal standing crop. It is not clear whether this phosphorus limitation is universal. Schindler (1977) maintains, however, that all freshwater lakes will eventually be phosphorus limited because other nutrients have an atmospheric pathway in their biogeochemical cycles and are thus more subject to internal regulation, whereas phosphorus cycling is strictly geologic and thus more sensitive to external factors. Relatively low concentrations of phosphorus in surface waters may create eutrophication problems. Sawyer (1947) estimated a critical level of 0.01 mg of soluble inorganic phosphorus per liter (0.01 parts per million [ppm]); other investigators have not been as ready to assign specific critical levels (Viets, 1975), although a range of 0.01 to 0.03 mg/liter (0.01 to 0.03 ppm) seems to be accepted (Baker et al., 1978). SOURCES OF PHOSPHORUS Phosphorus can enter surface water from a variety of sources including municipal wastes, industrial wastes, animal feedlots, and runoff from croplands. Point Sources Point sources of pollutants, such as municipal wastewater treatment facilities or industrial wastewater outlets, were formerly the major sources of phosphorus input to surface waters, with agricultural and other diffuse or nonpoint sources playing a relatively minor role (Bjork, 1972; Sawyer, 1947). In nonindustrialized countries where sewage treatment is limited, this dominance of point sources is still the case (Gilliam et al., 1985); but in the United States and Canada, nonpoint sources are increasingly important because of more effective point source control. Overall trends for U.S. rivers indicate that there are about equal numbers with increasing and decreasing phosphorus loads. In general, the decreases are linked to point source reductions, whereas the increases appear to be due to nonpoint source increases (R. A. Smith et al., 1987). The increases in total phosphorus loads were associated with

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture increased suspended sediment loads and with some measures of agricultural land use, such as the proportion of fertilized land and cattle population density. By 1978, about 45 to 50 percent of the total phosphorus load to the Great Lakes was attributed to diffuse sources, primarily agricultural activities (Groszyk, 1978; Johnson, 1978). In 1979, an estimated 28 to 40 percent reduction in the diffuse phosphorus load to Lake Erie was required to meet water quality goals (Logan et al., 1979). Agricultural Sources of Phosphorus The potential for phosphorus delivery to surface waters varies widely among different agricultural practices, and cost-effective solutions should target the systems with the greatest potential phosphorus delivery reductions per dollar spent on control measures. The general categories of agricultural phosphorus sources are croplands, lands in pasture or forage crops, and livestock wastes. Most of the phosphorus load to surface waters is due to row crops, particularly on fine-textured soils near watercourses (Groszyk, 1978). In one intensive study in Canada (Coote et al., 1978), soil clay content and the area of a watershed that was in row crops were two of the most important variables explaining the total phosphorus load in the watershed. Most of the phosphorus lost from croplands is not in solution but is bound to eroded soil particles. Sediment-bound phosphorus is not 100 percent available for plant uptake, but sediment control in itself is desirable from a number of standpoints, including the fact that many of the pesticides lost from fields are sediment-bound (Johnson, 1978). Sediment and total phosphorus loads from pasturelands are generally lower than those from croplands, but more of the phosphorus lost is in the more available dissolved form (Baker et al., 1978). This result has been ascribed to lack of fertilizer incorporation and leaching of phosphorus from foliage and animal wastes on pastures (Baker et al., 1978; Viets, 1975). Manure from livestock waste disposal may be a significant source of phosphorus loads in water; one estimate (Moore et al., 1978) is that about 5 percent of the phosphorus excreted by livestock annually ends up in surface waters. If manure is spread on frozen ground, losses of phosphorus through runoff from manure may be severe. In the Great Lakes region, 30 to 38 percent of the total livestock waste phosphorus load is lost through runoff from manure on frozen ground (Moore et al., 1978). Moore and colleagues (1978) found that most of the rest of the

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture phosphorus load from wastes (44 to 50 percent) in water was due to runoff from dairy cattle operations. In other regions, such as the southeastern United States, swine waste is a potentially large source of phosphorus (K. R. Reddy et al., 1978). In some cases, reduction of sediment phosphorus losses can result in increases of soluble phosphorus loss (Sharpley and Menzel, 1987), so the answer to phosphorus loading problems is not as simple as sediment control and is likely to involve trade-offs. Forms and Bioavailability of Phosphorus Phosphorus occurs in many forms in both the solution phase and, in particular, the solid phase. These forms are little understood, even though there are many data in the literature concerning the chemistry of phosphorus in water, soils, and sediments. The relative bioavailability of various forms of phosphorus varies, but there is no standard method of determining this important quantity. Soluble Phosphorus Soluble phosphorus is arbitrarily defined as phosphorus that will pass through a 0.45-µm-pore filter. Soluble reactive phosphorus is that fraction of phosphorus that is reactive with molybdate, according to the Murphy-Riley procedure or its variants. This fraction has been assumed to consist of orthophosphate, but there is evidence that some organic phosphorus is included (Rigler, 1968); for this reason, molybdate-reactive phosphorus is usually referred to as soluble reactive phosphorus or dissolved reactive phosphorus rather than orthophosphate. Not all of the dissolved reactive phosphorus in lake water is completely available for algal growth (Sharpley and Menzel, 1987). The relative difference in dissolved reactive phosphorus and bioavailable phosphorus in water is greater in waters with low levels of phosphorus and is less in solutions with higher dissolved reactive phosphorus concentrations (Sharpley and Menzel, 1987). Even though not all of the phosphorus in water is available to algae, there is often a close relationship between the total amount of phosphorus in water and the standing algal crop (Dillon and Rigler, 1974). Particulate Phosphorus Phosphorus is strongly bound to sediments by anion adsorption reactions. These reactions probably account for the rapid removal from

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture water of phosphorus that is in contact with lake sediments (Syers et al., 1973). Much of this adsorbed phosphorus is not easily desorbed, and the amount that is desorbable decreases with the age of the sediment-adsorbed phosphorus complex (Syers et al., 1973). Particulate phosphorus is associated with iron, aluminum, and manganese in sediments (Bortleson and Lee, 1974; McCallister and Logan, 1978; Syers et al., 1973), although the association with manganese may be artifactual because of the coprecipitation of manganese and iron in nodules (Syers et al., 1973). In this regard, iron seems to be most commonly associated with phosphorus, aluminum and manganese are less so, and calcium carbonate is not commonly associated with phosphorus (Syers et al., 1973). The fraction of iron phosphorus seems to be associated with is the oxalate-extractable fraction, referred to as short-range-order or amorphous oxides. Oxalate extraction reduces or eliminates the phosphorus sorption capacity (Syers et al., 1973). For the most part, discrete phosphorus compounds have not been found in lake sediments, although there are exceptions (Syers et al., 1973). Most emphasis has been on the phosphorus fractions removed by a number of extractants that remove phosphorus that is more or less tightly bound. The nonspecificities of extractants for phosphorus removal and potential reprecipitation of phosphorus make this work difficult to interpret (Syers et al., 1973). Estimates of the fraction of sediment-bound phosphorus that is available for biological uptake vary according to the methods used to obtain the estimate, and the estimates obtained by different methods are difficult to interpret, making some standard means of obtaining bioavailability estimates desirable (Sharpley and Menzel, 1987). The bioavailability of phosphorus in sediments as measured by a variety of methods usually does not exceed 60 percent of the total phosphorus in the sediment (Sonzogni et al., 1982) and varies with the source of the sediment (Logan et al., 1979). A bioassay that measures phosphorus uptake by algae is the standard by which most chemical extractants are measured. The various assay methods include exchange with a hydroxy-aluminum-coated resin or phosphorus-32 or extraction with ammonium fluoride, sodium hydroxide, or nitriloacetic acid (Sharpley and Menzel, 1987). Opinions vary, but the best chemical extractant for measuring bioavailable phosphorus seems to be 0.01 M sodium hydroxide (Dorich et al., 1985; Sharpley et al., 1991; Williams et al., 1980), even though none of the chemical extractants appears to remove the specific fraction of phosphorus removed in an algal assay (Dorich et al., 1980). The relevance of the algal assay to total potentially bioavailable

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture phosphorus may be questioned for several reasons. The phosphorus uptake mechanisms of algae may be different from those of rooted aquatic plants, and in many cases, rooted aquatic plants are a more serious consequence of eutrophication than algal blooms (Sharpley and Menzel, 1987). In terrestrial systems, some plants can solubilize phosphorus from sources usually considered to be unavailable to plants (Jayman and Sivasubramaniam, 1975). Similar mechanisms may exist in aquatic systems. Use of phosphorus sediments by rooted plants also releases the phosphorus bound in sediments to the water column. In one study (Carignan and Kalff, 1980), rooted aquatic plants derived 72 to 100 percent of their phosphorus nutrition from sediments, making them potential nutrient pumps to the open water. Algal assays also do not account for possible phosphorus release when sediments are subjected to anoxic conditions. Anoxic conditions cause the release of phosphorus, which is thought to be due to the reduction of iron ions (Fe3+ to Fe2+) (Sharpley and Menzel, 1987; Syers et al., 1973). The amount of phosphorus in solution may increase manyfold when the sediments are subjected to reducing conditions (Mortimer, 1940, 1941). Regardless of the validity of assay techniques, the availability of sediments for nutrient exchange with the overlying water is important. Physical mechanisms such as the rate of settling of phosphorus-containing particles affect the availability of sediments, as does the thickness of the sediment layer that interacts with the overlying water. This layer may be only a few millimeters or up to a few centimeters thick (Sharpley and Menzel, 1987) and can be affected by physical mixing or aquatic organisms burrowing in the sediment (McCall et al., 1979). One study in a shallow, well-mixed area of a lake noted significant reductions in the sediment phosphorus concentrations in association with spring algal blooms (Wildung et al., 1974). Phosphorus associated with sediments may remain a problem years after excess phosphorus inputs cease. Lake Trummen in Sweden experienced nuisance algal blooms 10 years after nutrient inputs were reduced. The algal blooms were eliminated only after removal of the enriched sediments (Bjork, 1972). Desorption of phosphorus from sediments is estimated to contribute about 10 percent of the total phosphorus load to Lake Erie (Sharpley and Menzel, 1987). Total Phosphorus Although soluble and particulate phosphorus are discussed separately, they are closely related. The equilibrium concentration of soluble

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture phosphorus is controlled by the concentration of sediment-bound phosphorus. Furthermore, both soluble and particulate phosphorus contribute to water quality problems. A focus solely on reductions in soluble or particulate phosphorus can lead to trade-offs because, in some cases, reductions in sediment-bound phosphorus losses can result in increased soluble phosphorus losses (Sharpley and Menzel, 1987). Total bioavailable phosphorus is a more useful measure of phosphorus loadings, but it can be difficult to estimate because of the problems with estimating bioavailability discussed earlier. Total phosphorus can serve as a useful proxy for total bioavailable phosphorus, and reductions in total phosphorus loadings should be the goal of phosphorus control programs. PHOSPHORUS IN THE SOIL-CROP SYSTEM Like nitrogen and other plant nutrients, the phosphorus added to the soil-crop system goes through a series of transformations as it cycles through plants, animals, microbes, soil organic matter, and the soil mineral fraction. Unlike nitrogen, however, most phosphorus is tightly bound in the soil, and only a small fraction of the total phosphorus found in the soil is available to crop plants. The Phosphorus Cycle Figure 7-1 is a simplified illustration of the phosphorus cycle in the soil-crop system. Most of the phosphorus in soil is found as a complex mixture of mineral and organic materials. Organic phosphorus compounds in plant residues, manures, and other organic materials are broken down through the action of soil microbes. Some of the organic phosphorus can be released into the soil solution as phosphate ions that are immediately available to plants. Much of the organic phosphorus is taken up by the microbes themselves. As microbes die, the phosphorus held in their cells is released into the soil. A considerable amount of organic phosphorus is held in the humic materials that make up soil organic matter. A portion of this organic phosphorus is released each year as these humic materials decay. The phosphate ions released from the decomposition of organic phosphorus compounds or added directly in inorganic phosphorus-containing fertilizers readily react with soil minerals and are immobilized in forms that are unavailable to plants for growth. Phosphorus retention in soils is generally considered to be due to adsorption, although some evidence of direct phosphorus precipitation from solution exists (Martin et al., 1988).

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture FIGURE 7-1 The phosphorus cycle. Source: H. O. Buckman and N. C. Brady. 1969. The Nature and Properties of Soils, 7th Ed. London: Macmillan. Reprinted with permission from © Macmillan Inc. Aluminum appears to be more widely involved than sediments in phosphorus sorption in soils. Amorphous aluminum compounds (Kawai, 1980), peat-aluminum complexes (Bloom, 1981), and aluminum-substituted goethite (an iron hydrogen oxide) (Karim and Adams, 1984) have been implicated in phosphorus adsorption in soils. Phosphorus adsorption is related to iron oxides as well, but the more crystalline forms of the oxides (citrate-dithionite-bicarbonate [CDB] or CDB-extractable iron) (Karim and Adams, 1984; Solis and Torrent, 1989a) rather than the amorphous (oxalate-extractable) forms that are important for sediments appear to be involved. CDB-extractable iron and phosphorus are correlated (Solis and Torrent, 1989b), and by direct observation phosphorus has been seen to be enriched in some iron-enriched nodules (McKeague, 1981). Phosphorus is relatively enriched in finer soil fractions, so it is perhaps expected that phosphorus adsorption is correlated with the clay content of soils (Solis and Torrent, 1989b) and with the soil surface area (Olsen and Watanabe, 1957). Some organic acids reduce the adsorption capacity of phosphorus, perhaps by competition for anion adsorption sites (Kafkafi et al., 1988; Lopez-Hernandez et al., 1986; Violante et al., 1991). Solubilization of phosphorus and aluminum by organic acids has been noted (Fox et al., 1990a,b; Jayman and Sivasubramaniam, 1975) and appears to be related

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture TABLE 7-1 Phosphorus Inputs and Outputs in the United States, 1987 Input or Output Metric Tons of Phosphorus Inputs Fertilizer-P 3,570,000 (79) Manure-P 655,000 (15) Crop residues 272,000 (6) Total inputs 4,500,000 (100) Outputs Harvested crops 1,320,000 (29) Crop residues 272,000 (6) Total outputs 1,600,000 (36) Balance 2,900,000 (63) NOTE: Values in parentheses are percentage of total phosphorus input mass. See the Appendix for a full discussion of the methods used to estimate phosphorus inputs and outputs. to the stability of the aluminum-organic acid complex (Fox et al., 1990a,b). Under anaerobic conditions, phosphorus release to solutions low in phosphorus concentration is increased, the adsorption capacity of the anaerobic soil increases when the solution has a high phosphorus concentration. This increase is thought to be due to the increased surface areas of the reduced iron oxides (Khalid et al., 1977; Patrick and Khalid, 1974). Phosphate ions added to soils from either organic or inorganic sources enter into this complex series of precipitation or sorption reactions. These reactions greatly reduce the amount of phosphate ions that are in the soil solution and available to plants. The equilibrium level of dissolved phosphorus in the soil solution is controlled by the chemical environment of the soil (Nelson and Logan, 1983). Mass Balance Phosphorus is added to croplands in crop residues and manures, in synthetic fertilizers, and from phosphorus-bearing minerals in the soil (Figure 7-1). Part of the phosphorus entering the soil-crop system is removed with the harvested crop; the balance is immobilized in the soil, incorporated into soil organic matter, or lost in surface or subsurface flows to surface water or groundwater. Table 7-1 provides estimates of national phosphorus mass balances as the mass of phosphorus applied to croplands as synthetic fertilizers

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture (fertilizer-P) and crop residues and voided in manures (manure-P). (See the Appendix for a full explanation of mass balance estimates.) Much of the total mass of phosphorus voided in manures is not economically recoverable for use as an input in annual crop production systems because it is deposited on pasturelands or rangelands, from which collection is impossible. Furthermore, a substantial portion of the phosphorus voided in manures is lost in surface runoffs from pasturelands, rangelands, and handling and storage facilities. Only that portion of total phosphorus voided in manures that can be economically recovered for use on croplands was used in Table 7-1 as an estimate of phosphorus inputs from manure. The difference between phosphorus inputs and phosphorus outputs in crops and crop residues is reported as phosphorus balances. A more detailed analysis of phosphorus inputs and outputs to croplands helps to identify opportunities for reducing phosphorus loadings to surface water from farming systems. Phosphorus Inputs The phosphorus in fertilizer-P is the single most important source of phosphorus added to croplands in the United States (Table 7-1). The majority of this fertilizer-P was added to annual crops, and the amount of phosphorus added in synthetic fertilizers varies from crop to crop and region to region. Corn consumes more phosphorus than any other single crop, followed by wheat, soybeans, and cotton, in that order. Phosphorus application rates also vary between crops, with corn again receiving the greatest rates of phosphorus application per unit area; this is followed by cotton, soybeans, and wheat, in that order. Of the phosphorus applied to croplands in the United States, 42 percent is applied to land planted in corn, and 67 percent of the total phosphorus applied to U.S. croplands is planted in four crops: corn, cotton, soybeans, and wheat. These differences in phosphorus application rates combine with regional differences in crop mixes to produce the state-to-state variability in the total amount of phosphorus applied in synthetic fertilizers (Tables 7-2 and 7-3). The amount of recoverable manure-P is small compared with that supplied in synthetic fertilizers at the national level (Table 7-1). The phosphorus in manure represents only 15 percent of phosphorus inputs. The total mass of phosphorus voided in manures is much larger than that which is economically recoverable. The recoverable phosphorus represents less than half of the total mass of phosphorus in manure. Locally, the proportion of phosphorus supplied by manures can be

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture large. Recoverable phosphorus in manure, for example, supplies 65 percent of total phosphorus inputs in Vermont (Table 7-3). Phosphorus Outputs The fraction of total phosphorus inputs lost to erosion and runoff can be substantial, but it is difficult to estimate the amount. Larson and colleagues (1983) estimated that 1.74 million metric tons (1.92 million tons) of phosphorus, or about 50 percent of the estimated total phosphorus balance in Table 7-1, was lost in eroded sediments in 1982. Additional phosphorus can be lost in solution (see below). The importance of animal manures as a potential source of phosphorus loadings can be seen in the difference between total phosphorus and that which is recoverable in manure. The 662,000 metric tons (730,000 tons) of recoverable phosphorus accounted for in Table 7-2 represents only 49 percent of the total estimated 1,349,000 metric tons (1,487,000 tons) of phosphorus excreted in animal manures. A substantial fraction of this difference between the total amount of phosphorus excreted in manures and the amount that can be recovered for use in crop production may represent direct losses of phosphorus in runoffs from pastures, feedlots, and manure storage facilities. The majority of the total and recoverable phosphorus balance on agricultural lands is immobilized in either the mineral or the organic fractions of the soil. The potential for buildup of phosphorus levels in soil over time has important implications for efforts to reduce phosphorus loadings to water (see below). Phosphorus Buildup in Soils Relatively small annual additions of phosphorus may cause a soil buildup of phosphorus as illustrated in (Figure 7-2) (McCollum, 1991). Some of the phosphorus added in excess of crop needs remains as residual plant-available phosphorus, but not all of the added phosphorus will be available to crops; the amount of extractable phosphorus declines with time because of the slow conversion of phosphorus to unavailable forms (McCollum, 1991; Mendoza and Barrow, 1987; Sharpley et al., 1989; Yost et al., 1981). The rate of decline in extractable phosphorus (discounting plant uptake) varies with the phosphorus adsorption properties of the soil and the initial level of phosphorus in the soil (that is, the relative saturation of adsorption capacity) and with the amount of applied phosphorus. The phosphorus level in the soil is the critical factor in determining

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture negative log-log relationship between ER and soil loss (Massey and Jackson, 1952; Sharpley, 1985). Several other findings regarding ERs are notable. The ER is relatively greater for bioavailable forms of phosphorus than it is for less available forms (Sharpley, 1985), and ERs increase with increasing additions of fertilizer (Sharpley, 1980). Changes During Transport During transport from agricultural fields to streams and lakes, various changes in the forms of phosphorus are likely to occur (Sharpley and Menzel, 1987). The changes depend on, for example, the relative phosphorus concentrations in water and sediment and EPC. There is an inverse linear correlation between the soluble phosphorus in stream water and the logarithm of the sediment concentration in runoff (Sharpley and Menzel, 1987). The general trend is for phosphorus to be converted to less available forms during transport from a field's edge to a lake (Sharpley and Menzel, 1987). POSSIBLE MANAGEMENT METHODS FOR PHOSPHORUS LOSS REDUCTION There are two primary ways to reduce the amount of phosphorus lost from agricultural production: reduce phosphorus levels in the soil and reduce erosion and runoff from croplands. Procedures to Establish Threshold Levels The level of phosphorus in soil (soil-P) is an important determinant of the amount of phosphorus lost to surface water and groundwater from cropping systems. Higher soil-P levels lead to increased losses of both soluble and particulate phosphorus. Phosphorus management to reduce unnecessarily high soil-P levels should be part of efforts to reduce phosphorus losses from cropping systems. Phosphorus applications in excess of that harvested in the crop leads to the buildup of soil-P. In addition, some soils have naturally high levels of phosphorus, and the addition of phosphorus to these soils can exacerbate already high levels of soil-P. Phosphorus applications at levels that lead to a buildup in soil have been encouraged, in part, because of difficulty in predicting crop responses to phosphorus applications. Although the response of crops to phosphorus additions has been known for well over a century and attempts to define the crop-available

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture soil-P date at least from 1894, there is still no universally useful extractant for soil-P, and crop responses to recommended phosphorus applications are erratic. Crop yields and/or cumulative phosphorus uptake are often not predicted well by soil tests (Prabhakaran Nair and Mengel, 1984; Yang and Jacobsen, 1990; Yerokun and Christenson, 1990). Kamprath and Watson (1980) note that there is a problem in the interpretation of soil test results as a result of various soil-P buffer capacities, usually referred to as quantity-intensity relationships; that is, soils vary in their ability to replenish soil solution phosphorus when it is depleted by plant uptake. For example, in North Carolina, a threefold greater phosphorus level, as determined by soil tests, is needed to supply a crop's needs in a sandy soil than is needed in a finer-textured soil (Cox et al., 1981; Kamprath and Watson, 1980). This difference in a soil's capacity to supply phosphorus to a growing crop is related to the soil's phosphorus adsorption capacity. Some findings that may be of use in improving recommendations for phosphorus applications are summarized below. The texture effect noted above occurred in a greenhouse study of phosphorus test predictions (Lins and Cox, 1989). In that study, the clay content and surface area of soil were the variables that best improved phosphorus soil test yield predictions. As noted above, the clay content and surface area of soil are correlated with phosphorus adsorption capacity. Another study (Kuo, 1990) found that the best variable that can be used to predict plant phosphorus uptake by several soils that vary in their amorphous aluminum contents was the fraction of phosphorus adsorption sites in soil that were occupied. As in the previous study (Lins and Cox, 1989), the variables related to the soil's phosphorus adsorption capacity improved the accuracy of the predictions. It may be useful to include some measure of a soil's phosphorus buffer capacity in routine soil tests. Recommendations for fertilizer use include a safety factor to compensate for the uncertainty of predictions of crop responses to phosphorus. Because of either a history of phosphorus applications in excess of that harvested or naturally high soil-P levels, or both, soil-P levels have increased in many U.S. soils (Thomas, 1989), and many now have high levels of phosphorus. Table 7-4 lists the percentage of soil tests in each state reading high to very high or medium or less for phosphorus. Soil-P is often at levels above which a crop yield increase from additional phosphorus would be predicted (McCollum, 1991; Novais and Kamprath, 1978; Yerokun and Christenson, 1990). Mallarino and colleagues (1991) cited several studies reporting that increases in soybean or corn yields are small or nonexistent when soil test levels for

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture phosphorus are within the medium category (Grove et al., 1987; Hanway et al., 1962; Million et al., 1989; Obreza and Rhoads, 1988; Olson et al., 1962; Rehm, 1986). Phosphorus additions to soils with high soil-P test results should not produce increased corn and soybean yields in the Corn Belt (Bharati et al., 1986; Hanway et al., 1962; Olson et al., 1962; Rehm, 1986). This phenomenon suggests that applications of additional phosphorus to 56, 63, 78, 68, and 35 percent of the soils tested in Iowa, Illinois, Indiana, Ohio, and Missouri, respectively, would be expected to produce no increase in yields. Similar situations exist in the southeastern United States. Kamprath (1967, 1989) and McCollum (1991) have shown that corn and soybeans grown on Piedmont and Coastal Plain soils testing high in available phosphorus do not respond to phosphorus fertilizer additions. On the basis of the soil test data presented in Table 7-4, no response to phosphorus would be expected on approximately half of the soils in the southeastern United States. In North Carolina, phosphorus recommendations for soybeans grown on soils testing medium for phosphorus are greater than the amount of phosphorus removed in the grain (Kamprath, 1989). Thus, current recommendations will lead to soil-P levels greater than those needed for corn or soybean production. The rate of increase in soil-P with fertilizer additions is either linear or quadratic (Figure 7-2) (Cox et al., 1981; McCollum, 1991), and the rate of decrease is exponential. The rate constant is soil dependent (Cox et al., 1981) and increases with higher initial soil-P levels (McCollum, 1991). The decline approximates the kinetics of a first-order chemical reaction (McCollum, 1991). The increasing loss rate with increasing initial level essentially means that overfertilization is a waste of money because the phosphorus is converted to unavailable forms. That is, doubling of the initial phosphorus application rate does not double the residual phosphorus effect (McCollum, 1991). Mallarino and colleagues (1991) studied the effect on yields of phosphorus additions to a soil testing high for phosphorus. They reported occasional positive yield responses to fertilization, but these positive responses were not, in most cases, sufficient to pay for the cost of the added phosphorus. In the 11 years of the study, phosphorus applications to this soil that tested high for phosphorus showed appreciable positive economic returns in only 1 year for corn and for no year for soybeans. The addition of phosphorus resulted in negative returns in most years for both corn and soybeans, with losses in 1 year being greater than $49/ha ($20/acre) for corn (Figure 7-3). Some of the phosphorus added in excess of crop needs remains as residual plant-available phosphorus, but not all of the added phosphorus

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture TABLE 7-4 Soil Tests Reporting Very Low to Medium or High to Very High for Soil-P (percent) State Very Low to Medium High to Very High Alabama 65 35 Arizona 49 51 Arkansas 86 14 California 59 41 Colorado 57 43 Connecticut 49 51 Delaware 35 65 Florida 55 45 Georgia 62 38 Idaho 40 60 Illinois 37 63 Indiana 22 78 Iowa 44 56 Kansas 61 39 Kentucky 58 42 Louisiana 63 37 Maine 49 51 Maryland 36 74 Massachusetts – – Michigan 27 73 Minnesota 24 76 Mississippi 66 34 Missouri 65 35 Montana 59 41 Nebraska 69 31 Nevada 52 48 New Hampshire – – New Jersey – – New Mexico – – New York 62 38 North Carolina 33 67 North Dakota 70 30 Ohio 32 68 Oklahoma 52 48 Oregon 51 49 Pennsylvania 56 44 Rhode Island – – South Carolina 40 60 South Dakota 56 44 Tennessee 51 49 Texas 63 37 Utah 40 60 Vermont 75 25 Virginia 42 58 Washington 46 54 West Virginia – – Wisconsin 34 66 Wyoming 62 38 NOTE: Dashes indicate no data reported. SOURCE: Adapted from Potash and Phosphate Institute. 1990. Soil test summaries: Phosphorus, potassium, and pH. Better Crops with Plant Food 74(2):16-18.

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture FIGURE 7-3 Economic returns on investments of annual applications of phosphorus (P) fertilizers. Source: A. P. Mallarino, J. R. Webb, and A. M. Blackmer. 1991. Corn and soybean yields during 11 years of phosphorus and potassium fertilization on a high-testing soil. Journal of Production Agriculture 4:312–17. Reprinted with permission from © American Society for Agronomy, Crop Science Society of America, and Soil Science Society of America.

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture is available to crops because the amount of extractable phosphorus declines with time because of the slow conversion of phosphorus to unavailable forms (McCollum, 1991; Mendoza and Barrow, 1987; Sharpley et al., 1989; Yost et al., 1981). The rate of decline in extractable phosphorus (discounting plant uptake) varies with the soil-P adsorption properties and the initial soil-P level, that is, relative saturation of adsorption capacity, and with the amount of applied phosphorus. Several studies have investigated the buildup of soil-P under continuous phosphorus fertilization conditions (McCallister et al., 1987; Schwab and Kulyingyong, 1989), and others have documented the loss of soil-P under continuous cropping with only residual phosphorus available for crop uptake (Novais and Kamprath, 1978). Both buildup and decline phases have been studied as well (Cope, 1981; McCollum, 1991; Meek et al., 1982), but relatively few studies (Cope, 1981; McCollum, 1991) have been conducted over long time spans (several decades). Results of these few studies may provide some of the best information that can be used to aid in predicting residual phosphorus effects and actual phosphorus fertilization needs. Many soils can be cropped for a decade or more without the soil-P reaching a level at which fertilizer additions would result in a crop yield increase (Figure 7-4). A crop yield increase would not be expected until soil-P levels fall below 22 g/m3, which would occur only after several years of cropping, depending on the initial soil-P level. The data and studies available suggest that the amount of phosphorus added to cropping systems could be reduced without decreasing crop yields on a significant portion of the nation's croplands. In those soils testing high for soil-P, phosphorus applications other than for starter fertilizer could be suspended without yield losses, depending on the soil, crop, and climate. Despite weaknesses in the ability to predict crop responses to phosphorus applications, most states have soil test procedures that, although not perfect, can be used to establish the threshold levels of soil-P beyond which no crop response is predicted. Given the importance of reducing soil-P levels for reducing phosphorus losses to surface waters, such thresholds should be established. Applications of additional phosphorus, except for small starter applications, should be discouraged once that threshold level of soil-P is reached. In extreme cases, when damages to surface water are great, a second threshold level of soil-P—beyond which no additional phosphorus should be applied—should also be established. Once established, such threshold values should become a routine part of phosphorus application recommendations supplied by public and private organizations.

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture FIGURE 7-4 Decrease of soil-P over time, measured as Mehlich 1-extractable phosphorus, on Portsmouth soil during the residual phase. A, initial soil-P (Pi) at year 0 established by one broadcast application; B, Pi at year 0 is the result of 8 previous years of active buildup via annual, banded applications. Source: R. E. McCollum. 1991. Buildup and decline in soil phosphorus: 30-year trends on a Typic Umprabuult. Agronomy Journal 83:77–85. Reprinted with permission from © American Society for Agronomy, Crop Science Society of America, and Soil Science Society of America.

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture Methods for Reducing Erosion and Runoff The most studied methods for reducing phosphorus losses from fields involve a variety of reduced-tillage methods. There is usually a trade-off between sediment and solution phosphorus losses, and the chosen management method should be one that reduces the total bioavailable phosphorus load in runoffs (Sonzogni et al., 1982). A number of reduced-tillage, no-tillage, and other alternative tillage systems have been tried; the general result is that total phosphorus loss is less with any system that reduces soil exposure (Andraski et al., 1985; McDowell and McGregor, 1984; Romkens et al., 1973; Sharpley and Menzel, 1987). Concentrations of soluble phosphorus in runoffs are often higher, and total losses of soluble phosphorus are sometimes higher. The amount of bioavailable phosphorus, which should be the true measure of a system's effectiveness (Sonzogni et al., 1982), is not always measured, but it generally tends to be lower in reduced-tillage systems (Andraski et al., 1985). If manure rather than chemical fertilizer is applied, the advantage of reduced-tillage may be negated because of the leaching of phosphorus from the unincorporated manure, and bioavailable phosphorus losses may be greater than those from conventional tillage systems (Mueller et al., 1984). Other potential adverse effects of reduced-tillage practices include the stratification of phosphorus near the surface because of a lack of incorporation and the cycling of phosphorus to the surface by plants (Mackay et al., 1987). This phenomenon may increase surface losses of phosphorus because of the increased amount of available phosphorus at the surface and may also restrict nutrient uptake and rooting to surface layers, particularly early in the season. Reduced-tillage systems or other systems that increase soil cover and that effectively reduce both runoff and soil erosion generally reduce total losses of bioavailable phosphorus to surface waters. Reduced-tillage systems do, however, appear to increase the proportion of bioavailable phosphorus lost in soluble form. The effectiveness of reduced-tillage systems in reducing bioavailable phosphorus losses would be increased if parallel efforts were undertaken to reduce phosphorus concentrations in surface soils. Effective efforts to reduce phosphorus loadings to surface water should simultaneously reduce soil-P levels, erosion, and runoff. Efforts to reduce any of those three factors without reducing the others may exacerbate trade-offs between soluble and particulate phosphorus loadings. A number of the findings summarized above, as well as a number of others not mentioned, have been incorporated into models that predict soil-P adsorption properties, crop responses to fertilizer-P, and surface losses of phosphorus. These models and their supporting data have

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture been developed by Sharpley and coworkers over a number of years. The models are empirical rather than mechanistic, but they appear to provide useful predictions of phosphorus losses (Sharpley et al., 1991; Smith et al., 1991). These models may be helpful in choosing which of several phosphorus loss management methods will be most effective. Buffer Strips Buffer strips may be helpful in reducing both the particulate and soluble phosphorus fractions, especially when vegetation or crop residues are present. Nutrient load reductions of more than 70 percent have been achieved with several types of buffer strip-surface cover combinations (Alberts et al., 1981; Thompson et al., 1978). Under high flow conditions, the efficiency of the buffer strips diminishes and phosphorus losses may be greater than those under control conditions. Even after the runoff leaves the field, nutrient loads, particularly sediment-bound fractions, may be dropped by sedimentation near the field's edge above the watercourse. In one study (Cooper et al., 1987), 50 percent of the total sediment load was deposited within 100 m (109 yards) of the field's edge, and 80 percent was removed from above the creek floodplain. Another study involved a managed distribution system that applied beef feedlot runoff to a small wooded watershed (Pinkowski et al., 1985). More than 99 percent of the input phosphorus was retained in the watershed, although phosphorus concentrations and total losses increased relative to those under the baseline conditions. Also, there was a large net loss of nitrogen from the watershed, and the investigators encountered problems with tree mortality because of excessive soil moisture. Once in the watercourse, nutrient loads may still be reduced. Wetlands have been proposed to act as nutrient filters, but they may be only small sinks for nitrogen and phosphorus and may, in fact, be net exporters of some nutrients (Peverly, 1982). Buffer strips and protection of riparian areas should help to reduce phosphorus loadings to surface waters. These measures, however, cannot substitute for efforts to reduce soil-P levels, runoff, and erosion. Use of buffer strips and protection of riparian areas should increase the effectiveness of programs to reduce phosphorus loadings if they are part of a larger effort. Inclusion of Extreme Weather as Loss Factor All potential management options must be considered in the context of natural events (Sharpley and Menzel, 1987), that can cause large

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture losses of soil and nutrients from even the best managed systems. Intensive individual events may cause greater losses of phosphorus and sediments than those that occur during years of normal runoff. For example, Cahill and colleagues (1978) attributed 86 percent of total phosphorus loss to a 1-month period in late winter. Nelson and colleagues (1978) found that large runoff events contributed 58 to 78 percent of the total sediment phosphorus load over 2 years. Burwell and colleagues (1975) found that the period that included the planting date plus the 2 months that followed caused the greatest sediment and total phosphorus loss for conventionally tilled systems. Schuman and colleagues (1973) found that a few major rainfall events caused 80 percent of the total phosphorus loss. Hubbard and colleagues (1982) found that 64 to 86 percent of total sediment loss was due to one storm. These examples emphasize the fact that no management technique, no matter how well designed and implemented, will be 100 percent effective. Augmenting phosphorus management with buffer zones or using different cropping systems that provide greater soil protection can reduce the damage caused by extreme events. New Cropping Systems Immediate gains in reducing phosphorus loads to surface waters can be accomplished by simultaneous efforts to reduce soil-P levels, erosion, and runoff from current cropping systems. These improvements can be increased by efforts to incorporate buffer strips and protect riparian vegetation to trap phosphorus in the sediments and runoff from current cropping systems. In the long-term, it is unclear whether such improvements will be sufficient to meet water quality goals. Changes in cropping systems, such as the use of cover crops, multiple crops, or other approaches to increasing soil cover and reducing soil-P levels, may be needed to further reduce phosphorus losses from cropping systems. Cover crops, where applicable, appear to hold the promise of substantially reducing phosphorus losses from cropping systems. Sharpley and Smith (1991) reported that the addition of a winter rye cover crop to conventionally tilled corn reduced total phosphorus losses by 70 percent in Georgia. The combination of an alfalfa-timothy hay cover crop with no-tilled corn in Quebec reduced total phosphorus losses by 94 percent from that with conventionally tilled corn with no cover crop. In Alabama, phosphorus losses were 30 percent less in no-tilled than in conventionally tilled corn. A winter wheat cover crop in combination with no-tillage reduced phosphorus losses by 68 percent.

OCR for page 283
Soil and Water Quality: An Agenda for Agriculture This page in the original is blank.