Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
APPENDIX 431 NITROGEN AND PHOSPHORUS MASS BALANCES: METHODS AND INTERPRETATION The flux of nutrients through an agroecosystem is an important determinant of the productivity of the farming system and the potential for water pollution from losses of nitrogen and phosphorus. Mass balances can be used to assess the transformations and transfers that occur in and between components of the farming system and to assess the efficiencies of nutrient use in the system. Often such assessments point out where further study or quantification of transformations or transfers is needed. In most natural ecosystems, nutrient inputs and outputs are limited and most nutrients are cycled and recycled through the system. External inputs and overall exports or losses are minimal, and inputs and outputs remain in relative balance. In modern agricultural systems, however, the amount of external inputs, such as fertilizer, added to the system is very large to enable large outputs of the crops that are harvested for food and fiber. Nutrient budgets and mass balances have been approached on various scales, ranging from experimental plots to estimates of global balances (i.e., Follett et al., 1987; Hauck and Tanji, 1982; Meisinger and Randall, 1991; Power, 1981; Thomas and Gilliam, 1978). Nutrient budgets can be approached at various levels of detail and with varying degrees of completeness. Some budgets even attempt to characterize the inputs, storage, and processing by insects and soil and plant microbiota (Stinner et al., 1984). The general form of a nutrient budget can be viewed, in simple fashion, as a simple equation: the nutrient inputs to the ecosystem minus
APPENDIX 432 the nutrient outputs must equal the change in nutrient storage within the ecosystem. This is deceptively simple because quantifying all of the inputs and outputs is difficult, and even defining the "system" in space and time is problematic. Simplifying assumptions and partial budgets still provide important insights, depending on how the balances are to be used. Partial nutrient balances are used, though often only implicitly, in establishing nutrient- crop yield response models and fertilizer recommendations for crop producers. In farming systems, nutrient budgets can be used to review the balance of major inputs and outputs to assess where the opportunities lie for improvements in efficiency. Although the nature and amount of nutrient inputs and outputs vary among farming systems, regions, and even among fields, the mass balance concept provides a framework that can be applied systematically across a diversity of farming systems, as the field and farm scale balances discussed in Chapters 6 and 7 illustrate. The point of presenting partial balances here is to illustrate that this approach can be used even at the state and regional levels, as part of the analysis of farming systems, to guide program development and targeting to improve input use efficiency. As summarized by Meisinger and Randall (1991), agricultural watersheds with the largest nitrate losses are associated with excess nitrogen inputsâthat is, fertilizer, manure, and legume nitrogen inputs greatly exceed the nitrogen that is taken up by the crop. As Meisinger and Randall (1991:85) note: "These are also the sites where improved N[nitrogen]-management practices will have the greatest chance of improving groundwater quality." ESTIMATION OF STATE, REGIONAL, AND NATIONAL BUDGETS The committee did not attempt to estimate complete mass balances of the nitrogen and phosphorus flux in U.S. agriculture. The committee's intent was to present a partial balance for harvested cropland, focusing on the manageable nitrogen and phosphorus inputs and the resulting outputs of nitrogen and phosphorus in the harvested crops. The purpose of the committee's estimates is to illustrate (1) the opportunities that are available for improving the efficiency with which nitrogen and phosphorus are used in farming systems and (2) large- scale approaches that can be used to target and evaluate programs to improve nutrient use efficiencies across a diversity of farming systems. The committee's estimates do not include inputs or outputs from range-or pastureland or various set-aside or idled lands. Hence, these partial balances take on a different formula than the general form
APPENDIX 433 presented above: major inputs of nitrogen and phosphorus to croplands minus the major outputs of nitrogen and phosphorus in harvested crops equals the balance or residual nitrogen and phosphorus. The balance, therefore, is an estimate of the amount of the nitrogen and phosphorus inputs that cannot be accounted for in the harvested crops. These nitrogen and phosphorus balances may represent (1) storage in the soil or (2) losses from the farming system into the environment. The magnitude of the balance, and the relative magnitude of the inputs, provide insights into the opportunities to improve the management of nutrients. Data Used The committee estimated nitrogen and phosphorus balances for 1987 using data on crop and livestock production from the latest available Census of Agriculture (U.S. Department of Commerce, Bureau of the Census, 1989). The other primary data source was total fertilizer nutrient data by state, compiled by the Tennessee Valley Authority, National Fertilizer Research Center (Hargett and Berry, 1988). In perspective, 1987 is probably a representative, and perhaps a conservative year for calculating nutrient balances. Fertilizer use was down from peak usage and crop acreage for some commodities was below-average because of annual set-asides. Enrollment of cropland into the Conservation Reserve Program was just expanding, and while 1987 was the beginning of a drought period in the midwest, crop yields were still above-average in the midwest and nationally. The methods used to estimate nutrient inputs and outputs are outlined below. Estimation of Inputs The nitrogen and phosphorus inputs estimated by the committee include only the major, primary inputs of nutrients to cropland including the nitrogen and phosphorus in commercial fertilizer, the nitrogen and phosphorus in manure, dinitrogen fixation of nitrogen (and/or nitrogen accumulation) by legumes, and the nitrogen and phosphorus content of crop residues. Estimates of inputs were limited to primary sources since these nitrogen and phosphorus inputs can be directly affected by management and since it was desirable to limit the amount of computation required. There are other input sources, such as nutrients in precipitation and dry deposition, crop seed, foliar absorption, and nonsymbiotic fixation of nitrogen. These are minor or secondary inputs and they are not
APPENDIX 434 typically manageable, seldom measured, and are explicitly ignored in most studies and management systems. There are two other inputs that are important but highly variable, spatially and temporally. The nitrogen in irrigation water, generally available as nitrate, can be an important management consideration locally (see Chapter 6, and Schepers and Mosier, 1991). It was not possible to characterize irrigation water nitrogen inputs for this analysis. Other important inputs are the nitrogen and phosphorus contributed from mineralization from the soil. Although important, this factor is difficult to estimate at the scale of this analysis and has become a relatively small component in many farming systems. Of greater importance, given the need to manage the annual variability of nutrient availability to improve efficiencies, is an assessment of the nutrients available from mineralization plus the nutrients available as residual from inputs in previous years. The nutrients in the balance unaccounted for in the harvested crop (output term) may carry over as inputs in following crop years. The buildup of soil test phosphorus levels, presented in Chapter 7, is an example of the accumulation of phosphorus in the farming system from this balance term over time. Fertilizer Nutrients Estimates of fertilizer inputs are the most reliable of the estimates of inputs and outputs calculated for nitrogen and phosphorus balances. The nutrient inputs from commercial fertilizers were estimated directly from the sales and tonnage records of state agricultural agencies compiled by the Tennessee Valley Authority, National Fertilizer Research Center (Hargett and Berry, 1988, and unpublished data). These data were provided to the committee by the U.S. Geological Survey and had been adjusted for estimates of sales that crossed county and state lines (Fletcher, 1991: personal communication). The values in the adjusted data set generally vary by less than 2 percent from the unadjusted Tennessee Valley Authority compilations. Manure Nutrients The amount of nitrogen and phosphorus in manure applied to cropland was estimated using standard assumptions and values for the nitrogen and phosphorus content of livestock manures developed by the American Society of Agricultural Engineers (see Chapter 11, and American Society of Agricultural Engineers, 1988; Midwest Planning Service, 1985). First, the total amount of nitrogen and phosphorus voided by livestock was estimated by using the following calculation:
APPENDIX 435 TABLE A-1 Factors Used to Estimate Total Nitrogen and Phosphorus Voided in Manures NOTE: NA, not applicable. (total number of animals) * (estimated average weight of the livestock or poultry species/454 kg [1,000 lbs]) * (production, or residence period of the livestock or poultry species, in days) * (kg of nitrogen or phosphorus voided/ day/454 kg of animal weight). The factors used to estimate the amount of nitrogen and phosphorus voided in manures are given in Table A-1. The production period for dairy cattle and hens or pullets was assumed to be 365 days, and the year-end inventory data in the 1987 Census of Agriculture were used to estimate the number of dairy cattle and hens or pullets that were present during 1987. Annual sales data for beef cattle, swine, sheep, and poultry were used rather than inventory numbers. The sales data were adjusted using standard ratios of breeding stock to sales for animal numbers; and standard estimates of the number of days in production were used to estimate the number of beef cattle, swine, and sheep that were present in 1987. Not all manure voided by livestock and poultry are economically recoverable for application to croplands. Manure voided on pasture or rangeland, for example, cannot be collected for use. The committee included estimates of the nitrogen and phosphorus in only those manures that are collectable and recoverable. The proportion of the total nitrogen and phosphorus voided in manures that can be economically recovered varies depending on collection, storage, and application methods. Therefore, the manure nutrient production of each state was adjusted by
APPENDIX 436 TABLE A-2 Nitrogen Voided in Recoverable Manures State Percent Recoverable State Percent Recoverable Manure-N Manure-N Alabama 34 Nebraska 31 Alaska 0 Nevada 14 Arizona 40 New Hampshire 50 Arkansas 41 New Jersey 57 California 36 New Mexico 22 Colorado 35 New York 64 Connecticut 57 North Carolina 39 Delaware 50 North Dakota 22 Florida 21 Ohio 40 Georgia 37 Oklahoma 17 Hawaii 17 Oregon 23 Idaho 31 Pennsylvania 57 Illinois 32 Rhode Island 0 Indiana 33 South Carolina 24 Iowa 32 South Dakota 26 Kansas 33 Tennessee 27 Kentucky 28 Texas 25 Louisiana 18 Utah 32 Maine 69 Vermont 67 Maryland 65 Virginia 36 Massachusetts 67 Washington 33 Michigan 49 West Virginia 33 Minnesota 43 Wisconsin 57 Mississippi 26 Wyoming 20 Missouri 26 Montana 15 United States 34 SOURCE: D. L. Van Dyne and C. B. Gilbertson. 1978. Estimating U.S. Livestock and Poultry Manure Production. Report ESCS-12. Washington, D.C.: U.S. Department of Agriculture, Economics, Statistics, and Cooperative Service. estimates of the proportion of manure that is recoverable for cropland use, and for nitrogen the value was further reduced for storage and handling losses. The state totals were derived from the analysis of Van Dyne and Gilbertson (1978) and are given in Tables A-2 and A-3. Only about one-third of the total nitrogen voided in manures was estimated as recoverable, while about one-half of the total phosphorus voided in manures was estimated as recoverable. Though developed in the 1970s, Van Dyne and Gilbertson's review is the most thorough of its kind, and the basic assumptions have not changed. Using their 1978 values for recoverable manure provided a conservative estimate for 1987 because livestock production had, by then, generally become more concentrated,
APPENDIX 437 with more cattle and swine in confinement operations and less on pasture and range. Hence, a greater proportion of manure would have been collectable in 1987 than in 1974. TABLE A-3 Phosphorus Voided in Recoverable Manures State Percent Recoverable State Percent Recoverable Manure-P Manure-P Alabama 45 Nebraska 46 Alaska 0 Nevada 25 Arizona 45 New Hampshire 100 Arkansas 54 New Jersey 50 California 54 New Mexico 30 Colorado 42 New York 87 Connecticut 100 North Carolina 61 Delaware 100 North Dakota 25 Florida 33 Ohio 64 Georgia 57 Oklahoma 19 Hawaii 50 Oregon 27 Idaho 42 Pennsylvania 79 Illinois 57 Rhode Island 0 Indiana 58 South Carolina 43 Iowa 56 South Dakota 36 Kansas 48 Tennessee 35 Kentucky 38 Texas 31 Louisiana 20 Utah 43 Maine 100 Vermont 100 Maryland 83 Virginia 47 Massachusetts 100 Washington 44 Michigan 69 West Virginia 50 Minnesota 64 Wisconsin 81 Mississippi 39 Wyoming 27 Missouri 42 Montana 20 United States 49 SOURCE: D. L. Van Dyne and C. B. Gilbertson. 1978. Estimating U.S. Livestock and Poultry Manure Production. Report ESCS-12. Washington, D.C.: U.S. Department of Agriculture, Economics, Statistics, and Cooperative Service. Legume Nitrogen The importance of symbiotic dinitrogen fixation has been known, and that knowledge used to enhance crop production, since ancient times. Yet estimates of rates of fixation by legumes (legume-N) vary widely, depending on the species, the age, density, and vigor of the crop, the amount of nitrogen in the soil, and the number of years the legume
APPENDIX 438 stand remains in the field before being turned under. A summary of estimates for various legumes is given in Table A-4. TABLE A-4 Estimates of Nitrogen Fixation by Legumes Nitrogen Fixation Rate (kg/ha/yr) Legume Low Estimate High Estimate Alfalfa 70 600 Soybeans 15 310 Midwest 55 95 Southeast 70 220 Dry Beans 2 215 Peanuts 40 60 Cowpeas 80 100 Chickpeas 25 80 Clover (various) 100 200 Sweet Clover 4 130 Fava Bean 175 200 Lentils 165 190 Lupins 150 215 Peas 55 195 Vetch 90 120 SOURCE: Data derived from Evans and Barber, 1977; Follet et al., 1987; Heichel, 1987; Meisinger and Randall, 1991; Peterson and Russelle, 1991; Schepers and Fox, 1989; Schepers and Mosier, 1991; Thurlow and Hiltbold, 1985; Tisdale and Nelson, 1966. The lower values for fixation by perennial legumes given in Table A-4 generally reflect fixation during the first year of growth. The higher values reflect fixation that has occurred in stands of legumes that have been in place for 2 or more years. The estimates given in Table A-4 are estimates of the total fixed legume-N and are greater than values often estimated as available to crops that are planted after the legumes are harvested. The total fixation value, however, includes the amount fixed and taken up in the legume biomass, most of which is subsequently harvested and unavailable to succeeding crops. Estimates of the amount of nitrogen actually fixed by a particular species of legume are problematic because there are no unequivocal methods for measurement. LaRue and Patterson (1981) summarized published research and concluded there is not a single legume crop for which valid estimates of the nitrogen fixed in agricultural production were available. They did report consistent ranges for some legumes that emerged from various research. Part of the measurement problem occurs because legumes use nitrogen
APPENDIX 439 in the soil as well as fix nitrogen from the atmosphere. Generally, legumes will fix nitrogen only after taking up much of the nitrogen that is available in the soil (see, e.g., Phillips and DeJong, 1984). In fertile soils, with substantial soil organic matter and soil nitrogen, they will not fix as much as in soils of low fertility. For example, in Table A-4, the different values for fixation by soybeans under midwestern and southeastern conditions reflect the general differences in soils between those regions. Some studies suggest that some legumes (particularly annual grain legumes) may remove more nitrogen from the soil than they fix and, hence, the legumes may represent a net loss of nitrogen (Schepers and Mosier, 1991; Follett et al., 1987). Although unequivocal estimates of the nitrogen input from symbiotic nitrogen fixation are still open to discussion, crop rotation with legumes consistently produces a yield benefit to the succeeding crop, with reduced nitrogen inputs. This undoubtedly reflects various rotation effects as well as any nitrogen residuals supplied from true fixation or accumulation from other sources. To minimize environmental losses of nitrogen and to optimize crop yields, some estimate of the legume contribution to crop rotations must be made. To account for the combined effects of rotation and fixation, legume benefits are often estimated as a fertilizer nitrogen replacement value or a fertilizer nitrogen equivalence. Based on consistent results from many experiments throughout the United States, Schepers and Fox (1989) summarize that the fertilizer nitrogen equivalence of a 2- to 4-year old ''good" alfalfa stand is at least 100 to 150 kg/ha for the first succeeding crop and 30 to 50 kg/ha for the second crop; and nitrogen fertilizer applications to crops following soybeans should be reduced by approximately 15 to 17 kg/ha per Mg/ha of soybean yield (∼1 lb/acre/bu soybeans). In the committee's estimates of nitrogen balances, the input values for legume-N are balanced by estimates of the output of nitrogen in harvested alfalfa and soybean crops. The result is an estimate of the total nitrogen that may accumulate and remain as a nitrogen replacement value. In some cases, this replacement value may also reflect rotation benefits other than legume-N. Another problem in estimating the nitrogen supplied by legumes is that some legume-N, particularly from perennial forages such as alfalfa, may be available for succeeding crops several years after the forage crop is harvested and plowed down. For the first crop year the nitrogen replacement value of the legume may equal 150 to 200 kg/ha. Calculated over a 2-to 5-year period following plow down, some estimates of the nitrogen replacement value range as high as 450 kg/ha (Peterson and Russelle, 1991; Schepers and Mosier, 1991). The committee used only estimates of the hectares of alfalfa hay and soybeans harvested in 1987 and the tons of
APPENDIX 440 alfalfa hay or soybeans harvested in 1987 to estimate a replacement value for legume-N in 1987. Alfalfa and soybeans are the two major legume crops, but many others (see Table A-4) also contribute nitrogen inputs. Alfalfa and other legumes grown in pasture were not considered in the committee's estimates. Data on the land area planted to legumes and their yields were derived from the 1987 Census of Agriculture. TABLE A-5 Estimated Rates of Nitrogen Accumulation and Nitrogen Replacement Value for Alfalfa and Soybeans in Low-, Medium-, and High-Fixation Scenarios Legume Estimate Total Nitrogen Nitrogen Scenario Nitrogen Harvesteda (kg/ Replacement Fixed (kg/ha) ha) Valueb (kg/ha) Alfalfa Low 230 185 45 Medium 250 185 65 High 380 185 195 Soybean Low 175 165 10 Medium 200 165 35 High 220 165 55 aIncludes nitrogen in the harvested portion of the legume and in crop residue. b Includes the amount of fixed nitrogen available to a succeeding crop and the reduced need for supplemental nitrogen that may be a result of rotation effects. Because of the uncertainty in estimates of fixation and nitrogen replacement value, the committee generated three scenarios using a low, medium, and high estimate of the nitrogen supplied by alfalfa and soybeans in 1987. The input value used was considered to be an estimate of total fixation and accumulation of nitrogen by these legumes. The amounts of nitrogen harvested in alfalfa hay and in soybean grain and residues were entered on the output side of the balance. The difference between the total alfalfa or soybean nitrogen input and the alfalfa or soybean nitrogen output was used as an estimate of the residual nitrogen replacement value potentially available to a succeeding crop. The magnitude of the estimates of this nitrogen replacement value was compared to standard estimates of the nitrogen replacement value of alfalfa and soybeans derived from other research. The nitrogen fixation rates, estimates of nitrogen harvested in alfalfa hay and soybeans, and the resulting replacement values used in the committee's estimates of nitrogen balances are summarized on a per-hectare basis in Table A-5. The fixation values used in the committee's estimates agree well with those used by other scientists. The medium nitrogen fixation value for alfalfa (250 kg/ ha/yr) is approximately the same as that used by Peterson and Russelle (1991) (252 kg/ha/yr). The maximum value (380 kg/ha/yr)
APPENDIX 441 results in an estimated replacement value to a subsequent grain crop of 195 kg/ ha, slightly higher than the ranges suggested by Schepers and Fox (1989). However, only about 30 percent of alfalfa acres are typically rotated to another crop in most years. The low value for alfalfa nitrogen fixation produces an estimated resultant value for the total alfalfa hectares harvested in 1987 that approximates the replacement value expected if 30 percent of the alfalfa acres in 1987 were rotated to another crop in 1988. For soybeans, the estimates of fixation rates and harvest values used result in nitrogen replacement values equivalent to 5, 16, and 25 kg/ha/yr per Mg/ha of soybeans harvested (or about 0.3, 0.9, and 1.5 lb/acre/per bu soybeans). The replacement values, which are of most interest here, are in line with most recent estimates (see Meisinger and Randall, 1991; Schepers and Mosier, 1991; Schepers and Fox, 1989), with the medium to low estimates providing the best estimates for an annual balance. Crop Residues Crop residues are the mass of plant matter that remains in the field after harvest. The committee estimated the volume of crop residues using published estimates of the amount (ratio) of residue produced related to the amount of harvested grain (Larson et al., 1978). The phosphorus and nitrogen content of crop residues was derived from the United States-Canadian Tables of Feed Composition (National Research Council, 1982). The nitrogen content of residues was calculated assuming crude protein as 16 percent nitrogen. The grain-to-residue ratios used to estimate crop residue values the percentages of nitrogen and phosphorus in residues are given in Table A-6. (Crop residues are further discussed below). In an operational system, to use such balances for analysis, the legume residual from one year would become the input term for the next: that is, to assess opportunities for improved input management for 1987, the legume residual for 1986 should actually be used. In the committee's illustrative analysis, the 1987 legume residuals are used, for simplicity, but this also provides a conservative estimate; the area of soybeans and alfalfa was greater in 1986 than 1987, while the yields were greater in 1987. Estimation of Outputs Only the primary desirable nutrient outputs were estimated by the committee, that is, the nitrogen and phosphorus taken up in the harvested crops and in crop residues. Other outputs include undesirable losses into the environment through ammonia volatilization, denitrification, soil erosion and runoff losses, and leaching losses. As discussed
APPENDIX 442 above, the committee made no attempt to estimate the magnitude of these losses. The unaccounted for balance includes both storage and undesired losses that may be subject to improved management. TABLE A-6 Factors Used to Estimate Nitrogen and Phosphorus in Crop Residues Crop Ratio of Residue to Percent in Nitrogen Percent in Phosphorus Grain Residue Residue Alfalfa haya NA NA NA Barley 1.5 0.64 0.09 Corn 1.0 0.89 0.09 Cotton 1.0 0.59 0.05 Dry beansb 1.5 0.73 0.05 Haya NA NA NA Oats 2.0 0.65 0.06 Peanutsc 0.14 1.11 0.14 Potatoesc 0.15 0.37 0.03 Rice 1.5 0.48 0.07 Sorghum 1.0 0.74 0.12 Soybeans 1.5 0.74 0.05 Tobaccoa NA NA NA Wheat 1.7 0.51 0.04 NOTE: Conversion factor coefficients for bushels, bales, and hundredweight were obtained from the U.S. Department of Agriculture (1988). NA, not applicable. a Total plant harvest was assumed for alfalfa hay, hay, and tobacco. b The soybean residue/grain rate was used to estimate dry edible bean residue. c The peanut and potato ratio of residue to grain were obtained from J. W. Gilliam, North Carolina State University, personal communication, 1991. Other nutrient outputs from farming systems may include other gaseous losses such as N2O evolution during nitrification, decomposition of nitrous acid, or losses directly from maturing or senescent crops (see Bremner et al., 1981; Meisinger and Randall, 1991; Nelson, 1982). Some nutrients are taken up by weeds or immobilized by microbes and enter the storage pool. These outputs are small relative to other outputs and typically have been implicitly included in nutrient-crop yield response models. Harvested Crops The desired nutrient output from an agricultural ecosystem is the nutrient taken up in the harvested crop. The nitrogen and phosphorus
APPENDIX 443 in the harvested portion of the crop are effectively removed from the farming system, unless they are fed to livestock. Estimates of the yield of harvested crops in the 1987 Census of Agriculture were used by the committee to estimate phosphorus and nitrogen outputs. The phosphorus and nitrogen content of the harvested crops was derived from the United States-Canadian Tables of Feed Composition (National Research Council, 1982); nitrogen was calculated assuming crude protein as 16 percent nitrogen. The estimates of the nitrogen and phosphorus content of the harvested crops used to estimate nutrient outputs are given in Table A-7. TABLE A-7 Nitrogen and Phosphorus Content of Harvested Crops Crop Percent Nitrogen Percent Phosphorus Alfalfa hay 2.8 0.17 Barley 1.9 0.34 Corn 1.5 0.26 Cotton 1.6 0.11 Dry beansa 3.6 0.52 Hayb 1.9 0.27 Oats 1.9 0.33 Peanuts 4.4 0.30 Potatoes 0.35 0.06 Rice 1.3 0.28 Sorghum 1.8 0.29 Soybeans 6.3 0.60 Tobaccoc 2.7 0.29 Wheat 2.3 0.37 NOTE: Conversion factor coefficients for bushels, bales, and hundredweights were obtained from the U.S. Department of Agriculture (1988). a Percent nitrogen for navy beans was used for the dry edible bean. b To estimate percent nitrogen for hay, the average of five crops (Kentucky bluegrass, brome, fescue, oats [sun-cured hay], and timothy hay) was used. c Percent nitrogen for tobacco was obtained from J. S. Schepers and R. H. Fox. 1989. Estimation of N budgets for crops. Pp. 221-246 in Nitrogen Management and Ground Water Protection, R. F. Follett, ed. Developments in Agricultural and Managed-Forest Ecology 21. Amsterdam: Elsevier. Only the crops listed in Table A-7 were considered by the committee in estimating outputs of nitrogen and phosphorus from croplands. Analysis of national production statistics for all crops show that the
APPENDIX 444 crops included here account for approximately 90 percent of the total mass of harvested crops and more than 97 percent of the harvested nitrogen and phosphorus. Crop Residues The nitrogen and phosphorus in crop residues were estimated as described above. For the committee's purposes, crop residues were considered as both inputs and outputs and hence offset each other in the calculation of balances. In reality, however, residues from the previous year should be estimated as the input and the residues of the current year in the output term. Over time, however, most of the residues remain in the system and the nitrogen and phosphorus in those residues would appear alternately as inputs and outputs. Crop residues are often not counted as inputs in cropland balances. Estimation of Balances The committee's estimates, as discussed earlier, are partial nutrient balances for harvested cropland. The balance, or residual term represents an estimate of the amount of nitrogen and phosphorus inputs that (1) may go into storage or (2) may potentially be lost from the system (outputs) into the environment. The magnitude of the balance, then, provides insights into the potential for water pollution that may be created by nutrient fluxes through farming systems aggregated at the state, regional, and national levels. The magnitude of the balance and the relative importance of various inputs also provide insights into the opportunities to improve the management of nitrogen and phosphorus. These implications and opportunities are discussed in more detail in other portions of the report, as are comparisons with prior published balances (see Chapters 2, 3, 6, and 7). A summary of the committee's balance estimates is given in Table A-8. BALANCE ESTIMATES IN PERSPECTIVE The methods and values chosen for any nutrient balance, at the scale of this analysis, are clearly equivocal and subject to error. Despite the assumptions the committee was forced to make to estimate nitrogen and phosphorus balances, the committee's estimates are similar to other past published budgets (see Chapter 6). For the most part, the choices of methods and values made by the committee would be expected to produce a conservative estimate of the unaccounted for balance of
TABLE A-8 Inputs and Outputs of Nitrogen and Phosphorus on Croplands in the United States, 1987 Nitrogen Low Scenario Medium Scenario High Scenario Phosphorus APPENDIX Inputs and Metric Tons Percent of Metric Tons Percent of Metric Tons Percent of Metric Tons Percent of Outputs Total Inputsa Total Inputs Total Inputs Total Inputs Inputs Fertilizer 9,390,000 47 9,390,000 45 9,390,000 42 3,570,000 79 Manure 1,730,000 9 1,730,000 8 1,730,000 8 655,000 15 Legumes 6,120,000 30 6,870,000 33 8,560,000 38 NA NA Crop residues 2,890,000 14 2,890,000 14 2,890,000 13 272,000 6 Total 20,100,000 100 20,900,000 100 22,600,000 100 4,500,000 100 Outputs Harvested crops 10,600,000 53 10,600,000 51 10,600,000 47 1,320,000 29 Crop residues 2,890,000 14 2,890,000 14 2,890,000 13 272,000 6 Total 13,500,000 67 13,500,000 64 13,500,000 60 1,600,000 36 Balance 6,670,000 33 7,420,000 36 9,110,000 40 2,900,000 63 NOTE: NA, not applicable. a Input or output as a percent of the total mass of inputs. 445
APPENDIX 446 nitrogen and phosphorus. Some perspectives on the effect of the assumptions the committee made are worthy of review. Although data for only 1 year were used, as noted, 1987 appears to have been representative for the committee's purposes. It would be better, of course, if annual data on inputs and outputs were used to account for year-to-year variation in the crops planted, inputs used, and crop yields. The estimate of balances over time would be a particularly useful way to assess progress in improving input use efficiency and would be particularly useful since crop outputs can vary tremendously from year to year. Also, some inputs and outputs are not in steady state and should be factored over a period of years. Some of the nitrogen fixed by legumes, for example, is available for more than 1 year. Also, the nitrogen in manures is released over time, and multiyear decay constants are sometimes used to estimate the annual contribution of nutrients from manures (Schepers and Mosier, 1991). The committee's estimates of inputs are probably low, particularly the estimates for phosphorus and the estimates in the low and medium nitrogen scenarios, since no effort was made to account for the multiyear contributions or buildup of nitrogen and phosphorus over time. Standard assumptions were used by the committee to estimate nitrogen and phosphorus inputs from manures. Some of the manure produced undoubtedly would have been applied to pastures and other land, resulting in an overestimate of that applied to cropland. As discussed, however, the values used to estimate recoverable manure were conservative; they were derived for 1970s conditions, and by 1987 livestock production had become more concentrated, with more cattle and swine in confinement operations and less on pasture and range. Hence, a greater proportion of manure would be collectable in 1987 than 1974. Given the uncertainties for legume-N inputs, the assumptions made by the committee have been conservative. On the input side, only estimates for accumulation/fixation by alfalfa and soybeans were used. However, on the output side of this balance, the nitrogen harvested in other important legumes is included, notably dry beans and peanuts (Table A-6). Legumes account for more than 35 percent of the total nitrogen harvested in crops. Balances estimated on a crop-by-crop basis, then, would result in much higher ratios of inputs to outputs for some crops than is estimated by aggregating all crops together by state, region, or for the United States as a whole. The committee, however, did not estimate the nitrogen and phosphorus harvested in citrus and vegetables, undoubtedly resulting in too low an output value for some states, such as California, Florida, and Texas.
APPENDIX 447 The most important factor leading to conservative estimates of the difference between inputs and outputs was the committee's decision not to correct estimates of the harvested crop for moisture content. The committee did not correct for moisture content because of uncertainties in how the yields of several crops were reported. Correcting for the moisture contents of harvested crops would reduce the harvested nutrient output by about 15 percent on average across the crops considered. IMPLICATIONS OF BALANCES Even with the relatively conservative assumptions used, subtracting harvested crop outputs from inputs results in a residual balance equal to more than 60 percent of the phosphorus inputs and from 33 (low scenario) to 40 percent (high scenario) of the nitrogen inputs. The proportion of the residual nitrogen balance contributed by legumes was about 6, 9, and 16 percent less than that of the low, medium, and high scenarios, respectively. Although the nutrient balance after harvest may go into storage or environmental losses, as discussed in Chapters 6 and 7, many studies have indicated that the magnitude of such balances are directly related to the magnitude of environmental losses. Phosphorus is relatively immobile and may build up in the soil over time, and the concentrations in the surface soil, in particular, are directly related to runoff losses. Nitrogen is much more mobile in the environment, and few farming systems show any increase in soil nitrogen over time, even where large balances are found. There are several facets of the nitrogen budget and balances that need further discussion. As many large-scale and multicrop nitrogen balances would note, the nitrogen removed in the harvested crop is slightly more than the fertilizer nitrogen input. However, more than 35 percent of the total nitrogen in harvested crops is accounted for by legumes, which receive very little nitrogen fertilizer. Major commodities such as corn, cotton, potatoes, rice, and wheat account for more than 80 percent of the nitrogen applied in fertilizers, but the nitrogen harvested with these crops accounts for only 57 percent of the fertilizer nitrogen input. If all legume inputs and outputs are taken out of the estimated balances, the nitrogen harvested in the remaining crops is only about 35 to 40 percent of the nitrogen fertilizer and manure inputs. Nutrient balances constructed on a crop-by-crop basis, and analysis by input category can provide important guidance to programs seeking to improve the efficiency of nutrient use in farming systems.