EWG Analysis of Manure and Fertilizer Nutrient Loading in Minnesota

By Sarah Porter, Senior Geospatial Analyst and Project Manager

Special thanks to David James, from the USDA Agricultural Research Service, and Jack Acomb, for his help with aerial photography dating of Minnesota animal feedlots.

Introduction

Contamination of water resources is a growing concern in Minnesota. The Minnesota Pollution Control Agency, or MPCA, estimates that 56 percent of the state’s surface waters that have been assessed do not meet basic water quality standards,¹ and that non-point source pollution contributes to as much as 85 percent of this pollution.

Adequately addressing water quality issues in Minnesota requires a better understanding of the relative contribution of the major nutrient sources and their distribution. As concentrated animal feeding operations have grown in size and density over the past 25 years, so has concern over the capacity of the surrounding cropland to sustainably handle the byproducts of animal agriculture.

Consolidation of animal operations has created landscape scenarios in which large quantities of manure are produced in geographically clustered areas. Losses of nitrogen, or N, and phosphorus, or P, the two primary nutrients in manure, can contaminate surface water and groundwater, leading to a wide range of environmental and economic consequences.²

The following EWG analysis attempts to improve our understanding of the relative contribution of commercial fertilizer and livestock manure to nutrient loading in Minnesota.

• First, EWG used aerial photography to document the spatial and temporal characteristics of large concentrated animal feeding operations, or CAFO, growth in Minnesota between 1991 and 2020.
• Second, EWG used geospatial techniques to quantify the capacity of the landscape to uptake manure nutrients while identifying areas of manure saturation in the state.
• Third, commercial fertilizer sales data were combined with manure-sourced nutrient estimates to assess nutrient application relative to crop fertilizer recommendations at the statewide and county level.

History of Large CAFO Construction

Animal Feedlots

The Minnesota Pollution Control Agency maintains a database of feedlots³ housing 50 or more animal units in the state. Animal units, or AU, are used to compare differences in the production of animal manure among livestock types (1 AU equals 1,000 pounds of animal weight).⁴ MPCA publishes this feedlot registration as a geospatial dataset, which was downloaded from the Minnesota Geospatial Commons on February 3, 2020. The dataset included general information for 23,725 active operations housing a dominant livestock group, including poultry (layers, broilers and turkeys), swine, and beef and dairy cattle.

EWG tracked new and expanded CAFOs between 1991 and 2020. MPCA flags operations as CAFOs if they meet the large CAFO definition⁵ set by the U.S. Environmental Protection Agency, which relies on the type and number of animals housed at each facility. Swine operations with more than 2,500 hogs weighing over 55 pounds, or 10,000 hogs weighing less than 55 pounds, are considered large CAFOs, compared to facilities with 55,000 turkeys and 700 mature dairy cows.

Facilities were selected if they housed more than 1,000 AU or included a “CAFO” flag in the feedlot database. 1,557 operations met this definition. Sixty locations were omitted because we could not locate these operations using aerial photography. Reasons for this include that the facility had not yet been built, incorrect location information, duplicate facility or some other recording factor. The final count for large CAFOs was 1,497 operations.

Historical Imagery

The Minnesota IT Services Geospatial Information Office maintains a large collection of statewide aerial photography. EWG used this collection, along with recent Planet satellite imagery, to document the history of large CAFO construction and expansion in Minnesota between 1991 and 2020. Each CAFO was analyzed at six time stamps, listed below, along with the source of imagery used:

• 1991 – NAPP USGS black and white aerial ortho-photography
• 2003 – NAIP aerial photography; 1-meter horizontal resolution
• 2008 – NAIP aerial photography; 1-meter horizontal resolution
• 2013 – NAIP aerial photography; 1-meter horizontal resolution
• 2017 – NAIP aerial photography; 1-meter horizontal resolution
• 2020 – Planet satellite imagery

Facility Attribution

Several attributes were recorded for each facility, including the year it was first observed in imagery and years of expansion, if any. A facility expansion included the addition of or increase in size of livestock barns, or in the case of some cattle operations, the addition of or increase in size of feedlots or paddocks. Facility attribution was performed using the best judgment of the GIS analyst. An additional round of quality control was performed by a separate reviewer.

CAFO Dating Results

Between 1991 and 2020, the number of large CAFO operations in the state more than tripled, from 468 operations in 1991 to 1,497 operations in 2020. This represents a 219 percent increase. Most of this growth came from the swine industry, accounting for 889 of the 1,024, or 86 percent, of the new facilities constructed. This is followed by dairy (7 percent of new facilities constructed), poultry (5 percent) and beef (2 percent) (Figure 1).

Swine accounted for most new operations, but the number of large CAFOs for all animal types grew over the time period. Twenty-two large beef facilities were constructed after 1991 (a 28 percent increase), 67 dairy facilities (a 210 percent increase), 51 poultry facilities (a 36 percent increase) and 889 swine facilities (a 413 percent increase). Thirteen facilities decreased in size between 1991 and 2020, as buildings were removed or reduced in size.

The largest period of growth occurred between 1991 and 2003. The number of large CAFOs increased almost threefold during this period, from 468 facilities in 1991 to 1,114 in 2003. This trend of rapid CAFO growth and expansion in the 1990s has been observed across the country and is documented in EWG reports on CAFO growth in Iowa and in Ohio’s Maumee River basin.

EWG estimates that the 1,029 CAFOs added since 1991 have produced an additional 10.3 million tons of manure per year, 63,000 tons of land-applied nitrogen, and 15,000 tons of land-applied phosphorus, in the state. This does not reflect overall manure nutrients added to the landscape since 1991, since many smaller animal operations, dairy cattle in particular, have seen declines over the same period.

Figure 1. The number of large livestock operations in Minnesota tripled since 1991

Source: EWG via Minnesota Pollution Control Agency.

Trends in CAFO Size

The size of CAFO facilities, measured by the number of AU housed at each location, was compared for facilities existing in 1991 to those constructed after 1991 (Table 1). CAFO size remained remarkably consistent across time periods for swine and beef operations, whereas poultry facility size increased slightly, from an average of 1,502 AU in 1991 to 1,713 AU for facilities constructed after 1991. We observed an increase in size for dairy facilities, with a more than 50 percent increase in the average number of AUs housed at facilities in 1991, compared to those constructed after 1991.

The number of AUs housed at each facility was recorded from current (2020) permit data. This may not accurately reflect operation size in 1991, considering that nearly half, or 44 percent, of the 1,497 facilities present in 2020 have expanded at least once since their original construction. Two hundred and twenty-eight facilities have expanded more than once.

Table 1: Size and number of large CAFOs built before and after 1991

Source: EWG via Minnesota Pollution Control Agency.

Geographic Trends in CAFO Construction

More than half of new operations were constructed in just 12 counties. The top counties for CAFO growth included Martin (10 percent of all new operations) and Nobles (6 percent), followed by Blue Earth, Jackson, Rock, Waseca, Watonwan and Renville counties (tied for 4 percent of all new operations) (Figure 2). Martin County is currently home to the most CAFO operations in the state, at 148. This is almost twice as many as the next densest county (Nobles) at 79.

Figure 2: Share of CAFO Growth by County

Source: EWG via Minnesota Pollution Control Agency.

Modeling Landscape Capacity For Manure Nutrients

EWG employed a novel GIS program developed by Porter and James⁶ to spatially model the capacity of neighboring cropland to incorporate manure nutrients. Detailed estimates of manure and nutrient excretion were made for all 27,325 active feedlots in Minnesota.

Nitrogen loss was accounted for during manure storage and handling, and upon application to determine the amount of manure nutrients available for land application on an annual basis. The N requirement of agricultural fields was estimated using six-year crop rotations and commonly used guidelines on fertilizer recommendations for agronomic crops. Manure was applied to fields to meet the N requirement of growing crops. P was applied at the same time according to the calculated N to P ratio of each feedlot.

A distance-limiting measure was built into the program to limit the maximum distance manure can travel from its source. This measure helped reveal areas of pressure for manure overapplication when it was applied at recommended N rates. Lastly, manure nutrients were combined with county-level commercial fertilizer sales to allow for an estimate of nutrient supply relative to crop requirements at a county level.

Results were reported across a range of nutrient values that assumed different average weights for a finishing hog, both to analyze the sensitivity of this approach to changing nutrient parameters as well as to better represent MPCA values.

Estimating Nutrients From Feedlots

Animal counts listed for each facility represent the maximum number of animals that may be housed at that location and may not accurately reflect the actual number of animals at any given time. This is particularly true for smaller operations. To account for this, animal counts were reduced for facilities below 300 animal units using the following adjustment factors⁷: 90 percent for dairy and swine, 70 percent for beef, 80 percent for turkey and 85 percent for chickens.

The amount of manure, N and P excreted annually from each active feedlot was calculated using book values from the Midwest Planning Service (MWPS) 18 Second Edition (2004).⁸ Each type of animal in the Minnesota feedlot database was matched to an animal type from MWPS (Table 2) to calculate daily excretion values for manure, N and P. MWPS values were averaged when an exact match to an animal type could not be made. Daily excretion amounts were multiplied by 365 days to estimate annual excretion amounts. Although the MWPS is a common resource for estimating manure nutrient content, book values provide estimates only, and the actual characteristics of manure can vary by 30 percent from table values due to genetics, diet and farm management.

A study by Porter and James (2020)⁶ applied the average of MWPS finishing swine categories (weights from 150 to 300 pounds, with an average of 220 pounds) to all medium swine in the Minnesota feedlot database. However, MPCA routinely assigns a 150 pound weight to medium swine, assuming a grow-to -finish operation with pigs averaging between 55 and 300 pounds. As the manure and nutrient values for a 220 pound hog are approximately 40 percent greater than values for a 150 pound hog, these differences can be significant, particularly for the state’s dominant swine production system. To account for this, EWG ran the model twice, and results in this report are listed as a range of values representing a low end (150 pounds for all grow-to-finish swine) and a high end (220 pounds for all grow-to-finish swine) for nutrient values.

A range of nutrient values was not modeled for other animal types, as these values align better with MPCA estimates. It should be noted, however, that manure nutrient contents can vary widely among all livestock types.⁹ Ideally, a sensitivity analysis using different parameters would provide a full range of manure application scenarios.

Porter and James (2020)⁶ assessed the sensitivity of this approach to changes in N application rate and travel distance in Minnesota. They found that the extent of land area receiving manure, as well as the potential risk of overapplication, will expand and contract depending on which parameters are used, although the spatial patterns remain consistent. When varying manure-haul distance and N application rate, Porter and James (2020)⁶ found that total N applied in Minnesota (from both manure and commercial fertilizer sources), was between 110 percent and 155 percent of recommendations.

Table 2: Manure and Nutrient Excretion Values

Source: EWG via Midwest Plan Service.

Many cattle feedlots in Minnesota use a combination of confinement buildings and pasture. To account for the proportion of manure that is applied to pasture and is therefore unrecoverable, as-excreted amounts of manure, N and P were reduced by 50 percent for all cattle operations below 300 AU and with a “pasture” flag in the feedlot database. This factor was applied to over 12,000 beef and dairy cattle operations in the state and significantly reduced the manure contribution from smaller cattle and dairy operations, particularly in areas with large amounts of pasture and other grasses or hay.

Much of the N excreted in manure is lost to the atmosphere (primarily as ammonia) during manure storage, handling and field application. Although not accounted for in this study, there is evidence that much of this volatilized N is redeposited to the land surface within 1 kilometer of the confinement.¹⁰ The amount of N loss depends on several factors, including the type of manure and how it is stored and applied.

Nitrogen loss from each facility during storage and application was estimated using guidelines¹¹ developed by University of Minnesota Extension and Minnesota Department of Agriculture. A 2014 survey of Minnesota’s 2014 corn crop¹² was used to inform the prevalence of various manure application methods, with details provided in Porter and James (2020).⁶ Values for N loss used in this study are shown in Table 3. Annual N excretion amounts were reduced by the percent N loss during manure storage and handling, then subsequently reduced by the percent N loss during application.

Although the inorganic N in manure is available to crops immediately, the organic N portion acts as a slow-release fertilizer, becoming available to the crop throughout the growing season and up to several years following application.¹³ Nitrogen availability in this study was considered to be the sum of manure N available to the crop during the first, second or third year after application.¹¹ Similar studies¹⁴ estimating the contribution of manure nutrients to meet crop N needs have also used the sum of three-year availability. Additionally, Sawyer and Mallarino (2008)¹³ suggest that 90 to 100 percent of swine manure N is available to the crop in the first year after application.

Although some second-year N may be taken up by soybeans in a corn-soybean rotation, it is assumed that all N used by soybeans is obtained entirely through N fixation. This assumption may lead to a modest overestimate of manure N application relative to crop needs, as research has generally indicated that if mineral nitrogen is present in the soil, the plant will use this nitrogen to support its growth and development.¹⁴

Using methods in this study and across all animal types, an estimated 26 percent of the manure N applied each year will become available during the second year following application (compared to 68 percent in year one and 6 percent in year three). Methods presented in this study would benefit from future accounting of the proportion of year two N availability that will be taken up by soybeans.

Many areas with high livestock densities are trending toward continuous corn rotations, however, which requires taking credits for second year manure nutrients. In Martin County, for example, 40 percent of fields in a rotation including both corn and soybeans grew corn in four or more years of a six-year rotation (2012-2017). Additionally, a problem in temperate climates such as Minnesota is that mineralization of residual organic N may occur at times when crops are not present or are small, such as in the spring or fall, which can lead to increased leaching of manure N rather than uptake by growing crops. In response, some states are recommending that fields reduce manure application rates to the point that total manure N applied is approximately equal to projected crop demand, particularly for those with a history of regular manure additions.¹⁵

As-excreted amounts of P205 were multiplied by .44 to convert to elemental P. The amount of manure P applied annually from each feedlot was assumed to be the same as that excreted, assuming that P loss is negligible for all but open-lots and lagoons.¹⁶ Purdue University suggests that 20 to 40 percent of P can be lost to runoff and leaching from an open lot, though much of this can be prevented using runoff collection systems. Additionally, 50 to 85 percent of lagoon P may settle to the bottom and be unavailable until agitated, at which point P in lagoon sludge can be applied to cropland.¹⁶ In contrast to potential N loss to the atmosphere, P loss from animal operations will directly enter the land environment, posing a risk to water resources through runoff or buildup of soil P over time.

Table 3: Nitrogen Loss During Manure Storage and Field Application

Source: EWG via University of Minnesota Extension and Minnesota Department of Agriculture.

Estimating Field Nutrient Needs

Field boundaries with crop rotation history were obtained from the Agricultural Conservation Planning Framework (ACPF) database¹⁷ and were used to estimate the average annual nutrient requirement for each field in the state. ACPF provided six years of land cover history, from 2012 to 2017. Fields smaller than 15 acres were considered too small for reliable classification and were excluded from the analysis.

The use of a six-year crop rotation is important for several reasons. First, it allows for fertilizer recommendations to vary based on the previous crop grown. A legume credit, for example, results in a lower suggested N rate for corn following a soybean crop, as opposed to corn following a corn crop. Second, analysis of nutrient application should focus on the long-term loading of N and P over a crop rotation, rather than a single year.

For each field, a six-year nutrient requirement was calculated by summing the nutrient requirement for each of the six years. This value was divided by six to represent the average annual nutrient requirement for each field. Manure application is modeled in this study to meet the average annual N fertilizer requirement of each field. Manure application to meet the N need of growing crops is the most common approach throughout much of Minnesota, although P-based rate requirements may be used in certain sensitive situations.

The example below illustrates the difference in the average annual N requirement for a field in a continuous-corn versus an alternating corn-soybean rotation. This example assumes an N fertilizer rate of 150 pounds per acre for corn following a soybean crop, 195 pounds per acre for corn following a corn crop, and a soybean N fertilizer rate of zero. The ACPF six-year rotation allowed for the previous crop to be identified for all years of corn except year one, in which case a previous crop of soybean was assumed.

Example: Corn-soybean rotation (CBCBCB)

(150 pounds * 3 years) + (0 pounds * 3 years) / six years = 75 pounds/acre average annual N requirement

Example: Corn-corn rotation (CCCCCC)

(150 pounds * 1 year) + (195 pounds * 5 years) / six years = 187.5 pounds/acre average annual N requirement

The ACPF allows for a detailed look at crop rotation across Minnesota. Of the 265,453 fields with a N fertilizer requirement over a six-year period, 56 percent, or 147,380 fields, were classified as a strict corn-soybean rotation, with at least two but not more than three years of each crop type. The next most common rotation included corn as the dominant crop, with 21 percent, or 57,035 fields, of fields growing corn in four or more years of the six-year rotation. In these corn-dominated rotations, soybeans were the other crop 80 percent of the time.

A soybean-wheat rotation was the next most common on 7 percent of fields, followed by a corn perennial rotation on 4 percent of fields. Remaining rotations included a variety of extended rotations that include a mix of corn, soybeans, sugar beet, wheat and perennial crops. Nearly 3 percent of all fields, or 7,312 fields, grew corn in all six years of the rotation.

The methods of modeling manure application, described in detail below, represent an ideal scenario, in which each field surrounding a feedlot that has a fertilizer N requirement is available for manure application. However, many farmers may hesitate to accept manure because of variability in nutrient content and issues with soil compaction. State setback requirements and manure restriction areas further reduce the land available for manure application.

Estimating Field Nitrogen Need

Land cover was grouped into seven categories considered to be the major crops grown in Minnesota. The .05 Maximum Return to N (MRTN) rate for Minnesota¹⁸ was used to determine the N rate for all years planted to corn. This rate is suggested by MPCA as the maximum N rate for corn production¹⁹ and equates to 150 pounds per acre for corn following a soybean crop and 195 pounds per acre for corn following a corn crop. University of Minnesota Extension fertilizer guidelines²⁰ were used to assign crop N requirements for crop types other than corn (Table 4).

As with Andersen and Pepple (2017),¹⁴ the N requirement for both soybeans and alfalfa was set at zero, assuming that all N removed by these crops is obtained entirely through N fixation. Although some manure application to alfalfa does likely occur in Minnesota, this is assumed to be a small proportion of overall manure application in the state, with MPCA estimates ranging from 10 to 50 percent of manure from small cattle operations applied to alfalfa. The 50 percent reduction in manure for all cattle operations below 1,000 AU and with a pasture flag in the feedlot database was assumed to partially account for this fraction.

Table 4: Major Minnesota Crop Types and Modeled Nitrogen Recommendations

Source: EWG via Minnesota Pollution Control Agency and University of Minnesota Extension.

Estimating Field Phosphorus Need

The P requirement of each field was calculated as the amount of P required to replace what is removed with the harvested crop. University of Minnesota Extension and MDA guidelines¹¹ on the amount of P removed per unit yield were used to estimate by-field P removal rates (Table 5). Five-year county yield averages, from 2014 to 2018, from NASS surveys²¹ for each of the major crop types were assigned to agricultural fields based on the county where the center of the field is located. Alfalfa was differentiated from pasture in the ACPF field database by the inclusion of at least one year of row crop in the six-year crop rotation. The weight of a bushel of dried peas (60 pounds) was used to convert from P removal listed on a per-bushel basis for edible beans to average country yields listed in pounds per acre.²²

Table 5: Phosphorus Removal per Unit Yield for the Major Crop Types in Minnesota

Source: EWG via USDA National Agricultural Statistics Service, University of Minnesota Extension and Minnesota Department of Agriculture.

Commercial Fertilizer Sales

County-level commercial fertilizer sales were obtained from the MDA 2016 Crop Year Fertilizer Sales Report.²³ Statewide farm-use N and P205 sales were reported as 779,078 tons and 321,035 tons, respectively. Fertilizer sales for both N and P205 have remained steady over the past five years, with 2016 sales data falling right below the five-year average (2012-2016) for both N (781,962 tons) and P205 (330,053 tons).

Manure Allocation Process

Manure allocation was modeled from the point of production to neighboring agricultural fields, using the methods detailed in Porter and James (2020).⁶ Briefly, the program runs as a series of manure application loops, moving outward from each feedlot and applying manure based on the average annual N requirement of each field. As manure is applied to meet the N requirement of each field, P is also applied using the ratio of N to P that was uniquely calculated for each feedlot.

Each loop starts with a selection, for each feedlot provided, of the single nearest field (based on a straight-line distance measure), with an average annual N requirement greater than zero. A simulation of manure application subtracts the amount of N required by the field from the available N of the feedlot. Tracking accounts for the amounts of N and P applied during each loop of manure application. Once a field has met its total N requirement, the field is no longer eligible to receive N from any feedlot. Once a feedlot has disposed of the entirety of its manure, it is removed from the analysis.

Distance Limiting

A distance-limiting measure was built into the program to account for the cost of manure haul distance. A unique transport distance was assigned to each feedlot and specifies the maximum distance (defined as the shortest separation between two features) between the feedlot and a field identified for manure application. If all fields within the specified distance have received their N requirement, overapplication of manure must occur.

To begin modeling potential overapplication, the N requirement of all fields is reset to their original N requirement. The manure application loop then starts over by selecting the nearest field to each feedlot with manure remaining, then moving outward until either the N requirement of all fields is met a second time, or the feedlot has disposed of its manure.

If manure is applied to meet the N need of crops a second time, P is also applied, using the N to P ratio of each feedlot. As many overapplication loops take place as are necessary to dispose of all the manure from feedlots within the specified distance.

EWG relied on a 2014 MDA survey¹² and internal discussions with MPCA to inform the travel distances used in this study (Table 6). A maximum distance of five miles was used for all dairy, beef and swine facilities, and a maximum 25-mile travel distance was used for all poultry facilities.

Table 6: Maximum Manure Haul Distance for Each Animal Type

Source: EWG via Minnesota Pollution Control Agency and Minnesota Department of Agriculture.

Results

Manure Production

The methods described in this report show that Minnesota’s 23,725 active animal operations produce an estimated 49,347,638 tons of manure annually. For comparison to human population, Spellman and Whiting (2007)²⁴ report an estimated 0.518 tons of waste generation per person per year, which puts the Minnesota animal-to-human waste equivalent at approximately 95 million people.

Contribution by Size of Operation

Table 7 lists the manure generated on an annual basis by size of operation – small (fewer than 300 AU), medium (300 to 1,000 AU) and large (more than 1,000 AU). Operations larger than 300 AU account for only 4 percent of operations while producing nearly a third of the manure statewide. Although this is significant, operations between 300 and 1,000 AU account for nearly half of all manure produced statewide, suggesting that manure management for medium-size operations deserves additional focus.

Table 7: Percent of Statewide Manure Production by Operation Size

Source: EWG via Minnesota Pollution Control Agency and Midwest Plan Service.

Contribution by Type of Animal

Table 8 lists the total allowable number of animals in Minnesota by animal type, along with their percentage contribution to statewide manure N availability. In contrast to manure produced, N availability is the primary determinant of the amount of land area needed for manure application from each animal type. Animal counts listed in Table 8 are following the adjustment for overreporting in facilities smaller than 300 AU.

Table 8: Percent of Statewide Applied Manure Nitrogen by Animal Type

Source: EWG via Minnesota Pollution Control Agency and Midwest Plan Service.

Manure Allocation

Nitrogen

Using the calculations described in this report, an estimated 410,021 to 465,377 tons of N are excreted annually by all animal feedlots in Minnesota, with the low end assuming a 150-pound finishing hog and the high end representing a 220-pound finishing hog. Nearly half of this N (175,502 to 193,216 tons) is considered lost to the atmosphere during manure storage and upon field application. What remains is an estimated 234,519 to 272,161 tons of manure N to be land-applied to cropland in Minnesota each year.

There is ample cropland in the state requiring N fertilizer, with 18,853,295 acres (265,453 fields) having an annual N requirement greater than zero, based on a six-year crop rotation, with a total annual N requirement of 774,277 tons. On a statewide scale, this suggests that manure can satisfy between 30 and 35 percent of the crop N requirement in any given year.

Results show manure application to between 5,153,172 and 5,958,523 acres, or 27 to 32 percent of cropland in the state. Manure nutrients can fully satisfy crop fertilizer N needs on these acres. Figure 3 illustrates fields likely to receive manure when applied to meet the fertilizer N need of crops, using the low-end manure rate.

Figure 3: Fields Receiving Manure To Satisfy Nitrogen Fertilizer Recommendations

Source: EWG via Minnesota Pollution Control Agency, USDA-ARS Agricultural Conservation Planning Framework Database, Midwest Plan Service, University of Minnesota Extension and Minnesota Department of Agriculture.

Many regions throughout the state were identified where manure alone can fully satisfy the crop fertilizer requirement of all fields within a several-mile radius. These regions suggest a high degree of manure saturation and are widespread throughout southwest and central Minnesota, as well as parts of south-central Minnesota (in Martin County, in particular). Additional areas of manure saturation are located in west-central and southeast Minnesota, such as Winona County and the southern half of Stevens County.

Figure 5a shows the percent of county N requirement that can potentially be supplied by manure, assuming the low end of manure N availability (150-pound finishing hog). Counties with a higher hog density will have a larger potential range for total N satisfied by manure in this study, with the high end representing nutrient values calculated for a 220-pound hog. Eight counties can likely supply more than 60 percent of the crop need with manure alone and are listed in Table 9.

Table 9: Counties With a High Degree of Manure Saturation

Source: EWG via Minnesota Pollution Control Agency, USDA-ARS Agricultural Conservation Planning Framework Database, Midwest Plan Service, University of Minnesota Extension and Minnesota Department of Agriculture.

Landscape saturation with manure nutrients is a result of concentrated livestock density, insufficient cropland, or a combination of these factors. In three of the counties identified above (Morrison, Todd and Winona), the amount of cropland with an N fertilizer recommendation is less than 30 percent of total county acreage. This is compared to more than 75 percent for Rock, Nobles, Martin and Pipestone counties. In Stearns County, cropland with an N fertilizer recommendation makes up 49 percent of total county acreage.

Fields are modeled as receiving manure based on their average annual N requirement, which is not necessarily the actual N requirement of that field in any given year. For example, a field with a strict alternating corn/soybean rotation will have an average annual N requirement of 75 pounds/acre (three corn years at 150 pounds/acre + three soybean years at 0 pounds/acre divided by six years = 75 pounds/acre).

To estimate the acreage that might receive manure in any given year, the number of manured acres in soybean or pasture/alfalfa in an average year (using a six-year average from 2012 to 2017) was removed from acres receiving manure. This suggests that in an average year, between 2,916,442 and 3,151,195 acres, or 15 to 17 percent of cropland acres, in Minnesota are likely to receive manure.

At the low end of manure nutrients, minimal (0.9 percent of manured fields) yet isolated regions of potential manure overapplication are seen when applying at N rates and using the five and 25-mile haul distances. This suggests that, in general, there is enough cropland in Minnesota to uptake manure nutrients at the haul distances specified. This does not account for any commercial fertilizer application.

Isolated regions of manure disposal pressure are seen in southwest Minnesota, as well as in the southeast and central parts of the state. In these areas, there is likely significant pressure to dispose of manure above agronomic rates. In Nobles County, 1,244 acres, or 0.3 percent of cropland in the county, were identified as high risk for manure overapplication. This compares to approximately 8,000 acres in northeast Winona County, which represents 7.9 percent of cropland in the county. In Stearns and Morrison counties, 4,200 acres, or 1 percent of county cropland, and 2,600 acres, 1.6 percent of county cropland, were identified as high risk, respectively.

Phosphorus

Using the same calculations, an estimated 77,552 (150-pound finishing hog) to 87,808 tons (220-pound finishing hog) of manure P are applied to cropland in Minnesota each year. On average, 63,766 to 74,227 tons of P are removed annually from these same manured acres. This suggests excess P on manured acres ranging from 18.3 to 21.6 percent statewide, although the geographic distribution is highly regional.

A major concern with manure application to meet crop N needs is the imbalance between manure N and P relative to crop demands. For example, the University of Minnesota Extension states that turkey manure applied to meet the N needs of corn supplies more than five times the phosphate needed, and fields that receive turkey manure each year often show high levels of P buildup.²⁵

Long-term research in South Dakota showed that cattle manure applied to meet the N need of crops raised soil P levels substantially over a period of seven years (dairy manure) and 12 years (beef manure).²⁶ Over the twelve-year period of applying beef manure at N recommendations in a corn soybean rotation, soil P levels increased from less than 20 parts per million to more than 100.

Figure 4 shows P excess on manured acres when it is applied to meet the N recommendation of the crop, using the low end of manure nutrients (150-pound finishing hog). High-risk regions are geographically located in central and southeast Minnesota (and to a lesser degree, southwest Minnesota), in counties with a high proportion of poultry and cattle manure. In these counties, average P excess on manured acres ranged from 15 to 25 pounds per acre.

Figure 4: Phosphorus Balance on Manured Fields When Applied at Nitrogen Rate

Source: EWG via Minnesota Pollution Control Agency, USDA-ARS Agricultural Conservation Planning Framework Database, Midwest Plan Service, University of Minnesota Extension and Minnesota Department of Agriculture.

On 2,684,581 acres, or 57 percent of all fields receiving manure to meet the N need of crops, more P was applied than removed. On 1,463,057 acres, or 33 percent of manured fields, this excess was more than 10 pounds per acre. On 590,697 acres, or 14 percent of manured fields, this excess was more than 25 pounds per acre. More than half of all manured fields with a P excess greater than 10 pounds per acre were located in nine high-risk counties, all in central and southeast Minnesota, and all with high densities of poultry and dairy operations. These nine counties are shown in Figure 4 and listed in Table 10.

Table 10: High-Risk Counties for Excess Manure Phosphorus

 Source: EWG via Minnesota Pollution Control Agency, USDA-ARS Agricultural Conservation Planning Framework Database, Midwest Plan Service, University of Minnesota Extension, Minnesota Department of Agriculture, and USDA National Agricultural Statistics Service. 

Phosphorus overapplication from manure is less of a risk in areas dominated by swine manure, as the N- to-P ratio of swine manure is more balanced for a corn/soybean rotation than other manure types. Counties with a high proportion of swine manure tend to have a lower amount of P excess per manured acre and can even have a negative P balance in counties with the highest corn yields. It should be noted, however, that manure applied to meet the N rate of crops, assuming only year one availability of manure N, will increase total P applied, exacerbating P excess relative to crop removal rates.

To limit buildup of P in the soil, manure application should be based on replacement of P that will be removed from the soil by the upcoming crop (in the harvested portion).²⁷ Adoption of these P-based strategies will require an assessment of the increased area of land required for manure application at reduced fertilizer rates, which will introduce additional logistical and economic constraints.

The sum of P removal from all cropland in the state is estimated to be 221,663 tons. This suggests that manure can supply between 35 and 40 percent of the total P requirement in the state, with significant county-level variability. Figure 5b shows the percent of county P requirement that can potentially be supplied by manure alone, assuming the low end of manure nutrient estimates (150-pound finishing hog).

Manure P can supply more than 100 percent of the P removed from harvested crops in four counties: Morrison, Todd, Stearns and Winona. Excess P from manure in Morrison County is substantial, ranging from 166 to 171 percent of the total P removed by crops. This range is 118 to 120 percent in Todd County, 115 to 117 percent in Stearns County, and 103 to 105 percent in Winona County. In an additional two counties, Wadena and Benton, manure P can supply greater than 75 percent of total P removed.

Figure 5: County Nitrogen (L) and Phosphorus (R) Need Met by Manure

Source: EWG via Minnesota Pollution Control Agency and Midwest Plan Service.

Commercial Fertilizer Sales

Commercial fertilizer sales data may not accurately reflect fertilizer use in the same county where it was sold. MDA reports that clustering sales data from neighboring counties can reduce this geographical bias, and that statewide sales data is considered very accurate. To account for this geographic bias, farm-use fertilizer sales for both N and P205 were summed within the state of Minnesota’s Agriculture Best Management Practice (BMP) regions,²⁸ then redistributed back to counties based on each county’s respective share of N and P fertilizer requirement within that BMP region. P205 sales were multiplied by 0.44 to convert to elemental P.

Differences in soil parent material and climate have resulted in the development of a unique set of nitrogen BMPs for each region,²⁸ which guided the assumption that fertilizer use will be similar for counties within each BMP region. The reallocation of fertilizer sales resulted in substantial differences between original and reallocated sales at a county level.

However, what appeared to be even larger discrepancies existed prior to reallocation. MDA describes such a discrepancy,²³ with Brown County (182,500 acres of cropland) reporting 106,000 tons of fertilizer sold, compared to neighboring Redwood County, with less than half the fertilizer sales but nearly twice the amount of crop acres. Winona County and neighboring Olmsted County represent another example. Commercial N fertilizer sales in Winona County (103,100 acres of cropland) were reported as 10,808 tons, compared to 5,993 tons of N sold in neighboring Olmsted County, with nearly double (197,732) the cropland acres.

As a result of aggregation, all counties within each BMP region will have the same percentage N or P need met by commercial fertilizer. The availability of manure nutrients at a county level was not considered in the reallocation of commercial fertilizer sales, which may improve future estimates. Additional information on fertilizer use at a sub-county level, which is scarce, is necessary for understanding fertilizer use within counties and, ideally, within agricultural fields.

Nitrogen

2016 commercial N farm-use fertilizer sales in Minnesota totaled 779,078 tons. Remarkably, at a statewide scale, commercial fertilizer N sales almost perfectly match (100.01 percent) the statewide N requirement of all cropland. That is, 100 percent of state crop N recommendations are met by commercial fertilizer alone, before accounting for any land applied manure.

The proportion of county N need satisfied by commercial fertilizer varies across the state (Figure 6a). Only the southwestern and central BMP regions can supply less than 100 percent of crop N need through commercial fertilizer sold. These regions also have the highest percentage of N need met by manure, suggesting some negative correlation between the availability of manure nutrients and commercial fertilizer sales. The southeast BMP region had the highest percentage of N need met by commercial fertilizer alone, at 122 percent, whereas the southwest BMP region had the lowest percentage at 87 percent.

Phosphorus

2016 commercial phosphate (P205) farm-use fertilizer sales in Minnesota totaled 320,496 tons. This value was multiplied by 0.44 to convert from P205 to elemental P, or 141,018 tons of commercial P.

Figure 6b shows the proportion of county P need satisfied by commercial fertilizer. This is highest for northwest counties, where manure P is the lowest (less than 25 percent of P needs). The southern half of the state is steady at around 60 percent of county P requirement satisfied by commercial fertilizer. Commercial P sales are lowest in the central part of the state (40 to 50 percent of P removal), where manure P is the highest, again suggesting some negative correlation between the availability of manure nutrients and commercial fertilizer sold.

Figure 6: County Nitrogen (L) and Phosphorus (R) Need Met by Commercial Fertilizer

Source: EWG via Minnesota Department of Agriculture 2016 Crop Year Fertilizer Sales Report.

Commercial Fertilizer Plus Manure

Nitrogen

At a statewide scale, results suggest N application may be occurring at rates far above what is recommended by agency guidelines. When combining commercial N sold (779,078 tons) with manure N applied (234,519 to 272,161 tons), results show between 239,320 and 276,962 tons of N applied in Minnesota above agency recommendations. This is 31 to 36 percent above the economically optimum needs of all crops in the state, which is estimated at 774,277 tons. This is significantly higher than previous agency estimates of 10 to 15 percent (approximately 100,000 tons), which relied on extrapolating 2014 survey data to a statewide scale.

Figure 7a shows the county distribution of total N applied from manure plus commercial fertilizer sources relative to crop needs. Three counties (Grant, Chippewa and Washington) fall below 100 percent, and an additional 29 counties fall between 100 and 125 percent of crop N need. Twenty-six counties show applied N between 125 and 150 percent of N need, 11 counties between 150 and 175 percent, and three counties (Martin, Winona and Houston) show N applied between 175 and 200 percent of crop N need. The 13 counties with total N applied greater than 50 percent of N recommendations are shown in Table 11, below. Clearwater County was excluded because of the very low amount of cropland in the county, around 7 percent of total county acreage.

Table 11: High-Risk Counties for Excess Nitrogen From Manure Plus Commercial Fertilizer

Source: EWG via Minnesota Pollution Control Agency, USDA-ARS Agricultural Conservation Planning Framework Database, Midwest Plan Service, University of Minnesota Extension and Minnesota Department of Agriculture.

Although results suggest major N overapplication in some counties, the number of crop acres must also be considered. Percent N overapplication may be more pronounced in counties with less cropland, and absolute tons of N excess may be less than in counties with more cropland yet a lower percentage of N overapplication.

In Winona and Houston counties, for example, two of the highest counties for percent N overapplication, total cropland acreage is estimated at 103,000 acres and 69,000 acres, respectively. This compares to Martin County, with more than 380,000 acres of cropland with an N fertilizer need. When comparing absolute tons of estimated N overapplication, Winona and Houston counties show an excess of 4,977 and 2,476 tons, respectively, compared to 14,368 tons in Martin County.

It should be noted that of the 13 counties identified with N applied at more than 50 percent above recommendations, only five showed reduced commercial N sales following a reallocation of sales within each BMP region. This suggests that the presence of manure nutrients is likely a factor when reallocating county-level sales and may result in an overestimation of commercial N in counties with high levels of manure nutrients.

Six counties showed moderate increases in commercial fertilizer N sales following reallocation, ranging from a 12 to 45 percent increase. Todd and Wabasha counties showed increases in commercial N sales of more than 70 percent following reallocation. Winona, Goodhue, Waseca, Rock and Stearns counties showed decreased levels of commercial N sales following reallocation at the county level.

Phosphorus

The sum of P removal from all cropland in the state is estimated to be 221,663 tons. When combining 141,018 tons of commercial P sold with the estimated 77,552 to 87,808 tons of manure P applied in Minnesota each year, this indicates a relative balance of P inputs (218,570 to 228,826 tons) and P outputs from crop harvest (221,663 tons) at a statewide level. However, county P budgets vary substantially.

Figure 7b shows the county distribution of total P applied from manure plus commercial fertilizer sources relative to crop needs. Well over half of all counties show a P deficit, where more P is removed than applied on an annual basis. Twenty-five counties show a likely P excess, where more P is applied than removed. Fifteen of these counties fall between 100 and 125 percent of crop P need, six counties between 125 and 150 percent of crop P need, and three counties (Winona, Stearns, and Todd) fall between 150 and 175 percent of crop P need. One county (Morrison) falls at just over 200 percent of P need. Counties with excess P are usually characterized by dense concentrations of poultry and dairy operations, whose manures are higher in P content than other animal types.

Figure 7: County Nitrogen (L) and Phosphorus (R) Need Met by Manure Plus Commercial Fertilizer

Source: EWG via Minnesota Pollution Control Agency, Midwest Plan Service and Minnesota Department of Agriculture 2016 Crop Year Fertilizer Sales Report.

Conclusion

As the animal agriculture industry grows, questions surrounding the capacity of the landscape to uptake manure nutrients will become increasingly relevant. At a minimum, state agencies should evaluate alternative management scenarios using a cumulative, spatial approach such as that presented in this report. Additional focus should be placed on total fertilizer application from both commercial and manure sources in areas with high livestock density, which will require improved spatial information on fertilizer application at a sub-county level. Such analyses are necessary to allow for future adaption of livestock and cropping systems in Minnesota that may be necessary to address increasing water quality concerns.

References

1. ¹ Minnesota Pollution Control Agency. Minnesota’s Draft 2020 Impaired Waters List. St. Paul, Minnesota, 2019. Available Online: https://www.pca.state.Minnesota.us/water/Minnesotas-impaired-waters-list.
2. ² Long, C.M.; Muenich, R.L.; Kalcic M.M.; Scavia. D. Use of manure nutrients from concentrated animal feeding operations. J Great Lakes Res. 2018, 44(2), 245–252.
3. ³ Minnesota Pollution Control Agency. Feedlots in Minnesota; St. Paul, Minnesota, 2019. Available Online from the Minnesota Geospatial Commons: https://gisdata.Minnesota.gov/dataset/env-feedlots.
4. ⁴ Office of the Revisor of Statutes, State of Minnesota. Minnesota Administrative Rules; Chapter 7020, Animal Feedlots; Minnesota, 2019. Available Online: https://www.revisor.Minnesota.gov/rules/7020/.
5. ⁵ Environmental Protection Agency. Regulatory Definitions of Large CAFOs, Medium CAFOs, and Small CAFOs. Available Online: https://www3.epa.gov/npdes/pubs/sector_table.pdf.
6. ⁶ Porter, S.A.; James, D.E. Using a Spatially Explicit Distribution Model to Assess the Contribution of Animal Agriculture to Minnesota’s Agricultural Nitrogen Budget. Agronomy. 2020, 10(4), 1–15.
7. ⁷ Minnesota Pollution Control Agency. Nitrogen in Minnesota Surface Waters. Conditions, Trends, Sources, and Reductions. St. Paul, Minnesota, 2013.
8. ⁸ Lorimor, J.; Powers, W.; Sutton, A. Manure Characteristics; MWPS-18, Section 1; Midwest Plan Service: Ames, Iowa, 2004.
9. ⁹ Wilson, M. Manure Characteristics. University of Minnesota Extension: St. Paul, Minnesota, 2018. Available Online: https://extension.uMinnesota.edu/manure-land-application/manure-characteristics.
10. ¹⁰ Loubet, B.; Asman, W.A.H.; Theobald, M.R. Ammonia deposition near hot spots: processes, models and monitoring methods. Atmospheric Ammonia. 2009, 205–267.
11. ¹¹ Minnesota Department of Agriculture; University of Minnesota Extension. Nutrient and Manure Management Tables; St. Paul, Minnesota, 2012.
12. ¹² Minnesota Department of Agriculture. Commercial Nitrogen and Fertilizer Selection and Management Practices Associated with Minnesota’s 2014 Corn Crop: St. Paul, Minnesota, 2017.
13. ¹³ Sawyer, J.E.; Mallorino. Using Manure Nutrients for Crop Production; Iowa State University Extension: Ames, Iowa, 2016. Available Online: https://store.extension.iastate.edu/product/Using-Manure-Nutrients-for-Crop-Production
14. ¹⁴ Andersen, D.S.; Pepple, L.M. A County-Level Assessment of Manure Nutrient Availability Relative to Crop Nutrient Capacity in Iowa: Spatial and Temporal Trends. Trans. of the ASABE. 2017, 60, 1669–1680.
15. ¹⁵ Pettygrove, G.S.; Heinrich, A.L.; Crohn, D.M. Manure Nitrogen Mineralization. University of California Cooperative Extension: University of California, Davis, California, 2009. Available Online: http://manuremanagement.ucdavis.edu/files/134367.pdf.
16. ¹⁶ Sutton, A. L.; Jones, D.D.; Joern, B.C.; Huber, D.M. Animal manure as a plant nutrient resource. ID-101. Cooperative Extension Service, Purdue University, West Lafayette, Indiana. 1999
17. ¹⁷ North Central Region Water Network. Agricultural Conservation Planning Framework. 2019. Available Online: https://acpf4watersheds.org/.
18. ¹⁸ Sawyer, J.; Nafziger, E.; Randall, G.; Bundy, L.; Rehm, G.; Joern, B. Concepts and Rationale for Regional Nitrogen Rate Guidelines for Corn; Iowa State University Extension: Ames, Iowa, 2006.
19. ¹⁹ Minnesota Pollution Control Agency. Manure Nitrogen Rates for Corn Production; St. Paul, Minnesota, 20
20. ²⁰ Kaiser, D.E.; Lamb, J.A.; Eliason, R. Fertilizer Guidelines for Agronomic Crops in Minnesota; University of Minnesota Extension: St. Paul, Minnesota, 2011.
21. ²¹ USDA-NASS. County Agricultural Production. United States Department of Agriculture-National Agriculture Statistics Service: Washington, D.C., 2019.
22. ²² USDA-ERS. Weights, Measures, and Conversion Factors for Agricultural Commodities and Their Products. United States Department of Agriculture-Economic Research Service: Washington, D.C., 1992.
23. ²³ Minnesota Department of Agriculture. 2016 Crop Year Fertilizer Sales Report. St. Paul, Minnesota, 2017.
24. ²⁴ Spellman, F.R.; Whiting, N.E. Environmental Management of Concentrated Animal Feeding Operations (CAFOs). CRC Press, 2007. 1–496.
25. ²⁵ Modderman, C. Avoiding Phosphorus Buildup in Soil from Turkey Manure. University of Minnesota Extension: St. Paul, Minnesota, 2019.
26. ²⁶ Bly, Anthony. Manure Management Rate Effects on Soil Health in South Dakota. South Dakota State University Extension. 2014. Available Online: https://www.manureexpo.org/uploads/4/1/3/4/41345563/4._manure_management_rate_effects_on_soil_health_in_sd.pdf
27. ²⁷ Wilson, M. New Manure Application Rate Guidelines for Minnesota. University of Minnesota Extension: St. Paul, Minnesota, 2019.
28. ²⁸ Lamb, J.; Randall, G.; Rehm, G.; Rosen, C. Best Management Practice for Nitrogen Use in Minnesota. University of Minnesota Extension: St. Paul, Minnesota, 2008.