Special thanks to Sarah Porter, previously with EWG

Introduction

Adequately addressing water quality issues in Minnesota requires a better understanding of the relative contribution of the major nutrient sources and their distribution.

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. This leads to a wide range of environmental and economic consequences.

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

EWG used geospatial techniques to quantify the capacity of the landscape to uptake manure nutrients while identifying areas of manure saturation in the state.

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.

Animal feedlots

The Minnesota Pollution Control Agency, or MPCA, maintains a database of feedlots1 in the state housing 50 or more animal units. 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).2 The MPCA publishes this feedlot registration as a geospatial dataset that EWG downloaded from the Minnesota Geospatial Commons on September 17, 2024.

The dataset included general information for 23,503 active operations housing a dominant livestock group, including poultry (layers, broilers and turkeys), swine, and beef and dairy cattle.

Modeling landscape capacity For manure nutrients

EWG employed a GIS program developed by Porter and James3 to spatially model the capacity of neighboring cropland to incorporate manure nutrients. Detailed estimates of manure and nutrient excretion were made for all 23,503 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.

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.

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 factors: 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, or MWPS, 18 Second Edition (2004).4 Each type of animal in the Minnesota feedlot database was matched to an animal type from the MWPS (see 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 commonly used to estimate manure nutrient content, book values provide estimates only. The actual characteristics of manure can vary by 30 percent due to genetics, diet and farm management.

A study by Porter and James (2020)3 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, the 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. EWG ran the model using values representing the low end (150 pounds for all grow-to-finish swine) for nutrient values.

But manure nutrient contents can vary widely among all livestock types.5 Ideally, a sensitivity analysis using different parameters would provide a full range of manure application scenarios.

Porter and James (2020)3 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)3 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: Minnesota manure and nutrient excretion values (in pounds/animal/day) as adapted from MWPS-18, 2004.

Animal type MWPS animal type Manure N P Dairy cattle, large Combination lactating (305 days) and dry cow (60 days) (1,400 lb) 142 0.92 0.203 Dairy cattle, small Combination lactating (305 days) and dry cow (60 days) (1,000 lb) 101 0.65 0.145 Dairy heifer Average of 750 and 1,000 lb dairy heifer 53 0.265 0.0396 Dairy calf Average of 150 and 250 lb dairy calf 16 0.085 0.007 Beef steer/stock Finishing cow (1,100 lb) 54 0.4 0.053 Beef feeder/heifer Average of finishing cow (750 lb) and cow in confinement 64.5 0.31 0.057 Beef cow/calf pair Sum of cow in confinement and beef calf (450 lb) 140 0.55 0.119 Beef calf Average of 450 and 650 lb beef calf 58.5 0.245 0.048 Swine, large Average of all swine > 300 lbs 13.4 0.113 0.034 Swine, medium Average of 150 to 300 lb finishing swine (Porter and James, 2020) 11 0.128 0.02 150 lb finishing swine (MPCA) 7.4 0.09 0.0132 Swine, small Average of 25 and 40 lb nursery swine 2.5 0.025 0.004 Turkey, large Male turkey (20 lbs) 0.74 0.0111 0.003 Turkey, small Female turkey (10 lbs) 0.47 0.0078 0.002 Layer, large Layer (3 lbs) 0.15 0.0026 0.0004 Layer, small Broiler, large Broiler (2 lbs) 0.19 0.0021 0.0006 Broiler, small Chicken liquid manure Layer (3 lbs) 0.15 0.0026 0.0004

Source: EWG via Midwest Plan Service

Many cattle feedlots in Minnesota use a combination of confinement buildings and pasture. Manure that is excreted in a pasture is assumed to be 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 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.6 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 guidelines7 developed by the University of Minnesota Extension and the Minnesota Department of Agriculture, or MDA. A 2014 survey of Minnesota’s 2014 corn crop8 was used to inform the prevalence of various manure application methods, with details provided in Porter and James (2020).3 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.9 Nitrogen availability in this study was considered to be the sum of manure N available to the crop during the first, second and third years after application. 7 Similar studies10 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)9 suggest that 90 to 100 percent of swine manure N is available to the crop in the first year after application.

Some second-year N may be taken up by soybeans in a corn-soybean rotation, but 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.10

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. 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.11

As-excreted amounts of phosphorus pentoxide, or P 2 O 5, were multiplied by .44 to convert them 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.12

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.12

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

Animal type % N loss during manure storage % N loss during manure application Dairy 35% 15% if >= 300 AU, 30% if = 300 AU, 30% if < 300 AU Swine 20% 15% Poultry 35% 20%

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

Estimating field nutrient needs

EWG obtained field boundaries with crop rotation history from the Agricultural Conservation Planning Framework, or ACPF, database13 and used them to estimate the average annual nutrient requirement for each field in the state. The ACPF toolbox provided six years of land cover history, from 2018 to 2024. EWG used the ACPF ManureMap Toolbox to identify fields where manure was likely to be applied, based on proximity to feedlots and crop nutrient needs. Using six years of crop history, nitrogen and phosphorus rates were determined using the toolbox.

Commercial fertilizer sales

EWG obtained county-level commercial fertilizer sales from the MDA 2018-2023 Crop Year Fertilizer Sales Reports.14,15,16,17,18 Fertilizer sales for each county were averaged for the five years to normalize sales figures for each county. Then average fertilizer sales for counties in each Minnesota BMP region19 were combined and reallocated based on the average N need determined by the ManureMap Toolbox.13

Manure allocation process

EWG modeled manure allocation from the point of production to neighboring agricultural fields, using the methods detailed in Porter and James (2020).3 Briefly, the program runs as a series of manure application loops, moving outward from each feedlot. Manure is applied 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 individually 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, it is no longer eligible to receive N from any feedlot. Once a feedlot has disposed of all its manure, it is removed from the analysis.

Results

Manure production

The methods described in this report show that Minnesota’s 23,503 active animal operations produce an estimated 48,991,052 tons of manure annually. For comparison to human population, Spellman and Whiting (2007)21 report an estimated 0.518 tons of waste generation per person per year, which puts the Minnesota animal-to-human waste equivalent at approximately 94.6 million people. That is over 16 times the 2024 estimate of Minnesota’s population, 5.79 million humans.

Contribution by size of operation

The manure generated on an annual basis is broken down by size of operation in the 2020 report and this report (see Table 4): small (fewer than 300 AU), medium (300 to 1,000 AU) and large (more than 1,000 AU). Operations larger than 1,000 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, which suggests that manure management for medium-size operations deserves additional focus.

Table 4: Percent of statewide manure production by operation size

Size of operation Percent of all facilities % of total manure % of total manure (2020 report, 150 lb finishing hog) (2024 report, 150 lb finishing hog) = 1,000 AU 4% 30% 33%

Source: EWG via the MPCA and Midwest Plan Service

Contribution by animal type

Table 5 lists the total allowable number of animals in Minnesota by animal type, along with the breakdown of facilities by dominant animal type and 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.

Table 5: Percent of statewide applied manure nitrogen by animal type

Animal type Number of animals % of facilities % of total applied manure N (from MPCA registration) Dairy cattle 1,321,621 17% 31% Beef cattle 2,631,361 57% 32% Swine 11,336,178 19% 28% Poultry 58,949,312 3% 8% Other 349,713 3% .1%

Manure allocation

Nitrogen

Using the calculations described in this report, an estimated 404,095 tons of N are excreted annually by all animal feedlots in Minnesota. Nearly half of this N (172,734 tons) is considered lost to the atmosphere during manure storage and upon field application. What remains is an estimated 231,361 tons of manure N to be applied to cropland in Minnesota each year.

There is ample cropland in the state requiring N fertilizer, with 19,784,374 acres (281,756 fields) having an annual N requirement greater than zero, based on a six-year crop rotation, with a total annual N requirement of 836,304 tons. On a statewide scale, this suggests that manure can satisfy over 27 percent of the crop N requirement in any given year.

Results from the ManureMap Toolbox show manure application to 4,684,390 acres. Manure nutrients can fully satisfy crop fertilizer N needs on these acres.

Commercial fertilizer sales

Commercial fertilizer sales data may not accurately reflect fertilizer use in the county where it was sold. The 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 P 2 O 5 were summed within the state of Minnesota’s Agriculture Best Management Practice (BMP) regions, 14,15,16,17,18 then redistributed back to counties based on each county’s respective share of N and P fertilizer requirement within that BMP region. P 2 0 5 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,19 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. (See Table 6.)

Table 6: High-risk counties for excess nitrogen from manure plus commercial fertilizer

Minnesota county % Commercial fertilizer N % manure N % of total N Need Cass 84.1 74.5 1.587182 Carlton 84.1 72.14 156.2 Morrison 84.1 69.9 154.0 Martin 89.4 61.2 150.7 Winona 100.9 49.6 150.6 Hubbard 84.1 61.4 145.5 Rock 77.6 66.2 143.9 Clearwater 126.8 15.9 142.7 Stearns 84.1 57.6 141.7

Source: EWG via the 2019-2023 MPCA fertilizer sales reports, MPCA, and Midwest Plan Service

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 but a lower percentage of N overapplication.

Animal operations and manure pressure in vulnerable groundwater areas

Minnesota maintains a shapefile of “vulnerable groundwater areas”20 where nitrate can more easily move through soil and into groundwater. There are over 100,000 fields within these vulnerable areas and the model output indicated that almost 33,000 received manure. The Central Sands and Southeast karst regions are especially concerning as those are the areas where we see concentrations of both manure feedlots and vulnerable groundwater areas.

A total of 6,480 facilities are located within the vulnerable groundwater area borders. Another 6,044 are within a mile of a border, meaning over half of the feedlots in Minnesota are within 1 mile of an area of vulnerability. For feedlots within 3 miles of vulnerable groundwater areas, the number grows to 18,355, or 78 percent of the feedlots. Figures 5 and 6 show the CAFO and medium-sized facilities within the vulnerable groundwater areas, along with the vulnerable groundwater areas themselves.

Figure 5. Large facilities by animal type within MPCA vulnerable ground water areas

Image

Source: EWG via MPCA

Figure 6. Medium facilities by animal type within MPCA vulnerable ground water areas

Image

Source: EWG via the MPCA

Conclusion

As the animal agriculture industry grows, questions about 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.