SWINE AND DAIRY WASTE MANAGEMENT USING PRECISION AGRICULTURE

Scott A. Shearer, Biosystems and Agricultural Engineering

 

Richard I. Barnhisel, Agronomy
Joseph L. Taraba, Biosystems and Agricultural Engineering
Stephen F. Higgins, Biosystems and Agricultural Engineering

 

The intent of this project is to apply precision agricultural practices to the management and land application of animal nutrients. The main thrust of the work will be conducted at Worth and Dee Ellis Farms in Shelby County, Kentucky with other operations to be involved as management tools are developed. The Worth and Dee Ellis Farms site was selected because of the integrated swine, dairy and crop production enterprises. The project will initially involve identification of land resources, and the distribution and quantification of nutrient levels. These data will be stored in a Geographic Information System (GIS) database. A series of application rules will be developed to identify waste application rates and volumes. The rules are intended to limit phosphorus levels based on local soil test values, or nitrogen based on the needs of the subsequent crop. In latter phases of the project waste application uniformity will be address, and the nutrient content of the injected waste will be assessed in an effort to move towards variable-rate application of waste. In addition, near surface water quality will be assessed using suction lysimeters placed according to soil type, soil test values, and slope.

• Assess and quantify the variability in nutrient content of swine and dairy waste at the time of application.
• Develop calibration procedures and assess the uniformity of distribution of liquid-manure injection application equipment.
• Develop a prioritized set of rules for precision agriculture mapping packages to aid in selection of land areas within the farmstead that are best suited to handle the nutrient loading associated with animal waste application to minimize impacts on water quality.
• Quantify water quality impacts of traditional versus "site-specific" approaches to animal waste application.

Work in the first year of the project focused on developing a GIS database and rules to govern waste application at the Worth and Dee Ellis Farm site. The Worth and Dee Ellis site in Shelby County is comprised of approximately 4300 acres of grain crops, and swine and dairy enterprises. Approximately 1,000,000 pounds of milk is marketed each year through the dairy operation. The farrow to finish swine operation produces approximately 4,000 market weight hogs per year. Both animal operations utilize manure storage pits integral to the animal production facilities. A nearby earthen storage facility, design by NRCS, is also available to store animal wastes. Recently, NRCS designed and cost shared a lagoon to be used for storing dairy waste. Pits located under the dairy facility will be phased out in favor of the lagoon. Prior to field application, the pits are agitated to insure that the solids are suspended, and to provide a homogeneous slurry. All animal waste is injected to provide a source of nutrients to grain and silage crops, reduce run-off, and to minimize odor and volatilization of nutrients. Fields are selected for application in accordance with crop rotation and proximity to the animal enterprises. In cooperation with the farm operators the following fields were identified as potential sites to receive injected waste. Included as potential application sites are Fields 16, 21-42, 44 and 45. These 25 fields include a total of 1151.4 acres of cropped land.

Field boundaries were driven using an ATV equipped with Differential Global Positioning System (DGPS) receiver (Onmistar model 7000). The DGPS receiver used wide area differential correction. The system accuracy was reported to be 2 to 3 meters horizontal for 95% of position fixes. Once the boundary was established a one-acre grid (208 feet square grid) was positioned over the boundary maps. The technician then navigated to the grid intersections. Composited soil samples were collected at the grid points by pulling five soil cores at a 15 feet radius of the ATV to a depth of four inches. The four-inch sampling depth was selected for consistency with AGR-1 recommendations and no-till farming practices. The soil cores were placed in a sample box and labeled with the grid point identifier from the FieldLink software (Agris, Inc. Roswell, Georgia). The soil samples were submitted to the University of Kentucky Regulatory Services for Test 1 with optional organic matter analysis.

Yield monitoring continues to be a significant portion of the effort undertaken at Worth and Dee Ellis Farms. Three combines have been equipped and instrumented with mass flow sensing devices in the clean grain elevators and moisture sensors in either the bin-loading auger or clean grain elevator. DGPS capabilities are provided using Omnistar 7000 model receivers with wide-area C-band correction. Yield monitoring data is downloaded from the combines at harvest and transferred to the farm office personal computer (PC) for processing. Expanded files are imported into AgLink version 5.3 (Agris, Inc, Roswell, Georgia) for display and location verification. Data from various combines were merged within the AgLink package to form yield maps on a field basis.

A GIS database was constructed utilizing soil grid sample analyses results for the cropland surrounding the animal enterprises. The GIS was constructed using SSToolbox (SST Development Group, Inc.) which is based on the ArcView GIS engine (ESRI, Inc.). In keeping with traditional agricultural mapping practices all data were entered on a field basis. The agronomic data entered in this database included soil tests for phosphorus, potassium, organic matter content, soil water pH, soil buffer pH, zinc, magnesium and calcium. Additional items of importance that were mapped included state highways, county roads, waterways, and city corporation limits. Features remaining to be mapped include sinkholes, water dens, springs, ponds and other environmentally sensitive areas.

A set of rules for applying animal wastes were developed based on environmental policies in the Commonwealth of Kentucky (Bowden, 1978; and Hoag and Roka, 1995) and the nutrient content of the waste (Collins et al., 1995; Sutton et al., 1979; Krider, 1995; and Wells, 1996). The rules were established based on historical cropping and manure application, and the need to reduce impacts on water quality. Application rules were constructed in variable and fixed rate formats. As application technology evolves, variable-rate waste application will become commonplace. For many Kentucky producers the rules will be used to select application areas within a field that lend themselves to dead reckoning for navigation. That is to either flag the field using GPS, or provide a map with landmarks that will guide the applicator.

Much of the cropland at Worth and Dee Ellis Farm has been instrumented with weirs and automated water quality sampling equipment under the Phase II of S.B. 271. We will continue to utilize this equipment to assess and monitor "site-specific" animal waste management versus practices currently in place. The geology of the Outer Bluegrass soils lends themselves readily to monitoring as much of the near-surface ground water becomes surface water as it flows from wet weather springs in the area. A combination of the near-surface and surface water flows over the weirs that are in place. Storm events will be sampled for N and P levels, with monitoring continuing at periodic intervals as long as flow is present. As an indication of long-term nutrient movement suction lysimeters are being installed at 35 and 70 cm depths to permit monitoring of near surface water quality from fields where animal wastes have been or will be applied. Areas of high and low fertility will be selected to contrast the mineralization or volatilization of N and P nutrients. A total of 18 lysimeter sites in each field will be selected for monitoring.

Grid soil sample results are reported in Table 1 for the study area identified at the Worth and Dee Ellis Farms. Of particular interest is the soil test P (phosphorus) levels in Fields 16, 23, 36, 37, 39, 40, 41 and 42. These fields consist of areas that are low in soil fertility when considering the P soil test levels, and therefore are excellent candidates for animal waste application. Furthermore, Fields 21, 32, 38, 44 and 45 have some potential as at least a portion of these fields are below the threshold P level of 60 in accordance with AGR-1 (University of Kentucky, 1998). The correlation between the field averages of soil test P and K (potassium) values was found to be only 0.213. This indicates that it will be necessary to look at soil test K levels as another rule for the GIS classification scheme. Although, average soil test P level will give a good indication of which fields offer potential for accepting swine and or dairy waste. Obviously a combination of soil test P level and haul distance must be considered when identifying the best location.

Rules were developed to guide producers at Worth and Dee Ellis Farms in the application of animal wastes. The foundation of the rules coincides with Federal and State (including emergency swine) regulations for setbacks, AGR-1 (1999) for nutrient application levels, and, as a last resort, on a mass balance relationship to coincide with crop removal of P and K. The primary nutrient components considered in this rule-based approach were phosphorus and nitrogen.

Federal laws are applied by region with Kentucky being assigned to Region 4. Kentucky statutes tend to be more stringent, especially in the case of recently enacted swine regulations; Emergency 401 KAR 5:009E. These pertain to "swine-feeding operations" where the number of "animal units" exceeds the equivalence of 1,000 or more swine units. While the swine feeding operation at Worth and Dee Ellis Farms does not fall under these regulations, if the owners plan an expansion of 10% or more of the existing operation, they must abide by the emergency regulations. Unique to these regulations, and aside from the requirement that producers follow sound agronomic practices, are the setbacks for waste injections. The setbacks are as follows: 1,500 feet to any dwelling not owned by the applicant; 3,000 feet to incorporated city limits; 150 feet to lakes, rivers, or karstic feature; 150 feet to the property line; 1 mile to downstream water as listed in 401 KAR 5:030 as other than use protected; and 5 miles to downstream public water supply surface water intake. These setbacks apply to injection. For other surface application methodology, these values are increased by between 50 and 100 percent. Although these setbacks are not directly applicable, they will remain as part of the rule-based approach to site selection for waste application.

Land application of waste via injection will be governed by phosphorus and nitrogen content of the waste, local soil test levels, AGR-1 recommendations, and setbacks as specified above. In the event soil P levels are considered "high," waste will be applied in accordance with crop removal rates. From Ohio State (1998), the nutrient levels associated with crop removal are summarized in Table 2. An indication of nutrient content of swine and dairy waste is given in Table 3 (Ohio State, 1998). Although the latter table lists nutrient contents, it must also be recognized that nitrogen and phosphorus may not be readily available for crop use.

Baker (1996) estimated that 95% of the soluble form of nitrogen (urea) is available to the crop in the first year for injected manure. However, only one-third of the organic nitrogen is available in the first year. If the manure is stored in a lagoon, the available organic nitrogen is increased to 50%. Thereafter, the remaining organic matter is made available at the rate of 5% per year. Nearly all of the P and K in stored animal waste is available to the crop in the first year after application. Baker (1996) and Wells (1996) reduce these values to 80% of the total when liquid manure is stored in a lagoon and applied via injection.

Figure 1 depicts the relationship of the fields identified by the farm operators as being suitable for animal waste application. Boundaries of individual fields or production units are outlined in black and filled in dark gray. Field numbers are used to label each unit. Field 41, lower center of the farm map, was selected to illustrate the application of rules for P management.

Figure 2 shows a gray scale map depicting crop performance. The smaller grid cells have dimensions of 54 feet by 54 feet. Each cell represents 0.067 acres, or roughly one-sixteenth of an acre. The yield data file contained in excess of 28,000 data points that were reduced by averaging the yield data values over the 54 feet by 54 feet cells. This reduced the data set to approximately 1,400 data points, each associated with a single cell. This grid resolution was maintained for the remaining analysis.

Eighty-nine grid point locations were soil sampled within Field 41. Using a linear interpolation technique, each of the 54 feet by 54 feet cells were assigned a soil test P value. This rasterized version of a soil test P map was then used to estimate the quantity of swine waste to be applied. Data for the corn P₂O₅ recommendations in AGR-1 (1999) were regressed to determine a continuous relationship that could be used as a rule to guide phosphorus application. The relationship was as follows:

where W_P2O5 is the weight of P₂O₅ to be applied per acre and P_{Soil Test} is the soil test value as determined using Mellich III extraction. Using this relationship and 80% availability, and a P₂O₅ content of 27.0 lb/1000 gallons of swine waste as pumped from storage pits, the application map in Figure 3 was generated. Integrating the application rate over the surface area of the field results in a total applied waste volume of 134,471 gallons, or 39.6 slurry wagon loads of manure. This represents approximately one-half of the annual waste production from the finishing floors of the swine operation.

A second approach might be to limit waste application to crop removal rates. Assuming that the '98 corn crop in Field 41 extracts 0.37 lbs of P₂O₅ per bushel of grain harvested, Figure 2 can be changed to a swine manure application rate map as shown in Figure 4. Again, integrating the swine waste application rate over the field area results in a total applied waste volume of 136,064 gallons, or 40.0 slurry wagon loads of swine waste. This approach is less desirable than applying in accordance with AGR-1 as suggested in Figure 3. The reality is that "low" soil test P areas will remain low using this approach. However, application in accordance with crop removal should limit any build-up in soil test P levels.

This example illustrates the potential of GIS to solve waste management problems associated with the swine and dairy enterprises at Worth and Dee Ellis Farms. The intent will be to place animal waste in locations in need of additional nutrients. This should also address the concerns relative to water quality as N application will be limited to crop needs while P application will be limited by crop removal for high fertility regions, or by AGR-1 when soil test P levels fall below the threshold value of 60.

Data collected to date illustrates the opportunities for Worth and Dee Ellis Farms to manage animal nutrients similar to the way inorganic sources of nutrients are managed using precision agriculture. Grid sampling has revealed opportunities for a return to the cropping enterprise based on the identification of those areas of a field best suited for animal waste application. The rules-based approach to identification of these areas and specification of the quantities of waste that the land will accept based on limiting factors including infiltration, slope, soil type, and productivity potential, are easily implemented in a GIS. From the GIS software pacakage, application rate maps can be generated to guide producers in application using dead reckoning, or generation of application rate maps for automatic control of waste application. The remaining problem to overcome is the specification of nutrient content of animal waste to be applied. Investigation of nutrient content of animal wastes and ground water quality will provide the basis for subsequent work to be conducted in this project.

References

Bowden, J.P. 1978. Livestock Waste Management – Questions and Answers Concerning Laws and Regulations. Extension Publication AEN-44. College of Agriculture, University of Kentucky, Lexington.

Baker, J.C. 1996. Livestock Waste Sampling, Analysis and Calculation of Land Application rates. North Carolina Cooperative Extension service. Publication Number EBAE 111-84.

Collins, E.R., Jr., J.D. Jordan and T.A. Dillaha. 1995. Nutrient values of dairy manure and poultry litter as affected by storage and handling. Animal Waste and the Land-Water Interface. pp. 343-353.

Hoag, D.L. and F.M. Roka. 1995. Environmental policy and swine manure management: waste not or want not? American Journal of Alternative Agriculture. 10:163-166.

Krider, J.N. 1995. Innovative utilization of animal waste. Proceedings of the National Livestock, Poultry and Aquacultural Waste Management Conference. pp. 82-87. St. Joseph, Michigan: ASAE.

Ohio State University. 1998. Ohio Livestock Manure and Waste Water Management Guide. Bulletin 604. The Ohio State University, Extension Service.

Sutton, A.L., D.H. Vanderholm, and S.W. Melvin. 1979. Fertilizer value of swine manure. Pork Industry Handbook. Extension Publication No. ASC-80. College of Agriculture, University of Kentucky, Lexington.

University of Kentucky. 1998. 1998-1999 Lime and Fertilizer Recommendations. University of Kentucky Cooperative Extension Service. Publication number AGR-1.

Wells, K.L. 1996. The agronomics of manure use for crop production. Extension Publication No. AGR-165. College of Agriculture, University of Kentucky, Lexington.