What is the Experimental Prescription Design Tool?

As part of a research effort to improve yield and protein of winter wheat for farmers in Montana, a variable rate application software was developed to generate experimental nitrogen trials. These experimental trials follow a random stratification strategy across the field.

How do you use it?

1. Field information

The first step is to provide data regarding the field for which the experimental trial has to be created. This consists of a field name and four different .csv files (* indicates optional files):

• Field name: enables easy identification of fields. Because of its purpose, field names have to be unique.
• Field shape file: tells the application the shape of the field using the designated ‘WKT’ or ‘geometry’ column.
• Existing grid file: if a grid already exists, this allows for that grid to be read in.
• Previous year yield file: yield values read in from the previous year the same crop was applied. This will be used to assign bins to the grid cells.
• Previous year protein file: same as the yield file.

If a field already exists, it can be selected from the dropdown menu. Users are only able to see the fields that they created/own, unless they have administrative privileges.

2. Prescription information

Binning information
The yield and protein values from the previous step are used to determine which “bin” a specific grid cell belongs to. To decide the size of a single bin, two different strategies can be applied:

• “equal yield values”: If the yield values for a field range from 20 – 100, and the user wishes to create 2 yield bins, bin 1 would contain all data points with yield values from 20 to 60, and bin 2 contains those with values from 61 to 100.
• “equal data points”: If the bins are created based on the number of data points in each bin and there are 100 data points available, both bins would have to contain 50 data points, regardless of the yield values.

The resulting bins for each grid cell are then used to assign stratified experimental rates across the bins.

Nitrogen values

Grid shape information

• Cell height and width: To create a new grid to overlay the field
• Field Buffer: How much room should be left between the grid and the field boundary

Optimization

• Minimize rate jumps?: Runs a Genetic Algorithm that minimizes jumps between consecutive cells while maintaining stratification
• As applied fertilizer rates: Needed to determine the order of the cells

What is the Experimental Prescription Design Tool?

The goal of DIFM is to revolutionize farm management, working with farmers and crop consultants to implement scientific experiments on their own farms, enabling them to increase profits by making data-driven management decisions. Farmers and consultants can use this tool to help design their on-farm experiments.

How do you use it?

1. Field information

The first step is to provide data regarding the field for which the experimental trial has to be created. This consists of four different .shp files (.shp,.dbf,.shx,.prj), that give the boundary of your field.

2. Rates Specification

Next, the input type is chosen that you would like to design your experiment for and rate information.

Type of Design

• jcl (jump-conscious latin): Designs the Latin Square so that the rate will not go from the highest rate to the lowest rate. Specifically, does not allow for a jump of more than two rate levels.

  • Number of rates: Can specify the number of rates you desire for an input.

• ejca (extra jump conscious): Able to specify how much they want to restrict the jump in rates as the applicator is driving

  • Max Jump: specification of the maximum you want the jump in rates to be. The ejca will determine the number of rates needed to successfully design a trial when the max jump specified. The number of rates given will be ignored if it does not enable the specified maximum jump

3. AB-Line

Third, if you have them, can upload AB-Line .shp files (.shp,.dbf,.shx,.prj). If you do not have AB-line Files, one can be created by specifying what direction should your created AB-Line run.

AB-Line Type

• Lock: trial design is created so that the plots are alined correctly to the very ab-line provided by the user. It is intended for those who use the ab-line when implementing the trial.

• Free: Unlike the lock option, only the heading of the ab-line is extracted and then trial designs are created following the heading. This option creates an ab-line that is different from the one provided by the user.

• Non: The same as Free, but does not output an ab-line

4. Plot Specification

Next, enter in details about your experimental plots. Please see example plot below for a reference. Minimum and Maximum plot length can be entered in order to maximize the area.

5. Machine Specification

Finally, information about your machine is then entered.

6. Once all necessary information is given, click “Generate Map” to create your trial design.

Once this is created, you will be able to view your map, and download shape files of the map.