On-farm crop tests can be powerful tools for individual farmers
Simple, low-cost trials on your own farm can help you decide if a new practice will work for you.
November 21, 2013 - Author: Jim Isleib, Michigan State University Extension
Every new product, seed variety, fertilizer or plant health product looks great in the company advertising, and new technology developed at universities may have a lot of potential. But will it really do all that on your farm?
A key figure in the history of technology transfer to farmers is Seaman A. Knapp, the father of “farm demonstration.” He developed and promoted farmer demonstration, mostly in the southern United States, during the 1880s until about 1910, before the establishment of the Cooperative Extension Service. Knapp said, "What a man hears, he may doubt; what he sees he may also doubt; but what he does, he cannot doubt." American agriculture and the American farmer have changed dramatically since then, but there is still value in this approach.
With some basic planning and maybe some assistance from a local Michigan State University Extension educator, conducting your own on-farm project is possible. It will take some extra time at planting and harvest when you may be hurrying to get your normal work done, so it should be thought through thoroughly. Possible projects might include testing new crop varieties, fertilization products or practices, plant health products or other agronomic practices.
“Demonstration” of a practice or product implies that no scientific analysis is done on the resulting information. A non-replicated “side-by-side” trial is a common example of this approach. Demonstrations have value and can help with decision-making, but are not as convincing as a replicated trial.
To have much validity, even a simple on-farm test should be replicated at least three times. A higher number of replications is better. That is, each “treatment” – seed variety, fertilizer rate, fungicide – should be repeated at least three or four times. The treatments should be randomized within each replication. This can be done without sophisticated equipment using the “draw the name from the hat” technique. Most on-farm trials are planted in strips. If you have four seed varieties you wish to compare for yield, then 4 varieties x 3 replications = 12 strips.
Strips must be wide enough to accommodate your harvest equipment and you need a good way to measure product harvested from each replication. A weigh wagon with built-in scale and auger is an ideal tool if one is available locally. Your local MSU Extension educator or the nearest elevator may know where to find one. The site of a trial should be on as uniform a field as possible. Avoid changes in soil type or management history. Care must be taken to avoid mixing seed during planting. Cleaning out a drill or planter between varieties takes time.
At harvest time, it is essential to be ready to record information as strips are harvested. If your project is a demonstration, your harvest data is all you need. If you try a replicated on-farm trial, you will need to have your information analyzed using statistics. This can be a little intimidating, but a simple online tool designed especially for on-farm trials called AgStats can be used by those with limited knowledge of statistics.
The AgStats program was developed by Oregon State University, Washington State University and the University of Idaho for the Pacific Northwest Conservation Tillage Systems Information Resource. It is somewhat dated, but still works fine. You simply select the number of replications and treatments, type in a title, select a “probability level,” usually 5 percent for strong statistical results, which means you are 95 percent sure that the difference is due to treatment effects and not to natural variability. The program allows choices up to 20 percent, or 80 percent sure. The AgStats program contains good background information about on-farm testing.
Early decisions for those interested in on-farm tests:
- Am I willing to make time for this?
- Do I have access to good equipment for the things I want to test?
- Do I have a good way to measure yields?
- Is there a local resource person, like MSU Extension, I can work with on this?
- Should I consider an on-farm meeting to go along with the project?
- Should I look for outside funds like NCR-SARE to help pay for the project expenses?
In my own personal experience working with farmers in the Upper Peninsula, I have helped farmers with on-farm trials on oat and barley varieties, perennial forage species and mixtures, use of industrial hardwood ash, fungicide use on oats and other projects. Every project doesn’t work perfectly. On-farm trials are subject to the same challenges as normal farm crop production and are more vulnerable to problems that will negatively affect final results.
There is merit in repeating a test over a number of years for consistency. The information gained can make a real difference and there is a lot of satisfaction in completing the on-farm project. Sharing this information with other farmers may add to this satisfaction.