Ensure Success With On Farm Trials

There is an over abundance of research that is done every year and presented to you to prove which product your farm needs to ensure success. Whether it is seed, chemical, agronomic enhancement or a miracle herbal additive, the amount of information is monstrous. There are many private research trials and public trials for all the different types of products that are available to your farm. Many of these trials are regionalized to provide more accurate information for your farm.

The one type of trial that is the best way to find out what works best on your farm is the on farm trial. When doing an on farm trial there are several things that you must remember and keep in mind to ensure successful and accurate results.

  1. Decide what you actually want to accomplish–You must set up your trial for the based on the information you need. If you want to test the drought tolerance of certain varieties then make sure the trial is not on irrigated land.
  2. You must have a check strip–You must have a standardized check to compare your test back to. It is important that the check be relevant to your test and is a fair evaluation of the normal circumstance. The check should be included in this years trial. Comparing a new seed variety to what you did last year is not a fair comparison.
  3. Replicate if possible–More samples is always better than one. If you have the time and energy, your trial results will be considered more accurate if you replicate the trial over a coupe different locations or within the same location.
  4. Test multiple products–Since you are going through the effort to have on farm trials I would recommend that you try a couple different products to see what works best for your farm. Usually there is several competitive products that justify being tested against each other.
  5. Plan to properly collect the data–Data collection is a very important step and takes time. This is usually where most on farm trials fall apart. Harvest time is very busy which can lead to mistakes as it pertains to trials. For example, I have seen where a farmer was doing a yield trial and at swathing time he swathed across the trial instead of parallel to the strips. The trial was destroyed and a lot of work and time had proved to produce no data and benefit.
  6. Ask for financial assistance–There is the possibility of getting financial assistance from the companies that own the product in the trial. There could be a donation of product, financial payment or help in running the trial itself.

There is no better way to make sure that you are using the products that work best for your farm than an on farm trial but you need to make sure that it is done right or the information is misleading and useless to yourself or anyone else.

If you think I have missed a key point please add it in the comments section.

2 thoughts on “Ensure Success With On Farm Trials

  1. I agree that these trials can be valuable tools for analyzing inputs for any operation. I am also excited because many of the newer tools in crop agriculture such as parallel tracking, variable rate technology, satelite imagery, yield monitors, and yield mapping can each make on farm testing easier and more effective.

    One point I would build on is replication and interpretation. Repitition is important so you know if the results are repeatable or just random. I maintain that if you split any field there will be a difference in yield between the two parts and when you conduct a test, you need a mechanism to determine if the difference that you observed is due to the treatment (input you are testing) or just field variability.

    I would also add that replication is of little value without statistics. If you are just averaging the results, you would get a similar result to just splitting the field. You need to look at the variability within the data. This is where statistics can help you determine if the range of outcomes is different that we may expect from a random variable. Statistics will help you interpret your results. You can find a simple tool to do the stats for you at http://pnwsteep.wsu.edu/agstatsweb/index.html.

    I should also point out that it has been my experience that even these classical statistics are flawed in many field situations. The reason for this is that the yield of each strip is largely determined by the productivity of the soil in the strip or the proportion of soils of mixed productivity. Consider just two strips that are testing a product that will yield a 5 bushel increase, but overlay a transitional area with 2 soils that may have a 20 bushel difference in yield potential (no one has seen a yield map with only one number/color). If the untreated contains more of the more productive soil, it will yield more than the treated strip with the 5 bu jump. Be sure to check your previous yield maps or drainage ways that do not run directly across all strips (even in relatively flat fields).

    This leads me to two last points. Field Selection is critical – not every field is suitable for on farm testing. The other is that I expect spatial statistics to overcome these issues in future on farm testing – innovators should look forward to that.

  2. Hate to say it BUT don’t forget about the “Stats”. We all hated the Stats Classes in U but there is a point to “Statistical Significance”. We can get running wild with enthusiasm on “face value” numbers but if those numbers are just “the odds” we could be running down the wrong alley.

Leave a Reply


This site uses Akismet to reduce spam. Learn how your comment data is processed.


Register for a RealAgriculture account to manage your Shortcut menu instead of the default.