On-Farm Research Using Yield Monitors

Kurt LawtonEducation, Harvesting, Research

Watching those yield numbers register as data on a map as you go round by round during harvest will help aid decision making this winter. To this end, John Fulton, Extension Specialists, Biosystems Engineering, Auburn University, wrote a good piece on how to use such data on the Alabama Cooperative Extension System’s Precision Ag Blog.

While yield monitors or maps can provide beneficial data to make informed farm management decisions, one must ensure they are properly implementing yield monitors so the correct decisions are made. Yield monitors are excellent tools to conduct field-scale research and spatially document yield differences across fields. Yield maps can identify issues within fields where low yields exist, support the implementation of site-specific management, and make comparisons between varieties, seed treatments, and new management practices.  However, data needs to be collected over whole fields or several fields in order to obtain a sufficient amount of yield data to make the right decision(s) for your farm.

Caution is extended for those using yield monitors to harvest variety trials or plot-scale studies. While this data can provide useful information and support plot-scale work,yield monitors are not a replacement for weigh wagons which should be used to obtain and compute yields for making comparisons in plots. Variety trials and plot studies are generally too small (100, 300 or even up to 900-ft long) in order for yield monitors to accurately measure the accumulated yield over 1 or 2 passes.  The potential errors of yield monitors in these small areas are too high.  Therefore, a measured +/- 5 bushel difference might not really exist between treatments being compared.  While proper calibration is always recommended for yield monitors and the calibration process indicates only a few percent error, the error in yield data generated will be higher for plot size studies and can actually be in the range of 10% to 20%.  For field-scale yield data, these errors will be “averaged” out and the mean yield for a treatment (e.g. variety) can be accurate (1% to 5% error depending upon operating conditions).

As an example, lets say a yield monitor has an actual error of 5%.  If yields are running around 150 bu/ac for the plots, then a 7.5 bu/ac error exists,  Therefore, greater than a 7.5 bu/ac difference must be measured before a variance can be concluded between treatments.  Further, if the actual error is 10% or 20%, then a 15 and 30 bu/ac difference, respectively must exist before any substantial conclusions can be drawn saying one treatment is better than the other.  In any of these cases, one cannot say that one treatment is better than the other if only a 3 or 5 bu/ac difference was measured; a weigh wagon would be required to indicate whether such a few bushel difference.

Remember, while yield monitors can provide useful information for small-scale studies, they are not a replacement for weigh wagons.  Solely relying on yield monitors for providing the necessary performance data at this scale, can lead to incorrectly analyzed yield differences and ultimately the wrong conclusion or management decisions being made. Yield monitors are excellent tools to support management decisions but must be used properly.