Omission Research

Today’s high soybean yields that some producers are achieving have caused heightened interest in how to raise yields from lower yielding sites to the level of the higher yielding sites.

One technique that can be applied in research studies to address this issue is an omission trial treatment structure (Ruffo et al., Agronomy Journal, 2015; Univ. of Ill.), or a treatment structure that contains systems of production where input factors are separately withheld from a high technology (HT) production system or are separately added to a low technology (LT) production system. In a trial or research study that uses this technique, the treatment controls are the HT system having all designated input factors applied, and the LT system having all of the same input factors withheld.

This treatment structure is designed to allow three comparisons: 1) LT vs. HT controls; 2) effect of the addition of selected individual treatments or input factors vs. their counterpart LT control; and 3) effect of the exclusion of selected individual treatments or input factors vs. their counterpart HT control. To evaluate the effect of a factor, the effect of that factor removed from the HT production system is compared to the HT control that contains all other designated factors, and conversely, the effect of a factor added into the LT system is compared to that system where none of the other factors are added.

The table below shows how this setup appears when the same five treatments or input factors are used in both HT and LT production systems.

 





































































































































Example omission trial design with five input factors.
Input Factors  
System*    A    B    C    D    E Cost
HT YES YES YES YES YES ??
HT - A NO YES YES YES YES ??
HT - B YES NO YES YES YES ??
HT - C YES YES NO YES YES ??
HT - D YES YES YES NO YES ??
HT - E YES YES YES YES NO ??
LT  NO NO NO NO NO ??
LT + A YES NO NO NO NO ??
LT + B NO YES NO NO NO ??
LT + C NO NO YES NO NO ??
LT + D NO NO NO YES NO ??
LT + E NO NO NO NO YES ??
*HT = high technology system; LT = low technology system.


 

Note in the above table there is a column for cost associated with each input factor. This should always be accounted for since each factor will have a cost associated with its exclusion from the HT system or its inclusion in the LT system. Thus, any response of a system to the inclusion or exclusion of a factor must be evaluated in relation to the cost of that factor.

An example experiment where this technique was used was conducted in Ohio, and the results are reported in Agronomy Journal, Vol. 107, Issue 5 in 2015. The article is titled “Soybean yield response to Rhizobia inoculant, gypsum, manganese fertilizer, insecticide, and fungicide”, and is authored by Bluck et al. More details about the research are contained in a thesis entitled “Soybean yield response in high and low input production systems”.

The Ohio experiment was designed to evaluate the effect of selected inputs (Rhizobia inoculant, gypsum for sulfur, manganese foliar fertilizer, and foliar insecticide and fungicide applications) on soybean seed yield in HT and LT production systems. Interestingly, there was little response to the subtraction (HT) or addition (LT) of any of the selected input factors in this study. These lack of responses are attributed to there being either no deficiencies that were being targeted by the inputs (inoculant, sulfur, Mn fertilizer), or no conditions that were sufficient for remediation responses (foliar insecticide and fungicide applications).

Thus, it is important in this type of research that 1) experimental sites are selected that in fact have deficiencies that will be targeted by input factors such as fertility amendments, or 2) a field history and/or crop scouting history have determined that remedial measures will most likely be required to effect a response to input factors such as foliar insecticide and fungicide applications.

For this research to be effective, special attention must be given to ensuring that all production factors in the experiment other than those chosen for addition or subtraction are either present or applied at the level required for expected production from each system. Otherwise, a true measure of the effect of the management factors that are the most influential in achieving crop yield potential may not be obtained.

Composed by Larry G. Heatherly, Sept. 2015, larryheatherly@bellsouth.net