Rationale for and Benefits of Sharing on-Farm Data.

Yield monitors provide farmers with a revolutionary technology to rapidly increase efficiencies in crop production practices. Yields can now be measured easily, accurately and inexpensively at the field scale, and because yields are the primary yardsticks for assessing crop practice practices, farmers can assess practices on a routine basis. However, farmers working alone cannot generate sufficient data to increase efficiencies in crop production practices. That is because crop production practices are greatly affected by environmental conditions, and large amounts of data are required to accurately describe the effects of environmental conditions on crop production practices. 

Farmers are cooperating in networks with scientists, crop consultants, agency personnel, and commodity organizations to realize the great potential of yield monitors to improve crop production practices. The Indiana Department of Agriculture’s INfield advantage program and the Iowa Soybean Association’s On-Farm Network are two examples of farmer networks. Networks enable completion of replicated strip trials on a field scale, and quick analysis and summary of the strip trial results for greater learning from the results by farmers. The full potential of yield monitors, however, is being impeded by a lack of clear guidelines for data stewardship, which is the sharing, aggregating and accessing of the huge amount of trial results collected by farmer networks.

Guidelines for data stewardship would greatly increase the value of the results of strip trials. The value would increase because guidelines would facilitate combining strip trial results across farmer networks, which would allow creation of large collections of strip trial results with the associated meta-data. Large collections of results with the meta-data are needed to enable decisions about crop production practices based primarily on data and not primarily on limited data sets that cannot provide adequate solutions to crop production practices, which results in the decisions often being made based on experience, unreplicated trials, and/or expert opinion.

Basing decisions about crop production practices primarily on scientifically robust data is needed for two reasons. First, society is demanding better solutions to problems like excess N and P in water bodies caused by agricultural production; and second, our current methods of research that do not permit the accumulation of results from large numbers of trials evaluating productions practices are woefully inadequate for creating solutions to problems like excess N and P in water bodies. These two reasons are linked. Without large collections of data with thousands of trials across many years, better solutions to problems caused by crop production practices are impossible to create.

Large collections of results from field-scale replicated strip trials are required to create solutions to problems caused by crop production because crop production is fundamentally a plant growth process. Plant growth processes are biological processes driven by interactions among plants, soils, and environmental conditions that are difficult to predict. Due to the complexity of these interactions, massive amounts of data are needed to describe the probability that a new crop production practice will be superior to an existing practice. The only way to obtain the amount of data needed across years and fields is by cooperation among farmer networks to create large collections of the results from their trials with the associated meta-data.