On-farm precision agriculture experimentation allows farmers to better select crop varieties and allocate needed resources to targeted zones in their fields, saving money and increasing profit margins.

Luciano Shiratsuchi, LSU AgCenter associate professor and precision agriculture state specialist, is studying the relationship between plant and soil spatial variability and crop varieties in on-farm trials across the state.

“On-farm precision experimentation featuring different soybean varieties and corn hybrids uses soil electrical conductivity sensors to map soil variability as well as drone imagery to capture crop performance,” he said.

Sometimes the highest-yielding variety ranked No. 1 is not the most profitable, so variety selection should be site-specific based on parameters beyond traditional yield data, Shiratsuchi said.

Because soil variability has to be measured, a greater understanding of soil variability and varietal performance in a trial conducted by the AgCenter precision agricultural team can assist farmers in evaluating key varieties on-farm, he said.

“The best way to extrapolate research results is by participating in an unbiased network of on-farm precision experimentation,” Shiratsuchi said.

The approach differs from traditional small plot research because commercial farm machinery with variable-rate controllers is being used in planting, and yield monitors are harvesting long strips without any flag or stake. To learn more about the network, go to www.onfarmprecisionag.com.

The continuation of the project is vital in providing funds for personnel, small sensors and drone maintenance to maintain relevance of this applied research. The COVID-19 pandemic has resulted in little interference with the project’s on-farm trials because interference in farming practices was nominal and only small labor inputs were needed to plant strips and plots.

“Many farmers often have some e-technology such as sensors and devices that are by default available on modern farm machinery, but they are not using the data that are being recorded,” Shiratsuchi said. Other examples of misuse include owning yield monitors to make yield maps without having the GPS connections to make the maps, data loggers with telemetry not being used to transfer data and in some cases growers paying for an unused cloud account.

Moving forward, Shiratsuchi plans to generate a web database with on-farm precision experimentation beginning with the varieties used by cooperator farmers to generate spatial data and upload new data from future variety trials, commonly referred to as core blocks.

Images submitted by LSU AgCenter associate professor and precision agriculture state specialist Luciano Shiratsuchi

NDVI images show that plant vigor measurements and other characteristics, such as plant height, can be acquired by unmanned aerial vehicles, or drones, and other active sensors available on modern farm machinery can be correlated with soybean yield. Each square width equals one planter pass with a different variety, illustrated by v1, v2, v3. Some soil spatial patterns indicate low yield, NDVI is low and the plant is smaller, illustrating that the data is coherent and that some varieties perform better than others in different zones within the field. While data is still under fine analysis, including the collection of more soil data, the research project promotes the use of farm equipment currently owned by growers but not being utilized effectively.

This story is featured in the Louisiana Soybean and Grain Research and Promotion Board 2020 Report.