The economic performances of different trial designs in on-farm precision experimentation: a Monte Carlo evaluation

On-farm precision experimentation (OFPE) has expanded rapidly over the past few years. While the importance of OFPE trial design efficiency has been recognized, existing studies have primarily used statistical measures of that efficiency to compare designs. The current article is motivated by the surety that farmers are more interested in economic results than statistical results; Monte Carlo simulations of corn-to-nitrogen (N) response OFPEs were used to compare economic performances of 13 different types of OFPE trial design. Each design type’s economic efficiency was measured by the monetary profits resulting from applying the site-specific economically optimal N rates estimated from the data generated by the design type. Results showed that trial design affects the economic performance of OFPE. Overall, the best design was the Latin square design with a special pattern of limited N rate “jump,” which had the highest average profit and lowest profit variation in almost all simulation scenarios. A particular type of patterned strip design also performed well, generating average profits only slightly lower than those from the best design. In contrast, designs with gradual trial rate changes over space were less profitable in most situations and should be avoided. Results were similar under various scenarios of nitrogen-to-corn price ratios, yield response estimation models, and field sizes used in the simulations. It was also found that the designs’ economic performances were roughly correlated with the spatial property measures of trial designs in existing literature, though much remains unexplained.