A Method for Ascertaining and Controlling Representation Bias in Field Trials for Airborne Plant Pathogens

Area of Science:

• Agricultural Science
• Plant Pathology
• Computational Biology

Background:

• Field trials are crucial for evaluating treatment efficacy in agriculture.
• Discrepancies between experimental and real-world treatment effects are termed 'representation bias.'
• Inter-plot interference, especially from wind-dispersed pathogens, can cause significant representation bias in plant disease field trials.

Purpose of the Study:

• To quantify representation bias in plant disease field trials.
• To explore the relationship between field design parameters and representation bias.
• To emphasize the impact of plot dimensions, spacing, and treatment allocation on experimental accuracy.

Main Methods:

• Development of a computer simulation model for plant disease dispersal.
• Simulation of various experimental scenarios to assess representation bias.
• Analysis of the influence of plot design elements on bias magnitude.

Main Results:

• The computer simulation successfully quantified representation bias under different conditions.
• Key field design parameters, including plot dimensions, spacing, and treatment allocation, were identified as significant factors influencing representation bias.
• The study demonstrated the potential for substantial bias in trials involving wind-dispersed pathogens.

Conclusions:

• Representation bias can significantly impact the reliability of field trial results for plant disease treatments.
• Optimizing field trial design, particularly plot configuration and treatment allocation, is essential to minimize representation bias.
• The developed simulation provides a valuable tool for understanding and mitigating bias in agricultural field studies.