Skip to main content
APS Plant Health 2025 Main banner
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.

POSTERS: Analytical and Theoretical Plant Pathology I

P-005 - A Causal Mediation Approach for Evaluating Planting Date Effects on Soybean Yield Considering Disease Pressure

Monday, August 4, 2025
3:45 PM - 4:30 PM HST
Location: Exhibit Hall 3

    Presenting Author(s)

  • SS

    Santosh Sanjel

    Postdoctoral Scholar
    Pennsylvania State University
    University Park, Pennsylvania, United States

    • Author(s)

  • DS

    Denis Shan

    Kansas State University
    Manhattan, Kansas, United States

  • SC

    Shawn P. Conley

    University of Wisconsin-Madision
    Madison, Wisconsin, United States

  • SM

    Spyridon P. Mourtzinis

    University of Wisconsin-Madision
    Madison, Wisconsin, United States

  • PE

    Paul Esker

    Penn State
    UNIVERSITY PARK, Pennsylvania, United States

Abstract Text: Meeting the optimal planting date is critical for maximizing soybean yield. Current optimal planting date recommendations for a region are based on local field trials and models. However, isolating the effect of planting date on yield is challenging due to the confounding effects of localized management practices in traditional field trials. Causal inference methodology applied to farmers’ field survey data offers a complementary approach to estimating planting date effects. We applied causal mediation analysis, which permits the decomposition of the total effect of planting date into direct (independent of disease) and indirect (modulated by disease) effects. We analyzed survey data from approximately 5,000 soybean fields (2014 to 2016) across 11 North-Central states. Using Directed Acyclic Graphs (DAGs), we identified appropriate adjustment sets and applied multiple causal estimators to quantify planting date effects on yield. Because actual disease data were unavailable, we simulated soilborne disease severity values (1–100%) based on agronomic conditions and literature on temperature and soil moisture thresholds and used the simulated data to model high and low mediation effects. Our results illustrate that when the indirect effect via disease is substantial, disease management becomes critical to maximizing yield benefits derived from optimal planting dates. This study demonstrates the feasibility of integrating causal mediation analysis into agronomic research, providing a novel approach to incorporating plant disease into decision-making frameworks.