From Scalar Summaries to Functional Comparisons: A Framework for Analyzing Plant Disease Progress Curves

Area of Science:

• Plant Pathology
• Epidemiology
• Quantitative Genetics

Background:

• Disease progress curves (DPCs) are crucial for assessing plant disease management and host resistance.
• Scalar summaries like AUDPC can mask important variations in epidemic timing and shape.
• A need exists for methods that capture the full trajectory of plant disease epidemics.

Purpose of the Study:

• To introduce a curve-based framework for comparing plant disease epidemics as epidemic phenotypes.
• To enable trajectory-based comparisons of disease epidemics beyond traditional scalar summaries.
• To provide a tool for enhanced host resistance phenotyping and comparative epidemiology.

Main Methods:

• Utilized a hierarchical generalized additive model (HGAM) to estimate environment-adjusted mean epidemic curves.
• Quantified hybrid similarity using a functional distance metric over the epidemic time domain.
• Employed hierarchical clustering to identify epidemic phenotypes based on curve shape differences.

Main Results:

• The curve-based framework identified distinct epidemic phenotypes in southern corn leaf blight that AUDPC missed.
• Comparisons revealed differences in epidemic trajectory shapes not apparent from overall disease levels.
• Breeder-defined resistance classes showed systematic differences in epidemic trajectory shapes across environments.

Conclusions:

• The proposed framework offers a complementary approach to scalar summaries for analyzing DPCs.
• Treating DPCs as epidemic phenotypes allows for nuanced comparisons of disease trajectories.
• This method enhances host resistance phenotyping and comparative epidemiology by capturing temporal dynamics.