Weather-based coffee leaf rust apparent infection rate modeling.

Area of Science:

• Plant Pathology
• Agricultural Meteorology
• Crop Science

Background:

• Brazil is a leading coffee producer, with significant cultivation of Coffea arabica and Coffea canephora.
• Coffee leaf rust (CLR) significantly reduces crop yield by damaging leaf area.
• CLR incidence is strongly influenced by meteorological conditions, necessitating predictive models.

Purpose of the Study:

• To identify the optimal linearization model for estimating CLR apparent infection rates.
• To establish correlations between CLR infection rates and key weather variables.
• To develop and evaluate a weather-based CLR disease warning system.

Main Methods:

• Analysis of 88 site-seasons of CLR epidemic data.
• Assessment of various linearization models (linear, monomolecular, logistic, Gompertz, exponential) for apparent infection rate determination.
• Correlation analysis between infection rates and weather variables, followed by multiple linear regression modeling.

Main Results:

• The Gompertz growth model demonstrated the best fit for CLR progress curves.
• Minimum temperature and relative humidity were identified as the most significant weather correlates of CLR infection rate.
• A predictive model for high-yield conditions with narrow row spacing achieved a low false negative rate of 9.4%.

Conclusions:

• A robust weather-based system for forecasting coffee leaf rust has been developed.
• The model effectively utilizes minimum temperature and relative humidity for predicting CLR infection.
• This system offers a valuable tool for proactive disease management in coffee cultivation, particularly under specific high-yield conditions.