Rotten Hazelnuts Prediction via Simulation Modeling-A Case Study on the Turkish Hazelnut Sector

Area of Science:

• Agricultural Science
• Plant Pathology
• Computational Modeling

Background:

• Hazelnut quality defects, including off-flavors like the rotten defect, are influenced by environmental conditions and fungal pathogens, primarily Diaporthe species.
• Accurate forecasting of the rotten defect is crucial for both consumers to identify high-quality sources and for farmers to anticipate and manage pathogen risks.

Purpose of the Study:

• To develop and apply a novel rotten defect forecasting model for hazelnuts in Turkey's main producing regions.
• To simulate the epidemiological cycle of Diaporthe spp. and modulate plant susceptibility based on phenology.

Main Methods:

• A process-based simulation model was created to reproduce the Diaporthe spp. epidemiological cycle.
• Model calibration utilized weekly phenological observations and post-harvest rotten incidence data (2016-2019) from 22 orchards.
• Sensitivity analysis under varying weather conditions identified key parameters for model calibration.

Main Results:

• The rotten defect simulation model achieved a mean absolute error below 1.8% for rotten incidence in calibration and validation datasets.
• Model validation across 321 locations showed variability in correlation (R² = 0.4 and 0.21) due to unconsidered environmental and agronomic factors.
• The model successfully differentiated rotten incidence across municipalities and reproduced interannual variability, highlighting the impact of precipitation and plant susceptibility.

Conclusions:

• The developed model accurately predicts hazelnut rotten defect incidence, demonstrating a strong dependence on precipitation patterns and plant susceptibility.
• Further improvements will incorporate additional environmental and agronomic factors to enhance predictive accuracy.
• The model shows potential for operational, in-season forecasting after application in other hazelnut-producing regions.