Botryosphaeria branch dieback (BOT) threatens macadamia production across all growing regions in Australia due to recent climatic conditions that promote the occurrence and increased aggressiveness of Botryosphaeria. Nine species within the Lasiodiplodia and Neofusicoccum genera have been described as the causal agents of BOT in Australian commercial orchards. The disease manifests through symptoms such as brown or dried foliage interspersed among healthy foliage, wedge-shaped necrotic lesions in the wood and eventual death of infected trees.
Botryosphaeria infects macadamia trees through pruning wounds using rain-splashed conidia. Current BOT management involves pruning during winter, when the risk of infection is lower. However, the optimal timing for pruning may vary across years and regions depending on climatic conditions. This study aimed to develop a model to predict airborne conidia concentrations of Botryosphaeria based on weather conditions, providing a tool to assist in determining the best timing for pruning in Australian macadamia orchards. We hypothesised that weather-based models could accurately predict airborne Botryosphaeria conidia concentrations in orchards.
Airborne conidia were collected from macadamia orchards with corresponding weather data obtained from the Bureau of Meteorology (BOM) between 2018 and 2021. The data were split into training (80%) and testing (20%) sets for model development and evaluation. The model was then validated in orchards across the three main production regions: central-eastern Queensland, southeast Queensland and northern New South Wales during the 2023 and 2024 production seasons.
A Gaussian generalised linear model (GLM) revealed that airborne Botryosphaeria conidia concentrations were significantly influenced by rainfall (R² = 77%), relative humidity (R² = 30%) and maximum temperature (R² = 20%). The inclusion of rainfall and maximum temperature in the GLM model predicted conidia concentrations with 89% accuracy. Independent evaluation on the testing data set demonstrated a 97% match between predicted and observed conidia concentrations. The model accurately reproduced the trends, amplitude and shape of Botryosphaeria conidia progression across all studied regions (R² > 95%). This predictive tool can optimise pruning timing and improve BOT management, offering a more precise approach compared to the current calendar-based approaches.