AUTHOR=Sharma Pankaj , Rai Pramod Kumar , Meena Prabhu Dayal , Sharma Hariom Kumar , Singh Vijay Veer , Sanyal Shravani , Prasad Niranjan , Kachchwaha Jitendra , Gupta Navin Chandra , Sharma Anubhuti , Bharadwaj Nitish Rattan 

TITLE=Predictive modelling and epidemiological forecasting of sclerotinia rot in Brassica juncea under climatic variability in Indian conditions

JOURNAL=Frontiers in Plant Science

VOLUME=Volume 16 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/plant-science/articles/10.3389/fpls.2025.1650230

DOI=10.3389/fpls.2025.1650230

ISSN=1664-462X

ABSTRACT=Sclerotinia rot (SR), caused by Sclerotinia sclerotiorum, poses a significant threat to Indian mustard (Brassica juncea L.), cultivated across major oilseed-growing regions in India. A long-term field study was conducted from 2009–2010 to 2021–2022 to investigate the role of key agrometeorological parameters on influencing SR incidence under three sowing windows, namely, 8 October (early), 29 October (timely), and 19 November (late sown). Weekly meteorological variables, including maximum and minimum temperature (°C), relative humidity (RH) (%) during morning (07:20 h) and afternoon (14:20 h), rainfall (mm), wind speed (km/h), evaporation (mm), and bright sunshine hours (BSSH), were collected and used to develop regression-based weather indices and random forest models to develop robust predictive models for effective forecasting. Results revealed that the 29 October sowing window was consistently associated with the highest predicted SR risk (up to 39.4%), when maximum temperature hovered at approximately 18–20 °C, RH exceeded 94% in the morning, and BSSH fell below 3.8 hours. A strong negative correlation (R2 = 0.86) was observed between BSSH and SR incidence, particularly in the 29 October sowing window. Petal infestation studies confirmed early colonization pressure, with percent petal infection peaking at 20.7% during the second week of January area under the petal progress curve (AUPPC), which provides condensed weekly petal infestation trajectories into a single measure of inoculum pressure and depicts the highest epidemic pressure in the mid-sowing window. Disease forecasting models incorporating weighted weather indices demonstrated high predictive accuracy with R2 values of 0.75, 0.76, and 0.78 for early, timely, and late sowing dates, respectively, when validated with 2022–2023 observations. Future predictions using the random forest model (2025–2030) indicated that the 29 October sowing remains the most vulnerable, while the 19 November sowing consistently exhibited lower disease risk due to less favorable microclimatic conditions to support apothecial formation and ascospore release. The study emphasizes that sowing time, in conjunction with real-time meteorological variables, significantly governs the epidemic potential of SR. The predictive models developed herein offer a reliable decision support system for major mustard growing states of the country, enabling proactive disease forecasting and sustainable crop protection strategies.