Evaluating influence of weather on Ascochyta blight severity in chickpea using correlation network, regression and principal component analysis

Aim: Weather parameters play a critical role in the survival and propagation of the causal pathogen Ascochyta rabiei. In the absence of vertical host resistance, the present study was designed to explore the weather-disease interrelationship for effective disease management. Methodology: Field experiments were conducted with three chickpea cultivars and three sowing times for two years. The relationship between the percent severity index (PSI) at weekly interval and observed weather data was explored using multivariate statistical techniques. Results: The correlation networks revealed that temperature and relative humidity parameters were negatively and positively associated with the blight PSI, respectively. Humid thermal ratio (HTR) significantly correlated with PSI among all three chickpea cultivars. The linear regression of PSI with weather variables exhibited a significant negative slope (β) with all thermal variables whereas a strong positive β was displayed with relative humidity and HTR. The Best subset multiple regression models further affirmed that PSI in three chickpea cultivars can be strongly predicted with models constituted with temperature, humidity and HTR. Principal component analysis showed that PC1 and PC2 explained 88.2% variability with weather variables affecting the PSI. The clustering of thermal and humidity condition parameters around the first two PCs explained their pivotal role in disease severity progress. Interpretation: The multivariate analysis revealed that air temperature, relative humidity and HTR had significant influence on periodic Ascochyta blight severity in chickpea crop. Present exploration will be vital in developing more targeted disease prediction and management strategies to secure chickpea production. Key words: Ascochyta blight, Best subset regression, Chickpea, Correlation network, Weather variables