Weather-based prediction module: A promising tool for Spodoptera litura anticipation in groundnut

Spodoptera litura has been causing considerable damage to groundnut and other crops across India. Anticipation of pest incidence using a prediction module under changing climate is need of the hour. The Spodoptera litura male moth caught by sex pheromone trap from 1991-20 (30 years) during kharif season is considered for the study. The 30 years of data clearly indicated that moths
trapped were concentrated in 34 th and 35 th SMW, which falls during the second fortnight of August and the first week of September. There was steep decline in the number of moths trapped from 2015 onwards indicating a consistent decrease in Spodoptera incidence during kharif at Dharwad despite of normal rainfall and other weather factors. The multiple regression model fitted had multicollinearity of independent variables. Stepwise regression, Morning RH was found to be the major deciding factor of Spodoptera population. Finally, simple linear regression model was fitted and the predicted trap catch data was validated with actuals for four consecutive kharif seasons starting from 2017. However, the predicted was found to be far higher than actual number of moths trapped during respective years. The presentanalysis clearly indicated need for a better understanding the influence
of biotic and abiotic factors on Spodoptera under field conditions.