Weather based prediction model for beet armyworm (Spodoptera exigua Hubner) in chickpea

Aim: The study aimed to develop and validate weather based prediction model for beet armyworm (Spodoptera exigua) population in chickpea through adult catches in pheromone traps. Methodology: The data on adult trap catches of S. exigua were recorded daily and weekly means were computed. Log transformed trap catches data were used for correlation with weather parameters of current week, 1-lag, 2-lag and 3-lag weeks. Thereafter, multiple-linear regression analysis was done and a model was developed. The prediction model of S. exigua was validated with the appropriate statistical tools. Results: Peak incidence of S. exigua was recorded during 45th standard meteorological week (SMW) with 15.6 moths per trap per week. Amongst current, 1-lag, 2-lag and 3-lag week weather parameters, the male moth population had significant positive correlation with maximum temperature (Tmax) and minimum temperature (Tmin), and negative correlation with morning relative humidity (RH1) of 2-lag week. The sunshine hours/day (SSH) of current week had a significant negative association with S. exigua male moth catches, while the soil temperature (ST) of 2-lag week had highest positive correlation with trap catches. Regression equation was computed by regressing male moth catches of S. exigua against weather data of weeks with highest correlation coefficient. Interpretation: Often, pest-weather models are developed based on current week weather factors. However, it has been witnessed in this study that weather of preceding weeks (up to 3-lag) may also influence the pest population, and thus it needs to be considered for proper understanding of pest dynamics