Unraveling Pest-Weather Dynamics: A Stepwise Regression Approach on Major Pests in Rice
Annepu Jhansi *
Department of Statistics & Computer Applications, Agricultural College, Bapatla, Acharya N.G. Ranga Agricultural University, India.
B.V.R.Ch. Ravi Kumar
Department of Statistics & Computer Applications, Agricultural College, Bapatla, Acharya N.G. Ranga Agricultural University, India.
Shaik Shameem
Department of Statistics & Computer Applications, Agricultural College, Bapatla, Acharya N.G. Ranga Agricultural University, India.
P. Lavanya Kumari
Department of Statistics & Computer Applications, Sri Mekapati Gowtham Reddy Agricultural College, Udayagiri, India.
*Author to whom correspondence should be addressed.
Abstract
Rice (Oryza sativa L.) is a major staple crop of the Godavari delta region of Andhra Pradesh, where its productivity is severely constrained by key insect pests, namely Yellow Stem Borer (YSB), Brown Planthopper (BPH), Gall Midge, and Leaf Folder (LF). Understanding long-term pest–weather relationships is essential for improving pest forecasting and developing climate-responsive management strategies. This study analyzed long-term weekly pest and weather data spanning over two decades (2002–2023 for kharif and 2003–2023 for Rabi) recorded at the Regional Agricultural Research Station, Maruteru, using stepwise regression analysis to identify critical climatic drivers influencing pest dynamics.
The results revealed strong season-specific pest responses to weather variables. During kharif, YSB incidence was primarily driven by minimum temperature (TMIN), evening relative humidity (RHE) and sunshine hours (SSH), while during Rabi, TMIN and RHE were dominant. Gall Midge outbreaks were mainly associated with SSH in kharif and TMIN and morning relative humidity (RHM) in Rabi. Leaf Folder infestation was significantly influenced by RHE and maximum temperature (TMAX) in kharif, whereas RHE, RHM and TMIN governed its occurrence in Rabi. BPH population during kharif responded to RHM, SSH and TMAX, while during Rabi, RHE emerged as the principal driver. The peak pest pressure was consistently observed during the kharif season, indicating it as the major risk period for yield loss in the region.
Although the developed models exhibited relatively low R² values, reflecting non-linear behaviour and high ecological heterogeneity in pest populations (a major limitation of the study), they provide critical insights into the climatic triggers of pest outbreaks. Based on these findings, the study recommends weather-based pest surveillance, timely advisories, and the adoption of climate-responsive integrated pest management (IPM) practices, particularly during the vulnerable kharif season, to effectively mitigate pest damage and enhance rice productivity in coastal Andhra Pradesh.
Keywords: Rice, pest forecasting, weather parameters, multicollinearity and stepwise regression