Advanced Dendrolimus punctatus forecasting through Grey-Markov integration

Dendrolimus punctatus Walker (Lepidoptera: Lasiocampidae) is a major defoliator that causes significant ecological and economic damage to forests in China. Accurate forecasting of D. punctatus outbreaks is essential for effective pest management. This study develops hybrid models that integrate BP neural networks or stepwise regression with Grey and Markov chain models to improve the prediction accuracy of the peak period of D. punctatus for the first generation of eggs. Model performance was evaluated using the Mean Relative Error (MRE) and the Mean Square Error (MSE). The results indicate that the hybrid models, particularly the BP-Markov model, outperformed the single models except for the Regression-Markov model. The BP-Markov model yielded the lowest MRE (0.2422%) and MSE (0.0197), indicating high predictive accuracy. This study demonstrates that integrating different mathematical models can improve forecasting precision, suggesting potential utility for supporting proactive forest pest management. The proposed model can provide timely early warnings of potential D. punctatus outbreaks, thereby helping forest managers optimize monitoring schedules, determine appropriate intervention timing, and allocate pest-control resources more efficiently.