关键词: 二化螟, 越冬期, 虫源基数, 气象因子, 多元线性回归

Abstract:

The striped stem borer (Chilo suppressalis) is a major target for control in Chinese rice production. Its overwintering population directly determines the intensity of infestation in the following year. Based on field survey data and concurrent meteorological data from 2004 to 2022 for the overwintering population of C. suppressalis in Nanchang county of Jiangxi province, this study analyzed the correlation between trends in the overwintering population and pre−winter meteorological conditions. Key meteorological factors were screened to construct multiple linear regression models for both the average and maximum overwintering populations. Standardized regression coefficients (β) were used to evaluate the effects of selected meteorological factors. This study aimed to elucidate the impact of pre−winter meteorological conditions on C. suppressalis and to provide a reference for dynamic pest monitoring under changing climatic conditions. The results indicates that: (1) from 2004 to 2013, the overwintering population of C. suppressalis in Nanchang county exhibited a decreasing trend. After 2014, however, the population increased sharply, with a statistically significant rise (P<0.01). The average overwintering population exceeded 30.00×10⁴ individuals·ha⁻¹, while the maximum overwintering population remained consistently above 130.00×10⁴ individuals·ha⁻¹. Between 2015 and 2022, the occurrence level of C. suppressalis in Nanchang county remained at the highest damage level (Level 5). (2) Key meteorological factors influencing the average overwintering population were the average temperature in mid−October, the minimum temperature in mid−November, and the average relative humidity from late November to early December of the same year, with absolute correlation coefficients all exceeding 0.40 (P<0.05). Similarly, key factors influencing the maximum overwintering population were the minimum temperature in mid−October, the minimum temperature in mid−November, and the sunshine duration in November of the same year, with absolute correlation coefficients all exceeding 0.40 (P<0.05). (3) Multiple linear regression models for the average and maximum overwintering populations were established based on these key factors. Both models fit the training dataset well (average: R²=0.55, P<0.05; maximum: R²=0.56, P<0.01), but their generalization capability was unstable. Analysis using standardized regression coefficients (β) revealed that the most significant pre−winter factor affecting the average overwintering population was the average temperature in mid−October (β=−0.62), showing a significant negative correlation. The main pre−winter factors affecting the maximum overwintering population were the minimum temperature in mid−October (β=−0.40) and the minimum temperature in mid−November (β=0.39), with the former showing a negative correlation and the latter a positive correlation. This study indicates that warming in mid−October before winter suppresses the overwintering base population of C. suppressalis, while higher temperatures in November favor population expansion. This study clarifies the stage−specific and comprehensive impact of pre−winter meteorological conditions on the overwintering population of C. suppressalis and highlights the critical role of pre−winter temperatures in controlling the pest source. It provides a reference for developing dynamic monitoring and control strategies for pests under changing climatic conditions. 

Key words: Chilo suppressalis, Overwintering period, Initial insect population, Meteorological factors, Multiple linear regression