关键词: 大豆食心虫, 气候因子, 复合指数, 风险区划

Abstract:

Provincial−level climate risk zoning indicators for the soybean pod borer, which can provide technical support for disaster prevention and mitigation of agricultural pests and diseases, promote the improvement of quality and efficiency in agricultural production, and the high−quality development of meteorological services. This study utilized data on the rate of damage caused by Leguminivora glycinivorella from 25 meteorological stations in Heilongjiang province from 1980 to 2021. Characteristic climatic factors were screened by employing methods such as correlation analysis, path analysis and collinearity analysis. A comprehensive climate risk index for Leguminivora glycinivorella was defined and constructed. The BP artificial neural network method was used to analyze the relationship between the rate of damage and the characteristic climatic factors. Based on the K–means clustering method and typical disaster years, the critical values and grades of comprehensivc climate risk were determined, and the climate risk zoning indicators and models for the occurrence of Leguminivora glycinivorella in Heilongjiang province were established. After verifying the accuracy, climate risk zoning of Leguminivora glycinivorella was conducted for the periods of 1961–1990 and 1991–2020. The results indicated that the five significant characteristic climatic factors that affected the damage rate of Leguminivora glycinivorella in Heilongjiang province were: average temperature of the last ten days of September last year, the monthly average of the daily average temperatures in December of last year, the average of the daily average temperatures in the first ten days of April, the average of the daily average temperatures for May and June and the average daily relative humidity in August. The critical thresholds of the comprehensive climate risk index for the low–risk, medium–risk, high–risk and extremely high–risk zones of Leguminivora glycinivorella in Heilongjiang province were [0, 0.47), [0.47, 0.58), [0.58, 0.68), and [0.68, 1.00], respectively. The rates of damage were [0, 3.5%), [3.5%, 7.0%), [7%, 10%), and [10%, 43%], respectively. The verification results of 40 reserved samples showed that the proportion of samples with fully consistent risk grades calculated by the indicators and the actual risk grades was 67.5% and the proportion of samples with a difference of one grade was 27.5%. This demonstrates that the indicators had indicative significance. Both high–risk regions in 1961–1990 and 1991–2020 were concentrated in the Songnen and Sanjiang plains. The medium and low–risk areas were mainly distributed in the Greater and Lesser Xing'an ranges and the semi–mountainous area of Mudanjiang. Compared with the period of 1961–1990, the period of 1991–2020 saw the emergence of extremely high–risk areas, with the range of high–risk areas was significantly larger than that in 1961–1990, while the range of medium and low–risk areas were decreased significantly.

Key words: Leguminivora glycinivorella, Climatic factors, Compound index, Risk zonation