基于PERT与仿真技术的橘小实蝇传入定量风险评估  

Quantitative risk assessment of oriental fruit fly based on PERT and Monte Carlo simulation

作　　者：李白尼[1] 侯柏华[2] 郑大睿[1] 马骏[3] 张润杰[1] 

机构地区：[1]中山大学有害生物控制与资源利用国家重点实验室,昆虫学研究所,广州510275 [2]广东省昆虫研究所,广州510260 [3]广东出入境检验检疫局检验检疫技术中心,广州510623

出　　处：《植物保护》2008年第5期32-39,共8页Plant Protection

基　　金：国家“973”计划项目(2002CB111405);国家科技支撑计划项目(2006BAD08A13);国家自然科学基金(30471162,30671394)

摘　　要：检疫性害虫传入风险定量评估是外来入侵有害生物控制的重要组成部分。本文应用计划评审法(PERT)和Monte Carlo仿真技术分析了2007年春夏两季(5～8月)入穗外地水果(香蕉、芒果、番石榴)经过采摘、运输、销售、丢弃等环节后,传入广州地区的橘小实蝇的数量及其风险概率。结果表明,橘小实蝇累计传入数量为590～598头,能够形成295～372对潜在交配成虫,经典PERT模型和Monte Carlo仿真的计算结果很接近,传入风险确实存在,为100%。同时灵敏度分析显示,遭受橘小实蝇危害的水果比例、1kg受害水果所含橘小实蝇幼虫的数量、有限时间内3龄幼虫的逃逸比率是影响害虫传入数量的3个最主要的因素。Quantitative risk assessment of quarantine pests is an important part of integrated pest management （IPM）. In this paper, both PERT and Monte Carlo simulation were used to assess the risk of Bactrocera dorsalis （Hendel）, which entered Guangzhou from May to August in 2007 through the imported fruits （bananas, mangoes and guava） after picking, transporting, marketing, and discarding. The outcome revealed that the total number of B. dorsalis invading to Guangzhou was about 590, which could form about 295-372 mating pairs, and there is little difference between the results calculated by Classic PERT and Monte Carlo simulation. Therefore, the importing risk really exists. The results of sensitivity analysis showed that the infestation rate of the fruit goods, the number of larvae in fruits per kilogram, and the rate of B. dorsalis escaping from the hosts in finite time are three most important factors affecting the number of pests entering Guangzhou city through the transportation of fruit goods.

关 键 词：PERT MONTE Carlo仿真 橘小实蝇 风险分析 

分 类 号：S436.669[农业科学—农业昆虫与害虫防治]