昆虫学报 ›› 2022, Vol. 65 ›› Issue (10): 1389-1400.doi: 10.16380/j.kcxb.2022.10.015

• 综 述 • 上一篇    下一篇

自我重复取样技术与多项式定理在年龄-龄期两性生命表研究中的应用

建宇1, 陈燕婷1, 傅建炜2, 史梦竹2, 齐心3,4, 尤民生3,*   

  1. (1. 福建省农业科学院植物保护研究所,福建省作物有害生物监测与治理重点实验室, 福州 350013; 2. 福建省农业科学院农业质量标准与检测技术研究所, 福州 350003; 3. 福建农林大学应用生态研究所, 闽台作物有害生物生态防控国家重点实验室, 福州 350002; 4. 山东农业大学植物保护学院, 山东泰安 271018)
  • 出版日期:2022-10-20 发布日期:2022-11-27

Application of the bootstrap technique and the multinomial theorem in the research of age-stage, two-sex life table

LI Jian-Yu1, CHEN Yan-Ting1, FU Jian-Wei2, SHI Meng-Zhu2, CHI Hsin3,4, YOU Min-Sheng3,*    

  1. (1. Fujian Key Laboratory for Monitoring and Integrated Management of Crop Pests, Fujian Engineering Research Center for Green Pest Management, Institute of Plant Protection, Fujian Academy of Agricultural Sciences, Fuzhou 350013, China; 2. Fujian Key Laboratory of Agroproducts Quality and Safety, Institute of Quality Standards and Testing Technology for Agro-Products, Fujian Academy of Agricultural Sciences, Fuzhou 350003, China; 3. State Key Laboratory of Ecological Pest Control for Fujian and Taiwan Crops, Fujian Agriculture and Forestry University, Fuzhou 350002, China; 4. College of Plant Protection, Shandong Agriculture University, Tan′an, Shandong 271018, China)
  • Online:2022-10-20 Published:2022-11-27

摘要: 年龄-龄期两性生命表(age-stage, two-sex life table)简称两性生命表,是种群生态学研究与害虫治理中常用的重要理论与分析工具。根据两性生命表理论而设计的方便用户的软件TWOSEX-MSChart近年来被越来越多国内外学者用于昆虫种群研究的数据分析。两性生命表软件的分析功能是由许多的统计技术与计算机模拟方法作为数据分析的支撑,其中自我重复取样(bootstrap)是其重要技术之一。本文详述了bootstrap技术的基本原理、方法、优缺点及其在两性生命表分析中的应用,并介绍了其理论基础多项式定理(multinomial theorem)在生命表研究中的应用。与常用统计方法相比,bootstrap不需要数据分布假设就可以对数据总体的分布特性进行统计和推断。在两性生命表分析中,bootstrap不仅可以估算种群参数或一般统计值的方差和标准误,同时利用paired bootstrap test还可以比较不同处理间的差异,准确显示种群的变异性。利用相同的自我重复取样样本(same bootstrap samples)可以正确计算昆虫的孵化率与不同繁殖型对种群参数的贡献,并可连接天敌的生命表与捕食率分析,以正确分析昆虫种群的繁殖力或天敌种群的捕食潜能等。此外,本文介绍了bootstrap数学理论基础的多项式定理,其在两性生命表研究中的应用进一步证明了bootstrap重复取样技术能得到较稳定和可靠的总体参数估计,并分析了生命表研究中考虑无效bootstrap样本的重要性。近年来,关于两性生命表及bootstrap的应用相关研究较多,但是关于bootstrap技术的原理及方法很少报道。本文将有助于从事昆虫学和生态学的研究人员理解bootstrap技术和多项式定理的基本理论与原理及其在两性生命表分析中的应用,以更好地将其运用于相关科研工作中。

关键词: 生命表, 年龄, 龄期, 性别, 两性生命表, 自我重复取样, 多项式定理, 种群预测

Abstract: The age-stage, twosex life table,called two-sex life-table for short, is an important theory and an analytical tool that are commonly used in population ecology and pest management. The user-friendly TWOSEX-MSChart program, which had been designed based on the twosex life table theory to help researchers for data analysis in insect population studies, has been more and more widely used by increasing numbers of scientists around the world. There are many statistical techniques and computer simulations embedded in the TWOSEX program, and the bootstrap technique is one of the major procedures included in the program. In this article, we describe the principles, methods, advantages/disadvantages, and the application of the bootstrap technique in the twosex life table analysis, as well as the application of the multinomial theorem in life table research. Compared with the general statistics, the bootstrap technique can be used to estimate and infer the distribution characteristics of data without the assumption of data distribution. In the twosex life table analysis, the bootstrap technique can not only be used to estimate the population parameters or the variances and standard errors of general statistics, but also to be used to assess the differences between treatments by paired bootstrap test to accurately show the population variability. The same bootstrap samples can be used to calculate the hatching rate and the contribution of different reproductive forms to population parameters, and to link the life table and predation rate analysis of natural enemies for an accurate analysis of the reproduction and predation potential of natural enemies. In addition, we also introduce the multinomial theorem, i.e., the mathematical basis of the bootstrap technique. The application of the multinomial theorem demonstrates that stable and reliable estimates can be obtained by using the bootstrap technique. We also elaborate the necessity of considering the ineffective bootstrap samples in the life table research. In recent years, although the twosex life table and the bootstrap technique had been widely adopted in research, few reports discussed the principles and methodology involved. This article will help interested researchers in entomology and ecology understand the basic theories and principles of the bootstrap technique and the multinomial theorem, and their application in twosex life table analysis, so as to better apply them in the related scientific research projects.

Key words: Life table, age, developmental stage, sex, twosex life table, bootstrap, multinomial theorem, population prediction