Sampling techniques for IPM decision-making of the silverleaf whitefly (Bemisia argentifolii Bellows & Perring) on net-house tomato (Lycopersicon esculentum Mill.)

• Main Authors: Po-Yao Wang, 王柏堯
• Other Authors: Chiou-Nan Chen
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/72142923444289934204

碩士 === 國立臺灣大學 === 昆蟲學研究所 === 92 === I conducted studies to develop sampling plans for the decision-making of control action of silverleaf whitefly (Bemisia argentifolii Bellows & Perring ) on tomato (Lycopersicon esculentum Mill.) cultivated under net-house in the Asia Pacific Farm, Guansi Township, Hsinchu County. Using sequential sampling plans can help agriculturist to judge the damage of target pest and to decide whether control is needed. Before developing sequential sampling plans, I need to examine the distribution model of target pest in the grove. Random leaf samplings were taken weekly for four crop seasons, namely, from April to June, 2002, October, 2002 to January, 2003, March to June, 2003,and November, 2003 to February, 2004. In terms of vertical distribution, adults and nymphs were abundant on leaves at the middle and the bottom stratum of tomato (80%) than on leaves at the top stratum. Using X2-test to examine the fitness of the negative binomial distribution to the frequency data collected from the field, I concluded that negative binomial distribution fitted well to most of data sets. Thirty-three out of 47 data sets of adults and 26 out of 47 data sets of nymphs were in compliance with negative binomial distribution. Because both adults and nymphs infested tomato, data of adults and nymphs were pooled together to develop the sequential sampling plan. Aggregation patterns measured by Taylor’s Power Law revealed that adults were aggregative (b = 1.288) as well as nymphs (b = 1.481). The same conclusion was reached by using Iwao’s mean crowding-mean regression as the parameters of adults (?= 1.130, β = 1.223) and of nymphs(?= 7.457, β = 1.303) indicated. Using the parameters from Taylor’s Power Law (a, b) and from Iwao’s mean crowding-mean regression(? β) and set the precision level (D) at 0.1 and 0.2), optimal sample size was estimated for specified population density. Using Common K (Kc) derived from the pooled data sets fitting the negative binomial distribution and assume that class limits m1 = 32 (no./per leaf ) , m2 = 64 (no./per leaf) and Type I error?= 0.1, Type II errorβ = 0.1. I developed the sequential sampling plan as USV = 44.423 n + 91.4 and LSV = 44.423n - 91.4. Using Iwao’s mean crowding-mean regression for the pooled data and assumed the class limits m0 = 64 (no./per leaf), I developed the sequential sampling plan as USV： ,and LSV： (q：number of leaves examined). Based upon the ratio of infested leaves, a binomial sequential sampling plan was developed such that LSV： , USV： ( n：number of leaves examined and a leaf with ≧ 1 silverleaf whitefly body is defined as the infested leaf). However, the sequential sampling plans as developed above help judge the damage level of target pest remain further evaluation in practice.