Development and validation of binomial sampling plans for estimating leafmine density of Liriomyza trifolii (Diptera : Agromyzidae) in greenhouse tomatoes

Abstract

The dynamics of leafmines developed by Liriomyza trifolii (Burgess) larvae in greenhouse tomatoes were studied in Korea during 2003-2004 in order to construct and validate binomial sampling plans. An empirical P-T-m model, expressed as: ln(m)= alpha+beta ln[-ln(1-P-T)], was used to relate between the proportion of infested leaves (P-T) and mean densities (m) at tally thresholds (T, the minimum number of leafmines present before a leaf is considered infested) of 1, 2, 3, 4 and 5 mines per leaf. To ensure the consistent selection of sample units, a sequence of reference pictures, with various levels of leafmine formation, was measured in advance by image analysis. The mines < 0.4 cm(2) in area were excluded from counting in the greenhouses. There were no significant relationships between the total numbers of leafmines and individual mine areas. The binomial sampling plans were validated using resampling simulations with seven independent data sets. In an estimation of the density, the sampling precision (SE/mean) was found to increase with higher Ts; however, there were negligible improvements in the precision with T >= 3 mines per leaf. Using T=3, over a wide range of mine densities, as few as 30 samples were necessary to achieve a precision of 0.30. In comparing binomial models with T=3 and 5, using seven independent data, the model with T=3 was a robust and relatively unbiased predictor of the mean density, whereas the T=5 model was generally biased towards over-prediction of the mean density. The binomial sampling plans presented here should permit rapid estimation of the mine density and enhance development for a damage assessment program in greenhouse tomatoes.