Premium
Binomial sampling plan for tomato russet mite ( Aculopslycopersici (Tryon) (Acari: Eriophyidae) in protected tomato crops
Author(s) -
Moerkens Rob,
Vanlommel Wendy,
Reybroeck Eva,
Wittemans Lieve,
De Clercq Patrick,
Van Leeuwen Thomas,
De Vis Raf
Publication year - 2018
Publication title -
journal of applied entomology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.795
H-Index - 60
eISSN - 1439-0418
pISSN - 0931-2048
DOI - 10.1111/jen.12529
Subject(s) - eriophyidae , statistics , sampling (signal processing) , resampling , pest analysis , sample (material) , mite , binomial (polynomial) , mathematics , sample size determination , range (aeronautics) , biology , integrated pest management , toxicology , horticulture , agronomy , ecology , computer science , physics , thermodynamics , materials science , filter (signal processing) , composite material , computer vision
The tomato russet mite ( TRM ), Aculops lycopersici , is a worldwide pest of cultivated tomatoes. Currently, no effective biological control agents are available on the market. Therefore, chemical spray applications are required. Fast and reliable detection, monitoring and evaluation of interventions are a challenge, slowing down the development of an appropriate integrated pest management ( IPM ) strategy. This study describes a binomial sampling plan with the aim to reduce the efforts and costs for an accurate monitoring of A. lycopersici . Sampling was performed by taking pictures of the upper leaf surface with a smartphone through an attached magnification lens. A binomial sampling plan was developed based on the linear relationship between ln(mean TRM densities) and ln(−ln(1‐ P T ), where P T is the proportion of samples with more than T (tally threshold) mites. The minimum precision threshold of 0.30 was determined for the different models. A resampling for validation of sample plans ( RVSP ) programme with a fixed sample number was used for validation of the model on an independent data set. The binomial sampling plans were validated at tally thresholds of T = 9 and T = 15 with fixed sample sizes of 15, 20, 25 and 30. Precision levels were satisfying within a range of P T ‐values from 0.29 to 0.97 for T = 15 at a fixed sample size of 20. This range was much smaller for T = 9, where the P T ‐values range between 0.40 and 0.92 at the same sample size. A binomial sampling model with T = 9 with a fixed sample size of 15, which has the lowest time investment, is feasible for glasshouse tomato growers in practice. However, for the development of pest management programmes, a more intensive and more accurate binomial sampling plan with T = 15 and a sample size of minimum 20 is suggested.