How many kilograms of grain per sample unit is big enough? Part I - Comparison of insect detection and density estimation between manual probe sampling and Insector^® system

Sampling and trapping are important methods to detect insect infestation and estimate the abundance of species (species richness) within stored grain. Sampling and trapping were conducted inside a hopper-bottom bin holding 50-t wheat with 10.5% moisture content at 25±5°C. Adult rusty grain beetles, Cryptolestes ferrugineus (Stephens), were introduced into the bin to create overall insect densities of 0.1, 1.0, or 5.0adults per kilogram of wheat (A/kg). Twenty Insector® traps (OPI Systems Inc., Calgary, AB, Canada) were installed at four layers within the grain mass. Every week, 0.5- and 15-kg grain samples were manually extracted near five Insector® locations within the top layer of the grain mass. The success of insect detection, accuracy of insect density estimation, and characterized insect distribution pattern were evaluated using data collected by the Insector® system and the manual samples. It was found the Insector® system could detect adults within 6h after adults were introduced. The 0.5-kg samples could not detect the existence of insects or estimate insect densities when insect density was 0.1A/kg. Depending on the insect density, up to ten 15-kg samples (sampled at the top layer) were required to detect the existence of adults. The estimated insect densities by the Insector® system were closer to the introduced insect densities than the manual sampling methods when the introduced insect densities were 0.1 or 1.0A/kg. There was no significant difference in estimated insect densities between the Insector® system and the 15-kg sampling method when introduced insect density was 0.1 or 1.0A/kg. The characterized adult distribution pattern was the same using either the data collected by the Insector® system or 15-kg samples, while 0.5-kg samples characterized different insect distribution patterns.© 2013 .