Miticide bioassays with spider mites (Acari:Tetranychidae): Effect of test design and sample size on the precision of lethal concentration estimates

The effect of interval space between the concentration (symmetric or geometric), number of concentrations and overall sample sizes were investigated using a Petri dish bioassay, the miticide propargite and two spider mites species. The precision (width of the 95% confidence intervals) of the LC(50) and LC(90) estimates was shown to be influenced by the test design and total sample size. The results suggest that a symmetric five-concentration design using a minimum total sample size of 480 mites is required to produce reliable LC estimates using the Petri dish method. Asymmetric designs did not increase the reliability of LC estimates. When less than 480 test subjects are available, a symmetric three-concentration design would give adequate LC estimates, but the precision would be less than with a five-concentration design.