Skip to main content
ARS Home » Research » Publications at this Location » Publication #68454


item Thomas, Donald
item Mangan, Robert

Submitted to: Journal of Economic Entomology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/24/1996
Publication Date: N/A
Citation: N/A

Interpretive Summary: Heat treatments can be used to disinfest fruits and vegetables of insects. Such treatments can be used to ensure that imported commodities do not spread quarantined pests without the application of pesticides. In order to assure that such treat- ments are reliable, it is necessary to identify the right combina- tion of temperature levels and treatment times that will kill the insects but not damage the fruit. Estimating the reliability of such treatments requires a mathematical analysis of experimental data in which insect survival is measured over a range of temperatures and treatment times. The present article compares four different mathematical formulas for determining death rates in the Mexican fruit fly using one large set of experimental data. The analysis concludes that a mathematical formula called the "Thermal Kinetic Model," widely used to measure the effec- tiveness of pasteurization of liquids, and a formula called "Probit Analysis," long used to measure the effectiveness of pesticides, give similar and reliable estimates. Two other formulations gave less reliable estimates.

Technical Abstract: A large set of mortality data for the Mexican fruit fly from induced thermal stress was fit to four mathematical models which are used to estimate thermal death points for quarantine level security. Two of the models, the Probit and Kinetic formulas, estimated death point intercepts in close agreement at the tail ends of survivorship curves. Based on the criteria of chi-square values, coefficients of determination, and fiduciary limits, the Kinetic model had the best fit to the data. The Complementary Log-Log Model produced disparately lower death point intercepts with wider fiduciary limits compared to the other three models with the same data set.