Peer review

The importance of predicting the seasonal occurrence of insects has led to the formulation of many mathematical models that describe development rates as a function of temperature. Yet many of the widely used models do not provide acceptable results for predicting development times. After a careful review of the literature, we believe the biophysical model of Sharpe and DeMichele (1977; J. Theor. Biol. 64: 649–670) is the most suitable for this purpose. This model provides an excellent description of development rates over a full range of temperatures, and can be modified easily to describe rates over a portion of that range. Herein we review the literature on modeling insect development rates, describe the Sharpe and DeMichele model, and present easy instructions for its use. A computer program, assembled from the Statistical Analysis System (SAS) Library, determines the correct number of parameters to be used in the model for a given data set, selects starting values of these parameters for nonlinear regression, and computes least-square estimates of the parameters by using Marquardt techniques.

Text in English