Comparison of non-linear temperature-dependent development rate models applied to in vitro growth of entomopathogenic fungi.

Five non-linear models with three to five parameters, built to quantify the effect of temperature on insect development and microbial growth, were tested to describe the influence of temperature on in vitro-measured growth rates of entomopathogenic hyphomycetes. Data from two isolates of each of the four fungal species, Paecilomyces fumosoroseus, Beauveria bassiana, Metarhizium anisopliae, Metarhizium flavoviride, were used to assess the features of each model. Criteria for model evaluation included the statistical quality of parameters estimates, the goodness of fit to data, as well as the ability to provide estimates of several key parameters: the upper and lower development thresholds, the thermal optimum and the maximal growth rate at thermal optimum. The second model proposed by Brière et al. (1999) was found to be the best, and Ratkowsky's model (1983) also exhibited good features.