Toward a better interpretation and use of thermal time models

Thermal time models express the temperature dependence of plant or organ development and are mostly used to predict phenological events over time. The most general thermal time model is the accumulation of daily mean temperatures above a threshold temperature [theta] during a development phase. The predictive quality of the model depends on the value of the parameter [theta]. However, the physiological and mathematical bases for [theta] estimation are sometimes forgotten, leading to questionable interpretation. Threshold temperatures [theta] have been estimated for the flowering to harvest phase of two tropical fruit species. For strawberry guava, a large and variable set of data, collected in orchards at different elevations, was available, and [theta] was estimated by five different statistical methods. For mango, few data were available and [theta] was estimated as a parameter of the fruit growth curve. Estimated [theta] values were - 3.9°C for strawberry guava and 20.0°C for mango. For strawberry guava, discrepancies between the model and data were significantly correlated with orchard elevation, fruit set ratio and flowering to harvest duration, but not with the mean temperature of the phase. These results are discussed with a special emphasis on three points which could improve thermal time models and their use: the significance of the threshold temperature [theta], the domain of validity of the model, and the influence of factors other than temperature on plant development. (Resume d'auteur)