“…These modifications of the original way to calculate thermal time, however, are still not enough to overcome the empiricism of the approach because it is composed of a combination of linear equations, which introduces abrupt changes at the transition points of the response function. The response of plant processes to temperature is smooth and continuous, which can cause a significant departure from the linearity (SHAYKEWICH, 1995). The thermal time approach has also other disadvantages: (i) it is confusing how people implement the calculation of thermal time on a daily basis, i.e., sometimes the average daily temperature is subtracted from T base and sometimes the minimum and maximum temperatures are subtracted from T base and then averaged (McMASTER & WILHELM, 1997); (ii) it lacks generality because either the end points of the function, defined by T base , T opt , and T max , or the total number of °C for completing a given process or developmental phase are often dependent on genotype, and for many genotypes these values are unknown (McMASTER & SMIKA, 1998;KINIRY & BONHOMME, 1991).…”