Bartlett’s test was applied to establish homogeneity of variance, following which the
data were subjected to ANOVA as a factorial combination of cultivar irrigation
regime measurement date. A separate ANOVA was conducted for each measurement
date and each main factor (cultivar or irrigation regime); in this case, means were separated
on the basis of an LSD test. To define the response of the trend between eco-physiological
parameters, yield, and plant dry weight, analyses of regression (linear and quadratic) were
performed. The significance of differences between regression coefficients was evaluated
by a parallelism test (Ottaviano, 1977).