2.5. Statistical analyses
Statistical analyses were performed using STATISTICA 10.0 (StatSoft Inc., STATISTICA data analysis software system, www.statsoft.com, USA). A two-way analysis of variance (ANOVA) was used, and differences were estimated using Fisher’s LSD test at the P < 0.05-significance level. Multiple regression analysis (at P < 0.05) was used to develop model equations describing the effects of air temperatures and photosynthetically active radiation (PAR) on the marketable yield of cauliflower. The following data were taken into account for yield modeling: mean, maximum and minimum air temperature up to curd initiation by the plants (when more than 50% of plants in the treatment had curds of at least 0.5 cm diameter) and air temperatures and PAR values from curd initiation to harvest culmination. The values describing thermal conditions up to curd initiation involved temperatures maintained at different levels during the LT treatment of the transplants. Harvest culmination was defined as a day when the majority of curds were harvested from particular experimental treatments. Data used for analyses are presented in Tables 1 and 2. Modeling was performed using the “general regression models (GRM)” module of the STATISTICA software to conduct a stepwise regression procedure with backward elimination to remove the least significant variables. Coefficients of determination (R2) and standard errors of estimation (SEe) were calculated to assess the accuracy of the model.