Statistical analysis
The relationship between plant height and DAP was fitted using a three-parameter sigmoidal model, and when DAP for individual height approached 50% of its maximum height, (H50) was calculated and designated as a proxy of growth rate. Leaf AT data were fitted to a quadratic function and the optimum temperature for photosynthesis (Topt, oC) and Asat at Topt were estimated (Gunderson, O’hara, Campion, Walker, & Edwards, 2010). We were limited to four glasshouse bays for our experiment, so we did not have replicate bays for the growth temperature treatments. Therefore, acknowledging pseudo-replication as a function of our complex experimental design, we were limited to considering individual plants within each bay as independent experimental units (i.e. replicate), and subsequently tested the treatment effects with a general linear model, as has been done in our previous studies (Lewis et al. 2013; Duan et al. 2014), as well as in other studies (Apgaua et al. 2019). We used a two-way ANOVA to test the main and interactive effects of temperature and water deficit stress on gas exchange variables, biomass and AT response parameters. We also used a two-way ANOVA to assess the effects of water deficit stress and nocturnal warming within each growth temperature regime (i.e. cool and warm). A three-way ANOVA was used to test the instantaneous effects of growth temperature, water deficit stress and heatwave on leaf gas exchange variables during the heatwave and recovery stage, as well as on harvested biomass at the end of the experiment. The differences in physiological and growth variables across treatment combinations were assessed using Tukey’s HSD post hoc . Data were tested for normality and homogeneity of variance before ANOVA was performed. Statistically significant differences were considered if P ≤0.05. All data and statistical analysis were performed in R 3.5.3 statistical computing environment (R Development Core Team 2014). ANOVA was performed using the lm() in combination with the Anova() function in thecar package, and multiple comparisons were conducted with the HSD.test() function in the agricolae package.