2.6 Calculation and statistical analysis
The relative height increment rate (RHI) was calculated based on height measured on 30 May, 2018 (A), on 10 October 2018 (B), and on 8 October 2019 (C), using [(B – A) / A] × 100% for the 2018 growth, and [(C – B) / B] × 100% for the 2019 recovery growth. The NSC concentration was calculated as the sum of the concentration of soluble sugars plus that of starch for each sample.
All data were checked for normality with the Kolmogorov-Smirnov test and for homogeneity of variance with Levene’s test. The effects of species, drought duration, N-fertilization, and their interactions on the parameters measured on each date were analyzed using linear mixed effect models, with block and main plot as random effects. Within each species, the effects of drought duration, N-fertilization, and their interactions on the parameters measured on each date were also analyzed using linear mixed effect models, with block and main plot as random effects, followed, if significant, by Tukey’s post-hoc test. Regression analysis was used to test the relationship between 2019 relative height increment rate and tissue NSC levels (both pre- and post-winter). All analyses are carried out by the package “LME” in R v.3.2.5 (R Core Team).