Firstly, we run a one-way ANOVA on the data of growth-in-height
in the base treatment with population as the fixed factor. The purpose
of this analysis was to assess the effect of provenance on the
performance of plants grown under common warm temperature. Secondly, we
run a nested two-way ANOVA on the data of growth-in-height with climate
nested within geographic area (i.e., northern core area and southern
marginal areas). To this aim we grouped the four populations in two
groups, i.e., cold sites (NOR + SOB) vs. warm sites (SWE + PRA). The
purpose of the nested ANOVA was to assess effects of biogeographic
history, associated with geographic location, vs. putative local
adaptation, associated with climate at the location sites. Thirdly, we
run six two-way ANOVAs on growth-in-height and five ecophysiological
response variables (i.e., net photosynthesis, gs, WUE,
chl content and Fv/Fm) with population
and treatment as fixed factors. We also run a generalised logistic
regression model with logit-link function to determine differences of
mortality across treatments, populations and their interactions.
Fourthly, phenotypic plasticity of the four populations was tested by a
modified version of the relative distance plasticity index (RDPI;
Valladares et al. 2006). The RDPI, ranging from 0 (no plasticity) to 1
(maximum plasticity), was calculated by formula (1) for the variables
presenting significant effects of population in the two-way ANOVAs of
the third step described above.
RDPI = ∑(dij→i′j′/(xi′j′ +
xij))
dij →i′j′ is the absolute value of the difference
xi′j′ - xij;
xi′j′is the value measured in the base treatment;
xi′j′is the value measured in the extreme treatment;