The growth rate in the base treatment was highest for NOR, followed by SWE, and lowest for SOB and PRA. The growth rate differed both between geographic areas, with higher values in the northern area, and between site climates, with higher values in cold sites. However, the difference was stronger when the populations were grouped by geographic area than by climate (Fig. 3). When comparing growth rates of the four populations in the two treatments, there were significant effects of both population and treatment. However, the growth rate differed strongly between treatments for SWE, and to much a lesser extent for SOB. Both presented higher growth rate in the extreme treatment than in the base treatment (Fig. 4a). Net photosynthesis did present significant main effects either among populations or between treatments, but net photosynthetic rate was higher for the SWE plants grown in the extreme treatment than in the base treatment, as shown by a significant population × treatment interaction (Fig. 4b). The plants grown in the extreme treatment had lower stomatal conductance, higher water use efficiency and higher chl content compared to the base treatment, though the patterns of differences between treatments for these response variables were not consistent among populations (Fig. 4c-e). Similar to net photosynthesis, WUE was higher in the extreme treatment that in the base treatment for SWE and to a lesser extent for SOB. The Fv/Fm was high (> 0.75) in both treatments (Fig. 4f). However, there was a significant effect of treatment because of higher mean Fv/Fm values in the extreme treatment. The Fv/Fm did not differ among populations. Although the difference in Fv/Fm values between treatments was weaker in PRA compared to the other populations, there were no overall significant effect of population neither significant treatment × population interaction (Fig. 4f). Mortality was significantly higher in the extreme treatment, but it was not affected by population or by treatment × population interaction (Fig. 4g). Four variables presented significant effects of treatment when comparing the base vs. extreme response, i.e. growth-in-height, gs, WUE and chl content. Based on the RDPI calculated for these four variables the four populations presented a weakly (p = 0.08) significant difference only for WUE that had highest plasticity in SWE (Table 2).