2.1 Experimental design and treatments
The experiment was carried out in the greenhouse of the Swiss Federal Institute for Forest, Snow and Landscape Research WSL (47°21’48” N, 8°27’23” E, 545 m a.s.l.), Birmensdorf, Switzerland. On February 27, 2018, 3 years old sessile oak and European beech saplings (~30-40 cm in height) were planted into 10-liter plastic pots (26 cm in diameter). After transplanting into the pots, the plants were grown for 4 weeks under well-watered conditions (irrigation every 2-3 days) in the greenhouse to recover from the transplant shock. The cultivation soil consisted of semi-decomposed humus and commercial potting soil. The initial soil nitrogen and carbon contents were: 39 mg N kg-1 soil for NH4+-N, 573 mg N kg-1soil for NO3--N, 22.42% for soil C, and 0.82% for total soil N. The greenhouse temperature and humidity during the period of drought duration treatment are shown in Fig. S1.
A split-plot experimental design with three blocks was employed in this study. Each block was divided into two main plots, one of which was randomly assigned for oak (48 individuals) and the other one for beech (48 individuals) (Fig. S2). Each plot was then divided into two sub-plots, one of which was randomly assigned for fertilization (24 individuals) and the other one for the non-fertilization treatment (ambient, 24 individuals) (Fig. S2). Each sub-plot was further divided in to four sub-sub-plots (rows) randomly assigned for one of the four drought duration treatments (6 individuals each). Therefore, a total of 288 plant individuals (144 individuals for each species) was included (Fig. S2).
The fertilization treatment was conducted on 10 June 2018 (just prior to the drought duration treatment) (Fig. S2). The fertilizer (Osmocote Exact 3-4M Standard, 7.0 % NO3--N, 9.0% NH4+-N, 9 % P2O5, ICL, Suffolk, UK), equal to 1.68 g N kg-1 dry soil (0.945 g N kg-1 dry soil NH4+-N, 0.735 g N kg-1 dry soil NO3--N) was added to each pot assigned for fertilization.
After the N fertilization, plants were exposed to four drought duration treatments for 4 months in 2018 (Fig. S3). Plants in the well-watered treatment (D0) were watered thoroughly once a week. There were two moderate drought duration treatments, one of which was watered thoroughly biweekly (D1), and the other one was watered thoroughly once a month (D2) (Fig. S3). The treatment of the longest drought duration in the present study was watered thoroughly once two months (D3) (Fig. S3). We focused on the duration between two watering events and thus did not measure the soil water condition. After the harvest occurred on 1 October 2018, all remaining plants were treated (e.g. well-watering) in the same way until the end of the experiment (Fig. S3).