Trait measurements
We measured 27 traits (Table 2) which together describe plant morphology, leaf structure, plant gas exchange and its temperature dependence, and plant water use strategy. For the thirteen of these traits which we measured at more than one elevation, we calculated plasticity indices. We made these trait measurements on 16 target species (Table S1), chosen to represent several plant families, to be widely present in the turfs in most treatments, and to have leaves which fitted into the gas analyzer.
Traits describing plant morphology and leaf structure were measured at the source elevation (2000 m) in August 2019 and 2020. To measure leaf area (LA), leaf mass per area (LMA), leaf thickness (LT), and leaf dry matter content (LDMC), we collected leaves of the target species, scanned them and measured their fresh and oven-dried weight. LT was measured with calipers from the fresh leaves. LA was estimated from the leaf scans with the software ImageJ (Rasband 1997). Canopy shape index (CSI) was measured as the ratio of mean canopy width to height (ranging from near zero for tall plants to >1 for prostrate plants). Rooting depth (Rootd) and width (Rootw) were measured from excavated root systems.
To characterize plant gas exchange, we measured photosynthesis (A), nighttime respiration (R), stomatal conductance during day and night (GS, GSnight), transpiration during day and night (E, Enight; for the importance of night-time water flux, see Snyder 2003; Daley & Phillips 2006), instantaneous water use efficiency (WUEinst) and intrinsic water use efficiency (WUEint) with an infra-red gas analyser (CIRAS-2, PPSystems, Hitchin, UK). To quantify how photosynthesis or respiration varied after transplantation to lower elevation, measurements were taken from two sites: +3°C (1400 m) and +0°C (2000 m). Temperature in the leaf chamber was kept constant at either 24°C (day measurements) or 14°C (night measurements). Gas exchange measurements were taken in 2019 and 2020, 3-4 years after the transplantation.
To characterize the instantaneous response of gas exchange to temperature, we measured photosynthesis and respiration response to temperature at 2000 m (+0°C) using both a CIRAS-2 and Licor-6800 (LICOR, Lincoln, Nebraska, USA). Photosynthesis was measured at 14°C – 34°C, at 4°C intervals, and respiration at 5°C – 30°C, at 5°C intervals. After, the leaves were scanned, oven-dried, weighted and milled to measure carbon (C) and nitrogen (N) content either with combustion analysis (for large samples) or with an elemental analyzer coupled to a mass spectrometer (for small samples, Werner et al. 1999; Werner & Brand 2001). Leaf scans and weights were used to estimate change in LMA and LA resulting from transplantation. Photosynthetic nitrogen use efficiency (PNUE) was calculated by dividing mass-based photosynthetic rate (A/LMA) with the N content of the same leaf.
Plant water use strategy was characterized by measuring stable isotope ratios in 13C/12C and18O/16O (‰) from leaf cellulose to determine long-term stomatal conductance and water use efficiency (δ13C and δ18O; Scheidegger et al. 2000; Moreno‐Gutiérrez et al. 2012). δ13C is a proxy for intrinsic water use efficiency (A/GS) and δ18O correlates negatively with stomatal conductance (GS), both integrated over the leaf life span. Leaf samples collected at each site were oven-dried and cut into small pieces for cellulose extraction. Measurements of δ13C and δ18O from the extracted cellulose were performed using a PYROCUBE (Elementar, Hanau, Germany) connected to a Delta Plus XP mass spectrometer (Thermo Finnigann, Bremen, Germany) at the laboratory of stable isotopes at the Swiss Federal Institute of Forest, Snow and Landscape (WSL), following Weigt et al. (2015).
We measured the depth of water uptake by comparing the δ18O of root crown water to the δ18O of water from different soil depths (Barnard et al. 2006; Moreno‐Gutiérrez et al. 2012). Root crowns were collected into airtight vials and pooled together per species and site to ensure enough water per sample. Soil was collected from each site with soil corers (max 35 cm depth) and separated every 5 cm into airtight glass vials. Water was extracted from samples in the lab by using a water bath and cooling traps. δ18O was analysed from the extracted water with a High Temperature Combustion Elemental Analyzer (TC/EA; Thermo Finnigan, Bremen, Germany) and a mass spectrometer (Delta plus XP; Thermo Finnigan) (Saurer et al. 2016). Since each site was sampled on only one occasion (and the measurements could be influenced by weather), we used the average depth across all four sites as the species-level trait in the analyses. As source water δ18O can drive variation in leaf cellulose δ18O (Roden & Farquhar 2012), we used δ18O enrichment (δ18O leaf cellulose – δ18O root crown water) as the estimate of long-term stomatal conductance (Scheidegger et al. 2000; Moreno‐Gutiérrezet al. 2012). See Supporting Information Table S2 for sample sizes for all traits.