Water potential measurements
Psychrometers (PSY1, ICT International, Armidale, NSW, Australia) were applied to obtain water potential values from the leaves that were simultaneously measured with the optical and pneumatic method. Sand paper with a grit size of 400 was used to carefully remove a small area (10 mm2) of leaf cuticle under a stereomicroscope, while paying special attention not to abrade the vascular bundles and introduce air-entry into the xylem. Vaseline was used to seal the psychrometer chamber to the abraded surface of the leaf. Water potential measurements were taken every 10 min, and the measurements were stopped when leaves were completely desiccated, or the water potential showed no further decrease over a long period. The duration of measurements was between one and two days.
At the beginning of each dehydration experiment, the leaf water potential dropped rapidly within one hour. After that, the water potential decreased slowly and steadily. To test the accuracy of the xylem water potential values taken with psychrometers, we applied two different approaches. Firstly, for leaves that were attached to long branches, we measured not only leaf xylem water potential, but also stem xylem water potential with another stem psychrometer (Fig. S1). This comparison provided a reasonably good match between both organs, although differences became more pronounced at high levels of dehydration for some species, with xylem water potential decreasing more quickly for leaves than stems. An exception was F. sylvatica , which had more negative xylem water potentials for leaves than stems when water potentials in leaves were less negative than -5 MPa. Secondly, water potential values of detached leaves that were measured with the Pneumatron and optical method were compared to a second set of leaves. Therefore, two leaves with a similar size were excised from the same branch at the same time. One leaf was attached to a psychrometer, while the other one was left to dry under similar conditions (same light intensity, temperature, and humidity). The water potential of the second leaf was measured with a pressure chamber (PMS Instrument Company, Albany, OR, USA) at an interval of 20 to 60 min to validate the accuracy of the stem psychrometer measurements during the first hours of dehydration. These tests revealed an overall good agreement between both methods for three species tested (C. betulus , F. sylvatica , and Q. petraea ; Fig. S2).