Experiment 2: Does embolism spreading also depend on the conduit size?
The percentage of cumulative embolised xylem area in leaves with cut-open minor veins and intact leaves was plotted against time for the six species studied (Figure S5). For leaves with several artificial cuts in the 3rd or 4th order veins, no immediate embolism formation was detected after the cutting, except for the vessels that were cut open. In all six species, however, we observed embolism events in minor veins adjacent to the cuts during the first hours of desiccation (Figure 3, Figure S3). These embolism events in minor veins, with relatively narrow and short vessels, occurred prior to embolism formation in the wide and long vessels of major veins. Since the amount of embolised pixels near the cuts had a local effect only and was relatively small compared to the total amount of embolised pixels that could be detected in the entire leaf area, there was no large overall difference in embolism spreading between intact and cut leaves. Therefore, the embolism spreading patterns did not differ when the percentage of embolised pixels of the entire scanned leaf area was plotted against time. After the artificially induced embolism events near the cuts, embolism spreading occurred in the major veins, and then spread to the minor veins in both intact and cut leaves (Figure 3).