Vessel length
Maximum vessel length of stems and leaf petioles was determined by
applying the air injection method (Greenidge, 1952). After connecting
the basipetal part of a stem or petiole to a syringe, a ca. 150 kPa
pressure was applied, while the acropetal part was kept under water.
Successive cuts at the proximal stem/petiole end were made under water
with a razor blade at intervals of 10 mm for stems and 2 mm for leaves
until the first continuous stream of air bubbles could be seen emerging
from the cut end. The corresponding length was then measured and
recorded as the maximum vessel length of a stem or leaf petiole
(MVLstem and MVLpetiole, respectively).
At least six stems or leaves were taken for these measurements (Figure
S4).
Hydraulic segmentation between short (0.5cm in length) stem segments and
leaf petioles was also tested for six samples per species based on the
air injection method. The syringe was connected to the short stem
sample, and the leaf was shortened until bubbles could be seen emerging
from the cut end.
The vessel length distribution of leaf xylem was measured with a
Pneumatron device (Pereira et al. , submitted). Instead of
injecting air (Cohen et al. , 2003; Wang et al. , 2014; Panet al. , 2015), the amount of gas that could be sucked up via
cut-open conduits allowed us to measure the air conductivity of open
vessels while shortening leaf petioles. We then plotted the air
conductivity of the cut-open vessels against the petiole length. The
average vessel length was obtained by fitting the vessel length equation
from Sperry et al. (2005) to our data.
We defined a segmentation index as the maximum vessel length at the
petiole end divided by the petiole length. This index indicated to what
extent the longest vessels from the petiole end run into the leaf blade.
Values < 1 indicated that vessels ended before the leaf blade
started, while values > 1 suggested that at least some
vessels starting at the petiole end run directly into the midrib of the
leaf blade.