Figure legends
Figure 1 . Leaf xylem vulnerability curves of six deciduous
species using the optical method applied to different samples: a
detached leaf (red), leaf attached to a short stem segment (blue), and
to a long branch (green). For each sample type, four to five replicates
were measured (pale colours). Bright lines are regression lines for each
type of sample, and grey lines indicate 95% confidential intervals.
PEP50 values (MPa) of the three types of samples were
presented with mean ± standard deviation (n = 4 or 5), with different
small letters showing significant difference between each sample type
(P <0.05). Note that the curve fitting may give the
impression that embolism occurs at positive xylem pressures, while the
raw data make clear that embolism spreading never occurs at a xylem
water potential that is less negative than -0.1 MPa.
Figure 2 . Relationship between the segmentation index and the
difference in PEP12 (red), PEP50(yellow) and PEP88 (green) values between detached
leaves and leaves attached to a short stem segment. A segmentation index
was defined as the maximum vessel length divided by the petiole length,
with vessels running into the leaf blade when the value is
> 1, and the longest vessels ending before the leaf blade
when the value is <1.
Figure 3 . Maps of embolism events between an intact leaf (A-F)
and a leaf with cuts in minor veins (G-L) of Quercus petraea . The
cut and intact leaf pair were attached to the same long branch, with
images taken for both leaves after a certain desiccation time (shown on
the left). White solid lines represent artificial cuts of the
3rd or 4th veins, and white arrows
point at embolism in minor veins near the cuts, which happened much
earlier than embolism in major veins.
Figure 4 . Xylem vulnerability curves of leaves of five
deciduous species based on the optical and pneumatic method. Both
methods were applied to the same detached leaf. Pale and bright red
lines represent raw data and regression lines of the optical method,
respectively. Blue lines represent raw data (pale blue) and regression
lines (bright blue) based on the pneumatic method. For each species,
four leaves were tested, and P50 values (MPa) of both
methods were presented.
Figure 5 . Correlation between paired values of
P12 (a), P50 (b), and
P88 (c) as measured with the optical and pneumatic
method. Each dot represents a single leaf, solid black lines indicate
the regression line, blue dashed lines indicate the 95% confidence
intervals, and black dashed lines indicate the 1:1 line.
Figure 6 . Illustration of the optical method and pneumatic
measurements on detached leaves and possible embolism spreading during
dehydration after 5 and 10 hours. The Pneumatron was connected at the
basipetal end of the petiole, while the leaf area scanned with the
optical method (pale yellow square) is separated from the cut-open
vessels (white) by at least several end walls. (a) The initial status
when leaves were fresh, and all intact vessels were water-filled (blue).
(b, c) Embolism spreads to various intact vessels (green), from which
gas is extracted with the Pneumatron, and which are observed with the
optical method when these are within the scanned leaf area. It is
assumed here that the Pneumatron extracts gas across two end walls.
Arrows indicate gas diffusion through intervessel walls towards the
Pneumatron. Embolism formation in vessels only captured with the optical
method are coloured yellow. Adapted from Jansen et al. (In
press).