Experimental site and plant materia

Experimental site and plant material
The experiment was conducted in a research greenhouse
near Murcia, Spain, over a 2-week period
between November 2 (Day 306) and November 14
(Day 318) 1997. Three-year-old apricot trees (Prunus
armeniaca cv. Búlida, about 2 m high and 1.5 m wide
at the base of the crown, total leaf-area of about 6.4
m2) were grown in 70 L plastic pots containing soil
mixture of 2 parts clay loam, 1 part peat and 1 part
vermiculite.
Experimental conditions and irrigation regime
Air temperatures inside the greenhouse ranged
between 20 and 25 ◦C during the daytime and fell to 10 to 15 ◦C at night. The vapor pressure deficit (VPD)
reached a daytime maximum of between 0.7 kPa (Day
307) and 1.2 kPa (Day 314) (Figure 1). Photosynthetically
active radiation (PAR) varied from day to day,
reaching a daytime peak value of between 170 µmol
m−2 s−1 (Day 307) and 340 µmol m−2 s−1 (Day 314),
on about 1/3 of full sunlight (Figure 1).
Four trees were irrigated once per day with nutrient
solution (3.75 mmol L−1 NH4NO3, 1.3 mmol L−1
NH4PO4H2, 0.01 mmol L−1 Fe and 0.01 mmol L−1
Zn). This irrigation was supplied using a single lateral
line that had two compensating drippers (qe =2Lh−1)
for each tree. The irrigation was controlled automatically
by a head unit programmer which maintained
soil matric potential at about −30 kPa (monitored with
tensiometers placed at 20 cm depth). The soil water
content was estimated gravimetrically along the experiment.
Beyond November 9 (Day 313), with the
soil mixture initially at field capacity (θm=31.3%),
water was witheld for four days. Soil water content
declined steadily over this period to reach a value close
to wilting point (θm = 15.2%) by day 317. After these
four days (Day 318 at 10.30 a.m.) pots were then
re-irrigated to their full water holding capacity.
Estimation of transpiration
The diurnal course of transpiration was measured in
two apricot trees growing in 70 L pots. Each pot was
placed on top of a weighing balance (capacity 150
kg and, resolution of 5 g, Ohaus, model 3150510).
The top surface of each pot was covered with plastic
to eliminate soil evaporation. The mass of each pot
was recorded every half an hour using a Campbell
data logger (Model CR10, Campbell Sci, Logan, Utah,
USA).
Measurements of sap flow in the trunk
Sap flow was monitored routinely in the trunk of the
two trees described above. We used the compensation
heat-pulse technique (Swanson and Whitfield,
1981), with one set of heat-pulse probes per tree. Each
set consists of a heater needle of 1.8 mm diameter
and two thermocouple probes also of 1.8 mm dia.
Heater and thermocouples were installed into parallel
holes drilled radially into the trunk, at a height of
0.3 m above the soil surface. Sap velocity was measured
following the procedure of Green and Clothier
(1988), using the theoretical calibrations of Swanson
and Whitfield (1981) to account for the probe-induced
effects of wounding. Once the heat-pulse velocity,