Classical water relations theory predicts that predawn plant water potentia
l should be in equilibrium with soil water potential (soil Psi(w)) around r
oots; and many interpretations of plant water status in natural populations
are based on this expectation. We examined this expectation for two salt-t
olerant, cold-desert shrub species in glasshouse experiments where frequent
watering assured homogeneity in soil Psi(w) and soil-root hydraulic contin
uity and where NaCl controlled soil Psi(w). Plant water potentials were mea
sured with a pressure chamber (xylem Psi(p)) and thermocouple psychrometers
(leaf Psi(w)). Soil Psi(w) was measured with in situ thermocouple psychrom
eters. Predawn leaf Psi(w) and xylem Psi(p) were significantly more negativ
e than soil Psi(w), for many treatments, indicating large predawn soil-plan
t Psi(w) disequilibria: up to 1.2 MPa for Chrysothamnus nauseosus (0 and 10
0 mM NaCl) and 1.8 MPa for Sarcobatus vermiculatus (0, 100, 300, and 600 mM
NaCl). Significant nighttime canopy water loss was one mechanism contribut
ing to predawn disequilibrium, assessed by comparison of xylem Psi(p) for b
agged (to minimize transpiration) and unbagged canopies, and by gas exchang
e measurements. However, nighttime transpiration accounted for only part of
the predawn disequilibrium. Other mechanisms that could act with nighttime
transpiration to generate large predawn disequilibria are described and in
clude a model of how leaf apoplastic solutes could contribute to the phenom
enon. This study is among the first to conclusively document such large dep
artures from the expectation of predawn soil-plant equilibrium for C-3 shru
bs, and provides a general framework for considering relative contributions
of nighttime transpiration and other plant-related mechanisms to predawn d
isequilibrium.