Physiological and morphological response patterns of Populus deltoides to alluvial groundwater pumping

We examined the physiological and morphological response patterns of plains cottonwood [Populus deltoides subsp. monilifera (Aiton) Eck.] to acute water stress imposed by groundwater pumping. Between 3 and 27 July 1996, four large pumps were used to withdraw alluvial groundwater from a cottonwood forest along the South Platte River, near Denver, Colorado, USA. The study was designed as a stand-level, split-plot experiment with factorial treatments including two soil types (a gravel soil and a loam topsoil over gravel), two water table drawdown depths (similar to0.5 m and > 1.0 m), and one water table control (no drawdown) per soil type. Measurements of water table depth, soil water potential (Psi(s)), predawn and midday shoot water potential (Psi(pd) and Psi(md)), and D/H (deuterium/hydrogen) ratios of different water sources were made in each of six 600-m(2) plots prior to, during, and immediately following pumping. Two additional plots were established and measured to examine the extent to which surface irrigation could be used to mitigate the effects of deep drawdown on P. deltoides for each soil type. Recovery of tree water status following pumping was evaluated by measuring stomatal conductance (g(s)) and xylem water potential (Psi(xp)) on approximately hourly time steps from before dawn to mid-afternoon on 11 August 1996 in watered and unwatered, deep-drawdown plots on gravel soils. P. deltoides responded to abrupt alluvial water table decline with decreased shoot water potential followed by leaf mortality. Psi(pd) and percent leaf loss were significantly related to the magnitude of water table declines. The onset and course of these responses were influenced by short-term variability in surface and ground water levels, acting in concert with physiological and morphological adjustments. Decreases in Psi(pd) corresponded with increases in Psi(md), suggesting shoot water status improved in response to stomatal closure and crown dieback. Crown dieback caused by xylem cavitation likely occurred when Psi(pd), reached -0.4 to -0.8 MPa. The application of surface irrigation allowed trees to maintain favorable water status with little or no apparent cavitation, even in deep-drawdown plots. Two weeks after the partial canopy dieback and cessation of pumping, g(s), and Psi(xp) measurements indicated that water stress persisted in unwatered P. deltoides in deep-drawdown plots.