The ecophysiology of drought is bracketed by strong principles of plant gas exchange and hydraulic function and robust descriptions of the impact of climatic conditions on past plant performance. The field is weaker in the prediction of drought impacts, where approaches have largely relied on empirical relationships rather than mechanistic models. Important exceptions have sought to link climatic conditions to plant hydraulic status and the probability of mortality. A key variable in such models is plant water potential (ψ), a synoptic variable that integrates soil and atmospheric moisture stress and is a control point for stomatal regulation and the driver of hydraulic dysfunction. We posit that our knowledge of in situ ψ is inadequate; traditional methods limit measurement frequency, likely cause important transient events to be overlooked, and restrict development of a richer understanding of the impacts of extreme and integrated water stress on plant function. Moreover, although the relationship of loss of hydraulic conductivity to ψ is well known for many species based on short-term laboratory studies of excised tissue, we lack an understanding of the ψ conditions (magnitude, duration, dose, etc.) that drive variation in hydraulic conductivity in situ.
Results/Conclusions
To advance understanding of in situ ψ, we installed stem psychrometers in temperate forest, semi-arid woodland, chaparral and desert species. Psychrometers and pressure chamber measurements showed strong agreement at water potentials as low as -9.5 MPa, and performance was stable for periods of weeks to months. A resampling of half-hourly measurements over multi-week time series showed that traditional “midday” measurements frequently missed the true daily minimum, and biweekly or longer measurement intervals missed important departures from interpolated trends. In Larrea tridentata, psychrometer measurements revealed that maximum daily ψ sometimes occurred after sunrise. In the isohydric species Sequoia sempervirens, ψ exceeded the published threshold for embolism formation during periods of high VPD, but the relationship between minimum daily ψ and maximum daily VPD was weak and depended on predawn ψ and light level. Periods of very low VPD associated with rain and fog sometimes resulted in ψ above the gravitational potential. We also report on seasonal variation in branch hydraulic conductivity in relationship to antecedent ψ. We conclude that high resolution in situ ψ measurements can advance understanding of drought ecophysiology and the development of predictive models of plant response to a changing climate.