A review of whole-plant water use studies in tree.

Weighing lysimeters, large-tree potometers, ventilated chambers, radioisotopes, stable isotopes and an array of heat balance/heat dissipation methods have been used to provide quantitative estimates of whole-tree water use. A survey of 52 studies conducted since 1970 indicated that rates of water use ranged from 10 kg day(-1) for trees in a 32-year-old plantation of Quercus petraea L. ex Liebl. in eastern France to 1,180 kg day(-1) for an overstory Euperua purpurea Bth. tree growing in the Amazonian rainforest. The studies included in this survey reported whole-tree estimates of water use for 67 species in over 35 genera. Almost 90% of the observations indicated maximum rates of daily water use between 10 and 200 kg day(-1) for trees that averaged 21 m in height. The thermal techniques that made many of these estimates possible have gained widespread acceptance, and energy-balance, heat dissipation and heat-pulse systems are now routinely used with leaf-level measurements to investigate the relative importance of stomatal and boundary layer conductances in controlling canopy transpiration, whole-tree hydraulic conductance, coordinated control of whole-plant water transport, movement of water to and from sapwood storage, and whole-plant vulnerability of water transport to xylem cavitation. Techniques for estimating whole-tree water use complement existing approaches to calculating catchment water balance and provide the forest hydrologist with another tool for managing water resources. Energy-balance, heat dissipation and heat-pulse methods can be used to compare transpiration in different parts of a watershed or between adjacent trees, or to assess the contribution of transpiration from overstory and understory trees. Such studies often require that rates of water use be extrapolated from individual trees to that of stands and plantations. The ultimate success of this extrapolation depends in part on whether data covering short time sequences can be applied to longer periods of time. We conclude that techniques for estimating whole-tree water use have provided valuable tools for conducting basic and applied research. Future studies that emphasize the use of these techniques by both tree physiologists and forest hydrologists should be encouraged.