Mountain ash (Eucalyptus regnans F.J. Muell.) forest catchments exhibit a strong relationship between stand age and runoff, attributed inter alia to differences in tree water use. However, the tree water use component of the mountain ash forest water balance is poorly quantified. We have used the sap flow technique to obtain estimates of daily water use in large mountain ash trees. First, the sap flow technique was validated by means of an in situ cut tree experiment. Close agreement was obtained between the sap flow estimate of water use and the actual uptake of water by the tree from a reservoir. Second, we compared the variability in sap velocity between a symmetric and an asymmetric tree by using multiple sap flow loggers. In the symmetric tree, velocity was fairly uniform throughout the xylem during the day, indicating that accurate sap flow estimates can be obtained with a minimal number of sampling points. However, large variations in sap velocity were observed in the asymmetric tree, indicating that much larger sampling sizes are required in asymmetric stems for an accurate determination of mean sap velocity. Finally, we compared two procedures for scaling individual tree sap flow estimates to the stand level based on stem diameter and leaf area index measurements. The first procedure was based on a regression between stem diameter and tree water use, developed on a small sample of trees and applied to a stand-level census of stem diameter values. Inputs to the second procedure were tree water use and leaf area of a single tree and the leaf area index of the stand. The two procedures yielded similar results; however, the first procedure was more robust but it required more sampling effort than the second procedure.

Sap flow measurement techniques, such as the heat pulse (compensation) method, are practical means for estimating the water use of individual trees and are often the only reasonable alternative for measuring forest and woodland transpiration in complex heterogeneous terrain. The need to scale estimates of water use from a sample of individual stems to a stand (population) of known area may be satisfied by applying scalars of flux based on tree size or domain. We estimated the aggregate errors in applying the heat pulse technique to the estimation of stand transpiration in a poplar box (Eucalyptus populnea F.J. Muell.) woodland in southeastern Queensland, Australia, by a combination of precision analyses, experimental validation and Monte Carlo simulations of sampling errors. Errors in sap flux density measurements were approximately 13%. The potential error in the flux estimates for individual stems with stratified sampling of sap flux density with depth and bole quadrant based on four sensors was an additional 25%. Conducting wood area, diameter at 1.3 m, leaf area and domain based on Ecological Field Theory all proved excellent scalars of flux at the stand level. With a sample size of six trees stratified by diameter, coefficients of variation in scaling to the stand level were approximately 5% for any of these scalars. The greatest potential source of error in estimating stand transpiration by the heat pulse method was in the measurement of the fluxes of individual stems; scaling these measurements to a homogeneous stand of trees involved less uncertainty.

There is a need to generalize water use behavior of eucalypts to facilitate bioengineering and landscape remediation programs in a wide range of Australian environments. A critical question can be stated as a null hypothesis: tree water use per unit leaf area (leaf efficiency) is independent of eucalypt species. This is implicitly equivalent to the hydrological equilibrium hypothesis that states that leaf area is a function of climate, at least in cases where transpiration and growth are limited by soil water. Failure to reject this null hypothesis simplifies (a) the selection of tree species for water balance management, (b) the generation of regional-scale expectations of leaf area index, and (c) the estimation (monitoring) of the effectiveness of plantations in controlling site water balance. The hypothesis was tested with tree water use data collected in natural multi-species stands across Australia, including sites in the wet-dry season tropical woodlands of the Northern Territory, the Mediterranean climate forests of Western Australia, and a woodland system in southern New South Wales receiving an even distribution of rainfall throughout the year. We also tested the hypothesis in a multi-species tree plantation growing on a saline gradient. In each case, we could not reject the hypothesis of constant leaf efficiency among eucalypts. In every case there was a common, strong, linear relationship among tree leaf area and mean daily water use by all tree species in a sample. Single factor (species) analysis of variance did not detect significant differences between leaf water efficiencies of species. For the jarrah forest (Eucalyptus marginata J. Donn ex Sm., E. calophylla R. Br. ex Lindl.), the null hypothesis held in both spring (wet) and autumn (dry) conditions. The null hypothesis held in the mixed species woodland of New South Wales (E. macrorhynca F.J. Muell. ex Benth., E. blakelyi Maiden., E. polyanthemos Schauer.) under summer and autumn conditions, and across five species in the wet-dry tropical woodland (E. miniata A. Cunn. ex Schauer, E. tetrodonta F.J. Muell., E. porecta S.T. Blake, Erythrophleum chlorostachys F.J. Muell., and Terminalia ferdinandiana Exell.). The null hypothesis also held for a plantation of E. occidentalis Endl. and provenances of E. camaldulensis Dehnh. growing on a shallow saline gradient; i.e., leaf water efficiency remained constant across species and varieties despite obvious effects of salinity on the size of individual canopies. We conclude that there is little evidence for rejecting the hypothesis that leaf efficiency does not vary significantly among sympatric eucalypt species in rainfall-limited (soil-water-limited) systems. These findings open the way for useful bioengineering generalities about the hydrological role of trees in the Australian landscape.

No abstract

No abstract