A COMPARISON OF HEAT PULSE AND DEUTERIUM TRACING TECHNIQUES FOR ESTIMATING SAP FLOW IN EUCALYPTUS-GRANDIS TREES

Sap flow rates were measured simultaneously by the heat pulse and deut erium tracing techniques in nine Eucalyptus grandis W. Hill ex Maiden. trees at two sites (1) to compare results from the two techniques and (2) to assess the impact of the assumptions underlying the deuterium tracing method on the calculation of sap flow for a range of tree size s. The trees ranged in height from 4 to 14 m with leaf areas of 5 to 3 5 m(2). In all trees, sap flow estimated by the deuterium tracing tech nique was higher than sap flow estimated by the heat pulse method: wit h differences of 11 to 43% in eight of the trees and 113% in one tree. The largest difference was attributed to errors in the heat pulse met hod, as indicated by aberrant relationships between sap flow measured by the heat pulse method and tree size characteristics (i.e., diameter , sap wood area, leaf area) for that tree compared with the other expe rimental trees. Drilling holes in the trees to allow injection of deut erium had no significant effect on sap flow, even when 32 holes were d rilled. Sap flow measured by the heat pulse method was only lower afte r drilling than before drilling in three trees, and the difference onl y persisted for about 1 h. Deuterium concentrations of water collected from the tree canopies had not returned to background values 17 days after injection. Twenty-one days after injection, sapwood and heartwoo d samples taken from trunks near the injection sites contained conside rable concentrations of deuterium, indicating that some of the deuteri um injected into the trees was still present. An experiment performed on two trees showed that deuterium was stored in the heartwood and sap wood throughout the trees, and its distribution within the trees four days after injection was similar whether it was injected into only the sapwood (where it should mix with sap and be transported from the tre e most readily) or into both the sapwood and heartwood, indicating tha t there was considerable movement of deuterium between the heartwood a nd sapwood. Deuterium storage was accounted for by an approximate mean s in the sap flow calculations, and may have resulted in an error of a bout 10% in sap flow estimated by this method. We conclude that the he at pulse and deuterium tracing techniques can be used simultaneously t o increase the number of sap flow measurements obtained from a forest, thereby increasing the precision of forest water use estimates. Their combination would be most effective in stands with a wide range of tr ee sizes and sap flow rates, where the relative differences in sap flu x estimates between the methods is small compared with differences in sap flow between trees.