|dc.description.abstract||This thesis examines the application of the heat pulse velocity technique (HPV) to plantation-grown Eucalyptus grandis in the Eastern Transvaal, South Africa. The work addresses the application of the technique per se and is ultimately focused on improving the prediction of the hydrological impact of afforestation, to assist in the equitable management of South Africa's limited water resources. The verification of the HPV technique on E. grandis against the cut-tree method showed that the technique accurately reflected the water uptake in four three-year-old trees and a sixteen-year-old tree. It was found that accurate measurement of wound size and probe separation was essential for accurate water use estimates. The optimal probe allocation strategy for accurate measurements of transpiration in individual trees and stands of trees was examined. Stratifying the depths of implanted probes resulted in greater precision and repeatability in the HPV-derived estimates of sap flow in E. grandis. Given a limitation in the number of probes available to estimate stand transpiration, the results showed that sampling many individuals with a low
sampling intensity (few probes per tree), rather than sampling few individuals intensively, improved the estimate of stand transpiration. An examination of the influence of tree age and season on transpiration rates showed that the transpiration rate per unit leaf area of E. grandis declined with age. Also, transpiration rates were higher in summer than under equivalent conditions of evaporative demand in winter. A seasonal change in the response of transpiration to VPD was implicated as the primary cause of this shift. A number of models were derived to predict transpiration from E. grandis. The variables vapour pressure deficit
(VPO) and photosynthetically active radiation (PAR) were found to account for a large proportion of the observed variation in transpiration from the age sequence of trees studied. The models developed are applicable to trees of varying age, but are valid only for conditions where minimal soil water stress is experienced. The derived models were tested against two sets of independent data. This confirmed that a simple linear multiple regression adequately describes the relationship between transpiration and the two driving meteorological variables, PAR and VPO, in E. grandis.
The application of a selection of the developed models on a sample data set from Sabie showed that transpiration from a three-year-old stand of E. grandis in summer may be more than double that for a sixteen-year-old stand under the same conditions. Simulated results also showed that transpiration in summer was about 25 to 50% higher than that from the same stand during winter conditions. Simulated transpiration rates from the young E. grandis stands were high, suggesting that further validation
of the estimated rates is required before the models are applied. It is concluded that the HPV method is an ideal technique to estimate water use in E. grandis trees. The models developed represent a major advancement on previous models used to predict the hydrological impact of afforestation on mountain catchments.||en