RAINFALL INTERCEPTION LOSS AS A FUNCTION OF TREE SPACING

Available from UMI in association with The British Library. Requires signed TDF.; This thesis reports on a study of rainfall interception loss as a function of tree spacing. The experiment was carried out at Cloich, 32 km south of Edinburgh. The literature related to the role of interception loss in the water balance of afforested catchments is reviewed and the controversy surrounding the subject is outlined and discussed. A need for detailed experimental work on rainfall interception loss from tree stands of agroforestry systems is identified. The characteristics of the experimental site as well as the various experiments carried out are described in detail. Particular attention is given to the modelling aspect of interception loss as a function of tree spacing. Two methods, the volume balance method and the mass exchange method, were employed to derive the model parameters from four tree spacing treatments of 2 m, 4 m, 6 m and 8 m of Sitka spruce (Picea sitchensis (Bong.) Carr). The experiment spanned a period of two years from 1 July 1987 to 31 August 1989.; The results obtained showed that, on average, interception loss in the closed stand (2 m spacing) was 33% of the gross rainfall, against 24%, 15% and 9% in the 4 m, 6 m and 8 m spacing treatments, respectively. The stand parameters, the free throughfall coefficient and the boundary layer conductance, are identified as responsible for the variation amongst the spacing treatments. The boundary layer conductance per unit ground area was found to decrease with spacing as the number of trees decreased but increased per tree as the result of the increase in the ventilation rate and turbulent exchange within the canopies in widely spaced stands. Spacing dependent functions were derived for the free throughfall coefficient, the boundary layer conductance, the canopy and trunk storage capacities, the free stemflow coefficient, the drainage rate from the canopy and the coefficient of evaporation from the stem. The storage capacities of the canopy and the trunk, the free stemflow coefficient and the drainage rate from the canopy were unaffected by the spacing and were found to be the properties of the individual trees.; The measurements of interception loss from the four treatments are compared with predictions made by two models, the Gash model and a modified implementation of the Rutter model. The latter, modified by inclusion of the derived spacing-dependent functions, is shown to be a good model for predicting interception loss in agroforestry systems.