Soil water storage in a semi-arid Eucalyptus populnea woodland invaded by woody shrubs, and the effects of shrub clearing and tree ringbarking.

Soil water was monitored over a six year period in an intact shrub invaded semi-arid Elrcalyptlts popztlrlea woodland (control) and on areas which had been treated by either shrub-clearing, or by ringbarking of trees and shrub-clearing. Measurements were made under both the shrubby thicket areas near the eucalypt, and the sparsely shrubbed interthicket areas more distant from the trees. Average soil water storage over the six years for all treatments was only 26 nun. Much of this water was stored in the upper 500 Inm of the profile and hence was susceptible to direct evaporation from the usually bare soil surface. In the intact n.oodland and following wet weather, significantly more soil water was stored under thickets than under the interthicket areas. With the return of dry weather this cxtra soil water was rapidly depleted, and thicket soils would often become drier than interthicket soils. After pro- longed dry weather, soil matric potentials of - 10 to -1 2 MPa were recorded at a depth of 500 mm. Matric potentials by this time were least negative under thickets. Shrub clearing without rinpbarking increased thicket and interthicket soil water storage by 17% and 2396 respectively. The ring- barking and shrub clearing treatment increased thicket profile storage more than that of the interthicket (81% and 64% respectively). The effect of ringbarkinp lvas often pronounced at a distance of 25 rn from the tree. The contrasting soil water response to the two treatments indicated that in this semi-arid environment only a relatively srnaU change in soil water balance may accrue from incomplete clearing. The ren~oval of both shrubs and trees is probably necessary to make a large difference to soil water storage.