Application of the process-based model BIOMASS to Eucalyptus globulus subsp. globulus plantations on ex-farmland in south western Australia: I. Water use by trees and assessing risk of losses due to drought

Abstract The process-based model BIOMASS (McMurtrie, R.E., Rook, D.A., Kelliher, F.M., 1990a. Modelling the yield of Pinus radiata on a site limited by water and nitrogen. For. Ecol. Manage. 30, 381–413) was applied to simulate the water balance at five plantations of Eucalyptus globulus growing in the Mediterranean climatic region in the southwest of Australia. Seasonal patterns of variation of soil water content, measured using a neutron moisture meter, were in good agreement with simulations for two of the sites. At these sites the capacity of the soils to store available water was well defined. In one case the soil water storage capacity was limited by the depth of soil to basement rock, while in the other it was limited by the depth to impenetrable siliceous hardpan. At another two sites, where the soils were deep, measurements of seasonal changes in soil water content showed that roots extracted water from at least 6 m depth. The soil water contents simulated for the surface 3 m depth agreed well with measurements for the winter/spring period but were lower than measured values for the late summer/autumn period. This suggests that the model was not completely successful in simulating partitioning of soil water extraction from different depths. There were small differences between the simulations of variation in soil water content using the earlier version of the BIOMASS model (McMurtrie et al., 1990a) and the modified version (McMurtrie, R.E., Leuning, R., Thompson, W.A., Wheeler, A.M., 1992. A model of canopy photosynthesis and water use incorporating a mechanistic formulation of leaf CO 2 exchange. For. Ecol. Manage. 52, 261–278). The differences are due to different assumptions for estimating the average maximum stomatal conductivity ( G c ) of the canopy. In the earlier version of the model, a constant value of G c was assumed for all plantations. However, in the modified version G c varied with integrated C assimilation for each plantation, which in turn depended on the proportion of shaded leaves in the canopy. Simulations using the modified version agreed more consistently with measured soil water contents. At a fifth site, where roots would have access to a watertable at 3 m depth, the agreement between the simulated seasonal pattern of soil water content and values estimated from measurements was poor. Measured soil water contents were much higher than those simulated. This observation is explained by upward movement of water in the capillary fringe above the watertable. The success of the BIOMASS model in simulating soil water contents at plantations where the soil depth, and therefore soil water storage, can be defined, suggests the model has potential for assessing the risk of tree mortality due to drought on these sites. The risk is discussed in relation to effects of rainfall and its seasonal distribution, soil water storage and leaf area index.