Forest-scale sap flux responses to rainfall in a dryland eucalyptus plantation

Dryland salinity is caused by rising saline water tables, the result of relatively recent landscape-scale clearance of deep-rooted vegetation. One obvious solution to this problem is the reintroduction of deep-rooted vegetation into these landscapes, most likely non-deciduous trees. Ideally, continually-transpiring deep-rooted trees would remove moisture from throughout the soil profile, increasing the capacity of the soil to store water, thus lowering water tables by effectively reducing the number of rainfall events that contribute to groundwater recharge. In this study, we examined how water use by a Eucalyptus sideroxylon A. Cunn. ex Woolls plantation, growing in a salinity-prone landscape, varied in response to rainfall events across four years of sap flux monitoring. Responses of the plantation were observed across multiple seasons, from above average to well below average rainfall. We observed that the plantation forest, while capable of continuous water use during drought, was also quite responsive to rainfall events. During the driest periods, during which shallow soil moisture was reduced to a stable minimum, the forest continued using water at around 1 mm/day. Generally we observed increases in forest water use following only 5 mm of rainfall, in contrast to 20 mm for neighbouring native vegetation. We compared a range of plausible empirical models for describing forest water use responses to rainfall. The best model demonstrated that rainfall size, post-rainfall PET and the interaction between rainfall size and antecedent soil moisture made significant contributions to variation in forest water use across rainfall events. Interestingly, the model showed that all else equal, higher antecedent soil moisture tended to reduce potential increases in forest water use in response to rainfall.