Savannas and seasonal drought: The landscape-leaf connection through optimal stomatal control.

다국어 초록 (Multilingual Abstract)

At a California blue oak savanna site, this research performs the first, full continuous partitioning for a savanna of evapotranspiration from the understory and tree layers through sap flow and eddy covariance measurements over seasonal cycles. Advanced statistics are used to identify the non-linearities in responses of canopy stomatal conductance, G<italic>_V</italic>, to meteorological drivers, and in particular the critical soil moisture points for blue oak water stress are identified. Apparent coordination in the seasonal use of water and energy resources in the out-of-phase phenology of the grasses and trees hints at optimization between growth forms for the annual useage of water in this Mediterrean ecosystem.
Applying the probabilistic ecohydrological optimality model of hydrologists Rodriguez-Iturbe and co-workers (2001), the different levels of statistical water stress experienced by the grasses and trees are differentiated during the different seasons helps explain their out-of-phase growth periods. However, this hydrologists' model's lack of inclusion of the role of energy, the driver of transpiration, motivates an examination of the question of optimality as developed by plant ecophysiologists in leaf gas exchange models. It is found that their complementary lack of inclusion of a soil moisture budget in turn limits their leaf level optimality models, since there is no clearly specified constraint for the optimality criterion. Thus, this research proposes a complete optimality model that clarifies the necessity of linking both leaf and landscape scales, the work of both ecophysiologists and hydrologists. A new mathematical framework is required, that of stochastic optimal control, in which the plant's stomatal conductance is steered by expectations of future water availability at the landscape scale, which have been made by natural selection. Thus, this framework incorporates adaptation to climate as well as response to immediate environmental conditions.
This research adds to the database one of the more complete budgets of ecosystem fluxes of water vapor, increases coverage of arid and savanna ecosystems in the flux measurement community, and clarifies the meaning of optimality of water useage at leaf and landscape scales.