Integration of drought tolerance mechanisms in Mediterranean sclerophylls: a functional interpretation of leaf gas exchange simulators

In this article I investigate how Mediterranean sclerophylls integrate drought resistance mechanisms across organization levels and a functional whole plant interpretation of mechanistic leaf gas exchange models is provided. The feedback between transpiration, carbon gain and soil moisture through a plant gas exchange simulator coupled to a stochastic soil water balance model is described. In previous studies stomatal responses to environment are modeled through an empirical relationship, based on correlative behavior with respect to net assimilation rate, that has been shown to accurately predict gas exchange rates under a range of climatic conditions. In these models, however, the parameter that controls the level of conductance attained for a given assimilation rate (gF) is unknown under drought conditions, precluding their application to water limited ecosystems. In this article I propose an algorithm to compute this parameter based on the assumption that canopy conductance must convey to a strategy of long-term carbon gain maximization and avoidance of mortality by the plant, the unit of natural selection. Two possible negative feedback control strategies between gF and soil moisture, depending on whether the plant senses the environment via leaf or soil water potential are tested. It is found that a strategy operating via leaf potential is more effective from a whole plant perspective than direct sensing of soil water status. The parameter gF provides an integrative description of various processes (e.g. concentration of root generated compounds) leading to stomatal closure. Thus this result supports the view that stomatal control requires both hydraulic and chemical signaling to be effective. Simulations based on a negative feedback via leaf potential predict linear dependency between gF and soil water potential during the dry period as reported by detailed empirical studies on gF variation. Comparison of the whole plant simulator against a leaf based optimization model suggests that stomatal conductance represents an optimal solution of a hierarchically structured system that operates at two temporal scales. At short time scales (minutes to hours) stomata perform a sub-optimal system strategy that allows the plant to optimize water loss per unit of carbon gain. Simultaneously, adjustment of canopy conductance to soil water status allows the plant to maximize carbon gain and avoid mortality at longer time scales (hours to days). These results are relevant both to understand how plants integrate process across functional levels and to proper description of soil water resources in mechanistic biosphere vegetation dynamic models.