2013 ASHS Annual Conference

Mechanistic water use models represent a potential step forward for predicting irrigation scheduling in horticulture systems.  The mathematical framework for transpiration estimates centers around stomatal conductance (g_s) submodels.  Several of the g_s models contain a parameter that specifies the minimum g_s estimate (g₀).  Sensitivity analyses with a canopy flux model (MAESTRA) identified g₀ to have the greatest influence on transpiration estimates (seasonal mean of 40%).  A canopy spatial analysis revealed the influence of g₀ to vary (30% to 80%) with the amount of light absorbed by the foliage and to increase in importance as absorbed light decreased.  The parameter is typically estimated by extrapolating the linear regression fit between observed g_s and net photosynthesis (A_n).  However, our measurements demonstrate that the g_s-A_n relationship becomes nonlinear at low light levels and thus, extrapolating values from data collected in well-lit conditions resulted in an underestimation of g₀ in Malus domestica when compared to measured values (20.4 vs. 49.69 mmol·m^-2·s^-1 respectively).  In addition, extrapolation resulted in negative g₀ values for three other woody species.  We assert that g₀ can be measured directly with diffusion porometers (as g_s when A_n ≤ 0), reducing both the time required to characterize g₀ and the potential error introduced by statistical approximation.  Incorporating measured g₀ into MAESTRA significantly improved transpiration predictions (6% overestimation vs. 45% underestimation, respectively), demonstrating the benefit in g_s models.  Foremost, diffusion porometer measurements offer a viable means to quantify the g₀ parameter, circumventing g₀ estimate errors associated with linear extrapolation of the g_s-A_n relationship.