Mixed hydraulic responses to drought in six common woody species from a dry evergreen sclerophyll forest in South Africa.

Abstract

Despite the emergence of a general conceptual framework for woody tree response to drought, few studies link variation in functional traits of coexisting species to drought outcomes in diverse plant communities. We use a natural drought event to test an ecological prediction from the embolism avoidance hypothesis: that co-existing species of a single growth form (woody trees) will converge upon traits that avoid embolism during all but the most severe droughts. We evaluated hydraulic traits and drought responses of six common woody tree species from South Africa's Albany Subtropical Thicket. For each species, we measured laboratory-based xylem vulnerability and Pressure-Volume curves, and in situ minimum water potentials and four metrics of drought canopy damage during a dry period as well as a subsequent wetter period. We also quantified leaf construction and plant architecture traits, including tree height, Huber value and leaf mass per area (LMA). Species varied in the water potential associated with 50% loss of xylem function (P50), and turgor loss point, leading to between-species variation in stomatal and hydraulic safety margins. All species were shown to withstand leaf xylem water potentials more negative than -4.5 MPa before experiencing embolism. Predicted percent embolism during the dry period was associated with whole-plant drought damage but only following recovery. The LMA, modulus of elasticity, Huber value and tree height were also associated with drought damage, albeit less predictably so. Our results provide support for the embolism avoidance hypothesis and demonstrate how knowledge of species' hydraulic traits can predict canopy dieback during drought events. However, our study also reveals mixed functional responses to drought within a single major growth form (i.e., woody trees) within a community that is composed of multiple growth forms, highlighting the complexity of predicting drought outcomes in diverse communities.