Xylem recovery from drought-induced embolism: where is the hydraulic point of no return?

For terrestrial plants, maintenance of the integrity of the rootto-leaf water transport pathway is essential for sustaining photosynthetic 9as exchange and growth. The problem of maintaining long-distance water transport is especially challengin9 in trees because path-length resistances and gravity can result in steep gradients of increasing xylem tension from roots to terminal branches, potentially increasing the risk of tension-induced xylem embolism (Sperry and Tyree 1988). These emboli reduce xylem hydraulic conductance and, under certain conditions, can accumulate rapidly through a process that has been called runaway embolism, which can ultimately lead to catastrophic and irreversible hydraulic failure (Tyree and Sperry t988). This phenomenon has led to intensive interest in the role of plant hydraulic dysfunction in drought-induced mortality currently takin9 place in many tree-dominated ecosystems, which is expected to increase under future climate change scenarios (Allen et al. 2010, McDowell et al. 2011, Choat et al. 2012). The susceptibility of xylem to embolism has traditionally been characterized by generating the so-called xylem vulnerability curves, which are plots of the loss of hydraulic conductivity in relation to xylem tension or pressure (Figure 1, top). Key features of these curves that are often quantified include the embolism threshold (Pe), the xylem pressures corresponding to 50 and 88°1"o loss of conductivity (Pro and P88) and the slope of the steep, nearly linear portion of the curve. Thus, highly embolism-resistant xylem might be expected to show a highly negative value of Pe and a gradual slope beyond this point, which would ensure substantially more negative values of Pro and P88. However, vulnerability to embolism does not necessarily equate to risk of embolism in situ, which is partly determined by the influence of stomatal regulation of transpiration on the normal .operating range of xylem tension in a given plant organ. Isohydric species tend to show relatively constant maximum values of xylem tension, whereas the maximum xylem tension in anisohydric species varies according to environmental conditions such as soil water availability and vapor pressure deficit (Tardieu and Simonneau 1998, Schultz 2003, Rogiers et al. 2012). However, isohydry and anisohydry represent two extremes of a continuum of regulation of xylem tension. This limitation on relatin9 the characteristics of xylem vulnerability curves to the actual risk of embolism and hydraulic dysfunction in intact, field-grown plants has contributed to the development of the concept of hydraulic safety margins (Alder et al. 1996, Hacke and Sauter 1996, Pockman and Sperry 2000, Vilagrosa et al. 2003, Brodribb and Hotbrook 2004, Meinzer et at. 2009, Hoffmann et al. 2011, Johnson et al. 2012), which can be defined as the difference between the minimum xylem pressure normally attained in a given plant organ and a reference point on its hydraulic vulnerability curve (e.g., Pro). Intensive research on plant hydraulic architecture has revealed much about broad patterns of xylem vulnerability to embolism and hydraulic safety margins within plants (e.9., Martinez-Vitalta et al. 2002, Domec and Gartner 2005, Domec et al. 2006a, 2006b, Maherali et al. 2006) and across contrasting vegetation types (Pockman and Sperry 2000, Maherali et al. 2004, Jacobsen et al. 2007, Meinzer et al. 2009, Choat et al. 2012). Nevertheless, simple hydraulic predictors of a species’ performance and ability to survive in a specific type of