Search

The earliest vascular plants had stems with a central cylindrical strand of water-conducting xylem, which rapidly diversified into more complex shapes. This diversification is understood to coincide with increases in plant body size and branching; however, no selection pressure favoring xylem strand-shape complexity is known. We show that incremental changes in xylem network organization that diverge from the cylindrical ancestral form lead to progressively greater drought resistance by reducing the risk of hydraulic failure. As xylem strand complexity increases, independent pathways for embolism spread become fewer and increasingly concentrated in more centrally located conduits, thus limiting the systemic spread of embolism during drought. Selection by drought may thus explain observed trajectories of xylem strand evolution in the fossil record and the diversity of extant forms.

While nonstructural carbohydrate (NSC) storage can support long-lived woody plants during abiotic stress, the timing and extent of their use are less understood, as are the thresholds for cell mortality as NSCs and water supplies are consumed. Here, we combine physiological and imaging tools to study the response of Vitis riparia to a 6-week experimental drought. We focused on the spatial and temporal dynamics of starch consumption and cell viability in the xylem and phloem of the stem. Starch dynamics were further corroborated with enzymatic starch digestion and X-ray microcomputed tomography imaging. Starch depletion in the stems of droughted plants was detected after 2 weeks and continued over time. We observed distinct differences in starch content and cell viability in the xylem and phloem. By the end of the drought, nearly all the starch was consumed in the phloem ray parenchyma (98 % decrease), and there were almost no metabolically active cells in the phloem. In contrast, less starch was consumed in the xylem ray parenchyma (30 % decrease), and metabolically active cells remained in the ray and vessel-associated parenchyma in the xylem. Our data suggest that the higher proportion of living cells in the phloem and cambium, combined with smaller potential NSC storage area, rapidly depleted starch, which led to cell death. In contrast, the larger cross-sectional area of the xylem ray parenchyma with higher NSC storage and lower metabolically active cell populations depleted starch at a slower pace. Why NSC source-sink relationships between xylem and phloem do not allow for a more uniform depletion of starch in ray parenchyma over time is unclear. Our data help to pinpoint the proximate and ultimate causes of plant death during prolonged drought exposure and highlight the need to consider the influence of within-organ starch dynamics and cell mortality on abiotic stress response. [ABSTRACT FROM AUTHOR]