Advancing the pneumatic method to assess xylem vulnerability to embolism among distinct growth forms

Drought-driven plant mortality is closely linked to xylem embolism. Building useful, reliable datasets of xylem vulnerability to embolism requires methods that are practical, fast, accurate, widely accessible and robust across growth forms. We tested and advanced the pneumatic method for constructing xylem vulnerability curves (VCs) across contrasting growth forms to improve inference of drought resilience. Using an automated pneumatron, VCs were constructed for three species representing a small woody shrub (Erica monsoniana), a large woody shrub (Protea repens), and a reed-like graminoid (Cannomois congesta). For graminoid culms, we compared three approaches for estimating xylem water potential (Ψ) and developed a non-invasive method that couples repeated relative water content (RWC) measurements with Ψ-RWC models to obtain high-temporal Ψ estimates. Percent air discharged (PAD)-Ψ relationships were well captured by sigmoid functions. Cannomois congesta showed the steepest curves and the least negative thresholds overall (P₅₀ = -2.91 ± 0.09 MPa), indicating early, rapid embolism progression, whereas Erica monsoniana was most resistant (P₁₂ = -5.91 ± 0.74 MPa; P₅₀ = -6.78 ± 0.76 MPa) with higher variability; Protea repens was intermediate. P₅₀ estimates were the most comparable with prior optical, pneumatic and centrifuge estimates, whereas P₁₂ and P₈₈ showed greater divergence. Ψ_TLP was less variable between species, but ranked similarly (-1.49 ± 0.03, -1.53 ± 0.03, -1.59 ± 0.02 MPa for C. congesta, P. repens, and E. monsoniana, respectively). Such variation yielded systematically wider hydraulic safety margins for the three species. By demonstrating that the pneumatic method can generate reliable vulnerability curves across small and large woody shrubs and graminoids, this study broadens the comparative evaluation of xylem vulnerability across growth forms with contrasting anatomy. A practical advance is the use of repeated RWC measurements paired with Ψ-RWC relationships to improve Ψ resolution in graminoid culms while minimizing disturbance.