Abstract:

Objective: Plant embolism vulnerability is a hot issue in hydraulics studies of woody plants, and is generally assessed by establishing vulnerability curve (VC). Centrifuge method is one of the main method being applied to establish the VC. Previous studies have generally believed that when the Cavitron centrifuge method was used to establish the VC of a long-vesselled tree species, the vessels inside the centrifuged branch was cut open, and the presence of a large number of open vessels causing the embolism artifact, resulting in extremely vulnerable "r-shaped" curve, so significantly overestimated the cavitation vulnerability of plants. This study was intended to test whether the type of VC is affected by the proportion of open vessels at the inter-specific level, and to lay a foundation for the reliability of centrifuge method in assessing the plant embolism vulnerability, therefore, the plant's cavitation resistance can be more accurately evaluated, which provides a reliable basis for the selection of drought-resistant tree species in ecological construction. Methods: Nine tree species with different vessel lengths were selected. According to the Cavitron centrifuge rotor diameter (27.4 cm), nine tree species were divided into three categories:long-vesselled species (average vessel length > 15 cm, Robinia pseudoacacia, Morus alba and Lagerstroemia indica), medium-vesselled species (average vessel length 10-15 cm, Ligustrum lucidum, Amygdalus persica, Hippophae rhamnoides) and short-vesselled species (average vessel length < 10 cm, Acer truncatum, Corylus heterophylla, Populus alba 'I-101'×(P. alba×P. glandulosa '84K'). The VCs were established by Cavitron centrifuge. Maximum and mean lengths of vessels were measured by air injection technique. Relationships between open vessels fractions and characteristics of the VCs were analyzed. Results: The VCs of long-vesselled species and short-vesselled species of A. truncatum and C. heterophylla could be well fitted by single Weibull function, while the VCs of medium-vesselled species and short-vesselled species of I-101×84K could be well fitted by double Weibull function. The VCs of long-vesselled R. pseudoacacia and M. alba were similar, showing "r-shaped", while that of L. indica was "s-shaped"; the VCs of medium-vesselled H. rhamnoides and L. lucidum, and short-vesselled I-101×84K were double Weibull shape consist of two "s-shaped" curves, while the VC of medium-vesselled A. persica was closer to "s-shaped"; the VCs of short-vesselled A. truncatum and C. heterophylla were both "s-shaped". The mean vessel lengths (L_v) of long-vesselled species were close to 20 cm, the proportion of vessels with both ends opened (F_27.4) was about 20%, the proportion of vessels opened from one end to the center of centrifuge rotor (F_13.7) was about 60%; the L_v of medium-vesselled species was close to the radius of centrifuge rotor (13.7 cm), F_27.4 was about 10%, and F_13.7 was about 40%; the L_v of short-vesselled species was below 5 cm, F_27.4 was less than 0.1%, and the maximum value of F_13.7 was 2.68%. The embolism vulnerability indices P₁₂ and P₅₀(xylem water potentials correspond to hydraulic loss of 12% and 50%, respectively) of nine species were positively correlated with the proportion of open vessels (F_27.4 and F_13.7), but P₈₈(xylem water potential corresponds to hydraulic loss of 88%) was not significantly correlated with the proportion of open vessels. Conclusion: At the inter-specific level, this study verified that the proportion of open vessels significantly affected the types of embolism vulnerability curves established by Cavitron centrifuge. When the proportion of open vessels was very few, the VCs were "s-shaped", but with the increase of the proportion of open vessels, the "r-shaped" and "intermediate-shaped" curve may be formed. This conclusion further supports the hypothesis of open vessel artifacts, and provides a theoretical basis for the selection of tree species when measuring the embolism vulnerability by Cavitron centrifuge.

Key words: embolism, vulnerability curve, open vessel, Cavitron centrifuge, Weibull curve