Xylem vessel length and centrifuge measures of xylem cavitation resistance

Background/Question/Methods

Vessel length is a key trait in determination of plant hydraulic function and the structure of plant hydraulic pathways, yet this trait is not often studied.  Additionally, the reliability of centrifuge-based techniques in the generation of xylem vulnerability curves has recently been questioned and it has been suggested that species with long vessels may be susceptible to measurement artifact.   We compared centrifuge vulnerability curves to previously published dehydration curves in 10 chaparral shrub species with maximum vessel lengths ranging from 23 to 196 cm.  We also examined previously published comparisons of centrifuge-based techniques and non-centrifuge based methods, such as dehydration or air-injection.  This comparison included both long and short vesselled species.  Data on vessel length were compiled to evaluate the prevalence of species with long vessels that may require special methodological considerations.

Results/Conclusions

Centrifuge techniques were not equally susceptible to artifact when measuring species with long vessels.  Some techniques are robust to vessel length effects, even when there are open vessels in a measured sample.  In one technique, centrifuge curves matched previously published dehydration curves in 9 out of 10 species, including some with very long vessels.  Other techniques that involve the measurement of hydraulic conductivity while stems are being spun (spin techniques) were more sensitive to vessel size artifacts.  For instance, when a long-vesselled and ring porous Fraxinus species was measured, a spin technique produced an erroneous vulnerability curve whereas a different centrifuge technique displayed close agreement to curves generated using other methods.  This analysis suggests that, in some cases, spin techniques may not be appropriate for long vesselled species; however, such cases may be rare because most tree and shrub species have predominantly short vessels (mean ± se; shrubs 4.3 cm ± 0.3 and trees 8.3 cm ± 0.1).  Our results cast doubt on the existence of a long vessel artifact in centrifuge techniques that do not measure stems while they are spinning.