Vessel diameter and vulnerability to drought-induced embolism: within-tissue and across-species patterns and the issue of survivorship bias

Abstract

Summary Sherwin Carlquist was aware that wood represents a compromise between transport efficiency and safety. He considered wide vessels to be more vulnerable to drought-induced embolism than narrow conduits. He proposed that narrow vessels might provide redundancy and conductive safety. Research on vessel diameter and safety emerged in the 1970s and has dramatically grown in recent years showing the transformative impact these concepts have had on current research. We show how our experimental work largely supports Carlquist’s assumptions on the vulnerability of wide vessels. Although there is great variability in the vulnerability of narrow vessels, the data show that wide vessels are always vulnerable. We offer explanations for why narrow vessels differ in their vulnerability, and these explanations are rooted in Carlquist’s observations on the woody Californian flora. We recognize that there is disagreement among researchers on the correlation between vessel diameter and vulnerability to drought-induced embolism. In relation to this, we discuss the issue of flushing, i.e., whether embolism should be removed prior to generating a vulnerability to cavitation curve. Constructing a vulnerability curve from native samples that contain a substantial fraction of embolized vessels is an example of ‘survivorship bias’ because only those vessels that remained functional through a past selection process are included. Survivorship bias can lead to false conclusions on the relationship between vessel diameter and embolism resistance. This is especially the case because anatomical measures rarely incorporate functional information on the selection of measured vessels, leading to a mismatch between vessels evaluated for function and anatomy.