Die or survive from drought? The role of cambium integrity in saplings resilience

Hydraulic failure has been causally linked to cellular damage at the leaf level, yet the functional pathway connecting xylem dysfunction to whole-tree mortality remains unresolved. One recent hypothesis is that tree death ultimately depends on the loss of meristem vitality. However, because meristems are difficult to access and assess, their response to drought has rarely been investigated. Here, we aimed to identify whether reduced water supply resulting from hydraulic dysfunction directly compromises cambial integrity and determines tree survival. We subjected potted saplings of a gymnosperm species, Abies concolor, and an angiosperm species, Fagus sylvatica L. to severe drought by withholding water, generating different levels of loss of hydraulic functioning ranging from 30% to complete loss of conductivity (PLC 100). Prior to rewatering, water potential, percentage of embolism, relative water content, and level of cellular damage were quantified at the stem level, while cambial cell integrity was assessed using transmission electron microscopy. Sapling survival was monitored for one year following drought release.

Surprisingly, saplings of both species could display similar water potentials, relative water content status or level of hydraulic failure at the time of rewatering but exhibited different survival capacities. Saplings that maintained structurally intact cambial cells recovered, whereas those showing cambial damage died, regardless of their level of hydraulic status. Thus, our results provide direct evidence that cambium integrity represents a critical bottleneck linking hydraulic failure to tree mortality. They also evinced that the mechanisms behind loss of cambial cell integrity are mainly explained by the consequences of tree dehydration after hydraulic failure. Focusing on the water relocation towards cambial cells during a drought event could help understand the mechanisms associated with cambial cell death, and identify potential thresholds for improving the precisions of the mechanistic models aiming at predicting tree mortality.