Dissimilar stem and leaf hydraulic traits suggest varying drought tolerance among co-occurring <italic>Eucalyptus grandis</italic> × <italic>E. urophylla</italic> clones.

In South Africa, there is currently a phenomenon where a <italic>Eucalyptus grandis</italic> × <italic>E. urophylla</italic> clone is dying on a large scale, whereas co-occurring clones in the same plantation area remain healthy. No primary biotic pathogen has been isolated from these dying trees. However, the plantation region has endured severe drought conditions. We hypothesised that the specific <italic>E. grandis</italic> × <italic>E. urophylla</italic> clone dieback is due in part to the affected trees having an elevated drought vulnerability. We tested this hypothesis retrospectively by examining how the dying (branch dieback started) as compared with two healthy and co-occurring <italic>E. grandis</italic> × <italic>E. urophylla</italic> clones vary in branch xylem anatomy (water transport) and leaf stable carbon isotopes (gas exchange) across two levels of mean annual precipitation in the planted landscape. There was significant intra-hybrid variation in the range of studied hydraulic traits. It appeared that the dying clone is maladapted to drier field conditions relative to the two healthy growing clones. Individuals of the drought-susceptible clone had increased hydraulic conductivity, having both wider vessel lumens and larger lumen fractions, and also did not regulate leaf-level transpiration in the drier landscape as efficiently as the two unaffected clones (lower δ¹³C ratios). In turn, one of the healthy growing clones had a completely different hydraulic strategy, having higher wood density, lower lumen fractions, and higher δ¹³C ratios that are widely considered as hydraulically resilient during water stress. The other clone appeared to also gain drought tolerance via higher leaf-level water-use efficiency, although drought tolerance was less clear regarding xylem anatomy. We highlight how variation in relative drought vulnerability can be highly significant even between clones within a single hybrid group. Knowledge of this variation in plant hydraulics between commercially planted and closely related <italic>Eucalyptus</italic> trees would complement clone deployment programmes in a target landscape. [ABSTRACT FROM AUTHOR]