Evidence of xylem hydraulic sectoring in apple trees from a deuterium tracing experiment in a split-root system

Xylem is the plant tissue devoted to water transport. Its structure and anatomy vary among tree species, ranging from integrated (i.e., well-connected) to sectored networks. While xylem hydraulic sectoring has some advantages (e.g., reduced spread of pathogens and embolism), it also limits the exchange of water and nutrients between plant organs in different locations in the tree. In agricultural settings, where water and fertilizer inputs are often localized, preferential flow pathways in xylem could lead to non-homogeneous distribution of these resources within the trees. We therefore carried out an experiment to determine the degree of sectoriality in apple tree xylem, hypothesizing that differences in water availability at root level would influence this behavior.

To test our hypothesis, we potted young apple trees in a split-root system with four independent compartments. Soil compartments in different sets of trees were irrigated with water having different isotopic composition (enriched, δ²H ≈ 1650‰, or tap, δ²H ≈ -80‰) or left dry, obtaining five different treatments (100, 50_W, 25_W, 50_D, and 25_D, where the number represents the percentage of sectors receiving labelled water, and W and D indicate whether the remaining sectors were irrigated with tap water or left dry, respectively). Four days after the labelled irrigation, we destructively sampled shoots, trunk, rootstock, roots, and soil in each pot, every time collecting four samples corresponding to the respective sectors of the split-root system. Water was subsequently extracted from the samples by cryogenic vacuum distillation and its isotopic composition determined with IRMS. A two-end-member mixing model was applied to determine the contribution of labelled soil water in each tree organ.

In the trees receiving water in all sectors (100, 50_W, and 25_W), the average fraction of labelled soil water measured in the tree was consistent with that in the soil and reflected the number of soil sectors receiving labelled water in the respective treatment (100%, 50%, or 25% of enriched soil water). Conversely, when the labeled water was applied only to one or two soil compartments (25_D and 50_D), the average fraction of enriched soil water in the trees was higher than when the other compartments received unlabeled water (25_W and 50_W), indicating a higher water uptake by the roots in the irrigated sectors. Interestingly, in all treatments except the 100, we observed a high variability in the fraction of labelled soil water among different parts of the canopy within each tree. When soil water availability was homogeneous (50_W, 25_W), at least one sector of the tree canopy showed a negligible (<10%) contribution of labelled soil water, indicating that water flow was predominantly axial. When part of the soil was dry (50_D, 25_D), lateral water movement was enhanced, evidencing that hydraulic sectoring is affected by the water availability at root level. Therefore, when trees have access to water pools with different availability and isotopic fingerprint, the isotopic composition of water could be spatially variable also within the plant. This has consequences in ecohydrological studies.