Drought limits the long-distance transport of water in the xylem due to the reduced leaf-to-soil water potential difference and possible embolism-related losses of conductance and of sugars in the phloem due to the higher viscosity of the dehydrated sugary solution. This condition can have cascading effects in water and carbon (C) fluxes that may ultimately cause tree death. We hypothesize that the maintenance of xylem and phloem conductances is fundamental for survival also under reduced resource availability, when trees may produce effective and low C cost anatomical adjustments in the xylem and phloem close to the treetop where most of the hydraulic resistance is concentrated. We analyzed the treetop xylem and phloem anatomical characteristics in coexisting Scots pine trees, symptomatic and non-symptomatic of drought-induced dieback. We selected the topmost 55 cm of the main stem and selected several sampling positions at different distances from the stem apex to test for differences in the axial patterns between the two groups of trees. We measured the annual ring area, the tracheid hydraulic diameter (Dh) and cell wall thickness (CWT), the conductive phloem area and the average lumen diameter of the 20 largest phloem sieve cells (Dph). Declining trees grew less than the non-declining ones, and despite the similar axial scaling of anatomical traits, had larger Dh and lower CWT. Moreover, declining trees had wider Dph. Our results demonstrate that even under drought stress, maintenance of xylem and phloem efficiencies is of primary importance for survival, even if producing fewer larger tracheids may lead to a xylem more vulnerable to embolism formation.

This article describes a field- and laboratory-based framework that can be used to monitor the C balance in Mediterranean pine forest ecosystems under different management practices that determine their structure and function. By jointly monitoring stand structure, gas exchange, litter, and decomposition dynamics, this protocol enables the assessment of how management-driven changes regulate carbon uptake, turnover, and losses, thereby affecting carbon sequestration potential. As an example, we suggest the implementation of the protocol at ten (10) permanent monitoring plots across three study areas located in Greece. The first group of plots represents a post-fire chronosequence in pine stands with no management interventions. The second group includes pine stands that exhibit variation in overstory and understory density driven by differences in microclimate and management history. The third group consists of peri-urban pine stands subjected to thinning of varying intensity. The monitoring protocol is implemented across all plots and the collected data can be classified into three analytical domains: (a) demography, encompassing measurements of tree growth and mortality; (b) litter and decomposition dynamics, involving the quantification of litterfall and its seasonality and the estimation of its decomposition rates; and (c) gas exchange, focusing on measurements of leaf photosynthesis and respiration (including relevant leaf functional traits) and monitoring of soil respiration. These three data domains can be used to comparatively consider the effect of forest management on key ecosystem processes and to constrain local-scale vegetation dynamics models.