2.6 Respiration
We assessed girdling effects on respiratory physiology by measuring leaf, branch wood, and whole-crown respiration three days before (13 May) and six days after the trees were girdled (22 May). Two branches were selected at random for each chamber on each date, and branches were detached in the evening by cutting at the stem insertion point. All leaves were removed prior to cutting each branch in to 10 cm long segments. All segments of a single branch were placed into a large gas exchange chamber (3010-GWK1, Heinz Walz GmbH, Effeltrich, Germany) connected to an infrared gas analyser (IRGA; LI 6400-XT, LiCor, Lincoln, NE, USA). The chamber temperature was controlled such that the tissue temperature was 15 °C (±0.5). The CO2 concentration of the reference cell was 400 ppm and the flow rate was 700 µmol s-1. The flow rate was selected to cope with high amounts of water vapour released from stem segments. We recognise that sampling may have affected respiration rates or released CO2 dissolved in the xylem. However, measuring respiration of detached plant organs is a standard approach (e.g., Poorter et al. 1990; Tjoelker et al. 1999, 2005; Comas & Eissenstat 2002; Drake et al. 2016, 2017), and we expect any sampling effects to be systematic across dates and not affect our ability to detect girdling effects.
The leaves from each branch were mixed and three leaves were selected at random; the respiration rate of these three leaves were measured as a composite sample using a large, opaque leaf cuvette (Li-6400-XT with Li-6400-22 conifer chambers; Licor). The block temperature was controlled such that the leaf temperature was 15 °C (±0.5). The CO2 concentration of the reference cell was 400 ppm and the flow rate was either 350 or 500 µmol s-1, depending on moisture levels. This flow rate was lower compared to that used for stem segments, as less water vapour was released by the leaves. Leaves and branches were dried and weighed, and respiration rates were expressed per unit dry mass. Note that the pre-girdling leaf and branch respiration measurements were presented in Drake et al. (2016). The post-girdling data have not previously been published.
After the two branches were removed for the pre-girdling branch and leaf respiration measurement (13 May), we measured respiration of the entire aboveground biomass of each tree. For these measurements, each whole-tree chamber was manually transformed into a closed system that excluded soil and the rate of CO2 accumulation within the chamber airspace was measured every minute. The whole-tree chamber environmental controls were used to maintain the air temperature in all chambers at 15 °C. The rate of CO2 efflux from the entire crown was calculated as in Drake et al. (2016). Note that these measurements reflect the combined respiratory activity of leaves, branches, and the stem. These measurements were collected during night-time. The total mass of leaves, branches, and stem wood was measured via a full destructive harvest on 26 May 2014. All leaves, branches, and stem wood were manually separated for each tree, dried, and weighed. The total sum of leaf, branch, and stem wood dry mass was used to express the whole-tree respiration measurements per unit total aboveground mass. We assume that total tree mass did not change appreciably over the 13 days separating the first crown respiration measurement and the tree harvest.