4.1 Impact of girdling on H2O and C economy
Despite higher air temperatures and greater VPD during the post-girdling period, water use of the girdled trees declined. Reductions in transpiration have been observed in other tree girdling experiments and were attributed to feedback responses in Photosystem II (e.g., Sellinet al. 2013), photoinhibition due to high concentrations of leaf NSC (e.g., Lopez et al. 2015) and also secondary responses unrelated to photosynthesis and NSC (e.g., Asao & Ryan 2015). However, plant material in these earlier studies was kept under identical environmental conditions before and after application of the girdle. In the current study, girdled trees reduced water use under environmental conditions that should stimulate the opposite, and without any noticeable accumulation of NSC in leaves. This points towards another possible chain of events triggered by girdling.
More than 50 years ago, Kurtzman Jr. (1966) suggested that phloem girdling will lead to a reduction in conductive sapwood area as a result of vessel embolism. Since then, studies have investigated how removal of phloem tissue impacts water transport in xylem with some showing a reduction (e.g., Salleo et al. 2004; Zwieniecki et al.2004; Tombesi et al. 2014), while others did not (e.g., Domec & Pruyn 2008). However, none of these studies assessed if girdling changed the radial pattern of sap flow, which often displays a declining gradient from outer to inner sapwood. One study has directly assessed changes in this radial gradient using dye injections and showed that girdling had the largest impact on conductivity of the outer sapwood (Ueda et al. 2014). Combining the results of Ueda et al.with (a) the reduction in water use under higher VPD, (b) the absence of accumulation of NSC in leaves and (c) the strong water to air pressure gradient at the surface of exposed sapwood opens the possibility that girdling replaced water with air in rays and other capacitive stores of water in the outer sapwood. This effect would agree with Fick’s second law of diffusion of water in wood (Avramidis 2007). The connection of rays to outer xylem vessels has been documented empirically (Pfautschet al. 2015b), and anatomical analyses have shown that sapwood ofE. tereticornis can have more than 700 rays mm-2 (Treydte et al. 2021). This evidence points towards pronounced capacity of radial access to xylem vessels. Once rays are air-filled, the negative pressure in vessels would draw air bubbles into the transpiration stream leading to embolism. Experiments using a combination of stem girdling and x-ray or magnetic resonance tomography could be used to validate this interpretation. Regardless of potential effects from embolised vessels on transpiration, the gridling treatment helped separating hydraulic and osmotic processes to validate D G+ and increased our fundamental understanding of xylem-phloem interactions.
Increasing embolism in xylem vessels helps explain the observed reduction in whole tree transpiration, especially under rising VPD. Under these conditions, leaf water potentials would decline and reduce stomatal conductance. This sequence of responses to girdling is supported by the decline in whole-tree C uptake. In contrast to other studies (e.g., Johnsen et al. 2007; Maier et al. 2010; Lopez et al. 2015), NSC did not accumulate in leaves of E. tereticornis . Moreover, girdling did not lead to higher rates ofR in leaves. These results imply that girdling reduced assimilation of carbohydrates, but at least over the few days after girdling, loading of soluble sugars into the phloem continued. The lower rates of assimilation help explain the small yet significant decline of soluble sugars in leaves after girdling.
Girdling of E. tereticornis trees resulted in acceleratedD G+ above and collapse ofD G+ below the girdle. Further, girdling accelerated rates of R in stems and branches but not leaves. Similar up-regulation characteristics for stem growth and R soon after girdling have been observed in other girdling experiments (e.g., Daudet et al. 2005; Johnsen et al. 2007; López et al. 2015). These responses can be explained by changes in sink strength above and below the girdle. Accelerating D G+and R will reduce NSC concentration in the phloem, thus stimulate continued loading and prevent accumulation of NSC in leaves. This mechanism does not involve hydrolysis of starch in leaves and explains the stable concentration of starch in leaves before and after girdling. As we have shown, D G+ close to the source was greatest during the day when NSC are assimilated, producing a clear rhythmicity of D G+ at the top of trees. NSC, including limited amounts of starch from storage at the top of trees were used to fuel high D G+ and R , while at the base the absence of ‘fresh’ NSC induced a significant depletion of both stored starch and available NSC to maintain some metabolic activity but not D G+.