4.2 Effects of lateral fluxes
Previous theoretical and empirical studies have documented a synchronous radial movement of xylem and bark (e.g., Sevanto et al. 2011; Steppe et al. 2006, 2015; see also Scenario 1 in Fig. 1). This dynamic was interpreted as the result of diel rhythms of transpiration and associated changes in stem water potential that led to discharging and recharging of capacitive water stores. Field observations of asynchronous movement of the two tissues prompted the necessity for a refinement of this interpretation (Vandengehuchte et al. 2014; Zweifel et al. 2014; Pfautsch et al. 2015b; Mencucciniet al. 2017). In response, we have developed an empirically based method to separate most of the co-occurring fluxes in xylem and phloem from D G (Mencuccini et al. 2013, 2017). Here we progressed the model fromD Gempir toD G+ to fully account for viscoelastic movements in bark and xylem at high temporal resolution that have their origin in changes of osmotic and hydraulic pressure in both tissue types.
High conditional and marginal coefficients of determination of our mixed effects model indicated good similarity between observed and modelled data outputs. The patterns, amplitudes and slightly offset sequence of oscillations in thickness of bark (dD b) and xylem (dD x) reflect our current understanding of the effects of changes in xylem pressure potential (ΨP) and osmotic potential (ΨΠ) during the diel course of transpiration very well (e.g., Sevanto et al. 2003, 2011). Moreover, the changes in physiological functions after girdling, including the reduction of transpiration and shifts in the availability and use of NSC are all captured in our diel courses of dD b and dD x. It becomes clear that previous assessments of radial expansion and contraction of tree stems may have over- or under-estimated the magnitude of some effects, as they can act as a counterforce.
One example is the expansion of bark tissue during the night as consequence of capacitive recharge when pressure potential in xylem is close to zero (increasing D bΨPxat stable D xΨPx). The amplitude of this change in bark thickness prior to girdling was around +60 µm at the base and top of the stems. Simultaneously, and driven by the dilution effect of water flowing back into the bark, tissue contraction of around -20 µm can be expected during the early hours of the night (decreasing D bΨΠb at increasingD xΨΠx). These effects are in agreement with theoretical models that describe diurnal changes in turgor and osmotic potential in phloem and xylem (Hölttä et al.2006, 2017). Measurements of dD b alone that do not account for these changes would simply indicate a dimensional change of +40 µm. When assessing these effects during the post-girdling phase, it is clear that capacitive recharge and associated expansion ofD b is reduced below and above the girdle, reflecting the reduction in negative pressure exerted by the xylem in response to reduced rates of transpiration. The altered NSC dynamics result in a continuous contraction of D b during the night below the girdle in response to the decline in NSC. Yet, at the same time D b above the girdle continuously expands, potentially in response to accumulation of NSC in bark tissue that are not used to fuel D G+ (irreversible radial growth remains low during the night in the post-gridling phase, see Fig. 4f). Other similarly complex examples of counterforce effects can be deducted from our Figures 6 and 7.
The rhythmic pattern of high and low growth rates at the top of trees during the day and low growth during the night at the top of trees is not unusual. Using gene expression techniques, Solomon et al.(2010) provided evidence that stem growth in eucalypts is not necessarily a linear process limited to night-time hours. However, documenting a simultaneous occurrence of continuous growth at the stem base and a rhythmic growth pattern at the top of trees provides an important step forward in our understanding of the dynamic relationships between sources and sinks of NSC.