1. Introduction
Vertical transport of non-structural carbohydrates (NSC) in phloem and water in xylem of tree stems have been the subject of countless investigations. In contrast, radial transport processes remain understudied, and this limits our understanding of day-to-day growth dynamics in trees (Cuny et al. 2015; De Swaef et al.2015). The prevailing view is that trees accumulate NSC during the day in leaves and use them to support stem growth during the night when tissues are highly turgid (e.g., Hölttä et al. 2010; Steppeet al. 2015). This view is supported, at least partly, by time-series data documenting maximum stem expansion during the night in several tree species of the temperate zone of the northern hemisphere. However, examples from other biomes have now shown that the origin of radial movement in tree stems can be far more complex, involving irreversible (i.e., radial growth) and reversible (i.e., radial expansion-contraction) processes that can occur independently or simultaneously, during the day and/or night (e.g., Pfautsch et al. 2015a; Mencuccini et al. 2017).
At the cellular level, irreversible radial growth (D G) results from periclinal division of mother cells located inside (sapwood) and outside (phloem) a meristematic region of the secondary cambium (Schrader et al. 2004). A combination of exogenous and endogenous factors triggers the formation of primary cell walls. Subsequently, individual layers of new cell walls undergo biosynthetic processes that lead to densification, dehydration and lignification, which conclude the process ofD G. Not only do sapwood mother cells differentiate more than phloem cells – typically at a ratio of 4:1 to 10:1 in trees (Fromm 2013) – they are also permanently retained as wood, whereas phloem cells are eventually shed. It is widely accepted that NSC are the primary carbon source supportingD G in trees (e.g., Hartmann and Trumbore 2016; Martínez-Vilalta et al. 2016).
In contrast to D G, reversible stem expansion and contraction is the result of differences in hydraulic pressure and osmotic potential between xylem and capacitive water stores in sapwood and bark (note that bark in this context is the living section of bark that contains both phloem and cambium). The resulting bi-directional transport of water and osmotic substances is mostly organized through transport in rays that provide a functional link between bark and xylem (e.g., Sevanto et al. 2011; Pfautsch et al. 2015b; Pfautsch 2016). Contraction of stems during daytime and their expansion during night-time, similar to a sine-wave curve (Scenario 1 in Fig. 1), partly results from radial transport processes. It is widely accepted that this fluctuation in radial direction is primarily a result of mobilising and replenishing stored water in xylem, sapwood and bark, following shifts in vertical gradients of hydraulic pressure (e.g., Zweifel et al. 2001; Steppe et al. 2006; Hölttä et al. 2009; Mencuccini et al. 2013). Contraction and expansion of sapwood and bark is commonly observed to occur in synchrony, although movement of bark may lag slightly behind that of sapwood as a consequence of the indirect coupling through rays with the transpiration stream inside the xylem (e.g., Sevanto et al. 2003, 2011).
Studies of stem diameter changes in Eucalyptus species have revealed different, asynchronous patterns. Sapwood of E. globuluscontracted and bark expanded during the day (Scenario 2 in Fig. 1), which indicated asynchronous patterns driven by strong hydraulic decoupling between xylem and bark (Zweifel et al. 2014). Solomonet al. (2010) observed a temporal separation of processes related to biosynthesis and lignification of new cell wall materials in E. globulus , as well as temporal variation in regulation of genes that encode aquaporins, which facilitate influx of water in expanding, growing cells. These findings provide additional evidence thatD G and reversible expansion and contraction of stems are complex processes that can act at different time scales. An additional asynchronous pattern was observed in stems of E. saligna and E. tereticornis (Scenario 3 in Fig. 1), where sapwood expanded, and bark contracted during the day (Pfautsch et al. 2015a; Mencuccini et al. 2017). The two asynchronous patterns do not match widely accepted plant hydraulic modelling of a contraction of sapwood during morning hours as the result of increasingly negative hydraulic pressure in xylem (e.g., Steppe et al. 2015). A mechanistic basis for asynchronous radial expansion and contraction of stem tissues has not yet been fully established.
It is technically challenging to simultaneously document dynamic changes in multiple sinks to determine sink strength and preferential use of available carbon resources. Systematic isolation of potential carbon sinks from supply of NSC by using the phloem girdling technique can circumvent this issue. Phloem girdling has been used successfully to investigate carbon dynamics of trees (e.g., Hogberg et al. 2001; Binkley et al. 2006; Appel et al. 2012), and regulation of photosynthesis by turnover of sugars and starch in leaves (Nebaueret al. 2011). Importantly, measurements of girdling effects on stems, leaves, NSC metabolism and transpiration markedly increased our understanding of the interrelatedness of water and carbon economies in plants (e.g., De Schepper et al. 2011; Sellin et al. 2013; López et al. 2015).
Assessment of girdling effects on short-term dynamics of radial growth in stems is complex because of the simultaneous appearance of cambial cell development and reversible expansion and contraction of stems caused by hydraulic and osmotic effects. For example, with the onset of transpiration during morning hours, water stored in bark and sapwood can be drawn into xylem, where it mitigates increasingly negative water potentials in conducting vessels. The loss of capacitive water in the bark or sapwood will result in contraction of the affected tissues (e.g., Zweifel et al. 2001; Pfautsch et al. 2015a). Similarly, osmotically-induced radial fluxes of water can lead to expansion or contraction of bark. Thus, when recording the radial movement of stems at high temporal resolution, several signals are recorded concurrently (Chan et al. 2016).
Here, we use phloem girdling to manipulate source-sink relationships inE. tereticornis trees and employ point dendrometers with high temporal and dimensional resolution to document changes inD G and reversable expansion and contraction of bark and sapwood. However, deciphering the sequence of processes that result in stem growth cannot be achieved by observing effects of girdling alone because of the fundamental links between the carbon and water economies inside trees. It is necessary to also quantify the effects of girdling on carbon assimilation and related changes in the use of NSC, and loss of carbon (C) through respiration (R ), but also any effects on transpiration. Girdling has been shown to increase respiratory C loss from tree stems by 45% (Yang et al. 2019) and markedly reduce transpiration (Oberhuber et al. 2017). We developed a mechanistic model that can disentangle most of these underlying processes, including radial transport of water between bark and xylem. The present work refines previous models (Mencuccini et al. 2013, 2017) by including a term for the lateral flux of solutes and the associated changes in osmotic potential between bark and xylem that drive radial viscoelastic changes in tree stems. We hypothesise that daytime increases in osmotic potential in rays lead to sapwood expansion which significantly counter-balances tension-driven contraction in xylem and fibres. The observed high rates of leakage and uptake of soluble sugars by phloem (Epron et al. 2016; Furze et al. 2018), and the pronounced radial transport of water in the symplast of rays (Pfautsch et al. 2015b), support this concept. Use of a wide range of experimental data and output from the improved mechanistic model allowed us to generate empirical evidence of “pathway effects” (sensu Sellier & Mammeri 2019) of phloem loading, its impact onD G at different positions along the vertical stem axis and underlying asynchronous patterns of reversible expansion and contraction of bark and sapwood tissues.