Introduction
The oxygen isotopic composition has been used to trace water transport into and through plants such as the identification of sources water taken up by a plant (Dawson & Ehleringer 1991; Ellsworth & Sternberg 2014) and water movement through the leaf (Barbour & Farquhar 2004; Barbour et al.2017a). However, the use of oxygen isotopic composition to trace water transport is dependent on understanding the factors that influence water movement and the physical and biological processes that fractionate oxygen isotopes in water. For example, physiological factors such as transpiration (E ) and leaf anatomical structures can influence water flow within the leaf and the degree of fractionation, impacting oxygen isotopic composition in leaf water [δ18OLW](Yakir 1998; Barbour & Farquhar 2000; Barbour et al. 2000a; Barbour 2007; Cernusak et al. 2016; Holloway-Phillips et al. 2016). Because of these complex interactions of δ18OLW with leaf physiology and anatomy, it is presently difficult to draw clear conclusions from the experimental data (Barbour et al.2017a). Therefore, a better understanding of how water movement within a leaf interacts with the physical and physiological processes must first be understood to be able to fully exploit δ18OLW(Cernusak et al. 2016; Barbour et al. 2017a).
Multiple models have been developed to help explain how environment and physiological processes influence δ18OLW. For example, the two-pool (Roden & Ehleringer 1999), the string-of-lakes models (Gat & Bowser 1991), and the Péclet effect (Barbouret al. 2000b; Barbour et al.2004) have been developed to help explain how evaporatively-enriched water at the sites of evaporation and unenriched δ18Osw influence δ18OLW. In all of these models, unenriched water supplied through the xylem becomes enriched at the sites of evaporation based on a modified Craig-Gordon model, where enrichment is a function of relative humidity, leaf temperature, and oxygen isotopic composition of vapor (See Theory section; Craig & Gordon 1965; Farquhar et al. 1998). The two-pool model operates under the principles of mass balance where unenriched water in the xylem and enriched water from the sites of evaporation (δ18Oe) are considered two discrete water pools that mix to form δ18OLW. Alternatively, the string-of-lakes model used in grasses considers that instead of two discrete water pools there are multiple discrete pools in series up the leaf blade, where the source water to each pool is the water from the pool preceding it. This leads to a progressive enrichment of water from the first pool at the base of the leaf blade to the last pool towards the leaf tip. However, the two-pool and the string-of-lakes models do not address cases when δ18OLW varies with E (Farquhar & Lloyd 1993).
The Péclet effect model accounts for the continuous gradation from unenriched water moving through the xylem to enriched water at the sites of evaporation through opposing flows of advection and back-diffusion (Barbour et al. 2000a; Barbour et al. 2000b). These opposing flows in the Péclet model are dependent on E and the effective mixing length over which unenriched water from the xylem mixes with enriched water from the sites of evaporation (L ). Lcan be unique to each tissue type where these two isotopically-different water pools are mixing, such as in the mesophyll as water moves from the xylem to the inner-stomatal cavity [L m] or in the veins where enriched mesophyll water mixes with unenriched xylem water [L v] (Holloway-Phillips et al. 2016). In this manner, isotopic mixing occurs radially from the xylem out to the stomata and longitudinally in the veins where enriched water enters the veins through back diffusion and is carried toward the leaf tip (Farquhar & Gan 2003). Leaf traits such as cell wall composition that can affect the extent that mixing occurs by influencing the length and tortuosity od the path that water takes through the xylem and mesophyll (L ) and flow or conductivity (E ), which, in turn, affect δ18OLW beyond the effect that δ18OSW has. Each of these models have found empirical support under specific conditions and for certain species; however, none of them can universally explained the isotopic behavior of leaf water (Roden & Ehleringer 1999; Barbour & Farquhar 2000; Barbour et al. 2000b; Helliker & Ehleringer 2000; Barbour & Farquhar 2004; Loucos et al. 2015; Song et al. 2015; Holloway-Phillips et al. 2016).
However, the Péclet effect model is unique in that it can be used to test hypotheses on how leaf anatomical features affect the movement of leaf water and influence δ18OLW. Specifically, this model can be used to test how leaf traits influence the advection of water from the xylem to the sites of evaporation and back diffusion of water from the sites of evaporation into the mesophyll and xylem, which determines the influence of the enrichment of water at the sites of evaporation above source water (Δ18Oe) on leaf water enrichment above source water (Δ18OLW). In other words, anatomical features that potentially influence the mixing of water pools and the flow of water out of the leaf (E ) or that impact Lcan be tested on how they alter Δ18OLW. For example, greater vein density increased the oxygen isotopic enrichment along the grass leaf blade because it increased mixing of enriched mesophyll water with xylem water, leading to the progressive enrichment of xylem water and Δ18Oe up the leaf blade and subsequently Δ18OLW(Helliker & Ehleringer 2000; Farquhar & Gan 2003). In dicots, greater vein density had the opposite effect in that Δ18OLW was lower because the proportion of unenriched xylem water (f sw) was greater (Holloway-Phillips et al.2016). Stomatal density and pore size also influence Δ18OLW, where greater stomatal densities, even when E remained the same, had higher Δ18OLW because L decreased with stomatal density (Sternberg & Manganiello 2014; Larcher et al. 2015; Liang et al. 2018). These findings have led to increased interest in anatomical features such as cell wall composition that partly control the permeability of the cell wall and, in turn, influence the relative importance of apoplastic water movement from the xylem to the sites of evaporation, which ultimately affectsE and L (Barbour et al. 2017b).
Cell wall composition mutations modify how leaf anatomy influences water movement andphysiological properties that affect E , L , and, ultimately, Δ18OLW. For example, suberin in the suberin lamellae provides a barrier to apoplastic water movement and passive water loss from the vasculature (Mertz & Brutnell 2014; Vishwanath et al.2015).Zmasft double mutants, which were created in maize (Zea mays L.) by mutating two paralogously duplicated, unlinked maize orthologues of Arabidopsis thaliana ALIPHATIC SUBERIN FERULOYL TRANSFERASE , have a compromised suberin llamelae ultrastructure that increased cell wall elasticity and apoplastic water diffusion (Mertz et al. In review). Greater apoplastic diffusion and cell wall elasticity can increase available water for transpiration, while less barriers to water movement can increase the number of water channels and modify the relative importance of each channel that water takes in and out of the leaf vasculature and from the vasculature to the stomata, influencingL m and L v. Cellulose synthase-like F6 (CslF6 ) in rice (Oryza sativa L.) mediates the biosynthesis of mixed linkage glucan (MLG), a polysaccharide important in cell wall composition (Vega-Sánchez et al. 2012). Loss-of-function mutants do not synthesize MLG (Vega-Sánchez et al. 2012; Smith-Moritz et al. 2015) and have higher cell wall porosity, which has been show to affect the internal conductance of CO2 (g m) (Ellsworth et al. 2018). This increase in porosity affected the movement/conductance of CO2 and may also alter apoplastic water movement as well, which can alter water flow to the stomates (E ) or the length and tortuosity of water path (L ). Therefore, these mutants provide a means to evaluate the effect of cell wall composition on leaf physiology with respect to water transport through its impact on Δ18OLW.
To determine the influence of cell wall properties on Δ18OLW we tested how the loss of MLG and suberin changed water movement from the xylem to the sites of evaporation under environmental conditions (growth light and relative humidity) that have been demonstrated to influence how E andL impact Δ18OLW. The data presented here demonstrates that changes in both cell wall properties and stomatal density influence Δ18OLWthrough the impact of E and L on the Péclet effect.