Mechanisms of FWU
The presented results unequivocally show that open stomata contribute to FWU and highlight the role of the cuticle in the process. In both P. dulcis and P. communis , (‘fully’) open stomata by FC treatment resulted in increased water absorption rates across the entire Ψ range evaluated relative to a ‘cuticular only’ path (stomata ‘closed’) and to partially open stomata (Figure 3). Indeed, open stomata reduced the time for full Ψ recovery via FWU by up to half and enabled up to 3-4 times higher fluxes and hydraulic conductance (i.e. Qmax and Kmax, Table 1). Interestingly, the hydraulic resistance (Rmin) of P. dulcis leaves with open stomata, was the same as that estimated for the mesophyll or high-resistance leaf compartment when water was supplied via the petiole (260 MPa m2 s g-1; Guzmán-Delgado et al. 2018; Zwieniecki et al. 2007). It must be, however, mentioned that in that study leaves were younger (collected in May) than in the present study (collected in September), leading to differences in P. dulcis FWU parameters (e.g. Rmin of intact leaves was ~660 vs ~1,000 MPa m2s g-1 in May and September, respectively). Such a discrepancy might indicate that FWU is influenced by temporal changes in leaf surface properties, as reported by Cavallaro et al. (2020). Even if open stomata facilitate FWU, the rates are still orders of magnitude lower than the hydraulic suction via the petiole, which suggests that a simple flow of liquid water into the leaf via stomata does not occur.
To further asses the principles of water movement into a leaf from its wet surface across open stomata, we can first calculate the flux if liquid bridges covering the pore volume were formed. For simplicity we assume that open stomata are cylinders of radius equal to half of the maximum pore width recorded here (r = 1.27 µm for P. dulcis and r = 1.04 µm for P. communis ) and length equal to the average pore depth (l = 6.73 µm for P. dulcis and l = 10.00 µm for P. communis ). Considering a pressure difference equal to the Ψ at Qmax (ΔΨ = 1.11 MPa for P. dulcis and ΔΨ = 1.27 MPa for P. communis ) and using measured stomatal densities (ρ = 211.5 stomata mm-2 for P. dulcis and ρ = 179.8 stomata mm-2 for P. communis ), we can estimate the flux of liquid water (Ql) through the porous epidermis using the Hagen–Poiseuille equation (Ql = [π r4 ΔΨ / 8 ηw l] ρ; where ηw is water dynamic viscosity). The estimated liquid flux would be 4.0x104 and 1.2x104 g m-2 s-1 for P. dulcis andP. communis , respectively. Even with the highly conservative assumption of path resistance equal to half of the leaf thickness l≈150 µm, the respective flow rate would be 1.8x103 and 7.7x102 g m-2 s-1, yet the observed flux was in the range of ~2.5x10-3 g m-2s-1, i.e. about six orders of magnitude lower, thus precluding the possibility of flood type water penetration via open stomata. Hence, we have to consider another, more plausible scenario, where hydrophobic ledges over guard cells (Figure 4a, b) prevent the formation of liquid bridges through stomatal pores, but still allow for the formation of a water film over the pore, as intuited from ESEM analyses. In such a scenario, we can raise a question as to whether vapor diffusion across open stomata can account for the observed increase in flux over that of cuticular uptake only. Assuming the same parametrization (stomatal density, ρ, pore radius, r, and pore length, l) as for liquid flow, and using only the epidermis thickness as the barrier for vapor flux we can estimate the vapor flux (Qv) through stomata using the general diffusion equation based on Fick’s laws (Qv = [D ΔC / l] π r2 ρ, where D is the diffusion coefficient of water vapor in air, and ΔC is the concentration gradient between saturated air at the droplet surface and air at ~99% humidity in the leaf air space). By adding the contribution of the cuticle (Qmax for ABA leaves) to the calculated vapor flux through stomata, we obtain fluxes of 2.02x10-3 and 1.01x10-3 g m-2s-1 for P. dulcis and P. communis , respectively, which are within the same order of magnitude as those recorded (3.35x10-3 and 1.97x10-3 g m-2 s-1). If the path length is increased to 150 µm, the respective vapor fluxes are still 1.15x10-3 and 6.83x10-4 g m-2 s-1 (65% difference). This simple, conservative analysis provides strong support for the notion that, in the presented case, the main mechanism of water entry though stomata is vapor diffusion. There is, however, space for a marginal amount of water entering the leaf as submicron suspended water droplets (some fog droplets; Eichert et al. 2008), and as thin water films created along the pore walls of few stomata (Burkhardt et al. 2012; Eichert and Burkhardt 2001; Eichert et al. 2008). Additionally, even if epidermal surfaces are not wet, dehydrated organs can rehydrate from near saturated air (close to 100% RH, and higher than air RH of the stressed tissue; ΔC>0) at the surface, as previously predicted (Binks et al. 2020; Vesala et al. 2017) and experimentally found (Guzmán-Delgado et al. 2017; Laur and Hacke 2014).
Using chemical treatments on plants may result in equivocal responses of the physical processes under study if affected by the biological activity influenced by the treatment. Both FC and ABA were shown to modify membrane permeability, sometimes affecting leaf hydraulic properties (Blatt and Clint 1989; Coupel-Ledru et al. 2017). While this effect could not be avoided, our trial evaluating FWU via the adaxial surface of FC, ABA and control leaves of P. dulcis suggests that cuticular uptake was not affected by the treatments, so that the effect of open stomata leading to increased FWU is related to a new water pathway and not changes in membrane permeability. The lack of a significant effect on FWU by chemically induced variations in membrane hydraulic properties also suggests that FWU is dominated by the resistance related to the physico-chemical properties of the cuticle and by vapor diffusion, but not membrane facilitated liquid fluxes. Thus, the epidermal surface acts as a high-resistance pathway controlling the kinetics of leaf rehydration via FWU (Fuenzalida et al. 2019; Guzmán-Delgado et al. 2018). Most importantly, the pattern of leaf rehydration is independent from the absorption pathway, whether it is the cuticle or the cuticle with open stomata, suggesting a similar internal path for water redistribution.
Foliar water uptake (Kmax) increased significantly with increasing gs (R2=0.9999, p< 0.001, Figure 1c Supporting Information), but the slope was significantly greater in P. communis than in P. dulcis . The absence of a common relationship across species was also found by Limm et al. (2009), and may indicate that species-specific stomatal features, leaf rheology and tortuosity of the mesophyll path when stomata are open can exert further influence on the absorption process. Indeed, while both species show traits that could be associated with greater protection against excess water entry (and dehydration), such as outer and inner cuticular ledges and substomatal cuticles, P. communis – the species with lower FWU capacity (lower Kmax) – has more prominent inner ledges and a cuticle that extends farther into the substomatal cavity, even lining parenchymatic cells (Figure 4). Furthermore, changes in the properties of the chemically heterogeneous cuticle and other cell wall regions underneath can affect the species’ capacity for FWU through stomata, and when the cuticle is the dominant path (Boanares et al. 2018; Guzmán et al. 2014a,b; Guzmán-Delgado et al. 2018). These changes may be driven by the stages of cuticle hydration, starting with a partial wetting of the cuticle from the outer towards the inner side, followed by epidermal cell rehydration and then rewetting of the cuticle from its inner side (towards the leaf surface), shortening the diffusion pathways and creating liquid bridges (Guzmán-Delgado et al. 2018).