Soil hydraulics affect the degree of isohydricity.

Dear Editor,

In a recent inspiring paper, Knipfer et al. (2020) used the experimental relation between plant water potential at midday and predawn (and ) to derive characteristic values of water potential (WP) curves defining stomatal behavior. Based on this relation, they identified two characteristic points: the water potentials at stomatal closure and at leaf turgor loss. The relation between and is receiving increasing attention as it allows one to express the stringency of stomatal control to water potential and the degree of isohydricity, overcoming the shortcomings of the isohydric and anisohydric dichotomy (Martínez‐Vilalta et al., 2014; Charrier et al., 2018; Hochberg et al., 2018). Meinzer et al. (2016) used this relation to define new metrics of isohydricity, including a “hydroscape”, defined as the area enclosed by the and regression lines. Their hypothesis is that the Hydroscape area is broader for anisohydric plants. The analytical approach of Knipfer (2020) allows one to link stomatal regulation to other physiological thresholds such as hydraulic conductance and water content losses, as commented by Charrier (2020). In this letter, we provide an additional perspective on the and curves, which more comprehensively relates leaf water potential with transpiration rates, soil drying, and the loss of hydraulic conductance. The WP curves and the resulting hydroscapes not only reflect the stomatal behavior but also the loss of soil hydraulic conductance. Consequently, the degree of isohydricity would be shaped by the hydraulic conductance of the soil–plant continuum, and primarily by that of the soil.

We follow the hydraulic framework proposed by Carminati and Javaux (2020). The model is conceptually similar to others recently reviewed in Wang et al. (2020). Based on a simplified model (see parameters in Table 1) that includes the nonlinearity of the hydraulic resistances of soil, xylem, and root, they defined a hydraulic surface in which a demarcation line, the so-called stress onset line (SOL), separates the linear (green) and nonlinear (orange) zones (Figure 1A). The authors proposed that the SOL is primarily controlled by the loss of soil hydraulic conductivity. Figure 1, B and C illustrates the hydraulic surface and the SOL from the soil and the leaf perspectives, respectively. In Figure 1D the model is shown from the , perspective, which corresponds to the domain of the hydroscape, as defined in Meinzer et al. (2016). Indeed, if we assume no transpiration and negligible soil and plant redistribution fluxes, . Note that in dry soils, when the soil hydraulic conductivity is very low, even a small water flow (caused by night transpiration or tissue rehydration) might be sufficient to cause a deviation of from . The green domains in Figure 1, A–D represent the “safe” zone for plants in which an increase of transpiration (E) does not imply a large change of . The red line (SOL) represents the onset of hydraulic nonlinearity. Carminati and Javaux (2020) hypothesized the SOL is the upper limit to the stomatal conductance the plant can achieve. A follow-up hypothesis is that stomatal regulation is triggered by a disproportion between change in leaf water potential and transpiration, but it is not yet clear what hormonal or hydraulic signal would allow stomata to accomplish such function. This regulation takes place diurnally as the transpiration demand increases and changes over a week as the soil dries and plants grow. Similar to this concept, Sperry and Love (2015) hypothesized stomatal regulation should keep plants in the green zone to limit xylem hydraulic failure. It is striking to observe the similarity between the SOL in Figure 1D and the shape of the WP curve in Knipfer et al. (2020).

Table 1

Carminati–Javaux model parameters (for more details see, Carminati and Javaux, 2020)

Soil-plant system	Model	Parameters
Soil	k_soilh= k_sathh_o^-τ	h  0 = −10 cm, ksat = 0.001 cm s−1, τ = 2.5 (Figures 1 and 2, A)   h  0 = −10 cm, ksat= 0.01 cm s−1, τ = 3 (Figure 2B reference)
Xylem	K_xψ=K_rootψψ_0x^-τx	ψ0x = −3 MPa, τx = 5, Kroot=10−7 cm³ hPa−1 s−1
Domain		r0 = 0.05 cm, rb = 1 cm, lroot = 1,000 cm

Figure 1

Carminati–Javaux (CJ) model of the soil–plant hydraulics in the transpiration (E)-leaf water potential ()-soil water potential () domain. The SOL (in black) separates the soil–plant hydraulic surface into two zones, one linear in green and one nonlinear in orange. A, 3D view; (B) lateral view from the soil perspective; (C) lateral view from the leaf perspective and (D) top view, corresponding to the hydroscape domain. Model parameters are given in Table 1

Figure 2

The hydroscape as simulated by a soil–plant hydraulic model. A, The hydroscape domain in light green is enclosed by the SOL (continuous red line) and line (dotted black line). Iso-transpiration lines are shown as dashed lines ranging from low (blue) to high (yellow) E-levels. Two exemplary dry-down events are shown: a quick one in pink and a slower one in purple. Colored closed circles correspond to midday leaf water potential values. We assume Θ1 to occur when the midday leaf water potential reaches the SOL. B, Sensitivity analysis of the SOL and hydroscape to hydraulic properties of the soil–plant system. Dotted red: reference (see Table 1), dashed yellow: doubled soil hydraulic conductivity, continuous blue: soil hydraulic conductivity decreased by 50%. The black dashed line is the 1:1 line. When soil hydraulic conductivity is increased, soil limitation occurs at more negative

Figure 2 shows the hydraulic surface with potential transpiration isolines induced, for instance, by different vapor pressure deficit (VPD), light intensity, or canopy developments. In this conceptual representation of the soil–plant hydraulics, the green region between the 1:1 line at which and the SOL corresponds to the domain in which plant hydraulic status does not limit stomatal closure. The actual temporal trajectory within this domain is controlled by the transpiration demand. Figure 2 shows two possible temporal evolutions of , at two transpiration demands over several days: low (purple) and high VPD (pink). Rapid dry-down results in larger oscillations and more negative (Figure 2, pink curve). When the plant hydraulic status reaches the SOL, stomatal regulation is expected to maintain on the SOL. It is striking to observe the shape of the evolution (closed circles in Figure 2) corresponds to the observations by Knipfer et al. (2020; Figures 1, A and C, 2, 5).

ΘΘPhase I observed by Knipfer et al. (2020) corresponds to the conditions before the SOL—i.e. before any loss in the soil–plant hydraulic conductance occurs. In this phase, the slope of , is primarily affected by environmental conditions, such as VPD or light intensity. For constant atmospheric conditions and neglecting plant growth, the slope is expected to be 1 in this phase. Phase II starts when stomatal closure is induced by the loss in soil hydraulic conductance. In this phase the , trajectory follows the SOL. During this phase, the slope is <1 and can be even negative. As observed in Figure 2A, our theory implies that the threshold value Θ1 at which starts to be controlled by the SOL occurs at less negative values for quick than for slow dry-down conditions. However, differences in Θ1 values might be difficult to observe in real experiments as they are relatively close to each other. Phase 3 results from the leaf turgor loss, which is not accounted for in our model. Note also that this analysis neglects plant adaptations to soil drying, such as root plasticity, osmotic adjustments, and shift of turgor loss points, examples of processes that would impact the , curve.

Our hypothesis that the WP curves are shaped by the SOL has an important corollary: the WP curves and hydroscape do not depend only on plant traits and atmospheric conditions but are a function of soil hydraulic properties as well. Figure 2B illustrates a few examples of how the SOL is affected by changes in soil and plant hydraulic properties.

ΘIn summary, our concept shows , curves are a function of nonlinearity in soil–plant hydraulics, which implies their area (hydroscape), slope and threshold values (Θ1) are not a function of plant traits and atmospheric conditions only, but also of the soil hydraulic properties. Furthermore, it follows that the slope ,, which is assumed to be a proxy for the degree of isohydricity based on these metrics, is partly determined by soil properties. However, root and canopy development, which defines water supply and water evaporative demand respectively, might plastically adapt to different soils, such that the SOL and the , curves might eventually not be affected by the soil properties. More research is needed to assess the importance of soil properties on stomatal regulation. The analytical approach of Knipfer et al. (2020) coupled with the proposed soil–plant hydraulic model allows a quantitative assessment of hydraulic limitation and of how stomatal regulation depends on environmental conditions.