Module 1

Water Transport & Xylem Hydraulics

Cohesion-tension theory, Hagen-Poiseuille flow, cavitation, stomatal biophysics, and transpiration

1.1 Cohesion–Tension Theory

In 1894, Henry Dixon and John Joly proposed that water rises in tall trees by being pulled up under tension (negative hydrostatic pressure), driven by evaporation at leaf surfaces. This requires two physical properties of water: the enormous cohesion between water molecules (hydrogen bonding), and their adhesion to cellulose walls.

Water Potential and Its Components

The total water potential \(\Psi_w\) (J m\(^{-3}\) = Pa) is the chemical potential of water referenced to pure free water at ambient pressure and temperature:

\[ \Psi_w = \Psi_s + \Psi_p + \Psi_g + \Psi_m \]

Osmotic (solute)

\(\Psi_s = -iCRT\)

Van't Hoff relation; i = osmotic coefficient, C = molarity. Negative: solutes lower water potential.

Pressure

\(\Psi_p = P - P_{atm}\)

Gauge hydrostatic pressure. Positive in turgid cells; can be strongly negative in xylem (–1 to –4 MPa in tall trees).

Gravitational

\(\Psi_g = \rho_w g h\)

Equivalent to 0.01 MPa per meter height. At 50 m, gravity alone requires −0.5 MPa.

Matric

\(\Psi_m\)

Capillary and adsorptive interactions with cell walls, relevant in unsaturated soils and wood.

For a leaf at the top of a 50 m tree with \(\Psi_{soil} = -0.3\) MPa, a transpiring leaf maintains \(\Psi_{leaf} \approx -2.5\) MPa. The gradient:

\[ \frac{d\Psi_w}{dz} = -0.01 \,\text{MPa m}^{-1} - \frac{r_h \cdot E}{A_{xylem}} \]

where \(r_h\) is hydraulic resistance per unit length, \(E\) is transpiration rate, and\(A_{xylem}\) is xylem cross-sectional area.

1.2 Hagen–Poiseuille Flow Derived from Navier–Stokes

Consider steady, incompressible, laminar flow in a cylindrical conduit of radius \(r\) and length \(l\). The Navier–Stokes equation in cylindrical coordinates (axial symmetry, \(v_r = v_\theta = 0\),\(v_z = v_z(r)\) only):

\[ 0 = -\frac{dP}{dz} + \eta \frac{1}{r} \frac{d}{dr}\!\left(r \frac{dv_z}{dr}\right) \]

With \(-dP/dz = \Delta P / l\) (constant), integrate twice:

\[ v_z(r) = \frac{\Delta P}{4\eta l}(r^2 - R^2) \]

where the no-slip BC \(v_z(R) = 0\) fixes the integration constant. This is a parabolic (Poiseuille) profile. The volumetric flow rate is obtained by integrating over the cross-section:

\[ Q = \int_0^R v_z(r) \cdot 2\pi r \, dr = \frac{\pi R^4 \Delta P}{8 \eta l} \]

This is the Hagen–Poiseuille law. The \(R^4\) dependence is crucial for tree hydraulics: doubling vessel radius increases conductance 16-fold. Angiosperms (flowering trees) have wide vessels (\(R \approx 25\)–\(200\,\mu\)m) providing high conductance, while gymnosperms (conifers) have only narrow tracheids (\(R \approx 5\)–\(20\,\mu\)m).

The hydraulic conductivity of a single conduit:

\[ k_h = \frac{\pi R^4}{8\eta} \qquad \Rightarrow \qquad Q = k_h \frac{\Delta P}{l} \]

For a xylem vessel network with \(N\) parallel conduits of different radii, the total conductance is\(K = \sum_i \pi R_i^4 / (8\eta l)\). The community metrichydraulically weighted mean vessel diameter is \(\bar{D}_h = (\sum D_i^4 / N)^{1/4}\).

1.3 Cavitation & Embolism Physics

Xylem water under tension is metastable—thermodynamically it should cavitate (boil), but kinetic barriers prevent nucleation. The free energy to create a spherical cavitation bubble of radius \(r_b\) in water at tension \(|P|\):

\[ \Delta G = 4\pi r_b^2 \gamma - \frac{4}{3}\pi r_b^3 |P| \]

where \(\gamma \approx 72\) mN m\(^{-1}\) is the surface tension of water. The critical radius at the energy barrier is \(r_c = 2\gamma/|P|\). For \(|P| = 2\) MPa,\(r_c \approx 72\) nm—below this, bubbles collapse; above, they grow explosively.

Air-Seeding at Pit Membranes

In practice, cavitation in trees occurs by air-seeding through inter-vessel pit membranes rather than homogeneous nucleation. The pit membrane has pores of radius\(r_{pore}\); air enters when:

\[ |P_{xylem}| > \frac{4\gamma}{d_{pore}} \]

For a pore diameter \(d_{pore} = 100\) nm, the threshold is \(\approx 2.88\) MPa. Pore size distributions in the pit membrane determine the vulnerability curve(loss of conductivity vs. xylem pressure). The Weibull function provides an excellent fit:

\[ \mathrm{PLC}(P) = 100 \left[1 - \exp\!\left(-\left(\frac{P}{b}\right)^c\right)\right] \]

where PLC is the percentage loss of conductivity, \(P_{50}\) is the pressure at 50% loss (fitted from \(b\) and \(c\)), and acoustic emission detectors record the ultrasonic pops produced when bubbles form.

Xylem Conduit Architecture

Vulnerability Curve

1.4 Stomatal Biophysics

Stomata control the trade-off between CO\(_2\) uptake and water loss. Guard cell turgor is regulated by ion fluxes:

H\(^+\)-ATPase (AHA2): pumps protons out, hyperpolarising the membrane to \(\approx -180\) mV, driving K\(^+\) influx via KAT1 channels.
KAT1: voltage-gated inward K\(^+\) channel; opens below\(-120\) mV; Kd \(\approx 2\) mM K\(^+\). Each channel has 4 subunits with S4 voltage sensor.
ABA signalling: ABA binds PYR/PYL receptors, inhibits PP2C phosphatases, activating SnRK2 kinases that phosphorylate SLAC1 (Cl\(^-\) channel) for stomatal closure.

Guard cell volume change \(\Delta V\) drives aperture change. The turgor-aperture relationship is approximately linear: \(a \approx \alpha P_{gc}\). Stomatal conductance to water vapour:

\[ g_{sw} = \frac{D_w}{d + a_{min}/2} \cdot \frac{\pi a^2}{4 l_{gc}} \]

Penman–Monteith Transpiration

The Penman–Monteith equation derives latent heat flux from energy balance and vapour transport. Starting from the surface energy balance:

\[ R_n - G = \lambda E + H \]

where \(R_n\) is net radiation, \(G\) is soil heat flux, \(\lambda E\) is latent heat (evapotranspiration) and \(H\) is sensible heat. Sensible heat flux via aerodynamic resistance:\(H = \rho_a c_p (T_s - T_a) / r_a\). Latent heat via total resistance:\(\lambda E = \rho_a c_p (e_s - e_a) / (\gamma (r_a + r_s))\). Eliminating surface temperature:

\[ \lambda E = \frac{\Delta(R_n - G) + \rho_a c_p (e_s^*(T_a) - e_a)/r_a}{\Delta + \gamma(1 + r_s/r_a)} \]

where \(\Delta = de_s^*/dT\) is the slope of the saturation vapour pressure curve, \(\gamma = c_p P / (\lambda \epsilon)\)is the psychrometric constant, \(r_a\) is aerodynamic resistance, and \(r_s = 1/g_{sw}\) is surface (stomatal) resistance.

1.5 Lockhart Growth Equation

Turgor-driven cell expansion is described by the Lockhart (1965) equation. For a cell of volume \(V\), irreversible expansion occurs only when turgor pressure \(\Psi_p\) exceeds a yield threshold \(Y\):

\[ \frac{1}{V}\frac{dV}{dt} = \phi(\Psi_p - Y) \quad \text{for } \Psi_p > Y \]

where \(\phi\) (MPa\(^{-1}\) s\(^{-1}\)) is the cell wall extensibility. The growth rate is limited by both wall extensibility and water uptake capacity \(L_p\). Combining both:

\[ \frac{1}{V}\frac{dV}{dt} = \frac{\phi L_p A/V}{\phi + L_p A/V} (\Psi_w^{ext} - \Psi_s - Y) \]

The Lockhart equation forms the basis of all modern plant growth models and has been extended to include cell wall loosening by expansin proteins (pH-dependent), yielding the Ortega extension.

1.6 Python: Water Potential Profile & Vulnerability Curve

We model the water potential gradient from soil to leaf in a 50 m tree, accounting for gravity, frictional resistance, and transpiration rate. Then we fit a Weibull vulnerability curve.

Python

script.py84 lines

import numpy as np
import matplotlib
matplotlib.use('Agg')
import matplotlib.pyplot as plt

# ---- Water Potential Profile ----
H = 50.0        # tree height (m)
z = np.linspace(0, H, 200)
rho_w = 1000.0  # kg/m^3
g = 9.81        # m/s^2
Mpa = 1e6       # Pa per MPa

Psi_soil = -0.3  # MPa
# Resistivity: hydraulic resistance per unit length (MPa s m^-1 per m = MPa/m)
r_h = 0.04      # MPa m^-1 per unit transpiration
E = 1.0         # relative transpiration rate (dimensionless, scaled)

# Gravity component
Psi_gravity = -(rho_w * g * z) / Mpa
# Friction component (increases with height)
Psi_friction = -r_h * E * z
# Total water potential
Psi_total = Psi_soil + Psi_gravity + Psi_friction

# ---- Vulnerability Curve (Weibull) ----
P_arr = np.linspace(0, -8, 300)  # xylem pressure in MPa (negative)
b = 3.5    # scale parameter
c = 3.0    # shape parameter
PLC = 100.0 * (1.0 - np.exp(-(-P_arr / b)**c))

# P50 calculation
P50 = -b * (np.log(2.0))**(1.0/c)

# ---- Plotting ----
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(14, 6))
fig.patch.set_facecolor('#0a0f0a')
for ax in (ax1, ax2):
    ax.set_facecolor('#0a0f0a')
    for sp in ax.spines.values():
        sp.set_color('#34d399')
    ax.tick_params(colors='#a7f3d0')
    ax.xaxis.label.set_color('#a7f3d0')
    ax.yaxis.label.set_color('#a7f3d0')
    ax.title.set_color('#34d399')

ax1.plot(Psi_total, z, color='#34d399', lw=2.5, label='Total psi_w')
ax1.plot([Psi_soil]*200, z, color='#fbbf24', lw=1.5, linestyle='--', label='Soil psi_w')
ax1.axhline(0, color='#6b7280', lw=0.8)
ax1.axvline(Psi_soil, color='#6b7280', lw=0.8)
ax1.fill_betweenx(z, Psi_total, Psi_soil, alpha=0.15, color='#34d399')
ax1.set_xlabel("Water potential (MPa)")
ax1.set_ylabel("Height (m)")
ax1.set_title("Water Potential Profile: Soil to Leaf")
ax1.legend(facecolor="#0a0f0a", labelcolor="#a7f3d0", edgecolor="#34d399")
psi_leaf = Psi_total[-1]
ax1.annotate(f'Leaf: {psi_leaf:.2f} MPa',
    xy=(psi_leaf, H), xytext=(psi_leaf + 0.3, H - 5),
    color='#34d399', fontsize=9,
    arrowprops=dict(arrowstyle='->', color='#34d399'))

ax2.plot(P_arr, PLC, color='#34d399', lw=2.5)
ax2.axhline(50, color='#fbbf24', lw=1.5, linestyle='--')
ax2.axvline(P50, color='#fbbf24', lw=1.5, linestyle='--')
ax2.axhline(88, color='#f87171', lw=1.2, linestyle=':')
ax2.set_xlabel("Xylem pressure (MPa)")
ax2.set_ylabel("Percent loss conductivity (%)")
ax2.set_title("Xylem Vulnerability Curve (Weibull)")
ax2.annotate(f'P50 = {P50:.2f} MPa',
    xy=(P50, 50), xytext=(P50 - 1.5, 60),
    color='#fbbf24', fontsize=10,
    arrowprops=dict(arrowstyle='->', color='#fbbf24'))

plt.tight_layout()
plt.savefig("output.png", dpi=120, facecolor="#0a0f0a")

print(f"Leaf water potential at top of {H:.0f} m tree: {psi_leaf:.3f} MPa")
print(f"  Gravity component at top: {Psi_gravity[-1]:.3f} MPa")
print(f"  Friction component at top: {Psi_friction[-1]:.3f} MPa")
print(f"Vulnerability curve parameters: b={b}, c={c}")
print(f"  P50 = {P50:.3f} MPa")
P88 = -b * (np.log(1.0/0.12))**(1.0/c)
print(f"  P88 = {P88:.3f} MPa")
print(f"  Hydraulic safety margin (P50 - psi_leaf) = {P50 - psi_leaf:.3f} MPa")

Click Run to execute the Python code

Code will be executed with Python 3 on the server

← Foundations Next: Photosynthesis →