The cohesion-tension pull

A redwood leaf 100 m up needs water that started in the soil. How does the water get there? There is no heart and no pump in a tree. The answer, proposed by the Irish scientists Henry Dixon and John Joly in 1894, is the cohesion-tension theory: the water is pulled from the top, not pushed from the bottom.

Here is the chain of events:

• Evaporation. Inside a leaf, water evaporates from the wet walls of cells into tiny air spaces and leaves through the stomata (next lesson). The energy for this evaporation comes from the Sun, not from the tree’s own metabolism.
• Tension. As water leaves, the remaining water in the cell walls forms tiny curved surfaces (menisci) in pores that are only a fraction of a micrometre wide. Surface tension makes these curved surfaces pull the water behind them, so the water is under negative pressure, called tension.
• Cohesion and adhesion. Water molecules stick to each other (cohesion, from hydrogen bonds) and to the cellulose walls of the xylem tubes (adhesion). So the pull is passed down, through continuous unbroken columns of water, all the way to the roots.
• Uptake. Water enters at the roots to replace what was pulled out.

The numbers show why this is the only mechanism that can work for a tall tree.

Hydrostatic pressure. Holding a column of water up against gravity needs a pressure difference ΔP = ρ g h. With ρ = 1,000 kg/m³ and g = 9.81 m/s², that is about 0.098 MPa for every 10 m of height (1 MPa = 1,000,000 Pa, roughly ten times atmospheric pressure).

• 10 m of water: 0.098 MPa.
• 30 m: about 0.29 MPa.
• 100 m: about 0.98 MPa, just to balance gravity.

Water also has to flow through the narrow xylem tubes, and that friction needs a pull of about the same size again. Measurements at the top of very tall redwoods found midday leaf water potentials of roughly −1.5 to −2 MPa, which matches this picture.

Capillarity (water creeping up a thin tube) is too weak. The height is h = 2γ cosθ ÷ (ρ g r). With surface tension γ = 0.072 N/m, contact angle 0° and a xylem tube radius r = 20 µm, h = 2 × 0.072 ÷ (1,000 × 9.81 × 0.00002) ≈ 0.7 m.

Root pressure (roots actively pumping ions in, so that water follows by osmosis) is typically only about 0.1 to 0.2 MPa. That could in principle support roughly 10 to 20 m of water, and in many trees it is smaller or absent in summer. It explains drops of sap on leaf edges on a still, humid morning, not water at the top of a redwood.

Why does the water column not snap? Water in a thin, clean, closed tube holds together well, so it can survive being pulled hard. This state is metastable: it is fine until something breaks it. If a gas bubble forms and spreads in a tube (cavitation, the topic of lesson 4), that tube stops carrying water. A tree has millions of tubes in parallel, so one or a few failures do not stop the flow.

A useful comparison is a suction pump or a drinking straw. They lower the pressure at the top of a column so that the atmosphere pushes water up from below. The atmosphere can balance at most about 10 m of water (0.101 MPa). A tree gets past 10 m because the water is not pushed up by the air, it is pulled by cohesion from above, and the pressure in the xylem goes below zero.