Plant Biology

Cohesion-Tension Theory: How Trees Pull Water 100 Meters Up Under Negative Pressure

A coast redwood lifts water 112 meters into the air using no pump, no heart, and essentially no metabolic energy — powered instead by sunlight evaporating water from its leaves and by the sheer tensile strength of water itself, which inside the tree is stretched to a negative pressure of roughly −1.9 megapascals (about −19 atmospheres). That is a suction stronger than any human pump could sustain over such a column without the water boiling into vapor.

Cohesion-tension theory is the accepted physical explanation for how water ascends the xylem of vascular plants. It states that water is pulled from above (not pushed from below) as transpiration from the leaves generates tension, and that this tension is transmitted down an unbroken, thread-like column of water held together by hydrogen bonding (cohesion) and anchored to the xylem walls (adhesion). Proposed by Henry Horatio Dixon and John Joly in 1894, it treats the plant as a passive wick in a continuous soil-plant-atmosphere continuum.

  • TypePassive physical water-transport mechanism (no ATP)
  • LocationXylem conduits (tracheids & vessels), root to leaf
  • Key playersWater H-bonds, transpiration, stomata, aquaporins, pit membranes
  • Driving forceWater-potential gradient; leaf tension ≈ −1.5 to −2 MPa
  • ProposedDixon & Joly, 1894–1895 (Böhm 1893 precursor)
  • Found inAll vascular plants (tracheophytes); most extreme in tall trees

Interactive visualization

Press play, or step through manually. The visualization is yours to drive — try it before reading on.

Open visualization fullscreen ↗

Watch the 60-second explainer

A condensed visual walkthrough — narrated, captioned, under a minute.

What It Is and Where It Happens

Cohesion-tension theory explains the ascent of sap — the bulk flow of water and dissolved minerals from roots to leaves — through the xylem, the plant's non-living, pipe-like vascular tissue. The conducting cells are tracheids (long, tapered, in all vascular plants) and vessel elements (wider, stacked end-to-end into vessels, in angiosperms). Both are dead at maturity: their protoplasts have degraded and their end walls are perforated or pitted, leaving hollow lignified tubes 10–500 μm wide.

The key insight is that this transport is passive and cohesive. Unlike phloem loading or ion uptake, moving water up the xylem costs the plant essentially no ATP; the energy comes from the sun, which evaporates water at the leaf surface. The plant behaves like a wick embedded in the soil-plant-atmosphere continuum (SPAC), a single connected water system spanning a water-potential drop of roughly 95 MPa from wet soil to dry air.

  • Cohesion: water-water hydrogen bonding holds the column together.
  • Tension: transpiration at the top puts the whole column under negative pressure.
  • Adhesion: water clings to hydrophilic cellulose in conduit walls.

The Mechanism, Step by Step

The engine sits in the leaf, not the root. The chain of events:

  • 1. Evaporation at the menisci. Water evaporates from the wet cellulose walls of mesophyll cells into the leaf's air spaces, then exits through open stomata. This retreats the air-water interface into the nanoscale pores of the cell wall (~5–20 nm).
  • 2. Surface tension generates negative pressure. By the Young-Laplace relation, a curved meniscus in a small pore sustains large negative pressure: ΔP = −2γ/r, where γ ≈ 0.072 N/m. A 10 nm pore can theoretically hold tension well beyond −10 MPa.
  • 3. Tension propagates downward. Because the water column is continuous and cohesive, this pull is transmitted, molecule to molecule, all the way down to the roots — like tugging a rope.
  • 4. Water enters the roots. The tension lowers root water potential below that of the soil, so water flows in osmotically across the epidermis and cortex, past the Casparian strip of the endodermis.

Flow rate obeys a Darcy-type law: J = −K·(ΔΨ/L). At steady state the transpiration pull and root uptake balance, and water moves at 1–100 m/h in trunk vessels.

Key Molecules and Characteristic Numbers

The hydrogen bond is the hero. Each water molecule forms up to four H-bonds (~5 kcal/mol each), and although any single bond lasts only picoseconds, their collective, constantly-renewing network gives liquid water an enormous theoretical tensile strength of roughly −25 to −30 MPa — enough, in principle, to hold a column of water hundreds of meters tall.

  • Aquaporins (e.g., PIP1, PIP2 family): ~28 kDa membrane channels that speed water crossing living cells (radial path in root and leaf) at ~10⁹ molecules/s per pore; they fine-tune hydraulic conductance and gate under stress.
  • Stomata & guard cells: regulate the top valve; controlled by blue-light, ABA (abscisic acid) signaling, and K⁺/malate turgor changes.
  • Pit membranes: porous cellulose meshes (pores ~5–50 nm) between conduits that let water pass but block air seeding.

Concrete case: a coast redwood at 112.7 m needs to overcome a gravitational potential of −0.01 MPa/m ≈ −1.13 MPa plus friction, giving canopy tensions near −1.9 MPa. Measured mid-day leaf water potentials of −1.5 to −2.0 MPa match this closely.

How It Is Studied and Regulated

Because the water is under tension (metastable), it cannot be sampled with an ordinary gauge — insert a needle and the column snaps. Techniques instead include:

  • Scholander pressure chamber (1965): a cut leaf is sealed with the petiole poking out; gas pressure is raised until sap just returns to the cut surface. The balancing pressure equals the tension the leaf held. This is the workhorse field method.
  • Cavitron / centrifuge: spinning branches generates known negative pressures; used to build vulnerability curves showing the tension (P50) at which 50% of conductivity is lost to embolism.
  • Acoustic & optical embolism detection: ultrasonic clicks and micro-CT / X-ray imaging catch air bubbles forming in real time.

Regulation is largely at the stomata: as tension rises toward dangerous levels, ABA triggers guard-cell closure, throttling transpiration — a trade-off between water safety and CO₂ intake for photosynthesis. Species differ in this 'isohydric' (conservative) vs 'anisohydric' (risky) strategy. Aquaporin expression and xylem anatomy (conduit diameter, pit structure) tune the system over longer timescales.

Cohesion-tension is often confused with the older, minor mechanisms it largely replaced:

  • Root pressure pushes water up from below via active ion pumping into root xylem, generating positive pressure (up to ~0.2 MPa). It causes guttation and refills embolisms overnight but is far too weak to lift water more than a few meters — and vanishes in rapidly transpiring plants.
  • Capillary action alone in xylem-sized tubes (say 50 μm) raises water only ~0.6 m — negligible for a tree. Capillarity matters at the nanoscale menisci, not in the conduit bore.
  • Phloem transport (pressure-flow / Münch hypothesis) is the opposite sign: living sieve tubes under positive turgor push sugary sap from source to sink, an active osmotic process.
  • Transpiration pull is essentially cohesion-tension named for its driving force; they are the same phenomenon.

The distinguishing claim of cohesion-tension is negative pressure sustained by cohesion — water in a metastable, stretched state, which no other transport theory invokes.

Significance, Failure Modes, and Open Questions

Cohesion-tension theory sets the ultimate height limit for trees (~120–130 m): as trees grow taller, canopy tension rises, forcing stomata smaller and lowering internal CO₂, which starves photosynthesis and caps growth — as George Koch and colleagues showed in redwoods (2004).

The system's Achilles heel is cavitation: when tension exceeds the water's practical limit or air is 'seeded' through a pit membrane, the column ruptures and an air bubble expands into a gas-filled embolism, plugging the conduit (an audible click). Drought pushes tensions past the P50 threshold, causing runaway embolism and hydraulic failure — a leading mechanism of drought-induced tree mortality now widespread under climate warming.

  • Open question — embolism refilling: how do some plants refill gas-blocked conduits while the surrounding xylem is still under tension? Proposed roles for aquaporins, sugar unloading, and xylem surfactants (lipids that stabilize nanobubbles) remain debated.
  • Open question — how water sustains such tension without spontaneously cavitating; the answer lies in the tiny, defect-free pore geometry that suppresses bubble nucleation.

Applications span drought-resilient crop breeding, forest die-off forecasting, and biomimetic 'synthetic trees' that pump water by evaporation alone.

Water-potential (Ψ) values along the soil-plant-atmosphere continuum for a transpiring tree, and how tension builds toward the canopy.
LocationWater potential Ψ (MPa)Pressure stateNote
Moist soil−0.1 to −0.3Slightly negativeWater enters roots osmotically
Root xylem−0.4 to −0.8Mild tensionEndodermis / Casparian strip gate
Trunk xylem (base of tall tree)−0.8 to −1.2TensionContinuous water column under pull
Leaf xylem / mesophyll (canopy)−1.5 to −2.0Strong tension (metastable)Evaporation at cell-wall menisci
Leaf air spaces (inside stomata)≈ −10 to −30Very low Ψ~95–99% relative humidity
Atmosphere (50% RH, 20 °C)≈ −95Extremely low ΨThe ultimate 'pull' driving flow

Frequently asked questions

Does moving water up a tree cost the plant energy (ATP)?

No — the ascent of water in the xylem is passive and requires essentially no metabolic energy from the plant. The energy is supplied by the sun, which evaporates water from the leaves; that evaporation generates the tension that pulls the whole column up. The plant only spends ATP indirectly, on things like opening stomata and running aquaporins and ion pumps that fine-tune the flow.

How can water be under negative pressure without boiling or breaking?

Liquid water in the xylem is in a metastable, stretched state. Its extensive hydrogen-bond network gives it a theoretical tensile strength around −25 to −30 MPa, far beyond the −2 MPa trees actually need. It doesn't cavitate because xylem conduits are tiny and nearly defect-free, so there are no gas nuclei to seed a bubble — as long as air is not pulled in through a pit membrane.

What is the difference between cohesion and adhesion here?

Cohesion is the attraction of water molecules to each other via hydrogen bonds; it keeps the water column continuous so tension can be transmitted from leaf to root like pulling a rope. Adhesion is the attraction of water to the hydrophilic cellulose lining the xylem walls; it prevents the column from pulling away from the walls and helps maintain the tiny curved menisci that generate the tension in the first place.

How do we know the theory is correct if the water is under tension?

The strongest evidence comes from the Scholander pressure chamber, which measures the balancing pressure needed to return sap to a cut surface — matching predicted tensions of −0.5 to −2 MPa. Centrifuge (cavitron) experiments confirm water sustains negative pressures beyond −1.6 MPa without cavitating, and acoustic and X-ray micro-CT methods directly detect the air bubbles (embolisms) predicted when tension gets too high.

What is cavitation and why does it matter?

Cavitation is the sudden rupture of the tensioned water column, forming a vapor/air bubble that expands to fill and block a xylem conduit — an embolism. It happens when tension exceeds the water's practical limit or when air is aspirated through a pit membrane (air-seeding). Widespread embolism causes hydraulic failure and is a major mechanism of drought-induced tree death, making it central to predicting forest die-off under climate change.

How is cohesion-tension different from root pressure?

Root pressure is a push mechanism: roots actively pump ions into the xylem, drawing water in osmotically and creating positive pressure (up to ~0.2 MPa) that shoves water upward. It only lifts water a few meters and disappears in actively transpiring plants. Cohesion-tension is a pull mechanism driven by transpiration and negative pressure from above, and it is the only force strong enough to raise water 100+ meters in tall trees.