It is the soil, not the plant, that sets the limit on water uptake during drought

For decades, agricultural breeders have tried to engineer crops that pull water more aggressively from drying soil. The strategy seemed logical: pack more dissolved solutes into plant cells, generate stronger osmotic pressure, and the plant should be able to extract water that its competitors cannot reach. Substantial money went into these breeding programs. None of them worked.

A study published March 23 in Science explains why. The bottleneck is not in the plant. It is in the dirt.

Negative pressure and the physics of drinking through soil

Plants move water from roots to leaves through a system that operates under negative pressure - essentially, the leaves pull water upward by evaporation, creating a tension that propagates down through the plant's vascular system and into the roots. Scientists call this negative water potential, and it allows trees to lift water tens of meters against gravity without any pump.

But the soil has its own physics. Water sits in pores of varying sizes between soil particles, held there by capillary forces. As soil dries, the remaining water retreats into smaller and smaller pores where capillary forces are stronger. Extracting water from those tiny spaces requires the plant to generate increasingly extreme negative pressures - and at some point, the system hits a wall.

Soil physicists have long known that when soil water potential drops below roughly -1.5 megapascals, plants can no longer extract water fast enough to sustain themselves. This threshold - known as the permanent wilting point - has been a standard reference in irrigation science for decades. But the question of what exactly enforces this limit, and whether plants could be engineered to push past it, remained open.

Where soil physics meets leaf biology

Andrea Carminati, professor of soil physics at ETH Zurich, approached the problem from below ground. Tim Brodribb, professor of plant physiology at the University of Tasmania, approached it from above. Their collaboration produced a synthesis that neither discipline had reached alone.

Plants regulate water loss through stomata - microscopic pores on the undersides of leaves that open to admit carbon dioxide for photosynthesis and close to conserve water. When soil dries and water becomes harder to extract, stomata close. This keeps the plant alive but starves it of carbon dioxide, slowing growth.

The critical insight from Carminati and Brodribb's analysis is that the physics of capillarity in soil pores predicts not just when soil runs out of accessible water, but also what happens in the leaves. The same physical constraints that govern water release from soil pores determine the tension at which stomata must close. The plant is not making an independent decision about when to shut down. It is responding to a physical limit imposed by the soil.

Why breeding programs failed

This finding resolves a puzzle that has frustrated crop scientists for years. If the plant's own properties - the strength of its cell walls, the concentration of solutes in its cells, the architecture of its vascular system - were the limiting factor, then modifying those properties through breeding or genetic engineering should improve drought tolerance. Breeders tried exactly that, investing heavily in varieties with higher cellular solute concentrations that should, in theory, generate stronger osmotic pull.

But if the limit is in the soil, then no amount of plant modification will help. A plant with twice the osmotic pressure still cannot extract water from pores where capillary forces exceed the available gradient. The soil sets the ceiling, and the plant has already evolved to operate near it.

As Brodribb put it, the limiting factor lies not in the plants but in the soil. The breeding programs were trying to solve the wrong problem.

Convergent evolution across land plants

The study's model calculations suggest that the convergence is not coincidental. Across a wide range of land plants - herbs, shrubs, trees - the relationship between soil water potential and stomatal closure follows similar patterns. This implies that natural selection has already pushed plants close to the physical limits imposed by soil capillarity. There is no untapped margin for breeders to exploit because evolution got there first.

This does not mean drought-tolerant crop development is hopeless, but it redirects the effort. If the soil is the bottleneck, then improving drought performance means changing the soil environment - through irrigation timing, soil structure management, or selecting for root architectures that access deeper or wetter soil layers - rather than trying to make the plant itself pull harder.

A model, not a field trial

The study relies on model calculations of water potential rather than field experiments across multiple soil types and plant species. The theoretical framework is compelling, but its predictions need to be tested under real agricultural conditions with different soils, climates, and crop varieties.

The permanent wilting point of -1.5 megapascals is itself an approximation that varies with soil texture. Sandy soils and clay soils behave differently, and the threshold at which water becomes inaccessible shifts accordingly. Whether the convergence between soil physics and stomatal behavior holds across the full range of agricultural soils remains to be confirmed.

The study also does not address other drought survival strategies that plants employ - deep root systems, water storage in stems, leaf shedding, or dormancy. These adaptations may allow some plants to outlast drought even if they cannot out-pull it.