The Movement Is in the Material

A fallen pine cone can close when it gets wet and open again as it dries. The surprising part is not that a plant responds to water. It is that this movement can continue in dead tissue.

No root has to supply it. No new growth has to push the scales apart. The object retains a working relationship between water and structure after the living machinery is no longer needed for the motion.

That makes a pine cone a particularly good thing to be wrong about. It looks like a little organism making a decision: bad weather, shut the doors. The better explanation is less human and, to me, more interesting. The movement is built into the material.

Two lengths that cannot agree

The useful starting point is a bimetallic strip: two attached materials that expand by different amounts when warmed. If separated, each could take its preferred length. Joined together, they constrain one another; in a freely bending strip, the mismatch produces curvature.

In a cone scale, water can play the role that temperature plays in the metal strip. Tissues change dimensions differently as they take up or lose moisture. Their attachment converts that unequal response into bending.

In their 2009 study, physicists Étienne Reyssat and L. Mahadevan followed this connection from individual cells to moving scales. They described a strongly moisture-responsive tissue opposed by a much less responsive one, with much of the deformation concentrated near the scale's attachment to the cone's central axis. The part beyond that region amplifies the motion.[1]

Think of a long arm turning at its base. As a deliberately arbitrary illustration, a rigid arm 30 millimeters long needs to turn only ten degrees to move its tip about 5.23 millimeters in straight-line displacement between positions. That is geometry, not a measurement of a pine cone. It shows why a conspicuous tip movement need not mean that every bit of tissue has stretched conspicuously.

The researchers also made paper–plastic bilayers. Paper's moisture response, constrained by a less responsive polymer, produced bending without a plant at all. Arranged as a little flower, their artificial petals opened when wetted and closed on drying—the reverse of the cone's overall action. The shared principle is unequal expansion, not a rule that water must always close an object.[1]

That is already a satisfying explanation. It is also where an explanation can become too tidy.

The resisting side is not inert

Calling something a “passive layer” invites an easy picture: one side drinks, the other does nothing. A newer study makes that picture difficult to keep.

Kim Ulrich and colleagues investigated Pinus jeffreyi scales using tissue weighing, three-dimensional X-ray imaging and comparisons of wet and dry internal structure. Their 2025 paper distinguishes three tissue categories: a sclereid cell layer, an intermediate brown tissue, and reinforcing sclerenchyma fibers.[2]

The names are less important than the difference they expose. All three tissues took up moisture. And a tissue's response depended on which direction you measured.

In the sampled fibers, median shrinkage along the cells' length was much smaller than shrinkage across them. The sclereid samples, by contrast, showed appreciable longitudinal shrinkage. The relevant contrast was not “changes size” versus “never changes size.” It was how much each tissue changed along the directions that matter to bending.[2, Table 1]

There is a second complication. Those reinforcing fibers are not necessarily equivalent to an unbroken slab of stiff material.

The team compared computer models ranging from a simple two-layer block to structures with separate fibers embedded in an intermediate tissue and a more scale-like geometry. With their chosen material properties, the uniform bilayer bent less than the more structured models. Smearing separate stiff fibers into a continuous layer had changed the model's resistance to bending.[2]

This does not mean the old physical principle was wrong. It means the principle was doing a different job from a detailed anatomical model.

“Attached tissues expand differently” tells us why a bend can happen. To predict how far a particular scale will bend, we also need to know where the material sits, how stiff it is, and how it responds in different directions.

Nor did the newer models settle every detail. Their stiffness inputs came from earlier work; a temperature change served as a numerical substitute for moisture-driven expansion and was adjusted to bring bending into the biological range. It was not an experiment heating cones to that temperature. Their simulated sideways curvature also ran opposite to the biological direction. The authors were testing the consequences of geometry and material assumptions, not unveiling a finished virtual cone.[2]

The same humidity is not the same state

There is another reason a cone is not a simple weather dial. It matters how it got to its current conditions.

Ulrich and colleagues found different moisture contents on the wetting and drying paths at the same relative humidity. Scale angles also depended on the path. In one basal-scale example, the difference approached twelve degrees in the region around 50–60 percent relative humidity.[2, Figure 4]

The name for this is hysteresis. Here it means that a current humidity reading is not enough to specify a unique material state or scale angle. A previously wetter scale and a previously drier one can arrive at the same surrounding humidity without arriving at the same configuration.

This is physical history, not recollection. And it does not turn the cone into a rain forecast.

The limits matter. The newer study used four cones from one tree, with only one sample per internal tissue sampling location for its imaging-based deformation analysis. Its controlled humidity steps were not rain falling on an intact cone. The earlier study's wetting and drying experiments likewise show why timing depends on water transport and geometry, not just on the existence of a swelling tissue. Neither paper gives every cone a universal closing time.[1,2]

It is plausible that a slow, history-dependent response could keep a cone from reacting too abruptly to a brief humidity change. The newer paper discusses that possibility. But the measured path dependence is not itself a demonstration that hysteresis improves seed survival or dispersal. A mechanical result and an evolutionary benefit need different evidence.

What I wanted to understand was how something dead could still move. The answer is not a leftover spark of life. A tissue can stop living without losing the arrangement that makes water do mechanical work.

Now the next question is harder: in an intact cone, how much of the sequence of opening comes from each scale's own material response, and how much comes from overlapping scales changing one another's access to drying air? The newer paper points toward research on this orchestration; I have not followed those cited studies yet. The hinge is only part of the object. The object may also change the conditions in which its hinges work.

Sources and scope

[1] E. Reyssat and L. Mahadevan, “Hygromorphs: from pine cones to biomimetic bilayers,” Journal of the Royal Society Interface 6 (2009), 951–957. Sections 2–4 and Figures 1–5. Full text. DOI: 10.1098/rsif.2009.0184.

[2] K. Ulrich et al., “Beyond the bilayer: multilayered hygroscopic actuation in pine cone scales,” Beilstein Journal of Nanotechnology 16 (2025), 1695–1710. Experimental methods, Table 1, Figure 4, modeling methods and Discussion. Full text. DOI: 10.3762/bjnano.16.119.

This is a reading-based explanation, not a new cone-wetting experiment. The arm calculation is illustrative geometry. No supplementary movies, raw imaging data or simulation runs were analyzed here.