Chinese Money Plant Hides a Voronoi Diagram in Every Leaf

Thirty-four leaves. Six plants. One mathematical pattern hiding on every windowsill that hosts a Chinese money plant.

A team at Cold Spring Harbor Laboratory in New York, working with researchers at the University of Calgary and the University of Alberta, has shown that the looped vein network on Pilea peperomioides leaves traces a Voronoi diagram, the same partition geometry used to draw school catchment maps and assign cell tower coverage. The study, published May 12 in Nature Communications, is the first documented case of a true Voronoi pattern in living tissue with both the dividing edges and the central points visible at the same time.

Geometry in a Houseplant Leaf

Pilea peperomioides is the round, flat plant that became a houseplant cliché in the mid-2010s. Cuttings circulate between neighbors. Garden centers stock it as a starter species for new growers, the kind that earns its place alongside the basics of indoor gardening for beginners.

Under a magnifying lens, the leaf does something stranger than its reputation. Looping veins carve the surface into closed polygons. Scattered across each leaf are hydathodes, microscopic pores that release excess water, balance fluid pressure inside the tissue, and help defend against bacterial invaders.

The Cold Spring Harbor team mapped 34 leaves across 6 plants and measured how the veins and pores lined up. The headline number landed quickly: about 73 percent of the looped polygons formed by major veins contained exactly one pore.

Stats snapshot of what came back from the leaf scans:

• 73 percent of looped polygons contain exactly one hydathode pore.
• 8 degrees mean deviation between actual vein angles and the angles a Voronoi partition would draw.
• 72 percent overlap between actual vein regions and the predicted Voronoi regions.
• 34 leaves measured across 6 plants, with three independent geometric tests.

What a Voronoi Diagram Does

A Voronoi diagram is a way of dividing a flat surface around a set of points. Drop the points anywhere, and the diagram carves the surface into regions, one region per point, with every spot inside a region closer to its own point than to any other. The math has been formalized since the mid-1800s and turns up wherever someone needs to partition space cleanly.

Where it gets used in practice:

• City planners draw school catchments, fire station coverage zones, and electoral districts on Voronoi grids.
• Wireless engineers assign cell tower service areas the same way, so every phone connects to its nearest base station.
• Biologists have spotted Voronoi-like patterns in giraffe spots, dragonfly wing veins, and the packing of epithelial cells under a microscope.
• Materials scientists model crystal grain boundaries with the same geometry.

None of those biological examples carried both visible centers and visible edges in the same picture. Giraffe spots show the centers without dividing lines. Dragonfly wings show the lines without obvious central points. The money plant is the first case to display both at once, which is what made its leaves usable as a test specimen for the underlying math.

Three Tests, One Verdict

The team did not stop at eyeballing the pattern. Cici Zheng, a former Cold Spring Harbor graduate student who now works at the Allen Institute in Seattle, ran three quantitative checks against the leaves she had mapped.

1. Geometry comparison. The team compared angles and distances between adjacent polygons against what a Voronoi partition would predict. Vein angles came in roughly 8 degrees off on average, a small error for a surface that grew without instructions.
2. Forward prediction. They asked whether actual veins matched the lines a Voronoi diagram would draw between known pores. Region overlap landed at 72 percent.
3. Reverse prediction. They worked backward, using only the vein polygons to predict where the pores should sit. Pores landed significantly closer to the predicted Voronoi centers than to any random or alternative reference point the team tested.

Three independent tests pointed the same way. The pattern is real, not a visual coincidence dressed up in math.

Why the Textbook Vein Model Fell Short

Plant biologists have explained leaf veins for decades through a model called canalization. Auxin, the hormone that directs growth, flows from one part of the leaf toward an exit point. Cells along its path get better at moving the hormone. The path narrows into a channel, and that channel hardens into a vein. The model works cleanly for one shape: a branching tree.

Most flowering plants do not grow branching-tree veins. They grow reticulate veins, the looped, mesh-like networks visible on almost any garden leaf held up to the light. Canalization can explain a single drainage line. It struggles to explain why the line closes into a loop, then meets another loop, then another.

For decades, the question of how reticulate veins form has remained open, and finally we have a plausible answer.

That comment came from Przemysław Prusinkiewicz, the University of Calgary computer scientist who has spent his career modeling plant form and is a co-author of the Nature Communications paper. It is a significant admission from inside the field. The most common vein topology in flowering plants did not have a clean mechanistic story, and the money plant gave the team a place to look for one.

Auxin Waves, Colliding

The new model keeps auxin at the center of the story but changes its job. Each hydathode behaves as an auxin source. From every pore, the hormone spreads outward in a wave. When two waves from neighboring pores meet, they collide and build a ridge of high concentration halfway between the sources. That ridge is where the vein appears.

The description matches a Voronoi diagram drawn algorithmically. A Voronoi edge sits at the locus of points equidistant from two seeds. An auxin ridge sits at the locus of points where waves from two sources arrive at the same time. The geometry comes out identical, which is why the money plant’s veins land on the predicted lines so consistently.

Computer simulations from the team produced vein networks that closely matched real leaves. Lab tests added a second layer of support: staining showed auxin carrier proteins clustering around hydathodes in the pattern the wave-collision model predicts.

Saket Navlakha, the Cold Spring Harbor associate professor who runs the lab and is senior author on the paper, has built a research program around the idea that organisms run algorithms without knowing it. Bacterial colonies optimize foraging like network routers. Slime molds solve mazes. In Navlakha’s broader research on biological algorithms, the money plant fits a long line of cases where local rules produce global geometry no central planner ever wrote down.

The Pattern Holds Under Stress

A clean pattern under one set of conditions could still be coincidence, locked in by a fixed genetic recipe. To check, the team grew the plant under shade, intense light, and high heat, then compared the geometry of the resulting leaves to controls.

Leaves came out different. Smaller in shade, larger and crinkled in heat, paler under high light. Hydathodes shifted in size with the stress. The Voronoi structure stayed intact.

That preservation rules out a hard-coded blueprint. The leaf is not unfolding a stored map of where every vein should go. It is running a local rule, in real time, using whatever pores it happens to have where they happen to sit.

Different starting conditions, same answer. That is the hallmark of an algorithm rather than a recipe, and it is the reason the team thinks the wave-collision logic may carry over to other species with looped vein networks, which is to say most of the flowering plants on the planet.

On every windowsill that holds a money plant, the geometry is being recomputed quietly, leaf by leaf.

Frequently Asked Questions

What Is a Voronoi Diagram in Simple Terms?

A Voronoi diagram divides a flat surface into regions around a set of points, with every location in a region closer to its own point than to any other. Cell tower coverage maps, school catchment maps, and crystal grain boundaries are all built on the same idea.

Why Is the Chinese Money Plant Special?

Pilea peperomioides is the first living tissue ever documented to show a Voronoi diagram with both the dividing edges (its veins) and the central points (its hydathodes) visible at the same time. Giraffe spots show centers without edges. Dragonfly wings show edges without obvious centers.

How Many Plants Did the Researchers Actually Measure?

The team analyzed 34 leaves across 6 plants. About 73 percent of the looped polygons formed by major veins contained exactly one pore, and vein angles deviated by roughly 8 degrees on average from the lines a Voronoi diagram would draw.

What Is a Hydathode?

A hydathode is a microscopic pore on a leaf, distinct from the more common stomata that handle gas exchange. Hydathodes release excess water in a process called guttation, help balance fluid pressure inside the leaf, and play a role in defending against bacterial invaders.

What Does Auxin Do in the New Model?

The team proposes that each hydathode acts as an auxin source, with the hormone spreading outward in a wave. Where waves from neighboring pores meet, the collision forms a ridge of high concentration that hardens into a vein. The geometry of those collision lines matches a Voronoi diagram drawn around the pores.

Does This Overturn the Canalization Theory of Veins?

No. Canalization remains the leading explanation for branching, tree-shaped vein patterns. The wave-collision model adds a mechanism for reticulate, looped vein networks, which canalization on its own cannot produce. The two processes can coexist in different parts of a plant or in different species.

Where Can I Read the Original Study?

The paper appeared in Nature Communications on May 12, 2026, under the title Reticulate leaf venation in Pilea peperomioides is a Voronoi diagram. An earlier preprint version is hosted on bioRxiv for readers who want the full methodology and supplementary figures.