The Chinese money plant, seen here against the backdrop of Cold Spring Harbor, is helping CSHL biologists uncover the mathematical formulas underlying nature itself. (Credit: Nick Wurm/CSHL)
Scientists Found Advanced Math Hidden in a Common Houseplant’s Leaves
In A Nutshell
- Researchers found that the vein network on Chinese money plant leaves forms an approximate Voronoi diagram, a well-known mathematical pattern in which each region is defined by its nearest central point.
- Water-secreting pores called hydathodes act as those central points, with veins forming the boundaries between them.
- A plant hormone called auxin is proposed to build this pattern by spreading outward from each pore in waves that collide and harden into veins where they meet.
- Even under heat and light stress, the pattern held firm, suggesting it emerges from a self-correcting biological process rather than a fixed genetic blueprint.
Nature has been solving an advanced geometry problem on the leaves of a popular houseplant, and scientists just figured it out.
Researchers studying the Chinese money plant, known scientifically as Pilea peperomioides, have discovered that the web of veins on its round, coin-shaped leaves follows a mathematical pattern called a Voronoi diagram. In plain terms, the veins divide the leaf’s surface into closed, polygon-shaped cells. Most of these vein-enclosed polygons contain one water-secreting pore, called a hydathode, located near the polygon’s mathematical center.
In a true Voronoi diagram, every point inside a polygon is closer to that polygon’s pore than to any neighboring pore. According to the study, published in Nature Communications, the plant’s veins come surprisingly close to that rule, not as a chalkboard abstraction, but as an approximation playing out in living plant tissue, built cell by cell during leaf development.
Beyond one houseplant, the discovery touches on a problem scientists have long struggled to solve: how flowering plants grow their looping, net-like vein patterns. Most existing models produce tree-shaped branches, not closed loops. This finding gives researchers a geometric framework to describe what’s happening and offers a new explanation for how it happens, one that could apply to many other plant species.
What Is a Voronoi Diagram and Why Does It Show Up in a Houseplant?
It’s the same math used to draw school district boundaries around schools or map the territories of competing cell towers, assigning every point in a region to its nearest hub. Some natural patterns, such as giraffe skin, can look Voronoi-like. But in many cases, scientists can see the borders without seeing the hidden center points needed to test the math rigorously.
In the Chinese money plant, researchers had a rare advantage: both the dividing lines (the veins) and the central points (the hydathodes) are visible and measurable. Hydathodes are larger than the tiny pores plants use for gas exchange and release water in a single direction.
A team based at Cold Spring Harbor Laboratory and the Universities of Alberta and Calgary analyzed 34 leaves from six Pilea plants, developing a computational pipeline to extract every hydathode position and trace the major vein network on each leaf.
To verify the claim, the researchers put the pattern through three separate mathematical tests, checking whether the geometry of the vein polygons matched what a true Voronoi diagram would predict, and whether hydathode positions alone could forecast where the veins would fall. In all three tests, hydathode locations outperformed alternative reference points such as the geometric center of each polygon or a random interior point. Deviations from a perfect Voronoi pattern were roughly equivalent to adding about 15% random noise to an ideal diagram, a small and biologically expected level of imprecision. As a control, the team applied the same tests to air chambers in a liverwort species; that structure failed all three, confirming the framework could distinguish true Voronoi patterns from lookalikes.
Running simulations using actual hydathode positions from real leaves, the team then checked their predictions against the distribution of PIN proteins, the molecular pumps that move auxin through plant tissue. When they mapped where those pumps were most active, the concentration was highest in and around hydathodes, lower in the secondary veins, and cells sitting next to forming veins showed the pumps orienting toward the vein, largely consistent with what the model predicted.
The Plant Hormone That May Help Build the Pattern
Knowing what pattern the veins form is only half the story. The team also proposed a mechanism for why, centered on auxin, a plant hormone proposed to drive vein formation.
For decades, the dominant canalization model has proposed that auxin gets funneled from source points to sinks into narrow channels that harden into veins. This works well for branching, tree-like networks but doesn’t naturally produce closed loops.
In the new model, auxin spreads outward from each hydathode in all directions, like a ripple in a pond. When waves from neighboring hydathodes meet, they form concentration peaks, and those peaks are where veins form, producing many of the closed loops seen in the leaf.
Chinese Money Plant Vein Pattern Holds Up Under Stress
One of the more telling results was how durable the Voronoi pattern proved to be. Plants were grown under shade, high-intensity light, and high temperature for five weeks. Stressed plants showed visible changes, including significantly larger hydathodes in the high-temperature group and variation in leaf color and size. Despite all of that, the Voronoi relationship remained intact, suggesting the geometry emerges from a local, self-adjusting process capable of arriving at the same solution under shifting conditions rather than a pre-programmed blueprint.
For decades, scientists studying vein formation have worked within a framework that handles branching patterns well but struggles with loops. Evidence suggests the main central vein of Pilea leaves is patterned by the older canalization mechanism before hydathodes even appear, so the two models aren’t mutually exclusive. But for the secondary veins that form the looping network, the colliding-wave model offers a more compelling explanation. If it extends to other species with hydathodes spread across their leaf surfaces, a piece of math growing quietly on leaves may finally have met its match.
Paper Notes
Limitations
The computational model was built under the simplifying assumption that leaf size and shape remain fixed during vein development, meaning the potential influence of leaf growth on Voronoi pattern formation was not captured. The model performs better at reproducing secondary veins than primary veins; the researchers note that primary veins appear before hydathodes and are likely patterned by a separate mechanism, and that the largest mismatches between the model and real leaves occurred near primary veins and the petiole attachment point. Some vein polygons on real leaves contained two or more hydathodes without a separating vein, a deviation the model does not replicate. On the experimental side, the immunolocalization work used an anti-PIN serum developed for other plant species rather than tools specific to Pilea, and the researchers could not rule out that signals from non-polar PIN proteins may have obscured polar PIN localization patterns. Whether the Voronoi vein geometry offers any functional advantages over alternative network patterns remains unknown.
Funding and Disclosures
The study received support from multiple sources, including the William R. Miller Fellowship at Cold Spring Harbor Laboratory, NIH grant R35GM125003, NSF awards IOS2224874 and 2131631, NSF CAREER award DBI-1846554, Natural Sciences and Engineering Research Council of Canada Discovery Grant 2019-06279, and funding from the Simons Center for Quantitative Biology at Cold Spring Harbor Laboratory. The authors declare no competing interests.
Publication Details
Authors: CiCi Xingyu Zheng, Shirsa Palit, Matthew Venezia, Elijah Blum, Ullas V. Pedmale, Dave Jackson, Enrico Scarpella, Przemyslaw Prusinkiewicz, and Saket Navlakha. Author affiliations include Cold Spring Harbor Laboratory (Cold Spring Harbor, NY), the University of Alberta (Edmonton, AB, Canada), and the University of Calgary (Calgary, AB, Canada). | Journal: Nature Communications | Paper Title: “Reticulate leaf venation in Pilea peperomioides is a Voronoi diagram” | DOI: https://doi.org/10.1038/s41467-026-71768-3 | Received: July 1, 2024 | Accepted: March 30, 2026 | Published online: May 12, 2026 | Article number: 4111







