10 Cool Papers | February 2026

Carnegie author: Jeffrey Dukes
Publication: Nature Ecology & Evolution

As droughts grow more frequent and severe, ecologists want to know which traits separate the plants that persist from those that lose ground. This study tracked 661 populations of 421 herbaceous species (non-woody plants like grasses and wildflowers) across 63 grassland and shrubland sites on six continents—all part of the International Drought Experiment—through a year of extreme drought. The goal: figure out which traits help certain plants hold on while others decline.

The answer depended on the kind of plant. In forbs—broad-leaved herbaceous plants like sunflowers and clover—drought resistance was linked to traits like deeper or thicker roots and larger plant size, not just leaf chemistry. But no single trait told the whole story. Root traits only predicted performance in combination with other traits, suggesting that drought resistance is less about any one adaptation and more about how a plant's whole toolkit works together. Annuals in wetter sites appeared to benefit from strategies that let them complete their life cycle before conditions deteriorate, but that advantage disappeared under harsher drought. The takeaway? Predicting how a grassland will weather a drought requires a nuanced approach, matching the right traits to the right kind of plant.

Carnegie author: Chen-Ming Fan
Publication: Cell Reports

The interior of a tumor is not a hospitable place. It tends to be acidic, starved of oxygen, and low on nutrients—a chemical gauntlet that shapes how cancer cells behave in ways researchers are still working out. This study found that short-term acidity (24 to 72 hours) suppressed cancer cell growth, slowed their metabolism, and largely stopped them from migrating and breaking away from tumor clusters, both in lab cultures and in animal models.

That sounds like it could be useful—until low oxygen enters the picture. The researchers found that a lack of adequate oxygen reversed all of that suppression and restored the cells' ability to move and spread. Some cancer cells can also adapt to acidity on their own after prolonged exposure (more than three weeks), though that capacity varies by cell type and is not universal. For researchers trying to predict when and how tumors spread, the chemical context inside a tumor—not just its genetic profile—may be a critical variable.

Carnegie author:  John Urban
Publication: Genome Research

Stem cells can both renew themselves and produce specialized cell types—that much is textbook. What is far less understood is how they manage it inside a living organism, where signals from neighboring cells, layers of molecular packaging, and the timing of DNA replication all feed into a cell's sense of its own identity. Studying that machinery in living tissue rather than in a dish has been difficult, in part because stem cells are rare and the tissues around them are complex.

To get around that, researchers developed new tools tailored to specific cell types and applied them to the Drosophila (fruit fly) testis, a well-established model for studying adult stem cells. They measured three things at once in two stem cell populations: which genes were active, how the DNA was packaged (tightly wound and silent, or loosely coiled and accessible), and when different stretches of the genome got copied during cell division. The two populations—germline stem cells, which produce sperm, and somatic cyst stem cells, which support them—differed in revealing ways. Germline stem cells, for example, showed a distinct DNA-copying schedule that tracked closely with how their DNA was packaged. 

That kind of layered insight only emerges when you measure multiple things at once in the same cells within a living tissue, what biologists are calling a "multi-omics" approach. Altogether, the tools and datasets from this work provide a foundation that could inform stem cell research well beyond the fruit fly.

Carnegie author: Anirudh Prabhu
Publication: Applied Geochemistry

When a geochemical dataset includes dozens of elements measured across thousands of samples, the most scientifically important data points—unusual combinations of trace metals, unexpected clustering—are often the hardest to spot. Standard tools like scatter plots and principal component analysis work by compressing the data into fewer dimensions, which inevitably loses information.

GeoChemNet takes a different approach: it maps the relationships among samples as a network, preserving more of the original complexity. In two case studies from the same region, the tool revealed hidden structure within rock units and helped narrow the search area for copper deposits hosted in sedimentary rock. Because the results are visual and interactive, GeoChemNet lets geoscientists and data scientists work from the same picture—which could speed up everything from basic geological mapping to mineral exploration.

Publication: The Astronomical Journal
Carnegie author: Michael Greklek-McKeon

About 72 light-years away, two small, cool stars orbit each other at a relatively close distance—roughly the gap between the Sun and Saturn. The system, TOI-2267, was already known to host two Earth-sized planets. Now, using new observations from the 5.1-meter Hale Telescope at Palomar Observatory along with archival data from NASA's planet-hunting TESS mission and high-resolution imaging, astronomers have confirmed a third: TOI-2267 d, also roughly the size of Earth.

Here is the puzzle: the team cannot yet tell which of the two stars the new planet orbits. If it belongs to a different star than the other two, TOI-2267 would be the first known pair of small red dwarf stars where both have planets that pass in front of them as seen from Earth—what astronomers call transiting. Similarly, if all three planets orbit the same star, the system is extraordinarily compact. Either way, it offers a rare laboratory for studying how small rocky planets form and survive around tightly paired stars.

Carnegie-affiliated authors: Claire Zurkowski (Carnegie Alumna), Jing (Jill) Yang, and Yingwei Fei 
Publication: Geophysical Research Letters

Deep inside Earth's mantle most minerals have crystal structures in which each metal atom is surrounded by six oxygen neighbors. Push the pressure high enough, and those structures collapse into tighter arrangements, with eight neighbors instead of six, fundamentally changing the mineral's physical properties. While we may not reach it on Earth, that transition is predicted to happen inside super-Earths (rocky planets larger than our own), but at pressures that are hard to reach in the lab.

To get a preview, researchers squeezed magnesium iron oxide (MgFe₂O₄) past 65 gigapascals—about 650,000 times the pressure of Earth's atmosphere—and heated it to nearly 2,840 degrees kelvin. They found a stable phase with the tighter eight-neighbor arrangement at pressures roughly 400 gigapascals lower than where the same structure is predicted to form in a related silicon-based mineral. 

That gap suggests iron-rich chemistry could trigger these deep structural changes far earlier than expected inside super-Earth mantles and current models of what those interiors look like may need to be revised.

Carnegie author: Ian Szumila
Publication: Nature Astronomy

A magnetic field is one of the most consequential features a planet can have. It deflects radiation from its host star and shields the atmosphere from being stripped away, making it central to questions about whether a planet could support life. Earth generates its field because the liquid iron in its outer core is electrically conductive and constantly churning, a process called a dynamo. But Earth is not the only kind of planet that might pull this off.

On young rocky worlds, giant impacts can produce globe-spanning magma oceans. As those oceans cool, dense iron-rich melt may sink to the base of the mantle, pooling into a molten layer. If that layer conducts electricity well enough, it could drive a dynamo of its own. 

To test the idea, researchers used laser-driven shock experiments to measure how well magnesium oxide and iron-bearing magnesium oxide conduct electricity at pressures between 467 and 1,400 gigapascals. The two turned out to be essentially the same—a surprise, since they expected iron content to make a big difference. The results suggest that super-Earths larger than roughly three to six times Earth's mass might be able to sustain these deep magma ocean dynamos for billions of years, generating magnetic fields nearly 10 times stronger than those powered by a metallic core. If so, the range of planets capable of shielding their atmospheres—and potentially supporting life—may be wider than previously thought.

Carnegie authors: Mark Phillips, Nidia Morrell, Carlos Contreras, Christopher Burns, Eric Persson, and Anthony Piro
Publication: The Astrophysical Journal

Type Ia supernovae—explosions triggered when a dense stellar remnant called a white dwarf gains enough mass to blow itself apart—are among the most important tools for measuring cosmic distances. Their peak brightness follows a predictable pattern tied to how quickly they fade, a trick that has helped map the expansion of the universe and provided some of the first evidence for dark energy. But fast-declining Type Ia supernovae, which dim more quickly than typical examples, have long been treated as too unreliable to use.

Drawing on 43 fast-declining events from the Carnegie Supernova Project, the researchers showed that these objects can still serve as distance markers when characterized using a measure called color-stretch, which captures how the supernova's brightness changes over time. They also developed a simpler method that estimates distance from the supernova's colors at peak brightness alone, without needing to track how it fades. That technique works because the colors of fast decliners are set mainly by temperature differences at the surface of the explosion rather than by dust between the supernova and Earth—a physical property unique to this subclass. 

The upshot: this class of supernovae, once written off as too quirky for the job, may be a new addition to the toolkit astronomers use to measure the universe's expansion.

Delayed Radio Emission in Tidal Disruption Events from Collisions of Outflows Driven by Disk Instabilities

Carnegie authors: Samantha Wu, Brenna Mockler, and Anthony Piro
Publication: The Astrophysical Journal

When a star strays too close to a supermassive black hole, the black hole's gravity shreds it—what astronomers call a tidal disruption event. Some of the debris falls back toward the black hole and powers a brilliant flash of light; some gets flung outward. That outgoing material can produce radio waves when it slams into surrounding gas, a signal that typically appears soon after the disruption. But in a growing number of cases, the radio signal does not show up for months or even years.

This paper proposes that the delay starts with the disk of gas that spirals around the black hole as material falls in. If that disk becomes unstable and briefly surges to a higher feeding rate, it can launch a fresh burst of material outward. The collision between that later burst and earlier debris or surrounding gas produces shocks traveling at 5 to 30 percent the speed of light—energetic enough to power the observed radio emission. The researchers' models reproduced the brightness and energy signatures of several real events, supporting disk instabilities as a unifying explanation for a growing and puzzling class of tidal disruptions.

Carnegie author: Sten Delos
Publication: Journal of Cosmology and Astroparticle Physics

Imagine dumping a huge bag of identical marbles across a field. No matter how carefully you spread them, some spots end up with a few extra and some with a few fewer. If dark matter is made of discrete chunks rather than a perfectly uniform fog, something similar should happen on the smallest cosmic scales—random clumps that arise simply because some patches of space ended up with slightly more dark matter than others. The authors call this built-in graininess "warm white noise."

What makes the idea interesting is its resilience. In standard models, the random motion of dark matter particles smooths out small-scale structure over time, a process called free streaming. But the researchers found that warm white noise can survive below that smoothing threshold and, under the right conditions, continue to grow as the universe evolves. Their mathematical predictions matched computer simulations well, both where clumps are still forming and where matter has already collapsed into dense, gravitationally bound structures. 

The framework offers a new way to test dark matter models—particularly scenarios involving extremely lightweight dark matter particles, where these effects might eventually show up in telescope observations. And, since the nature of dark matter remains one of the biggest open questions in physics, any new handle on its small-scale behavior matters.