Across 565 Pheidole species, antennae scale with body size within each caste and overlap completely between them — and manipulating the colony’s pheromones produced castes that normally never exist.
Body parts scale with body size, and they do so predictably enough that the relationship has a name. Allometry — the way a trait changes in proportion to overall size — is one of the oldest quantitative regularities in biology, and it is why a large animal is not simply a small one enlarged uniformly.
A study in the Proceedings of the National Academy of Sciences from the University of Ottawa reports a group of ants in which that relationship holds inside a caste and vanishes between castes. The same genome produces workers and soldiers of dramatically different sizes, and gives both the same antennae.
The lead author is Erica Vong, a graduate researcher at uOttawa, with Shannon Parisien and Hélène Orfali, in the Eco-Evo-Devo lab of Rajendhran Rajakumar, associate professor in the Department of Biology.
The pattern
In Pheidole ants, antennae scale with body size within each caste, exactly as expected. Across castes they do not: worker and soldier antenna sizes overlap completely, even though soldiers carry much larger heads and bodies.
Pheidole is the right genus to find this in. It is one of the most species-rich ant genera on Earth, and its colonies contain two sharply distinct worker castes — small minor workers that do the foraging and brood care, and large-headed soldiers built around the musculature for a pair of oversized mandibles. The size gap between them is among the most extreme in social insects.
The comparative work measured the pattern across 565 Pheidole species, drawing on laboratory experiments, field data, museum collections and comparative analysis. On that evidence the decoupling appears widespread through the genus, and may have originated near the common ancestor of Pheidole and Cephalotes turtle ants.
“Biologists have long assumed that traits such as antennae generally scale predictably with body size,” Rajakumar said.
Why antennae in particular should escape that assumption has a functional logic the study implies rather than tests. An antenna is a sensory organ, and its job — detecting pheromones, recognising nestmates, reading the chemical state of the colony — does not get easier or harder with the size of the ant carrying it. A sensory structure may have an optimal size set by what it has to detect, rather than by the body it sits on, in which case scaling it with the body would make it worse.
How the colony controls it
The developmental half of the study worked in one species, Pheidole dentata, and found the flexibility to be regulated rather than fixed: by the colony’s social environment, by juvenile hormone, and by signals from what the authors describe as an allometry-regulating organ.
Juvenile hormone is the familiar lever in insect caste determination — the titre a larva experiences at a sensitive window is what commits it to one caste or another — so finding it involved is expected. The social environment being a direct input is the more interesting part, because it means the decision is made with reference to what the colony currently needs rather than by the larva’s own trajectory alone.
Altering these factors, and especially the soldier inhibitory pheromone, produced what the team call nanoworkers and nanosoldiers: individuals with combinations of traits that normal development usually conceals.
The soldier inhibitory pheromone is the colony’s own regulator. Existing soldiers emit it, and it suppresses the development of more soldiers, which is a negative feedback loop that holds the caste ratio roughly constant. Removing or manipulating it is what let the researchers push development into configurations that do not occur in an intact colony.
“Sometimes new possibilities emerge when the expected connections between traits are broken and development is reorganized,” Rajakumar said.
Those unusual individuals are the strongest evidence in the paper. A comparative pattern across 565 species establishes that antenna size and body size are decoupled; producing novel trait combinations by manipulating specific signals demonstrates the mechanism that keeps them decoupled, and shows that the range of forms a Pheidole genome can build is wider than the range it normally builds.
“The antenna is not simply getting bigger because the ant is bigger,” Rajakumar said.
What “breaking a rule” means here
The release’s framing — that ants break a basic rule of biology — is worth moderating slightly, because allometric decoupling is not unknown.
Beetle horns, stalk-eyed flies and a range of sexually selected structures are standard examples of traits whose scaling has been altered by selection, and developmental biologists have studied how such decoupling happens for decades. What is distinctive here is not that scaling was broken but that it was broken selectively — preserved within each caste, abolished between them — and that the switch is under social control. A rule suspended in one context and maintained in another is a more interesting finding than a rule simply violated.
The evidence also sits at two different levels, and they should not be merged. The 565-species comparison is observational and largely morphometric, drawn substantially from museum specimens; the mechanism is from one species in a laboratory. Whether the same hormonal and pheromonal machinery regulates the pattern across the genus is an inference from the pattern’s distribution, not something the study measured in more than one species.
The evolutionary claim is the most provisional. Placing the origin near the common ancestor of Pheidole and Cephalotes is a reconstruction from the current distribution of a trait across a phylogeny, which depends on the tree used and on the assumption that the trait has not been lost and regained. The release reports no confidence measure for it.
Neither funding sources nor conflicts of interest are stated on the release, which for a PNAS paper with four named contributors is an omission worth noting.
What the work establishes is that a colony of ants has more developmental latitude than its finished products suggest, and that some of that latitude is held in reserve by a pheromone the ants themselves produce. The paper was published on 6 October 2026.
Notes
- Vong, E., et al. “Breaking trait scaling via the evolution and regulation of dimorphic decoupling in ants.” Proceedings of the National Academy of Sciences, 2026. DOI: 10.1073/pnas.2612172123.
- University of Ottawa. “Ants break a basic rule of biology to help them adapt to new environments.” EurekAlert!, 22 October 2026. https://www.eurekalert.org/news-releases/1148843.
Photo by Poranimm Athithawatthee on Pexels.com