Tschinkel made his first cast in 1985—a nest of fire ants, known as Solenopsis invicta, meaning “the unvanquished.” They create huge, long-lived colonies with a quarter of a million individuals, and queens that live for seven years. While most ants defend only their nests, fire ants ferociously defend surrounding territory, too, often over 1,000 square feet, and their stings are memorable even to mammals. Tschinkel had completed groundbreaking studies revealing “the behavioral rules governing the flow of food” in their colonies. He had explored their nests—first chloroforming the inhabitants, partly for his own safety but mostly “to knock them down where they stood so I could see how they were distributed in the nest”—and thought he had a good idea of the nests’ geometry. But when he poured dental plaster into one and then dug it out, the picture was much clearer. “The nests of fire ants are a lot more patterned and less randomly arranged than I had thought,” he says. “They were obviously organized, regular, predictable—so interesting. I got into the architecture.”

A few years later, he cast the nest of Odontomachus brunneus, the trap-jaw ant, named for its unusual facial structure. The trap-jaw’s gigantic mandibles protrude to the sides, giving it the look of a hammerhead shark. The jaws are remarkably strong: If the ant clamps something too smooth and round to hold on to and its jaws slip off, they snap shut with enough force to shoot the ant three inches backward. In this cast, Tschinkel recognized the same construction he’d seen in the fire-ant nest, “only here the internal nest consisted of a single unit—the shish-kebab unit.” That is Tschinkel’s description of chambers strung one after another along a single vertical tunnel, giving the cast itself a lumps-along-a-stick appearance. “So I got the idea of a basic, widespread architectural unit that might be fundamental to many ant nests.” 

Fire-ant nests are shallow; most of the chambers are linked closely to the core near the surface and come out of the ground pretty much intact. But the trap-jaw ants had built more of a sprawling nest, one that would lose real character—and data—if it was incomplete. Tschinkel had to retrieve all the pieces, and to see it whole, he had to devise a way of gluing and supporting an entire reassembled cast.

He was still pondering the problem when he got interested in the Florida harvester ant—Pogonomyrmex badius, casually known as the pogo. One of the more impressive ant species, the harvester constructs an elaborate, seven-foot-deep nest in less than a week, moving pounds of sand in the process. Then foragers search their territory for seeds, which are stored—as many as 300,000 of them—in subterranean chambers. Workers crush the seeds into pulp and feed it to the larvae. In turn, Tschinkel thinks, the larvae probably return a nourishing liquid to the workers, supplementing their diet of sweet plant exudates, aphid honeydew, and juices sucked from prey insects. Tschinkel’s early attempts to clearly describe the areas in the nests where all this happens were unsuccessful. But in the early 1990s, he found a freshly abandoned pogo nest, and he filled the entire thing with a single five-gallon pour of dental plaster. Once the plaster hardened, the cast came out of the ground—in 180 pieces.

“I cleaned them up, and they sat on my lab bench for three or four years,” he says. “Assembling it seemed daunting.” But Tschinkel, a hobby woodworker whose house is filled with elegant handmade furniture of his own design, devised a method of gluing the broken casting together with epoxy and mounting the cast in front of a tall plywood backboard, supporting it with projecting steel welding rods so that it would hang in space in the same orientation it occupied in the ground. “I started assembling subunits on the lab table,” he says, and over months—many times longer than it took the ants to build the nest—“I reassembled the cast into perhaps a dozen subunits and then figured out how these went together.” The nest of the harvester colony has 130 chambers connected by about 30 feet of vertical tunnels.

He did the same with other species, including Aphaenogaster ashmeadi and Pheidole morrisii, and some of those mounted casts occupy Plexiglas cases outside his office on the Florida State campus. They are, as Tschinkel describes them, “physically, intellectually, and aesthetically pleasing.”

Tschinkel believes that an ant colony grows just as a single organism does, by rules that guide interactions among its cells and between it and its environment, a process called embryogenesis. A colony is “produced from the single, mated queen through the rules and interactions of sociogenesis”—the process by which a society grows and changes according to its internal rules. “And just as mature organisms differ, reflecting the rules of embryogeny,” he says, mature ant colonies differ as well, reflecting variations in the rules of sociogenesis.

Tschinkel is trying to describe those rules. He studies, for example, how worker size, distribution, and labor patterns change as an ant colony grows, and how labor division by worker size and age helps shape the colony’s structure and habits. Such factors appear to organize the workforce the way a factory floor plan organizes personnel. Young workers start out low in the nest, looking after the brood and the queen, and then move upward as they age, taking on more-responsible jobs—“general nest maintenance, food preparation, seed storage. Finally, they move even higher to become guards and trash collectors and, at last, foragers.”

He is also documenting how new ant colonies begin, including some unusual variations on the model in which the queen digs a hole and starts things rolling. Although newly mated fire-ant queens usually found new colonies alone, sometimes they do it in cooperation with other newly mated queens that arrive on the scene simultaneously. That’s a puzzle because it would seem risky: Worker ants tend to kill all but one such queen. Sometimes a mated queen will settle in an orphaned, queenless colony, although she’s unrelated to the workers there, and take over as a kind of royal parasite. Tschinkel has no idea why the workers are willing to serve such a usurper. In addition, a new colony’s workers often steal a brood from other new colonies, whose workers steal it back, and so on, until one colony wins. Then all the workers go and live in the winning nest, thus abandoning a mother.

Ant-nest design has a basic theme, Tschinkel says: Vertical tunnels for movement and transport, and horizontal chambers for work, storage, and housing the brood. But nests differ in shape, number, size of chambers, and how they’re connected, depending on the species. With the Florida harvester-ant nest, for instance, the largest chambers are near the surface and closely spaced, becoming smaller and farther apart deeper in the ground. Small chambers are oval in shape; larger ones are multilobed and more complex.

But exactly how the workers “know” to generate these shapes is not so obvious. “As they’re doing the work, each worker responds to what needs to be done,” he says. “What are the properties of individual ant workers so that once each has made her contribution, the sum is a particular outcome?”

One of Tschinkel’s graduate students, Sasha Mikheyev, analyzed 17 nest casts of Formica pallidafulva. She consistently found that when the descending tunnels are vertical, the adjoining chambers are round, and when the tunnels are inclined, the chambers are oval or teardrop-shaped and lined up along the tunnel’s axis. In a simple way, this observation illustrates one of the rules for how nests are built, Tschinkel says: If a tunnel is vertical, the ants doing the digging tend to distribute themselves evenly as they work, and if it is sloped, they tend to collect in the lower end.

That’s a start, but it’s still unknown which workers do the digging, whether they have this directional bias individually or as a group, or how the number of ants may influence nest size and shape. “I can imagine if there are only a few, they might dig only a tunnel, because they wouldn’t be crowded. But if there are more, they might dig chambers too,” Tschinkel says.

Months later, on an August morning, Tschinkel is deep in the Apalachicola National Forest with a whole new idea packed into the bed of a pickup truck. Over the years, Tschinkel has cast ant nests with latex, plaster of paris, and dental plaster enhanced with glass fibers. Each has advantages, but none is perfect. So today he’ll be trying something new: molten metal. He has spent months fabricating a clever foundry based on a kiln of fireclay in a galvanized garbage can and an air blower made from an auto heater fan.

Tschinkel sets up the works, piles in charcoal, lights it, and then waits an hour for 30 pounds of scrap zinc to melt. Meanwhile, he builds a mud dam around the entrance of a pogo nest and blows away loose sand through a plastic tube. Finally, he pours in the molten zinc. It flows so smoothly that Tschinkel worries it might be disappearing down a subterranean rat hole. After waiting 10 minutes for it to cool and harden, he starts digging beside the nest with his favorite shovel.

“It’s like buried treasure,” says Kevin Haight, a graduate student, as gleaming metal emerges from the ground. Bristling from some of the tunnels are hairlike projections, perfectly captured—the tunnels of another ant species, the tiny, sneaky thief ant Monomorium viridum, which survives by raiding the broods of other ant species. Haight ties a rope to the heavy cast and helps haul it out of the ground. It emerges in just eight pieces. “Fabulous,” Tschinkel says.

But later, when he has time to think about it, he concludes that zinc is too dense. The metal cools and sets up before it reaches the bottom of the nest. Next time, he says, he’ll make a first pour with molten aluminum and a second pour in zinc.

He has many opportunities to perfect his technique for making 3-D casts. There are 50 ground-nesting ant species in the area alone, and roughly 5,000 worldwide, each with its own unique way of life and shape of nest. For instance, there’s the genus Atta, the leaf cutter, which builds the world’s largest nests, up to 35 feet deep and covering as much surface as a small house. “I’d love to do an Atta nest,” Tschinkel says, smiling, “but I’d need several tons of plaster.”