Now Reading
More Fun Than Fun: The Underground Architects and Engineers of the Ant World

More Fun Than Fun: The Underground Architects and Engineers of the Ant World

Fire ants. Photo: Stephen Ausmus, public domain


  • Walter Tschinkel has been exploring the relationships between ant nest size, architecture, ant size, colony size, lifestyle and division of labour.
  • His work has shown that we can eventually construct phylogenetic trees of ants based on their nests.
  • In doing so, we can trace the evolution of nest-building behaviour more powerfully than by merely studying the ants’ behaviours themselves.

It doesn’t take much to get me excited about the marvels of insect societies. But Walter Tschinkel’s Ant Architecture: the Wonder, Beauty and Science of Underground Nests (2021) aroused my admiration for the architects and engineers of the ant world to an altogether new high. But the ants must share the credit for this with the author. Walter Tschinkel has produced a paragon of science writing, reinforcing my pet idea that the process of science is as important as the product. The icing on the cake is that his process of doing science not only inspires admiration but also invites imitation – it’s the kind of science that will rightly make almost anyone feel that they can also be a first-rate scientist.

Walter Tschinkel, a well-known American entomologist, is professor emeritus of biological sciences at the Florida State University. Tschinkel is perhaps best known for his extensive research and encyclopaedic synthesis of the invasive fire ant (Solenopsis invicta). In his foreword to Tschinkel’s The Fire Ants (2006), E.O. Wilson wrote:

“Tschinkel has chronicled the story of a species well worth telling, but he has done much more. He has delivered a masterpiece. The Fire Ants is an example of how future biology will be written: a holistic account of a species or a group of related species across all levels of biological organization, from the molecules that regulate the behaviour of the organisms (in social insects, also colonies), to the life cycles, their role in the environment, their classification, evolution, and biogeography—and, in this case especially, their importance to humanity.”

Tschinkel’s second passion concerns the architecture of ant nests, including those of the fire ants. In my previous essay in this column, I admired the ability of weaver ants to construct nests by binding leaves together with silk donated by their larvae.

From the visually biased perspective of humans, the architects and engineers among Tschinkel’s ants seem even more admirable because they work entirely in the darkness of the underground. There are some 15,000 species of ants, and many of them construct elaborate underground nests. But we know so little about their underground architecture – as they say, ‘out of sight, out of mind’.

Not so for Tschinkel, who has made the most detailed and most innovative studies to date on subterranean ant nest architecture. But as he reminds us,

“…what evolved is not the ant nest—which is just hollow space in dirt—but the behavior of the ants that dig the nest. The nest is essentially the product, or “fossil,” of the ants’ behavior.”

So let us digress and focus on the ants’ behaviour and see how that might evolve. We say that natural selection acts on the phenotype, but it is the genotype that does the book-keeping of past selection. Analogously, natural selection acts on the nests, but it is the ant behaviours that do the book-keeping (in addition to the genes that influence the behaviour, of course) – another example of action at a distance.

While ‘genotype’ is the sum-total of all the genes present in an organism, ‘phenotype’ is the external manifestation of the genotype, such as eye colour, height, ability to digest milk, or the ability to sing a species-specific song. Genes interact with the environment to produce the phenotype.

The extended phenotype

In his runaway bestselling debut, The Selfish Gene (1976), Richard Dawkins argued persuasively that the gene’s-eye view is the best way to view natural selection: “if adaptations [by natural selection] are to be treated as for the good of something, that something is the gene”. When I met Dawkins in 1980-1981, he seemed more obsessed with the phenotype, having just finished writing his second book, The Extended Phenotype (1982). In the preface to the later Oxford paperback edition (1989), with an altered subtitle, Dawkins wrote, “I suppose most scientists—most authors—have one piece of work of which they would say: It doesn’t matter if you never read anything else of mine, please at least read this. For me, it is The Extended Phenotype.”

The altered subtitle of the Oxford edition, The Long Reach of the Gene, says it all. Dawkins had by no means taken his gaze off the gene. Indeed, he had found a new way of singing its praise. His argument now was that the phenotype, the external manifestation of genes, need not be restricted to the organism’s bodies “but may ‘extend’ far outside the body in which the gene sits, even reaching deep into the nervous systems of other organisms.”

The reference to “deep into the nervous systems of other organisms” points to the ability of certain parasites to modify the behaviour of their hosts to behave in a manner that benefits the parasite, even if it is to the detriment of the host itself. A well-known example is that of the liver fluke, Dicrocoelium dendriticum. This aptly called ‘brainworm’ makes the ants it infects climb to the top of blades of grass rather than go down into its nest so that it is more likely to be eaten by its next host, which may be a deer.

As Dawkins has argued, the upward moving behaviour of the ant is the phenotype resulting from the genotype of the worm, acting at a distance.

But we need not resort to anything so dramatic to appreciate the idea of the extended phenotype. Webs built by spiders and nests made by birds or ants may also be regarded as their extended phenotype. The webs and nests are the products of the interactions of the builder’s genes and its environment. The form of the web or nest surely affects the Darwinian fitness of its occupant, justifying the metaphor of “the long reach of the gene”.

Animal architecture has long aroused the fascination of naturalists and biologists, so that we have a great deal of detailed information regarding animal architects and engineers. More than 25 years ago, I discovered to my surprise and delight that Karl von Frisch, the decoder of the honey bee dance-language, had written a marvellous book, Animal Architecture (1974). Most recently, I was mesmerised by Mike Hansell’s Built by Animals (2007), which has come on the heels of a string of his more technical books, Animal Architecture and Building Behaviour (1984), Bird Nests and Construction Behaviour (2000) and  Animal Architecture (2005), revealing Hansell’s life-long passion for the subject.

Ant architecture

But Walter Tschinkel’s Ant Architecture breaks fresh ground altogether, taking us to a new frontier by making visible the unseen.

Field biology involves hard work! Walter Tschinkel digging up a large harvester ant nest. Photo: Walter Tschinkel

In researching for this article, I stumbled across Leon Vlieger’s most interesting blog, The Inquisitive Biologist, with this charming introduction:

“Having been fascinated with dinosaurs from a young age… I decided to study biology…at Leiden University. After three very diverse internships involving lizard hunting behaviour, plant programmed cell death, and lion ecology in Cameroon, I was seriously sold on evolutionary biology and continued studying for a PhD at the University of Helsinki, where I graduated in 2010… Realising that scaling the academic pyramid was not all I had hoped for, I have nevertheless remained passionate about communicating scientific knowledge. Currently I work for the world’s largest specialist environmental bookstore, NHBS, in rural Devon, England, where I am responsible for cataloguing all relevant new publications in the fields of wildlife, ecology, and conservation. This exposes me to a wealth of fantastic new books, some of which I review here.”

The Inquisitive Biologist is indeed a great place to get a wealth of reviews and excellent reading recommendations in ecology evolution and behaviour. In reviewing Tschinkel’s books, Vlieger says:

“Making groundbreaking scientific contributions on a shoestring budget has become a challenge in the 21st century. But there are still opportunities. Take American entomologist Walter R. Tschinkel. With little more than scrap metal, homemade portable kilns, and one almighty spade, he has been researching the architecture of ant nests, pouring molten metal into tiny holes in the ground and digging up the resulting casts.”

Making the unseen visible

Tschinkel was obsessed with the idea of making casts of ant nests from the earliest days of his career. He tried various substances to realise his dream, including latex, dental plaster and molten zinc, before settling on molten aluminium as the best compromise.

He rigged up a homemade kiln to melt the metal, using a 75-litre garbage can lined with sand and fire clay, later replaced by heat-resistant blankets, and hooked up to a 12-volt marine deep-cycle battery. The thrill of doing research by trial and error, rigging up ‘quick and dirty’ gadgets with locally available material, can never be matched by ordering expensive, professionally made brands of fancy equipment from the catalogues of well-known companies.

Walter Tschinkel (right) and a colleague pouring molten aluminium into a harvester ant nest. Photo: Walter Tschinkel

After the molten aluminium cools and hardens, Tschinkel would dig a pit on one side and excavate the cast from the side. Often, he had to cast successive portions of large nests and glue them together later. Having perfected his technique of making casts of ant nests using different materials, Tschinkel spent years trying his trick on different ant species. His latest count is 37 species!

Tschinkel has begun to use the resulting database to explore the relationships between nest size, nest architecture, size of the ants, size of their colonies, lifestyle and division of labour. Even more interestingly, he has initiated an exploration of the evolution of nest architecture, and shows how we can eventually construct phylogenetic trees of ants based on their nests.

In doing so, we can trace the evolution of nest-building behaviour more powerfully than by merely studying the behaviours themselves. This is why it is helpful to think of the nests as the ‘fossils’ of behaviour, the tangible manifestations of their behaviours – the extended phenotype.

Walter Tschinkel posing before a plaster cast of a complete harvester ant nest. Photo: Walter Tschinkel

Tschinkel has worked with many ant species, but the fire ant (Solenopsis invicta) and the harvester ant (Pogonomyrmex badius) have been his particular favourites.

The fire ant, also known as the imported red fire ant, is an invasive species in the US, Australia, China and Taiwan. Its presence has been especially devastating in the US, causing $750 million in damages annually and stinging tens of thousands of people, some of them fatally.

Not surprisingly, there is a great deal of interest and justification for studying this species. I have already mentioned that Tschinkel himself has been an important leader in this effort. As we have seen from one of the images above, their nest can be extremely complicated. Tschinkel says:

“Using a pretty low-tech process, I poured a half bucket of plaster slurry into a fire ant nest, dug up a 20 kg hunk of dirt and plaster, and took it home to wash off the soil with a hose to reveal the shape of the empty space. It was an eyebrow raiser, for I hadn’t imagined it correctly at all. In place of the rather random structure I had imagined, the cast revealed a clearly organized structure of repeated vertical shafts connecting horizontal chambers into many distinct vertical series, often with horizontal connections between chambers as well. I hung the cast from the ceiling of my office, where it amazed visitors for three decades.”

Walter Tschinkel’s first plaster cast of a fire ant (Solenopsis invicta) nest. Though incomplete, it gives an idea of the complex architecture. Photo: Walter Tschinkel

The Florida harvester ant (Pogonomyrmex badius) that Tschinkel also studies is a large ant species with highly polymorphic workers and deep underground nests with large granaries, where the ants store the seeds they collect. Upon casting and reconstructing the first complete harvester ant nest (see in the lead image above), Tschinkel says:

“The result, however, was magnificent, revealing an unsuspected beauty and order that made me want to reveal more such beauty in other nests and display it for others to see. There were two large barriers to realizing this desire—first, the daunting amount of work it took to reassemble a large plaster cast, and second, the obvious fragility of the reassembled cast.”

Tschinkel then overcame both barriers by switching from plaster casts to aluminium casts.

What about the builders?

But Tschinkel realised that, beautiful as they are, the shapes and sizes of the nests alone didn’t tell him the whole story. He needed to see how the nest architecture affected the social lives of the ants. He has developed an entirely different and rather tedious technique to get at the other half of the mystery.

Tschinkel has mastered the art of manually digging up ant nests, layer by layer, from top to bottom, repeatedly pausing to trace the outlines of the nest chambers on acetate sheets and collecting and sorting the ants and other coexisting creatures that he finds in each layer. With this additional information, he has been able to reconstruct the social life of the ants as it plays out in their three-dimensional underground nests.

Walter Tschinkel tracing the chambers of a harvester ant nest on acetate paper while digging up a live nest, layer by layer. Photo: Walter Tschinkel

Integrating the information gained by making casts and appreciating the architecture of the nests, and the additional information gained from the layer-by-layer excavation of live nests, Tschinkel has begun to understand the relationship between form and function.

It is probably more accurate to say that he has started to ask the right questions. For example, he finds that the size and complexity of the chambers decrease with the depth of the nest. Now we can wonder why this should be so. No such wonderment was possible before his painstaking work. Research raises more questions than it answers, and that is the beauty of basic science.

By cleverly applying his tricks of casting and excavating nests to harvester ants that have decided to move their nests to a new location, Tschinkel has discovered a great deal more about the structure and function of the nest architecture. He has asked and partly answered such questions as to why, when, where and how they move their nests. And what are the costs and benefits of moving?

By studying the process of digging a new nest from scratch, he has been able to ask such difficult questions as to how they know how deep they are and to test several alternate hypotheses.

Rather than say more, it will be well worth your while to read about his research in his book Ant Architecture, and elsewhere.

Walter Tschinkel is passionate about communicating his research and hobby to a broad audience. He has his own YouTube channel and has made many beautiful videos about his casting and excavation techniques. But I do want to draw attention to Tschinkel’s thoughtful meditation about the beauty of ant architecture:

“Is this beauty important to the ants? How would we possibly know? Science cannot ask such questions. Science postulates that this beauty has a practical purpose … [But] We hardly know how the architecture fulfils particular practical purposes, so the question simply hangs in the air. I am satisfied to let it hang there, though that will not stop me from trying to figure out how the architecture serves the colony that built it. I am aware that science is gradually chipping away at the mystery of beauty, but except for a few frayed edges, the mystery is still largely intact, as the nests of Pogonomyrmex badius attest.”

Making casts of ant nests, as many casts as possible, of as many species as possible, may well be Tschinkel’s hobby – but it is much more. His hobby is in the service of his larger goal to understand the function of nest architecture. If he had not spent an enormous amount of time and energy digging up live nests, he might have made the casts of many more than 37 species, but then he would have been a mere hobbyist. Doing science requires the passion and tenacity of a hobbyist but also the wisdom to pursue additional tasks to reap the benefits of the hobby.

It is in the nature of the collective and cooperative enterprise of science that every researcher inspires others to carry the torch of knowledge and curiosity into the future. Luckily, much new and even more innovative research is now already underway.

To take just one example, the ant researcher Michael Goodisman, from the Georgia Institute of Technology, has teamed up with a group of biologists, physicists, mechanical engineers and electrical engineers in his home institution and elsewhere to understand the behavioural and mechanical determinants of the fire ant’s nest-building by the fire ant.

In laboratory experiments, these researchers have offered the ants substrates composed of silica particles of different sizes and with different moisture content. Then, the researchers monitored their digging behaviour and nest growth using X-ray computed tomography.

In a parallel modelling study, they have shown that the “dynamics of collective motion in social living systems [ants] are consistent with dynamics near a fragile glass transition in inert soft-matter systems.”

In a third study, they specifically addressed the question of how ants deal with the inevitable problem of optimising traffic flow and avoiding clogging during the digging process. Combining mathematical modelling, observation of ant digging behaviour and the behaviour of computer programmed robots, the researchers have made some remarkable discoveries.

Ants seem to adopt a robust strategy to control the formation of clogs by counterintuitive behaviours – for example, some of them remain idle and some of them return without bringing back the loads they were supposed to carry. I hope managers of collective human action learn about the value of such selective unproductive behaviour!

Let us recall Tschinkel’s statement that his efforts have barely scratched the surface and that most of the mysteries of ant nests remain intact. To take this field forward, we need two rather different kinds of efforts. On the one hand, we need very detailed interdisciplinary studies that ask specific questions using a small number of selected species. On the other hand, we also need a large number of simpler studies documenting nest architecture of hundreds of species, spread across different parts of the ant phylogenetic tree and conducted in different parts of the globe.

Neither would be particularly difficult or very expensive, but the latter is especially ideal for “cutting-edge research at trifling cost” for scientists in the developing world.

I am optimistic that Tschinkel’s excellent book and his fascinating videos will inspire many young scientists around the world to help overcome his repeated lament that his “sample of species is far too small and limited to only some group of ants, to make reliable conclusions.”

I am pleased to see that he has already inspired my former student Annagiri Sumana and her students at the Indian Institute of Science Education and Research, Kolkata. PhD students Kushankur Bhattacharyya and Annagiri Sumana have published an impressive paper based on casting 77 naturally occurring nests of the local queenless ponerine ant, Diacamma indicum, deliberately made in different seasons.

Work in progress in Annagiri Sumana’s laboratory. Field assistant Basudev Ghosh melting aluminium (top left), PhD student Kushankur Bhattacharyya pouring molten aluminium into the nest of the local, queenless, ponerine ant (Diacamma indicum, top right) and the resulting cast of the simple nest (below). Photos: Annagiri Sumana

Sumana told me in an email:

“After trying to make wax casts, plaster of paris casts and cement casts with varying degrees of success, we found that aluminium casts were the best for our purpose. After reading Tschinkel’s work, my PhD student Kushankur Bhattacharyya and our field assistant Basudev Ghosh standardized the procedure for melting aluminium in a lead crucible at a temperature around 800º C, which was achieved using a regular mud lined iron stove and hard coke coal. But one shortcoming of the aluminium casts is that we lose information about how many ants live inside the nest. To overcome this, we made wax casts in addition to aluminium casts. As wax casts can be remelted, dead adult ants and pupae can be recovered and counted. The information of how many ants lived in the nest can then be correlated with different aspects of the nest’s architecture.”

I wish many more people follow in Tschinkel’s and Sumana’s footsteps.

Raghavendra Gadagkar is a Department of Science and Technology (DST) Year of Science Chair Professor at the Centre for Ecological Sciences at the Indian Institute of Science, Bengaluru.

Scroll To Top