Excavations

First task of this research project. Catalog the major scientific excavations of large ant nests, the people and institutions who did them, the methods, the headline findings, and every published reference that came out of those programs. This file is the foundation; downstream files (lifecycle, communication, collapse) draw their physical evidence from here.

Executive summary

Three independent research programs over the past 70 years are responsible for almost everything we know about the internal architecture of large ant nests:

1. Jacoby (1935–1955), Brazil — the methodological pioneer. Exposed Atta sexdens rubropilosa nests by direct excavation, then in 1950 became the first to pour cement into a nest to create a solid cast. Every subsequent “concrete cast” excavation descends from this method.
2. Walter R. Tschinkel (1990s–present), Florida State University — formalized the plaster, wax, and molten-metal casting method into a reproducible field technique. Cast hundreds of nests across dozens of species (Pogonomyrmex badius, Solenopsis invicta, Aphaenogaster, Formica, Pheidole, etc.). Synthesized in his 2021 book Ant Architecture.
3. Luiz Carlos Forti and collaborators (1990s–present), UNESP Botucatu, Brazil — scaled the cement-casting method to giant leafcutter nests. Their 10-tonne Atta laevigata concrete cast — the one popularized by the MSN article that prompted this research — produced a nest model 8 m deep, 67 m² in lateral extent, with 1,920 chambers including 238 fungus chambers. A parallel German–Argentine line led by Flavio Roces (Universität Würzburg) studies the closely related Atta vollenweideri in the Chaco, with a focus on ventilation, CO₂ regulation, and digging behavior.

The research is unified by one question: how does a colony with no central planner and no individual that grasps the whole, nonetheless build a structure whose spatial logic — fungus chambers separated from waste chambers, peripheral inflow tunnels feeding a central exhaust chimney, top-heavy chamber distribution — looks like the work of a civil engineer? The Forti and Roces programs frame this explicitly as a question about emergent functional architecture; Tschinkel frames it as the question of whether ants follow species-specific “architectural plans” or are responding to local cues that produce convergent shapes.

The headline excavation that triggered this project — sometimes captioned in popular press as a “hidden underground city” — is not a recent discovery. It is the long-running Atta laevigata program out of UNESP Botucatu, with field work going back to the late 1990s. The MSN article and similar pieces in 2024–2026 are popular-press reframings of work already published in Studies on Neotropical Fauna and Environment (2004) and Insects (2017). What is genuinely scientifically interesting is not the existence of the cast — that is a known result — but the engineering interpretation it permits.

Status of this file. Comprehensive on the three programs above. Gaps to fill in later sessions: (1) Jacoby’s original Portuguese-language papers (only known via secondary citation here); (2) the full Roces/Würzburg publication list; (3) Asian and African leafcutter / fungus-growing ant nest casts (currently unrepresented); (4) primary funding figures from FAPESP and CNPq grant databases.

1. The Brazilian Atta laevigata excavation (the trigger story)

What the popular press is describing

The MSN piece, the LADbible series, and several other 2024–2026 popular-science articles all describe the same field operation: a research team in Brazil pours roughly 10 tonnes of liquid cement into an abandoned giant ant nest, waits ten days for it to set, then excavates the surrounding earth to expose a 3D model of the nest interior. The exposed cast is reported to extend roughly 8 metres deep, span roughly 67 m² laterally, and contain on the order of 1,900 chambers including a few hundred fungus-garden chambers. The research was led by Prof. Luiz Carlos Forti at the Universidade Estadual Paulista “Júlio de Mesquita Filho” (UNESP), Botucatu campus, in the Laboratório de Insetos Sociais-Praga within the Faculdade de Ciências Agronômicas (FCA).

Two things matter that the popular pieces obscure:

• The cast is not new. The Atta laevigata cast and its quantitative description were published in 2004 (Moreira, Forti et al., Studies on Neotropical Fauna and Environment 39:2, pp. 109–116). The dramatic photographs that circulate online have been around for two decades; they go viral periodically as new outlets discover them.
• The colony was already abandoned. This is critical to the ethics and the science. The cement is poured into a dead nest — a colony that had already collapsed for unrelated reasons — so the excavation does not kill millions of ants. The “abandoned” framing in the MSN headline is not a mystery to be solved; it is a precondition for the study.

What was actually found

Quantitatively, Moreira, Forti, et al. (2004) studied three nests of Atta laevigata in Botucatu, São Paulo. Two were exposed by direct excavation and mapping; the third was cement-moulded for 3D visualization. Headline numbers from the paper and its successors:

Atta laevigata nests are, by these measurements, the deepest within the genus Atta and contain the highest number of chambers so far recorded for the genus. The species has therefore become the de facto model organism for studying scale limits in eusocial nest construction.

Engineering interpretation

The interpretive layer — the part that makes this interesting beyond “wow, big hole” — comes from later analyses by the same group and from comparative work on A. vollenweideri (next section). The two key claims:

• Thermo-regulatory ventilation (the “chimney effect”). Tunnels are not arranged randomly. Upper-central tunnels function as outflow pathways; lower-peripheral tunnels function as inflow. Heat generated by the millions of metabolizing ants and by the fungus gardens drives a convection current up through the central chimney; surface wind across the central turret pulls additional warm/CO₂-laden air out, which draws fresh air in through the peripheral inflows. Roces and colleagues showed for A. vollenweideri that this is a true wind-induced ventilation system — the first such mechanism described for a social insect colony (Kleineidam, Ernst, & Roces, 2001, Naturwissenschaften 88:7).
• Pathogen segregation. Waste chambers are placed at the periphery or bottom of the nest, structurally separated from the fungus gardens. Forti, Andrade et al. (2017) showed for A. capiguara that the total volume of waste chambers is proportional to the total volume of fungus chambers — i.e. the colony scales its sanitation infrastructure to match its production infrastructure. This is functional architecture by any reasonable definition.

People, institutions, funding

• Principal investigator: Prof. Luiz Carlos Forti, UNESP Botucatu, FCA, Departamento de Produção Vegetal.
• Lab: Laboratório de Insetos Sociais-Praga (LISP).
• Key collaborators on these papers: Sandra H. Moreira, Roberto da Silva Camargo, Nadia Caldato, Nilson S. Nagamoto, Carlos A. O. de Matos, José Cola Zanuncio, and others across UNESP campuses and partner institutions in Brazil and Germany.
• Funding: Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) — Forti held grant 301917/2009-4 covering this research line — and Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), which lists multiple Forti grants in its public database (bv.fapesp.br/en/pesquisador/5969). Brazilian leafcutter research is also institutionally relevant to agribusiness pest control because Atta species are major defoliators of eucalyptus plantations — this is the stated applied agenda that secures continued funding alongside the basic-science publications.

Stated thesis / agenda

The published agenda of the Forti program, as readable across the abstracts and discussion sections of the 2004, 2017, and 2020 papers, has three interlocking components:

1. Document. Quantitatively characterize the architecture of giant Atta nests across species (A. laevigata, A. capiguara, A. bisphaerica, A. sexdens rubropilosa) so that comparative claims become possible. This was largely missing before the program began — most prior literature was anecdotal or single-species.
2. Explain. Identify the rules by which workers construct the architecture, given that no individual ant has access to the global plan. Forti’s 2014 work and later collaborations explore how relocating brood and fungus triggers the excavation of new chambers — i.e., the architecture is built by local stigmergic responses to the presence of fungus and brood, not by a global blueprint (Pinto-Tomás et al.; see §5 for the PLOS ONE citation).
3. Apply. Translate the architectural understanding into better leafcutter pest control — specifically, into more effective chemical-bait targeting of fungus chambers and into agronomic practices that disrupt nest expansion.

The second component — emergent architecture from local rules — is the thesis that makes ant nests interesting as a model for the broader research questions in this project (communication, civilizational scaling).

2. The methodological lineage

The “pour something in, dig it out” technique looks simple but took half a century to mature. The lineage is worth recording because it explains why the Atta laevigata cast looks the way it does — it inherits design choices from earlier generations.

Jacoby (1935–1955) — the inventor

The Brazilian entomologist Martin Jacoby is credited as the originator of cement casting for ant nests. His 1955 paper for Atta sexdens rubropilosa is the first systematic exposure of a giant Atta nest interior. He had reported the cement-moulding technique itself in 1950. Jacoby’s work is published in Portuguese-language journals and is now hard to retrieve outside Brazilian libraries, but every modern Atta nest paper opens by citing him. Gap in this dossier: the original Jacoby papers themselves are not yet in our reference list — to be located via UNESP library or SciELO Brasil retrospective digitization.

Tschinkel (1990s–present) — the systematizer

Walter R. Tschinkel, professor emeritus at Florida State University, took the casting technique and turned it into a reproducible scientific method applicable to dozens of ant species. His innovations:

• Dental plaster (Williams & Lofgren 1988; Tschinkel adopted and refined) for nests too small for cement — gives a much higher-resolution cast.
• Molten paraffin wax and aluminium / zinc for nests of intermediate size, with the metal casts being striking enough to enter popular consciousness through Bert Hölldobler’s photography.
• Censusing protocol. Casting alone gives shape; Tschinkel’s protocol additionally requires collecting and counting all the chambers’ contents (workers, brood, seeds, fungus, refuse) to map vertical distribution of colony function.

His most cited single nest study is “The nest architecture of the Florida harvester ant, Pogonomyrmex badius” (Tschinkel, 2004, Journal of Insect Science 4:21), which established the now-standard descriptive vocabulary: shafts (descending, helical, 4–6 cm diameter, 15°–70° angle) and chambers (horizontal, top-heavy distribution). The book Ant Architecture: The Wonder, Beauty, and Science of Underground Nests (Princeton University Press, 2021; PROSE Award winner) is the definitive synthesis.

Forti and Roces (2000s–present) — the scalers

The Forti program in Brazil and the Roces program in Germany–Argentina are the two labs that took the casting method to the giant nests of leafcutter ants, where the cast can weigh 10 tonnes and a single nest can house 5–8 million ants. These are the only nests in the world that can plausibly be described as “underground cities” without exaggeration. Both programs publish actively; both treat the casts as raw data for an architectural-emergence research question rather than as endpoints.

3. Comparative table of major published nest excavations

The picture this table paints: everything we quantitatively know about the inside of a giant ant nest comes from fewer than a dozen labs, and most of it from two — Forti’s at UNESP and Roces’s at Würzburg. This concentration matters; it shapes which species we have data for and which we don’t.

4. The 40-tonne soil-movement figure

Almost every popular article on Atta nests asserts that the colony “moved 40 tonnes of soil” over its lifetime. This figure deserves an honest provenance note because it travels independently of any single excavation.

The original 40-tonne estimate is for Atta sexdens, not A. laevigata. It comes from the older Brazilian literature (frequently cited to Autuori, 1942 / 1947 in the Brazilian Arquivos do Instituto Biológico) and was reaffirmed in Hölldobler & Wilson’s The Ants (1990) and The Superorganism (2009). The arithmetic is: a mature A. sexdens mound averages roughly 40 m³ of excavated material at the surface, and the underground voids add perhaps another 20 m³; converting volume to mass at typical Cerrado soil density yields the round-number 40 tonnes.

The A. laevigata cast is comparable in scale to A. sexdens — possibly larger in depth — but the 40-tonne figure attached to it in popular pieces is borrowed, not independently measured. Treat it as an order-of-magnitude estimate, not a direct measurement of the cast nest.

5. Key references

This is the working bibliography for the excavations file. Citations are organized chronologically within each program. Where DOIs or stable URLs are known, they are included; where only the secondary citation is reliable, that is noted.

Jacoby — methodological origin

• Jacoby, M. (1950). A new method for studying the cuttings of Atta sexdens rubropilosa Forel. Brazilian entomological literature; original Portuguese, citation incomplete pending library retrieval. Gap.
• Jacoby, M. (1955). Estudo sobre a digestão dos sauveiros de Atta sexdens rubropilosa. Brazilian Ministry of Agriculture publications. Gap — full citation pending.

Tschinkel — Florida State University

• Williams, D. F., & Lofgren, C. S. (1988). Nest casting of some ground-dwelling Florida ant species using dental labstone. In Advances in Myrmecology (Trager, ed.), pp. 433–443. Brill.
• Tschinkel, W. R. (2004). The nest architecture of the Florida harvester ant, Pogonomyrmex badius. Journal of Insect Science 4(1):21. https://academic.oup.com/jinsectscience/article/4/1/21/886534
• Tschinkel, W. R. (2017). Lifetime nest excavation and contents of mature colonies of Pogonomyrmex badius. Ecology (the “Scientific Naturalist” series).
• Tschinkel, W. R. (2021). Ant Architecture: The Wonder, Beauty, and Science of Underground Nests. Princeton University Press. ISBN 978-0-691-17931-5.

Forti — UNESP Botucatu

• Moreira, A. A., Forti, L. C., Andrade, A. P. P., Boaretto, M. A. C., & Lopes, J. F. S. (2004). Nest architecture of Atta laevigata (F. Smith, 1858) (Hymenoptera: Formicidae). Studies on Neotropical Fauna and Environment 39(2):109–116. https://www.ingentaconnect.com/content/tandf/snfe/2004/00000039/00000002/art00003
• Forti, L. C., Andrade, A. P. P., Camargo, R. S., Caldato, N., & Moreira, A. A. (2017). Discovering the giant nest architecture of grass-cutting ants, Atta capiguara (Hymenoptera, Formicidae). Insects 8(2):39. https://www.mdpi.com/2075-4450/8/2/39 · PMC: https://pmc.ncbi.nlm.nih.gov/articles/PMC5492053/
• Camargo, R. S., et al. (2020). Nest architecture and colony growth of Atta bisphaerica grass-cutting ants. PubMed: 33137875.
• Pinto-Tomás, A. A., et al. (2014). Nest enlargement in leaf-cutting ants: relocated brood and fungus trigger the excavation of new chambers. PLOS ONE 9(5):e97872. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0097872

Roces — Universität Würzburg

• Kleineidam, C., Ernst, R., & Roces, F. (2001). Wind-induced ventilation of the giant nests of the leaf-cutting ant Atta vollenweideri. Naturwissenschaften 88(7):301–305. https://link.springer.com/article/10.1007/s001140100235 · https://pubmed.ncbi.nlm.nih.gov/11544898/
• Halboth, F., & Roces, F. (2011). The construction of turrets for nest ventilation in the grass-cutting ant Atta vollenweideri: import and assembly of building materials. Journal of Insect Behavior 25:296–310. https://link.springer.com/article/10.1007/s10905-011-9290-8
• Halboth, F., & Roces, F. (2017). The construction of ventilation turrets in Atta vollenweideri leaf-cutting ants: carbon dioxide levels in the nest tunnels, but not airflow or air humidity, influence turret structure. PLOS ONE 12(11):e0188162. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0188162
• Pielström, S., & Roces, F. (2014). Soil moisture and excavation behaviour in the Chaco leaf-cutting ant (Atta vollenweideri): digging performance and prevention of water inflow into the nest. PMC: 3991694.

Other excavations cited above

• Moser, J. C. (2006). Complete excavation and mapping of a Texas leafcutting ant nest. Annals of the Entomological Society of America 99(5):891–897. https://academic.oup.com/aesa/article/99/5/891/12176

Synthesis works

• Hölldobler, B., & Wilson, E. O. (1990). The Ants. Belknap Press of Harvard UP.
• Hölldobler, B., & Wilson, E. O. (2009). The Superorganism. W. W. Norton.

Popular-press articles that cited this work in 2024–2026

• LADbible (2024-08-27, 2024-01-19, 2026-03-12): “Scientists discovered hidden mega-city after pouring tonnes of cement into abandoned ant hill” and reposts.
• MSN / inkl: “Hidden underground ‘city’ built by ants discovered in Brazil after shocking concrete test” (re-aggregated multiple times; the AA22mAnx URL the user shared is one of these).
• ClickPetróleoeGás: “Scientists pour 10 tons of cement into an abandoned anthill in Brazil and reveal an ‘underground city’…”
• medtigo: “Discovery of a Lost Underground Ant City in Brazil”
• YouTube: “Secret Megalopolis of Ants Uncovered” (re-uses the original Forti footage).

6. Open threads to investigate later

These are deliberately not closed off — flagging them here so future sessions can pick up.

• Jacoby’s original Portuguese papers. Need to retrieve from UNESP library / SciELO Brasil retrospective digitization. The methodological lineage is incomplete without them.
• Funding figures. FAPESP and CNPq grant amounts and date ranges are publicly searchable but I have only summary references here. Worth pulling if we want to assess how much money has actually been spent on this research line over 30 years.
• Asian / African fungus-growing ants. The literature is heavily Atta-centric. Macrotermes termite mounds (very different organism, similar engineering question) are the obvious comparison case and are not covered here.
• Turner, J. Scott — The Tinkerer’s Accomplice (Harvard UP, 2007). Cross-species treatment of “extended physiology” and how organisms construct their environments. Should be added when we open the communication and lifecycle files.
• Comparative architecture across the Atta genus. A consolidated table of A. sexdens, A. laevigata, A. capiguara, A. bisphaerica, A. vollenweideri, A. cephalotes with measurements would be a natural next deliverable inside this file.
• The “civilization went extinct” framing. Why this Atta laevigata nest was abandoned is not explained in the published papers — it is taken as a precondition. Worth chasing because the user’s lifecycle and collapse questions both depend on it.