Australasian Science: Australia's authority on science since 1938

Cathedrals in the Desert

Figure 1. The relatively small size of the trees in the Top End’s savanna environments may not have been able to support termite colonies, leading them to evolve mounds whose size was not restricted by tree diameter. Credit: Jan Sobotnik

Figure 1. The relatively small size of the trees in the Top End’s savanna environments may not have been able to support termite colonies, leading them to evolve mounds whose size was not restricted by tree diameter. Credit: Jan Sobotnik

By Nathan Lo

Termite mounds populate the northern Australian landscape like miniature skyscrapers, yet genetic analyses reveal that the first termites that rafted to our shores originally built their nests in trees. Why did they do this?

In the opening scenes of Stanley Kubrick’s 2001: A Space Odyssey, a 3-metre high monolith mysteriously appears outside a cave, and causes much bewilderment among its proto-human observers. If we fast-forward three million years, the towering termite mounds of northern Australia inspire similar curiosity among those who behold them. Where did they come from? How long have they been here? Why are some much taller than others? How did they evolve? To answer these questions, we are carrying out DNA-based studies on Australian termites and their overseas relatives.

Australia has one of the most diverse range of termite mounds of any continent, and also boasts the tallest mounds on Earth. Known as cathedral mounds, these structures reach up to around 6 metres in height and 2 metres in diameter, and are built by millions of blind, grass-feeding termites (Nasutitermes triodiae). There are no architects, no engineers, no plans, no foremen. The instructions for building the mound are literally in the DNA of the colony’s members. Given a worker N. triodiae stands 3 mm in height, these structures are the human equivalent of more than four Burj Khalifa-sized skyscrapers stacked on top of each other, and can survive for more than 80 years.

Termites and their nests play a key role in northern Australia’s ecosystems, and have earned termites the title of “ecosystem engineers” and “keystone species”. Whereas savanna habitats in other parts of the world have large mammals to process most biomass, termites fulfil this role in northern Australia. In these habitats, termite biomass can reach as high as 20 g/m2, making them the dominant animal group. Their subterranean tunnelling creates soil structure, enhancing aeration, water infiltration and root penetration. Assimilation of nitrogen from the air into protein by the symbiotic bacteria in their guts improves soil fertility. Termites also act as a year-round food source for many other species, from invertebrates to vertebrates. Without them, the northern Australian landscape and its biodiversity would look vastly different.

Termite nests are central to their biology. The queen and king usually mate for life, and must be protected from the outside world. Their offspring, thousands of which can be produced each day in some species, must be fed and raised by their sibling workers. Most offspring will become workers and soldiers but, once a colony is established, a select group each year is raised as winged reproductives. These will fly out of the nest during summertime and attempt to start a new colony of their own with a prince or princess from another nest.

The function of termite mounds therefore includes protection from invertebrate and vertebrate predators, storage of food, and maintenance of a relatively constant internal environment that is conducive to colony growth, also known as “homeostasis”. The latter aspect has been relatively well-studied in Australian magnetic termite mounds, whose north–south orientation is believed to provide favourable temperatures in different areas of the mound throughout the day. Many termite mounds, including those in southern Australia, have internal temperatures in the high twenties, even when ambient temperature is well below that.

Although the function of mounds has been studied in some detail, the evolutionary origins of termite mounds have, until recently, received very little attention. To understand how these mysterious structures evolved, and how long they have been in Australia, we constructed family trees of different termite species, some of which build mounds and others that form different kinds of nests. These include nests inside rotten wood, underground in the soil, inside living trees or on the outside of trees, sometimes up to 10 metres in the air.

To work out these family trees, we sequenced the same DNA region in each species, and compared the sequences with each other. Once we had the trees we could infer how nest types have changed throughout the evolution of the group.

We also used what is known as a “molecular clock” to estimate when various branches appeared in the trees over the past few million years. To do this we used termite fossils of known age, and calculated how fast termite DNA changes. Rates are typically quite slow – in the order of about three changes per 100 DNA nucleotides (C, G, A, T) per million years. We were particularly interested to know if termite mounds have dominated our landscapes prior to Australia’s split from other parts of Gondwana around 55 million years ago, or if they evolved more recently.

We looked at two particular groups of termites. The first is the genus Coptotermes, which contains Australia’s main pest species C. acinaciformis, as well as the northern hemisphere pest C. formosanus, on which billions of dollars each year are spent on damage and control. We showed that, like many other Australian animals, including goannas, Coptotermes colonised Australia from Asia around 12 million years ago. This followed the collision of the continental plates upon which Australia and South-East Asian countries lie.

Like their Asian ancestors, Coptotermes continued to build their nests inside living trees once they arrived in Australia. However, they transitioned to mound-building multiple times independently. Although Coptotermes species are also found in South Africa, South America and Asia, they build mounds only in Australia.

Coptotermes mounds are made with a similar domed structure and are usually 1–2 metres in height. The outside of the mound consists of a hard clay outer wall that can be up to 30 cm thick. Under this, the mound is filled with thick “carton” made from partially digested wood and faecal matter. This carton layer looks a little like the human brain due to its high level of reticulation.

Under the carton, often at around ground level but sometimes lower, is the nest itself. In some species this is the size and shape of a basketball, and consists of a vast network of chambers, tunnels and galleries made of very thin carton material that is like brittle paper.

One case of mound evolution involves C. acinaciformis, which is a living-tree nester in the south-east of its distribution and a mound-builder in northern Australia. We think that the relatively small size of the trees in the Top End’s savanna environments may not have been able to support C. acinaciformis colonies, leading them to evolve mounds whose size was not restricted by tree diameter.

Building a nest in a mound compared with inside a tree has additional potential advantages, including a greater ability to regulate gas exchange in order to maintain homeostasis. This is because a mound is more exposed to air movement through its outer layer than a nest inside a tree, and can therefore take advantage of evaporative cooling. It’s unfortunate that southern C. acinaciformis did not evolve mound-building too, because it would make their detection and eradication much easier when they attack human structures.

The second group of termites we studied were nasute termites, which include the cathedral mound-builders N. triodiae (Fig. 1). Nasute termites get their name from the snouts on their soldiers, which spray chemicals at their enemies or competitors (Fig. 2). We compared the DNA of 42 Australian nasute species with dozens more species from overseas, and found that this kind of termite has colonised Australia on three separate occasions between 10 and 20 million years ago.

For one of these colonisation events we can infer that the last common ancestor of the Australian species constructed arboreal nests on tree branches (Fig. 3) because all of their overseas relatives have this characteristic. These arboreal nesters are often found in coastal areas, including in Central America and Asia.

To get to Australia, their nests may have been swept into the ocean during storm or tsunami events, and have made it to Australian shores due to favourable currents. Once in Australia, these termites continued to form nests in trees, and indeed two species along the eastern coast of Australia still do this. The arboreal nests of one species, Nasutitermes walkeri, can be easily seen when walking along Sydney’s foreshore.

During the period in which nasute termites arrived in Australia, the rainforests that once covered the continent were receding, and dry woodland, scrubland and desert habitats began to take over. In response, the arboreal nasutes that had established themselves in Australia evolved the ability to form their nests on the ground rather than in trees. Drier conditions may have made arboreal nest-building impractical, perhaps due to an inability to maintain the right levels of humidity or temperature in nests up in the trees. The formation of mounds may have allowed better regulation of these environmental variables.


Our analyses indicate that the transition from building nests in the trees to building them on the ground happened on multiple occasions (Fig. 3). In some cases the transition was to underground nesting rather than mound-building.

The evolutionary pathway followed by these termites therefore follows that of humans in some respects. On the basis of fossil bone analysis, ancestral proto-humans that lived six million years ago in Africa are believed to have been tree-dwelling. As was the case for tree-dwelling proto-humans, the descendants of arboreal nesting termites would go on to create large metropolises featuring millions of individuals, with a relative size larger than any structure on Earth.

Our work shows that termite mounds have been a part of our landscape for the past 15 million years or so, and are the result of ancient colonisation from other areas of the world, particularly Asia. We have shown that ancient Australian termites were able to adapt and survive in the face of significant environmental change, which has enabled them to become one of the dominant animals in the north of the country.


Nathan Lo is an Australian Research Council Future Fellow in the School of Life and Environmental Sciences at The University of Sydney.