Eucalyptus histories and futures: Living and dying with hydrological infrastructures

Abstract

The Camden White Gum (Eucalyptus benthamii) is a threatened species of Eucalypt found in the Sydney region, Australia, on the ancestral lands of the Gundungurra and Dharawal peoples. While the species faces a range of threats, in recent years its future has become even more uncertain with a proposal by the state government to raise the wall of the Warragamba Dam, and in so doing drown the largest remaining population found in the Blue Mountains World Heritage Area. A few scattered stands of these trees can also be found elsewhere, growing along the Nepean River in what is now a predominantly rural and suburban area in Western Sydney. But they too are threatened in a range of ways, including by the changing hydrology of this landscape which has impacted on the reproduction of these flood-reliant trees. This article attends to the biology of this tree: its life history and ecological relationships with the wider landscape, in particular with the flows of water that are at the heart of both the life and death of this species. Drawing on environmental humanities and STS research on infrastructures and the layered more-than-human histories that produce landscapes, this article explores how past and contemporary settler Australian relationships with water, and its control and regulation, are unravelling possibilities for ongoing life for the species in a way that is likely to only increase in an era of escalating climatic uncertainty.

We made our way along the creek, scrambling over uprooted trees washed down by a recent flood. It had taken us about two hours to hike down into the valley, and a further half an hour moving slowly along the creek, but we had now almost arrived at our destination. As we rounded a bend in the creek, the tall slender trees we saw in front of us had the features we were looking for: pale white trunks with dark brown bark peeling off them in ribbons towards the ground. We, Thom and Emily, had come to this place to find these trees, known as Camden White Gums (Eucalyptus benthamii), one of the roughly 700 species of eucalypts found in Australia. While eucalypts evolved about 60 million years ago, the last 2 million years has been a time of major diversification. As a group, they are notoriously difficult to tell apart, with close observation of fruiting and flowering bodies often necessary for reliable species identification. At this stage though, we had only a passing knowledge of eucalypt identification–and it hasn’t improved a great deal since–so, we’d relied on a map from a vegetation survey carried out almost 30 years earlier to lead us to what we were moderately confident was the last significant wild population of the species, growing along the bank of Kedumba Creek in the Blue Mountains National Park.

At the time of our visit in early 2023, these trees were in danger of being flooded as a result of a proposal from the New South Wales (NSW) State Government to raise the wall of the Warragamba Dam. We’d come on this pilgrimage of sorts to see these trees. Not just to see them as we already had growing in the National Arboretum hundreds of kilometres away, or as we much more readily could in the semi-rural areas around Camden in Western Sydney, but to see them in the closest thing left that approximates their landscape. The landscape in which they had evolved and come to be in relation to the local rivers and creeks, the alluvial soils, other plants and animals, and the wider more-than-human world.

This article explores the history and possible futures of the Camden white gum. In doing so, it argues that to understand this tree is to understand it in its landscape. This is true of all organisms, but it is perhaps particularly the case for plants and others who don’t move around (at least in the same ways as many other creatures). Standing in place, the resources for life have to come to plants or already be in situ. Arguably, this requires us to attend to the landscape and its flows in a different way. But the Camden white gum draws our attention even more closely to a particular place, as it evolved in and with a very confined geographic region, moulding itself in relation to the nuances and specificities of its forms and possibilities for life, while also contributing to the shaping of that place and its community of life. This region is now widely referred to as the Hawkesbury-Nepean Valley. For reasons of historical accident, the river at the heart of this valley is called the Nepean in its southern half, and then, at the cobbled, braided channels at Yarramundi, its name changes to the Hawkesbury for the second part of its travels. These were names given to the river by British colonists in the late eighteenth century. To local Aboriginal people, it is known as Dyarubbin, and it is by that name that we will refer to it (Karskens, 2020: 3–4, 45, 81–6).

Part One of this article situates this species within its more-than-human landscape, exploring its close connection to rivers and floodplains. It tells a story of both emergence and decline: of the species’ evolution and its more recent reduction down to two remnant populations—on the lands of the Gundungurra and Dharawal peoples—as a result of the extractive land and water use practices that came to dominate, and reshape, this region post British colonisation and settlement. In telling this story, we aim to attend to the ways in which landscapes emerge through the sedimentation of historical, more-than-human, processes. These processes are geological, morphological, and ecological, of course, but also thoroughly cultural, political, and economic (Tsing, 2024). Attending to landscape is about holding together the human and the nonhuman, the biotic and the abiotic, the material and the discursive, and doing so across vastly different temporal horizons, from the slow erosion of an ancient river valley to the short-term thinking of elected governments responding to a worsening anthropogenic climate crisis. All of these diverse forces and agencies are layered into one another—albeit in ways that aren’t always neatly linear—to shape the contours of not only the land itself, but of the lives and ways of life that are possible here today.

Building on this ground, Part Two slows down with changes to this river system. This is a story of hydrological transformations and the large-scale construction of dams and weirs. Through these engineering projects, a concerted effort has been made to convert Dyarubbin and its lands into an infrastructure for securing the freshwater and dry land necessary for the agricultural, residential, and industrial projects of the settler state. But the Camden white gum is a flood dependent species. This makes it particularly vulnerable to these changes in river flow; but it also makes the species an invaluable lens through which to view these landscape changes. While alterations to river flows are not the only threat faced by this species, now or in the past, they are a very significant one–and one that cannot readily be rectified or teased apart from larger threatening processes. As we will see, the impacts of these hydrological transformations are far from being contained between the banks of the river: they spill out into a wider landscape, in particular the floodplain forests that relied on these waters.

In examining these processes of hydrological transformation, we aim to explore the multispecies possibilities for life and death that are opened up and foreclosed by shifting flows of water. In taking up this topic, this article contributes to the substantial literatures on how infrastructures of different kinds, including hydrological infrastructures, reorder worlds: how values, priorities, understandings, and broader imaginaries are built into infrastructures, given material force and stability by them, even if always imperfectly and incompletely (Appel et al., 2018; Bowker and Star, 1999a; Hetherington, 2018; Hommes et al., 2022). The article engages in particular with Ashley Carse's notion of “nature as infrastructure” (Carse, 2012). The story told here, however, is both about those bits of ‘nature’—beings, places, and processes—that are subsumed within infrastructural systems, and those that are actively excluded from the new system to become “orphans of infrastructure” (Light, 2007) in the reconfigured landscape. We propose to understand these excluded parts of natural infrastructural systems as “decommissioned natures,” and we explore some of the forms and consequences of their exclusion in this particular case.

Part Three returns to the proposal to raise the Warragamba Dam, the threat that initially drew these trees to our attention. While this proposal has recently been shelved, it continues to haunt these trees and their futures. This dam is situated on the western edge of Dyarubbin's valley and provides roughly 80% of Sydney's drinking water, capturing water from several smaller tributaries of Dyarubbin as they flow down from the Blue Mountains. The proposal to raise the dam wall by roughly 14 metres would allow it to also function as a flood mitigation infrastructure, protecting the increasingly urbanised floodplain below. The added capacity behind the raised wall would ordinarily be left empty so that in times of heavy rainfall it could catch and store floodwaters to be slowly released over the coming weeks. Doing so, however, would inundate up to 1300 hectares of world heritage listed bushland in the Blue Mountains, home to numerous Gundungurra sacred sites as well as threatened species like the Camden white gum and the Regent Honeyeater (Anthochaera phrygia). Importantly, as we will see, serious questions remain about the extent to which this engineering project would actually reduce flood risk. Despite this, until a change of government in early 2023, this proposal looked like a certainty, with the former Premier of NSW adamant that his party would put “people before plants” (Rose and Cox, 2022).

The new Labor government campaigned with a promise to halt this development and since their election have publicly shelved it. We’re hopeful that it will stay that way, but this is far from guaranteed. Certainly, there is still considerable pressure from some corners for the proposal to go ahead, and a future change of government could see this plan back on the table again. With this uncertainty in mind we think it is important to continue to explore this proposal, both because it would have such a far-reaching and profound impact on the Camden white gum and all the other lives and relationships sustained by this place, and because–even as a possibility rather than the inevitability that it seemed to be until so recently–it remains an instructive site for understanding the diverse agencies and forces that shape more-than-human landscapes.

This article was co-authored by three environmental humanities scholars of largely European descent: two historians whose work is focused on the environment, science, and Indigenous connections to Country, and an environmental philosopher/STS scholar whose research focuses on biodiversity and extinction. As such, it draws on a combination of research methods, including both archival research and interviews and participatory research with biologists, conservationists, archaeologists, hydrologists, environmental managers, and the Indigenous communities whose ancestral lands and waters we are discussing here. Drawing these sources together, we aim to not only better understand the contemporary plight of the Camden white gum and the transformation of this place, but to situate these dynamics within the longer more-than-human histories (O’Gorman and Gaynor, 2020) that have produced them. This research is part of an Australian Research Council funded project with a larger multidisciplinary team that is exploring questions of narrative, power, and justice in relation to the Warragamba Dam.

More-than-human landscapes and histories

Dyarubbin rises south-west of Sydney in the Southern Highlands near Goulburn. From here, the river travels north, along the base of the Blue Mountains, skirting the westward edge of greater Sydney. Dyarubbin has many interesting geological features, one of which is the way in which it flows through alternating landscapes of sandstone and shale. In sandstone country, the river has cut narrow, deep, winding, passages, referred to by hydrologists as “incised meanders”. When it moves over shale, however, it spills out into broad floodplains, depositing alluvial material that has created fertile soils. The river alternates, back and forth, being funnelled into narrow gorges and then discharged into wide floodplains, as it makes its way north.

These differing geological conditions allow the river to support a variety of distinctive ecological communities. A botanical survey of this area conducted in 2000 paints a vivid picture of this patchiness: flowing north, the river moves first through the scrubby vegetation found on sandstone, then spills out onto floodplains where it is lined by tall cabbage gum and blue box eucalypts, then back into sandstone scrub, before flowing into another landscape of tall eucalypts reaching up from a dense understory. The cabbage gums pick up again here, and are joined by another tree, the Camden white gum. They are a localised species, and for reasons still unclear, in their natural habitat they reach only as far north as the Nepean Gorge (Benson, 1985: 371; Benson and Howell, 1995: 22).

Sadly, alongside this diversity of ecological communities, a close examination of Dyarubbin's landscapes reveals a story of incredible loss. While the sandstone scrub ecosystems have fared better, having less fertile and “desirable” soils, all of the diverse floodplain-based ecological communities across this vast area have been decimated, reduced to remnant vegetation in patches along the riverbank and the rugged creeks. Today, it's thought that only about 3% of the river flat forests of the Cumberland Plain remain (Howell et al., 1994: 259). As such, while the Camden white gum is amongst the most threatened of the plant species found in this region, it's decline is also emblematic of a much larger landscape-level process of loss.

Dyarubbin's other unusual and compelling feature is that it emerges from the ranges to the south, but then re-enters ranges to the north, at Durumbuluwa, Sackville. From there the river flows north through its sandstone labyrinth, and at Wulumarrang, Wisemans Ferry, it turns abruptly south and east before becoming an estuary that opens into the ocean at Broken Bay. This geomorphology creates a “bathtub effect” that causes Dyarubbin to flood severely. In times of heavy rainfall, water cannot escape the narrows at Durumbuluwa. As a result, it backs up and drowns the whole valley at this point, as well as at other sandstone “choke points” like Wallacia and Castlereagh to the south (discussed further below) (Infrastructure NSW, 2017: 16; Karskens, 2016).

Dyarubbin's floods lie at the heart of the intimate connection between the Camden white gum and this river. All of the remaining, naturally occurring, populations of this species are found growing on alluvial flats where past floods have deposited nutrient-rich silt. When flood waters inundate these areas, Camden white gums are quite happy standing in water for a few weeks at a time—something that will kill some other species. More than this though, they seem to rely on intermittent flooding to bring forth the next generation in significant numbers, what botanists call ‘recruitment’. Like other eucalypts, Camden white gums produce seed which they then unceremoniously drop at their own feet. While fire prepares the conditions for young eucalypts in some Australian landscapes, along these parts of the river it is flooding that clears away the understory, soaks the ground, and deposits a rich layer of nutrients in which seeds can germinate and grow. At the same time, flooding may also play a role in dispersing seeds along the river, ensuring the mixing of populations that prevents inbreeding. Camden white gum and its associates have a reciprocal relationship with the river, too. They help manage salinity, shade the waters, buffer and stabilise the riverbanks and so maintain the channel (Howell et al., 1995: 48).

Aboriginal people have lived with Dyarubbin in these patchy, fertile landscapes for at least 50,000 years. They co-created these places through specific and deliberate cultural burning practices that helped to clear out limited patches in the dense, generally mesophilic riverside forests and rainforests through “cool fires”, especially at the confluences of rivers and creeks, and at stony fords (Karskens, 2019: 404–413). Away from the riverbanks, cultural burning in the open forests on the shale soils was probably more extensive and would have created access for women to dig for yams and fern roots, as well as greenpick for animals (Barrett, 2015; Karskens, 2020: 45; Mathews, 2003: 29). Indeed, in Gundungurra language, “Dyarubbin” is closely related to the word for yam, and yams were plentiful in river flat forests, on river and creek banks and in the lagoons. These areas were also hunting grounds which settlers often called “kangaroo grounds”. Aboriginal people along the river also shaped landscapes through a range of other agricultural and aquacultural practices, all of which encouraged particular animals and plants (Gammage, 2011; Karskens, 2019: 404–415).

Gundungurra and Dharawal people had close knowledge of Camden white gums and ties to them within their own naming and kinship systems. The trees were prominent along the river reaches in their traditional lands. Gundungurra people call them Durrumbyang, and for Dharawal people they are Kai’yeroo (Bodkin, 2000; Russell, 1914: 12). ¹ The name Durrumbyang appears to be linguistically related to both generic and specific names for eucalypts across the wider region as far north as Port Macquarie. ² Knowledge of the trees was continuous, too. Over 120 years after the invasion, Gundungurra man Werriberrie, Willian Russell, recalled that the “big white gums”, Durrumbyang, grew at his birthplace on Werriberrie Creek, near the colonial settlement of The Oaks. In fact, the area was known as Durrumbyang for them: it was their Country, to which they belonged (Russell, 1914: 12). Gundungurra and Dharawal-speaking people may have seen Camden white gum and other tree communities as having links to Country akin to their own. Certainly, Darug and Darkinyung people on Dyarubbin further north also named places for the trees that lived on them. ³

Trees played active and important roles in Dreaming stories. For Gundungurra people, haunted trees lamed people who accidentally trod on their butts. Burl-encrusted trees were hiding places for Dhurramulan, the ‘evil half-spirit of the water’. A shady tree protected the famous Bulan brothers by deflecting the clubs of their enemies. And one tall tree offered the ancestral lizard Bunburang safe refuge from the smoke and flames of a great bushfire. This tree was Dhurrumbee, undoubtedly the Camden white gum (Mathews, 2003: 31, 33, 34, 35, 17–18).

If Gundungurra people knew Dhurrumbee/Durrumbyang intimately, it took colonists over 120 years to accurately identify and name the tree within their own scientific knowledge systems. In the 1850s Sir William Macarthur of Camden Park attempted to bring Australian trees to order, assembling a vast collection of New South Wales timber samples for the Paris and later London Exhibitions, and publishing a gazetteer describing each tree. Macarthur included the trees’ Aboriginal names where he knew them. He must have been familiar with Camden white gums on his vast riverside estate at Camden Park, but the species had not yet been scientifically named, and so he listed it simply as ‘the flooded gum of Camden’. Colonists read trees in much the same way they read men: according to character, bearing and moral worth, practically inseparable from economic worth. Hence, the ‘flooded gum of Camden’ may have been a “fine looking” tree, but it was a “worthless sort; the timber weak and not durable” (Macarthur et al., 1861: 63).

In subsequent decades, the Camden white gum was incorrectly catalogued as Manna gum (E. viminalis) by none other than the famous English botanist George Bentham in his monumental Flora Australiensis (Bentham and von Mueller, 1863–1878: 239–240). Botanist R. H. Cambage mistook a stand of Camden white gums on Werriberrie Creek near The Oaks for the distinctive Camden Woollybutt (E. macarthuri) in 1911 (Cambage, 1911: 553). Only three years later, Cambage and Joseph Maiden published an article which finally named the Camden white gum as Eucalyptus benthamii, evidently honouring the man who misidentified the tree fifty years earlier (Maiden and Cambage, 1914). It was late 1914, the same year that Werriberrie, William Russell, published his reminiscences, including the stories about Durrumbyang on Werriberrie Creek. Had Maiden and Cambage read Werriberrie's account? In their paper they hastily corrected the identity of the Werriberrie Creek trees as Camden white gums. But they did not acknowledge Werriberrie's book or include the name Durrumbyang.

Dyarubbin was first invaded by settlers in 1794, after the arrival of the British First Fleet in Australia. In the intervening 230 years, less than the lifetime of a single Camden white gum, a devastating process of transformation and loss has been unleashed across these river landscapes. With that transformation, the Camden white gum has become an endangered species. Today, there are two main remnant populations left: a semi-rural population scattered along the Nepean River around Camden, Bent's Basin, and Wallacia (Dharawal Country), and the population growing within the National Park along Kedumba Creek that we have already mentioned (Gundungurra Country). These populations have different histories and possible futures, and so in what follows we will briefly discuss them one at a time.

Small groups of settlers began arriving on the river around Camden by 1807 (Atkinson, 1988: 11, 14–17). They cleared the dense forests on the alluvial lands for agriculture to feed the colony. The area around Camden is Dharawal Country, and continued settler expansion on Dharawal people's lands, as well as that of the Gundungurra people to the west, often led to violent confrontations, as had occurred on Dyarubbin further north from 1794. Dharawal and Gundungurra resistance was fierce, particularly in 1814–1816. Governor Lachlan Macquarie responded with a massive military campaign, which resulted in the massacres of at least 14 Aboriginal people at Appin in April 1816. There followed eight months of relentless raids and killings along Dyarubbin. Many of the Aboriginal resistance leaders gave themselves up after an amnesty was declared in November 1816 (Karskens, 2020, chapter 5).

The war on forests and trees continued too. While earlier settlers who occupied land along Dyarubbin from 1794 had cleared their farms gradually, the small settlers around Camden by all accounts quickly cleared the riverbank farms of trees (Atkinson, 1988: 29; Karskens, 2020: 193–195). By the mid-1820s gentleman farmer James Atkinson observed that the ‘greater part of the alluvial lands upon the Hawkesbury and Nepean have been cleared, and are under cultivation’ (Atkinson, 1975: 8; Karskens, 2020: 200).

The trees felled in this process, including Camden white gums, were likely used opportunistically by colonists for furniture, housing, fences, firewood, and so on. ⁴ The English common name for the Camden white gum reflects this first site of encounter between British settlers and these trees. In 1805, a vast area of 5000 acres (2023ha) along the Nepean River was taken by British ex-military officer, politician and grazier John Macarthur and his wife Elizabeth Macarthur. They named it Camden Park, after their patron, Lord Camden (Atkinson, 1988: 16–19). Interestingly, cattle had made their way to this area long before British colonists. The legendary escape of five Afrikaner cattle from the camp on Sydney Cove in 1788, and their rediscovery seven years later as a fat, sleek herd on the Nepean at the ‘Cow Pastures’ led to the realisation that cattle could survive and thrive on their own in the bush. This knowledge ultimately shifted the colony's future away from the original agrarian vision and towards grazing (Karskens, 2020: 87–88; Thompson, 1990: 28ff).

But the vast monopolisation at Camden Park also allowed trees some respite from the relentless clearing on the smaller farms. John Macarthur's horticulturally inclined son William was an admirer of eucalypts, albeit largely for their economic uses, and he conserved some native trees on the estate, as well as sections of riverflat forest on its long river frontage (Benson and Howell, 1995: 73–74; Macarthur et al., 1861). ⁵ By contrast, clearing of both original and regrowth forests on Dyarubbin's alluvial banks intensified over the nineteenth century, especially in the wake of immigration booms from the 1830s and 1840s onwards. Werriberrie wrote of the clearing of Durrumbyang on an estate called Victoria Park in his birth Country on Werriberrie Creek. He even recalled who did the work: a man named “Parramatta Jack did most of the clearing of the paddock” (Russell, 1914: 12). By the 1890s, this estate was probably typical: it was almost all cleared, apart from corridors of remnant forest used as windbreaks (Anonymous, 1897).

Despite decades of clearing, reduced populations of Camden white gums continued to occupy their Nepean homelands over the nineteenth century, where ongoing cycles of floods and freshes allowed them to reproduce. They escaped the fate of sought-after timbers like red cedar, which had almost disappeared from the area by the 1860s. ⁶ Camden white gums were scattered on the banks or grew in the rugged and more inaccessible places that farmers avoided, like Bent's Basin, or they managed to regenerate in riverside places which later became local reserves and parks (Benson, 1985; Douglas and Scott, 2022).

But, in the early twentieth century, another significant threat to these trees emerged. While much of the land on which they grew had already been claimed for agriculture, in this period the rivers that they relied on were increasingly dammed to meet the needs of the growing population of Sydney. Drinking water supply had been a constant concern for successive governments from the late nineteenth century, as indeed it continues to be today. Water engineering projects like levies and weirs were used to address this issue from that time. Beginning in the early twentieth century a series of dams were constructed in the Upper Nepean, culminating in the Nepean Dam, opened in 1935. These dams radically altered the flood regimes of the river. The impacts of this change for Camden white gums seem not to have been appreciated, let alone studied, until many decades later (Benson, 1985).

Today, only a few hundred Camden white gums survive in this region in small, scattered populations along the Nepean River as it passes through parks and industrial and agricultural areas. Even more concerning than their low numbers, however, is the fact that they are almost entirely older trees. As the recent Conservation Assessment of the species by the NSW Threatened Species Committee put it, “throughout this area, recruitment of the species is poor or absent, and where present, significantly compromised by a range of threatening processes” (Douglas and Scott, 2022: 4–5). Those threatening processes are manifold, including the widespread presence of a range of weeds that outcompete young trees, and of deer, goats, and other animals that graze them down. But amongst those threats, and connected to them in complex ways, is also the changed flood regimes of the river. The committee goes on to note that poor recruitment is “likely a result of less frequent flooding,” that might not only deposit silt but help to suppress weeds and other competing vegetation (Douglas and Scott, 2022: 13).

As you travel west from the Nepean River, the land begins to rise, slowly at first and then all at once as you enter the region known as the Blue Mountains. Here, on Gundungurra Country, at Kedumba Creek which lies 50 kilometres west of Camden, the second significant remaining population of Camden white gums is found. Many trees in this area survived the initial period of land clearing for agriculture. But they were not to escape the growing demand for new sources of drinking water in the twentieth century. In the 1940s, the Burragorang Valley, through which runs a tributary of Dyarubbin, was confirmed as a site for a major water supply reservoir. The Warragamba Dam was built between 1948 and 1960.

At that time, there was a small agricultural township in the valley that had been established in the 1820s and included European and Aboriginal families (Smith, 2016). Some of the Gundungurra families established farms on Aboriginal reserves in the valley. By the late nineteenth century, however, many of these families moved or were pushed out of the valley, with some moving to ‘The Gully’ in Katoomba in the Blue Mountains where they maintained strong connections with their ancestral lands (Smith, 2016: 100–07, 180–81, 197, 209). When the dam was being built, some members of the Gundungurra families who had stayed opted to join the labour force working for the Water Board on the dam project, at least in part so that they might stay on Country as long as possible (Brown et al., 2025).

Much of the valley floor had already been cleared for agricultural use when construction began on the dam. Before the valley could be filled with water, however, workers were hired by the Water Board to remove remaining vegetation to the new waterline. It was dangerous and underpaid work – in the steepest parts the men were “lowered into the … gorge on ropes and had to chop trees with one hand, holding onto the rope with the other” (Smith, 1991: 9). Any trees not cleared would simply have been drowned, along with a number of significant Gundungurra sites and old farms and buildings, as the newly forming Lake Burragorang filled with water (Jackson, 2019: 241; Smith, 2019: 136–141). The wider valley had been sacrificed for the water that was viewed as an essential part of the ongoing growth and economic prosperity of the metropolitan city of Sydney. As historians Margaret Cook and Peter Spearritt summed the situation up: “Rampant Sydney suburban development in this era [from the 1960s], egged on by builders, land subdividers and councils eager for ever more rate income, knew no bounds. And the dam stood as a continuing symbol of how Sydney could grow and grow” (Cook and Spearritt, 2021: 216).

Today, this valley and the surrounding parts of the Blue Mountains are a protected catchment area. It is here, deep within this protected area, that the last major population of Camden white gums can still be found. These trees grow in the floodplain along the banks of Kedumba Creek, the waters from which feed into Lake Burragorang, which was formed by Warragamba Dam. To date there has not been a thorough survey of the trees, so their precise location and numbers remain unknown, with estimates varying between a few thousand and up to about ten thousand trees (Douglas and Scott, 2022: 3). As we have seen, however, the future of these trees is now also somewhat uncertain, with many—perhaps all—of them lying within the territory that would be inundated if the Warragamba Dam wall was raised.

Infrastructures of hydrological transformation

The decline of the Camden white gum over the course of the last two hundred years cannot be understood outside of the larger story of the hydrological transformation of this landscape. Over this period, Dyarubbin has been thoroughly transformed by its enrolment into projects of colonial settlement, wealth production, and nation building. Thinking with Ashley Carse, we might even say that this landscape has become an ‘infrastructure’ for these projects. Some features of this place are clearly infrastructural in the traditional sense of this term, most obviously the massive concrete dams that have transformed the flows of rivers to make water available as a resource for domestic, agricultural, and industrial use. But in an important sense, the wider landscape has itself also been subsumed within this infrastructural project. Carse encourages us to pay attention to the many ways in which nature becomes infrastructure as environments are “managed for the purposes of service delivery” (Carse, 2012: 544). Of course, diverse landscapes and nonhumans frequently make vital contributions to the healthy functioning of infrastructures, but Carse draws our attention here to those sites in which they are actively managed to make these contributions.

As we have seen, one vast area of land was drowned to become Lake Burragorang. This reservoir has a capacity of over 2000 gigalitres, enough fresh water to generally supply roughly 80% of Sydney's demands. This area is not so much managed as sacrificed—transformed into a giant water tank—for service delivery. But other areas of land have been rendered infrastructural in precisely the way Carse describes, and managed to perform vital functions in this hydrological endeavour. In particular, huge areas of bushland around Warragamba Dam and the other dams in the ‘integrated catchment system’ that provides Sydney's drinking water have been designated as ‘Special Areas’ and are now managed by WaterNSW to protect water quality. Access to these areas is either highly restricted or prohibited altogether. A total of 364,000 hectares of “mostly unspoiled native bushland” is designated in this way in the greater Sydney Catchment Area – more than half of which lies around Lake Burragorang (WaterNSW, 2023a). When the dam was built, the important role that the forest played in this infrastructure was recognised, and rather than being simply taken for granted—an invisible contributor—the forest came to be actively managed to ensure its continued contributions.

Of course, this active management takes place within particular scientific and broader cultural paradigms that shape and constrain what will count as nature, with significant consequences for all involved. This is perhaps nowhere as clear as it is in the way in which Gundungurra people have been severed from many of their sacred sites in the name of ‘protecting nature’. While many Gundungurra sacred sites were drowned by the building of the dam, others survived but now lie within the restricted Special Area. As a result, they too have effectively been lost to the community, in line with the inflexible, universalising, Western logic of Nature set apart from Culture, according to which protection of the former is best achieved by the exclusion of the latter. The ramifications for Gundungurra people are significant, including ruptures in intergenerational knowledge sharing for which access to these areas is essential. As Gundungurra woman Kazan Brown noted when reflecting on the consequences of this situation in an interview: “I don’t think I know enough about my culture, I don’t know how to mix the paint, I don’t know what they mixed it with, whether or not it's a men's site or a woman's site or whether a woman should be out touching things up, or anything like that. So, I wouldn’t even think about doing it [even if it were legally permissible]. And that's part of the legacy of being locked out” (Brown et al., 2025).

Alongside these forested Special Areas, the floodplains are also intimately connected to this infrastructural transformation. For millions of years, these plains have functioned as something like an overflow reservoir in times of abundant rainfall, a key part of the healthy functioning of this river. But today every effort is being made to excise them from this hydrological system: not to allow them to flood. As farming and towns became established on floodplains, these dams and an associated system of weirs sought to control and regulate the flow of water to prevent damaging floods. The end result, according to an environmental assessment, is a river whose waters have been so “severely regulated” as to effectively segment it “into a series of ‘weir lakes’ rather than a freely flowing river” (Morgan, 2014).

Over the past several decades, the infrastructural demands on this landscape have become even more significant, as the suburbs of Sydney have sprawled further and further westward. In the words of the state government department Infrastructure NSW: “The Hawkesbury-Nepean Valley is changing from a semi-rural landscape to an urbanised floodplain” (Infrastructure NSW, 2017: 19). Over 130,000 people now live on this floodplain, with estimates that this number will double in the next 30 years. Reporting these facts as though they have no power to influence the inevitable onward momentum of urban sprawl, Infrastructure NSW note in their flood risk management strategy for the area that: “Over 25,000 residential properties and two million square metres of commercial space are currently subject to flood risk, and this will significantly increase in the coming years” (Infrastructure NSW, 2017: 19). The Insurance Council of Australia estimates the valley to already have the “highest single flood exposure in New South Wales, if not Australia.”

This situation is being further exacerbated by the impacts of climate change which is contributing to the increased frequency and severity of flooding in the region. This fact was driven home with devastating force during the 2020–2023 la Niña event. In this short span of time, parts of this region flooded and reflooded numerous times, causing hundreds of millions of dollars of damage. It is this larger situation that the proposal to raise the wall of the Warragamba Dam claimed to respond to. At present the dam is utilised for water supply, but this increased height would have allowed it to also perform a flood mitigation role. In short, additional land above the wall would be periodically drowned in an effort to keep the areas below the wall reliably dry.

In this effort to disconnect the floodplain from the flows of this river system, these plains become a kind of ‘decommissioned nature.’ In contrast to Carse's helpful notion of ‘nature as infrastructure,’ these decommissioned natures are the bits of the world that were once part of a natural system but that are actively excluded from the infrastructuralisation of that system (we might also think of a mangrove behind a seawall, or a river that's waters have been diverted into a stormwater drain). Importantly, this exclusion does not necessarily equate to the safeguarding or preservation of these places. Instead, as in this case, it can be a severing that comes with its own damaging consequences, and one that may very well be motivated by the desire to incorporate the place into an alternative, more profitable, system or mode of use (i.e., floodplain development). This kind of decommissioning is only possible within a context in which a dynamic river system has been (imperfectly) reduced to an infrastructure for the production of fresh water and dry land. In this way, specific parts of the system are managed and stabilised to perform their specific functional roles in the infrastructure.

In our effort to attend to the ways in which nature is made infrastructural or decommissioned, there is an implicit appeal to systems-thinking. To understand a floodplain as a backup reservoir is to refer to a functional role it plays within a larger river system. Foregrounding the systemic nature of this relationship is an essential part of understanding and responding to the significant impacts that arise from its breakdown, from the exclusion of watery flows from these places. We need systems-thinking to do this work. ⁷ This does not mean, however, that we need to embrace understandings—and certainly not modes of management—in which everything is reduced to its function in a given system. Instead, what is needed are modes of knowing, describing, and inhabiting complex landscapes and ecosystems (not to mention all sorts of other systems) that acknowledge that all of their ‘components’ always exceed their allotted function(s): they have other lives, other meanings, other relationships and significances. ⁸

The decline of the Camden white gum is intimately tied to the infrastructural transformation of the rivers and lands it relies on. While this is far from being the only threat the species faces, it is a fundamental challenge. Today, the ambiguity and shifting watery possibilities have been wrung out of the land. For the most part, the watery places where these trees once were found, have today either been drowned or dried. Each place has been required to become one thing or the other. To have one human use, or the other. In such a world, there is simply no room for a species that evolved with, and relies on, intermittent flooding and its dynamic spaces of ambivalence and change. These species, like the Camden white gum, might be understood in Leigh Star's terms as ‘orphans of infrastructure,’ living beings who do not fit in the world that is being imagined and created by the infrastructures in whose presence they must somehow endure, if they are to survive at all. ⁹

Importantly, this process of orphaning is about more than the loss of a place. As we are seeing, it is about the alteration of a system of relationships, such that life might become impossible, or greatly diminished, even while a particular patch of land remains in some senses ‘available’ to the trees. The anthropologist Franz Krause is helpful here in his effort to understand the land-water nexus as a spatiotemporal phenomenon. Krause (2022) notes that places are not either wet or dry, and it is not enough to simply acknowledge that there are places that are a bit of both in some way. Instead, what is required is fine-tuned attention to the temporalities of wetness and dryness in particular places: to the shifting patterns and configurations of “water, mud and dry ground.” Krause's emphasis is on how human communities, their practices and possibilities, are bound up in these hydrological and geological processes, but there is ample room in his approach for attending to the way in which the lives of trees, animals, and various other nonhumans are caught up in—partially constituted by and constitutive of—these watery entanglements. Here, as we are seeing, there can be life and death consequences to particular rhythms of watery movement, and it may not be enough to simply safeguard a particular ‘place’: we might instead, or as well, be required to learn to read and protect (or in some cases even alter or restore) the particular temporal flows of water that enable this place to support lives in the way that it does, or once did.

Over the past several decades there has been a growing awareness of this need along Dyarubbin, as there has been in other parts of the country. Environmental groups, Aboriginal communities, ecologists, and others have championed the importance of ‘natural’ variability and fluctuation in rivers, leading to calls for the creation of ‘environmental flows’. In the upper Dyarubbin (Nepean), the first of these flows took place in 2001 (Growns, 2016). While there have been some efforts since this time to create regular, substantial, environmental flows that would replicate something like the natural variability of the river, it is generally agreed that the outcome to date has been thoroughly insufficient (Warner, 2014).

But the problem is actually even more complex and insidious than this. As lands along Dyarubbin (Nepean) have been cleared for agriculture, they have also been chemically transformed. One of the key culprits is superphosphate, a high phosphate fertiliser that began to be extensively used in this region in the mid twentieth century. Use on the river flats to produce lush grass for grazing cows and sheep has been particularly prevalent. Today, it is also widely used in the lawn farms that supply growing suburban developments. From here, of course, it readily washes down into rivers and creeks. Foremost amongst the plants that superphosphate helps to thrive and proliferate are the woody weeds that now dominate the riverbank. With this situation in mind, one environmental manager we spoke to worried that providing more regular, moderate, flood waters would simply expand the range of the weeds that are choking out Camden white gum saplings. As he put it, “we could kill the Camden white gums by fixing the floods and creating more regular inundation.” ¹⁰ As such, any helpful environmental flows would need to be very carefully calibrated, both in their source and for their likely contaminants based on recent weather and fertilizer and pesticide use. Sadly, the takeaway point is that it might be even harder than it seems at first to get these trees the water they need. Doing so would require much more fundamental changes to the way that people currently live with and generate profit from this watery landscape.

While the story that we have told in this section is one in which the river and its lands have been increasingly segmented and controlled, it is important to note that this has been–and will continue to be–an engineering dream that cannot ever be perfectly realised. Sometimes the flooded lands behind the dam wall begin to dry out, and Sydneysiders are called upon to limit their water consumption while the controversial desalination plants swing into action to try to cover some of the shortfall. In other cases, the floodplain lands that are supposed to now be dry, lands that people have built homes and businesses on, are reclaimed by the water, sometimes with disastrous consequences. But these moments of lost control also don’t necessarily equate to opportunities for species like the Camden white gum. The massive flooding that has occurred in this area in recent years is not the kind of flooding that regenerates. Instead, these large, and increasingly frequent, floods, tend to uproot, drown, and kill trees. All floods are not equal, from the perspective of a Camden white gum. Instead, the species can only survive and thrive within certain flood regimes, characterised by particular frequencies and intensities of flooding. There is some flexibility in what they can accommodate, but once things are pushed beyond that limit—in either direction—their numbers begin to decline.

Anthropocene modernisers: the proposal to raise the dam wall

The proposal to raise the wall of the Warragamba Dam—to add a flood mitigation capacity to this water supply dam—is the latest instalment in this long running effort to make this watery landscape more predictable and amenable to particular kinds of human use. As noted above, with the change to a Labor government in 2023, this proposal has now been shelved. Nonetheless, it continues to provide an important lens into the complex and consequential entanglement of the Camden white gum in this shifting landscape. What's more, there remains a strong possibility that this proposal might be resurrected as we continue to head toward an increasingly climate insecure future.

It is hard to read this wall-raising proposal as anything other than a ‘doubling down’ on the modernist projects of mastery that have created many of the challenges now faced in this place. Of course, the history of efforts to order and control the watery landscapes and relationships of Dyarubbin's valley is part of a much larger modernising project. In fact, the dams and the suburban sprawl that are at the heart of this story are both iconic features of the post-war development that is emblematic of modernity. In our current period of increased climatic instability, efforts to stabilise, order and control landscapes are taking on increasingly severe and desperate forms. In Australia, it seems that we face a future in which we swing from one extreme to the other, from drought and fire to storms and floods. The increased insecurity of this situation leads to growing calls to take control, to reinforce defences, to ‘future proof’ the country: we must build higher dam walls to store water for the dry times and hold back floods in the wet times.

Even in the context of generalised climate anxieties, however, flood risk in this particular valley looms large as a current and future problem. As described above, the landscape of alternating sandstone gorges and floodplains found here has enabled the larger mosaic of forest and scrub ecosystems within which the Camden white gum evolved, a resident of specific silt-rich plains. But this geology also produces extreme flood risk: it means that at various points along the river's course, all of the waters that have spilled out onto a wide floodplain must squeeze themselves back into a narrow gorge. And then, at Sackville Gorge, they remain confined in this way for the final long stretch to the ocean. While most river valleys widen as they approach their discharge point, this one does the opposite. The end result is a phenomenon that hydrologists call the ‘bathtub effect’. In times of heavy rain, the river and surrounding floodplains fill rapidly with waters flowing in from a variety of small and large tributaries. But then there is nowhere for that water to go, at least not quickly, so it backs up at the sandstone gorge choke points. As a result, floods happen quickly, and big floods can reach 3–4 times the height of those on otherwise comparable floodplains in NSW (Infrastructure NSW, 2017).

In this context of significant and growing vulnerability, many people, including the former government, sought an engineering solution in the form of a raised dam wall. One of the many tragedies of these kinds of efforts to control landscapes, however, is that they so often fail to achieve their desired outcomes while inflicting significant harm on those caught up in the process. Many expert critics of the Warragamba Dam proposal made precisely this point, arguing that because this valley floods from multiple sources (not all of which are above the Warragamba Dam), a raised wall would be unlikely to solve the problem in any meaningful way (although it would likely reduce and slow the rising of flood waters). At the same time, they worried that this proposal was crowding out other options, including buying back particularly risky land, improving evacuation routes and flood warning systems, and perhaps even lowering the regular water level behind the existing dam wall so that it might serve the same flood mitigation role. Worse still, critics of the plan saw it being used to create a false sense of security that was enabling the government to continue to support growing urban development on highly flood prone land. ¹¹

Unsurprisingly, if this plan were to go ahead, it would most likely have further dire consequences for Camden white gums. Many or all of the trees that we visited at the outset of this article, the Kedumba Valley population, would simply be drowned. The research necessary to understand the scale of this loss hasn’t been done. We don’t even know how many of these trees are growing in the valley, let alone precisely where they are, because this was not included in the environmental impact assessment for the proposal. Even with this information, it would be very difficult to predict future inundation levels across the complex topography of the area, but without this information it is simply impossible to know what the impact would be on this last remaining self-sustaining population of the species.

Despite this lack of both spatial and temporal specificity in their knowledge of the species, in making the case for the increased dam height the former government often deployed strategic spatio-temporal reframings of the situation to downplay potential impacts. To do so, they noted that any inundation of protected lands would be temporary, that a relatively small area of land would be affected, and it would be more than made up for by the creation of “offsets proportionate to the impacts” (WaterNSW, 2023b). In this way, they sought to minimise both the temporal and spatial impact of inundation: the areas won’t be permanently lost, and any temporary loss will be made up for by protecting other areas. The insufficiency of such arrangements for the Camden white gum are thoroughly obvious. These trees either can or cannot survive the ‘temporary’ inundation, and the reality is that we don’t know enough about their needs and locations, let alone the likely scale and period of inundations in the coming years, to know how likely survival would be. And of course, they cannot simply use other lands. Their requirements are thoroughly place specific.

In short, what this framing fails to recognise, or acknowledge, is that for the Camden white gum, this proposal is not just about a little bit of land made inhospitable for a little while. Rather, it is the next iteration of a process in which the whole of Dyarubbin's valley, the entire world of the species, has been systematically transformed and rendered inhospitable, and made so, as far as we have the power to predict, indefinitely. There simply are no other lands in the historical range of the species where the river systems haven’t already been made completely unliveable for them. In such a context, this “little bit” of extra taking is radically significant; a significance that can only really be appreciated when we situate this tree species in the long history of hydrological transformation of this landscape. ¹²

Furthermore, when we do so, it becomes clear that while this might be the most obvious impact of this proposal on the species, it is not necessarily the only one. If, as some critics worry, the raising of the Warragamba Dam–or even the promise of it–fosters further urban development in this area, what will this mean for the urban and rural Nepean population of Camden white gums? It's unlikely that a heightened dam wall would lead to much additional development in Camden and the other main areas where these populations of the species are now found because they all lie upstream of the Warragamba's confluence with the Nepean River. As such, any flood waters held back by the dam don’t impact significantly on these areas which, when they flood, tend to do so with waters coming down from the Upper Nepean. Instead, as one local land manager pointed out to us, the impact on these trees will likely be more indirect. If tens of thousands of additional homes are built in Dyarubbin's valley in the coming decades, they too will need to be supplied with drinking water. While the raising of the Warragamba Dam wall aims to add a flood mitigation capacity—that may or may not work, but either way might spur further development—it does not propose to add any further water storage capacity to the dam to provide for the needs of new development. As such, new development will put even more strain on the existing integrated catchment system that supplies Sydney, ensuring that there is even less, clean, high-quality, water available for the environmental flows that these populations of Camden white gums and the many other species they share these landscapes with need.

Over the past couple of decades there has been some effort to turn around the situation of the Camden white gum along the Nepean River, with a programme of tree planting taking place on government and private land. Sadly, most of the major planting has happened well away from the river and the flood zone, often for the simple reason that there is so little public land adjoining the river. As such, even if these trees do establish themselves, it's unlikely that they can become a self-sustaining population. This situation reminds us that the health of the species cannot be viewed outside of its landscape contexts, reduced to a simple matter of counting trees. Especially not if those trees are unable to produce offspring; if they are effectively, as one eucalypt biologist put it to us in an interview, “a population of the living dead.” ¹³

Coda: storytelling trees

Trees have often been powerful storytellers about whether and how floods move through landscapes. The debris of past floods, lodged high up in the branches, has served as a warning sign to people new to an area. Similarly, trees flattened in and along rivers and floodplains can be read for their messages, the size of the impacted trees and their root systems offering some, albeit tentative, insight into the scale and force of the floodwaters. As one of us, Grace, has shown, from their first arrival in Dyarubbin's valley, colonists have been reading these signs (in combination with other information, like that provided by Aboriginal people, and then their own slowly accumulating historical records) (Karskens, 2016). Trees standing in place can’t run or hide from floods, and so they stick around and provide an invaluable record of the changing landscape.

The Camden white gum has its own stories to tell about watery landscapes. They’re complex stories; not as simple to read as a flattened forest. They are stories that must be pieced together based on a slowly developing knowledge of the species’ life history, distribution, and relationship with its particular lands and waters; the kinds of soils and flood regimes it prefers or can tolerate; the extent to which it relies on flooding, and more. But if we take the time to learn to hear these stories, they surely have much to offer us in understanding how to live well in these places.

To put it simply, these trees announce to us that these are lands of flooding. They have always flooded, and they will always flood. Try as they might, water engineers and urban planners will not be able to “decommission” these floodplains, but the ongoing effort to do so—as well as the stories we tell ourselves about how this has been, or might be, achieved—remain highly consequential, nonetheless. The same geology of alternating sandstone gorges and alluvial floodplains that made this place so hospitable for Camden white gums, also makes it a dangerous and unpredictable place for permanent human dwellings. Perhaps amongst its many other stories and meanings, this particular tree species is a signpost that only some kinds of human inhabitation really work in this place. Perhaps we need to look to what the biology of these trees tells us about this place, and to ask if it is compatible with how we want, or are willing, to live here.

Instead, the history that has unfolded in this valley in recent centuries, and that we seem to be accelerating into at present, is one that is not likely to work out very well for any of us. In the effort to control this landscape and its waters, to render it profitable and inhabitable in the particular ways that appeal to decision makers, we seem to be producing a situation in which a good life on the floodplain is no longer really possible for anyone. Human communities are expanding into increasingly unsafe areas, in growing numbers; while the Camden white gums continue to lose the lands and the variable, clean, water flows that they require.

Perhaps, in this context, the lesson that these trees offer is one of unlearning the dream of mastery and control. This is a difficult line to walk in these Anthropocene times when more and more seems to be getting out of control. But perhaps this is precisely when relinquishing the desire for control—which is always also about relinquishing the illusion of control—is most important. Perhaps another lesson of the Camden white gum is that crafting sustainable ways of living with other species, in landscapes that are always more-than-human achievements, requires us to learn to accept and live with uncertainty, variability, and change.

Highlights

• Landscapes emerge through layered, more-than-human, processes that mix together diverse agencies.

• Endangered species with their specific, and often threatened, habitat requirements, can offer an insightful and vitally important window into these processes of landscape formation and dissolution.

• While nature can become an infrastructure in certain contexts (Carse), it is also possible for parts of the more-than-human world to become ‘decommissioned natures’ through their exclusion from the infrastructuralisation of an ecological system.