In 2010, a fifth-generation farming family in western Sydney approached the New South Wales government with an unusual proposal. The Downes family, whose land sat on the urban fringe of one of Australia’s fastest-growing cities, offered their worst paddocks for an experiment with no local precedent. They proposed removing olive trees, planting thousands of native trees, and attempting to restore a woodland ecosystem that had largely disappeared.
At the time, the land was worth millions as residential real estate. Sydney’s expansion had reached their boundary, and only five or six working farms remained in the area out of an earlier 150. The family instead signed a conservation agreement, began removing invasive African olive trees, and planted around 5,000 native trees. Fifteen years later, the property has become one of the most valuable in the region, not because of what was built on it, but because of what was allowed to regrow.
The woodland at the center of the project is the Cumberland Plain woodland, an ecosystem most Sydney residents have never heard of. Before 1788, it covered an estimated 100,000 to 125,000 hectares west of the harbour. A 2009 decision by the New South Wales Scientific Committee described it as an open, grassy, well-lit landscape, with grey box, forest red gum, and narrow-leaved ironbark scattered widely over native grasses, plus threatened understorey plants.
Three Aboriginal groups, the Darug, Dharawal, and Gundungurra, deliberately burned this country for more than 60,000 years to keep it open. The fires were low, intentional, and seasonal. They reduced shrubs, cleared seedlings that would otherwise close the canopy, and kept the grass layer open enough for groundcover species to hold. Without fire, even native shrubs cluster together and shade out the grasses. Fire is what prevented the entire ecosystem from collapsing into dense forest.
When European settlers stopped the burning, the woodland began to decline. Then the trees were cleared. Camden, where the Downes property is located, became the colony’s food bowl. Woodland was converted to fields at a rate of roughly 1.2 percent per year over two centuries, according to calculations based on a 2003 analysis. Only about 9 percent of the original woodland remains intact, with another 14 percent in the form of scattered trees standing alone in paddocks.
Then African olive arrived. It was not a random weed. Its recorded history traces directly to John Macarthur and his sons at Camden Park, who planted it as a hedge around their vineyard in the 1820s. Some of those original plantings, now around 200 years old, are still standing. No one expected the hedge to take over the landscape. For over a century it did not, largely because cattle grazed and trampled the seedlings. When farms began closing in the 1980s and cattle disappeared, the olives lost their only check.
A 2010 state report, based on satellite data, identified more than 1,900 hectares of dense olive in a 721-square-kilometre study area in the southern Cumberland Plain. Across western Sydney, the total exceeded 4,000 hectares. The olive’s main weapon is shade. Measurements published in 2013 found that only 4 percent of light reaches the ground under an olive canopy. Native plant species dropped by 78 percent. The olive itself maintained an 88 percent survival rate under its own shade. One plant thrives while everything else dies beneath the same trees.
There was early concern that olive trees permanently altered soil chemistry, but a 2024 study suggested these effects do not persist once the trees are removed. Shade, rather than chemical legacy, appears to be the main mechanism of harm.
The species that depend on these woodlands are running out of time. The regent honeyeater, once moving through southeastern Australia in flocks of hundreds of thousands, now numbers fewer than 300. The swift parrot has roughly 500 remaining and, at current decline rates, may face extinction by 2031. The glossy black-cockatoo, the Cumberland Plain land snail, and small bats that need tree hollows are all affected. Of 59 woodland bird species tracked across the plain over six decades, 20 have declined, and nearly half could be extinct within decades without habitat protection.
The Downes property traces back to an 1812 land grant. Jeremiah Downes leased it in 1859 and bought it in 1875, with an area of roughly 1,300 hectares. Edgar Downes, the fifth generation, inherited a property under pressure. Dairy deregulation in 2000 collapsed milk prices. The Millennium Drought dried out the country. Coal seam gas companies were granted exploration licences over the land, and the family fought them in court, eventually securing a heritage protection order.
Less well known is that Edgar had already changed his farming methods in 1985, a quarter-century before biodiversity banking existed. He replaced synthetic fertilisers with poultry litter, horse manure, wood shavings, and straw, installed subsurface drip irrigation, and stopped using chemicals entirely. He gave no treatments to cattle. A soil scientist who visited described the soil changing from a clayey texture to a crumbly, cake-like consistency, with doubled water-holding capacity.
So when New South Wales launched its biodiversity banking scheme in 2010, the family was well prepared. The concept is straightforward: landowners agree to permanently protect and manage native vegetation on their property, generating biodiversity credits. Developers who are legally required to offset habitat destruction buy those credits. The Downes family allocated a 23-hectare site as the pilot project, the first test case in the state. They sold their first credits within five months.
The protection takes the form of a Biodiversity Conservation Agreement, a permanent covenant registered on the land title. If the family sells the land, the covenant remains. A portion of the sale price goes into a state-managed fund that pays the landowner annually, in perpetuity, for weed control, pest management, and fire management. The family made it impossible for anyone, including themselves, to reverse the protection.
Removing African olive is difficult. Mature trees are dense, and cutting alone makes things worse because it triggers vigorous regrowth. The successful method is cut-stump treatment: the tree is cut, the fresh surface is painted with herbicide within minutes, and the site is monitored for years. For comparison, the Australian Botanic Garden at Mount Annan, a specialist institution with contractor teams, cleared 17 hectares in one year and still had 25 hectares awaiting funding. The Downes family has done it across 450 hectares on a working farm.
Research published in 2015 examined how to rebuild this woodland from bare ground. The studies concluded that Cumberland Plain woodland is not a forest in the conventional sense but a grassy woodland. The ground layer matters more than the canopy. After years under olives, that ground layer had disappeared. Native grass seeds were absent from the soil in the worst sites. The researchers changed their approach, working from the bottom up: grass first, then shrubs, then trees.
They tested fire alone, direct seeding alone, and both together. Fire alone stimulated germination of some native seeds that had survived 15 years under olives, but shrubs did not establish well. Direct seeding restored grasses and resisted weed invasion. Fire combined with direct seeding outperformed everything, producing more native cover, fewer exotic species, and fewer olive seedlings returning.
The seed used for restoration must be harvested by hand from the remaining patches. The Royal Botanic Gardens at Mount Anan had to create dedicated seed-production areas for Cumberland Plain grasses because, as their website documents, there is not enough wild stock for large-scale restoration. This creates a circular problem: the less remnant vegetation that is protected, the less seed is available to restore what has been cleared.
By 2021, the Downes family reported five active sites covering 140 hectares under permanent agreement, with another 163 hectares in process. They collaborated with external groups to build release pens for kangaroos and wombats at restoration sites. One of the first native shrubs to return was Bursaria spinosa, which can close the canopy and block sunlight from grasses if not controlled by fire, doing much the same thing the olive did. A native plant performing the work of a weed is a paradox that Cumberland Plain restorers discuss regularly. The woodland does not want to become a forest. It evolved under fire with an open canopy and light on the ground. Keeping it from closing into dense scrub is an ongoing problem in itself.
When light returns to the ground, grasses fill the space. Behind them come shrubs, then invertebrates, then small birds. Ecologists watch one indicator in particular. The noisy miner, a native honeyeater, thrives in overly fragmented and open woodlands, a degraded structure that invasive species often leave behind. When it dominates, the system remains damaged. The true sign of recovery is quiet, the small insectivorous and nectar-feeding birds that need an ordered environment. The noisy miner means the system is broken. A quiet woodland means it is recovering.
Some species may have passed the point where habitat restoration alone can help them. The regent honeyeater numbers fewer than 300. When songbird numbers fall that low, juvenile males cannot find older males to learn from. Research published in 2021 found that 12 percent of remaining males, about 18 birds, could only sing the songs of other species. Females avoid males that sing incorrectly. Fewer birds, worse songs, less mating, fewer birds.
In March 2026, researchers reported that they had trained captive-bred males at Taronga Zoo to sing the traditional song using recordings and two wild-born tutors. More than 50 percent of the zoo-bred males now sing correctly. The zoo is the only place on Earth where that song still exists. This is what the threat of extinction looks like up close, not just a number but a culture fading away.
Whether habitat rehabilitation can bring such species back remains uncertain. The University of New South Wales has monitored plantings in Western Sydney Parklands for over 20 years. After 10 years, there was no convergence with the composition of remnant woodland, statistically indistinguishable from retired pasture. After 20 years, plant composition still had not converged. Moths and butterflies showed early signs of recovery, and pollination networks were reassembling. No one yet knows whether the system is slowly converging or settling into a new form. The data are insufficient and may remain so for decades.
The biggest obstacle is physical. Ecologists generally estimate that a eucalypt needs 80 to over 100 years to form a functional nesting hollow, the kind that forms when branches die, hardwood rots, and termites carve inside. That hollow cannot be built. A glossy black-cockatoo or a small bat needs a tree that was standing before anyone who planted it was born. That is the timeline, not seasons or election cycles but generations.
Stand on the edge of a biobank site on a low brown hill, and you see two worlds. On one side, a woodland moving in the right direction. On the other, the Sydney suburbs. Property data for 2026 shows the average house price in Camden Council is over one million dollars. Vacant land sells for around $750,000. A 2024 valuation found industrial land values in Camden rose 20 percent in a single year, driven by the new Western Sydney Airport and metro line.
The Downes family could have subdivided their land. Instead they placed a covenant on it. By April 2021, one single Cumberland Plain woodland credit had reached $33,000, tripling in 18 months. Transactions across the scheme exceeded $100 million. But the same scheme that pays the Downes family also allows developers to clear habitat elsewhere. More than 10,000 hectares have been approved for development, and 1,014 hectares of Cumberland Plain woodland have been licensed for removal, more than 10 percent of what remains. Government documents acknowledge the methodology is flawed, a parliamentary inquiry heard the plan was not effective, and an audit office raised transparency concerns. Reports indicated consultants, rather than farmers, were making substantial profits from credit transactions. The Downes family was not named in any of that.
The gap between the value of a credit and what cattle earn on the same land is wide enough to change incentives in directions no one fully controls. Both things appear to be true at once: the scheme protects real land and enables actual clearing.
The sixth generation has divided the work. Henry farms the land. Claire diversified into filmmaking and a festival called The Secret Garden that ran for 11 years and raised over half a million dollars for charity. Max runs the biodiversity bank. There are another 163 hectares in process. Two businesses on one property, each covering what the other cannot. The Biodiversity Conservation Trust paid nearly $37 million in annual management payments to stewardship sites across New South Wales in the year to mid-2025. That covers weed and pest control, permanent management. But that permanence depends on a fund that depends on a market that depends on a government. If any link weakens over fifty years, the forests protected under agreement become weeds protected under agreement.
What this story demonstrates is that a farmer can be paid to rebuild an ecosystem instead of clearing it. Removing an invasive organism and letting light back in triggers measurable chains of reaction. Economic viability is possible. It also demonstrates that planting trees does not create a forest, that even native plants become a problem without fire, that the same system that funds good work also funds clearing, and that a eucalypt only becomes hollow after it is older than everyone who planted it. A bird may forget its song if its numbers drop too low.
Sometimes the most powerful thing done for a landscape is not dramatic. It is a family noticing what is still alive under the olives and deciding to let the light back in.
