‘Almost certain to fail’: the high-risk NSW plan to offset huge new urban growth areas

The Perrottet government plans to use a controversial reconstruction technique to offset up to a quarter of the endangered bushland that it will clear to make way for a massive housing expansion around Sydney. The plan has been branded by experts as "high risk at best, if not almost certain to fail".

The government announced last year it had approved the Cumberland plain conservation plan, a major planning policy that will guide the development of four new urban growth areas and up to 73,000 houses from Penrith to Wilton over 35 years.

Environmental offset requirements are usually met by conserving and restoring existing bushland in other areas to compensate for clearing in the development area. But under the CPCP, up to a quarter of the offset requirements could be met by instead attempting to reconstruct the cleared bushland - including critically endangered Cumberland Plain woodland - virtually from scratch.

David Keith, a professor of botany at the University of New South Wales who has assessed attempts to reconstruct Cumberland Plain woodland in western Sydney for two decades, has told Guardian Australia it is unlikely to be possible.

Keith said reconstruction is difficult to achieve, particularly on farmland where soils have been fertilised and disturbed both physically and with chemicals that inhibit native plant growth.

"The success could only come with an enormous investment sustained over many decades and possibly more than a century," he said. "I'd be thinking investments of up to six-figure sums, even for small areas, over cycles of five years.

"It really is high risk and if you were in a business and looking at an investment like that, you'd be looking elsewhere pretty quickly."

The CPCP estimates that some 1,754 hectares of native vegetation will be cleared by the expansion of suburbs to house Sydney's growing population out to 2056. This includes some of the last remnants of critically endangered Cumberland Plain woodland and shale sandstone transition forest, as well as habitat for Sydney's largest healthy koala population.

The state government has proposed compensating for this destruction through a series of biodiversity offsets made up of a combination of new reserves, purchase of biodiversity credits and other measures such as revegetation to meet an offset requirement of 5,325 hectares.

But serious issues are being raised with a plan to meet up to 25% of the total offset requirement for native vegetation through reconstruction of ecosystems from scratch on cleared lands, including flood-prone grazing land in the Hawkesbury region.

If all of the housing proposed within the CPCP is developed, remnant Cumberland Plain woodland and shale sandstone transition forest will be cleared in Wilton, Appin, Cobbity and the Penrith region. Less than 10% of the Cumberland Plain woodland's original extent remains intact and what remains is highly fragmented.

The plan, which still requires approval from the federal environment minister, Tanya Plibersek, notes there is so little of the ecosystems remaining that it will be "challenging" to meet the necessary offset targets. It claims this is one of the reasons reconstruction will form part of the offset strategy.

Much of this work would take place within three proposed new reserves, including the Georges River koala reserve, which a former NSW public servant has criticised as a "double dip" on earlier conservation commitments.

The other two reserves under "investigation" are the proposed Gulguer reserve in the Wollondilly local government area and the proposed Confluence reserve on Eastern and South creeks in the Hawkesbury region, which can't be developed for housing because it is flood-prone.

In its plan, the government says these two potential reserves, which will be made up of blocks bought from private owners, were prioritised because they will connect other existing nature reserves.

Much of the land targeted in the proposed reserves, particularly in the Confluence investigation area, is highly degraded from past agricultural activities and the application of chemical fertilisers, which remain in the soils.

But the plan says reconstruction of almost 1,000 hectares of ecosystems, including Cumberland Plain woodland and river-flat eucalypt forest, could occur across these two sites.

Campbelltown and Penrith councils questioned this approach, as did some members of a community reference group that was established to advise the government while it was drafting the CPCP. Campbelltown council wrote in its public submission that reconstruction was "known to be high risk, expensive and have high failure rates".

The government's public reports state it was told during public consultation that replanting Cumberland Plain woodland "has a low success rate and should not be relied on to deliver conservation lands".

It was also told the Confluence as a reserve area was not supported because it "does not have high biodiversity value and should not be a priority for conservation".

It's an example of what Keith refers to as the "worthless land syndrome that has played out globally".

"The leftover lands are given to conservation because they're not of value for other uses," he said.

Guardian Australia sent a series of questions to the NSW government. A spokesperson for the planning and environment department said the CPCP was one of the largest strategic conservation plans to be undertaken in Australia.

"The plan will support the delivery of housing, jobs and infrastructure, while protecting important biodiversity - including threatened plants and animals - in Western Sydney," the spokesperson said.

The grassy Cumberland Plain woodland contains hundreds of native plant species.

The trees themselves make up less than 10% of the plant species, which are dominated by ground covers, including grasses and wildflowers such as daisies and lilies.

For 20 years, Keith and a team of scientists have monitored a site at the Western Sydney Parklands at Hoxton Park where 26 plant species found in the Cumberland Plain woodland, including some ground covers, were planted from the early 1990s.

They have tracked the plantings to measure two things: how closely they developed into something resembling remnant Cumberland Plain woodland found in the vicinity of the parklands and how much they diverged from nearby abandoned pastures, which were in similar condition to the parklands site at the start of the project.

Their studies have found the trees have survived well and are beginning to resemble those in the remnant woodlands and the site has grown into something different from the abandoned pastures.

But they have not found evidence the plant community is developing into Cumberland Plain woodland. Keith said the ground covers and some of the herbs and shrubs, in particular, had proven difficult to bring back to the level of diversity seen in Cumberland Plain woodland.

"It's like creating a different kind of ecosystem altogether," he said. "We're creating something but it ain't Cumberland Plain woodland."

Keith said his team's work suggested that reconstruction as an offset for clearing of intact bushland within the CPCP area was "high risk at best, if not almost certain to fail" because it involves trading certain loss for uncertain gains.

There is another method that involves scalping of soils and direct seeding to reconstruct grassy woodlands but its success is contested within the scientific community. In the Cumberland Plain it has been trialled on sites of a few hectares and has not been tested on the scale proposed for offsets in the CPCP.

Trials of the technique in Victoria monitored restoration of grassy woodlands on roadsides over a 13-year period. By 2021 the proportion of native species had increased to 52% at one site and 70% at a second site, according to one study co-authored by Paul Gibson-Roy, one of the restoration ecologists who developed the technique.

Gibson-Roy is optimistic about what this method of reconstruction can achieve but said for it to be realised at a landscape scale, there needed to be "proper financial, planning and policy support" in place.

Martine Maron, a professor of environmental management at the University of Queensland and an environmental offsetting expert, said whether reconstruction could be used to compensate for clearing of woodlands depended on the condition of what was being lost. She said it could work as an offset for degraded ecosystems but there was no evidence it could work as a replacement for intact Cumberland Plain woodland.

"The things in the landscape that are vanishingly rare are good condition, remnant ecosystems and large old trees," she said. "We can't replace them. Those are the things you can't offset".

The CPCP is not explicit about what methods it plans to use to reconstruct ecosystems or how it will try to overcome the documented challenges.

As part of the plan, the government has proposed a 35-year research strategy for which it has committed a small amount of funding - $1.8m - to Western Sydney University for the first four years.

Paul Rymer is the lead researcher on the Cumberland Plain project and was also a member of the community reference group that advised the government in the drafting stages of the CPCP.

He said reconstruction had "identified barriers to success, such as the soil nutrient loads, weed loads and ability to establish a diverse, functioning, resilient ecosystem". Part of his team's research will explore whether and how those barriers could be overcome.

Rymer said standard methods of planting trees had not produced the desired results. He said the direct-seeding method trialled in Victoria showed more promise but its success appeared to depend on the land's past use.

The Wilderness Society's Tim Beshara, who worked for a decade in ecological restoration in the Sydney basin, said the government's plan amounted to an experiment where the destruction of existing woodlands was certain while any future gains were uncertain.

"Because we've lost so much bushland in Australia, we urgently need to be scaling up our ecological reconstruction, but it's nonsense to think an experimental attempt to reseed a grassy woodland makes up for bulldozing any existing remnant," he said.

"The need for widespread ecosystem reconstruction is well established. But on the question of whether a reconstructed ecosystem is or could ever be a legitimate replacement for western Sydney's grassy woodlands, the answer at the moment is clearly no."

The department's spokesperson said the proposed reconstruction work could play an important role in improving the biodiversity of western Sydney.

They said the government had set vegetation benchmarks that would have to be met at the various sites, and if they weren't, the reconstruction efforts would not count towards the offset targets.

"An independent review of the CPCP will be undertaken every five years to ensure it is meeting its objectives," they said.

"Annual monitoring and reporting will also be carried out over the course of the CPCP, to ensure the biodiversity offsets delivered align with any development impacts that may occur."

Wayne Olling, the flora and fauna manager for the Blacktown and District Environment Group, said the decision to allow up to a quarter of the environmental offsets to come from reconstruction was a "poor outcome".

Olling was a member of the community reference group and has previously sounded the alarm over biodiversity offset failures associated with other western Sydney developments, including earlier suburb developments and the airport.

"I've lived in the heart of the Cumberland Plain, played in the bushland as a boy and as an adult I've sought to preserve it," he said.

"To see it become the target of the government's strategy for economic development sickens me."