(Image: Zennie/Private Media)

This is part two in a series. For the full series, go here.

Nearly 30% of Australian carbon credit units (ACCUs) — purported reductions in or storage of CO2, used to offset emissions from carbon polluters — are derived from what’s called human-induced regeneration (HIR).

The principle behind HIR, and about 37 million ACCUs derived from it, is that vegetation left to regrow back to forest will lock in CO2 for the long term, at a rate that can be reasonably estimated (using a method known as the full carbon accounting model, or FullCAM), for the allocation of credits for that CO2.

And there are essentially two forms of HIR: projects on landscape previously completely cleared (either by farmers for pasture or crops, or by fire), or landscape that has not been cleared but is used for grazing.

No one has any issues around ACCUs derived from regrowth from full clearing: the scientific consensus is that woody vegetation will regrow and, all things being equal, draw down CO2. Credits issued for farmers who would otherwise have maintained that land in a cleared state are genuine increases in sequestered carbon additional to what would have otherwise occurred.

But only around 5% of the area credited under HIR projects — which are supposed to contain even-aged forest regeneration — has been cleared in the past. Most projects relate to land that has not been cleared, but where credits are being generated because farmers are reducing grazing pressure, supposedly enabling it to regenerate back to forest because cattle and sheep are no longer consuming the vegetation.

This is where the dispute over the science starts.

In 2020, the Clean Energy Regulator (CER), which oversees the emissions reduction fund, approached University of Queensland Professor Rod Fensham, a veteran ecologist with expertise in woody vegetation and grazing dynamics in arid lands, to prepare a literature review of how livestock and feral animals can suppress vegetation. With the CER’s agreement, Fensham focused on mulga. Mulga dominates most arid and semi-arid land covered by carbon estimation areas (CEAs) used for carbon credit projects, and because animals eat it, it is the vegetation most likely to be affected by removing animals from CEAs for the purposes of regenerating land into forest.

Fensham found that the scientific literature showed “the dynamics of mulga vegetation are strongly aligned with climate cycles. It grows rapidly in wet times and is prone to mass mortality during droughts. Under some, but not all circumstances, sheep browsing can suppress mulga regeneration, and retard, but not limit the recovery of biomass potential. However, the substantial body of research on mulga suggests that under general land-use practices browsing cannot suppress mulga biomass.“

That is, the primary mechanism for driving the sequestration of carbon in HIR projects makes little difference — it’s the amount of rainfall that’s important. Any regrowth will be accelerated by above-average rainfall, and then mostly die back once drought occurs. The result is the release of sequestered carbon from decaying wood.

But after seeing Fensham’s conclusions, the CER refused to accept, or pay for, his report. In his subsequent submission to the Chubb review, which was commissioned by the Albanese government to examine the integrity of ACCUs in 2022, Fensham wrote:

Mulga recovers and thickens up during average and wet seasons and then collapses during droughts. The ‘management’ component of changes in biomass is minor compared to the ‘climate’ component and is impossible to attribute accurately. The semi-arid Australian landscape is characterised by boom-and-bust cycles and there is little opportunity for carbon sequestration in vegetation that has not been mechanically cleared.

The CER insists that grazing is central to regeneration: in June it released a statement along with the Department of Climate Change claiming “scientific literature supports the view that grazing animals can stop trees reaching a forest, such as a study by Eldridge et al (2016), which found that overall ‘ecosystem structure, function, and composition in rangelands are negatively affected by livestock grazing’.”

Problem is, the author cited by the CER, Professor David Eldridge, said no such thing. In response, Eldridge criticised the CER, saying:

I’m very disappointed the Clean Energy Regulator has been using the results of my research to support the notion that removal of grazing leads to an increase in woody plant cover and density … What we found was that when you remove grazing, you increase the density and cover and structure of the herbaceous material, the material that’s less than about a metre tall. There is nothing in that research that relates to woody plants … The notion that removal of grazing is going to lead to an increase in cover or density of woody plants is just not consistent with the literature.

In fact, Eldridge noted, grazing might actually help woody plants “by removing competition from the grasses — grasses compete with woody plants, and if you can over-graze or graze those grasses out then you’re more likely to get woody plants …”

“On so many levels this is stupid,” Eldridge told Crikey. “It’s nonsensical.” He points out that for most rangeland involved, there’s limited capacity for woody vegetation anyway, and that lack of grazing is of no relevance once plants are high enough to no longer be eaten by cattle or sheep: “We don’t have giraffes in Australia.”

Fensham, Eldridge and other ecologists who see rainfall as crucial aren’t alone. The Australian Academy of Science prepared a report for the Chubb review on the different forms of producing ACCUs. On HIR, it noted that one of the “concerns or limitations” was

difficulty in determining the carbon sequestration attributable to human activity (particularly the removal of grazing) as opposed to rainfall in the regions where most HIR projects occur. Variable patterns in rainfall are the dominant drivers of fluctuations in woody biomass in these systems, with the proportion attributable to human activity small and variable … It is not clear how changes in carbon sequestration in HIR projects can be convincingly differentiated between human and climatic changes.

The report goes on to suggest “restricting new HIR projects to areas with higher rainfall and showing clearer signals of human activity”.

In another submission to the Chubb review, the respected Wentworth Group of Concerned Scientists wrote that while HIR worked for cleared land:

Most HIR projects have, however, been directed to arid and semi-arid regions where vegetation has never been cleared. In these boom or bust systems, rainfall is the key driver of vegetation change, and drives both increases and decreases in biomass. While reducing grazing pressure can result in increased tree and shrub cover in these landscapes, from a carbon sequestration perspective this effect is small relative to cyclical climatic drivers. It is difficult to disentangle the influence of the project activity on carbon stocks from other drivers including rainfall. If causation cannot be confirmed (i.e. that the project activity directly results in increased carbon stocks) and additionality can’t be guaranteed (i.e. reliable estimation of the amount of increased carbon due to project activity alone), then it is not possible to establish credibility of the method. 

A large team of ANU and UNSW experts with extensive experience of carbon markets made a similar submission.

The scientific view is strong and widely shared: rainfall drives regeneration across 95% of the area credited under HIR projects, not human decisions. Only the CER and the Department of Climate Change disagree. And as Australia enters another El Niño event after years of above-average rainfall, the role of rainfall and drought in regrowth in hundreds of HIR projects across the country is about to be tested to the limit.

Ultimately this is a scientific debate between a large number of scientists and a regulator with a vested interest in the current system. But the discussion moves out of the complex world of mulga revegetation and into public policy when it comes to independent verification and accountability — and it turns out that is as lacking as the scientific basis for HIR.

Next: how dodgy carbon credits were allowed into the system despite their lack of credibility.