Removing 23,000 Sheep Restored Australia's Arid Wilds — The Before⁄After Doesn't Look Real.

Planet in Balance 🌎
1 follower
1 Video View·Oct 6, 2026
AI Generated·May contain errors
In 2008, a conservation group paid $4 million for 2,164 square kilometers of South Australian
desert. An old sheep station west of Roxby Downs. Rainfall, about 150 millimeters a year.
The first thing they did was take the sheep off. Flocks there had once topped 23,000. Then they
started closing water points and pulling up roads. They made the land less usable than it had been
in 150 years. The theory was that a desert with nothing pulling on it repairs itself.
400 kilometers east, there was good reason to doubt that. A 400 hectare patch of the same
country has been fenced off from sheep since 1925. 100 years of protection. Some of the old
bluebush in there still has not come back. So people watched Bon Bon Station. Year 1, nothing.
Year 3, still nothing. By year 7, there was nothing on that ground a person could point at
and call recovery. For 8 years, the skeptics were right. Then researchers from the University of
Adelaide and Bush Heritage Australia stacked up 22 years of free satellite data and compared the
destocked reserve against the neighbor's still-running stock. A straight line appeared in the middle of
an empty desert. The line was the fence. On the protected side, bare ground had dropped 2.1%
relative to the country around it. Some of the worst ground out there may never come back
without machinery. The animals people most want to see are still missing, for a reason that has
nothing to do with sheep. And the lake at the center of the property fills about once every
10 years. Subscribe, because all three of those are coming. Everything alive out there is built
around one question. Can you hold on long enough to still be alive when the water finally arrives?
Sheep broke that three ways. First, what a sheep eats. It does not graze evenly.
It works down a preference list, and at the top sit the long-lived perennial plants,
deep roots, decades old, holding the soil and banking the seed. Those are the plants that
carry a desert through the dry years. Sheep take the savings account before they touch the income.
Second, what a hoof does. Much of that ground is sealed by a living skin a few millimeters thick,
made of moss, lichen, and bacteria. It holds the dirt down in the wind. It traps water.
It feeds the soil. One hoof breaks it. Growing it back takes years, and the evidence from
deserts elsewhere suggests it can take very much longer than that. Third, the water stops soaking
in. With the perennials gone and the skin broken, rain runs off instead of down. So the boom arrives
and the land cannot use it. Australia worked this out early. South Australia passed a Sand
Drift Act in 1927. In 1936, an ecologist named Francis Ratcliffe rode through this exact district
and wrote a report for the national science body, the CSIR, on soil drift in the arid pastoral
country. Sand moving like water. Fences buried. A Soil Conservation Act followed three years later,
and the whole time the law was pushing the other way. Pastoral leases carried a condition requiring
the holder to run a minimum number of sheep or cattle per square mile. Not a ceiling. A floor.
The ground was blowing away, and resting, it could cost you the lease. Bon Bon was taken up in 1878.
Between 1970 and 1989, it carried an average of about 15,000 sheep. The grazier who sold it was
Paul Blight, and he had already been running the property below what it could carry, on purpose,
so the vegetation could regenerate. When he sold, he wanted it managed sensitively. The man who put
this land into recovery was the man who had been grazing it. $4 million Australian dollars across
2,164 square kilometers works out at roughly $18 a hectare. Australia's National Reserve
Acquisition Program had been averaging about $10.61 a hectare, so Bon Bon was not a bargain
by that measure. It was still country the size of a small nation for the price of a house in Sydney.
The money came from the Commonwealth's National Reserve System Program,
the South Australian government, and Bush Heritage's donors. The split between the three
has never been made public. Then came the part that outlives everyone involved. A conservation
covenant? Permanent. I-U-C-N Category 2. And it is not an island. The reserve links Yelabina
Regional Reserve to Wabma-Kadabu Mound Springs Conservation Park. What recovery actually means?
To read those eight silent years, you have to meet the plant that gives this country its name.
Mulga is a wattle, Acacia anura. Australian Acacia woodland covers around 9 million square kilometers,
close to an eighth of the continent, and Mulga is the tree that runs it. It belongs here. The
plant causing trouble in this story is buffelgrass, brought in deliberately as feed, now spreading
along the road corridors. Mulga has almost no true leaves. What look like leaves are flattened green
stems, thin and rolled and turned to bounce heat away. The branches point up rather than out,
so rain runs down the branches, down the trunk, and into the ground at the base of the tree.
The root is the real animal. A seedling 10 centimeters tall can already have a root
three meters down. It is a legume, like a pea, so it pulls nitrogen from the air and feeds some of
the poorest soil on earth. In a dry spell, it drops its leaves and mulches itself. It feeds bees,
birds, lizards, and mammals. Aboriginal people ground its seed for food and cut its hard timber
into shields, spears, and digging sticks. The word Mulga comes from the name for a long, narrow
shield, and it is slow. A young Mulga in a good run puts on about one meter every 10 years.
Lifespan estimates range from over 50 years to two or three hundred, and that spread probably
tells you nobody has watched a single tree from seed to death. Pastoralists did not hate this
plant. They depended on it. In drought, when the grass finished, Mulga leaf held its feed value,
so graziers lopped branches or pushed whole trees over to get stock at the canopy.
Scrub feeding, they called it. It kept a lot of families going. The damage was underneath. In the
1980s, researchers ran a four-year trial on young Mulga at five levels of sheep grazing, from none
to heavy. The ungrazed trees more than doubled in height, and over nine in ten climbed past about
roughly as high as a sheep can reach. The grazed trees were not meaningfully taller after four
years. Not under heavy grazing, not under the lightest grazing either. That trial was not run
at Bon Bon. Same plant, same animal, so it tells you what to look for, but it is not a measurement
of this property. Sheep did not kill the Mulga. They paused it, generation after generation,
just below the height where it turns into a tree. Which is why recovery here does not start with
green. It starts underneath, with the skin knitting, roots going down before anything goes up,
and seed piling into the soil, waiting for a wet season and low rabbit numbers to land in the same
12 months. In 2024, Retallack and colleagues published the numbers in the Journal of Applied
Ecology. 22 years of satellite fractional cover data, 15 of them with the sheep gone, measured
against still-grazed neighbors. That comparison is what controls for rainfall. A wet year lifts
everybody. Their word for the result was unequivocal. Green plant cover rose about half a
percent. Dry plant cover, dead stalks, brown litter standing where it fell, rose 1.1 percent,
more than twice as much. Bare ground fell 2.1 percent, and the drop sat on the fence line.
The signal held through drought. In a desert, dead plant matter is scaffolding. It breaks wind,
catches seed, shades dirt, and slows water. So the first question is not whether the place has
gone green. It is whether it has stopped being bare. Three things sit beside that result.
Go back to Coonamore, the 100-year-old patch. Some of its old bluebush still has not returned.
And kangaroos and emus were never shut out of it because they jumped the fence. Even the cleanest
experiment in the country has a hole in it. Then there is work by Harby and Brandle across the
South Australian rangelands, removing stalk. Worked in mulga woodland that was in fair shape.
In chenopod shrubland already in poor shape, they found no meaningful difference.
That study was regional, not done at Bon Bon. What it points to is that badly scalded ground
may need machines, seed, and earthworks rather than patience. And the Western Myol,
the other great tree out there, produces new young trees roughly once every 20 to 25 years.
Harvester ants take nearly all its fallen seed within hours. One South Australian study measured
about 99 percent. And they eat it rather than plant it. The researchers propose that a seedling
only gets through when three rare things coincide. Very heavy rain, seed scuffed open by water
sheeting across the ground, and seed buried just deep enough to hide. They publish that as a
hypothesis and it still is one. What is not in doubt is the pattern. Individual Myols are centuries
old and there is a long gap in young ones across the last 100 years, put down variously to sheep,
rabbits, and kangaroos. So $4 million did not buy a recovery. It bought time and the legal right for
the land to spend it. The crust went first, creeping back over dirt that had been bare.
Above it, the grasses. Bush Heritage's scorecard for 2009 to 2019 reports the long-lived grasses
sheep used to hunt down turning up again across the country that was hit hardest.
Soil cover improved, plant condition improved, and young plants came through, which is the part
that counts. A next generation rather than old survivors holding on. The animals showed up as
juveniles, which is what ecologists watch for. Young sanguanas, bearded dragons, emu chicks,
quails, Australian bustards. In 2018, a trapping survey caught 176 animals across 35 species,
and among them was the first record at Bon Bon of a Giles's planegal,
a meat-eating marsupial that weighs about 16 grams. It may well have been there all along,
with nobody catching one. What it proves is that somebody is finally looking properly.
In 2010, a bush blitz survey came back with species new to science, including six new
native bees. On a 150-year-old sheep station, there were animals that did not yet have names.
The tally now stands at 21 native mammal species, 55 reptile species, and 144 bird species,
18 of them threatened. No satellite produced the last piece of evidence.
When the pastoral board visited, Dean Willis, an Antakirinja Matu-Yankunitja jaraman,
described native insects and plants coming back that had not been seen out there for decades.
One cascade has not happened. The small mammals, the diggers, the seed movers,
the animals that built the soil structure of arid Australia have not returned with the grass.
Something is waiting for them. According to the Australian Government's Environment Department,
feral cats alone kill more than 1.5 billion native mammals, birds, reptiles, and frogs
a year in Australia, plus roughly 1.1 billion invertebrates.
They threaten over 200 nationally listed species. Professor Sarah Legge attributes to cats at least
20 of Australia's mammal extinctions over the last 200 years, about two-thirds of the total.
Those are national estimates, not counts taken at Bon Bon. They are the reason the work there
looks like this. Since 2016, 1,865 permanent bait stations, 96 camera traps, sustained
fox reduction inside a 40-square-kilometre core. Rabbit warrens mapped across more than 36% of the
reserve, then ripped and fumigated. AI-assisted tools for targeting cats, 96 cameras, 2,164
square kilometres, one camera per 22 square kilometres. Down the road, Arid Recovery runs
123 square kilometres of fenced country with about 60 square kilometres held free of cats,
foxes, and rabbits. Inside that fence, bilbies, burrowing bettongs, greater stick nest rats,
western barred bandicoots, western quolls. Bon Bon is roughly 17 times larger and it is not fenced.
The bilby and the burrowing bettong there remain an aspiration. They are not back. Which raises
the obvious question, if the plants are the point, why not plant them? Because of the arithmetic,
2,164 square kilometres. Seedlings that need a root metres deep before their first summer.
A germination window that may open twice in a working lifetime. So the trees get no treatment
at all. Nothing planted, nothing watered. What gets treated is everything around them.
Old tracks and surplus water points are decommissioned, so animals stop crowding
where the pipes used to be. Brush is packed into eroding ground and low banks are built to slow
water and hold it in the landscape. Landscape rehydration in a desert, which stops sounding
like a contradiction once you accept that the problem was never the lack of rain. It was the
rain leaving too fast. There is a fire break network and no wildfire since the purchase.
That matters because Mulga is severely fire sensitive and does not reliably come back from
a burn. And there is a standing fight against buffel grass along the highway corridor,
because buffel carries fire into country that never evolved to burn.
Nobody out there is growing trees. They are removing the reasons the trees kept failing.
And then the turn almost nobody sees coming. Take 23,000 sheep off and grazing pressure does not
necessarily end because overabundant native kangaroos can graze a reserve down, much as
livestock did. The research on that has become solid enough that the simple version of this
story no longer holds. Removing the sheep was not the job finished. It was a change of problem.
The traditional custodians here are the Antakirinja Matuyankunitjatjara people.
In May 2011 at Kuba Pedi, the federal court recognized non-exclusive native title over
78,672 square kilometers on a claim first lodged in 1995. It was the first native title settlement
in South Australia to which the Australian government was a party. Set against tens of
thousands of years of custodianship, 150 years of sheep is an episode. Ecologists have a name for
what that does to us. Shifting Baseline Syndrome. Each generation takes the world it inherits as
normal. Drive through arid South Australia, think that is simply what a desert looks like.
And you may be looking at a photograph of an injury and calling it a face.
Only one group of people out there holds a reference image older than the wound. Which is
why Nikki Willis, a young Antakirinja Matuyankunitjatjara woman, recording her elders'
stories on that country, is not a cultural footnote to the ecology. She is part of the instrumentation.
And once in a while the desert shows what it was built for. Near the center of the reserve is a
lake called Lake Puckridge. Bush Heritage describes it as roughly circular and a few kilometers
across. Other sources make it considerably larger. It is dry almost all the time and fills
on average about once a decade. When it does, black-winged stilts arrive. Red-necked avocets.
Waterbirds come out of an empty sky over the ground that has offered them nothing for years.
And for a few weeks the middle of an old sheep station is a wetland. No one rebuilt that. There
was no engineering in it. They took the weight off and waited. That is what recovery means in a
desert. Not being healed. Being able to use the next good year when it comes. For eight years at
Bon Bon it looked like nothing was happening. Everything was. It was moving at the speed a
desert moves. Near Braidwood in New South Wales, a racehorse breeder named Peter Andrews and a
landholder named Tony Coote spent years putting leaky weirs into a creek to push the water back
out into its own floodplain. In 2018, seven months into a drought, Maloon Creek was still
running green. Green enough that the Prime Minister came out to look at it.
That video is on the channel. See you next time.
