Sir Charles Burrell must have a lot of nerve. I recently went on ‘safari’ at his Knepp Castle ancestral estate in Sussex and the place is seriously neglected. In fact it looks as though he has walked away, shut the door and thrown away the key.

This 3500-acre estate was, until 15 years ago, a mixed arable and dairy farm on the Wealden clay, an hour south of London. However it was only turning a profit two years in ten; that clay, it seems, is difficult to farm being a slippery mess in the wet, and bone hard when it is dry. Charlie Burrell (as he prefers to be called) had always been interested in wildlife so, inspired by a visit to the Dutch rewilding project at Oostvaardersplassen, he decided to try something similar at Knepp. He removed all the internal boundaries on the estate, introduced longhorn cattle, Exmoor ponies, fallow and red deer and Tamworth pigs – as proxy species for our original wild herbivores – and let them loose to feed and roam as they pleased across 2500 acres. On the face of it that was a pretty crazy thing to do with an estate your family had been farming for hundreds of years. I take my hat off to him, not just for nerve but also vision.

I grew up with the notion that the original vegetated landscape of Britain, pretty well everywhere below 2000 feet, was high forest – a kind of majestic Mirkwood with an overarching canopy of big trees. Then about 20 years ago, the Dutch ecologist Frans Vera challenged this model, suggesting that it had underestimated the effect of the large wild herbivores that were common at that time – cattle, horses, deer, pigs and beaver – plus their attendant predators – wolves, lynx and bear. He postulated that the interaction between these animals and their environment would have created something more akin to savannah or parkland with clumps of trees, thickets of scrub and areas of grassland dotted with individual trees. Not all ecologists agree with him and in truth we don’t really know what the original ‘wildwood’ looked like before humans started clearing it 8000 years ago. Finding out is part of what Knepp is about; they call it “process led conservation”.

What confronted me amongst the fine old estate oaks was a mess of sprawling hedges 30 feet wide at the base, big blocks of sallow and large open areas dotted with patches of bramble and scrub – some of which sheltered young oaks from the browsing animals, just as Frans Vera predicted. This landscape is prompted, manipulated and sculpted by cattle, horses, deer and pigs. Some of it is low grade habitat at present, particularly the overgrown fields covered in ragwort and fleabane; but to everyone’s surprise populations of some nationally declining species such as turtle dove, nightingale, cuckoo and purple emperor butterfly are increasing at Knepp. What this landscape has, in effect, is lots of woodland-edge, which would have been plentiful in Vera’s vision of the ‘wildwood’. This may account for these increases, and why so much of our wildlife seems to thrive in the ‘edge’ habitats that have become scarce in intensively farmed landscapes.

Looking at it I was struck by how big it seemed – 2500 acres of unfenced land in lowland England is substantial. To my eye it looked like a transitional landscape on its way to something else and clearly it is still on the move, but perhaps not towards continuous woodland. The staff at Knepp just don’t know and are willing to sit on their hands and wait to find out. One thing has become clear: some areas are, unpredictably, developing differently from others, there is a continuity of process but outcome is unforeseen. It seems that the changing landscape has personality, is wilful. A dynamic process has been let loose by the very act of not acting; the mice have indeed come out to play.

There are many limitations to comparing this project with a truly wild ecosystem, one of which is the absence of large predators to limit the populations of introduced herbivores. There is just not enough room for them at Knepp, and anyway lynx and wolves might be a stretch too far for the neighbours in Sussex. So the estate has to accept the responsibility of being the apex predator and cull them. Animal welfare apart, what is “overgrazing” and “too many” are fascinating questions in a project that claims to have no desired (or undesired?) outcome. How will they react if the numbers of purple emperors or turtle doves start to go down? Hold their nerve and do nothing, I hope.

Another fascinating twist in this story is that Charlie Burrell has certainly not thrown away the key – the estate now regularly makes a profit. This has been achieved in various ways: renting out the farm cottages and redundant agricultural buildings; running a safari and camping enterprise and selling those culled animals as organic free-range meat at premium prices. Also expensive inputs such as fertilizer, herbicide, stock management, and farm machinery are no longer required. In fact Knepp is still a farm, if a rather eccentric one, which qualifies for a good deal of public money: £200,000 a year in Single Farm Payments alone we were told. I doubt they could survive without that, so Brexit is making them understandably nervous.

I hope this imaginative and courageous experiment can continue to thrive. Knepp Castle estate has come up with an innovative version of rewilding suited to lowland England which will be fascinating for naturalists and ecologists to watch unfold over the years. It is already a reservoir from which wildlife can colonise the surrounding countryside and a source of inspiration and research, as well as a haven of wildness for visitors in this very crowded corner of Europe.

‘A breath of fresh air for the spirit’ is what Charlie Burrell calls it.