"I’M not sure we should reveal the location,” says Kate Holl. “What if it attracts loads of people?” That, I tell her, is unlikely. We’ve driven three hours from the central belt, taken a bouncy ride in a small boat to an island, tramped for four miles along roads and tracks, changed into wellies to wade across several hundred yards of sea-scoured mud and shell sand, scrambled up rocks, and fought our way through brambles, blackthorn, heather and sprouting hazel.

We’re now in the cool, damp interior of a wood. It takes time, determination, map-reading and ingenuity to get here, and when you do, it looks to the unaccustomed eye like nothing more than trees.

But for Holl, woodland adviser to government agency Scottish Natural Heritage, the place – a wood-covered tidal island ridge, rearing out of the sea and sand, whose top is no more than 15 metres wide and perhaps a couple of hundred long – is something like a cathedral. Our journey has brought us to a tiny sample of temperate rainforest – the habitat that once cloaked Scotland’s west coast from Argyll to Wester Ross, but which in this pristine state is now vanishingly rare.

Holl shows me the species: the tall, waving birches, the sturdier ash, holly growing in little spriggy bushes, sycamore saplings, and lots of clumps of hazel, growing in tight, straight stands of trunks, ranging from this year’s twigs to the thick vertical tree-stems. In their very old age the stems are bowed and gnarled, dying under the weight of ivy. One set of stems of all ages covers two or three square metres. She explains it is all the same plant: recent studies have found these trees could be thousands of years old.

There are mosses and lichens, fungi and ferns, saplings, flower-stems, ivy and honeysuckle. We’re here in autumn but Holl first visited the place the year before, in May. "It was fantastic,” she says, pointing to ranks of green-fronded plants. “You had a carpet here of wild garlic with beautiful white flowers, thousands of them, mixed in with the bluebells, then in here there was wood sorrel and the geraniums, pink flowers and white flowers. It was stunning.”

For Holl, the sheer richness of what she called the “infill”, the mid-story of shrubs, saplings and brambles and the lower levels of flowers and ferns, is vital. She has worked for SNH for 30 years, preaching to the rest of the organisation about woodland, and maintains a quiet but unstoppable enthusiasm. Here she is in her element: “It’s like a hidden lost world. There are very few places like it left in Scotland.”

That makes this island potentially an important resource in the effort to save a rare and dwindling habitat that is among the richest in the British Isles: Scotland’s Atlantic woodlands, a rainforest with the potential to be every bit as rich and diverse as that of the tropics.

There are historic reasons for the forest’s disappearance, including clearance for agriculture and later felling and management for charcoal and other industrial products, but talk to any conservationist and they will say grazing animals – largely deer – have played a major role. The red deer, which lives on Scotland on mountains and moors, is by nature a woodland animal, whose favourite food is trees.

As landowners have seen less need to manage woodland over the past century, walls and fences have toppled and herbivores have flooded into the woods. The problem is getting worse: it is widely accepted that Scotland has far too many red deer – there are estimated to be around 400,000, compared to less than half that number in the 1960s.

That means such woodland will never recover unless deer impacts are cut down dramatically. How that should be done is open to debate, and Scottish Natural Heritage is working hard to see deer numbers are reduced and deer are better managed, but it’s a complex problem. Holl has studied woodlands elsewhere in Europe, including Norway and the Isle of Wight, after winning a Churchill fellowship. She has seen what a rich deep layer of “infill” means: “There are places with similar climate, soil, and environment to Scotland that have productive woodland as the default ecosystem."

Why don’t we have it here? “It’s simple. It’s because we have higher herbivore numbers than pretty much anywhere else in Europe,” she says.

Holl and her colleagues have developed a tool for assessing the grazing levels in woodland, and she can tell as soon as she walks into a wood how much grazing is going on. On the trip here she instantly dismissed a couple of woods we drove past as “overgrazed”; she admits Scottish holidays with her family can be difficult as she gets irritated at the thin woods and bare hills.

But on our little island she is happy. Since her last visit there have been no incursions by sheep or deer. In fact, she believes it has rarely if ever been grazed, which is the reason why the woodland is in such good condition, with its multiple layers of canopy and its thick dense infill. The lack of deer on the neighbouring big island, the soft, muddy tidal flats, the rocky approach and the tangled, almost impenetrable scrub, guarding the island are likely factors.

As a result it has survived as an example of what our rainforest could be like everywhere, full of earthy smells, dappled light, green fronds, thick fungi and birdsong. Holl wants to see such woods spreading, with their dense undergrowth and biodiversity. But what can be done to promote it, and why should we save it?

Adam Harrison is a tall, well-spoken Englishman with an unruly mop of dark hair, and ruggedly handsome features that remind me of Desperate Dan: there’s a touch of TV gardening expert Monty Don in there too. He worked until recently for the Woodland Trust, and put together the Atlantic Woodland Alliance to campaign for this habitat. It was launched at an event attended by MSPs and representatives from its member groups, including conservation charities, SNH, Forest and Lands Scotland – what used to be the Forestry Commission – and landowners.

Everyone applauded Harrison’s presentation, and listened attentively as forestry experts emphasised the importance of the habitat. As well as grazing animals there is another major problem, non-native invaders, mainly rhododendron ponticum, but also species such as the widely-planted sitka spruce. The “rhoddies” blanket the forest floor and stop new trees getting established: over decades they take over woodlands entirely.

They and the deer numbers need action if the rainforest is to recover, but the challenge will be to find the money and political will. Six months on from that launch, Harrison tells me already Brexit has cast its long shadow over the alliance’s efforts. The uncertainty around funding has limited the ambitious ideas it had for connecting surviving patches of rainforest across Argyll.

One area they are still working on is Ardnamurchan and Morvern. Linked to the mainland by a 13-mile neck of land in the north, from Loch Eil to Loch Ailort, it is what Harrison calls a “defensible peninsula”. Deer could be stringently managed and rhododendron eradicated, as they have been on smaller scale on the isolated Knoydart peninsula. Torridon, with a group of conservation-minded landowners, is another possible.

Among the weapons, he says, in the alliance’s armoury to persuade Government – and the public – that action and funding is necessary, is the idea that getting local people and communities involved in restoring and managing the rainforest helps the communities. Rather than giving an area handouts, locals can be paid to work on management of rainforest, bringing economic benefits. In Knoydart for instance, £250,000 was spent over ten years on rhododendron eradication, effectively one full-time job for someone who would spend their money locally. In a place with around 100 residents, that made a difference.

There are other benefits for communities. “The fact they are working together to solve one problem can give them more strength to solve other problems,” says Harrison. “It might be transport or something completely unconnected with the environment, but it strengthens communities.”

Land managers have to be persuaded of the benefits of protecting woodland from grazing, and that can be hard: part of the value of a sporting estate is based on the number of stags shot each year, and stalking is easier with plenty of deer. Culls are organised in the Highlands by around 45 deer management groups and Richard Cooke chairs their association, the ADMG. He tells me deer should not take all the blame, as wild goats, rabbits and hares all have impacts, but accepts that keeping deer out would help regeneration. Rather than the conservation solution of reducing numbers, he sees fencing as the answer: “If you want to get major change the easy way to do it is to enclose land. Fencing is not popular in some quarters but it’s an easy way to achieve multiple objectives... you’ve got deer tourism on one side of the fence and on the other you have habitat regeneration. It’s not a choice of either or, the two go pretty well together in the right numbers.”

Ministers and MSPs are expected to consider measures to reduce deer numbers soon after a report by a working group is presented, and that could include the compulsory culls that conservation groups favour. Harrison sees this as an opportunity to win politicians over to lower deer densities.

He believes hunting estates should get on board with the idea of fewer deer and richer, more extensive woodland, providing more shelter for those left, because it can deliver a higher yield for stalking: “You get bigger animals, 30% larger, and as many available for shooting even if the overall herd is smaller, because far fewer die each winter. It strikes me as a win-win.”

Another argument for saving these rainforests is that more trees help towards carbon capture, and for landowners can form part of the carbon offsetting economy. But the rhododendron problem is an example of how hard it is to get enough cash to make a difference.

In Glen Creran, north of Oban, Harrison and I meet forestry expert Gordon Gray Stephens, who is leading a project to remove invasive ponticum. In the cool sunlight the woods at Glasdrum Wood National Nature Reserve have a rare beauty, and Gray Stephens shows me another lost world, this time small enough to hold in your hand. Using a hand lens lets me see the detail of the mosses, lichens and other tiny plants which cluster on the gnarled trees.

We peer at a sample of a scapenia liverwort, which looks at first like a moss but on closer inspection has tiny leaves. It is one of the specialised plants of the rainforest, found almost exclusively on Scotland’s west coast. Gray Stevens, a woodland consultant and rainforest enthusiast with a touch of the David Bellamy, says there is an international obligation to look after what is special.

“Britain signed up to the Rio de Janeiro convention on biodiversity and one of the things it was based upon was that each country should look at what was unique to it. The way we persuade people in India to look after tigers that are eating their goats is that we in Western countries, where most of the pressure for conservation is coming from, should also look after the things that are unique to here.”

The Glen Creran scheme has had grants of about £95,000 from SNH, what was the Forestry Commission, and the National Lottery to clear rhododendrons. Removing them is hard work but the knottiest problem isn’t cutting them down: it’s getting all 16 landowners here to agree to get rid of the plants, removing the source of seeds and preventing further infestation.

Last year the Forestry Commission, as it was, admitted that ridding Scotland’s landscape entirely of rhododendron ponticum, regarded as the only way of completely stopping the march of the plant, would cost £40 million a year over ten years. Just £2m a year was being spent, by FCS on its own land and in grants to landowners. FCS sources have indicated to me this isn’t even enough to stop the ponticums spreading further.

It illustrates the gap between what conservationists and the Government’s agencies think is needed to protect habitats such as temperate rainforest, and what’s actually happening on the ground. Harrison has an uphill battle, but he believes the powers that be will respond. “I think there’s a window of opportunity. We have to work out what the cost of this is and what the return is ... what impact will having more woodland have on the carbon budget and climate change? It can be aligned with better economics in rural areas, costing it and making the business case.”

The urgency is growing: studies by the Royal Botanical Gardens in Edinburgh have indicated that, while many specialised species can retreat north, west, and to higher ground as the climate warms, the rare lichens and other species of the Scottish rainforest will have nowhere to go.

So does Harrison expect taxpayers and politicians to spend money to save a lichen? “It might not be as iconic as a tiger or an orang utang but it is just as important in terms of global wealth and biodiversity because they are unique... There’s a lichen that has only been found in hazel woods on the west coast, the white script lichen. It’s the only lichen of its kind in the world.

“If we don’t do anything to make the environment more robust, getting rid of rhododendrons and the like, if we don’t make the remaining woodland better for this wildlife, then we are going to lose it. Given the global importance of these woodlands, Scotland has a responsibility to do something about this. It is the last stronghold.”

Back on our little island, Holl explains that what remains of the west coast rainforest is mainly oak or hazel woods, with little infill and diversity, and its boundaries are contracting because seedlings are browsed out. The oak-dominated forest is a result of intensive management during the start of the industrial revolution when the charcoal they produced was in high demand. For the last 150 years or so they have seen little management and grazing has stripped out the understory. Similarly hazel woods have lost their “infill”, and been left with just single hazel trunks, unable to pass on their community of lichens to the next set of stems on the same rootstock.

But plants and animals from our island and tiny pockets like it could spread into revitalised woodlands if grazing were stopped, to bring back the rich biodiversity that was once widespread across the west coast of Scotland.

When Holl is not enthusing about woodland she’s working on her smallholding near Edinburgh, which includes new woodland, with its own pesky deer problems. And while she believes we need to spend more on woodland at the moment than we will get back, it will eventually yield rich returns in terms of timber, shelter for livestock, fruit production, and other natural products.

“We know that ecosystems that are healthy and diverse will be more resilient to [climate] changes so it can only be a good thing to improve the conditions of a woodland,” she says. “We owe it to the next generation to hand over an environment that’s no worse than what we inherited, and we can and should make it better.”