In a government case study on gov.uk, Holnicote Estate is presented as a practical example of how wildlife control is shaped by access law, protected site duties and woodland recovery goals. The National Trust estate covers 5,059 hectares in Exmoor National Park, much of it classed as access land under the Countryside and Rights of Way Act 2000, with 150 miles of public footpaths and bridleways crossing the holding. That setting matters because management choices are not being made on closed forestry ground. Holnicote includes open moor, coast and woodland, including Horner Wood National Nature Reserve and 351.8 hectares of ancient oak woodland within a Site of Special Scientific Interest. Any control programme therefore has to sit alongside public access, habitat protection and the condition of sensitive woodland.
According to the government case study, grey squirrels are the most serious current threat to woodland habitat across the estate. Bark stripping affects both young and mature trees, particularly in the canopy, and in some instances squirrels can remove 60 to 70 per cent of the bark. The result is weakened trees that are more exposed to wind damage and less likely to thrive over time. The range of available responses is narrower than it first appears. Since pine martens were confirmed on the estate, lethal traps for grey squirrels have no longer been used there because of the risk to that predator. Live trapping is described as technically possible but operationally poor at this scale, as effective lines would need 40 to 50 individual traps and at least three to four checks each day over the course of a month before any clear effect could be expected.
Holnicote has instead joined a wider grey squirrel control effort with neighbouring landowners in Exmoor National Park, using a syndicate of trained volunteers with low-powered air rifles. The case study records that one trained volunteer, operating at fixed bait stations, removed about 150 grey squirrels between October 2024 and June 2025 in an area that had been suffering heavy damage. Because the woods are open to the public, this method depends on site design and risk management. The estate uses formal risk assessments to decide where bait stations can be placed, with operators positioned away from public routes and steep valley sides used as backstops. The government case study says a 16.27-joule air rifle is presented as a low-risk option in that arrangement. Holnicote is also monitoring pine martens and Eurasian goshawks as natural predators of grey squirrels, and the case study notes that, at the time of writing, a further 10 pine martens were due to be released into Exmoor National Park within the following year.
The deer question is handled through a different set of trade-offs. While fencing is a common way to protect young trees, Holnicote decided not to make it the main response after considering cost, displacement and upkeep. The concern was that fencing one block would protect that site only temporarily while pushing browsing pressure into adjoining areas, and would then require a plan for maintenance and removal within around 10 years. The case study gives a direct comparison at Riverlands Wood in Porlock Vale. Fencing 17 hectares against deer would have cost about the same as employing a full-time Wildlife Manager for a year, including the initial equipment bill. Government grants have therefore been used to support a more flexible mix of targeted deer and squirrel control, alongside overplanting of selected tree species so that the estate can rely less on guards and fencing.
The estate also distinguishes between deer species rather than treating browsing pressure as a single problem. Red deer have a recognised role in the local visitor economy, but roe deer are identified in the case study as a more immediate threat to young trees and coppice regeneration, particularly in areas where Eurasian beavers are present and woodland recovery is already under pressure. At Riverlands Wood, Holnicote has been using Trico, a deer repellent derived from emulsified sheep fat, in places where fencing and tree guards would be costly and impractical. Each tree is sprayed twice a year, once in early spring and again in autumn. Initial results reported by the estate suggest roe deer browsing damage fell to below 20 per cent within the first year, although the case study also notes that deliberate overstocking of trees may have helped to produce that result.
Holnicote is not treating those early results as settled evidence. For new planting, the estate plans to use Forest Research's nearest neighbour technique to calculate percentage damage and test how much of the outcome is linked to repellent use and how much is linked to planting density. That distinction matters for future planting decisions and for how public grant support is directed. Habitat design is being used as part of the response as well. Glades and rides with highly palatable flora are being maintained along woodland edges to provide alternative browsing habitat. The case study says such areas once formed only a small share of the woodland, but are now built into management plans with deer control as a main objective. A secondary effect has been an increase in biodiversity value within those spaces.
Selective lethal control remains part of the estate's approach, with deer numbers currently managed at around 8 deer per square kilometre. The government case study presents this not as a stand-alone answer, but as one element within a broader package that includes habitat management, tree species choice and non-lethal measures. For policy readers, the main lesson is straightforward. Wildlife management on accessible, protected land is rarely a matter of picking one tool and applying it everywhere. At Holnicote, countryside access rules, protected site duties, public safety requirements, neighbouring landownership and grant funding all shape what can actually be done. The estate's approach is therefore best understood as a working model of conservation policy in practice, where legal duties and field operations have to be aligned before woodland recovery can be sustained.