How Holnicote Estate Manages Deer and Grey Squirrels

If you only looked at the headline version of this story, it would sound contradictory. Pine martens are being watched, grey squirrels are being shot, deer are being repelled and sometimes culled, and all of this is happening on land where people walk, ride and visit. But that is exactly why Holnicote Estate matters as a case study: it shows us that looking after nature usually means making several decisions at once, not picking one tidy answer. In the government case study, Holnicote Estate is described as a 5,059-hectare National Trust estate in Exmoor National Park, West Somerset. Much of it is open under the Countryside and Rights of Way Act 2000, with about 150 miles of footpaths and bridleways. It includes moorland, coast and woodland, including Horner Wood National Nature Reserve, a Site of Special Scientific Interest with 351.8 hectares of ancient oak woodland.

The woodland problem starts with grey squirrels. The National Trust says they are the biggest threat to woods across the estate because they strip bark from young and mature trees, often high in the canopy where damage is easy to miss until it is serious. In some cases, they can remove 60 to 70 per cent of the bark, leaving trees weaker and more likely to suffer in strong winds. **What this means:** this is not simply a row about one animal people find charming and another they do not. Grey squirrels are a non-native species in Britain, and the damage described here affects the condition of the woodland itself. When bark goes, the tree loses protection, and the effect spreads to habitats that depend on healthy trees.

The obvious response might seem to be trapping. But the government case study explains why that has changed. Pine martens have now been found on the estate, and because they need protecting, lethal traps for grey squirrels are no longer allowed there. Live trapping is still possible in theory, yet on an estate this size it becomes a heavy, labour-intensive job. The case study says an effective set-up would need 40 to 50 traps, checked at least three or four times a day for a month before a clear effect was likely. So the estate has joined neighbouring landowners in a wider Exmoor grey squirrel control effort. A trained volunteer now shoots grey squirrels at fixed bait stations using a low-powered air rifle, and the National Trust says about 150 squirrels were removed between October 2024 and June 2025 in an area where damage had been serious.

That detail may make some readers pause, and it should. Because this is public land, safety is not an afterthought. The case study says risk assessments are used to decide where bait stations go, with operators kept well away from footpaths and with steep valley sides used as a backstop. At the power being used, a 16.27-joule air rifle is described as low risk in this set-up. At the same time, the estate is not treating shooting as the whole answer. The National Trust says it is also monitoring pine martens and Eurasian goshawks, both of which prey on grey squirrels. Plans are also in place to release another 10 pine martens into Exmoor National Park within the next year as part of a wider reintroduction programme. **What this means:** predator recovery can help, but it does not remove the need for day-to-day management straight away.

Deer raise a different set of questions. Fencing is one of the most common ways to protect young trees, but Holnicote decided against making it the main fix. The National Trust's reasoning was practical as much as ecological: fencing may protect one patch in the short term while simply pushing browsing pressure into another. It also needs maintenance, and if it goes up there should be a plan to remove it within 10 years. Cost mattered too. According to the government case study, fencing the 17-hectare Riverlands Wood at Porlock Vale against deer would have cost about the same as employing a full-time Wildlife Manager for a year, including essential equipment. So the estate chose a more adaptable plan, backed by government grants, mixing targeted control of deer and squirrels with overplanting of some tree species to cut back on guards and fencing.

The deer story is not one-size-fits-all either. The case study notes that red deer are a major part of local tourism, but roe deer are the animals doing most of the harm to young trees and coppice regrowth, especially where Eurasian beavers are present. To reduce that pressure, the estate uses Trico, a deer repellent made from emulsified sheep fat. In Riverlands Wood, where fencing and tree guards would be costly and awkward, each tree is sprayed twice a year, once in early spring and again in autumn. The National Trust says browsing damage from roe deer fell to below 20 per cent within the first year. It also accepts that this result may partly reflect deliberate overstocking of trees, which is why the estate plans to use Forest Research's nearest neighbour method on new planting to measure the damage more carefully.

Another part of the plan is simpler to picture. The estate is keeping glades and rides with very palatable plants along woodland edges so deer have other places to browse. In older woodland plans these areas covered only a small share of the ground, but now they are being designed in from the start, mainly because of deer management. The case study says this has also increased the biodiversity value of those spaces. Selective lethal control remains part of the mix, with deer numbers being kept at what the estate describes as a healthy and sustainable level of about 8 deer per square kilometre. If there is one lesson here for the rest of us, it is that conservation arguments are often too thin when they focus on a single tool. Holnicote's approach only makes sense when we look at habitat management, tree choice, non-lethal methods and lethal control together, and when we ask a basic question first: what are we trying to protect, and how much damage are we willing to accept while doing it?