The honey bee, Apis mellifera, has existed far longer than humans have been building hives to house them. Before hives were available, and still to this day, honey bees have built their nests in hollow trees and sometimes in other natural cavities, including in the ground, in rock crevices and occasionally not in any kind of cavity. They also commonly nest in cavities in buildings.

Swarms do not select cavities at random. A nest cavity is chosen via a remarkable process carried out by scout bees from a swarm which locate a number of nearby options and of these select a good one (Seeley 2010). They are choosy, rejecting those too small or large in volume, or with entrances that are very small or large. Although bee hives are generally kept at ground level, swarms prefer cavities high above ground.

Swarms can travel several kilometres to a new nest site. Worker bees in swarms have enlarged wax glands and quickly build combs and rear brood from eggs laid by the accompanying queen. The new colony is challenged to build its population and food stores before winter. Many fail. Colonies founded in spring have a higher mortality rate in the following winter than colonies established in previous years.

The Ancient Tree Inventory

The Ancient Tree Inventory (ATI) is a citizen science project started in 2004 by the Woodland Trust (a charity promoting trees and woodlands), the Ancient Tree Forum and the Tree Register of the British Isles. Anyone can register a tree meeting the relevant criteria: ancient and veteran, based on probable age and old-tree characteristics, or notable, such as a good specimen. The listing is then confirmed by an expert and made public. In mid-November 2023 the ATI web site noted that “already more than 190,000 trees [are] listed but there are thousands more to add”.

Trees on the ATI can be referred to as Trees of Special Interest or TSIs but in this article we also refer to them as ATI trees. The ATI is far from complete. Based on a recent study by Nolan et al (2022) with the descriptive title “Distribution models calibrated with independent field data predict 2 million ancient and veteran trees in England”, there are 2 million trees suitable for inclusion on the list in England alone, which is about 15 per square kilometre. Given that England is 53.5% of UK area, there could be 3.7 million such trees in the UK. 

Ancient trees, wildlife, cavities and honey bees

As noted by the Woodland Trust, ancient and veteran trees are a valuable resource for wildlife. Therefore, protecting ancient and veteran trees does more than help just trees and woodlands. As trees get older and larger, they accumulate features valuable to other species. These include cavities, bark with a wider range of textures and pockets where it is partly detached, bark-free areas, deadwood, root buttresses, ivy and other climbers, epiphytes including mosses and lichens, bracket fungi, insect galleries, and in tropical areas termite nests. (For more features see Kraus et al. 2016. Catalogue of tree microhabitats – Reference field list1)

Cavities, the feature honey bee colonies need, take a long time to develop. Only 1% of oak trees, Quercus robur, under 100 years old in southern Sweden had cavities, rising to 50% at 200–300 years and 100% at over 400 years (Ranius et al 2009).

Tree cavities have many causes including fungal decay, which can be promoted by physical damage including that by humans. Some result from excavation by primary nesting animals. The black woodpecker (Dryocopus martius) is Europe’s largest woodpecker, although not found in Britain. It excavates tree cavities commonly occupied by honey bees. However, at approximately 10 litres, these are not quite large enough to accommodate the bees and their winter honey stores. Britain has three woodpecker species. The largest is the green woodpecker (Picus viridis). It makes cavities of approximately 7 litres (15cm wide x 40cm deep). In a study of 21 wild honey bee colonies in New York State, all were in cavities over 10 litres (Seeley and Morse 1976). Research using bait hives showed than cavities of 40 litres were selected over 10 or 200 litres (Seeley and Morse 1978).

Although woodpecker cavities are on the small size, a larger than average cavity or one that has increased in size via decay or by connecting to another cavity, would be a viable home to a honey bee colony. Honey bee workers have small, blunt mandibles adapted for moulding wax but not excavation, so they cannot enlarge a cavity. Many insects, including wood-boring beetle, wood wasp and moth larvae, make tree cavities although these are far too small for honey bee colonies.

LASI survey of ATI Trees

In 2021 and 2022 we surveyed 1052 trees on the ATI at 16 sites in South East England: in East Sussex, West Sussex, Kent, Surrey, and Greater London (Visick and Ratnieks 2023). These were mainly in landed estates with sufficient accessible trees to make surveying practical. Surveys were mainly in early spring, before leafing, to facilitate observing forager bee traffic at nest entrances. To find colonies, trees were scanned by eye and with the aid of binoculars on days and times suitable for foraging. To ensure that the traffic indicated a living colony, not just robbing or scouting, we confirmed that some bees had pollen in their baskets.

How many?

We found 22 honey bee colonies in the 1,052 trees, equivalent to an occupancy of 2% (noting that one tree had two colonies). This suggests that there could be 40,000 honey bee colonies living in ATI trees in England or 75,000 across the UK.

This estimate is based on an extrapolation from a relatively small survey in one part of the UK and should be used with caution. Perhaps interested beekeepers, bee and tree lovers, and naturalists will survey ATI trees in other regions.

The figure could be an overestimate. Maybe ATI trees in South East England have higher occupancy. However, the way in which we surveyed suggests that it may be an underestimate. For one thing, some wild colonies may not have been detected. More importantly, trees were mainly surveyed in early spring, after winter but before swarming. Many honey bee colonies die in winter and early spring with numbers building up in late spring and early summer via swarming. Most wild colonies that survive the winter will swarm, producing one prime swarm and possibly several after-swarms. If we reckon that each surviving colony, on average, makes one or two swarms, then surveys in summer would be expected to find two or three times as many wild colonies.

Of course, wild bee colonies do not only nest in ATI-quality trees. Many will be living in other trees as well as other cavities, particularly in buildings. Overall, this could be a substantial number. There are an estimated 288,000 managed hives in the UK. How many wild colonies are there? We do not know, but it is quite possible that the number of wild colonies is a high proportion, or even equals or exceeds those managed in hives. In another study we reviewed research on the density of wild honey bee colonies worldwide. The results indicate that wild colonies outnumber managed hives two- or three-fold, although in Europe hives outnumbered wild colonies (Visick and Ratnieks 2023).