Castell Henllys is an Iron Age inland promontory fort owned and managed by the Pembrokeshire Coast National Park Authority. The site delivers a range of the National Park Authority's key corporate objectives including those aimed at providing for the conservation, understanding and enjoyment of its natural and cultural heritage. The fort is a scheduled ancient monument and is one of many later prehistoric promontory forts within the National Park.

Castell Henllys is a unique visitor destination in the UK. It is the only Iron Age Fort to have roundhouses and other buildings reconstructed on their original, archaeologically excavated foundations. The award winning success of Castell Henllys, both as an educational resource and as a visitor attraction, is founded on the ability to bridge the gap between archaeological research and visitor understanding.

Castell Henllys occupies a spur of land overlooking the Nant Duad in a valley formed at the end of the last ice age from glacial melt water. The site was fortified by the local aristocracy of the time around the beginning of the Iron Age partly as a display of power and status and also to provide defence for that elite against the threat of attack from neighbouring forts. Towards the end of the Iron Age and the advent of the Roman period in the first century AD, the focus of settlement at Castell Henllys shifted. This was manifested in the establishment of a defended farm between the inner and outer great ramparts and in the abandonment, probably for economic as well as social reasons, of the fort itself. By the fourth century AD and the probable establishment of a new, Irish influenced local elite, the position of power shifted from Castell Henllys and the site was abandoned (Mytum 99). Throughout the Iron Age the surrounding landscape was formed, managed and dominated by successive generations of leaders at Castell Henllys.

Castell Henllys was occupied as fort and farmstead for nearly a millennia but evidence such as a Neolithic flint arrow head from the site and more recent nineteenth century land records demonstrate that it has always had an important relationship with people in its landscape. It is possible to suggest that, since the site was first surveyed in 1925, Castell Henllys has achieved an importance in the region greater than it had in the Iron Age.

The interpretation of Castell Henllys Iron Age Fort is inextricably linked to the development of its landscape. One could not exist without the other. The formation of the Castell Henllys landscape was determined by how the residents of the fort and their successors managed it. Successive probable landscapes of managed woodland, cultivated fields and open grazing present a palimpsest of the past which has been imaginatively interpreted through the dramatic reconstruction of the fort and its domestic buildings. In terms of interpretation, a roundhouse becomes the landscape, representative of the management required of it to provide the materials to build the house.

The archaeological record holds the key for interpreting not only people but also the landscape in which they lived and the habitats for the floral and faunal communities that were supported by it. The partial reconstruction of the fort has provided practical management prescriptions requiring the reconstruction of the landscape around it. These management objectives help provide for the conservation of bats, dormouse communities, ravens and dippers, otters and orchids, lichen communities and other flora and fauna already present at Castell Henllys, sometimes in large numbers (Castell Henllys has the second largest Pipistrelle nursery colony in Pembrokeshire). In short, the story of Castell Henllys can just as easily be related through the eyes of a dormouse as it can by a chieftain. The stories will differ but they will be of equal importance and, combined they highlight the diversity of wildlife and sustainable woodland and landscape management present in the Iron Age and Roman period and reconstructed and recreated at Castell Henllys today.

The long-term, structured programme of archaeological research and the subsequent attempted reconstruction of Iron Age buildings on their original, excavated foundations make this an important site. As an interpretative resource for visitors, school children and students, Castell Henllys is unique. The reconstructed area of the fort lies next to areas under archaeological excavation while nearby fields are used to demonstrate prehistoric farming practices with Iron Age breeds of livestock. Since its purchase by the National Park in 1991, Castell Henllys has become an important visitor attraction in north Pembrokeshire receiving around 30,000 visitors per year and providing a significant input into the local economy. The site is a model for innovative interpretation and a prime example of best practice in sustainable tourism.

As a result of an engineer's structural survey commissioned by the National Park Authority (NPA), the need to periodically replace or rebuild the structures at Castell Henllys was identified in order to avoid a perceived sense of decline especially by visitors returning to the site. In 1998, in partnership with the European Regional Development Fund, the NPA decided to attempt to reconstruct the largest known Iron Age roundhouse to have existed at Castell Henllys. The house platform chosen for the project had yielded comparatively little information during excavation by the University of York (Mytum 1986). Its probable diameter was gauged from a partially surviving wall foundation gully and the lack of internal postholes suggested that the roundhouse had a free-spanning roof. This paucity of information allowed a free approach to the project allowing the opportunity to experiment with construction type and methodology, whilst enabling reconstruction on the original site of a roundhouse with the same dimensions and maintaining a sense of shared space for visitors. By originating and recording every detail of material procurement and construction methodology (Bennett forthcoming), a focused attempt to consider the longevity of roundhouses and the resource implications of the construction of roundhouses on the local landscape in the Iron Age has been possible.

Castell Henllys dominates its local landscape not by being sited on the highest available point, but rather by appearing to be the centre of a natural amphitheatre to those parts of the landscape higher than it. The chosen house platform for this project is not only the largest at Castell Henllys, it is also the most dominant. With this as a consideration, as well as the demands of modern building regulations, our intention was to build a high status roundhouse with the best materials and craftsmanship available. The new roundhouse required the appropriate planning consents and visits from the Building Officer and Fire Officer at the various stages of its construction. The material requirements were massive. The new structure followed the foundation trench of its Iron Age antecedent and has an internal diameter of 13m with a roof apex of 8m. The wall is 1.2m in height and of wattle, woven around posts approximately 1m apart. The roof was constructed with oak rafters with a surface slope of c. 45 degrees to facilitate the run-off of rain and snow. The roof is thatched with water reed tied to hazel and ash purlins.

The primary strength and stability of the building is focused on three ring-beams. The first, a tension ring of mortise and tenon jointed split oak lengths was constructed at the top of the wall to bear the outward pressure caused by the weight of the roof and to provide a base to which the rafters could be joined. The second and main ring-beam is positioned just over halfway up the roof. A central post was erected at the beginning of the project to help in the positioning of rafters and the ring beams. The main ring-beam was constructed in the shape of a wheel with spokes radiating out from the central post to the principal rafters; the wheel's rim effectively joins all the rafters together and provided a surface to which secondary rafters were joined. The main ring-beam braces the roof and transfers the pressures on the roof due to weather from one particular point to the whole of the structure, increasing its strength considerably. A smaller, upper ring-beam was constructed just below the apex of the roof to facilitate the positioning of the tops of the principal rafters. When the ring-beams had been constructed and the roof rafters jointed and set in place, the central post was cut off just below the level of the main ring-beam leaving a free-spanning roof.

The water reed thatch was laid onto purlins of ash and hazel fixed to the rafters with hemp twine and wooden pins. Each of the purlins were joined together around the rafters, forming separate hoops 1m apart covering the area of the conical roof. The reed was attached to the purlins under strips of split hazel, which were then tied down with hemp twine. When approximately half the roof had been thatched, the wall was daubed both inside and outside the roundhouse. The daub mixture was made from six parts clay to one part cow dung together with partially rotted straw and water reed. The daub was pushed hard into the wattle then smoothed off. Around 15 tons of daub was used for the wall, which had to be patched up as the drying daub cracked. The finished building is impressive, particularly from the inside. The house is spacious, cool in summer yet remarkably warm in the winter when the door is closed. It is clearly visible and dominant in the surrounding landscape. Buildings of this size, complexity and sophistication should not be described as huts.

The acid soils in west Wales rot untreated timber very quickly. The internal oak supporting posts in another of the Castell Henllys roundhouses rotted out and had to be replaced after only ten years. The roof superstructure and thatch of a well-built roundhouse will always outlive the wall stakes and posts where they are set in the ground and yet there was, apparently, little evidence found at Castell Henllys to suggest that major maintenance and repair was undertaken to replace them. It would be easy to interpret this lack of evidence to suggest that the roundhouses had a short lifespan.

The positioning of a tension ring at the top of the wall might explain the lack of need for substantial wall supports below ground thus eliminating the need for their repair or replacement. Peter Reynolds has suggested that the weight of the building might provide stability even when wall stakes have rotted below ground. The engineer's report commissioned by the NPA maintains that outward pressure provided by the conical roof would eventually cause the walls to splay (Whitby and Bird 97). If, however, the structural integrity of the roundhouse roof can be maintained by the main ring beam then the tension ring should effectively stop any splaying of the rafters and wall posts joined to it caused by vertical and horizontal pressure from the weight of the roof. This could imply that the structural function of the wall is simply to support the tension ring at a height above ground convenient for those living in the roundhouse. The longevity of the roundhouse can thus be determined by the lifespan of the roof rafters, tension ring and the section of the wall above ground, rather than those surviving parts of the wall posts below ground level. If the roof has been carefully thatched by a craftsman and is subsequently maintained to a high standard the reed should survive for more than fifty years. If the roof rafters and tension ring timbers are large and protected from rain by good thatch above and smoked from the hearth below, they will actually get stronger with age. There is good reason to suggest that large roundhouses could have survived for generations.

By closely studying the material requirements for the construction of roundhouses we may be able to reconstruct aspects of the local landscape around areas of population in the Iron Age and consider the value of such resources. The timbers used in the construction of the latest roundhouse at Castell Henllys are all coppice products: that is they were all derived from trees, such as oak, hazel and ash, which have been regularly felled on a planned rotation to encourage the rigorous and straight re-growth of shoots. The time-span of a coppice rotation depends on the tree species and its intended use.

One hundred and twenty five bundles of hazel rods were used in the construction of the roundhouse, mostly for the wattle wall and the roof purlins. The bundles of hazel consisted of around twenty-five hazel rods each up to three metres long. Depending on the quality of the coppice, two hazel bushes might be required to provide one bundle of hazel. Such hazel requires a coppice rotation of up to eight years. One bundle of hazel was needed for each metre of wattle wall; therefore around ninety hazel bushes were needed to produce enough hazel for the roundhouse wall. Those hazel stools (stumps) would have been at risk from browsing deer, cattle, sheep, pigs and wild boar for at least four years out of the eight required within a given coppice rotation.

The roof required twenty-four straight oak rafters of up to 9m in length and twenty-two 2m-split lengths cut for wall posts and more for the ring beams. Coppiced oak may take up to twenty-five years to grow to the length required for the rafters. In all around thirty-four oak trees were needed for the building in addition to the hazel. The thatching of the roof used two thousand bundles of water reed which, ironically, were imported from Turkey. Many of the reed beds in Britain are now managed for nature conservation rather than for commercial purposes and supply can be problematic. Water reed also suffers from the high nitrate levels in many of our watercourses making the reed structure friable shortening its lifespan as a thatching material. Combed wheatstraw can be used for thatching but is not so long lasting as water reed.

The relationship between resources and roundhouse construction cannot be ignored. It is unlikely that the materials for such buildings would have been widely available in the landscape without extensive and long-term woodland management. Woodland work without modern tools is very labour intensive and one wonders whether such work would have been possible without the control and organisation of work forces by local elites. The large scale coppice requirements for roundhouses, palisades, fences and other structures places a value on those resources easily forgotten when considering the Iron Age. To what extent, for instance, would charcoal have a role as the primary fuel for heating and cooking? Again, without modern tools, the cutting of firewood into manageable lengths is hugely labour intensive while charcoal can be produced from long lengths of timber in large quantities in the woodland and is a more efficient and easier to transport fuel. As firewood, coppiced timber is an expensive use of a valuable resource. If our attempts at reconstruction indeed echo those of the Iron Age, it would be astonishing to think that the effort involved in cultivating and gathering coppice products would have been undervalued by a lack of subsequent woodland management. Similarly it is unlikely that large roundhouses might have been constructed for relatively short-term use.

The demand for coppice products in the Iron Age would have been influenced by the frequency, size and density of contemporaneous surrounding settlement. The longevity of roundhouses would, in part, have also determined the management of resources. The sustainability of managed woodland resources in the Iron Age would have been directly related to the value placed on these limited resources by local leaders. Whether or not Iron Age elites bothered about the human time consumption (conspicuous or not) related to woodland work, has to be considered against material requirements and the conservation and management of the resource. If we accept that the construction of Iron Age (and earlier) roundhouses were reliant primarily on coppice products, then the landscape implications would have been considerable. In this light, there seems little doubt that these coppice products were greatly valued by people in prehistory.