CHAPTER 5
Woodlands and their Management
WOODLAND IS A major habitat in and around the Lower Wye (Peterken, 2007). At 19% of the land area, it is not only the most extensive semi-natural habitat, but a high proportion is ancient and many individual woods have been designated as Special Areas for Conservation (under European legislation) and SSSIs. The gorge sides support almost continuous belts of woodland, which, as Gilpin recognised, form a major component of the landscape. The district contains some of the largest individual woods in southern Britain – Haugh Wood, Highmeadow Woods, Chepstow Park Wood, Wentwood – and on the fringes we have the Forest of Dean, an immense concentration of forested land by British standards. Moreover, this well-wooded state is of great antiquity: there has never been a time when the Lower Wye was open and bare, and in the gorge and its immediate surroundings it is quite possible that woodland has never dropped below 30% of the land area. In the seventeenth century, a colossal tract of woodland stretched from Wentwood to Chepstow and Monmouth on the Welsh side, and then, with scarcely a break, continued through the Dean to Ross and Dymock, and west of the Monnow to Grosmont and White Castle. Today, the Dean and the Lower Wye still stand out as one of the most densely forested parts of Britain, more like Continental Europe than the sparsely wooded English lowlands.
Woods are complex habitats. They occupy a range of site types that is almost as large as all other terrestrial habitats put together and they mix both open and tree-covered ground. Stands range in structure from ancient parkland to thickets of narrow-crowned saplings. Most woods include rides and forest roads – narrow corridors of bare ground and meadow – and some have patches of heath and marsh. The Lower Wye has many inland cliffs, most of which are still surrounded by native woodland. Streams commonly start as woodland springs. These subsidiary woodland habitats are considered in Chapters 7 and 8.
ORIGINS
Changes during the nineteenth and twentieth centuries
It is important to know the origins of the 19,600ha of woodland in the Lower Wye, and how it has changed over the centuries, for origins partly determine which species are present. We can start to measure the changes in the amount and distribution of woodland by comparing the pattern of woods on the first Ordnance Survey maps with those on the present map. The OS first surveyed the Lower Wye between 1812 and 1815 – the original survey sheets can be seen in the British Library – but the first published maps emerged several years later, by which time small revisions had been incorporated.
In 1815, woodland occupied 15,500ha, or 15.2%, of the Lower Wye. The main concentrations, then as now, were in the Dean (48% of the district) and the gorge (35%), but the Woolhope Dome (18%), Trellech plateau (17%) and southeast Herefordshire (16%) also contained well above the national average woodland cover. The great majority of this woodland still survives: in fact, only 4.5% has been lost in 190 years. Some woodland was destroyed by quarrying for limestone, such as at Penhow near Caerwent, Livox south of Tintern and Stowfield near Scowles. Open-cast mining for coal removed woodland from Howle Green in southeast Herefordshire and at Oakenhill Wood and The Tuffs in the Dean. Several woods in southwest Herefordshire have been lost to agricultural expansion, including Scudamore Wood near Orcop Hill, Mowaston Wood south of Little Dewchurch, the core of Bolstone Wood and the southern half of Warmhill Wood. Scarp woods have been lost at Woolaston Slade and the western fringes of the Trellech plateau, and many small woods have gone from the south Monmouthshire limestone and streamsides in the Trothy catchment. These losses were just the latest in a long attrition: for example, Tintern Abbey’s Harthill Wood in Hewelsfield disappeared over 300 years ago, leaving a name and a few small-leaved limes in the hedges.
Despite these losses, the overall trend has been to more woodland. Indeed, every district of the Lower Wye has more woodland today than it did in 1815. The most conspicuous expansion was on the former Trellech Common, where great expanses of heathland were afforested in the nineteenth century. Similar heathland afforestation took place on a smaller scale on Poors Allotment, Welsh Newton Common, Coppet Hill and the warren on Aconbury Hill. Fields near Staunton were planted in the early nineteenth century, and the largely treeless wood-pasture permeating and fringing the Dean was infilled with plantations and natural growth. New woodland also developed on abandoned small fields that had once been commons on the Hudnalls, the Doward, Leys Hill at Walford and Aconbury Hill. Streamside scrub was allowed to expand in several places, notably along the upper Mounton Brook; several patches of scrub on slopes developed into woodland; and plantations were inserted into Kentchurch and Stoke Edith parks.
The great majority (76%) of modern woodland originated before 1815, and a further 5.6% was already scrub or parkland at that time. Significantly, the new woodland was not isolated: some 81% of all post-1815 woodland was contiguous with pre-1815 woodland: in fact, 44% originated within 100m of an existing woodland, and the average distance from any bit of new woodland to an existing wood was only 294m, so woodland species did not have to travel far to colonise. The only place where new woodland stood a long way from existing woodland was between Trellech and the gorge, where the new woodland resembled the conifer plantation forests that spread over upland Britain in the twentieth century.
Ancient and primary woodland
How much of the 15,500ha of woodland in 1815 was actually ancient (originated in the Middle Ages or earlier)? Some certainly originated as plantations on formerly unwooded ground, but most woods can be identified by name in medieval documents. The precise proportion can only be established with detailed research, but I estimate that 80-90% of all pre-1815 woods were ancient, and thus that 65-70% of all existing woodland is ancient. These woods were all listed in the inventories compiled by the Nature Conservancy Council in the 1980s.
A more difficult question is: how much of the medieval woodland was primary, i.e. had never been completely cleared of trees since the original forests were broken up by prehistoric clearance? This raises many issues, including our inability to pinpoint exactly where and how much of the prehistoric woodland was cleared and whether the original woodland was dense or a mosaic of tree-covered ground and open areas (see below). The ancient woodland on Iron Age forts in Pierce Wood, Llanmelin Wood, Symonds Yat, Caplar Camp and elsewhere looks secondary (though Oliver Rackham has suggested that small-leaved lime woodland may never have been completely eliminated when the ramparts of the camp in Welshbury Wood were constructed) and woodland certainly spread over cleared land before the Romans arrived (Chapter 4). Even so, much of the ancient woodland seems primary, if only because large- and small-leaved lime (Fig. 66), both notoriously slow colonisers of secondary woodland, are abundant (Peterken, 2005a). These species were abundant in the pre-Neolithic woodland of southern Britain (Greig, 1982).
There will always be some ambiguity about which particular woods are ancient or primary, especially in the Lower Wye, where so much of the woodland was once a mosaic of open ground, scrub, old trees and dense groves, the common woods that descended from the prehistoric wooded waste. As an example we can consider the land that can be seen from the Symonds Yat Rock. Coppet Hill, the view to the north, was known in 1372 as ‘Coppyngwode’ when it was ‘worth nothing but for big timber’, i.e. it had been converted to open wood-pasture (Tisdall, 1998). Since at least the seventeenth century it has been open rough pasture, but even now two relicts of the medieval wood survive: protected from grazing, wild service-tree still grows on the conglomerate outcrop; and on a margin protected by a limestone quarry small-leaved lime grows with herb-paris. Lords Wood on the Doward, the view to the west, was also part of the wooded waste of the manor of Goodrich, but in 1718 the commoners agreed to give up their rights in the area now known as Lords Wood, stock were excluded
FIG 66. Flowering small-leaved limes stand out in the lower part of Bigsnap Wood on the Monmouthshire side of Bigsweir Bridge, illustrating the irregular distribution of trees in ancient, semi-natural woodland. The toll house was erected when the bridge was built in 1824.
by ditches and fences, and thereafter the woodland was treated as coppice until in 1840 it was sold to the Office of Woods and converted to oak-dominated high forest (Watkins, 1982). Today, Lords Wood is regarded as ancient, whereas the new woodland spreading over Coppet Hill is regarded as secondary, but the difference is a matter of degree.
MANAGEMENT OF ANCIENT WOODLANDS
Woodland survived because it was useful. Even woods on the steepest ground in the gorge and steep slopes such as Caplar were used as sources of timber, wood, leaf mould, honey, hawks, etc., and most were also used as pastures for at least part of the time. Under the guidance of Oliver Rackham (1980, 1986, 2006) we have come to recognise two main forms of traditional management, coppice and wood-pasture, and both played their part around the Lower Wye. Coppicing involved cutting trees close to the ground and allowing them to sprout again from the stumps to yield crops of brushwood and poles, so any grazing had to be limited. A few shoots were allowed to grow into large trees for timber, but their number was limited so that the thicket of underwood could grow strongly. Wood-pasture on the other hand was always available for pasturage, but trees could be lopped to yield small wood. Inevitably, regeneration was inhibited by cattle and deer, so wood-pastures tended to develop into open groves of pollarded trees with large trunks, and eventually into heathland. The contrast was not absolute, for informal coppices were integral parts of medieval wood-pasture, and formal, embanked coppices were fashioned from wood-pasture and continued to be grazed intermittently after new shoots had grown out of reach of cattle.
Wood-pasture
Prehistoric woodland became ‘wood-pasture’ as soon as it was exploited by prehistoric people. In the Lower Wye, enormous tracts of such woodland survived into historic times as wooded waste, subject to common rights. Whilst much of this was incorporated into forests, chases and parks, an unusually large number of commons survived as woodland, and some survive today.
Domesday Book gives few clues about the character of eleventh-century woodland, though we know that King Edward could hunt in Haye Forest before 1066. The Laws of Hywel Dda regulated hunting in forests and woods, and restricted usage in order to maintain timber stocks and ensure regrowth (Linnard, 2000), but pasturage remained one of the main functions of Welsh woods. Records assembled by Penny Oliver (1998) for a tract of southeast Herefordshire that included Penyard Park, Chase Wood, Howle Hill and Bishopswood record a dispute in 1286 between the bishop’s huntsmen and the king’s foresters over the right of the former to kill a stag, and an accusation made in 1347 and 1354 by the keeper of Ross and Malvern chase that the Bishop of Hereford stole the formidable total of 500 hares, 1,000 rabbits, 1,000 partridges, 500 red deer, 500 fallow deer and 300 roe deer. The bishop’s woodmen also illegally cut down oak, ash, beech, hazel, thorn and elm to make charcoal (Whitehead & Patton, 2001).
Early records of the Dean (Hart, 1966) imply a mosaic of underwood, timber trees and open ground. By 1282 the western fringes from Penyard, Monmouth, Hadnock and Tidenham had been privatised as chases and parks, and many Dean woods had been felled and wasted, leaving extensive clearings. However, as underwood was present at the same time, pasturage must have been light enough to allow regeneration. There seems to be no clear evidence of ancient pollards in the Middle Ages, but in 1565, Taverner describes several thousand acres in Staunton, Ruardean, Lea Bayley, Bearse and elsewhere as ‘oak and beech of great age, most part commonly shred well nigh unto the top’. A survey of 1634 counted 64,007 oak and beech trees that were too old or deformed to be timber trees, and which had the same value as offal wood (branch wood). In 1638, the Bearse contained one timber oak, one non-timber oak, 142 beeches and 45 stoggals, and the trees of Staunton and Bicknor woodwardships, most of which were oak or beech, were ‘for the greatest part decaying’. In 1680, 340 old decayed oaks and beeches were to be felled in the Lea Bailey.
Today, the main relic of unfenced wood-pasture with ancient trees and glades is at Speech House (Fig. 37), but it is a battered remnant of the stand described by Cooke (1913):
There are few statelier oaks than here, and nowhere is the woodland so thickly interspersed with ancient holly-trees, many coeval with the house itself, some possibly older still. Fully three thousand of these shining evergreens stand scattered around. It is easy to find hollies with stems five or six feet in circumference, though in many cases the trunk has forked close to the ground, whilst in others two or three stems which seem to grow from the same root are closely intertwined.
A huge wood-pasture also dominated the Trellech-Wentwood plateau, extending from Wentwood to Tintern and Penallt and onwards from the Buckholt northwards. Parts were clearly coppice with standards and a scatter of ancient trees: thus, when the Duchy of Lancaster surveyed the 594 acres (240ha) of Earlswood in 1608 they found 1,513 timber oaks, 4,269 great trees, 590 decayed trees, 314 timber beeches and 4,952 stools of beech (Courtney, 1983). Trees were steadily declining elsewhere, however. For example, Buckholt was so greatly spoiled in 1608 that it was worth nothing as timber. In 1581, commoners claimed rights of estovers, firebote, housebote, hedgebote, pannage, herbage and the right to dig lime to spread on their land in Wisewood (Bradney, 1913), which covered well over 400ha east of Trellech, and in 1697 it was ‘formerly well stocked with red deer and timber’ (Courtney, 1983). Later, recorders noted ‘where the wood doth now grow, or heretofore hath growne, and been cut down or spoyled’; and on 5 April 1714 some 8oha of ‘the best of ye woods’ near Tintern were burned by locals rioting against enclosures (Gwent RO, D749.311). When it was finally enclosed in 1810, Wisewood was open moorland with just a fringe of trees on the slopes down to the Wye. In addition to Wisewood, a 1676 survey records that the Lordship of Trellech included 2,208 acres (894ha) of common wood and 2,493 acres (1,009ha) of waste, probably formerly wooded, and the 1810 Enclosure Act covered 4,200 acres (1,700ha) of waste and wood in Penallt, Trellech, Tintern, Cwmcarvan, Llandogo and Llanishen. Latterly, Earlswood, Coed Llifos and much of Wisewood became farmland, Wentwood, Chepstow Park and Hale Wood became coppices, Beacon Hill, Ninewells and others passed through heath and grassland to become plantations, leaving just Cleddon Bog and smaller fragments as heathy common.
Hudnalls, a 500ha wooded common in St Briavels and Hewelsfield, was a detached part of the forest of Dean (Hart, 1966). Although in 1282 it had been cut down by the men of St Briavels, it was still largely wooded when it was first mapped in 1608, and in 1641 it was described as underwood, implying coppice, though it was also used as pasture. Fortunately, a fine description of its former state was preserved in a Wye Tour guide (Heath, 1806):
The wood, from being open to a number of families, does not arrive at any growth to render it of any much value. The poor who cut it, and to whom the right belongs, make it up into small faggots, and send them by water to Bristol, where it is purchased by the bakers.
Describing the parts visible from the Welsh bank: ‘the oak pollards were as numerous as the beech [on the Welsh side of the Wye] but from the high price which has of late years been given for bark, they have been so destroyed, as not to leave a twig behind from which a pound of bark was to be obtained.’ Remnants of this wood (Fig. 67) show that it was an open woodland of stubs and pollards of mainly sessile oak, pedunculate oak, beech and small-leaved lime.
FIG 67. Beech and small-leaved lime stubs in the common wood of the Hudnalls, growing on the Quartz Conglomerate (puddingstone) outcrop and the rock debris below. Until the early twentieth century, these were repeatedly lopped for firewood. Now uncut, they have grown tall and will eventually collapse: indeed, the multi-stemmed tree on the right has fallen since this photo was taken.
Moccas Park (Fig. 68) is now one of the finest ancient deer parks, but the first certain evidence of its existence dates only from 1665, when it was shown on Saxton’s map (Harding & Wall, 2000). Evidently, the ancient trees described by Kilvert were formerly just wooded waste in a remote corner of Dorstone parish, which was incorporated into the landscaped park around Moccas Court. It survived as ancient wood-pasture because it was remote from both Dorstone village and the great house. Many other parks are recorded from the Middle Ages onwards, but most have disappeared. For example, timber was being taken from a park at Trellech in 1330. Rotherwas near Hereford was once a deer park. At Fownhope in 1354, the Chandos family had a park, but only the name remains. In the later nineteenth century (Shirley, 1867), the ancient park of Holme Lacy still had magnificent timber and 130 fallow deer. A late Victorian inventory (Whitaker, 1892) recorded Lydney, St Pierre, Wyastone Leys, Haye, Moccas, Kentchurch, Holme Lacy and Haye parks, all stocked with fallow deer, with fine old trees at the last three of these. Wyastone, on Little Doward, is unusual as a new nineteenth-century deer park.
The Lower Wye retained an exceptional amount of wood-pasture from
ancient times. Until the early Middle Ages, it appears to have been a highly variable entity that included underwood, stands of timber trees and open pastures. Some commons survived in this form into the twentieth century and fragments were incorporated into parks, but most was either enclosed as coppice or allowed to become open rough pasture before it was cleared to farmland or stocked as plantations.
Coppice
We know from trackways in the Somerset Levels that coppicing was taking place across the Severn when the Sweet Track was built about 3920 BC (Coles & Orme, 1976), and the evidence from the Caldicot Level (Chapter 4) suggests that the same was true of the Lower Wye. Early references from the Dean (Hart, 1966) to ‘underwood’ confirm that coppicing was widespread in the Dean and carefully regulated. Thus, for example, in 1237, the constable was ordered ‘to take care that in the season when underwood should be cut [autumn and winter], it should be so cut to grow again and that no damage should befall the coppice’, a reference to the need to protect new growth from grazing animals by enclosure with a ditch and a hedge. In 1627 this was formalised in an agreement between the commoners and the Crown to enclose all coppices ‘well and sufficiently’ for nine years ‘so that the coppice receive not hurt from cattle or otherwise.’ A report in 1662 concluded that underwood should be allowed to grow for 14 years between cuttings, and that 30 ‘of the best and straightest young shoots of oak and beech be left as standards for timber in every acre and no more lest it prejudice the underwood’. Coppice management was not without problems: for example, in 1333, when much of the wood was converted on site into charcoal for the ironworks, one charcoal pit set fire to bracken and burned some underwood. Such accidents were rare, however, and the main problem was grazing animals. Thus, in 1565 Roger Taverner recorded that 10 acres (4ha) of Kydnall Collet (in Staunton woodwardship) was waste, and a report in 1692 noted that the Lea Bailey was then a ‘spring’ (coppice) of oak and beech of 4-6 years’ growth, ‘but much cropped and spoiled by cattle’ because the boundary banks and hedges had been ‘in the night several times pulled down and dest royed by persons unknown.’
In more recent times, coppice rotations were generally between 10 and 20 years. Clark (1794) recorded that Herefordshire ash, sallow and alder coppices were cut every 12-14 years and oak was cut at 18-20 years. The Bond family’s management records for Cadora Woods, held in the Gloucester Record Office, show a rotation of 9-14 years between 1709 and 1738, with most coppices cut after 11-12 years. In 1795, Haugh Wood was being cut on a 20-year rotation in association with neighbouring woods (David Lovelace in Peterken, 2007). Rudge (1813) reckoned most coppices around the Dean were cut every 14-15 years. Coppicing was not always regular: for example, the whole of the 308ha of Hadnock Wood was cut in four years from 1600 to 1603 (Peterken & Jones, 1987).
In theory, both underwood and timber trees should be cut close to the ground, but several allusions suggest that many trees in Wye Valley coppices were cut in other ways. Thus, the description of Kydnall Collet in 1565 shows that in 1555, when the underwood was last cut, the older oaks scattered among the underwood were shred, i.e. the side branches were cut for fuel, leaving a lollipop-shaped tree. (Shred trees are still visible in some nineteenth-century engravings.) In 1720, the coppice in Hembridge Grove in Newland had been damaged by the ‘height of the stubs by former ill cutting’, and a similar reason was given for the low value of David’s Grove (part of Cadora Woods) in 1722. In fact, cutting of coppice at a variety of heights from flush with the ground to over 2m above the ground appears to have been widespread along the gorge, and the remains of this practice can still be seen in the stub trees of Cadora Woods (Fig. 69).
The accounts of management in Cadora Woods show that the main costs were cutting, marking, hedging, carriage (to Bristol and Redbrook) and rinding,
FIG 69. A small-leaved lime stub in Cadora Woods. This ancient coppice was typical of the Lower Wye in having numerous high-cut stools. About 1970, it was largely converted to conifer plantations, leaving the ancient stools to die and rot in deep, evergreen shade. Fortunately, the wood was acquired by the Woodland Trust in time to save some, but by then many of the surviving stools were extremely weak. The conifers are being cut back, and the stools lopped to prevent them collapsing under their own weight, but they have to be cut even higher now to prevent fallow deer from browsing all the new growth.
i.e. removal of oak bark for tanning. The main income sources were rind and smart hoops, cordwood, backwood, stub wood, great and small faggots, bast without the wood, white rods and bark. The same uses were recorded a century later (Clark, 1794; Rudge, 1813), though the market for hop poles had grown. Larger oak was ‘split into laths, or sold for the use of the coal-pits, and a great deal of both is made into cord-wood, for the consumption of the iron furnaces, either in its natural state, or converted into charcoal.’ The lime, or ‘white rod’, was sold to the toy-manufacturers in Bristol, and the inner bark was made into ropes for cider-presses, draw-wells and fishing boats. The bark was stripped about midsummer, dried like hay in the sun, often made into ropes on the spot and sold in pairs 90m long. The demand for hoops for barrels from ash and hazel was a local speciality to Monmouthshire and Gloucestershire; they were exported from London and Bristol to the sugar plantations of the West Indies (Jones, 1961).
The visual impacts of coppicing were recorded by Heath (1806):
Every ten or twelve years, the woods [by the Wye] are cropped quite close to the ground, principally to supply the forges and furnaces with charcoal, &c and as they sprout again, this delightful verdure appears scarce distinguishable at some distance, from the most luxuriant crops. As in other spots their vigour is increased, or come to full growth, different tints and shades are seen, which constitute the wonderful variety so peculiar to these scenes.
Photographs record how cut material was dragged out by horses (Fig. 70), and Cooke (1913) left a fine description of charcoal burners at English Bicknor (Figs 71 & 72).
A forestry dissertation rescued from a second-hand bookshop in Edinburgh (Lord, 1934) not only shows that traditional management survived within living memory, but also demonstrates the attention that traditional woodmen gave to detail. The coppice in Caswell Woods, opposite Tintern Abbey, comprised mainly beech, ash, hazel and lime, with some old standard oaks retained for amenity (Fig. 73). The beech and much of the other material was evidently used as firewood, but some other species had specialised uses. Thus, any willow was used to make ladder rungs, and the lime, which was useless as firewood, was used for the manufacture of musical instruments. Ash growing on limestone was of high quality, so it found a ready market for the manufacture of sports work: longer lengths were sent to ‘the manufacturers of shafts of agricultural instruments’. Size mattered: material about 21/2 inches by half an inch (6 x 1cm) was used as chisel rods in metal and iron foundries; the next size went for use as spile rods
FIG 70. Horse hauling coppice poles from woodland above Catchmays Court, Llandogo, before World War II. To the left (east) of the Wye as it flows down to Brockweir is the scattered settlement and small fields on the lower part of St Briavels Common. This view from just below the road up from Llandogo is completely obscured now by mature woodland, and, even if these trees were cleared, most of the houses on the common would be invisible because trees have spread and grown taller. (Source: National Forest Park Guide to the Forest of Dean, Forestry Commission 1947. Original photo credited to Messrs Fox Photos.)
for barrels; while still larger material was used as crate rods for packing. Long coppice was used for bean sticks and rake handles, while shorter, thicker material was used for slats of crates. Logs 8-12 feet (2.4-3.6m) long of equal diameter throughout and containing no knots were used for barrel hoops. Those about 12 feet (3.6m) long of tapering diameter became putt rods, used in making salmon traps. Still larger material was used for turnery work of all kinds while smaller material was cut for firewood. Older coppice, where ash was 6-8 inches (15-20cm) in diameter at the small end and any length from 2 feet (60cm) upwards, if of good quality, went for the manufacture of helves (handles) of all kinds, for the wood from coppice shoots was much better than that from seedling or maiden trees. The wood might be straight or crooked, for it was cleft along the grain and, if crooked, it needed only to be strained and straightened.
FIG 71. Charcoal burning on the flanks of the Wye near English Bicknor early in the twentieth century: building the charcoal pit. (From Cooke, 1913)
FIG 72. Charcoal burning on the flanks of the Wye near English Bicknor early in the twentieth century: the charcoal burner’s hut. (From Cooke, 1913)
FIG 73. Tintern Abbey and Caswell Woods photographed before 1929. The multitude of postcards of Tintern dating from about 1900 onwards show incidentally the twentieth-century history of woodland management on the slopes behind. Here, part of Caswell Woods has been coppiced in the winter preceding the photo, and the rest was cut a year or two earlier, leaving larger trees only on the skyline.
FIG 74. Ancient semi-natural woodland below the Fiddlers Elbow in late spring. The variety of colours and tones show this to be an intimate mixture of beech, small-leaved lime, ash, gean, yew and others.
FIG 75. Giant pollard beech and an ash stool on a boundary bank in the Little Doward woodlands. Giant pollards are still a feature of the Lower Wye woods and nearby field boundaries, though the majority are small-leaved limes. These structures are immensely impressive, but they will eventually disintegrate unless the crown branches are cut again.
Traditional woodmen used what the woods produced and made little attempt to change woodland composition. As Nisbet and Vellacott (1907) recorded, the underwood in the 1,083 acres (438ha) of ‘old coppicewoods’ on the Sedbury Park Estate comprised ‘a mixture of oak, hazel, birch, alder, black withy, lime, maple, whitebeam and chestnut growing at random…[which] do not appear to have been planted, but seem to be the remnants of the ancient woodlands formerly existing in this part of the country.’ By this means the ancient coppices retained a mixture of tree and shrub species that descended directly from the original-natural forests (Fig. 74). Today, pollards, stubs, stools and multi-stemmed trees are still common, but are slowly disappearing (Fig. 75).
Other trees and woodland
Today, most woodland is ‘high forest’, i.e. closed stands of trees grown from seedlings, most of which were planted. High-forest structures and restocking by planting became prominent only in the nineteenth century, but both were known in earlier times. Some of the early Dean records imply patches of high-forest oak and beech, and Linnard (2000) records an example of high forest at Gilbert’s Hill, east of Abbey Dore, which at the Dissolution comprised ‘120 acres, whereof 13 acres be of 50 years growing, and the residue of 100 years growing and above’, all enclosed by a hedge. Earlier, he noted that the Welsh Laws of Hywel Dda made explicit reference to planting for shelter and regeneration.
MODERN CHANGES IN MANAGEMENT
Despite the prevalence of coppice and wood-pasture in the past, we see little sign of traditional management today. Coppicing was in decline from the late nineteenth century onwards, though it continued well into the twentieth century in many woods. Collections of old postcard photographs of Tintern Abbey show that almost the whole of the backdrop of Caswell Woods was coppiced in the 1920s (Fig. 73), and in many woods coppicing continued until the 1950s. Gilpin’s Wye Valley landscape of large patches of freshly cut and regrowing coppice was last seen in 1970 at the Fiddlers Elbow, where the woods were cut for the Sudbrook pulp mill, and again in 1993 when the wood was acquired by the Woodland Trust.
The first signs of coppice decline appeared in the Crown Woods and large woodland estates. When the Crown acquired Tintern Woods from the Beaufort Estate in 1901, they inherited oak-dominated coppices, but resolved to change these to beech high forest (Schlich, 1915; Spence, 2000). The product of this policy can be seen in the mature beech-oak stands in many woods by the Wye (Hael Wood, Bargain Wood and Coed Beddick), Angiddy (Buckle, Ravensnest and Lower Hale Woods) and Whitebrook (Pwllplythin and Manor Woods). A report produced by an expatriate Indian forester ‘home on furlough’ (Hill, 1897) recommended converting the coppices in Highmeadow Woods to beech high forest, and the woods were duly thinned to that end. About 1880, the Bigsweir Estate converted the coppices above the Florence and in Red Hill Grove to oak high forest.
Most coppices have not been cut for 50 years or more. Some now justify the old forestry appellation ‘scrub’, but most have grown into fine, tall mixtures of ash, lime, beech, oak, gean and others (Fig. 76). Nevertheless, the woods of the Lower Wyndcliff, Lancaut, Causeway Grove near Redbrook, and elsewhere can still be recognised as former coppices by the mixture of slender, multi-trunked limes, ash, field maple, oak, beech, alder, etc. (the former underwood) and larger oaks with spreading lower crown branches (the former standards). In most instances the unprecedented height-growth of the underwood has killed the lower crown branches of the standards.
FIG 76. Outgrown small-leaved lime coppice in Lower Martridge Wood, below the Wyndcliff, growing into graceful high forest. (Photo: Margaret Perring)
The decline of traditional management was complemented by the rise of plantation forestry. Courtney (1983) records ornamental conifer planting in east Monmouthshire from the early eighteenth century, and Hart (1966) gives the first introduction of conifers – Weymouth pine – into the Dean as 1781. Indeed, Herefordshire was ‘so well stocked with fortuitous timber and coppice woods, it is no wonder that the propagation of woodlands should be neglected, or avoided’ (Duncumb, 1805). However, by the early nineteenth century, Scots pine, Norway spruce, silver-fir and European larch were widely planted, and later in the nineteenth century the conifers of the Pacific Northwest of North America arrived in numbers. Even so, when the great Census of Woodlands was compiled in 1947, only 12% of all woodland in Herefordshire and Monmouthshire was stocked with conifer high forest (including mixtures with broadleaves), though, significantly, only 2.8% of these conifers had been planted before 1907. Evidently, the rate of conifer planting rose steadily through the first half of the twentieth century, when European larch, Japanese larch and Douglas fir were the main species being planted. Conifers were disliked from the start – thus Fosbroke (1818): ‘Will it never be known that firs in groups are like plumes on the graves of the Picturesque.’
After World War II, many neglected woods were felled and replanted. Initially the favoured species were native broadleaves, mainly beech and oak, but from the 1950s onwards conifers were widely planted into the ancient coppice woods, using herbicides to control native regrowth. The products of this policy are still widespread, especially in Forestry Commission woods (e.g. Haugh Wood, the Highmeadow Woods around Staunton, Chepstow Park Wood, Wentwood and half the Dean). Furthermore, many private woods were converted with state aid: I well remember my first professional visit to the Wye gorge in 1969, when I watched the felling and poisoning of Cadora Woods. The outcome was a dense, dark thicket, relieved only by a scatter of retained oaks and marginal broadleaved regrowth.
These changes are now being reversed in many woods, especially in the gorge and the large Forestry Commission woods. From 1985 the National Broadleaves Policy provided special protection to the ancient woods identified in the NCC’s Inventory, and once the timber resources of eastern Europe and Russia became accessible after 1989, the bottom dropped out of the softwood market, thereby enabling foresters to acceed to conservation and landscape demands for native woodland restoration. Fortunately, some of the original coppice constituents, notably small-leaved lime, had proved difficult to kill, so even in Cadora Woods there was much to save when the Woodland Trust embarked on restoration thinning.
In recent years, thoughts have also turned to the management of the native broadleaved woodlands, including the neglected coppices, mostly by utilising some timber, whilst maintaining the general appearance of mature woodland (Wye Valley AONB, 1992; Peterken, 2007). Foresters now call this ‘continuous-cover forestry’, which takes the form of perpetual thinning or very small-scale patch felling, and a prayer that natural regeneration will fill the gaps.
The remaining wood-pastures were also transformed, though in this instance the transformations were necessarily bound up with enclosure. Some, such as the Hudnalls, Fence and Lower Meend in St Briavels, were settled in the late eighteenth century by squatters, who built a new landscape of small fields. The woods were not totally removed: indeed, the Hudnalls woodland remained on the slopes over the Wye, where it was still roughly coppiced and depastured until the mid twentieth century. Now, however, it has grown up to a form of high forest in which the former practice of lopping can still be seen in the grotesque, contorted shapes of trees on the steep rocks and in 200-year-old walls (Fig. 67). Elsewhere in St Briavels, the Bearse was planted with oak, and more recently with conifers. Some former common wood-pastures remain: three in Hewelsfield still have massive lime pollards in a matrix of younger growth.
All these changes have brought several new factors into the management of Wye Valley woodlands. The height and density of trees have increased, thereby intensifying the shade cast on woodland habitats. Few visitors recognise this, but long-term residents know that former views are now obscured. Old photographs quickly highlight the changes: the westerly view from Symonds Yat Rock down the gorge was frequently photographed in the early twentieth century, but was obscured until recently.
In native woods, differential impacts on tree species have emerged. Admittedly, selectivity is not new – the coppice woodsmen promoted oak as standards, and the Crown estate managers promoted beech – but today, with the rise of continuous-cover management, the forester will constantly be deciding which species to cut and which to keep. Moreover, he or she will increasingly be guided by national recommendations, itself a new factor in the Lower Wye. A particular factor in selectivity will be the burgeoning populations of fallow deer and grey squirrels. Deer are offered a wide menu, but have favourites and dislikes. They go for hazel, lime and wych elm, but tend to leave beech. Grey squirrels on the other hand vigorously strip bark from sycamore, beech, sallows, field maple and oaks, but tend to leave ash and lime. Against this background, the future composition of the native woods will reflect more the willingness of people to control these two introduced mammals.
FIG 77. It is now difficult to witness traditional coppicing, but here in Cadora Woods the Woodland Trust has cut a coupe just above the A466, partly as a safety measure. The stand had been left uncut for perhaps 40-50 years and stool density has declined, but fresh shoots will spring from the stumps and the gaps will be filled for a decade with brambles.
There are also plans for a modest revival of coppicing. The Woodland Trust plans to retain some coppice on Fiddlers Elbow, the Gwent Wildlife Trust has made charcoal from coppice in Croes Robert Wood for many years, and the Forestry Commission has coppiced the lower parts of Caswell Woods and some other sites. Ride-side scrub is also being cut in Haugh Wood and other sites, and strips of trees beside the main valley road are being cut back for safety (Fig. 77), and both actions are tantamount to coppicing. Here, too, the ability and willingness to control deer browsing will be crucial.
TREE AGES
Most of the Lower Wye woods are ancient, and many stands are mature and must seem unchanging to most visitors, but how old are the trees? The most direct evidence comes from counts of annual growth rings, which are clear enough on oak stumps, but less so on birch and beech. Most ring counts are opportunist – one can hardly fell the tree to find out how old it is – though specialists can obtain estimates from cores. Unfortunately, even with a corer, the interesting trees – the largest and oldest – are either hollow or very difficult to count, so ages can only be estimated from size, form and increment rates.
The great majority of Wye Valley trees are less than 100 years old. Either they were planted after the foundation of the Forestry Commission in 1919, or they grew up after the last coppicing between 60 and 100 years ago. Since they generally date from immediately after a felling, most stands are strongly even-aged at canopy level. Most coppices contain two age classes: the former underwood dates from the last cutting, and the former standards generally date from the nineteenth century. Some of the largest standards are those in the remnant coppices of Lady Park and the rest of Highmeadow Woods, where wartime fellings revealed that many dated from about 1800. Along the A466 road through the gorge, the trees on the slopes above the road have all grown steadily over the last 60-100 years to a size at which they are increasingly likely to fall on the carriageway, so precautionary roadside fellings have now become commonplace (Fig. 77). Residents were previously inclined to travel with a bow saw in the boot.
Many Wye Valley trees are inherently short-lived. For example, the largest birches and geans are unlikely to last more than 100 years, and even large trees such as ash and beech rarely live much beyond 200 and 300 years respectively. The elms, which have been ravaged by Dutch elm disease, used to live for around 200 years; most have gone, but a few large wych elms dating from 1870-1900 have survived at least two virulent outbreaks. The lime pollards (Fig. 78), which are still such a feature of old hedges and woodland boundaries around the gorge, may not be as old as they seem. One of the most impressive is a monumental treble-pollard on the edge of Causeway Grove near Redbrook, whose trunks started as shoots on a coppice stool. Now with girths of 3.3-4.3m at breast height, they supported an immense overgrown crown until the Woodland Trust had it lopped to preserve it.
The largest and oldest trees are almost certainly the oak pollards in fields and the yews in churchyards, not the largest spreading beeches, which mostly date from the late eighteenth century, nor the impressive sweet chestnut avenue at Bigsweir, which, by analogy with Brampton Bryan Park (Whitehead, 2000), probably dates from the mid seventeenth century. The most celebrated oak was at Newland (Fig. 79), a hollow pollard that achieved 43.5 feet (13.3m) girth at five feet (1.5m) from the ground in 1906, and must have been well over 500 years old when it collapsed in an early May snowstorm in 1955 (Hart, 1966). Another vanished oak, the Moccas oak, reached 36.5 feet (11.1m) girth in 1891: Kilvert thought it was 2,000 years old, and more recent estimates date its birth to 1064, but in truth nobody knows (Harding & Wall, 2000). The Foresters’ oaks in
FIG 78. A recently lopped small-leaved lime pollard in Cadora Woods, sprouting vigorously, despite decades of shade. The trunk would have broken up without such treatment.
Wentwood marked the spot where the medieval Forest Courts were held, but by 1875, when they had achieved girths of 19 feet (5.8m) and 23 feet (7.0m), ‘time [seemed] to have gathered all his storms against them…The upper portion of each trunk [was] torn away, many branches [were] snapped, but the great bodies [stood] firmly, holding out tortuous arms covered with crisp foliage’ (Linnard, 2000). Today, the largest may be the Pontrilas or Jack o’ Kent’s oak at Kentchurch Park, which was measured in 1997 by the Royal Forestry Society at 11.35m girth and was estimated to be 954 years old (Fig. 80). The very act of repeatedly lopping the crown branches prolonged the lives of such trees, as we learn when they now collapse under the weight of overgrown crowns.
The largest churchyard yews may be even older than the oldest oaks, and we have their ages certified by David Bellamy to give us confidence. The great hollow yew at Hewelsfield, 6.20m in girth at breast height (gbh), could well be 1,300 years old, but what of the 7.20m gbh King Yew in Tidenham East Wood, the 9.10m gbh yew at Much Marcle, the Peterchurch yew that was 8.50m gbh in 1847 (TWNFC, 1866, p. 247), and the Kentchurch yew that Linnard (2000) measured at 11.32m in 1997?
Even yews as old as this are unlikely to be the oldest individual trees. Consider again the great lime pollard in Causeway Grove. Coppice stools can sprout repeatedly, and under traditional coppicing they may have been cut many times, each time growing larger, so the rootstock from which it grew could be any age. Oliver Rackham and Donald Pigott found that the cluster of lime stools in Silk Wood, Westonbirt, is genetically one individual, and the diameter of the cluster could only have developed over thousands of years of repeatedly resprouting after coppicing (Rackham, 2006). The ancient woods of the Lower Wye also include lime clusters: we cannot know their age, but some could be individuals that started to grow in the Bronze Age forests.
FIG 79. The Newland oak in the early twentieth century, with agricultural workers of all kinds ranged below. Once one of the largest oaks in Britain, it died in the 1950s, yet remains as a well-rotted hulk, beside which one of its progeny was planted as a reminder. Trees like this were once lopped for fuel and browse, but they have almost faded from the scene.
FIG 80. Jack o’ Kent’s oak in Kentchurch Park, a pedunculate oak, now the largest-girthed tree in the Lower Wye. It was once much taller, but the upper crown has died, leaving a bleached column of dead wood and a circle of vigorous lower-crown branches. This process of retrenchment can be seen in other oaks in the park, usually at a less advanced stage.
THE CAST OF CHARACTERS: TREE AND SHRUB SPECIES
The Lower Wye contains nearly 60 native tree and shrub species, but not all of these occur in woods. The count cannot be precise, because the line between small shrubs (e.g. red currant) and shrubby ground vegetation (e.g. bilberry) is difficult to draw; the whitebeams could be counted as a single variable species or as a cluster of distinct microspecies (see Chapter 9); and the status of some species is ambiguous. Whatever the uncertainties, the great majority of British natives grow in the Lower Wye, and each species exhibits a particular combination of characteristics that enable it to play a distinctive role in the woods and the landscape as a whole.
In order to understand how the woods work, it is useful to group trees and shrubs on the basis of their size, longevity, rate of growth and tolerance of shade. The most conspicuous are the long-lived forest dominants, beech, oak, lime, elm, ash and field maple. Together with the naturalised chestnut and sycamore, these make up the bulk of most woods, i.e. the most abundant constituents of the canopy layer. Then there is a group of fast-growing, short-lived, slender trees –birch, gean, alder, aspen and sallow – that are quite capable of growing into the canopy. These tend to be intolerant of shade, but they survive by quickly colonising clearings and neglected ground outside woods by means of copious seed production, efficient mechanisms of dispersal and rapid early growth. Thirdly, there is a group of small, shade-bearing trees that exceptionally grow into the canopy, but which usually hang on to the margins and lurk below the taller trees. These include the crab apple, hawthorns, hazel, holly, rowan, whitebeams, wild service-tree and yew.
Below and around the trees is a mixed bag of shrubs, with a wide range of behaviour. Some, such as alder buckthorn, privet, red currant and guelder-rose, tolerate shade and grow happily with bramble and tree saplings in the lowest layer of the woods. Others, such as broom, occupy clearings and survive shade as dormant seeds. Most – including roses, wayfaring-tree, blackthorn, dogwood, spindle and buckthorn – grow better in gaps and on margins, but hang on in shade, growing slowly. Then we have three woody climbers. Traveller’s-joy is a colonist of gaps and margins, like birch. Ivy, like forest dominants, reaches into the canopy of large trees in mature woods, whilst honeysuckle grows in shade, but flowers on margins, like a shrub. Finally, there are some trees and shrubs that have little place in modern woodlands, notably black poplar, grey poplar and a host of tree and shrub willows, which may once have been part of floodplain woodland.
The most important trees in the woods, ecologically, scenically and commercially, are the forest dominants. Three of these – beech, large-leaved lime and small-leaved lime – are shade species. They grow well in full light into huge trees, but are also capable of surviving in deep shade. Beech is the commonest forest dominant in the whole of temperate Europe, but in the Wye Valley it is near the edge of its European range. Curiously, the local beeches have a strong tendency to generate small shoots from the base of the trunk and suckers from roots near the trunk, rather like the American beech. The graceful small-leaved lime is infrequent in much of Britain, but in and around the Wye gorge it is common, and in some woods it is abundant. Its close relative, the large-leaved lime, is one of Britain’s rarest native trees, but the southern Welsh borderland is this species’ headquarters in Britain, and it is actually quite common in Chanstone Woods in the Golden Valley and in some woods on limestone, such as the gorge-side stands in Lady Park Wood and the Dowards. Except in the area around Trellech that was once Wyeswood, from which the commoners’ cattle have presumably eliminated them, the abundance of limes is a particular feature of Wye Valley woods.
The oaks represent the other end of the behaviour scale. Both the sessile oak and pedunculate oak require well-lit situations to regenerate: oaklings are most likely to grow into trees if they start in a clearing, on a margin, or outside woods altogether. In St Briavels, for example, oak saplings are virtually unknown in the woods, but abundant in neglected grasslands and difficult to keep out of rockeries. However, together with beech and lime, oaks form the largest trees and are capable of outliving their competitors. The majority of oaks in the gorge are sessile oaks, and originally this was probably the main native species, even on limestone. Pedunculate oak was widely planted in the eighteenth and nineteenth centuries, but is undoubtedly native on the deeper soils that are now largely used as farmland. Sweet chestnut, a close relative of oaks, bears more shade and thus regenerates better in mature woodland.
The remaining forest dominants are intermediate. Ash, wych elm and sycamore all regenerate in shade, but they are good colonists, producing copious seed, and their growth accelerates strongly in full daylight, so their behaviour in woods has much in common with birch and its like. Although individual trees can live to well over 200 years, they generally die younger than oak, beech and lime. Wych elm has been much reduced by disease: this reached the Wye Valley about 1971 and killed most mature elms, leaving just saplings and a few large wych elm in woods and on field margins (Peterken & Mountford, 1998). Its place has largely been taken by ash and sycamore. Sycamore is a naturalised introduction that has colonised most woods, mostly within the last 150 years, and continues to expand to the point where it dominates some woods, such as the southern end of Highbury Wood. Ash, too, has expanded since the nineteenth century (Armitage, 1914), although it is certainly a native. The final species in this group, field maple, forms medium-sized, moderately long-lived trees that grow in shade, but, like wych elm and ash, grow better in the open.
The woods have been embellished (or spoiled, depending on one’s point of view) by numerous introductions. Some came long ago, notably the sweet chestnut, which in the twelfth century was present as a mature stand of trees at The Chestnuts in the eastern part of the Dean. Other well-established species include not just sycamore, but also holm oak on Sedbury Cliffs, horse chestnut, Norway maple and several shrubs, notably the shade-bearing rhododendron and cherry laurel, and the light-demanding buddleia. In the Dean and Trellech plantations, larch, Douglas fir, western hemlock and several other introduced conifers have regenerated naturally.
The status of several Wye Valley species is ambiguous. We have already mentioned sweet chestnut and sycamore, which, though introduced to Britain, have naturalised in the Wye Valley to the point where their behaviour is indistinguishable from the shade-bearing group of forest dominants. Hornbeam may well be native to the district (Chapter 4), but all the trees I have seen appear to have been planted, even those in ancient woods. Scots pine, once a native in distant prehistory, also seems invariably to have been planted, or to have seeded from planted trees. English and narrow-leaved elms, which are infrequent in woods, but common in hedges (Chapter 8) and on the Wye and Lugg floodplain, may be native to floodplain woodland, but Richens (1983) thought they had been introduced in prehistory as a fodder plant.
WOODLAND TYPES
The Wye Valley woods occupy a distinctive position within British woodland. They are part of a ‘borderland’ zone between the concentrations of upland, oceanic types to the west and north and the lowland types to the south and east (Fig. 81). Borderland woods tend to be rich mixtures with a good deal of ash, wych elm, lime, oak and field maple, with hazel, holly and yew on dry sites. Sessile oak commonly occurs on limestone; hornbeam and beech are usually rare and doubtfully native; and the rare large-leaved lime is concentrated on borderland limestones. North of the Wye typical borderland woods are found in the Malverns, Severn Gorge, Peak District, the Magnesian Limestone outcrop and the North York Moors limestones. West, they extend well into Carmarthenshire, and southwest they extend to the Mendips and the limestone around Chudleigh. The particular significance of the Lower Wye is that it is in the only region where borderland woods overlap the native range of beech, and the only part of the region which has retained a high density of both lime species.
The woodland types described in the National Vegetation Classification (Rodwell, 1991a) are the recognised standard, and in most of Britain the NVC works well, but it struggles to do justice to some of the Wye Valley woods, partly because it is largely structured round upland and lowland types, between which many Wye Valley stands seem intermediate. The NVC also incorporates beech types that have counterparts in non-beech types, and this presents particular difficulties in the Wye Valley, where beech is often a fluctuating minority component of a mixture. Another reason is more practical: the distinctive Wye Valley combinations were not sampled as thoroughly as those in, say, the Peak District.
The woods of the Lower Wye were thoroughly surveyed by Ian Bolt in 1978-80. Ian was an amazingly energetic fieldworker who was the first to describe and map woodland types throughout the district. He went on to help me initiate
FIG 81. The distribution of borderland woods in Britain compared with the historic range of beech, as defined by Rackham (1997). With minor exceptions, the Lower Wye is the only area where they overlap.
the Ancient Woodland Inventory, now an essential tool for woodland conservation throughout Britain, but sadly he fell victim to cancer in his 30s. He used my stand types (Peterken, 1981) to describe the woods.
Using Ian’s records as a basis, I have summarised local diversity in terms of nine woodland types (Table 9), indicating for each the species that are usually present, and those that are generally common or abundant. Types 1-5 occupy dry and mesic (i.e. ordinary) sites, whereas types 6 and 7 are found on flushed and wet ground. Collectively, types 1-7 cover the full range of existing ancient, semi-natural woods. They were known by the pioneer ecologists (Moss et al., 1910; Tansley, 1939) as ‘oakwoods’ because most were coppices dominated by oak standards. Today, after a century of wartime fellings and coppice neglect, the density of oaks has declined and other trees dominate most stands, so we are now blessed with beechwoods, ashwoods, limewoods, oakwoods, alderwoods, birchwoods, elmwoods and combinations of these. The particular combination we see today depends on site conditions and management history.
The major portion of the great woods on the limestones of the upper and lower gorge falls within type 1, but the type also occurs sparingly on the Devonian woods and in the Cwm, Salisbury Wood and other woods in south Monmouthshire. The ground flora is inherently rich, but it has been impoverished in the last 25 years by excessive deer grazing, though the spectacular spring displays of wood anemone, bluebell and ramsons have survived. Similar stands are infrequent but widespread on the hard limestones throughout the borderland, but the Lower Wye stands are unique in containing beech as a common native tree.
Type 2 stands are essentially the drier forms of type 1. Concentrated on the cliffs and steep slopes of the upper and lower gorge, with scattered stands in Haugh Wood, they include both the highly distinctive mixture of beech, yew and whitebeam that in May provides such a striking contrast in foliage colour around the crags, and mixed broadleaved stands with a solid understorey of yew in Pierce Wood and elsewhere. The most extreme form of type 2 is the scrub on the cliff faces and rock towers themselves, where drought and instability limit the growth of ash and wych elm and allow shrubs and small trees to come into their own. This is where the rare whitebeams are concentrated, together with roses, privet and other shrubs. Large-leaved lime is also concentrated around these sites in the upper gorge (Fig. 11), where it frequently regenerates from seed, and sessile oak often dominates former coppice on the dry slopes above. The beechwoods on the south-facing steep slopes of the Little Doward, which must be almost the driest woods in Britain, resemble closely the woods of central European and central French limestones. Taken together, the limestone woods of the Lower Wye are some of the richest in Britain.
TABLE 9. Native woodland types in and around the Lower Wye Valley. Types 1-7 summarise the range of types found in ancient, semi-natural woodlands. Type 8 hardly exists, but could develop. Type 9 groups together all kinds of secondary semi-natural woodland. The woodland types in the National Vegetation Classification are fully described by Rodwell (1991a). Stand types are those described in the 1993 edition of Peterken (1981).
Type 3 stands are concentrated on the sandstone slopes of the middle gorge. These are the acid-soil counterparts of type 1, possessing the same dominants as the limestone woods, but they lack lime-loving species such as whitebeams, privet and large-leaved lime (Fig. 82). The bluebell displays reach their stunning peak in this woodland type, though brambles commonly obscure them. A poorer oak-hazel form of this woodland, lacking beech, ash and limes, is widespread in the gorge, the western margins of the Trellech plateau, Highmeadow Woods, Haugh Wood and the scattered woods of Herefordshire. The native mixtures of beech-lime-oak that constitute types 1-3 are virtually confined to the Lower Wye in Britain, but are widely scattered through Continental Europe.
On the impoverished, strongly acid soils of the Conglomerate outcrop around the gorge and margins of the Trellech plateau we find a heathy woodland in which sessile oak, birch, rowan and holly are particularly common, sometimes
FIG 82. Beech grove growing on sandstone in the common woods of the Hudnalls with holly and little else.
accompanied by beech. This is type 4, which was once extensive in Wentwood, Chepstow Park and Hale Wood. Parkhouse Rocks survives as an extreme example of bilberry-sessile oak woodland, but similar stands are quite common as far north as the Kymin, and appear again at Walford and in Haugh Wood. They resemble the ‘western’ oakwoods that are found so commonly on the hillsides of west Wales, though they are not as well endowed with mosses and liverworts. Similar woods can be found as far south in Europe as Galicia.
Ash-oak woods on heavy, poorly drained soils (type 5) are the ordinary woods of Midland England and lowland sites throughout Britain (Tansley, 1939). They are common on the Woolhope Dome (e.g. Lea and Pagets Wood), the Brownstones of Archenfield, Treville Forest and the western edges of the Trellech plateau. Hornbeam dominates many woods on such soils in southeast England and much of Continental Europe, so this is where we might expect any hornbeams in the Wye Valley to be native. As in the type 1 woodlands, bluebell, mercury, wood anemone and bramble usually dominate the ground flora.
Most woods also contain flushed ground, which is kept permanently moist by water flowing in the subsoil. Here, mixtures of ash, wych elm and alder thrive, and this constitutes type 6. Such stands are usually more fertile than surrounding woodland because they receive nutrients from above. Much of the woodland along the Cage Brook falls into this type, and many of the gently sloping woods in northeast Monmouthshire are flushed throughout. Even in the sandstone parts of the gorge, such as Hudnalls, streaks of alder-ash woodland fringing small streams are clearly visible from over the valley. The mini-ravines that these streams have cut into the slope are conspicuously rich in ferns, but they also offer a full range of niches from dry to sopping wet: unsurprisingly, almost any woodland herb species can be found there in small quantity. Throughout the Lower Wye, the slopes of woods that reach close to a floodplain are flushed, forming a zone sporting slippery sheets of ramsons in spring (Fig. 83). Even in the predominantly acid woods of the sandstones, flushing allows ash, field maple and wych elm to thrive in stands that are superficially similar to the limestone woods (type 1). In a few instances, such as the lower slopes of Cadora Woods, flushing is consistent enough to support alder even on steep slopes. Oddly, ramsons also dominates upper steep slopes on limestone around the Newland meander and elsewhere.
The wet woods (type 7) develop where the water table is always high in
FIG 83. Ramsons, a characteristic feature of woods on limestone and on lower slopes of sandstone woods, in the woodland above the Slade Brook. Spring displays of ramsons, anemone and bluebell have so far survived the persistent heavy grazing by deer, but primroses have probably declined, and the summer-growing herbs have considerably decreased.
depressions on floodplains, collection bays around headwaters, and seepage zones at the very base of slopes. In these situations, soils are not only wet, but usually organic and slimy. Most are fertile, for they receive nutrients running off the land. Wet alder woods are common as narrow strips along most headwaters, and they are particularly well developed along the upper Mounton Brook, which has wonderful displays of monk’s-hood (Fig. 84). The Slade and Mork Brooks have fine swampy alder woods (Fig. 85), fed by numerous springs and diversified by braided streams. Some of the drained marshes, such as Ailmarsh and Coughton Marsh, survive as small patches of wet alder woodland. The Big Bogs at Allensmore, for example, is mainly an outgrown alder coppice with pedunculate oak standards over a swamp dominated by meadowsweet, nettles, lesser pond-sedge, marsh-marigold and yellow iris. Wet woods are not confined to valley
FIG 85. Alder woodland by the Mork Brook, one of several well-developed wet woodlands by watercourses draining into the St Briavels meander. Snowdrops are scattered among the alder stools in a natural-looking pattern, and were obviously not planted, but they could have been washed downstream from gardens.
bottoms: even around the gorge, alder is also characteristic of steep slopes where water is permanently running above or near the surface. Many of the small streams draining from the Hudnalls flow through extremely swampy alder wood, even on the very lip of the gorge. More surprising still, alder was once characteristic of the limestone plateau around St Briavels, where alder still grows in hedges and around headwaters.
These seven types span the range of ancient semi-natural woodland. Versions of them presumably made up most of the original-natural forests. They are also the vegetation from which many of the non-woodland semi-natural habitats must have formed. Thus, heaths came from type 4, acid grassland from type 3, neutral grassland from types 5 and 6, limestone grassland from types 1 and 2, and marshes from type 7.
Two further broad types of woodland need to be recognised, of which the first is largely conjecture. Floodplains were partially wooded, but so little woodland remains on the Wye or any other floodplain in Britain that we must guess what its characteristics would have been by analogy with surviving original floodplain woodland in the temperate lowlands of Continental Europe. Depressions would be occupied by wet woodland of type 7, but most floodplain soils are well drained – alluvium that is wet only in floods – and here we might expect ash-pedunculate oak-alder stands with wych elm and English elm. Floodplains also include shoals along the main channel, deposited as the channel tries to migrate laterally across the floodplain. Here we would expect tall willows and perhaps black poplar to be dominant. Collectively, these form type 8 woodland. They no longer exist, but their components survive as black poplars and tree willows on river banks, as suckering elms in the hedges of south Herefordshire, as hops trailing over floodplain hedges, and as reconstituted stands (e.g. Coughton Marsh).
All secondary woodland has been bundled into a single type (9). They tend to be dominated initially by shrubs and then by well-dispersed, light-demanding species, exemplified by birches in the heathland plantations around Trellech and Tidenham, the oaks, hawthorns and blackthorn that colonise abandoned pastures, and the sallows and alder that colonise neglected wet ground. Conversely, they rarely contain limes, service and spindle, weak colonists that are generally confined to ancient woods. Secondary woods span the range of all the previous eight types, but their composition depends less on soil and more on which species happen to grow nearby: there is rarely a consistent match between the type of ground and the combination of species.
Finally, it is worth recognising that some stands do not sit comfortably in any type. For example, the woodland on Sedbury Cliffs is an odd mixture of secondary scrub with holm oak and ancient pedunculate oak coppice with hazel, holly, field maple, privet, aspen and a multitude of service and whitebeams, with ground vegetation containing wood sage, madder, southern wood-rush and goldenrod that fits somewhere between types 2, 3 and 6.
NATURAL WOODLAND
Every bit of Lower Wye woodland has been influenced by people. Even the great banks of native trees along the gorge have been cropped and manipulated for centuries, if not millennia. Suppose, however, that the influence of people had been minimal: would the woods be different? If we can answer this question, we might understand the circumstances in which the local wildlife originated, measure our influence on today’s woodlands, and predict how woods would develop if we left them entirely untouched.
In Chapter 4 we considered what various kinds of sub-fossil remains tell us about the character of primeval or original-natural woodland, i.e. before people became a dominant factor. To this we can add the experience gained from a long-running experiment in the upper gorge. In 1938, partly to allay criticism of afforestation in the Lake District, the Forestry Commission offered ecologists an opportunity to set up research reserves, but the war intervened and Lady Park Wood was the only place in Britain where a reserve was actually established. The driving force was Eustace Jones (Fig. 86), a plant ecologist in the Oxford University forestry department, who regularly brought his students to Highmeadow Woods. In 1944 the reserve was formally established and Dr Jones laid out nine sample transects in which every tree and shrub was to be mapped, measured and recorded. He duly recorded the transects in 1945 and 1955, but then other interests intervened. A further recording was completed in 1977 by Alan Orange, then a school-leaver living in Coleford, and in 1983-5 I completed a fourth recording. The baton then passed to Ed Mountford, who recorded in 1992-3 and 2001-2. As a result we now have curricula vitae of about 20,000 individual trees with which to understand how the wood works.
Lady Park Wood in 1945 consisted of ‘old-growth’ stands that were last coppiced in 1870 and thinned in 1902 and the 1920s, and ‘young-growth’ stands that had been felled for the war effort in 1942-3. After 1944, both were allowed to grow without any management, so that natural developments could be
FIG 86. Eustace Jones with the author inspecting the permanent transects in Lady Park Wood. While teaching as a forest ecologist at Oxford University, Dr Jones established Lady Park Wood as one of the most detailed records of the growth and dynamics of near-natural woodland in Europe. (Photo: Ruth Briggs)
studied. Analysed after 40 years, we found that the number of trees declined, the remaining trees got larger, and the shrub layer thinned out due to shade, ageing and browsing by deer (Peterken & Jones, 1987, 1989; Peterken & Mountford, 1995). As the trees became larger, so the impact of natural and semi-natural disturbances increased. In particular, the record-breaking drought of 1976 killed hundreds of birches in the young-growth and some of the largest beeches in the old-growth (Fig. 87). Moreover, even the beeches that survived stopped growing for a decade. Earlier, about 1970, Dutch elm disease reached the wood and killed most of the large elms on the lower slopes, thereby exposing tall ash trees to increased rates of wind damage. Even short-lived events had long-term consequences. Thus, in 1984, an unusually large summer population of voles destroyed all the small beech saplings that had started to thrive as the canopy opened after 1976. In 1983, a late April snowfall, which lay only for a morning, flattened many weak lime stems, and these were subsequently pinned to the ground by the falling remains of the birch that had died in 1976. If the deer had not browsed all the shoots, these would have rooted, and generated new lime trees. An event that lasted five hours could thus have changed the woodland composition for centuries. More generally, drought, disease and windthrow created gaps (openings in the canopy) where, but for the deer, saplings would now be growing vigorously.
Well before 2000, the old-growth was nearly natural in several respects (Fig. 88). The stature of the stand and the volume of deadwood had built up to levels found in so-called virgin forests in eastern North America and Continental Europe. The stand was punctuated by gaps and consisted of a mixture of large and small trees of many species, all mixed together. The woodland on the steep slopes was far more open because trees on the slopes were more likely to fall over, and it retained far more shrubs than the woodland on the gentler slopes. Examination of the time series showed that in quiet periods the beech increased at the expense of other species, but that ash and lime were least affected by the various disturbances. We concluded, therefore, that there was no single natural composition: rather, the natural mixture would fluctuate between beech dominance and ash-lime dominance according to the chance impacts of various disturbances.
Sadly, Lady Park reserve is overrun by fallow deer, so there is very little regeneration (Peterken & Mountford, 2005). A series of small exclosures shows that deer are now so numerous that brambles are being browsed to oblivion: the wood is becoming a wood-pasture. Moreover, grey squirrels are deforming the beech, thereby preventing it from maintaining its position as a forest dominant. Nevertheless, we have seen enough of natural processes to realise that we can
FIG 87. A late-nineteenth-century beech in Lady Park Wood, seven years after it was killed by the 1976 drought. Several years later, decayed beech snags are still frequent, and many old living beech have trunk scars and broken crowns dating from that natural event.
FIG 88. The old-growth stand in Lady Park Wood in the late 1980s. Only 110 years since this wood was last coppiced it has attained many of the characteristics of temperate ‘virgin’ forests, with large and small trees mixed together, canopy gaps, leaning trees and large volumes of deadwood. This may be the original-natural state of much of the Wye Valley woodland.
never be sure what will happen next, and that unmanaged woodland is far less tidy than managed woodland.
What does Lady Park Wood tell us about the original woods of 5,000 years ago, when much the same mix of tree species was widespread in the Lower Wye (Chapter 4)? One answer might be ‘very little’: after all, the wolves, red deer and aurochs have gone as completely as the Neolithic people, and the reserve is inevitably influenced by what happens in the managed land around it. Until recently, the general belief was that the bulk of the original-natural landscape was covered in dense forests of tall, slender trees, disrupted at intervals by storms, drought and disease that generated large volumes of deadwood (Peterken, 1996). Gaps created by disturbances rapidly filled with saplings, so grassland and other habitats were confined to permanent open spaces associated with cliffs, coasts, floodplains, rivers and marshes.
This view was recently challenged by a Dutch ecologist, Frans Vera (2000), who argued that, despite wolves and human hunters, large herbivores, such as deer and cattle, would have been numerous enough to keep woodland fairly open. In his view, natural forests were mosaics of glades, scrub and groves of trees, constantly fluctuating in pattern, more like wood-pastures than high forest, and it was in these glades that oak and other light-demanding trees were able to regenerate. Vera’s reinterpretation is still being debated, but the evidence from pollen diagrams (Svenning, 2002; Mitchell, 2005) suggests that open woodland was very limited and was most likely to have been present on poor soils and floodplains, leaving the rest of the land to dense forest. The fossil forests on the Severn margins (Chapter 4), where one might expect open areas and wide-crowned trees, tend to support the dense-forest hypothesis, for most trees were narrow-crowned, some were long-lived, many were blown down in storms, and trees dominated the pollen rain, though this may indicate only that their pollen spread further than herbaceous pollen. Moreover, remains of small mammals in the caves of the upper gorge indicated woodland with only small grassy openings. However, we have little to guide us from the bulk of the Lower Wye, and until more evidence becomes available, natural forest is perhaps best envisaged as limited tracts of wood-pasture on infertile, wet and rocky ground in a matrix of dense forest elsewhere.
Another interesting debate relates to ‘future-natural woodland’: what would develop if we left a woodland reserve to develop with minimal human influence? An ancient wood, with all the original constituents still present, might eventually regain the original-natural state, but secondary woods usually lack important original species, such as the limes. More significantly, our woods have been colonised by sycamore, cherry laurel and many other naturalised trees and shrubs; they are grazed by populations of deer that are no longer regulated by large carnivores (though one does wonder how the mysterious black cats stay alive!); and beech, one of the original forest dominants, is compromised by grey squirrels. When all these factors are taken into account, we might predict that any woods that are allowed to be natural in the future are more likely to develop into sycamore-ash-laurel woods than lime-oak woods. Even so, the woods of the Lower Wye stand a better chance of recovering an original-natural state than woods elsewhere, simply because so many ancient woods retain original mixtures of tree and shrub species.
Summing up, the ancient woods of the Lower Wye have been used for millennia. Their current structure has been largely determined by how they have been managed over the last 200 years, but those that have not been converted into plantations still comprise mixtures of the same tree and shrub species that were familiar to our Neolithic predecessors. The appearance of many woods may be more natural now than it has been at any previous historical period, but, even in ‘natural’ reserves such as Lady Park Wood, we will never be able to shut out the influences of modern life completely. Nevertheless, together with the landforms and the rivers, they remain our closest links with the primeval landscape.