sign in sign up
<<
Home , Armston, Assarting, Castle Ashby, Coppingford, Cranford St Andrew, Cranford St John

The History of Tree Health and Tree

Populations in since c.1550.

Tom Williamson

Gerry Barnes

Toby Pillatt

Acknowledgements A large number of people have helped with this project, providing access to documents or providing information and advice. We would like to thank, in particular, the staff at Archives and Local History, the Record Office, Record Office, and the various Record Offices (at , , , , , Sheffield, Doncaster, Barnsley, Hull and Beverley); and Crispin Towell at archives, Northamptonshire. Thanks also to Anne Rowe and Peter Austin, for information about Hertfordshire; Tracey Partida, for access to her research on Northamptonshire enclosure; and to Sid Cooper, Teresa Betterton, Rachel Riley, Richard Brooke, Jim Lyon, Andrew Falcon, Rod Pass, John White, Mark Pritchard, Rory Hart, Peter Clarke, Garry Battell, Nicola Orchard, Justin Gilbert, Andrew MacNair, Jack Langton, Rob Liddiard, Steve Scott and Patsy Dallas.

Contents

Part 1: Report Summary ……………………………………………………………………………… ...... 4

Part 2: Main Report: ...... 18

1. Introduction: trees, and landscapes …...... 18

2. Farmland Trees ...... 51

3. and -Pasture …………………………………………………………………… ...... 78

4. Tree Species in ‘Traditional’ Landscapes’ ...... 107

5. The Development of Trees and , c.1780-1880 ………………………… ...... 129

6. Trees and Woodland since c.1880 …………………………………...... 159

7. Health, Age and Management ………………………………………………………………… ...... 183

8. Conclusions ……………………………………………………………………………………………………………………204

Bibliography and Principal Sources …………………………………………………………………………………...214

2

Abbreviations

Barn - Barnsley Record Office, South Yorkshire

Bev – Beverley Record Office, East Yorkshire

Brad – Bradford Record Office,

Cal IPM – Calendar of Inquisitions Post-Mortem

Cald - Calderdale (Halifax) Record Office, West Yorkshire

CPR – Calendar of Patent Rolls

Don – Doncaster Record Office, South Yorkshire

FC –

FC archive – Forestry Commission archive, Santon Downham, Suffolk

HALS – Hertfordshire Archives and Local History, Hertford

HHA – archives, Hertfordshire

Hull - Hull Record Office, East Yorkshire

Kirk - Kirklees () Record Office, West Yorkshire

Lds – Leeds Record Office, West Yorkshire

NCC - Norfolk County Council

NHRO – Northamptonshire Record Office,

NRO – Norfolk Record Office,

Nrth - Northallerton Record Office,

Shf – Sheffield Record Office, South Yorkshire

TNA/PRO – The National Archive/Public Record Office, Kew

Part 1. Report Summary

This report, the result of a two-year research project jointly funded by DEFRA and the AHRC, is divided into two sections. This, the first, comprises a brief summary of the principal findings. The second, much larger section presents the arguments, evidence and data upon which these are based. The aims of the project were to:

• Examine the changing character of tree populations in England since c.1550 in terms of the numbers of farmland trees and the number and extent of woods; the distribution of both at various spatial scales;, their species composition and their management. • Cast light on past attitudes to tree health in England and on historic patterns of tree morbidity.

The prime, underlying motivation for undertaking this research was a belief that, in order to understand current threats to trees in England and evaluate their true scale, we need to view them in historical perspective. In particular, it is important to assess the extent to which rural tree populations are essentially ‘natural’ in character, as opposed to being culturally determined; and to identify any developments over the last century or so which may have rendered them more susceptible to the attacks of pathogens.

The project was based on a systematic examination of early printed texts on forestry, farming and land management. But it also – and more importantly – involved extensive archival research, using maps, deeds, estate accounts, surveys, descriptions, correspondence and much else. For this second aspect, our attention was firmly focussed on four counties, largely chosen because they represent a wide tranche of landscape types and of post-medieval social/economic systems, as well as displaying contrasting geological, climatic, edaphic and topographic characteristics.

Northamptonshire was chosen because it possessed an archetypical ‘champion’ landscape, of the kind found widely across the central in the medieval and earlier post-medieval periods. Most of the land lay, not in fields bounded by walls or hedges, but in ‘open fields’ comprising the unenclosed and intermingled strips of large numbers of farmers. Such fields were managed collectively, and farms and cottages were closely clustered in nucleated villages. These open landscapes were gradually replaced by networks of hedged fields in the period between 1550 and c.1760, and more rapidly thereafter, through various forms of enclosure. Northamptonshire was, and is, a county of varied geology, although largely dominated by heavy clays, on the most extensive and difficult tracts of which the three royal of Rockingham, Salcey and Whittlewood were found. Other aspects of the county’s landscape and agrarian history had already been exhaustively studied during an earlier AHRC –funded project based at the University of , and the data and GIS datasets produced by this provided an ideal context for our enquiries. Norfolk, the second county chosen for detailed investigation, was also in part a ‘champion’ county, although here such landscapes were mainly concentrated on the lighter and permeable geologies where, by the seventeenth and eighteenth century, extensive tracts of heathland also existed. The south and east of the county, in contrast, where heavier clays and loams occurred, experienced a considerable degree of early enclosure – even in the Middle Ages much land here lay in enclosed fields – and woods were abundant. In was, on the whole, a wealthier and more populous

county than Northamptonshire, and its county town, Norwich, was the second or third largest population centre in England. Hertfordshire was different again. Little open field survived here by 1550, so that hedges and hedgerow trees were abundant, and large amounts of both enclosed woodland, and wood-pasture, also existed. The county also lay within a short distance of , something which had an important influence on the management of its trees and woods. Lastly, Yorkshire – comprising its three traditional ‘Ridings’ – was, like Northamptonshire, a mainly ‘champion’ county, its clay Vales and chalk Wolds largely farmed as open-fields from nucleated villages. But it also included large tracts of the western Pennines, and the North York Moors, where typical ‘upland’ landscapes, featuring large areas of peat moorland, could be found. Yorkshire is also distinguished by its early industrialisation. By the sixteenth century the south Yorkshire coalfield was already being exploited on a large scale, as were the mineral reserves of the Pennines.

There were important geological, topographic, climatic and edalphic differences between these four counties, but of greater significance in shaping the character of their trees and woodlands were human influences. One of the principal arguments of this report is that tree populations were essentially, or at least substantially, artificial in character: there was relatively little about them that was ‘natural’. In a densely settled and intensively-exploited landscape trees were either deliberately planted or intentionally permitted to grown in particular locations. Trees and woods were thus intimately connected with complex social and economic systems and it was this, as much as the influence of purely natural factors, which explains the variations which they manifested not only over space, but also over time. Pre-industrial tree populations in woods, on commons or on farmland were, moreover, much more intensively managed than today, and in most cases trees were regarded exclusively or at least primarily in economic terms. Timber was required in large quantities, but more important was the need for wood, especially for industrial and domestic fuel.

Farmland tree populations before the later eighteenth century

Farmland trees in the period before the later eighteenth century were, by the standards of today, extraordinarily numerous in many districts. In the old-enclosed areas of Hertfordshire and Norfolk, where hedges formed a dense and intricate mesh, there were usually – by the time our records become abundant, from the later seventeenth century – an average of more than 15 farmland trees per hectare; often more than 20; while local densities of 50 or even more were not uncommon. Such trees were generally scattered fairly evenly across the landscape, growing either within hedges or in ‘rows’ around field margins. In ‘champion’ open-field districts, in contrast, there were far fewer hedges and thus fewer farmland trees, although rather more than is normally assumed. Very large numbers existed in the tofts and yards of the village ‘envelopes’, and others could be found growing in the wider landscape, on wastes and roadsides, in hedges planted around the margins of the great fields or on parish boundaries, and on floodplain meadows, where large numbers of willows often grew. Averaged across the area of individual parishes tree densities in unenclosed landscapes could reach 4 or 5 per hectare but for the most part, in the Midland ‘core’ of the champion – exemplified in this study by the county of Northamptonshire - they were generally between 1 and 2.5 per hectare. These trees, being strongly concentrated in particular parts of the landscape, provided rather lower levels of ‘connectivity’ than in old-enclosed districts.

In the course of the sixteenth, seventeenth and eighteenth centuries open fields were gradually enclosed, and replaced by networks of hedged fields. Yet this did not mean that the density of trees in these areas necessarily increased to the kinds of levels just described in districts of ‘ancient

5 countryside’ like Hertfordshire. Some early enclosures – because they led to the destruction of settlements to make way for sheep farms, and thus to the removal or neglect of the hedges around tofts and crofts – seem to have led to a net reduction in tree densities. Even where enclosures were associated with the continuation of arable-based farming systems, and the survival of villages, trees numbers seldom rose, within the main areas studied here – Northamptonshire and Yorkshire - much above 10 per hectare. Often there were less than 5 trees per hectare and in the Vales of York or Mowbray, less than 2.5. This was part of a wider pattern, and one which was not primarily related to any distinction between ‘woodland’ and ‘champion’ landscapes, for even long-enclosed districts in Yorkshire, on the fringes of the uplands or in the Dales, seldom boasted tree densities exceeding 4 or 5 per hectare. In other words, by the middle of the eighteenth century most parts of Norfolk and Hertfordshire boasted very high numbers of farmland trees; newly-enclosed areas of Northamptonshire medium numbers; and Yorkshire very few.

The management of farmland trees displayed both broad similarities across the various landscapes studied, and also regional variations which, to a significant extent, mirror those in overall tree densities. The broad similarities relate to timber trees. In all areas these were, before the midnineteenth century, almost invariably felled young. Almost all were cut down before they were seventy years old, and often long before: trees were frequently harvested when they were only thirty or forty years of age. This form of management seems to have been dictated by a desire to minimise sawing: trees were thus harvested at sizes appropriate to the dimensions of the timber required for a particular task. It was also probably motivated by a reluctance to tolerate very large, spreading trees on farmland or in village closes; by an awareness of the fact that, after eight decades, growth rates of most of the relevant species begin to slow; and, in the case of , by a desire to maximise the production of bark, which was most easily peeled, and was of better quality, from relatively young trees. Bark, used in the tanning industry, formed a major part of the value of timber trees, especially in Yorkshire.

It was in the numbers of pollards – trees regularly cropped to produce a harvest of ‘poles’ suitable for firewood, fencing and other domestic uses - that the most striking differences between sample areas are apparent. In the old-enclosed districts of Hertfordshire and Norfolk over 60 percent of trees were generally managed in this way, and in most areas at least 70 percent. In some places, such as south Hertfordshire, the numbers were higher still, reaching 80 percent or more on some properties. In Northamptonshire, in contrast, even after enclosure, the proportion of pollards was significantly less, generally around 25 percent and only rarely exceeding 50 percent. In Yorkshire the situation is more complex and the data less abundant: there are signs that in the sixteenth and seventeenth centuries there may have been quite high numbers of pollards in some districts, most notably the Dales, although in the ‘champion’ Vales the numbers of pollards was probably comparable to that in Northamptonshire. But by the eighteenth century pollards were a rare feature everywhere in Yorkshire, scarcely mentioned in documents. This rank order broadly mirrors that of the density of farmland trees, with parts of Hertfordshire (especially the south of the county) at the top of the score table, old-enclosed areas of Norfolk second, early-enclosed districts of Northamptonshire third and with Yorkshire coming a poor fourth.

The correspondence is explained by the fact that, to a significant extent, differences in the number of farmland trees were simply a function of how many pollards were present. Numbers of timber trees, that is, were broadly comparable in all enclosed districts, at between 2 and 6 per hectare, but to varying extents in different areas farmers and landowners added pollards to this base level,

6 sometimes to a degree that must surely have impacted on the yields of crops and grass in the adjacent fields. Their main reason for so doing was unquestionably to produce fuel: the character of local fuel economies critically shaped farmland tree populations, just as it had a determining (if often now neglected) influence on other aspects of the rural environment, such as the management of . Where other sources of fuel were abundant, the need for pollards was simply less.

Yorkshire thus had abundant areas of upland moors, and access therefore to substantial supplies of Sphagnum peat; large tracts of lowland peats existed in the Humber wetlands; while the county’s principal coalfields were already in production by the end of the Middle Ages. Yorkshire had little need for firewood, and thus few pollards, at least by the later seventeenth century. Northamptonshire communities also had some access to coal from an early date, especially from the Warwickshire coalfields, and some of them could exploit peat in the great Fens to the east, or the woods and wood-pastures of the three royal forests. Nevertheless, a significant proportion of firing needed to be supplied by trees growing in villages and, more sparsely, in the open-fields themselves; and as enclosure of open fields gradually proceeded through the sixteenth, seventeenth and early eighteenth centuries pollards were accordingly established in the new hedges, although at a declining rate as improvements to transport infrastructure progressively lowered the costs of bringing coal into the county.

In the old-enclosed areas of Norfolk and Hertfordshire, in contrast, wood played a far more important role in local fuel economies. In Norfolk, peat was locally available (even after the draining of the Fenlands in the seventeenth century) and heaths supplied a range of combustible materials, but coal was only available on the coast, or beside the principal navigable rivers, and the population was dense. Not surprisingly, pollard numbers were high, varying slightly according to the availability of alternative fuels. North-east Norfolk was reasonably well endowed with heaths and fens in the seventeenth and eighteenth centuries and accordingly had noticeably lower numbers of pollards than , where few such alternative fuel sources existed. Hertfordshire had no peat, few heaths, and was land-locked with only one navigable river. Although well-wooded, it made economic sense to fill the hedges with pollards. Here, however, there was another factor at work. The highest numbers of farmland pollards were to be found in the south of the county, reflecting the distorting effect on the market for wood and charcoal provided by demand from London, only 15 kilometres or so to the south.

The character of local fuel economies were thus crucial in shaping the management and density of farmland trees, although it is likely that other economic factors may have played a part, including local styles of vernacular architecture. Small pieces of oak wood, as opposed to timber, were used in the construction of timber-framed houses – for scantling and studs – and it is noteworthy that areas in which pollards were relatively rare also tended to be ones in which, by the seventeenth century, farms and cottages were widely constructed of local stone. Either way, it is evident that the character of farmland trees was intimately connected with that of the wider economies of particular regions and districts.

Woods and wood-pastures

The various counties, and their geographical subdivisions, also displayed much variation in the amount of woodland and ‘wood-pasture’, or grazed woodland, which they contained. To some

7 extent these were a function of complex social and economic developments during the millennium leading up to 1550, rather than of influences operating in the post-medieval period.

It is generally assumed that ‘champion’ areas were poorly-wooded whilst old-enclosed, ‘woodland’ districts were extensively wooded, but to some extent woodland of various kinds was simply more clustered in champion areas, especially in royal forests. In old-enclosed or ‘woodland’ districts, in contrast, such as Hertfordshire, it was more evenly distributed. Although estimates are difficult to make, given the nature of the evidence, ‘woodland’ Hertfordshire and ‘champion’ Northamptonshire may thus have had a broadly similar proportion of their land area – around 10 and 8 percent - occupied by woodland of various kinds in c.1550. Norfolk, for reasons largely lost in the medieval or earlier past, had around 7 percent and Yorkshire, perhaps in part once again because of an abundance of alternative fuels, probably as little as 3 -4 percent, although the different Ridings displayed significant contrasts in this respect, with the East Riding having very little woodland and the North and West Ridings rather more. Indeed, in all counties there were marked contrasts in the density of woodland from one district to another, although overall these followed broadly similar patterns. Areas of light chalk, or sands, deforested for the most part before the Middle Ages, had very little woodland of any kind by 1550 (the ‘champion’ areas of ; the Yorkshire Wolds; west Norfolk). Woodland was also sparse on the highest moorlands in Yorkshire, clustering instead on the more sloping ground on the margins of the main upland masses. In the Midlands and the south woodland was especially abundant on heavy boulder clays, although also - usually in the form of wood-pasture – on acid, gravelly soils, especially where these occurred on high, remote watersheds. It is almost impossible to estimate the extent of woodland and wood-pasture combined in c.1550 but, averaged across the four counties, it is unlikely to have amounted to more than c.5 or 6 percent of total land area.

The distinction between enclosed woodland, managed as coppice, and wood-pasture or grazed woodland, was well-established by 1550 in all the areas studied, and in many districts the two varieties of woodland displayed markedly different distributions, with the greatest concentrations of wood-pasture generally occurring on high watersheds, especially in Northamptonshire (in the three royal forests) and Hertfordshire (on the high ground in the south of the county, and on the crest of the Chiltern Hills), but with enclosed coppiced woods generally located in more convenient areas, closer to the principal settlements. Some wood-pastures were found within private deer parks but most were on common land. In the period after 1550 a significant proportion of the former were converted to agricultural land or to more open, ornamental grounds; while much of the latter gradually degenerated to open pasture, due to poorly controlled felling and the intensity of grazing. This said, the speed of the latter process has been exaggerated by many historians, influenced by myths relating to the poor management of communal resources. Organisational structures which allowed portions of common land to be fenced off to protect young planting existed in most if not all areas: common wood-pastures were, in relative terms at least, a sustainable resource. They survived best into the eighteenth century where particularly strong organisational structures existed, either under lordly direction or controlled by commoners; and also where there was a good market for fuel. South Hertfordshire, where both these circumstances applied, thus retained large tracts of wooded common into the eighteenth century. The wood-pasture sections of the Northamptonshire forests in contrast, poorly if not chaotically managed and regulated, became increasingly denuded of trees.

8

Private coppiced woodland was more intensively and commercially managed than wood-pasture, either by owners of by their lessees. Livestock were rigorously excluded for much or all of the time and this, as much as regular per se, crucially shaped aspects of their ecology, especially the presence within them of a particular range of vascular plants, the so-called ‘ indicators’. Here, as in the wider countryside, timber trees were usually felled when scarcely mature but in some woods, at least, a proportion were allowed to reach a greater age – 100 years or more – to provide large timber for specialised uses. Some of the material cut from the coppices provided domestic fuel but most probably had specialist uses and the character of woods – the length of coppice rotations, even the species composition - was thus strongly influenced by the character of the local economy. Rotations in Yorkshire were, for example, commonly twice the length of those practised in the southern counties: larger poles could be used for props in the local mines, and also produced better charcoal and whitecoal, and (in the case of oak) more bark, all in high demand in this industrialising area. Whatever the precise form of management, coppices seem mainly to have provided specialist, high-quality wood for local crafts and industries. Firewood faggots were, in most contexts, a secondary product.

In many districts there are signs that a reduction in the area of coppiced woodland took place over the two centuries following 1550, possibly amounting to as much as 20 percent of woodland area in Norfolk and Hertfordshire. Because woods were private property they could be grubbed out and the ground they occupied converted to other uses by their owners, in whole or part, in response to perceived short- or medium-term economic advantage. Wood-pastures likewise declined, through gradual degeneration or deliberate destruction, so that the proportion of the four counties combined occupied by woodland may have fallen from around 5 percent in 1550 to rather less than 4 by the second half of the eighteenth century. It must be emphasised that, given the nature of the sources, these figures are speculative: but they are broadly correct, and highlight very clearly the poorly-wooded character of the English landscape in the post-medieval period.

Species composition

Given this, it is unsurprising that coppiced woods and most wood-pastures were intensively managed environments, the former exclusively producing timber and wood, the latter also supplying grazing and other products. Nor is it surprising, given the scale of demand for fuelwood especially, that the wider countryside was also, in many enclosed districts, characterised by often dense populations of intensively-managed trees. But equally important is the fact that these economic imperatives ensured that the kinds of tree species found in both woods and on farmland differed in many respects from those which had been common in the ‘natural’ woodlands which had existed before the impact of farming. The range of species trees present in the countryside was largely the consequence of human choice, dictated by practical and economic considerations.

The most striking characteristic of English tree populations, considered in the long term, was their overwhelming dominance by just three kinds of tree, oak (either Quercus robur or Quercus petraea), ash and elm (Ulmus procera or U. carpinifolia). In most districts, at least since the seventeenth century, these together constituted between 85 and 100 percent of the trees recorded on farmland, ignoring commons and meadows; and almost all of the timber grown in managed woods. On farmland the relative importance of the three trees displayed a measure of regional variation. In particular, elm was infrequent in the north, for climatic reasons; and while oak was in most districts the most important species in Northamptonshire (and probably other Midland areas) it was

9 routinely the second or third most common. In all districts, however, oak was - by a wide margin - the dominant timber tree in coppiced woodland. Although pollen diagrams from the four counties studied indicate that oak, elm and ash had been important trees in the ‘climax’ vegetation before the adoption of farming, ash in particular was generally present at lower frequencies and elm was much more abundant, at least before the ‘elm decline’ of the early Neolithic. Oak (accompanied by ) dominated the vegetation in Yorkshire, but large areas of birch-alder woodland also occurred on the damper ground; while in the three southern counties small-leafed lime (Tilia cordata) was usually the most important tree and species like pine and yew were locally important, together with alder on wetter ground. The importance of oak, ash and elm in the post-medieval countryside was not simply a consequence of their ability to thrive, unaided, in farmed and settled landscapes. As already intimated, the documentary sources leave no doubt that the vast majority of trees were deliberately planted. Even where self-seeded, their survival was a consequence of deliberate choice. Maple illustrates the point well. It seeds freely in hedges but was seldom allowed to grow into a tree: instead it was plashed or coppiced with the other in the hedge. Oak, ash and elm were selected as farmland trees because they had the greatest range of practical uses, and could tolerate a wide range of conditions, in terms in particular of soils and drainage. Other species were more choosy in their requirements, had fewer uses, could offer little that these three could not provide, or were (like maple) thought to be better managed as coppice, in woods or in hedges.

The dominance of these three key species was not total. On farmland, in almost all areas, other trees were present in small numbers, either self-seeding in hedges and somehow surviving there or, more usually, deliberately planted or maintained because they provided some particular product required by landowners or farmers. The character of these ‘minority’ trees – which included in particular black poplar, aspen, crab, cherry, walnut, hornbeam, maple, lime and holly - seems, on the available evidence, to have displayed a measure of regional variation. Thus in Hertfordshire the most important in the west of the county, on the Chiltern dipslope, were aspen and cherry; but in the east, on boulder clay soils, maple, black poplar and hornbeam were more typical. In addition, lowlying wetlands had their own distinctive range of trees, varying according to type but perhaps influenced by economics. Meadows, especially extensive in Northamptonshire, were characterised by large numbers of willows, accompanied to varying degrees by black poplar; while alders tended to dominate where soils were more peaty and acidic, in parts of Yorkshire and Norfolk especially. More strikingly, wood-pastures – whether on commons, or in the form of deer parks – often had their own range of trees, markedly different from those growing in the surrounding landscape. These were selected and in some cases deliberately planted because they had good resistance to grazing and browsing, or because – in deer parks – they provided feed for the deer. Hornbeam and , usually rare as trees in the wider countryside and as timber trees (as opposed to coppiced underwood) in enclosed woods, were classic species of wood-pastures.

We have, perhaps, exaggerated the anthropogenic character of tree populations in the foregoing paragraphs. Soils and climate also played a crucial role in deciding which trees grew where. But the essentially artificial character of tree populations is worth emphasising simply because it is so often assumed, or implied, that the kinds of tree species found in woods, wood-pastures and on farmland were largely the consequence of environmental factors. In reality, even the underwood in coppiced woodland was deliberately modified in some places, even by replanting, or developed in particular ways as a by-product of human management. The hornbeam coppices so common in the woods of south Hertfordshire, and south Norfolk, were very probably the consequence of human intervention

10 and management, driven by a need to supply major population centres – London and Norwich respectively – with charcoal for fuel.

Although the time period covered by this research project begins in 1550, in reality much about our trees and woodlands remains unclear until the later seventeenth or even eighteenth century, due to the paucity of documentary sources. Given the close connections between the character of tree populations and local social and economic systems, it is quite possible that the particular balance of species, and the particular modes of management, which we have derived from maps and documents may have been different in earlier periods, under different social and economic circumstances. The gradual expansion of coal use, in particular, may well have engendered greater regional variations in the extent of pollarding than existed in earlier centuries: there are signs that the contrast between Yorkshire, and the other counties studied, in this respect was less marked in earlier times. Nevertheless, many of the patterns and processes suggested by seventeenth and eighteenth-century documents - such as the distinctive character of wood-pasture trees, or regional variations in the importance of oak as a farmland tree – do appear to have been true in earlier periods, and were probably stable over long periods of time.

Changes in tree populations during the late eighteenth and nineteenth centuries

The character of English tree populations changed radically in the period between the late eighteenth and the late nineteenth centuries. There was, from the start of the eighteenth century, a gradual but steady increase in the amount of new woodland planted, as landed estates became larger and more consolidated, and as much common land – usually occupying marginal ground, difficult to exploit in other ways – was enclosed. This ‘great replanting’ has been highlighted by a number of previous writers but to some extent its impact was cancelled out by the continued degeneration or deliberate destruction of wood-pastures, and by the grubbing out of longestablished coppiced woods as changes in agricultural techniques made it easier to cultivate the heavy but often fertile soils that they occupied - most dramatically following the enclosure of the Northamptonshire forests in the early nineteenth century. The overall result, averaged across the four counties, was probably a modest increase in the proportion of the total land area devoted to woodland of various kinds, from perhaps 3.7 percent to around 4.2 percent, in the period between c.1780 and c.1880. Woods tended to be lost from areas of fertile boulder clay, and they tended to proliferate on the kinds of light, chalky and sandy soils, and in upland districts, where they had formerly been rare; wood-pastures of all kinds went into sharp decline. Overall, the period thus saw – to some extent - an evening-out of the distribution of woodland across the landscape, with losses in ancient forests and old-enclosed areas, and increases in former ‘champion’ districts and in areas of marginal land formerly occupied by common heaths and moors. This was part of a wider process of landscape homogenisation which occurred in the period, with the loss through enclosure of extensive and continuous tracts of heath, fen and downland, and the expansion of farmed, domesticated countryside, featuring hedged fields and patches of woodland.

In addition to changes in the extent and distribution of woodland, the later eighteenth and nineteenth centuries witnessed major changes in its character. Except in a few limited locations, surviving tracts of wood-pasture were now destroyed; and while enclosed woods of coppice-with- standards type continued to be established, most of the new planting consisted of , without a coppiced understorey. These, moreover, were planted with a more diverse range of species than was found in old-established woods, including both none-native conifers like spruce and

11 larch, and indigenous or naturalised species which now spread being beyond their former ranges, especially beech, sweet chestnut and Scots pine. In some cases, stands of pure conifers were established, especially on the most marginal land. Much of this new planting was essentially commercial in character but in many cases it was motivated, in whole or part, by considerations of status, aesthetics and amenity, a clear shift in motivation.

The second key change of the late eighteenth and nineteenth centuries was the consequence of industrialisation, and of a growing enthusiasm for agricultural ‘improvement’. The numbers of pollards on farmland declined dramatically, as progressive improvements in transport ensured that coal became the principal domestic and industrial fuel in all districts. The overall numbers of farmland trees also fell drastically, partly although not entirely as a consequence of this change. In old-enclosed areas of Norfolk and Hertfordshire more than 50 percent of farmland trees commonly disappeared, although locally the losses could be much higher, sometimes exceeding 85 percent between c.1780 and c.1880. It is usually assumed that the twentieth century was the key period for the destruction of hedgerow trees but in many districts the scale of decline during the nineteenth century was probably greater. Few districts in the four counties studied could, by the 1880s boast more than 8 farmland trees per hectare, and in most the figure was between 3 and 7.

A third and equally striking change was a rise in the average age of timber trees, both in woods and on farmland, which seems to have occurred from the 1850s. The development of industrial sawmills and better transport systems, together with the declining value of bark as new forms of tanning (employing chromium salts) were developed, all discouraged landowners from felling trees when they were young, instead allowing most to grow until they were eighty years old or more.

Developments since c.1880

If the hundred years between the late eighteenth and the late nineteenth centuries saw radical changes in the character of England’s trees, developments in the course of the following century or so were, if anything, more far-reaching, although in many ways representing an intensification of established trends. Major shifts in social and economic organisation were associated, once again, with significant changes in the character of farmland trees and woods. Key drivers included the increasing scale of timber imports; the break-up of large landed estates and changing land-use priorities; growing urbanisation and an associated divorce from, and romanticisation of, the ‘rural’ and the ‘natural’ on the part of opinion-formers; and a steady expansion in the involvement of the State in land use and land management.

Perhaps the most important development was an acceleration in the establishment of new areas of woodland, initially as a consequence of the activities of the Forestry Commission, subsequently as a result of the bushing over of surviving areas of common land and other rough or derelict land and, in the last decades of the century, as the outcome of government-funded agro-environment schemes and private initiatives, intended to enhance the environment or provide cover for game. Although in the popular mind England’s woodlands may be shrinking, in reality the area of land across the four sample counties devoted to woods and plantations of various kinds has risen steadily since the late nineteenth century, from just over 4 percent of their combined land area to around 5 percent by the middle of the twentieth century, and reaching over 8 percent today. Such woodland – continuing earlier trends – was increasingly concentrated on the more acid, marginal soils.

12

It is striking how this growth in woodland area is seldom recoginsed as a conservation success. Of course, this is partly because, while the area of woodland in England may have increased markedly, its character has in many ways undergone an environmental decline. Much of the new woodland (although to a declining extent, over time) comprised pure conifer plantations, while the active management of coppices in ancient woodland gradually declined through the century and had largely ceased in most areas by the 1960s, with important impacts on the distinctive woodland flora and the species which depend upon it. This said, to some extent the environmental benefits or disbenefits of landscape change depend upon what is being measured and over what period of time: even the creation of pure conifer plantations has benefited certain species, such as the siskin. Either way, it is clear that, considered in historical terms, woodland per se is not in short supply in England. What we lack are, arguably, forms of planting and woodland management which are more clearly focused on the promotion of , and on the provision of a range of social benefits.

The fate of farmland trees over the last century or so has been more complex, and the interpretations set out in the main report are, to a large degree, dependent on definitions. Key factors were large-scale opportunistic fellings following the break-up of large estates in the first half of the century, and during two World wars; the impact of Dutch elm disease; and post-War field amalgamation and hedgerow loss. Less easy to document, but of immense significance, were wider changes in the treatment of farmland trees, with a general reduction in the planting and economic management of timber in hedges, as forestry became increasingly concentrated in woods and plantations. By the 1960s and 70s farmland trees were generally regarded – depending on perspective – as useless encumbrances, unnecessary hindrances to the real business of the countryside; or as timeless, natural adornments of the landscape, to be preserved at all cost. As a result of all this, not only did the numbers of farmland trees decline to very low levels – generally around 1 or 2 per hectare in most of the districts studied by the 1970s – but a high proportion comprised over-mature specimens.

The complication is that on some measures, but not others, this situation was reversed during the following four decades. While individual, free-standing trees in managed hedges have continued to decline, or have at least displayed no marked recovery, a reduction in or cessation of management has allowed some hedges to develop into lines of close-set trees. At the same time new forms of conservation and amenity planting – ‘linear features’ and ‘groups’, to use Forestry Commission terminology, clusters of trees too small to be counted as true ‘woods’-– have appeared in many areas. If the individual trees in these various features are included in the calculation, then the numbers of farmland trees have recovered sharply, returning in some places to their late nineteenth-century levels. As with so many aspects of arboreal history, much depends on what precisely is being measured and how. On a local basis the result of such recent changes in tree numbers can be strikingly counter-intuitive. It is possible to stand in a landscape now largely denuded of hedges, holding in hand a copy of a late nineteenth-century Ordnance Survey map which carefully depicts a densely-hedged landscape richly studded with trees. Yet, because the handful of remaining hedges have been allowed to grow into lines of close-set ash, oak and maple, in simple numerical terms there has been no significant loss – possibly even an increase – in the density of mature trees. The kinds of raw figures presented by some sources to illustrate the well-treed character of the English landscape thus need to be treated with caution. Even where tree numbers have returned to earlier levels, through recent changes in hedge management or deliberate planting, individual specimens generally have a much more clustered distribution within the landscape than in

13 the past, ensuring low levels of connectivity. All this said, in many places – as already intimated – wholesale removal of hedgerows has ensured continued loss of farmland trees, often to very low levels. The pattern on the ground in fact displays much local variation, relating in large part to patterns and types of ownership. Farms now well-endowed with trees, many of them recently planted, can exist beside holdings sparsely populated by over-mature specimens, mainly .

For a variety of reasons, the species composition of farmland trees appears to have grown more diverse over the last century and a half. Sycamore and beech were more widely planted in some districts by large landed estates in the nineteenth and earlier twentieth centuries, and the ‘conservation’ plantings of the twentieth century have often involved a more extended range of species than the oak, ash and elm which formerly dominated the landscape. Of particular note has been the increased importance of maple as a tree (as opposed to a hedgerow ) in the countryside, in part a consequence of deliberate planting, but mainly a result of the localised abandonment of hedgerow management. At the same time, the decline of farmland species like black poplar, aspen, and walnut, already underway in the nineteenth century, has continued apace. Overall, albeit to varying extents, farmland tree populations remain overwhelmingly dominated by oak and ash.

Perhaps the most important change in farmland trees in the course of the last century and a half has been the steady rise in the numbers of middle aged and old specimens. Other than pollards, regularly rejuvenated to some extent by regular cropping, the countryside before c.1850 contained few trees which were more than 70 years old. In contrast, standard trees over 150 years are now common in most areas and examples over 300 years old are by no means infrequent. This development is primarily a consequence of the decline in active economic management, coupled (arguably) with the triumph of sentimental and conservationist attitudes to trees.

In summary, trees and woods have certainly experienced a decline in England, in numbers and in terms of their environmental benefits, in the course of the twentieth century. But the situation is perhaps more complex and nuanced than is sometimes suggested, and the extent of gains and losses may depend, to some extent, on what precisely is being measured.

Health, disease and pathogens

Close examination of both documentary sources, and early texts on forestry, makes it clear that, within the period studied here, no serious, national epidemic of tree disease occurred in England prior to the outbreak of Dutch elm disease in the late 1960s and 70s. This said, it is also clear that poor health in trees was common in the past, and perhaps considered more normal than it often is today. Scolytus, Tortrix, and various forms of rot have always affected trees: but when populations were vigorously managed, diseased specimens were simply felled and sold with the first onset of serious symptoms. There are some signs that the development of forestry – the planting of large numbers of trees of one species on often marginal, acidic land – may have increased the incidence of disease from the early nineteenth century, but it is difficult to dissociate any real deterioration in tree health from the burgeoning interest in and knowledge of pathology associated with the rise of scientific forestry in the same period, and from the increasing optimism about what this could achieve. The twentieth century, however, did unquestionably see a steady decline in arboreal health. The onset of globalisation and the increasing speed with which plant materials and timber could be transported were the key influences, ensuring the arrival of new pathogens, starting

14 with oak mildew in 1908 and the first serious outbreak of Dutch elm disease in the 1920s. The rising scale, and speed, of international trade – possibly compounded by the effects of global warming – have led to further epidemics in the later twentieth and early twenty-first century. Equally important, however, has been a more general decline in the health of trees since the middle decades of the century, with the widespread appearance of stag-headedness, ash ‘dieback’ and oak ‘decline’.

An historical perspective on tree health, besides confirming that the current scale of epidemics is unprecedented (at least in the period studied), can contribute to our understanding of these issues in three key ways.

• It strongly suggests that some of the more general signs of ill health in trees are a function of management and perception, rather than of the arrival of new pathogens. Most of the chronic, ill-defined conditions like ‘decline’ affect trees which are more than 100 years old. The latter would have been rare before the late nineteenth century, when trees were either pollarded – and thus, in many ways, kept in a state of perpetual juvenescence – or were taken down when still relatively young. It is also clear, moreover, that when farmland trees were intensively managed, specimens suffering from ill health were simply felled and sold with the first onset of serious symptoms. Indeed, stag-heading and ‘bleeding’ were taken to be signs that it was time to fell, so that what we perceive as symptoms of illness may, to an extent, be the normal signs of ageing. In the ‘wildwood’, whatever precise form this may have taken, a very high proportion of trees may have exhibited such characteristics. In addition, it is arguable that the more romantic, conservationist attitudes to trees which developed in the course of the twentieth century, amongst a population which was increasingly divorced from the realities of farming or forestry, brought with it an expectation that all trees, left to their own devices, would be ‘healthy’, that this was their ‘natural’ condition: an attitude which we would not apply to members of our own species.

• Secondly, an historical perspective serves to highlight some neglected changes in the wider environment which may have contributed to the susceptibility of trees to disease, most importantly a marked increase in the quantity of dead wood in the countryside, and rising levels of soil nitrogen, potentially damaging to the mycorrhizal fungi so necessary to tree health.

• Thirdly, and most importantly, an historical approach highlights starkly not simply the essentially anthropogenic character of rural tree populations, but also their limited and restricted nature, massively dominated by only three species in the period up to the 1970s, and now by just two. This has ensured that our tree populations have very little potential resilience in the face of future invasive pests and pathogens, such as emerald ash borer. It is true that the artificial dominance of these trees in the landscape – the consequence, primarily, of past economic factors – has been reduced to some extent since the nineteenth century, with more diverse new plantings, and more recently with natural regeneration of species like maple where hedges have been neglected. But more radical attempts to diversify the character of farmland trees are now urgently required, not least in the face of Chalara which might, in the worst case, leave oak as the sole mature hedgerow tree across large parts of the English landscape.

15

Recommendations

Arising from these findings, we would suggest that the following actions might be considered by the appropriate agencies in the future, as resources become available.

• Although raw figures for the numbers of farmland trees present in particular areas often suggest a marked recovery since the 1970s, these are in some ways misleading. Increases are a consequence of small group planting and the outgrowth and neglect of hedges. More planting is urgently needed, of trees which are scattered more evenly across the landscape: this would not only increase biological ‘connectivity’, and would be more in keeping with the long-term character of the English countryside – captured, for example, in Constable’s iconic paintings. • Future plantings should be made more diverse, to provide greater resilience in the face of likely pathogens. In the past, tree populations were largely shaped by economic and practical considerations. We are now obliged to make radical changes to the composition of our farmland tree populations, due to increases in the incidence of invasive pathogens, but we are no longer constrained by such considerations. Amenity, biodiversity, and aesthetics can be the main drivers. • While future planting should continue to prioritise oak; and while we believe that we should also persevere with the planting of ash, at least until the scale of the impact of chalara becomes clear (using the more resistant strains as these become available); these trees should together make up no more than a half or perhaps two thirds of new planting. The remainder should comprise other species, in order to provide a measure of diversification and resilience. • Some authorities currently advocate that diversification should be achieved by the widespread establishment of non-native species, in anticipation of the likely effects of climate change. We would urge caution here: it seems unwise to make such a radical departure, with uncertain ecological effects and clear cultural disbenefits, given that many indigenous and long-naturalised species have ranges which extend far beyond the UK, and that the speed of warming remains uncertain. • Nor should a standard palette of native species be introduced in all contexts, including trees quite alien to the areas in question. We would suggest instead that emphasis should be placed on the planting of the ‘minority’ trees which have long been characteristic of particular areas and districts. Thus in Yorkshire sycamore might be widely planted, and holly, maple, willow and alder more locally: beech, although a relatively recent arrival, is widespread in many districts as a hedgerow tree and might also be more widely established. • In Hertfordshire, aspen, cherry, beech and perhaps apple might be appropriate trees in hedgerows in the west; on the boulder clays in the east, black poplar, hornbeam and maple (the latter anyway widespread as a hedge shrub). In other areas, different ranges of species might be planted, following a limited amount of documentary research. Such species are ‘tried and true’, and evidently well suited to the localities in question. Planting these, rather than an indiscriminate ‘diversity mix’, would ensure a continuing measure of local and regional distinctiveness in the character of farmland trees, so important for sustaining a ‘sense of place’. • In addition, attempts might be made to recreate some other long-term characteristics and features of ‘traditional’ treed landscapes. The willows which were for a long period a major

16

feature of the great Midland floodplains might be replanted on a large scale, and attempts might be made to restore, in some form, the long-lost stands of hornbeam trees which once existed on the uplands of south Hertfordshire. There are many other examples. • In general, our advice is thus that planting should be informed by history, but without simply replicating the character of past tree populations which, as noted, has long been dominated by a narrow range of species. This said, in particular circumstances trees which have never featured widely in farmland might also be increased, in areas where they are likely to flourish, such as small-leafed lime or wild service. • Scientific research might usefully be directed towards assessing whether, as we suggest in the main report, ill-defined conditions such as ‘oak decline’ may in part be artefacts of management, and of perception, rather than simply symptoms of pathogens or the consequence of specific environmental changes. Research might also explore the possible impacts of two other developments noted in this report: the general rise in soil nitrogen levels over the last two centuries, and the increase in the amount of dead wood present in the environment. • We suggest that commercially-managed populations of farmland trees are more rigorous and resilient than un-managed ‘natural’ ones. Some attention might thus be paid not simply to providing incentives to landowners to plant and manage for conservation and amenity, but (at least in part) for reasons of commercial forestry. In some ways, conservationist sentiment may achieve less in the long term than commercial management. Looked at in one way, devices like Tree Preservation Orders may be part of the problems we face, rather than part of their solution.

Concluding remarks Research on this vast and complex subject is continuing, and further funding has recently been received from the Woodland Trust to extend our investigations into Essex and Suffolk. A short description of our preliminary findings appeared in British Wildlife earlier this year, and University of Hertfordshire Press have agreed to publish a book (text to be completed by June) presenting ours data and conclusions in more detail. We are now keen to widen our enquiries into the impact of fuel extraction on the landscape more generally, examining in particular the role of hedges as a source of firewood in the past and the possible implications of this for future countryside management.

Tom Williamson Gerry Barnes Toby Pillatt

17

Part 2. Main Report

1. Introduction: trees, woods and landscapes

Contexts

The research described in this report forms part of a wider raft of projects commissioned by government agencies as a response to recent problems with tree health in Britain. The impact of Dutch elm disease, which devastated the country though the late 1970s and 80, has been followed by a series of further epidemics, including horse chestnut canker, leaf miner, oak processionary moth, larch disease ( Phytophthora ramorum) and – most recently, and most worryingly- ash chalara (Cheffings and Lawrence 2014). All are caused by invasive organisms – fungi, bacteria, or insects – and they have thus, with some justification, been viewed primarily as part of the wider biological impact of globalisation, which has led throughout the world to the spread of a wide range of plants and animals into new environments, often with negative consequences (Rotherham and Lambert 2011). There are also a number of further threats on the horizon including Emerald ash borer, Pine processionary moth and Citrus longhorn beetle. In addition, however, and quite apart from the impact of alien pathogens, there are signs that tree health in England is suffering a more general decline, something manifest in the identification of more complex and diffuse conditions such as ‘oak decline’ and ‘ash dieback’ (Denman and Webber 2009).

The main purpose of our research is to place current concerns about tree health within a broader historical perspective. We are not natural scientists, but we believe that by examining particular environmental relationships, not simply as they are today, but as they have developed over time, we can help to evaluate existing scientific ideas about tree disease, and perhaps help to formulate new explanations and hypotheses. One of our aims has been the simple one of assessing whether current threats to tree health really are as different from anything experienced in the past as is generally assumed; to assess, that is, whether or not similar large-scale epidemics of native trees have occurred during the past three or four centuries. But we have also attempted to model the character and development of tree populations in England over this same period of time, in terms of their age structure, management and species composition, in order to understand whether observable changes may have led to an increased susceptibility to pathogens, introduced or otherwise. The essence of our argument – not entirely new, but perhaps made with a greater emphasis than in previous work– is that there is nothing very ‘natural’ about trees in the countryside, at least in the period since the sixteenth century. Whether growing in farmland, or in woods, tree populations are essentially artefacts of management, and their changing character has always been related, in complex ways, to a broad raft of social, economic, and technological developments. This very artificiality, we further argue, together with a number of key changes in management over the last century or so, has served to render trees in England particularly susceptible to epidemic disease.

The scale of the difference between wild, and ‘economic’ tree populations is long-established, and should occasion no surprise, given that it a feature shared with almost all aspects of the English environment. Heaths, woods and meadows are, in most ways, no more ’natural’ than suburban gardens or inner-city waste grounds (Williamson 2013). Indeed, one indication of how far removed

Figure 1. , Norfolk, is still actively managed by the Norwolk Wildlife Trust.

Figure 2. Woodrising Wood, Norfolk: the overwhelming majority of coppiced woods are now outgrown and derelict.

19 we are from a truly ‘natural’ landscape in England, uninfluenced by human activity, is the fact that natural scientists argue over what precise form this might have taken, with many agreeing with Frans Vera’s ideas – that the pre-Neolithic landscape was essentially characterised by relatively open landscapes, similar to savannah, with only limited areas more densely occupied by trees (Vera 2000); while others still believe that it was largely occupied by closed-canopy (Hodder et al. 2009; Kirby and Baker 2013).

Figure 3. Ash trees, formerly pollarded, in Wymondham, Norfolk. They appear to have been planted when the common here was enclosed around 1815.

As farmland replaced the ‘wildscape’ – whatever precise character this had taken – there was an increasing need to manage trees more intensively, for both timber (i.e., larger elements, principally trunks, which could be used structurally, in the construction of buildings and the like) and wood (i.e., smaller pieces, suitable for use as fuel, fencing, tools, and minor elements of buildings) (Rackham 1986, 65-7; Rackham 2006, 2-5. The extent to which particular forms of management were adopted was a consequence of the character of the demand for woodland products, which in turn was a function both of demographic pressure, and of wider aspects of economic and technological

20 development. The greatest and most regular need in most periods and areas was for wood, rather than timber, and in the pre-industrial period this was principally produced by coppicing and pollarding. The former involved cutting trees or shrubs down to at or near ground level, on a rotation, in order to produce a regular crop of small, straight ‘poles’: it could only really be practised in woods and other locations from which grazing animals could be excluded, or controlled, as regular browsing suppressed the recovery of the plants, leading to their eventual destruction. It was for this reason that coppices were enclosed, usually by a bank topped with a live or dead hedge, less usually by a wall. Coppicing was usually combined with the cultivation of timber or ‘standard’ trees, although these could never be very numerous, or else their canopy shade would suppress the growth of the coppices beneath (Rackham 1986, 65-8). Management of woods by coppicing declined during the twentieth century: while a small proportion of old-established woodland is still coppiced, by conservation bodies, most are now outgrown and derelict (Figures 1 and 2). Pollarding was similar to coppicing but the trees were cut, not at ground level, but at a height of two metres or more, out of range of browsing stock. It was thus practiced in wood-pastures – grazed woodlands of various kinds - although it was also common in other contexts, as for example where trees were grown on farmland. In some areas, most farmland trees were in fact managed by pollarding; but in others, as we shall see, trees grown primarily for timber predominated.

It will be evident already that the growing of trees in cultural landscapes was something that involved conscious choices: trees interfered with other ways of using land while, conversely, other ways of using land might render tree planting problematic. Grazing thus not only made coppicing difficult or impossible; it also militated to varying degrees against the establishment or survival of any kinds of tree (Rackham 1986, 6). Trees in hedges also had to be carefully preserved against damage caused by routine hedge maintenance, and the numbers of trees which might be tolerated in such locations was a fine balance between the often extensive damage they did to crops in adjacent fields, and the economic importance of the wood and timber they provided. The choices made in this respect were rendered more complex by the fact that the income to be derived from planting might only be realised after many years, whereas damage sustained by crops might be more immediate. The fact that the profits to be derived from farming, and from hedgerow timber,often accrued to different individuals raises further complications and difficulties. The management of trees was, in innumerable ways, thus embedded from the earliest date in wider questions of ownership, management and exploitation.

We have approached our key research issues – the incidence of arboreal ill-health in the past, and the changing character of rural tree populations – in two main ways. Firstly, we have examined a range of early texts on forestry and land management, analysing changing attitudes towards trees, their management and health. Secondly, and more importantly, we have undertaken a systematic investigation of all relevant documentary and cartographic material relating to trees and woodlands in four sample counties: Norfolk, Hertfordshire, Northamptonshire, and Yorkshire, the latter divided traditionally into its three ‘Ridings’ (Figure 4). In all cases, the pre-1974 county boundaries were used in research and analysis. Four separate study areas were selected on account of the fact that systems of managing trees, woods and hedges, because they were closely associated with a wider range of economic and agrarian issues, displayed marked patterns of regional variation. As the forester Moses Cook observed in 1676, ‘our ordinary Husband-men will vindicate Their Countrey Husbandry to be better than the next, for indeed Countreys do differ much in the ordering of Trees and Hedges’ (Cook 1676, 101). More than this: different regions and districts of England, in tree

21

Figure 4. The four sample counties. (a), North Riding of Yorkshire; (b), West Riding; (c), East Riding; (d), Northamptonshire; (e), Norfolk; (f), Hertfordshire. management as in much else, followed their own particular trajectories of development, converging or diverging over time in response to two key influences.

The first was the scale of the demand for wood and timber, and of the relative balance between the two. Changes in demographic pressure, including urban growth in neighbouring districts; in patterns of national and international trade and transportation systems; and in the character of local industries; all radically influenced the numbers and management of trees in any district or region. One important aspect of this was the change over time in the demand for wood as an industrial and domestic fuel (Warde and Williamson 2014). As we shall see, coppicing took a wide variety of forms, associated with the specific demands made by each local economy, but in most areas a significant proportion of the cut material was used as fuel. Pollarding, in contrast, was mainly carried out to produce fuel. Regional variations in the availability of other sources of firing – and changes in fuel supply over time, with the advent of large-scale industrialisation in the eighteenth and nineteenth centuries - thus had a major impact on tree management, and on the numbers of trees tolerated in the countryside. And because some areas were more and others less accessible from the principal

22 coalfields, or other fuel sources, their inhabitants displayed very different attitudes to the numbers of management of tree populations.

The second relates to the character of the landscapes in which trees grew. The extent to which individual trees could be planted and maintained on farmland, and to which particular plots of land could be intensively managed to produce wood or timber, were contingent on wider patterns of land management and land ownership. In the Middle Ages coppiced woods were thus invariably privately-owned, part of the manorial demesne and subject to only very limited rights of use and access by the wider community; where woods existed on common land, in contrast, they took the form of wood-pastures. Even in post-medieval times, moreover, woods and plantations were associated with larger estates, and not with the properties of small freeholders. Only major landowners could afford to put tens or hundreds of acres out of cultivation for decades or longer, postponing immediate income for long-term gain. The character of ownership might also have an impact on trees growing in hedgerows and on pastures. Trees interfered with the cultivation of crops and might be less easily tolerated by the small freehold farmer than by the squire; the latter was particularly the case because, where estate land was leased by tenants, as it mostly was, the timber was generally reserve to the landlord, as was often the ownership of the trunk or bolling of the pollards, although the tenant had the right to the wood cut from them – a condition of most post- medieval leases, and a feature of most medieval customary tenures (Rackham 1986, 121).

Equally important was the character of land ownership – the extent to which particular parcels were owned as outright private property. It was difficult to establish trees, or maintain them in the long term, where land was subject to common grazing rights. Conversely, when commons were enclosed, and passed into private ownership, the marginal land they occupied was often peculiarly suited, in economic terms, for wood and timber production. And where open fields – comprising the intermingled unhedged strips of many proprietors, subject to communal grazing and access rights – were replaced with networks of hedged fields, occupied in severalty, the numbers of farmland trees almost invariably increased. The progress of enclosure, in other words, was a major influence on both the extent of woodland and the density of farmland trees.

The four study areas were selected because they are broadly representative of this range of factors and influences. Some contain, or lie close to, major markets of wood and timber; others were more remotely located. While some were, by the seventeenth and eighteenth centuries, characterised by large landed estates, in others the properties of small proprietors predominated. Some were densely settled, others were not. Above all, the counties chosen are representative of the principal landscape types found across England in the early modern period, in terms of the relative extent of open and enclosed ground; and of the chronology with which the former gave way, over time, to the latter. Regional variations in the landscape is a complex topic but it thus needs to be briefly discussed in general terms, in order that the various areas selected for more detailed study can be better placed in context.

Regional landscapes

The arrival of farming in England around 4,000 BCE was followed by several millennia during which – interrupted by sporadic reverses – much of the of grazed woodland was cleared and replaced by cultivated land. The archaeological evidence leaves little doubt that by the time of the Roman Conquest in the middle of the first century CE settlements could be found on almost all

23 soils, albeit concentrated on the lighter loams, especially on the chalklands of southern England. Settlement intensified still further during the Roman period and by the second century, to judge from the available evidence, few places in England can have been more than a kilometre from a farm, villa, or larger concentration of dwellings (Fowler 2002, 16-18; Roberts and Wrathmell 2002, 43). It is uncertain what this means in demographic terms but, while the population of Roman Britain was certainly less than that suggested by the evidence from (1086), of perhaps two and a half million, it may have run it close. Large areas of woodland and open grazing survived, but farmland probably dominated the landscape across large parts of the country. Settlement on this scale implies a sophisticated agricultural system and there is no real doubt that, on the larger holdings at least, a mouldboard plough, capable of turning a proper furrow, was in use (Brown and Foard 1998; Fowler 2002). In the immediate post-Roman period – the fifth and sixth centuries – there was a significant retraction of settlement, onto lighter soils and permeable geologies; but in the course of the Middle Saxon (7th-9thcenturies) and Late Saxon (9th-11th centuries) periods there was a gradual recovery of population, and a re-expansion of farming onto heavier soils (Hill 2000; Fowler 2002, 203; Banham 2010). By the time of the Norman Conquest in 1066 England was thus already, in Lennard’s famous words, an ‘old country’, long past the pioneer stage of colonisation (Lennard 1959, 11). Less than a fifth and arguably as little as a tenth of its area was by this time occupied by woodland, and even this was very different from the ‘wild’ vegetation which had existed before the advent of farming in the fourth millennium BC, having been heavily exploited for centuries for wood, timber and grazing. Large areas of the country, more than half the land area, was under arable, mostly cultivated using a heavy mould board plough pulled by six or even eight oxen.

The population of England was increasing rapidly at the time of the Norman Conquest, and it continued to rise – possibly as much as doubling – over the following two centuries. Tracts of unenclosed grazing and woodland contracted steadily. Much was converted to arable land but portions were also taken into private ownership by the social elite as deer parks – private wood- pastures which functioned as venison farms and hunting reserves – or, in the manner already described, as more intensively managed coppiced woodland. The remaining areas of open land survived as the commons or ‘wastes’ exploited by particular communities, or by groups of communities. By the thirteenth century, when local documents become abundant, we can obtain a reasonably clear view of the organisation of farming across much of England. Most of the land was exploited by peasant farmers. Some were effectively free proprietors, but most paid a rent to a manorial lord, in cash or as labour on his demesne, or ‘home farm’ – the land kept in hand and managed to produce a range of marketable commodities. The peasant farms comprised a house site and associated yards – tofts and crofts - and an area of defined land, mainly cultivated as arable, to which were appended rights to use the non-arable commons of the manor, for grazing and as a source of firing and raw materials. Manorial lords were, by the thirteenth century, recognised as the legal owners of common land but their ability to exploit it was limited by the rights enjoyed by their tenants. Most of the classic ‘semi-natural habitats’ found in England were then common land, including heaths, upland moors, fens and chalk downs, and in addition there still remained many tracts of grazed woodland or wood-pasture. It is a sign, however, of how far clearance of woodland had proceeded by the twelfth century that many of the most extensive tracts were preserved as royal ‘forests’, in which deer were encouraged for the royal hunt through special legal restrictions which served to reduce the rate of ‘assarting’ (woodland clearance), and to regulate other matters –

24 the keeping of dogs, the height of fences – which might impact on deer populations, or on the operation of the hunt itself.

As population continued to climb through the eleventh, twelfth and thirteenth centuries, patterns of fields and settlement developed along increasingly divergent lines in England, leading in time to the emergence of a range of distinct regional landscapes. The most dramatic contrast was between upland areas, where population levels generally remained low, the extent of arable land limited, and large areas of rough grazing survived; and the more cultivated and populous lowlands. But there was also, within the latter, a distinction between what early-modern topographers were later to describe as ‘champion’ and ‘woodland’ districts (Roberts and Wrathmell 2002, 1-3; Rackham 1986, 1-5; Williamson 2013, 125-46). The former, characterised by nucleated villages farming extensive communal open-field systems, occupied a great swathe of central England running from Yorkshire to the Channel coast. Farms comprised numerous separate unhedged strips, each around seven metres wide, which were scattered through the territory of the township, in some cases with such regularity that the same sequence of properties was repeated throughout the fields. For purposes of cropping the strips - ‘lands’ or ‘selions’ - were grouped into bundles called furlongs, and these in turn generally into two or three great ‘fields’, one of which lay fallow each year and was grazed by the village livestock, the dung from which replenished the nutrients depleted by repeated cropping (Hall 1982 and 1995). The other fields would not all be occupied by the same crop – the unit of rotation was actually the furlong rather than the field – but they generally contained crops in the same ‘season’, that is, one field would be autumn-sown and the other spring-sown. In many places these highly communal farming systems, cultivated according to rules articulated in manorial courts and village assemblies, survived into the eighteenth or even nineteenth centuries, when they were removed by large-scale enclosure, often through parliamentary acts. Thus was created the landscape of straight-sided fields, defined by walls or flimsy hawthorn hedges, which is often referred to – following the historical ecologist Rackham – as planned countryside, and which dominates this broad central belt of the country (Rackham 1986, 4-5).

To the south and east, and to the west, lay what early topographers referred to as ‘woodland’ landscapes, sometimes described as ancient countryside by modern scholars. Here settlement was more dispersed in character, with scattered farms and hamlets - many strung around commons and small ‘greens’ - as well as, or instead of, compact villages (Williamson 2013b, 125-46). Often some, occasionally the majority, of the land lay in hedged closes. But open fields of a kind usually existed, although they were rather different in their layout and operation from those found in ‘champion’ districts. They were generally ‘irregular’ in character: that is, the holdings of individual farmers, rather than being scattered throughout the lands of the township, were clustered in particular areas of the fields, usually near their farmstead (Campbell 1981; Martin and Satchell 2008; Roden 1973). Often, although by no means always, communal controls on the organisation of were less pervasive than in the champion. Such landscapes were often well wooded, but the term ‘woodland’ referred, in fact, to the number of hedges and hedgerow trees they contained, for even where open fields were prominent they were often numerous and small.

‘Woodland’ and ‘champion’ are simplified terms, each covering a range of landscapes. There were, in particular, two main types of ‘champion’ countryside (Williamson 2013, 125-146). One, found on light, well-drained land, overlying chalk or sand, boasted extensive tracts of unploughed ground – downland or heath – in addition to the open arable. This was because nutrients were washed out of

25

Figure 5. Landscape regions in England. Different authorities disagree over the details of regional landscape boundaries, but all concur in seeing the central parts of the country as being characterised by ‘champion’ landscapes with village settlement, extensive and highly communal open fields, and late enclosure. (a), the boundaries of Howard Gray’s ‘Midland System’; (b), Oliver Rackham’s distinction between the ‘planned’ and the ‘ancient’ countryside; (c), the ‘Central Province’, as defined by Roberts and Wrathmell (2000); (d), densities of dispersed settlement mapped by Roberts and Wrathmell (2000). these freely-draining soils with particular rapidity, and needed to be very regularly replenished. These were ‘sheep-corn’ districts, in which huge flocks were grazed on the heaths or downs by day, and were by night close-folded on the arable after harvest or when it lay fallow, treading in urine and dung and providing a much-needed injection of Nitrogen and other nutrients (Kerridge 1967, 4351). In such districts the open-field strips or ‘lands’ were usually ploughed flat, and separated by narrow, unploughed balks (Kerridge 1992, 25-30). In the Midland areas of England, in contrast,

26 champion landscapes were mainly associated with heavy clays. Here the individual ‘lands’ were ploughed in broad ridges, to assist drainage. These still survive in places, preserved under grass as the earthworks known to archaeologists as ‘ridge and furrow’. Landscapes like these lacked the great open pastures – the nutrient reserves of down or heath – of the sheep-corn lands. Heavy soils retained nutrients better than light, permeable ones, and close-folding would, for much of the year, have damaged the soil structure (Kerridge 1992, 77-9). The fallows were dunged by the village livestock, but in a less intensive manner.

‘Woodland’ landscapes displayed a corresponding degree of variation, especially in terms of the relative proportion of the cultivated land that lay in open fields and enclosures, and the extent to which farming was carried out on communal lines. In addition, many parts of the country, such as western East Anglia, displayed intermediate characteristics, making their definition as ‘woodland’ or ‘champion’ problematic. Moreover, while in general terms the two kinds of landscape displayed the broad pattern of regional distribution just noted, in detail pockets of ‘woodland’ countryside could be found deep within the ‘champion’, as for example in north , while (more rarely) islands of ‘champion’ could be found within the woodland (Brown and Taylor 1989). Generalised landscape terms like these , in short, a wide range of local and regional variation, although they do represent a useful way of thinking about the medieval countryside.

The origins of this broad division remain a matter of keen debate amongst landscape historians and others, and need not detain us here. The key point is that it was broadly correlated with variations in the numbers and distribution of both farmland trees and woodland. Although, as we shall see, the landscape of the ‘champion’ was by no means as devoid of trees as is sometimes suggested, there were far fewer than in woodland districts, in large measure because there were fewer hedges in which they could grow. They were also more concentrated, and with less overall connectivity, for a disproportionate number were clustered in the immediate vicinity of nucleated villages, where there were small networks of enclosed private crofts. For reasons which are more complicated, but which probably relate to the development of communal land-use systems during the early middle ages, champion areas were also normally less extensively wooded than ‘woodland’ ones. Woods did occur but they tended to be clustered in limited areas – usually on high clay interfluves between major drainage basins, where they often came to acquire the status of ‘royal forests’, with their special laws to sustain deer populations and restrict further woodland clearance.

The period of population growth which underpinned the development of the medieval landscape came to an end in the early fourteenth century, due to livestock disease and climatic deterioration; and then went into sharp reverse from 1348/9, with the advent of the Black Death. The population shrank by at least a third in the second half of the fourteenth century. This led, directly and indirectly, to significant changes in the organisation of agriculture. Depressed grain prices encouraged the laying of much arable land to pasture during the later fourteenth and fifteenth centuries, for prices of meat and other animal products held up better than those for grain; and in all areas farms tended to increase in size, in part as survivors of epidemics took over holdings left vacant by the demise of less fortunate families. These developments, initiated by the late medieval population decline, were also related to wider changes in the organisation of society, with complex causes – and they continued, and to some extent intensified, even as demographic growth resumed in the course of the sixteenth century. The old forms of customary tenure gradually evolved into a range of ‘copyholds’, some of which effectively recognised local lords as owners of their manors, and their tenants as tenants in the modern sense; others providing farmers with a greater degree of

27 security; and some giving them so many proprietorial rights that they effectively joined the ranks of the small numbers of freeholders who had always existed amongst the peasant population. Yeomen farmers, as well as the gentry and aristocracy, sought ways to increase their profits, and the size of their properties, as a more complex and market-orientated economy developed. Farms continued to increase in size, and production continued to become more specialised, leading to the emergence by the middle of the sixteenth century of a pattern of fairly specialized farming regions. While many lowland districts continued to focus on grain production, especially on the lighter land and more fertile loams, others – particularly areas of heavy clay - came to specialise in livestock farming: in dairying, or in the fattening of sheep and cattle which had often been reared elsewhere, in remote upland regions (Kerridge 1973; Thirsk 1987).

The extent to which particular kinds of farming economy could develop in any district was, however, contingent to an extent on the character of its field systems, for an emphasis on livestock husbandry was easier in a landscape of enclosed fields, and more difficult where, as in Midland areas, land lay in extensive and highly communal open fields, and holdings took the form of widely scattered unhedged strips. Where ‘champion’ villages were small and located on particularly difficult clays or other marginal ground, however, and their inhabitants held their land by insecure copyholds, they were sometimes depopulated by manorial lords, usually following a period of more gradual decline, when some farmers had moved to taken up more amenable holdings elsewhere. In many Midland townships large-scale sheep farming thus replaced arable agriculture, with the earthworks of ‘deserted villages’ preserved under the turf. These kinds of unilateral, depopulating enclosures declined during the sixteenth century, in part because the most vulnerable communities had by now been picked off. They were replaced by enclosure ‘by agreement’, in which the principal proprietors agreed to enclose, carried out the appropriate surveys, re-allotted intermixed arable land as consolidated holdings, and divided the commons in proportion to the rights formerly exercised over them (Yelling 1977; Reed 1981). From the middle of the seventeenth century confidence and stability were often given to the new dispensation by bringing a fictitious legal dispute to the court of Chancery, contesting the agreement: once the court had found against the plaintiff the enclosure was deemed secure.

Where – in ‘woodland’ districts – less ‘regular’ open fields existed, these were usually enclosed in a rather different way, in a gradual, ‘piecemeal’ manner. Individual proprietors bought and sold, or exchanged, strips in order to create consolidated holdings which could then be surrounded with a fence, hedge or wall and cultivated on an individual basis. Over time, the open arable was gradually whittled away, leaving only the areas of common grazing which, being areas of mixed use-rights, rather than intermixed properties, could not easily be removed in this kind of piecemeal, individualistic manner (Yelling 1977, 11-29). Piecemeal enclosure was more important in woodland than in champion areas (although it occurred to some extent everywhere) because communal controls on agriculture were less deeply entrenched and, in particular, the intermixture of holdings was less, so that each portion of the open fields contained the lands of few proprietors. Relatively few transactions were thus required to create compact blocks of land held in severalty. This kind of enclosure produced distinctive field patterns, for open-field strips had seldom been dead straight but instead exhibited a slightly sinuous layout, characteristically taking the form of a shallow ‘reversed S’, caused by the way in which the ploughman moved to the left with his team as he approached the headland at the end of the strip, in order to avoid too tight a turning circle (Eyre 1955). Because the new walls and hedges were established along the edges of bundles of strips, they

28 served to preserve, in simplified form, the slightly wavy lines of the earlier landscape. Whatever the precise method of enclosure, the hedges planted in the sixteenth and seventeenth century, like those established in the Middle Ages, generally consisted of a range of woody shrubs. This was partly because it was difficult to source large quantities of hedging thorn in the absence of a commercial nursery industry, and in part because the wood produced from mixed hedges, by routine maintenance (generally by laying or plashing – substantially cutting back the hedge and weaving the remaining stems at intervals of ten or twelve years), produced much firing and small pieces of wood with other uses on the farm (Barnes and Williamson 2006, 73-96; Warde and Williamson 2014).

By the seventeenth century many ‘woodland’ districts on the heavier soils in south-eastern and western England had become areas of mixed husbandry with a strong pastoral bias, with three quarters or more of their land under grass. In the champion Midlands, in contrast, although the soils were well suited to livestock farming, the difficulties of enclosure ensured that arable continued to dominate the landscape, at least in the period up to 1650. We should note, however, that not all ‘woodland’ areas developed pastoral economies. Some, like the Chiltern Hills, remained important cereal growing districts: there was no neat correspondence of medieval landscape structures and field systems, and post-medieval regional farming economies. Either way, during the course of the fifteenth and sixteenth centuries the number of hedges, and thus in all probability the numbers of hedgerow trees, increased, although in ‘woodland’ districts – where enclosure was easier – faster than in the ‘champion’. As already noted, in many of the latter areas enclosure only came with the widespread adoption of parliamentary enclosure acts in the later eighteenth and nineteenth centuries, something which also ensured the removal of the often extensive areas of common land which still survived in many ‘woodland’ districts.

Although there were thus striking variations in the character of the landscape in different parts of southern and Midland England, more dramatic were the contrasts between the countryside found in these lowland areas and in the uplands, or the ‘highland zone’ in the parlance of an earlier generation of geographers. In these the medieval landscape again featured varying mixtures of enclosed land and open fields, often of ‘irregular’ form; and here, as in ‘woodland’ areas, the latter had usually disappeared by the end of the seventeenth century through the gradual, piecemeal enclosure . But in upland districts there were particularly extensive tracts of common grazing on the poorer, higher land, mostly comprising various kinds of moorland, most of which survived until enclosure (usually by parliamentary act) in the eighteenth or nineteenth centuries. In general terms, highland areas of England boasted relatively few areas of enclosed, coppiced woodland, although they often retained tracts of wood-pasture well into the post-medieval centuries. This is probably because low population levels, coupled with the availability of alternative materials – peat and moorland vegetation for fuel, stone for walling and building – made the enclosure and management of coppices in the Middle Ages less economically appealing (Rackham 1976). A more important difference between many upland districts, and the lowlands, however – emerging by the start of the period studied here, but soon to take a more radical form – was the development of an industrial economy, involving the exploitation of water power but also the mining of coal, lead and other minerals. Already, by the middle of the sixteenth century, this was having a significant impact on the management of trees and woodland in districts close to the major coalfields, especially in the north east of the country.

The four counties chosen for study exemplify these various broad landscape types, although most contain more than one. Indeed, only Northamptonshire – a classic example of a Midland, clayland,

29

‘champion’ county – could be said to display a significant degree of homogeneity in this respect. Hertfordshire, it is true, was for the most part an early-enclosed, ‘woodland’ area, but it had a strip of ‘champion’ land, mainly on light soils, running along its northern boundary. Norfolk, as already intimated, was a more complex area – much of it difficult to classify in terms of these simple landscape dichotomies – but its southern and eastern half lay mostly in enclosures by the seventeenth century, while the west was a light soil, broadly ‘champion’’ district with extensive tracts of heath – especially in the arid district known as Breckland – which lay largely unenclosed into the eighteenth and nineteenth centuries. Yorkshire was large and diverse, with tracts of upland moors, interspersed with tamer, more cultivated Dales; but also significant areas of lowland Vales which were, for the most part, clayland ‘champion’ country. In the East Riding lay the chalk Wolds, a sheep-corn district similar in some respects to western Norfolk. Each of the four counties also displayed other significant economic idiosyncrasies relevant to the history of tree populations, including variations in the extent of urban and industrial development. It is, therefore, necessary to provide a very brief account of the landscape history of each in turn.

Hertfordshire

Hertfordshire is one of England’s smaller counties, with an area – before recent reorganisation – of only 1,638 square kilometres (throughout this study, the pre-1973 boundaries are used as the principal framework for analysis). As well as being an archetypical example of old-enclosed ‘woodland’ landscape, it also lay close to fuel-hungry London, providing as we shall see a useful insight into the ways in which the management of trees and woodland might be influenced by the proximity of a major market for fuel and timber. That proximity has, however, also led in recent centuries to extensive urbanisation and suburbanisation, especially in the south and west of the county and also in a broad band through the centre, the latter as a result of the construction of the ‘Garden Cities’ of Welwyn and Letchworth in the early and middle decades of the century, and of the ‘New Towns’ of Stevenage and Hatfield in the post-War years (Rowe and Williamson 2013, 268-94; Ashby and Cudmore 2011). Built-up areas now account for around a third of Hertfordshire’s surface area; only in the east of the county is the landscape still predominantly rural, a circumstance which sometimes renders difficult long-term comparisons of woodland area and farmland tree densities.

Most of the county overlies chalk, although this formation is only exposed along the steep escarpment of the Chiltern Hills (and their more muted north-easterly continuation, the ‘East Anglian Heights’). This line of high ground forms, but only approximately, the county’s northern margin (Catt 2010). Where it extends beyond the escarpment, onto lower ground, geological formations older than the chalk – principally the sticky mudstones of the Gault – are exposed to a limited extent: everywhere else in the county, the geology is characterised by Tertiary and Quaternary deposits. The land falls gradually in height to the south east of the escarpment, and the western half of the long dipslope is occupied by the deposits known as clay-with-flints and plateau drift, which give rise to soils (mainly of the hornbeam and Batcombe Associations) which are acidic, subject to some seasonal waterlogging, and relatively infertile. Before the high Middle Ages this was a sparsely-populated district, with most settlement concentrated in the major valleys cutting through these varied forms of drift, where the underlying chalk is exposed. In the east of the county, in contrast, the dipslope of the chalk is covered with a thick mantle of boulder clay laid down by the Anglian ice, giving rise to more fertile, calcareous but relatively impermeable soils, mainly those of the Hanslope Association. The landscape is, however, well-dissected by the rivers Ash, Rib, Stort, Quin and their principal tributaries, ensuring that much of the ground drains relatively easily, while

30 in the deepest valleys the chalk itself is exposed, all ensuring that even in late prehistoric times this was a well-settled district, and by the time of Domesday had population densities significantly above the national average.

Figure 6. Simplified soil map of Hertfordshire.

(a) Freely-draining, slightly acidic soils formed in sands and gravels and overlying silty drift (mainly Sonning 1, Hamble and Ludford Associations). (b) Soils formed in chalk and ‘head’ (Upton 1, Prior, Wantage 1, Charity 2 and Coombe 1 Associations) (c) Heavy clay soils formed in decalcified chalky till (Beccles Association). (d) Clay soils formed in chalky till (Hanslope Association). (e) Poorly-draining, acidic soils formed in London clay and Pebble Gravels (Essendon and Windsor Associations). (f) Freely-draining, slightly acidic loams. (g) Slightly acid light clays with impeded drainage (mainly Batcombe and Hornbeam 2 & 3 Associations). (h) Alluvium and peat.

The southern termination of the Chiltern dipslope is formed by a band of lower ground running east- west through the county – the ‘Vale of ’ - which extends from Rickmansworth, through Watford, St Albans, Hatfield and Hertford, and is followed by the rivers Colne and Lea. It contains a wide range of glacial and pre-glacial formations (one geologist has described it as a ‘wide trough scoured out of the chalk by running waters ages ago and used later by nature as a sort of rubbish or detritus tip’), giving rise to a diverse range of soils, although mostly well-drained, if not particularly fertile (Gardiner 1976, 100). To the south the ground rises again due to the presence of solid formations later than the chalk, of Tertiary date – the heavy, impermeable London clay - which on the highest ground is capped in turn by deposits of pebble gravel (Catt and Doyle 2010). This is the most agriculturally challenging part of the county, both the London clay and the pebble gravels

31 giving rise to soils (principally those of the Essendon and Windsor Associations) that are seasonally waterlogged, acidic, or both. Until nineteenth-century suburbanisation, much of the land here lay under permanent pasture.

These variations in soils and geology had an important impact on Hertfordshire’s landscape, and on the character and disposition of its trees and woodland. As already noted, a strip in the north, lying on and below the chalk escarpment - from Tring in the west to Barley in the far north east – formed the southern edge of the Midland ‘champion’ zone, and pre-enclosure maps like that for Ickleford, surveyed in 1771, show landscapes of clustered settlement and extensive open fields (HALS DE/Ha/P1). Small areas of this champion district occupied heavy soils, especially on the Gault clays to the north of Tring, and here the open fields were often enclosed and laid to pasture in the course of the fifteenth, sixteenth and seventeenth centuries. But in most chalkland parishes they were only finally swept away by parliamentary enclosure acts in the late eighteenth and early nineteenth centuries, to be replaced by landscapes of large, straight-sided fields in private occupancy, bounded by hawthorn hedges. One of the curious features of the county is that in a number of places - , Bygrave, Wallington and Clothall - open fields still survived, unenclosed, into the late nineteenth or even twentieth centuries, to be visited and discussed by historians like Gilbert Slater and Frederick Seebohm as interesting survivals from the medieval past (Slater 1907; Seebohm, 1883). Most of the county, however, was characterised by ‘woodland’ countryside: settlement took a relatively dispersed form and open fields, where they occurred, were multiple and ‘irregular’ in character, often of limited extent, and were generally enclosed before the middle of the seventeenth century. Hedges, and thus presumably hedgerow trees, must therefore have been numerous, in most parts of the county, at the start of the period studied here.

In addition to having had, for centuries, large numbers of farmland trees, Hertfordshire is (and has long been) one of the more densely-wooded counties in England. Much land was evidently cleared in late prehistoric and Roman times but this was followed, in the early Saxon period, by a significant retraction of settlement. At the time of Domesday vast tracts of wooded ‘waste’ appear to have existed in the county, especially on the poorly-draining London clays in the south, and on the comparatively infertile soils of the Chiltern dipslope in the west. Indeed, many places which had, by the later Middle Ages, developed into sizeable villages with their own parish churches do not appear in Domesday at all: Ridge, Northaw, Totteridge, Elstree and Barnet in the south, Bovingdon, Flaunden, Sarratt, Markyate and Harpenden in the west. The following three centuries saw a massive expansion of settlement and agriculture, however, the county’s population perhaps increasing from around 30,000 in 1086 to over 70,000 in 1307; while the demand of London for agricultural produce probably acted as a further incentive for the expansion of farmland (Bailey 1998, xxii). Any surviving areas of the ‘wastes’ were enclosed as coppiced woodland or as deer parks by the feudal elite, with the remainder becoming greens and commons, tree-covered to varying degrees, which were often surrounded by girdles of farms and cottages.

The embanking and enclosure of woodland was probably already occurring in the county on some scale by the time of the Norman Conquest, in the densely-settled north-east of the county at least, for while Domesday records most woodland in Hertfordshire in terms of the number of swine that could be grazed there, in this eastern district many manors had only ‘wood for fences’ – suggesting that their woods were limited in extent and managed as coppice, with pigs and other grazing livestock excluded. The earliest source to provide any reasonable indication of the overall distribution of woodland in Hertfordshire is Dury and Andrew’s county map in 1766 (MacNair et al.

32

2015) (Figure 7). This demonstrably omits a large number of woods, especially in the east where they were small and numerous, but is nevertheless useful in showing that the somewhat uneven distribution of woodland within the county apparent on later maps, such as the Land Utilisation Survey of 1940, was already firmly in place by the mid eighteenth century (Cameron 1941, 319). There was thus virtually no woodland on the chalk soils of the ‘champion’ districts in the north of the county. Elsewhere, woods were widely distributed, although with particularly strong concentrations on the more difficult clay soils of the Hornbeam 3, Beccles and Windsor Associations (Figure 7).

Figure 7: the distribution of enclosed woodland in Hertfordshire, as shown on Dury and Andrews’ county map of 1766.

Yet the determinants of woodland location were complex, and topography – the pattern of valleys and interfluves – was also important. Managed woodland tended to be located towards the edges of the drift-covered interfluves, close to major river valleys where the main settlements were located: the interiors of the interfluves, more remote, tended to be occupied by common land (Rowe and Williamson 2013, 71-8, 119-130).This pattern, which has been noted by other areas of anciently enclosed countryside - in Sussex and Kent, and in Suffolk - presumably reflects decisions taken by manorial lords or their representatives in the eleventh, twelfth and thirteenth centuries (Witney 1998; Warner 1987, 5-9). Clearance and colonisation were directed towards the interiors of the main upland masses as cultivation expanded, with areas reserved for manorial use, as coppiced woodland, being retained in convenient locations, close to the main existing centres of settlement. Poor soils towards the centres of interfluves – in more remote locations – were thus more likely to be left to the use of the peasants, and those areas which were not taken into cultivation survived as common land (Warner 1997, 20). This pattern is particularly clear on the London clays in the south of Hertfordshire, where woodland was concentrated towards the margins of that formation – beside the valley of the Lea and the Colne – whereas the higher ground was occupied by extensive open

33 commons (Figure 8). It is a strong sign of the extent to which, even in the early Middle Ages, woodland and its distribution were strongly structured by social and economic influences.

Figure 8. The distribution of ancient woodland in south Hertfordshire. Typically, the main areas of enclosed woodland were concentrated towards the margins of the main areas of intractable, upland soils. The less accessible interior areas were occupied by tracts of common land, some of it still wood-pasture at the end of the Middle Ages and some of it open heath.

As we shall see, Dury and Andrews suggest that around 6 percent of the county’s surface area was by then occupied by enclosed woodland and, in spite of its many inaccuracies, this overall total is probably approximately correct. Although there is good evidence that the area under woodland in the county declined during the previous two centuries it is unlikely that, even in 1550, more than 7 percent of the county’s land area had been occupied by coppiced woodland. This figure is nevertheless misleading, because it fails to include the various areas of grazed woodland, or wood- pasture, which still existed in the county at this time, especially on the larger commons.

In 1766 the largest areas of common land were located in two broad bands. One occupied the high ground of the Chiltern ridge, land lying above c.150 metres OD, where Tring common, Wigginton common, the conjoined commons of Aldbury and Berkhamsted, and the commons of Hudnall, Studham, Buckwood, Kensworth and Caddington, all formed a near-continuous tract of unenclosed ground which continued across the county boundary into and Bedfordshire. A second string of large commons could be found on the high ground in the far south of the county, associated with the pebble gravels overlying the London clay. It comprised Bushey Heath (covering c.100 hectares even in 1766), Aldenham Common (c.200 hecatres), Boreham Wood common (c.260 hectares) and Barnet Common (c.230 hectares) and the conjoined commons of Northaw, North Mymms and Cheshunt, which together covered no less than 1,900 square kilometres. These large

34 commons, both in the south and on the crest of the Chilterns, continued to carry significant amounts of grazed woodland into the post-medieval period. In the Chilterns, for example, tenants of the manors in Tring had the right, in the sixteenth century, to take ten cartloads of wood each year from the ‘common wood’ of Tring, while a sixteenth-century lease of lands in Studham included an entitlement to a load and a half of wood yearly from the ‘common wood of Studham’ (HALS DE/FL/17075). Significant tracts of such common woodland still survived when Drury and Andrews’ map was surveyed in the 1760s.

In part because the extent of its wooded wastes and the ample opportunities they afforded for park making, in part because of the residential and recreational attractions presented by proximity to London, medieval Hertfordshire was also particularly well endowed with deer parks (Rowe 2009). By 1300 there were around forty in the county, a total which fell to around thirty after the Black Death but which remained at that level through the later fourteenth and fifteenth centuries. The largest example was at Hatfield, where the bishop of Ely enclosed around 1,860 acres (c.750 hectares) of land some time before 1222 (BL Cotton MS, Tiberius Bii, fo. 140–1). By 1285 the bishop also had a park at Little Hadham which extended over some 260 acres (105 hectares) enclosed from the waste in the north-west corner of the parish (Rowe 2009, 166). But it was the abbots of St Albans who had the largest number of parks, including Eywood, Derefold, Childwick and Gorhambury – all close to St Albans – and examples at Redbourn, Tyttenhanger, Boreham Wood and Bramfield. A number of large parks belonging to the Crown also existed, by the late Middle Ages, at King’s Langley, Berkhamsted and Hertingfordbury. Most medieval parks in Hertfordshire were, however, the property of relatively minor local lords, and they were thus more numerous in the eastern half of the county than in the west, reflecting the higher population levels on the fertile boulder clay soils and the greater numbers of manors there. Parks were, conversely, in general larger in the west and south, where more extensive tracts of woodland pasture had survived into the twelfth and thirteenth centuries. At least a third of the deer parks known to have existed in the fifteenth century were still in use in the sixteenth, mainly the larger examples (all but three were over 200 acres (80 hectares) in size, and half covered more than 300 acres (c.121 hectares). Indeed, almost all the latter group continued in use as deer parks into and often beyond the sixteenth century (Rowe and Williamson 2013, 153).1Absence of hard evidence, sampling bias and other factors make it very difficult to assess the total area of ground which might have been occupied by wood-pastures of various kinds in Hertfordshire in c.1550. We can, however, reasonably assume that the 30 deer parks accounted for around 20 square kilometres of grazed woodland; that the area of common land in the county in 1766 was comparable to that which had existed in 1550, there being little evidence for the disappearance of any large tracts in the intervening period; and that at that stage most commons, around 70 percent, had been to some degree wooded. Allowing for a small number of other private wood-pastures (sporadically suggested on early maps) would indicate that around 5 percent of the county had been occupied by grazed woodland in c.1550. If we add to this the estimated c.7 percent of its land area occupied by coppiced woodland, then as much as 12 percent of Hertfordshire’s land area may have been occupied by woodland of various kinds at the start of the period studied here. It was unquestionably the most densely-wooded of the four counties studied,

1 The parks containing 300 acres or more were: Hatfield great, Berkhamsted, King’s Langley, Bedwell, The More, Hatfield Woodhall, Standon, Shingle Hall, Eywood, Benington, Walkern, Hunsdon old, Weston great, Hatfield Millwards, Knebworth great, Ware and Tyttenhanger. The only one of these which does not seem to have survived into the sixteenth century is Eywood park which belonged to the abbey of St Albans. The acreage of this park is not known but it may have contained as much as c.418 acres. 35 as well as boasting the highest density of farmland trees: John Leland, writing around 1540, described how ‘This Shire aboundeth in plenty of …woods’ (Toulmin Smith 1907).

Northamptonshire

Northamptonshire, in sharp contrast to Hertfordshire, was a classic ‘champion’ county. In the Middle Ages most of its area lay in unhedged ‘open fields’ which were cultivated from strongly nucleated villages, and although a significant degree of enclosure took place in the fifteenth, sixteenth and seventeenth centuries, around half the county remained open when the first parliamentary enclosure acts were passed in the 1730s. Much of it, in consequence, has a relatively recent fieldscape of straight-sided fields defined by hawthorn hedges. The county has a complex solid geology, although dominated by clays and mudstones, and to a lesser extent sands and limestone, of Lower and Middle Jurassic date (c.195-135 BP). These formations are arranged as inclined planes dipping towards the south east (Aveline and Trench 1860; Edmonds, Poole and Wilson 1965; Hains and Horton 1969; Hollingworth and Taylor 1946). But this simple pattern is rendered complex in part by the fact that many of these formations are comparatively soft and compounded by a relatively damp climate, especially in the west of the county, where average precipitation exceeds in places 700 mm per annum (Hodge et al. 1984, 29-30). Not surprisingly, as a more integrated agricultural economy developed in the course of the post-medieval period, large areas in this part of the county were progressively laid to grass, although this was a long-drawn-out process, dependent on the spread of enclosure. We should also note, however that areas of lighter soils, overlying sandstones and limestone, also occur in the county (those of the Elmton, Shernbourne, and Moreton Associations); and that the eastern extremities of the modern county – to the east of – extend out onto the level East Anglian Fens, where there are extensive deposits of peat and alluvium of post-glacial date.

In spite of the generally challenging quality of its soils, Roman settlement was extensive in the county, but as in most other parts of England there was a significant degree of contraction in the early Saxon period, with farms retrenching to areas of permeable geology (especially in the major river valleys), before expanding again through Middle Saxon and late Saxon times. Already, by the time of Domesday, the country was sparsely wooded and fairly densely settled, and by the end of the twelfth century already boasted a classic ‘champion’ landscape of extensive and highly-regulated open fields, the individual ‘lands’ ploughed as ridge and furrow (Hall 1995). As much as 66 percent of the surface area was under arable cultivation at the time of the medieval population peak in c.1300, almost all of it as open fields. Of the remaining c.34 percent, some comprised ribbons of unploughed land – managed as pasture and meadow – within the open fields, intermingled with the furlongs; some consisted of more extensive areas of hay meadow on the floodplains of the principal rivers; while some comprised villages with their associated closes (Williamson et al. 2013, 101-126). Areas of wooded ground existed, but the overwhelming majority was located in the three royal forests (Figure 10).

In medieval times the forests contained a number of distinct elements (Pettit 1968; Page 2003; Hall 2001). Their core areas, belonging to the Crown, were in part both extra-parochial and outside the boundaries of any township. They comprised blocks of enclosed woodland, managed as coppice-

36

Figure 9. Simplified soil map of Northamptonshire.

(a) Freely-draining loams over sandstone (Banbury Association). (b) Relatively light, freely-draining soils formed in limestone of clay overlying limestone (Elmton, Shernbourne and Moreton Associations) (c) Heavy clay soils formed in decalcified chalky till (Beccles and Ragdale Associations). (d) Clay soils formed in chalky till (mainly Hanslope Association). (e) Poorly-draining, slightly acidic soils formed in Jurassic and Cretaceous clays (mainly Denchworth and Wickham Associations). (f) Freely-draining, slightly acidic loams (Wick Association). (g) Deep loams over clay (Oxpasture and Ashley Associations). (h) Loams over terrace gravels (Waterstock Association). (i) Alluvium and peat. with-standards, which were separated by areas of common wood-pasture and open pasture (called ‘ridings’, ‘plains’ and ‘lawns’ ) (Figure 11). The forest bounds extended well outside these cores of woodland and pasture, however, with forest law being applied to the farmland of many surrounding townships (the precise extent of the areas so defined tended to change over time). Among other things, this meant that deer were allowed to graze undisturbed on the crops in the open fields, to the great inconvenience of local farmers. The plains and ridings in the ‘core’ parts of the forests still

37

Figure 10. Reconstructed medieval land use, c.1300, Northamptonshire. For sources and methodology, see Williamson et al. 2013. comprised wood-pasture in the early Middle Ages and much continued to carry a reasonable density of pollards even in post-medieval times, although gradually degenerating to open pasture under the pressure of grazing and poorly-controlled timber extraction (Williamson et al. 2013, 106-109). The inhabitants of villages lying around the periphery of the forest had the right to graze cattle (in the modern sense: sheep were excluded) in the plains and ridings, but not in the areas called ‘lawns’, which were fenced off for the exclusive use of deer and the cattle of forest officers. They could also usually graze the coppices during the later part of the rotation, and pigs could also be pastured in some parts of the forest during the pannage season. Over time these core areas of the forest also came to contain numerous enclosed deer-parks and some assarts, in the form of hedged fields; and many of the areas of enclosed woodland were alienated by the Crown to magnates and local lords. The whole complex system was maintained and administered through forest courts - the Woodmote, Swanimote, and Court in Eyre – and by a range of officials and officers - verderers, regarders, woodwards, and foresters or keepers. Each forest was divided into districts called

38 bailiwicks, each based on a royal manor. Rockingham Forest thus had three: Clive (King’s Cliffe), Brigstock and Rockingham. The bailiwicks were in turn subdivided into ‘walks’ consisting of a variable number of woods or coppices (Pettit 1968).

Figure 11. Part of Rockingham Forest, Northamptonshire: reconstructed medieval land use (for sources and methodology, see Williamson et al. 2013).

Some woods did exist in the wider landscape of the county, outside the forests, a number of which still remain, such as the Grove at Hargrave, or the various woods on the interfluve between the Northampton Nene and the Ise - Cransley Wood, Hardwick Wood, and Sywell Wood. Others survived into the post-medieval period but have since disappeared, as at Mears Ashby, where a small piece of woodland existed in the north of the township until at least 1577 (Hall 1995a, 136). But in general woodland was rare beyond the forest bounds, and a survey of Crick and Clay Coton drawn up in 1526 typically reported that ‘there are no woods and the houses are in great decay for want of timber’ (TNA/PRO E36 179). The value of the enclosed woods within the forests was thus significant and a survey of the parish of Rothwell on the edge of Rockingham forest, made in 1521, described ‘a goodly wood called Rothwell Wood wherein be many fine oaks and other building timber worth to be sold by estimate an hundred pounds ...some wood must of necessity be saved for maintaining of the Lordships of Naseby and Rothwell and other Lordships in those parts where there is no wood to be had for money’ (NHRO Buccleuch 224 Box X881). Indeed, the forest commons, with their grazing

39 and pollards, were of such value that many townships and manors lying outside the forest cores had access to them, as we shall see. Rather than being drastically short of woodland, compared with counties outside the Midland belt, it is perhaps more accurate to say that in Northamptonshire (and other Midland counties) woodland was heavily concentrated, in the manner described, within the forests. Indeed, as much as 25,000 hectares of the county, or 10 percent of its surface area, was probably still occupied by woodland and wood-pasture in 1300, at the height of the medieval population expansion (Williamson et al. 2013: archaeologydataservice.ac.uk/ archives/view/midlandgis_ahrc_2010). This probably changed only gradually in the following two centuries or so, so that at the start of the period studied here the county may not have been appreciably less wooded than ‘woodland’ Hertfordshire. There may actually have been some increase in woodland area in the period between c.1350 and 1550. No less than 32 of the woods in the county which are listed in the Ancient Woodland Inventory appear to overlie former medieval arable, ploughed as ridge and furrow, while a further ten partly do so. Most, however, are small: with a few exceptions, such as the 33-hectare Ashton Wood in Sutton, they generally cover less than three hectares. Surprisingly, perhaps, most examples are found within the already well-wooded forest areas, especially within Rockingham (archaeologydataservice.ac.uk/archives/view/midlandgis_ ahrc_2010).

Population decline in the late Middle Ages did not, therefore, lead to a significant degree of woodland regeneration in the county, although it does appear to have ushered in a new phase of land use history. From the late fourteenth century arable land was steadily, if gradually, laid to pasture. This began with depopulating enclosures and village clearances of the fourteenth and fifteen centuries, but continued almost uninterrupted as land was enclosed, at various rates and by various means, into the twentieth century. The Land Utilisation Survey of 1938 shows that more than 70 percent of the county was by then under grass, a striking reversal of the medieval situation.

Norfolk

Norfolk is a larger county than either Hertfordshire or Northamptonshire – with a traditional, pre- 1972 area of 5,268 square kilometres, and has if anything a more diverse geology. As in Hertfordshire, chalk forms the basic underlying formation but it is only exposed at the surface in limited areas in the west, where it forms a low, degraded escarpment. Towards the south-east it dips gradually and becomes buried ever deeper beneath much younger formations, the most important of which are the so-called ‘Crag’ deposits – a complex sequence of gravels, clays and shelly sands of late Pliocene and early Pleistocene origin (Chatwin 1961; Funnell 2005). Only in a small area in the west of the county are pre-Cretaceous deposits – mudstone and shales, similar to those found in Northamptonshire – exposed over limited areas. All this solid geology is, however, obscured across large areas by drift deposits of various kinds. The most important is the chalky boulder clay, laid down in the Anglian glaciations, which forms a slightly tilted plateau extending over much of the centre, south and south east of the county. This is dissected to varying extents by streams and rivers. In some areas, as to the north of the , extensive tracts of almost level land occur but elsewhere, as in the far south east, around Hedenham and Ditchingham, the degree of dissection is so great that that the countryside presents a gently rolling appearance. As on the boulder clays of Hertfordshire, where the terrain was relatively undulating, cultivation was easier; where it was most level, in contrast, extensive areas of unploughed ground, often in the form of extensive commons, were to be found well into the post-medieval period. In the north east of the county the Brown Till or Norwich Brickearth is the principal glacial deposit, but is mixed across wide 40 areas with wind-blown loess. Here very fertile soils – those of the Wick 2 Association – can be found, but interspersed with areas of Quaternary sands and gravels, many of which continued to be occupied by tracts of heathland into the nineteenth century. More extensive areas of such deposits, again traditionally associated with heath land, occur in the area immediately to the north of Norwich; between Holt and in the far north of the county, forming a ridge which is probably a glacial moraine of Devensian date; and, most extensively, in the arid, agriculturally marginal area of Breckland, in the southwest of the county. Here Aeolian sands directly overlie chalk and till, giving rise to a district of vast heaths which survived until enclosure and reclamation in the early nineteenth century, and in many cases even into the twentieth. The district is now characterised by the great conifer plantations of the Forestry Commission. In the area of west Norfolk lying to the north of Breckland is another area of light, freely-draining soils, appropriately labelled the ‘Good Sands’ by Arthur Young (1813), for here the glacial sand lies more thinly over the underlying chalk, making for a less agriculturally marginal countryside, although one in which, until the eighteenth century, much open heath similarly existed. Various drift deposits of glacial date thus account for most of the county’s surface geology, but in addition it includes extensive areas of wetland in which Holocene silts and peats occur, especially in the Fens, in the far west of the county, and in to the east.

Not surprisingly, given such geological diversity, Norfolk is characterised by remarkable contrasts in landscape, although their character and development cannot be fully discussed here. The evidence leaves no doubt that prehistoric settlement was extensive in areas of light soil, and especially on the calcareous soils of the Newmarket Association in the ‘Good Sands’ and Breckland. Analysis of pollen from the sediments from a number of the county’s natural lakes or ‘meres’ ( Mere, Sea Mere, Diss Mere, Mere), and from a smaller number of riverine and coastal locations, suggest that pine and birch woodlands of the early Mesolithic were succeeded, in the period after c.6500 BC, by mixed woodland in which lime, together with oak and elm, were generally dominant, although pine continued to be a major component of woods on dry, acid soils, especially in Breckland. Limited modification of the natural vegetation during the Mesolithic was followed, from around 4,000 BC, by the more extensive but nevertheless localised clearances affected by Neolithic farmers. Much woodland persisted, especially on high interfluves, into the Bronze Age, but by the Iron Age ‘open agricultural landscapes were widespread’ in the county (Fryer et al. 2005, 11; Rackham 1986b, 162-3). Pollen cores from Hockham Mere, Old Buckenham Mere, and Diss Mere all suggest rapid of their catchments in the course of the Iron Age (Wiltshire and Murphy 1999), and archaeological fieldwalking, aerial photography, chance finds and objects recovered by metal detectorists all show a dense spread of Roman settlement - by the third century, an average of between 0.5 and 1.5 settlement sites per square kilometre (Gurney 2005; Rogerson 1995; Davison 1990). Farms could be found in almost all locations, even on the difficult Beccles Association soils of the boulder clay plateau, although both here and in other districts the main areas of cultivation were probably concentrated in the principal valleys, with the higher ground continuing to be occupied by tracts of woodland and grazing.

There was – as in the other areas already discussed - a major contraction of settlement in the post Roman period. Yet while there is some evidence (most notably from Diss Mere and Staunch Meadow at Brandon) for a measure of woodland regeneration, mainly in the form of increases in the frequency of ash, birch and hazel pollen, elsewhere – as at Scole or Hockham Mere – pollen sequences suggest the maintenance of largely open landscapes, albeit with some decline in the area

41 under cultivation and a concomitant increase in pasture (Murphy 1993; Fryer et al. 2005, 11; Bennett 1983). The Middle and Later Saxon periods saw a significant re-expansion of settlement. The area under cultivation increased once more, and the heavier clays began to be settled and farmed again (Bennet 1983; Godwin 1968; Peglar et al. 1989; Murphy 1994; Rogerson 2005). By the time of Domesday Book Norfolk was, with Suffolk, the most densely settled region of England. Some rural districts probably contained more people in 1086 than they did in the nineteenth century (Skipper and Williamson 2005). Yet Domesday also attests the survival of extensive tracts of uncleared woodland, now concentrated on the heavier clays, and the most acid gravels, on the high watersheds between major rivers, and especially in a broad arc running through the north-east, centre and south-east of the county (Rackham 1986b).

Figure 12. Simplified soil map of Norfolk.

(a) Freely-draining acid loams formed in windblown drift (Barrow Association). (b) Light, freely-draining soils formed in chalk (Newmarket 1 and 2 Associations). (c) Heavy clay soils formed in decalcified chalky till (Beccles and Ragdale Associations). (d) Slightly acidic loamy clay soils formed in chalky till (mainly Burlingham 1 and 2 Associations). (e) Lime-rich clay soils (Hanslope Association). (f) Freely-draining, slightly acidic loams (Wick Association). (g) Acid, freely-draining soils formed in glacial sands and gravels (Newport, Worlington, Ollerton and Associations). (h) Alluvium and peat.

The eleventh, twelfth and thirteenth centuries, a period of rapidly rising population, saw the Norfolk landscape crystalise out into something like the forms we encounter on the earliest available maps.

42

Much of the land area was cultivated as open fields, but only on the light soils in the west of the county – in Breckland and the Good Sands – did these occupy very extensive and continuous areas of ground, and take the kind of ‘regular’ forms broadly resembling those already described in Northamptonshire, although with a less strongly nucleated pattern of settlement (Martin and Stchell 2009). Extensive heaths also survived in these areas, as already noted, covering as much as 40 percent of the land area of Breckland as late as the 1790s, to judge from the evidence of the county map published by William Faden in 1797 (Wade Martins and Williamson 1999, 13). Other tracts of heathy land, this time associated with the Cretaceous Greensand and related formations, extended northwards along the edge of from Mintlyn to Snettisham. These light-soils districts in the west and north of the county were thus classic areas of ‘champion’ sheep-corn husbandry, with a settlement pattern of loosely-nucleated villages, and few hedges or hedgerow trees. In the far wet of the county, on the edge of the Fenland, small areas of pre-Cretaceous formations carried landscapes more akin to those of Northamptonshire, with nucleated villages farming open fields on heavy, sticky clays, with the individual strips ploughed as ‘ridge and furrow’.

On the boulder clay soils in the centre, east and south-east of the county, but also on the well drained and fertile loams of the north east, medieval settlement was much more dispersed in character, often with the majority of dwellings in a parish strung out around the margins of commons - ‘moors’ and ‘greens’ on the poorly-draining clays, ‘heaths’ on the more acidic ground. Open fields were extensive on the fertile loams of the north east, although generally more ‘irregular’ in character than those found in the west of the county. On the clays they also covered much ground, although here hedged fields also formed an important component of the landscape, even in the Middle Ages, and the demesne lands of manors, in particular, often lay entirely in enclosures. Almost certainly, the heavier soils of Norfolk were already, in the Middle Ages, characterised by higher densities of hedgerow trees than the lighter ones, even the light but fertile loams of the north east.

The ‘irregular’ open fields which dominated the eastern half of the county were particularly susceptible to informal, ‘piecemeal’ enclosure and most disappeared in the course of the fifteenth, sixteenth and seventeenth centuries. On the claylands especially this development was associated with a growing specialisation in cattle farming and dairying, and by the early eighteenth century large tracts of land were under permanent pasture. Enclosure added to the existing pattern of hedges, to create an irregular and often dense mesh of hedges. Enclosure came rather later to the more open landscapes in the west of the county, in Breckland and the Good Sands. Here it was sometimes achieved gradually, as land was steadily amalgamated in the hands of the large landed estates which increasingly came to dominate this part of the county in the course of the post- medieval period, but often it came about through parliamentary enclosure acts. In these districts it was motivated less by a desire to specialise in livestock husbandry than by the adoption of the new farming practices of the ‘agricultural revolution’, and especially new crops and rotations, which were hard to adopt in the communal landscapes of the open fields. Today the countryside of these districts is characterised by straight-sided fields typical of the late, planned enclosure of open fields and commons. Parliamentary enclosure also saw the removal of numerous areas of common land, both in the west of the county – including most of the great Breckland heaths – and in the east, where they had survived in large numbers even as the open arable around them had gradually been replaced, piecemeal, by hedged fields.

43

The extent of woodland in the county at the start of the period studied here is uncertain, as we shall see, but was probably significantly less than that in Hertfordshire or Northamptonshire, probably amounting to around 7 percent of the land area, even if wooded commons and deer parks are taken into account. The most wooded areas were to be found on the claylands in the centre, south and east. The density of farmland trees was also high in these areas, similar to that in the old-enclosed districts of Hertfordshire.

Yorkshire

Yorkshire is the largest of the counties studied here, and the largest in England, its three ‘Ridings’ – North, South and West - covering in all some 15, 650 square kilometres. Following the 1972 Local Government Act, South Yorkshire was formed out of part of the old West Riding and the northern sections of , but in the analyses that follow the old county boundaries, and those of the Ridings, will be used as the basis for long-term comparisons. Unsurprisingly considering its size, Yorkshire comprises very varied physical environments and landscapes, related in large measure to an equally varied topography and underlying geology, and what follows represents, of necessity, a gross simplification of a complex reality. The county provides an important contrast with the others studied not only in terms of its northerly location and topography, but also by virtue of the fact that it became industrialised at an early date. The county is rich in rock and mineral resources: limestone, coal, lead, alum, jet, ironstone, the list of extractive industries is long. It is well supplied with fast flowing rivers which provided power to mills, while access to the sea ensures good transport connections to Newcastle, East Anglia and London. The populous industrial heartlands of the region, spanning South and West Yorkshire, around the cities of Sheffield, Leeds and Bradford are perhaps the most obvious legacies of the industrial explosion of the late eighteenth and nineteenth centuries. However, it is in the uplands that more striking changes occurred. Now relatively sparsely populated, from the seventeenth to the beginning of the twentieth century, these were bustling industrial landscapes. Whilst it was once thought that this industrial activity effectively decimated local woodland resources, as trees were felled for fuel, pit props and manufacturing materials, it is now accepted that industry played an important role in preserving ancient woodlands, through intensive management and protection. In this respect, a history of Yorkshire's industry is also a history of its woodlands.

Most of Yorkshire, even the higher and poorer ground, had been largely cleared of woodland by late prehistoric times, to judge from the archaeological evidence – especially that of aerial photography, which has revealed vast areas of prehistoric field systems along the edge of the Pennines, in the Wolds and around the edges of the Vale of Pickering (Stoertz 1997). The post-Roman period saw the by now familiar pattern of settlement retraction, accompanied by localised woodland regeneration (Gledhill 1994, 210–213). During and after the middle Saxon period, however, the more difficult soils came back into cultivation, and by the twelfth and thirteenth centuries much of the lower ground in the county was characterised by nucleated settlements farming open-fields broadly similar to those found in other champion districts, although with their own particular characteristics – notably the ‘bydales’, bundles of long strips which ran continuously for a kilometre or more (Hall 2014, 45-52). In the upland areas, where the soil productivity could not sustain such extensive areas of arable, dispersed settlement and infield-outfield farming were common, with the infield – generally in the form of intermingled strips - regularly cropped, while the outfields were farmed on a long rotation, phases of cropping being interspersed with several years of use as pasture.

44

It is difficult to estimate the extent of woodland cover in Yorkshire in the Middle Ages. Whereas in some parts of England it is possible to make guesses based on the information provided by the Domesday Book, in Yorkshire this is particularly problematic. Many settlements in the county are named but not described in the survey, instead being recorded simply as lying ‘waste’, possibly as a consequence of the ‘Harrying of the North’ in 1069-70 (McDonnell 1992); while Gledhill (1994,141– 5) has also suggested a wider pattern of under-recording, arising from vagaries of the survey process, of the politics of medieval taxation, and of the manner in which woodland was allocated to different vills. Uniquely, the Yorkshire entries in Domesday include the categories of silva pastilis and silva minuta, the former clearly wood-pasture but the second of uncertain significance, although probably meaning ‘scrub’ (Gledhill 1994, 160). Overall, it has been suggested that at the time of Domesday around 12 percent of the North Riding was occupied by woodland, 16 percent of the West Riding and perhaps 4 percent of the East Riding (Rackham 1990,50). It is also possible to discern a broad contrast between well-wooded areas, mainly but not exclusively in the uplands, and more sparsely-wooded ones, principally arable lowlands, which is also reflected in the distribution of Old English place-names.

Yorkshire is the only one of the counties studied that includes significant areas of upland countryside. The north-east of the county is dominated by the North York Moors, towering above the lowlands to the south and west. The vast moorland plateau, rising to over 400 metres OD, is formed in Jurassic sandstones, mudstones and shales, and is punctuated by a series of limestone outcrops and river valleys, the latter mostly but not exclusively draining towards the south and forming scars and dales which today contain improved pasture, a little arable, and some woodland. This is today one of the most wooded areas of the county, with woods and plantations covering around 30,000 hectares, or 22 percent of the land area. Although, as we shall see, much of this is of recent planting, woodland has long been concentrated on this south flank of the Moors, extending south onto the lower ground and mixed geology of the area known as the Howardian Hills. In the west of the county, separated from the Moors by the lowlands of the Vale of Mowbray, is the second great upland area, the Pennines, where the ground rises to a height of 700m OD. The gently undulating eastern flanks of these uplands, where the shales and mudstones of the Coal Measures overlie significant deposits of coal, became a major industrial area from the late eighteenth century and is today dominated by two great conurbations based on Leeds to the north and Sheffield to the south. A thin band of Permian Magnesium Limestone, running north-south the length of the county, overlies the Coal Measures on their eastern edge and here a largely rural landscape still survives, traditionally an arable district dominated by large estates, and which now forms the eastern edge of this industrial zone. To the west, in contrast, the ground climbs steadily and the topography is dominated by alternating strata of gritstone and limestone. The former widen as the ground rises, producing extensive tracts of heath and moorland, punctuated by hard rocky outcrops: much remained as common land into the eighteenth or nineteenth centuries. The bleak uplands are cut by ribbons of lower ground, however – the Dales – where glacial rivers have cut down through alternating strata of limestone and sandstone. Here mixtures of enclosed fields and open fields could be found, alongside steep valley sides where wood-pastures sometimes survived into the post- medieval period. In both of these main upland areas, woodlands of various kinds thus appear, in all periods, to have been relatively extensive but mainly along the margins of the great upland masses, especially on the steep slopes of sides of gullies and valleys. Melvyn Jones (1999, 11) has suggested that at the time of Domesday woodland may still have covered 22 percent of the land areas on the

45

Gritstones and Coal Measures around Sheffield, although in the Doncaster area, on the Magnesium Limestone, the figure was only 9 percent. In general, the land in and around the uplands was enclosed with walls, rather than hedges. This tended to preclude the establishment of hedgerow trees to some extent – they were more difficult to incorporate into a wall than a hedge – although trees were sporadically planted in such boundaries and also – given the predominance of pasture – grew free-standing in the fields.

Figure 13. Simplified soil map of Yorkshire. (a) Acid, peaty upland soils. (b) Other acid upland soils (e.g., Belmont and Dale Associations). (c) Freely-draining acid sandy soils (e.g. Newport Association). (d) Seasonally waterlogged acid sandy soils. (e) Freely-draining slightly acidic loams E.g. Bishampton and Rivington Associations). (f) Freely-draining, lime-rich loams. (g) Heavy, mainly acidic clays formed in glacial drift Holderness, Denchworth, Foggathorpe Associations etc). (h) Well-drained limy soils in chalk and limestone (e.g. Elmton, Andover, Coombe Associations) (i) Alluvium and lowland peats.

46

In marked contrast to these two upland districts was the great area of lowlands, comprising the Vales of Mowbray and York, which runs north-south through the centre of the county. Geologically, both of these districts comprise sands, lacustrine clays and alluvium, giving rise to varied but predominantly clayey soils. Both, in medieval times, were characterised by ‘champion’ landscapes of nucleated villages and extensive open fields which represented, in effect, a northerly continuation of the landscape of the ‘champion’ Midlands. Much of this land was only enclosed in the eighteenth and nineteenth centuries, here with hedges rather than walls. The Vale of Mowbray comprises the northern section of the lowlands, and is distinguished by its meandering rivers and generally undulating landscape: the Vale of York is much larger, and contains a wider variety of soils, including significant areas of freely-draining loams. A third, less extensive lowland, the Vale of Pickering, lies to the south of the North York Moors and was rather different in character. The drainage basin for much of the River Derwent, it is a level area of boulder clays which, before drainage in the nineteenth century, included extensive marshes and wetlands. Reflecting this fact, the principal medieval settlements were located around the edges of the Vale, where land begins to rise to moors in the north and Wolds in the south, or on spurs of slightly elevated ground projecting from the valley floor.

To the south of the Vale of Pickering, and forming a broad, curving band extending from Flamborough Head on the east coast to the Humber in the south, lie the Yorkshire Wolds, hills formed in Cretaceous chalk but overlain towards the east by areas of glacial sands and boulder clay. The main mass of the Wolds was the most northerly of England’s sheep-corn districts, characterised by light soils which were easily leached of nutrients, and whose cultivation thus required the regular folding of sheep, although towards the east (and generally towards the peripheries of the district) the overlying glacial deposits of clay gave rise to more fertile loams. This was an area of nucleated villages and extensive open fields, those on the bleaker uplands often cultivated in part on an extended rotation: as the Georgical Committee were informed in 1674, the farmers had ‘in many townes 7 feilds and the swarth of one is every yeare broken for oates and lett ly fallow till itts turne att 7 yeares end, and these seven are outffeilds’ (Harris 1961, 25). The majority of villages were found on the lower ground, in valleys or around the margins of the uplands. A smaller number were located in more marginal locations, on higher ground, and a proportion of these were abandoned, and their land converted to sheepwalks and rabbit warrens, in the late Middle Ages. In common with other areas of light, well-drained soil, woodland was sparse here from an early date. At the start of the eighteenth century one commentator described the landscape around Wetwang as having ‘scarce a bush or tree for several miles’ (Hey 1984, 74). The same was broadly true of Holderness, a low-lying region of boulder clays and heavy clay loams, interspersed with lighter sand and gravel deposits, which lies to the south and east of the Wolds. This, too, was a ‘champion’ open-field district, which was characterised in post-medieval times by mixed farming, with a significant pastoral bias, and which by the end of the Middle Ages contained only a few, scattered areas of managed woodland. Although both these districts experienced a degree of enclosure during the sixteenth and seventeenth centuries, they were predominantly areas of eighteenth and nineteenth-century enclosure by parliamentary act, and the Wolds in particular are today a classic example of ‘planned countryside’.

At the time of Domesday most of the woodland in Yorkshire, as in the other counties studied, was probably wood-pasture; but as in other regions the proportion managed as coppice increased over time, although it is possible that this was a rather later development than in the south, in some

47 districts at least. Gledhill has analysed over 900 separate twelfth and thirteenth-century references to 400 separate woodlands in North Yorkshire, of which 121 can clearly be identified as wood- pasture or coppice: of these around 70 percent were clearly in the former category (Gledhill 1994, 168). He has further suggested that pannage was the most frequent use of woodland in the period before 1200; that between 1200 and 1320 more general references to woodland grazing predominate; that and only in the period between 1320 and 1400 do references to coppicing start to become more prevalent. It is perhaps surprising that more intensive management should have developed following the Black Death, given that the palynological data from the region implies a period of woodland expansion at this time, but an increased emphasis on coppicing perhaps reflects rising prices for underwood products at a time of rising disposable incomes. However, there was evidently much variation in experience from region to region within the county, and in Nidderdale in the Dales, to judge from the research of Ian Dormor, an increase in the area of coppiced woodland was related, in a more understandable fashion, to the gradual degeneration – continuing through the fourteenth and fifteenth century – of wood-pastures, the result of intensive grazing by the granges of Fountains and Byland Abbeys. It was also related – in a manner that was to continue to be significant through the period studied here – to a growth in local industry, the produce of coppices being used to make charcoal, or employed directly for smelting. By 1450 the area under wood- pasture in the Dale had decreased by half (Dormor 2002, 18); and at the time of the Dissolution, 69 percent (571 acres, 231ha) of the abbey woodlands here were coppiced (Dormor 2002, 54).

As in the counties further south, wood-pasture survived best where it was delberatly preserved for hunting. Powerful landowners enclosed extensive tracts of surviving ‘waste’ as deer parks, a process which appears to have peaked in the thirteenth and early fourteenth centuries in both south Yorkshire (Melvyn Jones 1999) and the Dales (Muir 2006, 124). Yorkshire also contained a number of royal forests and private chases – over 130 are referred to in medieval documents, although some of these would have been sub-divisions of larger areas (G. Jones pers. comm.). Most were likewise in upland areas, encompassing both extensive tracts of moorland and areas of wood-pasture on valley sides, and on the sloping ground leading up to moorland plateaux (Fleming 1998, 82–99).

The most striking difference between Yorkshire, and the other districts studied here, was the early development of an industrial economy, mainly concentrated on the margins of the Pennines and in some of the Dales. The smelting of non-ferrous metals, especially lead, increased in importance steadily though the post-medieval period and, before the development and widespread adoption of the reverberatory furnace in the first half of the eighteenth century, much was fuelled using charcoal. While the development of the South Yorkshire coalfield, which was particularly rapid after improvements to the navigation of the Don in the 1720s and 30s, in some ways undermined the demand for wood it also provided an important market for woodland products, for as more sophisticated forms of mining developed in the later seventeenth and eighteenth centuries there was an increasing demand for pit props, mainly utilising small timbers or the poles of coppices cut on a long rotation.

Sources

In the report that follows we analyse a large number of printed texts, published between the sixteenth and the twentieth centuries, to throw new light on tree disease, and on changing attitudes to trees, woodlands and their management. These include books specifically dedicated to the subject of forestry, such as Moses Cook’s The Manner of Raising, Ordering and Improving Forest

48

Trees of 1676, and also volumes whose main focus is on the practice of agriculture, but which also deal to some extent with farmland trees and woods, such as John Worlidge’s A Compleat System of Husbandry and Gardening (1660) or John Mortimer’s Whole Art of Husbandry (1707). In addition, in tracing the development of tree populations, and of tree disease, within the four study counties we employ a wide range of manuscript and other documentary evidence, mainly preserved at the various country record offices. It must, however, be emphasised at the outset that on a number of topics our data are often limited, biased or misleading. There are thus comparatively few useful sources from the period before the late seventeenth century. After this, the most important information comes from timber surveys; from estate accounts, letters, leases and other records; and also from maps of various kinds, aerial photographs and (to a lesser extent) contemporary illustrations. All present the researcher with particular challenges. For example, in some cases those responsible for making timber surveys of properties made clear the purpose of the undertaking, and which trees were included and excluded: some thus explicitly record only those with timber value, and exclude pollards. Often, however, such limitations are not expressly stated, leading to possible errors of interpretation. Estate records of various kinds provide useful information about management, such as details of the length of coppice and pollarding rotations and the age at which timber trees were felled, as well as informing us about attitudes to trees and the economics of forestry more generally. Unfortunately, much of the information they provide is impressionistic – little comprises the kinds of data which can be used statistically - and their allusions to tree health, in particular, are generally vague and hard to interpret. Maps are a particularly problematic source. Although there are a number of early examples which clearly attempt to show farmland trees with considerable care – sometimes even detailing their species and management – in other cases it is unclear where, precisely, on the spectrum between accurate portrayal, and mere illustration, the representation of trees on a particular map may lie. This problem is compounded by uncertainties about what, exactly, different sources actually mean by a ‘tree’, as opposed to a shrub or a sapling – a difficulty which, in fact, bedevils the use of almost every source employed, and every kind of statistical analysis we might make. Some early surveys specifically record the number of young trees (often referred to as ’stands’ or ‘standills’), and distinguish them from mature trees and pollards, but many do not, leaving it unclear whether certain age sizes were being excluded, or included, in particular documents.

Even apparently straightforward and relatively recent sources can cause difficulties in these kinds of respects. The First Edition 6-inch and 25-inch Ordnance Survey maps from the later and, in Yorkshire, mid-nineteenth century ostensibly show the location of every mature free-standing tree in the countryside, and this information has often been taken at face value by historians in the past, even though the original instructions given to surveyors do not appear to have survived (Rackham 1986, 222-2). Where comparisons can be made with near-contemporary estate maps they suggest that the numbers depicted were in fact broadly, although not entirely, reliable. In Yorkshire, for example, a selection of five estate maps produced around 1867 can be compared with the Ordnance Survey six- inch first edition maps from the mid 1850s, and the 25-inch first edition from the 1890s, for the same areas.2 In most cases there were more trees shown on the six-inch than on the 25-inch, suggesting a real decrease in the numbers of farmland trees, and in four cases the estate maps fall into and confirm this pattern, showing numbers mid-way between the two. The exception, an estate

2 The relevant maps (all in Sheffield RO) are: Ecclesall, WMP/MP/119; Swinton, WMP/MP/126; Brampton, WMP/MP/115 R; Hooton Roberts, WMP/MP/122: Nether Hoyland, WMP/MP/123 R. 49 map of a property in Ecclesall, shows significantly fewer trees than appear on the Ordnance Survey, presumably because of differences in the size of ‘mature’ trees being included. Yet there are grounds for suggesting that in some contexts the depictions of trees on Ordnance Survey maps do need to be treated with more caution, not least because of the difficulties involved in showing trees where the map is otherwise crowded with features and symbols. A significant degree of omission thus seems to have occurred in settlements, where few trees are generally shown, but where other sources suggest they were often tightly clustered. Other possible problems include the extent to which hedgerow trees may have been omitted when tightly packed in hedges and, once again, how – and how consistently – the surveyors defined a ‘tree’, as opposed to a sapling or bush. This latter problem, which also affects our interpretation of such sources as the 1946 RAF vertical aerial photographs, makes comparisons over time of tree densities in particular areas particularly problematic. Yet it must, in the final analysis, simply be accepted as a reflection of reality. There is, in truth, no clear line between a shrub and a tree, and changes in land management over recent decades - the tendency in some circumstances to allow hedges to grow, uncut, into lines of mixed trees and shrubs – have served to blur any distinction still further, as we shall see.

In short, few of the sources employed in this study were designed to answer the kinds of question which we are asking of them, and this needs to be borne in mind in the pages that follow, even when not specifically highlighted in the text. This said, clear patterns of development in tree populations, their management, and their health can be discerned: and these, we believe, throw important new light on current debates about the state of tree health in England.

50

2. Farmland Trees

Trees in old-enclosed districts: Hertfordshire and Norfolk.

The complex pattern of variations in landscape and regional economies briefly described in the previous chapter were associated with major differences in the density, and in the management, of farmland trees, that is, trees growing in locations other than woods and wood-pastures. To some extent these differences arose, as we shall see, from the nature of local economies, and especially fuel economies. But they were also contingent upon the extent of enclosure, for most farmland trees grew in hedges and, where hedges were few in number, so too were farmland trees.

Tree surveys and maps leave little doubt that in the period before the late eighteenth century, and in some cases well into the nineteenth century, the old-enclosed districts of Hertfordshire and Norfolk, in common with adjacent areas of south-east England and East Anglia, contained vast numbers of farmland trees. We might start our discussion with the claylands of south Norfolk where early maps – such as that made of the Channons Hall estate in Tibenham in 1640 – generally show trees densely packed into hedges, and also in places scattered across pasture fields, or grouped in bands two or three trees deep around field margins (NRO MC 1777/1). The latter features, which continue to appear on maps from the district well into the eighteenth century, are referred to as ‘rows’ in contemporary documents, sometimes as ‘grovetts’. A 1633 glebe terrier for the parish of Denton typically describes ‘the Churchyard with a Pightell and a little Grovett above that Close towards the south’, and ‘Two Closes joineing together Westward called South Crofte – the first Close hath a Grovett above’ (NRO PD 136/35). Unfortunately, it is only really in the eighteenth century that surveys of timber on farms, and other sources, provide detailed information on the density, management and character of farmland trees. Of particular importance are the maps prepared by the surveyor Henry Keymer of three small estates in the parishes of , Beeston-by- and . All use different symbols to distinguish four main categories of farmland tree - ‘elm’, ash’, ‘oak’ and ‘pollard’ (with that for Scarning noting, in addition, the presence of alders). None of these properties lies on the heaviest of the local clay soils (those of the Beccles Association), but instead on the slightly lighter loams of the Burlingham 1 Association, where there was rather more arable. Indeed, the Bintree and Beeston properties both included some open-field strips, although mostly lying in hedged closes. There are few signs of the ‘rows’ or densely-treed fields of the kind shown on many seventeenth-century maps, but on all the mapped properties there was nevertheless a very high density of trees, in part because of the low average field size and the tight network of hedges. The Beeston property, surveyed in 1764, comprised 70 acres (28 hectares), of which 7 lay in the open fields, and the remainder in hedged closes with an average size of 3.5 acres (1.4 hectares) (NRO WIS 138, 166X3). The map shows a total of 531 trees, an average of 7.6 per acre, or 8.4 per acre (c.21 per hectare) if we exclude the open-field land. The estate at Scarning was surveyed in 1761 and was slightly larger, covering 80 acres (32 hectares), and lay entirely in enclosures with an average area of 4.2 acres (1.7 hectares) (NRO BCH 20). Keymer’s map shows a total of 1,095 trees, or 12.4 per acre (31 per hectare). Lastly, the farm at Bintree, mapped in 1756, consisted of 79 acres (32 hectares), 15 of which lay in open fields and 64 in hedged closes with an average size of 3.4 acres (1.4 hectares), although this figure is slightly skewed by the inclusion of some very small yards around the principal farm (NRO Accn. Field 18.3.82 P188B). There were a total of 529 trees on the enclosed land, or 8.27 per acre (20 per hectare). These tree densities do not appear to have been unusual on the Norfolk

51

Figure 14. Extract from Henry Keymer’s map of an estate at Beeston, Norfolk, surveyed in 1764. Unusually, Keymer recorded the position of all farmland trees. The features resembling palm trees are pollards. claylands, to judge from an undated mid-eighteenth century sketch map of two farms in , again on mixed although rather heavier clay soils (NRO HIL 3/34/1-27, 879X3). 3 One property covered 34 acres (14 hectares), had an average field size of 4.9 acres (2 hectares), and had 96 trees – 5.76 an acre (14 per hectare). The other covered 28 acres (11 hectares), and had an average parcel size of a mere 2.8 acres (1.1 hectares). Here there were 227 trees, or 8.1 per acre (20 per hectare). Taking the five properties together, there was an average field size (ignoring land lying in open fields) of around 3.75 acres (1.5 hectares), and an average tree density of 8.4 per acre (c.21 per hectare). This is high, but it is likely that on heavier land even more farmland trees would have been present, not least because some pasture fields were densely filled with pollards, almost like diminutive wood-pastures. A tenant at in 1731 asked permission to fell 67 pollards in an 8 acre close (NRO DCN 59/21/1). On heavy clay soils at Thorndon, just across the county

3 In fact three properties are mapped, but for one no acreage is provided. 52 boundary in Suffolk, one farm had as many as 29 trees per acre (c.72 per hectare) when surveyed in 1742 (WSRO BT1/1/16).

As explained in the introduction, the light loams in the north-east of Norfolk were also, in many places, enclosed before the parliamentary enclosure period, and often before the end of the seventeenth century. The best evidence for the density of farmland trees in this district comes from a survey of the estates of the Earle family in the area around Heydon and Woodrising, which was made in 1722. This included some land lying on the clay plateau to the west – especially in the parishes of Wood Dalling and – but the majority lay on light loams of the Wick 2 and 3 Associations. In total, the survey records no less than c. 29,000 trees, of which c. 24,000 were on farmland, mainly growing in hedgerows, which were recorded on a field-by-field basis. Unfortunately, the area of some of the fields is not given, but densities seems to range from around 7 to nearly 19 per acre (17 to 47 per hectare), averaging 11 per acre (27 per hectare). These figures, however, include - as well as mature timber trees and pollards - ‘storrells’ or ‘standrills’, that is, small young trees, a category which comprised 41.4 percent of the trees recorded. If these are omitted from the calculation, then there was an average of around 16 trees per hectare, significantly lower than on any of the clayland properties just discussed. A rather smaller survey of property at Burgh and Skeyon, dating from 1816, suggests similar densities (NRO BR 90/35/22), although in this case it is uncertain whether immature trees are being included. While it is possible that Keymer (and the unnamed surveyor of the Whissonset farms) likewise included some of the larger ‘storrels’ as timber, lessening the apparent scale of the difference, on balance the north east of Norfolk appears to have been a less densely treed district than the claylands in the south and centre of the county.

The paintings of these same landscapes made by the artists of the ‘The Norwich School’ – such as John Crome (1768-1821) and John Sell Cotman (1782-1842) - at the start of the nineteenth century ought to provide some indication of what these densely treed landscapes actually looked like, but they are not a good guide to the numbers or management of trees, framed as they are within a range of artistic conventions and modelled to a large extent on Dutch models, and to a lesser extent on the depictions of Italian landscape made by artists like Poussin and Lorraine. This said, they do generally show a landscape well endowed with trees which are, for the most part, in healthy condition and – with the exception of some famous ‘veterans’ like the Winfarthing Oak, depicted by James Sillett - of no great antiquity. As the trees are generally incidental to some other object, or shown as distant parts of wider views, it is difficult to identify well-grown pollards from timber trees, but recently cut pollards feature in a number of the paintings, including ‘Road with Pollards’, ‘The Beaters’ and ‘At Bixley’ by Crome, and ‘Landscape with Pollards’ by John Middleton (all at Castle Museum, Norwich).

Hertfordshire, as we have seen - except for the strip of ‘champion’ land on the chalklands in the far north of the county - had been largely enclosed by the end of the seventeenth century, and extensive tracts had probably always been farmed in enclosed fields. Here, once again, early maps show trees not only growing in hedges but also in the ‘hedge greens’ around the margins of the fields, thinly scattered across pasture fields, or – on occasions – densely filling them, forming diminutive private wood-pastures. The evidence is particularly good from the boulder clay districts in the east of the county. On Olives Farm in Hunsdon in 1556 there were 896 trees growing on 175 acres – five per acre, or 13 per hectare (TNA/PRO E315/391); on the enclosed portions of three farms on the manor of Great Barwick in Standon in 1778 there were 16 trees per hectare, and on the

53 demesne there, 13 per hectare (HALS E/2832 and 2833); on a property at Whempstead, Little Munden, in 1808 there were 19 trees per hectare (HALS 81750); while on Pearces Farm and Stockers Farm in Thorley and Sawbridgeworth in 1807 there were 15 per hectare (HALS DE/H/P16). Most recorded densities were thus in the range 13 – 19 per hectare, apparently inclusive of pollards, but lower and higher figures were also recorded. One farm at Puckeridge in 1778 had only 3.4 trees recorded per hectare (HALS 43754); at the other extreme, a 33-acre (c.13-hectare) property at Kelshall surveyed in 1774 had 24 trees per acre (59 per hectare) (HALS DE/Ha/B2112). It is noteworthy that an earlier survey of the same property, made in 1727, shows even higher densities, a staggering 99 trees per hectare, but this may include ‘standils’ or immature trees, in contrast to the documents just discussed, which seem to omit them (HALS DE/Ha/B2112). Ignoring this last document, the average density of farmland trees in these seventeenth- and eighteenth-century surveys appears to have been around 15 per hectare, similar to that found in north east Norfolk, but lower than that on the claylands in the south of that county.

The south of Hertfordshire, however – a district, as we have seen, of particularly poor soils formed in London clay and pebble gravels – had rather higher recorded densities. On seven farms on the Broxbournebury estate in 1784, covering 373 acres (151 ha), there were 3,311 trees, an average of 22 per hectare: the greatest densities were recorded on West End Farm in Wormley, where there were no less than 1,358 pollards and 148 timber trees, together with 569 ‘spars’ (saplings) and 18 apple and two walnut trees in the orchard, growing on a mere 38 hectares – a density of around 40 mature trees per hectare (HALS DE/Bb/E27). On an 89-acre farm at Hammond Street, Cheshunt in 1650 there were ‘927 lopt Pollardes of oake Ash and Hornebeame’, apparently growing on the 48 acres (19 hectares) of enclosed fields on the farm, a density of about around 47 trees per hectare (PRO/ TNA E317/Herts/24, p. 37). On other properties in the district for which we have information densities were lower although still, for the most part, higher than those recorded on the boulder clays in the east of the county. On the Broxbourne Mill estate in the late eighteenth century the density of trees in the fields on the higher ground – covering 22 acres (c.8 hectares) – was 19 per hectare, but on the meadows – comprising 12 acres (c.5 hectares) – it was as high as 27 trees per hectare, mainly consisting of poplars and willows (HALS B479); while as late as 1852, on a farm in London Colney, 1,265 trees were recorded on 55 hectares, a density of 23 per hectare (D/EB 944 E12 & P1).

Only in the west of Hertfordshire, on the Chiltern dipslope, do the numbers of trees on enclosed land appear to have been noticeably lower, although it should be noted that there is less data from this district. On three farms recorded in a survey of 1838 lying in , Flamstead and Redbourne, there were less than five trees per hectare; while a map and survey of ‘Three Fields called Church Lands’ in Watford, made as early as 1754, record fewer than three per hectare (HALS D/E Cr 55; HALS DP 117 25/7). On the other hand, an undated late seventeenth-century map of a 49-acre (c. 20-hectare) farm in Flaunden suggests a density of over 15 per hectare, possibly a more reliable indication of the wider situation (HALS DE/X905/P1).

Trees in the ‘Champion’

With the possible exception of the Chiltern dipslope, the old-enclosed districts of Hertfordshire and Norfolk thus appear to have boasted very high numbers of farmland trees. Much of lowland England, however, did not lie in enclosed fields before the later eighteenth or even early nineteenth centuries, as we have noted, and in these ‘champion’ areas – characterised by extensive, unhedged open fields

54

– there were far fewer trees. Indeed, early advocates of enclosure made much of the paucity of wood and timber within such landscapes, and the hardships which this engendered, and for the most part modern historians have followed them. However, even a cursory inspection of the evidence from Northamptonshire – the main open-field county in our sample - suggests that open-field countryside was by no means lacking in trees. In part this was because, in many parishes, a measure of partial enclosure had occurred while the open-fields remained, for the most part, functioning and open, producing islands of enclosed ground. But it was also because, even where they remained untouched by enclosure, these landscapes contained rather more trees than is generally assumed.

To begin with, we should note that in all open-field districts a small network of enclosed land – hedged (or walled) tofts and crofts – existed in the immediate vicinity of the nucleated settlements, and these appear to have been densely planted with trees. The 237 topographical illustrations made by Peter Tillemans between 1791 and 1721, commissioned by John Bridges of Barton for a proposed history of Northamptonshire, include many views of completely unenclosed landscapes, still dominated by arable open fields, common meadows and commons, which show trees packed very densely in the immediate vicinity of settlements (Bailey 1996). His Western View of … Northampton taken above Kingsthorpe, for example, shows the town with a landscape of open fields, with the vast featureless unenclosed land of Kingsthorpe in the foreground (the landscape here lay largely open until 1766). Yet large numbers of trees can be seen both within and immediately around the town. A similar contrast is shown in his views of the villages of Buckton, Brackley, Helpston, Nassington and elsewhere. Such concentrations of trees within settlement ‘envelopes’ had been a feature of the landscape even in the Middle Ages: in 1377 at Ravensthorpe a house with three tie beams was built entirely using timber growing on the toft belonging to the messuage (NHRO LT64). In Northamptonshire at least, the presence of trees in these enclosed ‘cores’ of land was apparently encouraged by landowners where farms were rented out. A lease drawn up in 1582 between Robert Andrewe Gent of Harleston and Morrice Miles for land in Crick, still entirely open at the time, typically stipulated that the latter was to plant five ash, oak or elm trees every year (NHRO A/089). Planting was not confined to the messuages of tenants, however, but might embrace more public areas of settlements. Morton in 1712 described the number of large elm trees growing in ‘in or near the streets of many of our small towns upon a green bank or some other convenient By Places’ (Morton 1712, 396); while in 1740 ash trees were planted in the churchyard at Drayton (NHRO 292P/001).

We should also note that the unenclosed land lying between settlements was not entirely devoid of trees. Tilleman’s View … taken from the Road between Northampton and Kingsthorpe embraces six unenclosed townships (Kingsthorpe, Upton, Duston, Dalington, Holdenby and Brington). The village ‘envelopes’ are tight-packed with trees in the manner just described, but in the surrounding countryside numerous others – presumably willows - are shown lining the banks of the , and similar stream or river-side planting is depicted in his views of Peterborough, , Fotheringay, and Elton. Documentary records, and maps, similarly indicate the presence of trees elsewhere in the wider landscape. Hedges often existed on the boundaries between adjacent parishes, even when both remained otherwise unenclosed. A document of 1735 thus describes the making of a new hedge between Wilby and Wellingborough, parishes not enclosed until 1801 and 1765 respectively, detailing payments not only for hedging thorns but also for willows (Hall 1995,264; NHRO 350P/90). A map of Denton made in 1760 shows ‘Coopers hedge furlong’ on the parish boundary with Brafield: Denton was enclosed in 1770 and Brafield only in 1827 (Castle Ashby

55 muniments). Hedges also sometimes surrounded the open fields themselves. They are referred to in seventeenth and eighteenth-century churchwardens’ accounts for Crick (Hall 1995, 244), and are shown on maps of Upper Boddington (1758) (NHRO Map 3133); Kings Cliffe (c.1640) TNA/PRO MR1/314); Broughton (1728), Brigstock (1734) (Boughton House archives); and – less certainly – on maps of Holdenby and of (NHRO FH 272). Hedges might also surround the parish meadows, areas of common land, or individual ‘cow pastures’ – that is, parcels of land taken out of the arable in the post-medieval period to increase the area of common grazing. In addition, early maps sometimes show discontinuous fragments of hedge lying within the fields themselves, as at Murcott (1771) (BL ADDMSS 78141 A ); Wollaston (1789) (NHRO 447); Brigstock (1734) and Broughton (1728) (Boughton House archives). An undated document relating to Ecton suggests their possible significance, referring to ‘arbours and shelter-hedges in the beasts and sheep pastures’ which were to be ‘preserved and kept for the sheltering of herdsman and cattle as according to the ancient and laudable manner’. The responsibility for their maintenance lay with the lord of the manor, who was to have ‘the benefit of the lopping popping and plashing of the said arbours and hedges …’ - suggesting both that the hedge was regarded as a fuel source in its own right, and that it was stocked with pollards (NHRO E(S) Box X1071.).

Figure 15. Single trees growing in the open arable at , Northamptonshire, in 1703.

In addition to trees present in hedges, some free-standing trees could be found on commons and cow pastures, and even occasionally growing within the furlongs themselves, to judge from the evidence of seventeenth- and eighteenth-century maps of Broughton (Boughton House archives), (NHRO 1349), Quinton (NHRO 2895), Easton-on-the-Hill (NHRO 6-p/504) and Wollaston

56

(NHRO 4447). Sometimes single trees are shown amongst the arable, as on a map of Ecton surveyed in 1703 (Partdia 2014, 166; NRO Map 2115) (Figure 15); but sometimes several are depicted. A map of Kirby, surveyed in 1580, shows the common called ‘Preste Leys’ with a scattering of trees, but also scattered examples growing within the arable furlongs; while a map of Woodford, surveyed in 1731, shows trees lining the road running across the main common, and growing in lines within one of the furlongs. Crab apples feature regularly in the names of furlongs, as at Cogenhoe, Aynho, Weekley, , Higham Ferrers, and Strixton: in the latter case the tree itself is illustrated by the surveyor (Partida 2014, 165-8). With its twisted and knotty grain crab was ideal for a number of specialist purposes, and its relatively small size perhaps made it ideal for planting on the ends of arable strips.

Many other examples of trees growing within open-field landscapes could be quoted. The enclosure map for Ecton, drawn up in 1759, shows trees lining the road from Northampton to Wellingborough, which passes through the parish: the enclosure award, confirming the course of the road and establishing its dimensions within the new enclosed landscape, specifically states that it should extend ‘three feet in breadth as well on the north as on the south side of the elm trees now standing or growing theron and belonging to Ambrose Isted so as to include the said trees into the said road’ (NHRO Inclosure Vol A p9; Partida 2014). Isted was the lord of the manor and the manorial customs of the township, as stated in 1743, retained to the lord the right to plant on the commons of the township (NHRO E(S) Box X1071). Many enclosure awards describe similar rights. But trees growing on the ends of arable strips, and probably those on ‘leys’ and on cow commons, presumably belonged to the tenants or proprietors of the lands in question. In a legal case concerning in 1807 one witness described how his uncle, Thomas Collis:

Was owner of some copyhold lands in the Open fields of Kettering near the Turnpike Gate on the Road to Barton Seagrave the adjoining Parish. On which Lands he T Collis in the winter about or between the years 1750 and 1760 cut down a very large and high ash tree and also a large oak tree. The oak being so knotty and rough the Bark could not be got off in the usual way was the reason for his cutting the oak in winter, I do not know that the Lords or their Steward knew of his cutting any down (NHRO GK10).

Piecemeal enclosure, although generally on a limited scale in this ‘champion’ region, also tended to add hedges to the landscape, so that all in all it is not surprising that some documents suggest surprising numbers of trees and hedges in still largely unenclosed countryside. The parish of Flore to the west of Northampton still lay mainly unenclosed in 1704, when a terrier was drawn up which provides a detailed description of ‘1 Yardland of arable, meadow, and pasture ground within the open and common fields of Floore’. The property included one headland in Offerlong ‘with the hedge’; land in Middle Field which included grass and meadow ground with three willows in 'Treemer '; seven willows in 'mernes ' and four in Church Hooke ‘feildermost at the south end of Enbrooke Hedge’; together with the hedge at 'outland' and another hedge on Collins Hill, as well as referring to land ‘west of … Collins Hill and the hedge and trees in the hedge’.

Although none of the pre-enclosure maps for Northamptonshire appears to provide a detailed and accurate record of trees in the landscape – of the kind provided by Keymer’s surveys in Norfolk – some seem to attempt a reasonably accurate and comprehensive representation. That for Cogenhoe, for example, surveyed in 1630, shows a landscape of unenclosed open fields but it depicts no less than 350 trees, apparently drawn with some care, growing on the boundaries of the

57 village tofts and crofts, and along the roadsides leading through the surrounding countryside, as well as along the banks of the Nene. The map suggests an average density, across the parish as a whole, of around 0.9 trees per hectare (NHRO Map 5695). A rather more detailed and reliable survey exists for Woodforde, still almost entirely enclosed when surveyed in 1731 (Boughton House archives). This shows 506 trees growing on the banks of the river Nene and its tributaries, and a further 74 in the neighbouring meadows and in the hedges around them; 94 on roadsides away from the village; 37 on the common; and 8 scattered within the furlongs. In addition, 319 are shown growing in the closes around the village, some of which appear like diminutive wood-pastures; this figure omits examples growing in what seem to be orchards. It should be noted that the surveyor clearly omitted many other trees, growing in the hedges surrounding the village closes, because they would have interfered with the depiction of the houses there, so the overall total for the parish, of 1,040, or just over 1 per hectare, is almost certainly a serious under-estimate (NHRO Map 1385). Even where maps show few trees outside the village ‘envelope’, the number present in the tofts and crofts was such that average densities, across the parish or township as a whole, could be surprisingly high. A map of Broughton surveyed in 1728 shows 529 trees in the village closes and although none are shown in the wider landscape this still suggests an overall density for the parish as a whole of around 0.8 per hectare (Boughton House archives).

The scatter of trees in open fields and on roadsides and commons, and the larger numbers on flood plains, is also reflected in the lists of claims made at parliamentary enclosure, to compensate owners for trees left isolated within the new fields allotted to others. That drawn up when the parish of Irthlingborough beside the river Nene, still around 90 percent open when enclosed in 1808, records a total of 3,055 trees, an average of 2 per hectare (NHRO ZA 906). No less than 62 percent were willows, presumably growing on the floodplain of the Nene. Given its purpose, this figure almost certainly excludes most of the trees in the village ‘envelope’, unaffected by enclosure or exchanges. A similar valuation was drawn up on the eve of the enclosure of Finedon in 1806, similarly almost entirely unenclosed at the time, which recorded 1,038 trees (although 55 were ‘thorns’), a density across the parish as a whole of around 0.7 per hectare, but which again must exclude many trees in the ‘ancient enclosures’ around the village itself (NHRO ZA898). Again, the large numbers of willows (410) is noteworthy, presumably growing in the meadows beside the river Ise, which ran through the parish. Another such valuation, for Wilby in 1802, lists 433 trees, within a parish extending over 466 hectares – just under one per hectare, again including numerous willows. An earlier valuation of timber in the parish, made in 1764, lists 762 trees - an average of 1.6 across the parish as a whole, nearly a third of which were willows (NHRO X1657). In neither case is it entirely clear what proportion of these were growing within the closes around the village, but the nature of the former document suggests that most stood out in the fields, some perhaps in the hedge on the boundary with Wellingborough which, as noted earlier, was planted in 1735 (NHRO 350P/90). Other documentary sources suggest similar densities. A survey of 1749, listing the timber on the Boughton estate, details 517 trees in the still unenclosed parish of Luttington, in which around 80 percent of the land was in the hands of the estate – around 1.2 per hectare - although other evidence makes it clear that much of the land within the village envelope itself remained in the hands of small proprietors, suggesting that this figure again needs to be adjusted upwards (Boughton House archives). The same source records 891 trees in the largely open parish of , where the estate owned around half the 678 acres, perhaps suggesting a density of around 2.6 per hectare across the parish as a whole.

58

Although the available evidence does not allow us to provide accurate estimates, most unenclosed Northamptonshire townships and parishes seems to have had average densities of between 1.0 and 2.5 trees per hectare, although as emphasised these were not evenly distributed, but were instead clustered, in particular, in the village closes and on floodplains. In other parts of the Midland plain, outside the areas studied, maps and documents suggest a similar picture with, in particular, phenomenal numbers packed into village closes. A detailed survey of the manor of Milcombe in Oxfordshire, made in 1656, includes a list or ‘what wood and timber is growing on the premises’ of the tenants. It describes how Thomas Burchall had ‘A dwelling house of 4 bayes and a Barne of 3 bayes both in very good repayer, in his backe sid there is growing 70 trees of Ash and Elm 18 of them be very small samplars and there be 48 willows and 2 maple trees.’ Thomas Strainke had a dwelling house, barn and other outbuildings, and a ‘shop house 4 bayes’: ‘In the shope close and orchard there be 126 trees of Ash and Elm whereof 30 of them be very small samplars, in these 2 places there be 52 withes small and great and about as many new planted’. John Pattan had a home close containing ‘117 trees of Ash and Elm of w’ch 44 of them be very small samplars and there is 8 willowes in that Close’; Robert Credwell ‘s home close contained ‘48 trees of Ash and Elm whereof 17 of them be very small saplings and 10 young ash newly sett’; while Edward Butcher had ‘Ash and Elm trees 8 of which there is 43 very small samplars And there be 7 willows. The greatest part of this wood he saith to have planted himself… In Mayton nixt his homestall there is 72 Ashes and Elms great and small, noe timber among them all’. The surveyors concluded that ‘there be very pretty orchards to all these farmes Except Robert Credwells’ (NHRO C(A) Box104 4 1656).

As already noted, in most open-field districts in the Midlands a significant number of townships were completely enclosed, in a variety of ways, before the advent of enclosure by parliamentary acts in the mid eighteenth century. Even in Northamptonshire, the archetypical ‘champion’ county, around half the land area lay in enclosed fields by c.1750. It might be assumed that enclosure, because it introduced a network of hedges in places of the wide expanses of open ploughland and commons, would of necessity have significantly increased the density of trees, perhaps to the kinds of levels we have described in the anciently-enclosed parts of Norfolk or Hertfordshire. But in some cases enclosure may – initially at least – have actually served to reduce tree numbers. This is because the earliest large-scale enclosures in the Midlands, those occurring before the middle of the seventeenth century, generally involved a shift of land use from arable to permanent pasture, and the practise of large-scale sheep-ranching in fields that were often extensive (Taylor 1975, 115-17). The Northamptonshire village of Papley, for example, was depopulated and laid to grass in the 1490s, and a map surveyed in 1632 shows an average field size of nearly twenty hectares NHRO Map 2221). Given that depopulating enclosures like this led to the neglect and eventual destruction of the tight meshes of closes around the village itself, there was often little overall increase in the length of hedges, and sometimes even a reduction, potentially leading to a fall in the numbers of farmland trees. On the other hand, in many cases the pasture fields might themselves be scattered, to varying extents, with free-standing trees, as indeed was the case of Papley, although the admittedly schematic map suggests that their numbers were limited, only around 280 being shown in the whole parish, less than one per hectare. Other maps of early-enclosed townships, similarly fairly schematic in character, hint at similar low densities. Deenethorpe was enclosed around 1585 and a map of 1650 shows only 55 trees in the township – around 0.1 per hectare; a map of Kirby made shortly after enclosure in 1586 shows 164 trees, or 0.3 per hectare; while one of Lower Radstone, surveyed in the late sixteenth century, suggests a density of around 0.45 per hectare (NHRO FH272; NHRO Fh272; NHRO Photostat 1026). Later and perhaps more reliable maps of such early-enclosed parishes

59 suggest, however, rather higher densities of trees, although not much higher than in many unenclosed townships. Loddington, enclosed in 1660, has a map of 1727 which appears to show only c.1 tree per hectare; Halse, enclosed before 1585, has a map of 1725 showing around 1,725 trees, or c.2.4 per hectare, while Newton-le-Willows, enclosed in 1605, has a map of 1717 showing 1,276 trees, or 2.6 per hectare (NHRO T214; Boughton House archives).

Enclosures made after 1660 were less likely to be imposed by large landowners, and less likely to be associated with village shrinkage and depopulation. Even when townships were enclosed in this period under aristocratic direction, the resultant landscapes were often used for mixed farming rather than large-scale sheep ranching. So far as the evidence goes, enclosures of this kind generally created landscapes which boasted significantly more hedges, in part because the villages themselves (and thus the densely-treed enclosures around them) survived, and in part because the mesh of hedges created by enclosure was often denser. This said, the evidence of both maps and surveys suggests much variation in tree densities. Nobottle was enclosed towards the end of the seventeenth century and a map of 1715, which seems to plot trees with some care, suggests a density of nearly 4.2 per hectare (BL ADDMSS 78143). Surveys of the timber on the Boughton estate, made in 1749, indicate that the two Barnwells, both enclosed in 1683, had together around 3 mature trees per hectare; Hemington, enclosed around 1657, had 3.3; while Armston, enclosed in 1683, had 5 per hectare and Kingsthorpe no less than 6 (Boughton House archives). In some places tree numbers were significantly higher. A survey of mainly enclosed land in the parishes of Duke of Powis' estate in Upper and Nether Heyford, Glassthorpe and Newbold from 1758, covering some 770 acres, records a total of 3,004 trees, or 9.6 per hectare, although even this seems to exclude willows as the surveyor made comments like ‘29 Ash and Elm and also many willows’ (NHRO ZB 1837). On John Darker's estate in Gayton, Tiffield, Kislingbury, Milton, Litchborough and Upper Heyford in 1791 no less than 10.6 trees per hectare were recorded (NHRO YZ 2183). All this suggests much variation from place to place, with uncertain causes, but average densities in enclosed parishes in Northamptonshire, by the middle decades of the eighteenth century, of between 5 and 10 trees per hectare, significantly lower than those discussed in the old-enclosed counties of Norfolk and Hertfordshire.

The pattern reconstructed for Northamptonshire and other champion areas of heavy clay appears, in broad terms, to have been shared by open-field districts in ‘sheep-corn’ districts, on light, free- draining land, within the counties studied. In West Norfolk, a map of West , surveyed in the late sixteenth century, shows large numbers of trees in village closes, and in demesne enclosures around the manor house, together with examples scattered across some of the village commons and a particularly dense concentration (presumably comprising alders or willows) on the banks of the (NRO MS 21128, 179X4). Even though it still lay almost entirely unenclosed, more than 1,230 trees are shown, a number which – if roughly accurate - suggests a density across the parish as a whole of around 2.6 per hectare. Similarly, a map of Ickleford in north Hertfordshire, surveyed in 1771, seem to show (albeit schematically) large numbers of trees growing on floodplains, on roadsides and commons, and in closes around the village, as well as in partial enclosures within the fields – perhaps as many as 5 per hectare, averaged across the whole parish (Figure 16) (HALS DE/Ha/P1).

60

Figure 16. Ickleford in the ‘champion’ district of north Hertfordshire in 1771. Although still largely unenclosed, trees were abundant around village closes, beside the river Hiz, and in roadside hedges.

There is, unfortunately, little reliable evidence concerning the character of tree populations in the ‘champion’ districts of Yorkshire prior to enclosure – in the Vales of York and Mowbray, the Wolds, or Holderness – although what there is suggests a similar picture to that in Northamptonshire: a map of Warmsworth, surveyed in 1726, thus shows a scatter of trees around the periphery of the open fields, probably growing in hedges. What is clear is that following enclosure densities of farmland trees were significantly lower than in former open-field districts in the Midlands. Maps of properties in Thornanby, East Harlsey (1762), Pickhill (1778), Birdforth (1770), Easby (1779), Tadcaster (1718), and Helmsley (1785) – all of which appear to mark the positions of trees with some care – are remarkably consistent in showing densities of 0.4 and 2.5 trees per hectare, averaging 1.4 per hectare (Nrth ZNS; Nrth ZIQ; Nrth ZDS M 2/12; Nrth ZMI; Lds WYL68/63). This, however, was part of a wider pattern, one which transcended any division between ‘champion’ and non-‘champion’ areas, for in upland parts of Yorkshire – in villages in the Dales, or on the edge of the Pennines, where open fields were less extensive and a higher proportion of land lay in enclosures by the early modern period – tree densities were higher, but not by a very large margin. Maps of Walburn (early eighteenth century), Leyburn (c.1730), and Downholme (1738) show densities in the range of 1 to 2.7 per hectare, averaging around 1.5 (Nrth ZAZ (M) 1; Nrth ZBO (M) ½; Nrth ZBO (M) 1/5). Muir’s research into the ancient pollards of Nidderdale suggests rather higher densities: approaching 8 per hectare on one map of Clapham Green, for example (Muir 2000, 108-9). But overall, trees densities in Yorkshire, at least by the eighteenth century, were significantly lower than in the other areas studied.

61

Although the character of the evidence is somewhat uneven, densities of farmland trees thus clearly displayed a considerable degree of variation in post-medieval England. In Midland open-field districts before enclosure, if we include trees growing in the village crofts, average densities of between 1 and 2.5 per hectare were common, rising after enclosure to between 5 and 10 per hectare. In marked contrast, anciently-enclosed areas in the south east and East Anglia, to judge from the evidence from Norfolk and Hertfordshire, boasted phenomenal densities of trees, generally above 15 per hectare, commonly in excess of 25 and not infrequently reaching 40 or more. In Yorkshire, however, there were few farmland trees. In former open-field districts in the Vales, following enclosure, there were seldom more than around 2.5 per hectare. In the Dales, and on the edge of the Pennine, this might approach 4 per hectare, but seldom if ever higher. Muir’s research into the ancient pollards of Nidderdale suggests rather higher densities: approaching 8 per hectare on one map of Clapham Green, for example (Muir 2000, 108-9). But overall, trees densities in Yorkshire, at least by the eighteenth century, were significantly lower than in the other areas studied.

The management of farmland trees: pollards.

Not only the numbers of farmland trees, but their mode of management, displayed marked regional differences. The precise proportion of trees managed as pollards is difficult to ascertain because, as already discussed, early surveys sometimes omit some, or all, of the non-timber trees. Some surveys explicitly excluded them all, such as one of Tarrants Farm in St Stephens, Hertfordshire, made in 1733, which simply states ‘Pollard trees not reckoned’; elsewhere their omission is suggested simply by the remarkably low densities of trees recorded, compared to neighbouring places at around the same time. Where surveys do appear to have been comprehensive, the numbers of pollards present on farmland in the seventeenth and eighteenth centuries is often remarkable. On the boulder clays of east Hertfordshire, for example, 85 percent of the trees recorded on three farms on the manor of Great Barwick in Standon in 1778 were pollards, while on the demesne the figure was 84 percent (HALS A/2832 and 2833). On land at Whempstead, Little Munden, in 1808 83 percent of the recorded trees were pollards; on Pearces Farm in Thorley and Sawbridgeworth in 1807 the figure was the same; while on a farm at Puckeridge in 1778 it was 79 percent (HALS 81750; HALS DE/H/P16; HALS 43754). Some places in this district, especially in the early eighteenth century, could boast even higher figures. On the Cowper estate around Hertingfordbury 91 percent of the trees recorded on a map of 1704 appear to have been pollards (HALS D/EP/P4); while the 1727 survey of Kelshall, already mentioned, not only reveals a phenomenally high density of trees, but also that a very high proportion – 93 percent - were pollards (HALS DE/Ha/B2112). Even in 1774, when a further survey was made of this 33-acre property and the numbers of trees had been significantly reduced, no less than 86 percent were still being managed in this manner (HALS DE/Ha/B2112).

On the London clay soils in the south of the same county the proportion of pollards was even higher. On seven farms on the Broxbournebury estate in 1784, covering 373 acres (151 ha), there were 3,012 pollards but only 299 timber trees (HALS DE/Bb/E27); the greatest density was recorded on West End Farm in Wormley, where there were no less than 1,358 pollards but only 148 timber trees. On a farm in Hatfield and North Mymms surveyed in 1723 pollards made up around 74 percent of the trees (Austin 2013); but an 89-acre farm at Hammond Street, Cheshunt in 1650 apparently contained no timber trees at all but as many as ‘927 lopt Pollardes of oake Ash and Hornebeame’ (PRO/TNA E317/Herts/24, p. 37). On the Broxbourne Mill estate in the late eighteenth century, over

62

80 percent of the trees on the higher ground seem to have been pollards, while on the meadows the figure may have been as high as 93 percent (HALS B479). The data from the west of Hertfordshire show fewer trees, as we have seen; and also a lower proportion of pollards, with just 7 percent recorded on the ‘Three Fields called Church Lands’ in Watford in 1723 (HALS DP 117 25/7). As emphasised earlier, however, the sources for this district are relatively sparse and relatively late, and overall it is the very high density of pollards which is the most striking features of the oldenclosed Hertfordshire landscape in the seventeenth and eighteenth centuries.

The situation was broadly similar in the old-enclosed parts of Norfolk. On the boulder clays in the south and centre of the county pollards commonly made up more than 70 percent of the farmland trees recorded on maps and surveys before the mid nineteenth century. On the farms mapped by Henry Keymer in 1760s at Beeston and Scarning, 79 percent and 76 percent of the trees were pollards, although at Bintree the figure was only 40 percent (NRO WIS 138, 166X3; NRO BCH 20; NRO Accn. Field 18.3.82 P188B) (Figure 17). On the two farms in Whissonsett recorded on an undated mid-eighteenth-century map 62 percent and 72 percent of the trees were pollards (NRO HIL 3/34/1-27, 879X3); on a property in Wendling in 1777 73 percent were pollards (NRO EVL 348/1/1 and 29); while a survey of ‘Mr Fromours Estate’ in Pulham made in 1751 recorded 41 timber trees but no less than 472 pollards: even if we count the additional 46 ‘stands’, or young trees, with the timber, the pollards still made up over 84 percent of the trees on the farm (NRO DN/MSC 2/2225). In 1803, on a farm at Loddon, 76 percent of the recorded trees were pollards; while a timber survey of 1813 suggests that 89 percent of the trees (excluding saplings) on a farm at , probably located within that portion of the parish extending onto the boulder clay plateau, were pollards (NRO HNR 149/3). Stray comments in letters, journals and valuations similarly often imply that pollards were more numerous than timber trees. In 1792, for example, the Diss attorney and maltster, Meadows Taylor, described how on his properties in Palgrave (in fact just over the county boundary in Suffolk) there were ‘several good timber trees and many pollards growing upon the lands’ (NRO MC 257/23/5, 638 x 8).

On average, around 72 percent of the trees on the Norfolk claylands appear to have been managed as pollards in the period before the mid nineteenth century. In those parts of the light loam district of north east Norfolk which had been enclosed into hedged fields densities were similar but slightly lower. The 1722 survey of the Earle estates, which details some 24,000 farmland trees, divides them into three categories: timber, pollards, and ‘storrels’ or young trees (NRO BUL 11/283, 617X2). Leaving the latter aside – for they might eventually be managed as either timber, or as pollards – around 64 percent of the trees were pollards. Similar figures are available from elsewhere in the district. On a property in Burgh and in 1816 63 percent of the trees were recorded as pollards (BR 90/35/22). A timber valuation drawn up in 1775 for an estate in Gunthorpe, in the north of the county, records rather lower numbers – 866, as opposed to 658 timber oaks and 815 ash (and large numbers of ‘standils’) - 37 percent - but it is uncertain whether only the pollards with

63

Figure 17. Extract from Henry Keymer’s map of an estate at Beeston, Norfolk, showing the dominance of pollards. a timber value were noted by the surveyor (NRO NRO BL/CS 1/20/3/1-5). Seventy percent of the trees recorded on ‘Mr Boorns estate at Coltishall’ in 1813 were pollards (NRO HNR 149/3). Although the proportion of pollards in this district may have been less than in the south or centre of the county, or than could found across most of Hertfordshire, they probably accounted for around 60 percent of all farmland trees in the period before the mid nineteenth century. Indeed, even in open- field districts within Norfolk, or in places where open-fields had only recently been enclosed, the proportion of pollards was often very high, accounting for 100 percent of the trees on a property in in 1651 (NRO WLS XXVI/4, 414X6); around 56 percent of the trees recorded on a farm at Hockwold in 1816 (NRO BR 90/34/16); 75 percent of the trees recorded on one farm at in 1805 (a parish enclosed by an act passed in the following year) (NRO MS 13751, 40E3); and a similar proportion of those valued in the exchanges when were enclosed in 1777 (NRO C/Sca 2/135).

Yet further into the heart of the champion, in Northamptonshire, the proportion of pollards was noticeably lower. A survey made in 1749 of the timber on the estates of the Duke of Montagu in the east of the county suggests that, on property lying within nine townships enclosed within the previous century or so, only between 9.5 and 25 percent of the trees were pollarded.4 In Oakley and Brigstock, parishes which were not enclosed until 1807 and 1795 respectively but where the majority

4 In the two Barnwells, entirely enclosed in 1683, 25 per cent of the trees were pollards; while Armston, Kingsthorpe, Stoke Doyle, Hemington and Newton, with 12.8 per cent, 9.5 per cent, 10.6 per cent, 17.3 per 64 of the estate land lay in enclosures made piecemeal from the fields, 22.8 percent of the recorded trees were pollards. Similar proportions are apparent on enclosed land elsewhere in the county. In 1791, 26 percent of the trees listed on John Darker's estate in Gayton, Tiffield, Kislingbury, Milton, Litchborough and Upper Heyford were described as pollards (NHRO YZ 2183). Only in some unenclosed parishes are higher densities occasionally to be found, although even then not invariably. Of the Montagu estate in 1749, 50.4 percent of the trees in Geddington, almost entirely open until enclosed by act in 1807, were pollards; but in , which likewise lay open until 1807, the figure was 29 percent; while in Luttington, largely unenclosed until 1807, it was 23 percent. Only in Hanging Houghton, a parish which remained substantially unenclosed until 1794, was the proportion of pollards – 78 percent of the total - comparable to those found in Hertfordshire or Norfolk, although the estate held only a single small farm there. In Great Doddington, similarly, largely unenclosed until 1766, a survey of 1764 lists 380 trees growing on one estate, of which 200 (53 percent) were dotterel ash and a further 85 (20 percent) dotterel ‘salley’, or sallow. This said, other parishes seem to have had few pollards by the time they were enclosed. The same survey also deals with land in Wilby, a parish which lay largely open until 1801, and here only 108 trees – 14 percent of the total – were recorded as ‘dotterals’, presumably pollards, all of which were ash (NHRO X1657). In 1806, when the parish of Finedon was enclosed, only 11, out of a total of over a thousand trees, were specifically described as pollards, although there are grounds for believing that most of the 410 willows listed were also managed in this way (NHRO ZA898). Overall, timber surveys from the eighteenth century thus suggest that in most Northamptonshire townships between 10 percent and 50 percent of hedgerow farmland trees were managed as pollards in the seventeenth and eighteenth centuries, averaging around a quarter. It is true that, by their very nature, such sources tend to highlight practice on estates and tenanted farms, and that on small freehold properties the proportion of pollarded trees may have been higher: but the contrast with places like Norfolk and Hertfordshire, where upwards of 70 percent of trees were routinely cropped, is nevertheless striking.

In the fourth of our study counties, Yorkshire, the proportion of pollards in the seventeenth and eighteenth centuries was even lower. Indeed, pollards are seldom mentioned at all in the surviving sources, and references to ‘dotterels’, while in most cases probably indicating pollards, may not always do so. A survey made in 1642 of timber at Helmsley, on the northern margins of the Vale of Pickering, thus shows that most of the trees so described were to be found on farmland, with virtually none in the woods, suggesting that they were hedgerow pollards, or possibly the remains of wood-pastures (Nrth ZEW IV 1/6). Either way, they were not numerous, accounting for only c. 14 percent of the trees recorded (1,269 out of 9,234). Against this, we should note Muir’s research in Nidderdale, which identified 186 former pollards standing or extant boundaries, or on ones surviving only as earthworks, albeit scattered over an extensive area of the Vale, from Nidd to Stonebeck (Muir 2000). Some of the early pollards noted by Fleming in Swaledale were likewise boundary trees, rather than relics of early wood-pastures (Fleming 1998). Moreover, leases and accounts from the cent and 22 percent, were completely enclosed in 1683, 1580, 1689, 1657 and 1605 respectively. The trees listed under ‘Boughton and Kettering’, where only 7.7 percent were pollards, probably lay mainly in Boughton, enclosed in c.1540 and the centre of the estate; while in nearby Weekley, where a mere 6.8 percent were pollards and the parish remained three quarters open until 1807, most of the trees listed presumably grew in the old-enclosed eastern portion of the parish, given that they are described as being in the ‘manor hedgerows’.

65 seventeenth century often refer to the right of ‘greenhew’, often interpreted to mean pollarding, although in some cases that right was restricted, as at Settrington, where the tenants could only cut ‘small writhing bands for tying up ther cattell & making Harrow withes’ (King and Harris 1963, 10). Willow pollards may have been a common feature in some Vale townships, as they were in the ‘champion’ districts of Northamptonshire, and examples still survive in places like Fishlake, to the north east of Doncaster (Jones 2012, 184). Nevertheless, there is little doubt that pollarding was a less important practice in Yorkshire than in the other areas studied, at least by the time that documents become abundant, from the eighteenth century: references to pollarding as a practise are virtually non-existent, although there are some to what may be former pollards, such as the description of ‘distorted rickety trees of elm and ash’ in late eighteenth-century correspondence from the Bolton estate, or the ‘166 oakes or hedgehogges’ noted in a 1772 valuation from the same estate (Cald H1/YT/1772/Jan 15). Tuke in 1800 could still describe how the tenants ‘on some estates’ were allowed to crop hedgerow trees, ‘in doing which, large boughs are frequently cut off in a very rough manner’ (Tuke 1800, 191). On balance, however, the evidence suggests that the pollarding of farmland trees declined rapidly in most parts of the county from the seventeenth century, and was a comparatively unusual practise by the nineteenth century; even in the seventeenth century it was probably carried out at a lesser scale than in the southern counties studied. The management of pollards in wood-pastures may have disappeared from most parts of Yorkshire even earlier.

The management of timber trees

The most striking difference between modern tree populations, and those found in rural areas before the second half of the nineteenth century, was that the latter included few old trees, other than pollards. Most surveys and timber books give volumes, rather than girths, of trees, and it is not easy to estimate age from the former; nor is it always clear whether measurement are in cubic feet or in the unit of Hoppus feet, standardised in 1736 by Edward Hoppus to take account of the losses incurred when a log or stem is converted into timber (a Hoppus foot is 1.273 times the size of a cubic foot). Very occasionally, however, documents provide both volume and girths. Most are, in fact, quarter-girths, a unit widely employed since the seventeenth century and which, when squared, provides a good approximation of the area of the square that can be fitted within the circumference of a tree, which is itself a good prediction of the volume of usable timber. Figures for oak, ash and elm trees growing on Mrs Shepherds estate at Harpenden in Hertfordshire are provided in a document from 1775 (HALS 51124). Oaks with quarter-girths of around a foot (11 inches and 12 ½ inches) contained 11.75 and 14 cubic feet of timber; those with quarter-girths of around 1 foot 6 inches contained between 32 and 57 ½ cubic feet. A much more extensive collection of figures, from Aspall in Suffolk (outside the areas studied), dating to the period between 1728 and 1763, suggests that oaks with quarter-girths of around a foot generally contained 20-25 cubic feet of timber, but sometimes as much as 35. From these figures we might tentatively suggest that most oak trees containing less than 40 cubic feet of timber will have been less than 50 years old, and often much less; and that trees containing 20 cubic feet were generally be less than 30 years old – figures confirmed by discussions with modern foresters. Unfortunately, we have less information for ash or elm, although the figures from Harpenden in 1775 are broadly in line with those for oak.

On this basis, the timber volumes provided by eighteenth and earlier nineteenth-century surveys suggest that countryside must have been full of very young trees, mostly felled before – and often long before - they were middle aged, generally before they were seventy years old and often before

66 they were 40. On the glebe lands of Merton in Norfolk, for example, in 1846 only eight of the 68 oaks contained more than 40 cubic feet of timber; even the largest, with 64 feet, was not necessarily much older than eighty years (NRO WLS LXX/22/1-41, 481x7). The average contents of 762 oak trees growing on the Stow Bardolph estate in Stow and Wimbotsham in 1815 was 18.1 cubic feet, and that of the oaks growing in Shouldham Thorpe, Shouldham, Marsham and Fincham a mere 10 cubic feet (NRO HARE 5500, 223 x 1). The 85 oaks growing at Hindolveston in 1704 had an average volume of around 25 cubic feet (NRO DCN 59/21/3); while on an estate in , and Brinningham in in 1739 the average volume of 2,983 oaks was just under 11 cubic feet (NRO NRS 16112, 32A6). Such average figures obscure the fact that individual oaks might attain 45 or even 60 cubic feet of volume; and on some estates the average volume was itself significantly larger, reaching 28 cubic feet on an estate in at , Bintree and Biillingford in 1796, and 31 cubic feet at Harpenden in 1775 (NRO 20757A). But overall, relatively few oak trees were allowed to grow larger than 40 cubic feet, that is, to attain an age of more than 50 or 60 years. On the Boughton estate in Northamptonshire only 1.8 percent of the trees growing in the hedges of Weekley in 1749 were oaks with more than 20 cubic feet of timber. On Warton manor the figure was about 5.5 percent; on Newton manor, 5.7 percent; in Stoke Doyle, 5 percent; in Little Oakley and Brigstock, 4.3 percent; in Geddington, 5.8 percent; in Kettering, 3.3 percent; in Kingsthorpe, 2.7 percent; in Hanging Houghton, 2.2 percent; in Barnwell, 1.8 percent; in Hemington, 0.9 percent; and in Armston a mere 0.8 percent (Boughton House archives).

The volumes of the ash trees in the various surveys are slightly smaller, averaging around 15 cubic feet but with individual examples sometimes reaching 35 cubic feet, although on the Boughton estate no examples at all were recorded, out of more than 20,000 farmland trees, with a volume in excess of 20 cubic feet. With elm the situation is less clear, as there are fewer available figures, but while some large examples are recorded (one with a volume of 64 cubic feet at Harpenden in Hertfordshire in 1775: HALS 51124) most were smaller, generally around 14 cubic feet, and at Boughton in 1749 no examples exceeding 20 cubic feet were noted on the whole estate. Such young populations were not confined to the south and the Midlands. In Yorkshire, analysis of the principal timber surveys, and reconstructions of tree size based on bark quantities, suggests a mean farmland tree size of only 16 cubic feet: Marshall thought that the average age of the farmland trees in the Vale of Pickering at the end of the eighteenth century was around 50 years, although whether he was thinking in terms of a mean, or median, is unclear (Marshall 1788). In some places, in the county, it is true, higher average volumes are recorded, but they were never much higher. A particularly detailed survey of farmland trees at Newburgh on the edge of the Howardian Hills in Yorkshire, made in 1774, can be used to plot frequency distributions based on volume (Nrth ZDV). The 126 oaks and 195 ashes displayed relatively broad distribution curves, reflecting a mixture of different sizes. For both species the most common volume was around 20 to 30 cubic feet, but some oaks were again allowed to grow larger, with a greater proportion of trees above 50 cubic feet (around seventy years), resulting in a mean average of 46 cubic feet. Clusters of trees around 100120 cubic feet may reflect a small number of even older trees, perhaps retained for particular purposes. A survey of the estate in Hertfordshire, carried out on a field-by-field basis in 1719, does not specify the species of tree but again suggests a rather wider range of volumes and ages (Austin 2008). While the average volume of timber was estimated at 29 cubic feet, the largest example reached 315 cubic feet, clearly a very large and old tree. Some older farmland trees thus existed, in all areas, but most – probably well over 90 percent - were evidently felled before they were sixty or seventy years old.

67

Explaining the numbers and management of farmland trees

So far as the evidence goes, the relatively low age of standard trees growing in farmland was a feature of all the districts studied. It was probably the consequence of a number of factors. The most important was that, prior to the development of large commercial sawmills in the nineteenth century, it was easier to select timber which were already roughly the correct size for a particular job, rather than to allow all or most trees on a property to grow to a larger size, and then saw them up to the dimensions required. Such an approach made sense given that, by around eighty years, the growth rate of oak, in particular, slows noticeably anyway. In addition, very large trees adversely affected the crops growing in adjacent fields, and in this context it is noteworthy that, where the evidence is sufficiently detailed, we can sometimes see that the largest specimens were to be found growing in meadows, parks and (to a lesser extent) pastures, rather than in the hedges around arable fields.

The fact that bark is more easily peeled from younger trees, and is of better quality for tanning, may also have been an important consideration. Bark from oak trees was an important influence on the economics of management from at least the late Middle Ages, and by the late seventeenth century the tanning industry was being organised on a large scale, especially in Northamptonshire, with large tanneries being established in places like Kettering and small local tanneries closing. Consumption rose dramatically during the eighteenth century, and especially after c.1750, as the population grew rapidly and living standards improved: in the 1720s around 50,000 tons were used by the industry each year in England, but by the early nineteenth century the figure was above 90,000 tons in most years (Clarkson 1974). Nicol commented in 1820 that ‘the very high price of bark for the last seven years has undoubtedly lead to many premature falls’ (Nicol 1820, 3). The consumption of homegrown bark probably then stabilised, and gradually declined, due to imports, and more dramatically from the 1850s, as alternative methods of tanning, using in particular chromium sulphate and other chromium salts, was increasingly employed by the industry. The best bark for tanning came from oak coppice poles around twenty years old, or young oak timber of similar age; the worst came from old trees, especially ancient pollards, which were also difficult to peel. In 1726, Daniel Eaton wrote that he had sold trees with an allowance ‘for the bark at 6s 6d in the pound for the saplins, without any allowance being made for the doderells…….for we are assured that a great number of them will not peel’ (Wake and Webster 1971, 100-103). Bark was thus a valuable commodity, and that stored at a tannery described in the inventory of widow Judith Day of Hertford, taken after her death in 1688, represented 16.5 percent of the total value of the business (£339 10s 6d), and was worth almost as much as all of her household goods (£58 12s 2d) (Adams 1997, 100–3).

Whatever the explanation, the early felling of standard trees appears to have been a feature of forestry practise shared by all the districts studied. This contrasts with other aspects of farmland tree management, especially the numbers tolerated in hedges and in fields, and the relative proportions managed as standards or pollards. To some extent variations in tree densities followed, as we might expect, the familiar distinction between open-field and old-enclosed districts – between ‘champion’ and ‘woodland’ areas: an abundance of trees was associated, not surprisingly, with an abundance of hedges. But this was true only to an extent, for even after open arable was removed in ‘champion’ areas in the sixteenth and seventeenth centuries, the numbers of trees usually remained lower than in districts more anciently enclosed, while in Yorkshire trees were equally sparse on enclosed as on unenclosed land.

68

What is particularly striking is that the density of farmland trees, and the mode of their management, appear to have been broadly correlated: the higher the density of farmland trees, on the whole, the higher the proportion cropped as pollards. At one extreme, Yorkshire thus had few pollards, and densities of trees which seldom rose above c. 4 per hectare. At the other, in old- enclosed parts of Hertfordshire and Norfolk, over 70 percent of trees were pollarded, and there were generally more than 15 farmland trees per hectare, and often more than 25. Northamptonshire, where pollards were present in moderate numbers – generally comprising around 20 - 25 percent of the trees – and where, even in enclosed parishes, there were rarely more than 10 trees per hectare, falls somewhere between these extremes. In fact, differences in the overall density of farmland trees across the enclosed land in the various districts studied are largely, although not entirely, accounted for by the numbers of pollards; the density of timber trees across different regions displays much less variation.

What mainly needs to be explained in thus the numbers of pollards present in different areas, and this in turn raises the question of what the material cut from pollards was actually used for. A number of historians and ecologists have highlighted the importance in the pre-industrial economy of ‘leafy hay’, and in some northern districts of England ash, in particular, continued to be cut to supply winter browse even in the twentieth century (Fleming 1997). In general, however, there is rather less evidence for this practise than is sometimes suggested, at least by the early modern period. Gledhill's (1994, 305–6) analysis of leases and accounts in Swaledale, for example, implies that leaf foddering had all but ceased by the seventeenth century. Early agricultural writers describe how branches might be cut for cattle feed in the winter - but not in the summer, and then stored. Thomas Tusser for example refers to the practice of feeding the cuttings from pollards – and the trimming of timber trees – to cattle during the winter months, but his wording implies that this was as a side-product of cutting for fuel, the cattle presumably eating the twiggy growth and leaving the rest for firing (Tusser 1573, 4). Most seventeenth- and eighteenth-century commentators simply discuss pollards in terms of fuel. Nourse for example described how hedgerow oaks, elms, ashes ‘and the like’ could be lopped for fuel (Nourse 1699, 28); Blome discussed ‘those Trees not fit for Timber, but such that are designed for present use, for Fewel &c.’, which could be ‘lopped or shredded, at convenient seasons’ (Blome 1686, 254). Hedgerow trees provided a supply of ‘Fewel, so deficient in many Champaine Countreys’ (Blome 1686, 250). Poles cut from pollards were clearly used in other ways, for fencing and the like; but fuel was evidently uppermost in the minds of contemporary writers. Local documentary sources give a similar impression. A survey of the Blickling estate drawn up in 1756 described one farm in Wymondham on which there was ‘Little or no Timber & scarcely Pollards enough for hedge stakes the Tenant is therefore obliged to buy Coals. It lies within two miles of Norwich… ‘(NRO MC 3/252, 468X4).

It should be emphasised that seventeenth- and eighteenth-century writers commonly noted the disadvantages to growing trees in hedges. Smith for example suggested that they were not good ‘either for Corn or Grass’ in the adjacent fields, and weakened the growth of the hedge plants beneath (Smith 1670, 30). Nourse went to far as to assert that open fields were generally held ‘in greater esteem’ than enclosures, ‘where every Man’s property is secur’d by Fences’, partly on the grounds that ‘Corn never ripens so kindly, being under the Shade and Droppings of Trees; the Roots likewise of the Trees spreading to some distance from the Hedges, do rob the Earth of what should nourish the Grain’ (Nourse 1699, 27). Hedgerow trees, he suggested, also provided a home for birds which would predate on grain crops. But where there was a particular need for fuel such disadvantages were clearly ignored, and as well as the scatter of timber trees, found throughout the 69 country at a density of between around 3 and 7 per hectare, local people filled their hedges densely with pollards. Conversely, where alternative sources of domestic fuel were available, the incentives to do this were less. Variations in the numbers of pollards thus need to be understood primarily within the context of regional fuel economies.

In the period before the nineteenth century a wide range of material was used for domestic or industrial fuel in England. William Harrison for example, writing at the end of the sixteenth century, argued that the shortage of wood in the vicinity of London would soon drive the inhabitants of the city to burn ‘fenny bote, broom, turf, gall, heath, furze, brakes, whins, ling, dies, hassocks, flags, straw, sedge, reed, rush and also seascale’ (Edelen 1994, 281). Many of these materials were derived from heathland, especially broom, gorse, and heather, the latter cut in the form of turves dug to a depth of at least 2.5 cm, which thus included both the vegetation and, more importantly, a portion of matted root material. All were used as domestic firing, and especially for heating ovens, but were also employed industrially, especially for firing brick kilns. When Blickling Hall in Norfolk was constructed in 1617-21 more than a million bricks were fired in kilns entirely fuelled with gorse and broom faggots brought from the nearby heaths, and well into the nineteenth century most of the kilns on the Bedfordshire brick fields were likewise fired using heathland vegetation (NRO MC3/4; Cox 1979, 27). It has been suggested that the vitrified bricks so common in seventeenth, eighteenth and even some nineteenth-century buildings throughout England, and often used in a decorative fashion, may have been the consequence of using such fuel: ‘…the high proportion of vitrified headers was probably a consequence of using firing materials, such as heather and gorse, which gave off fumes containing potash’ (Cox 1979. 28-9).

Where demand was high or alternative firing in short supply, heathland vegetation might be carefully nurtured. Indeed, in the southwest of England gorse was often grown in private enclosures. Even in 1801 it was said of Sithney in Cornwall that:

Here are, it is almost literally true, no trees; consequently a considerable part of every estate is under furze, which would frequently, with proper cultivation, produce whatever the cultivated lands now produce (Turner 1982, I, 33-4).

Most of the areas where these materials grew were common land, however, where even the production of heather might be limited by the intensity of grazing. Careful management, often involving the ‘doling’ of certain areas to individual commoners, might therefore be necessary in order to preserve areas of combustible vegetation. In the early seventeenth century Thomas Blenerhasset could comment of Horsford Heath near Norwich that ‘This heathe is to Norwich and the Countrye heare as Newcastle coales are to London’ (Barrett-Lennard 1921, 20). But on many lowland heaths grazing clearly limited the production of fuel. John Norden, writing in 1618, described the gorse in the West Country, which grew ‘very high, and the stalke great, whereof the people make faggots’. He continued:

And this kind of Furse groweth also upon the Sea coast of Suffolke: But that the people make not the use of them, as in Devonshire and Cornwalle, for they suffer their sheep and cattell to browse and crop them when they be young, and so they grow too scrubbed and lowe tufts, seldome to that perfection that they might be (Norden 1608, 235).

70

Harrison’s ‘fenny bote’, or peat, was a far more important fuel source. In East Anglia the progressive drainage of the Fens during the seventeenth century greatly reduced the availability of peat, although smaller deposits continued to be exploited in areas like the Norfolk Broads, and even in south peat continued to be dug on some scale well into the twentieth century (Day 1999, vi). In Lincolnshire, in contrast, and on the Isle of Axolme especially, it remained a significant fuel source, Eden reporting in 1791 that it was ‘the usual fuel consumed by labourers’ in Lincolnshire (Eden 1797, 566). But it was in upland districts, where raised bogs supplied ample reserves of sphagnum peat, that it really retained its importance throughout the post-medieval period (Rotherham 2011). At Bolton in Westmoreland in the early nineteenth century, for example, it was argued that if the common was not enclosed it would soon be completely ruined by peat digging (Whyte 2003, 33). Peat was extracted on a commercial scale in the mosses of south-west Lancashire well into the eighteenth century and supplied to Ormskirk and Liverpool in a mixed economy alongside the output of the local coalfields (Langton 1979, 56-7). Like heathland vegetation, peat was employed industrially, as well as domestically, usually in the form of peat charcoal. It was used to fire lime kilns and, alongside wood charcoal, for smelting tin and other non-ferrous metals, only being finally replaced by coal in the first half of the eighteenth century following the development of the reverberatory furnace (Palmer and Neaverson 1994, 128). The extent to which peat remained a major source of fuel in upland districts is clear from the way that parliamentary enclosure awards often allocated, to those receiving allotments, a ‘moss dale’ where they could cut peat: ‘these usually show up on award maps as relatively small areas divided into large numbers of narrow rectangular strips’(Whyte 2003, 76). Such allotments continued to be made into the middle decades of the nineteenth century, as at Troutbeck in Westmoreland in 1840. They were exploited by the farmers, as much as by smallholders and cottagers.

The other key fuel source in post-medieval England was of course coal. This was unquestionably in widespread use as a domestic as well as an industrial fuel across much of northern England, and in parts of the Midlands, by the middle of the seventeenth century, but its adoption elsewhere was gradual, and at first often restricted to the wealthier elements in society. In Herefordshire in 1805 for example it was believed that ‘Coal is in general use as fuel by as many of the inhabitants as can afford the purchase of it’ (Ducumb 1805, 141); while a government survey of 1791 described how it was the most important fuel in Devon in the ‘Houses of Creditable People, but the poor burn no Coals, and very little Wood, on Account of the Expense... most of their Fuel is Turf or Peat’. The same report makes it clear that by this stage, most areas of the country were at least partly coal-using, and that in Durham, Yorkshire, Nottinghamshire, Staffordshire, Lancashire and Cheshire other fuels played little part in the local economy (House of Commons Papers 1792). But this penetration of coal into local markets, throughout the country, appears to have been a relatively recent development, connected with the progressive improvement of transport systems which occurred in the middle and later decades of the century. Most places in England and Wales lie at some distance from ready supplies of coal - at the start of the eighteenth century some three-quarters came from mines in Shropshire, Staffordshire, Yorkshire and the north-east of England5 - and so the nature of transport systems was clearly of crucial importance in deciding the distribution of its use, and thus the economic viability of other, less attractive local fuels. Contemporaries agreed on the crucial

5 Calculated from Hatcher, British Coal Industry, p.68. The proportion is much the same if we take instead the possible minima Hatcher provides for output. The north-east produced up 1.3 million tons, Shropshire 230, 000, Staffordshire 150, 000, and Yorkshire 150 000 out of a total of 2.3 million tons. 71 importance of water transport in structuring the use of coal. William Harrison, writing in the 1570s, noted how coal use was at that time just beginning to spread ‘from the forge into the kitchen and hall, as may appear already in most cities and towns that lie about the coast, where they have little other fuel except it be turf or hassock’ [our italics] (Edelen 1984, 281). This importance of water transport was still being emphasised by Pehr Kalm in 1748, who observed that coal could be found in London, and was widely burned in villages within a fourteen mile radius, but ‘in places to which they have not any flowing water to carry boats loaded with coals’ the population continued to burn wood - mainly from ‘trees they had cut down in repairing hedges’ - or ‘fuel of some other kind, as bracken, furze etc’.

The character of fuel supply, then, was almost certainly a crucial factor in determining the numbers of pollards which people were prepared to tolerate on farmland, and thus to a large extent the overall density of farmland trees. Where coal or peat were abundant – and perhaps where heaths were extensive and lightly grazed – farmers and landowners would have been less keen to fill hedges with trees. But the situation was further complicated by the presence of large urban centres, and the demand for fuel which they provided: places like London, or even Norwich, critically shaped the character of land management in their hinterlands. The proximity of London, for example, almost certainly explains the very high density of pollards in the south of Hertfordshire. Complex patterns of supply and demand, variable over time and space, thus underlie the variations in the character of tree populations revealed by maps and documents. This said, looked at in these terms, the variations in the numbers of pollards which we have noted between and within the four counties appear less mysterious.

Yorkshire, where pollards were thin on the ground, thus had ample supplies of other fuels. Peat and moorland vegetation like heather and gorse were widely available in the Pennines, the Dales and the North York Moors, and were locally used as fuel well into the nineteenth century, and perhaps later in remote districts (Rotherham et al. 2004). In Helmsley, for example, the manor court regularly dealt with cases of illegal peat and turf cutting as late as the 1860s (Nrth ZEW III 7/2). The use of these materials for firing extended to lowland sites, where bogs and moors remained unenclosed, and even to the cities, with sixteenth and seventeenth-century York being served by peat brought from Inclesmoor (Hatcher 1993, 124). More importantly, by the sixteenth century the output of the South Yorkshire coalfield was expanding rapidly: and while the east of the county, around the Wolds, lay some distance away, it was serviced from an early date by boats bringing coal from the coalfields of Durham and the north east. Not surprisingly, perhaps, as early as the 1530s John Leland was able to describe how, although wood was plentiful across much of the county, many people were burning coal. Population levels were rising rapidly during the seventeenth and eighteenth century, but these sources evidently kept pace with demand to such an extent that farmland trees never became a major source of fuel, although presentments for taking wood, cutting trees and breaking hedges in manorial courts, as at Newburgh in the mid-seventeenth century, or at Helmsley in the mid- eighteenth, may indicate that in some areas there were some fuel shortages, amongst the local poor at least, and that firewood continued to have an importance at a local level (Nrth ZDV V 30). This said, many of the references to hedge breaking probably relate to social protest and enclosure disputes, rather than to the actual gathering of fuel (Blomley 2007; McDonagh 2009).

Northamptonshire is a more complex case. Although there were no mineable coal seams in the county, the Warwickshire coalfield lay no more than 22 kilometres beyond the county boundary, and

72 the Staffordshire field less than 50. Both were producing coal on a significant scale by the seventeenth century. Morton in 1712 emphasised how ‘the remotest part of the county is not twenty miles from the seats of coal, either of the inland or of the Newcastle sort, which is brought so nigh as Stanford, Bedford and by water’ (Morton 1712, 16). Even in the sixteenth and early seventeenth centuries leases sometimes included a stipulation that the tenant would, in addition to any cash paid for the use of a particular property, agree to transport coal for the leasor. One example, relating to land at Crick in 1582, described how the tenant was bound to cart one load of coals each year to the proprietor’s house in Harlestone; another, from 1630 and for land in Hellidon, obliged the tenant to carry one load every other year ‘from any place William Bradgate appoints not above 17 miles distant from Helidon to any place in Helidon or any place not more than 2 miles from Helidon one load of coals’ (NHRO A/089; NHRO HOLT 655). These parishes were, perhaps significantly, located in the west of the county, close to Warwickshire. But references to sea coal are also abundant, at least from the early eighteenth century. In 1725, David Eaton typically wrote from that ‘as soon as the roads are better, I shall order the sea coal to be brought in’ (Wake and Webster 1971, 23). Over the following century, as land was being enclosed and hedgerows more widely planted, access to coal steadily improved in the county by the development of transport infrastructure. Acts were passed in 1714 and 1724 to make the Nene navigable from Peterborough to Northampton: navigation was improved as far as in 1730; and in 1761 as far as Northampton; all of which made sea coal more easily available. But road transport also made better, with the turnpiking of major routes like that from Warwick to Northampton (via Southam and ) in 1765. In 1750 the inhabitants of Northampton were paying 30d per hundredweight for coal that cost only 4d per cut at the pithead in Warwickshire; by the 1760s the price had fallen to 14d a bushel, around 21d per hundredweight (Hatley 1980). During the early nineteenth century, as the enclosure of open fields was being completed in the county, there were further improvements in transport. The Grand Junction Canal was completed in 1805, passing within four miles of Northampton: coal could now be obtained, at the canal side, for between 10 ½d and 12 ½d a hundredweight (Hatley 1980). By 1813 Pitt was able to note that, because pit coal was now being brought by canal from Staffordshire and Warwickshire, the demand for firewood from coppiced woodland had declined (Pitt 183, 146). A branch canal was constructed from to Northampton in 1815, making transportation to this key market even easier.

In East Anglia and the Home Counties, in contrast – represented in this study by Norfolk and Hertfordshire – demand for fuel was high and supplies, other than fuel wood, were much more restricted and expensive. Norfolk in the seventeenth century was more populous than either Northamptonshire or Yorkshire and Norwich was still England’s second or third largest city. Two of its largest towns, King’s Lynn and Yarmouth, were located on the coast, the river Yare allowed coal to be brought into Norwich with relative ease, and the Fenland rivers allowed some penetration into the west of the county. But most of the county lay at a significant distance from navigable water. Not surprisingly, although the resources of the local heaths and fens were exploited, the market for firewood remained buoyant, leading to the high numbers of pollards on farmland which we have described. In Hertfordshire the situation was different again. At the start of the period studied the Lea was the only navigable river in the county, limiting coal use; while London lay some 40 kilometres to the south of the county boundary, providing a vast market for fuel and charcoal. Hardly surprising, then, that pollard numbers were high, and that the highest densities of farmland trees, and of pollards, of all the areas studied were in the south of this county.

73

Indeed, in both of these old-enclosed districts not only the pollards growing in hedges, but the hedges themselves, were regarded as a source of fuel. William Marshall noted in the 1780s how, in north-east Norfolk, the ‘old hedges, in general, abound with oak, ash and maple stubs, off which the wood is cut every time the hedge is felled; also with pollards, whose heads are another source of firewood’. The entire supply of wood in the district, he added, ‘may be said, with little latitude, to be from hedge-rows’ (Marshall 1787, I, 96). To an extent hedges were regarded as a fuel source everywhere for, whether they were plashed or layed, or coppiced, routine maintenance of necessity produced an abundance of wood which would be used for firing rather than burnt in situ.6 Dormor has estimated (2002, 279–81) that the fuel requirements of eighteenth century households in Wensleydale and Nidderdale could be met by the wood cut from 14-25 metres of hedge. Thomas Tusser typically contrasted the open-field or ‘champion’ districts of England with the early-enclosed ‘woodland’ areas, in which ‘in every hedge’ there was ‘plenty of fuel and fruit’ (Tusser 1573, 94). Advocates of early enclosures in open-field areas often cited the improvements in fuel supply which would result from the planting of hedges; Walter Blith claimed if all land was enclosed, ‘No man almost in the Nation would be either at want of Firing’ (Blith 1649, 113; Reed 1981).. But it was in areas where other kinds of fuel were in short supply that hedges were most assiduously exploited as a source of firing, and in some cases at least their composition modified accordingly. Arthur Young thus described in the early nineteenth century how in Hertfordshire the need for firewood had ‘induced the farmers to fill the old hedges everywhere with oak, ash, sallow and with all sorts of plants more generally calculated for fuel than fences’ (Young 1804, 49); while Pehr Kalm noted how in the Hertfordshire Chilterns hedges were planted with a mixture of hawthorn and sloe but that in addition the farmers ‘set here and there, either at a certain distance or length from each other, or just as they please, small shoots of willows, , ash, maple, lime, elm, and other leaf-trees’ (Lucas 1892, 108). Seventeenth-century leases, such as one for land in King’s Walden in Hertfordshire, sometimes mention the tenant’s right to take the ‘offal’ – the material cut from the hedge when it was laid – for ‘firebote’ (HALS DE/B513/E1). But for the larger landowner, and the larger tenant farmer, the firing cut from a hedge seems to have been regarded largely as the means of covering the costs of maintaining it, although they too would have used a proportion of the wood, if only for oven firing. Randall Burroughs, an educated Norfolk farmer who kept a detailed diary of his agricultural work in the 1790s, regularly paid his labourers for maintaining the farm hedges, in part, with the ‘firing’ derived from them, usually valuing this at 4d per rod (c.5 metres), a figure which exclude the wood cut from the pollards in the hedge, which he kept for his own use. As he was paying, at the same time, 30 shillings for a cauldron (c.1.5 tons or 1,420 kg) of coal we might estimate, very roughly, that the wood cut from 300 metres of hedge was worth roughly the same as a ton of coal (Wade Martins and Williamson 1996).

6 Many hedges were managed by plashing or laying. A row of living, upright stems, spaced roughly 3–4 feet apart, was left behind when the bulk of the hedge was cut and cleared away. These provided a sturdy framework: other stems, cut and trimmed to a uniform height, were kept as vertical hedge-stakes but the remainder were almost (but not quite) cut through near the base, then bent over to a near-horizontal position and woven between the hedge-stakes. Small twigs called ‘edders’ or ‘ethers’ were woven along the top of the completed hedge to bind the hedge-stakes together. Both upright and horizontal stems remained alive and quickly sprouted new growth, so that the hedge soon became a dense, green wall, impenetrable to stock. Plashing was by no means universal practise. In some districts, including the northern parts of East Anglia, hedges were - at least by the seventeenth century - routinely coppiced, managed like the underwood in enclosed woods, probably because this produced more fuel. 74

It has long been appreciated that older hedges tend to contain more species than those planted in the eighteenth and nineteenth centuries, either because they have gradually acquired them through succession, over the centuries, as Max Hooper argued; or because they were created by managing strips of residual vegetation when land was assarted, in medieval times, from woodland, an argument advanced by Hooper’s colleague Ernest Pollard (Pollard et. al 1974).7 But in fact, as Wendy Johnson suggested many years ago, old hedges were also rich in woody shrubs because they had been planted that way, quoting various early writers who, like Young in 1804, had described the planting of mixed hedges (Johnson 1981, 197-9; Fitzherbert 1533, 53; Norden 1608, 201). Smith, writing in the late seventeenth century, believed that ‘it be and hath been a general custom in England, to have several sorts of wood growing in hedge-rows’ (Smith 1670, 27); Nourse similarly described how it was normal practise for farmers to sow, with the thorn, ‘Acorns, Ash-keys, Crab- quicks and the like’, not as timber, but to form the body of the hedge (Nourse 1699, 62); while Blagrave advocated the planting of hedges with species including elm, ash and oak. A number of studies of the character of hedges surviving in the modern landscape have similarly concluded that mixed-species planting was normal before the eighteenth century (Willmott 1980; Hall 1982, 105; Barnes and Williamson 2005, 73-96). It is true that, by the seventeenth century, writers like John Worlidge and Smith advocated planting with thorn alone, in part because mixed planting made for a weak and gappy hedge (Smith 1670, 28; Worlidge1681, 101). But even where such advice was followed, the material removed during routine maintenance would be made, according to Smith, into ‘Brush faggots good for Brewing, Baking, or to be sold to make or mend other fences’ (Smith 1670, 28). Hedges, irrespective of the pollards growing within them, thus made a major contribution to the fuel requirements of old-enclosed districts, but this was especially the case in areas where other sources of firing were in short supply, as well as fulfilling their primary role, as barriers to livestock.

Just as the presence of London and its insatiable hunger for fuel may have served to increased the density of farmland pollards in Hertfordshire, and other Home Counties, so it may have intensified the pressure on hedgerows. One example of this is that leases in such districts, to a far greater extent than in East Anglia or the Midlands, specifically stipulate the character of hedgerow management, perhaps to prevent over-cropping. They also show that hedges, and the pollards within them, were usually cut at the same time, as a single management operation, accompanied by the digging out of the accompanying ditch. One example, drawn up in 1693 for a farm in Aldenham in south Hertfordshire, typically stipulated that the tenant ‘shall not lopp or cutt or cause to be lopped or cutt any of the pollards growing upon the premises but when the hedges shall be new made and ditches scoured where the sayd pollards do grow’ (HALS D/EAm/E3). Young recorded that hedges in the county were generally plashed every twelve years, the rotation normally suggested by seventeenth- and eighteenth-century writers like Blome, but leases show that intervals of between nine and twelve years were common and in the late eighteenth century copyhold tenants of the manor of Barnet, close to London, were instructed ‘Nor to Plash any hedge or Lop any Pollard, till it has attained at least seven years growth, and then to scour and cleanse the ditch at the same time’ (Young 1804; HALS DE/B983/E1).

7 Interestingly, Pollard’s arguments were anticipated nearly three centuries earlier by Smith, who described how, when the extensive woods had been ‘approved and converted into Pasture and Tillage … the Improvers troubled not themselves to do all anew, of any one kind of stuff, but …. Where they stood convenient for such use, left remaining not only the stubs and roots of trees, but several whole trees…’ (Smith 1670, 27). 75

Another probable effect of high fuel demand from London was the way in which some hedges in the county were managed as wide strips of coppiced woodland, documents from the area making a clear distinction between hedges – managed by plashing – and these ‘hedge rows’. Sir John Parnell asserted in 1769 that he knew:

No part of England more beautiful in its stile than Hertfordshire. Thru'out the oak and Elm hedgerows appear rather the work of Nature than Plantations generally Extending 30 or 40 feet Broad growing Irregularly in these stripes and giving the fields the air of being Reclaim'd from a general tract of woodland (LSE Coll Misc 38/3 f.8).

Some of these features were large enough to be described in terms of the area that they covered. In 1556 there were 22 acres of ‘hedgerows’ on St Lawrence farm in Cheshunt, ranging in area from one to four acres (PRO/TNA 315/391, fol. 144). At nearby Broxbourne in the mid-eighteenth century ‘Whitestubbs Hedge Row’, ‘Pimbridge Hedge Row’ and ‘Palmer’s Hedgerow’ were all incorporated into the same twelve-year coppicing rotation which applied to the woods and wood-pastures on the Broxbournebury estate; while a short distance away, a map of Earl Cowper’s estate at Hertingfordbury in 1704 depicts ‘hedges’ with green lines and ‘hedge rows’ as broader green strips containing trees (HALS DE/Bb/E26; HALS D/EP/P4). While ’hedge rows’ of this kind were most frequent on the poor London clay and pebble gravels in the south of Hertfordshire, as they were in the adjacent areas of Essex (numerous examples are shown on surveys of West Horndon Hall of 1598 and Walthambury of 1643: Hunter 1999, 137-41)), but they could be found elsewhere, especially on the Chiltern dipslope. In 1543–4, Henry VIII’s woodward in Hertfordshire recorded the sale of 8 acres of underwood ‘from divers hedges in the parish of Flampsted called Hedgerowys’ worth £10 13s 4d (PRO/TNA E315/457). An early eighteenth-century plan of an estate in Great Gaddesden shows broad hedgerows bordered by trees extending around three fields, with narrower boundaries around the others; while a plan of Herons Manor Farm in Wheathampstead, drawn up in 1768, similarly distinguishes these different kinds of field boundary (HALS 15594; HALSA DE/V/P2).

Fuel economies thus played a significant part in shaping the rural landscape of early modern England, and especially in generating variations in the proportion of trees in each district managed as pollards, and thus in the overall numbers of farmland trees. It is nevertheless likely that other factors had a role in shaping the management of farmland trees. Some of the material cut from oak pollards, for example, was employed as minor elements in timber-framed buildings, and whereas in both Hertfordshire and Norfolk an absences of good building stone ensured that timber-framed houses continued to be built in some numbers into the eighteenth century, in much of Northamptonshire and Yorkshire stone building had become the norm at vernacular level by the early seventeenth. The extent to which land was owned in large continuous estates, rather than as freehold farms; and in particular the amount of control exercised by landlords over tenants; were also crucial factors. Timber trees were a hindrance to the tenant, pollards an asset. In Pearce’s words, ‘It is in the farmer’s interest, to make every tree a pollard’ (Pearce 1794, 56).

Conclusion

Certain aspects of farmland tree populations in the period before the early nineteenth century appear to have been shared by most if not all , most notably the early age at which timber trees were usually felled. There were, we have argued, a number of straightforward,

76 practical and economic reasons for this: an aversion to unnecessary sawing, so that trees were taken down at the size of the timber required; a realisation that growth rates slow in middle age; the impact that very large trees might have on the growth of crops in adjacent fields; and the fact that young oaks produced bark of better quality, and which was more easily peeled, than older trees. Others aspects of pre-industrial tree populations, however, displayed more variation, from district to district: especially the density of trees present in hedges and pastures, and the percentage of these which were managed as pollards. To a significant extent, these two aspects appear to have been related, in that where trees were numerous the proportion managed by pollarding was generally high. Variations in the overall density of trees were largely accounted for by the variations in the numbers of pollards, the density of timber trees in most enclosed districts being broadly comparable, at between five and seven per hectare, more or less regardless of average field size. Given that a high density of farmland trees inhibited the productivity of both pasture and arable land, as many seventeenth- and eighteenth-century commentators emphasised, it appears that it was only tolerated where the market for wood was especially lucrative. Most early writers suggest that the main purpose of pollards was to produce fuel and, as we have argued, the density of farmland pollards in the period before the middle decades of the nineteenth century appears to have been inversely related – both spatially and chronologically – to the availability of alternative sources of fuel, such as peat and, in particular, coal. Where the numbers of pollards were high, moreover, the hedges they grew in also often constituted an important source of firing. All this said, the poles cut from pollards were also used in other ways, especially in the construction of houses and outbuildings (as minor elements of frames, etc); and it may be significant that the two areas studied where densities were low - Yorkshire, and to a lesser extent Northamptonshire – both had access to building stone in a way that the other counties examined, Norfolk and Hertfordshire, did not. Either way, the numbers of hedgerow trees, and their management, were the outcome of economic and, to a lesser extent, social and tenurial factors. They had little to do with the ‘natural’ world.

77

3. Woodland and Wood-Pasture

Coppiced woods in Norfolk and Hertfordshire

In terms of their early-modern woodland history, the counties of Hertfordshire and Norfolk are usefully considered together because, despite some differences in the character of local markets for woodland produce, they exhibit broad similarities in management which were probably more widely shared, by other counties in East Anglia and the south east. Hertfordshire, as we have noted, was largely occupied by heavy clays and acid drift, and largely lay in hedged enclosures by the sixteenth century, with the exception of the strip of ‘champion’ countryside running along its northern boundary. Much of Norfolk – especially the south and east – was likewise occupied by heavy clays and largely enclosed by the sixteenth century. Woodland was relatively abundant in both counties but it is, in the absence of comprehensive map coverage, impossible to estimate its extent before the second half of the eighteenth century, when the first reasonably reliable county maps were surveyed. Drury and Andrews’ map of Hertfordshire, published in 1766, is not in fact particularly accurate in its depiction of individual woods but the overall total for enclosed woodland it suggests, of 10,096 hectares, around 6 percent of the county’s land area, is probably broadly reliable: it is true that on occasions it confuses woods with orchards, nursery grounds or even well-treed countryside, but to a large extent these erroneous inclusions are balanced out by the omission of many small woods (MacNair et al. 2011, 103-110) (Figure 7).

For Norfolk, the county map published by William Faden in 1797 is more accurate, and shows that woodland accounted for around 2.4 percent of the land area, but comes at a relatively late date – after at least five decades during which large landowners, adopting a fashionable interest in forestry, had planted many new areas of woodland in and around their parks, and sometimes more widely on their estates (Figure 18). When these are identified and removed from consideration, the remainder amounts to around two thirds of the woodland shown on the map – around c.1.6 percent of the total acreage of the county. It is not possible to estimate, except in the broadest terms, how these totals had fluctuated in the period from the mid sixteenth to the mid/late eighteenth century. There are signs that rising agricultural prices in the sixteenth and early seventeenth centuries led to the removal or truncation of some woods, especially in Norfolk, but the areas involved are generally small. In the parishes of Great and Little in mid Norfolk, for example, surveys made in 1603 and 1604 refer to a number of small land parcels, generally covering less than a hectare or two, as ‘late converted from wood to pasture’, while a lease of 1637 describes the 3.25 acres of Annyells Grove as ‘being of late woodeground and stubbed up’ (NRO MS 13167 40A7; NRO MS 13279 40C2; NRO MS 13159 40A6; NRO MS 13326 40C2). Some larger areas of woodland were certainly lost or truncated in the early modern period: in nearby Whissonsett an indenture dated 25th August 1655 refers to ‘…a close adjoining [Whissonsett Hal] containing 20 acres of pasture called the Home Wood…and close called Felled Wood containing 40 acres of pasture’ (Carthew 1879, 267). But overall, when the area of woodland shown on seventeenth-century maps is compared with that shown on maps from the later eighteenth century, reductions are modest due to the fact that losses were frequently balanced to some extent by new planting. In the parishes of Morley St Botolph and Morley St Peter, for example, the area of woodland shown on Thomas Waterman’s map of 1629 totalled some 35 acres (14 hectares) (NRO PD3/108). Some ten acres (4 hectares of this had been grubbed out by the early nineteenth century but a further 4 acres (1.6 hectares) had been planted,

Figure 18. Enclosed woodland shown on William Faden’s survey of Norfolk, published in 1797. suggesting a net loss of perhaps 17 percent (PD 3/108 and 109). Sometimes, it is true, reductions were greater. At Shotesham in south Norfolk, for example, the loss of the northern extension to Little Wood, Palmers Grove and anther small area of woodland, a total of 9.6 hectares, in the later seventeenth or eighteenth century was only partly made up for by the establishment of a new coppiced wood, Ringer’s Grove, covering 2.5 hectares (NRO FEL 1076; NRO BR139/1). Overall, however, it is unlikely that the area of enclosed woodland in the county declined by more than 20 percent between c.1550, and the time that Faden’s map was surveyed in the 1790s, given the scale of new planting in this period. In Hertfordshire there are similar examples of piecemeal loss: in the parish of Bushey, for example, the marked cluster of woods in the centre of the parish (Bourn Hall Grove, Knights Grove and Curburds Grove) shown on an undated early eighteenth-century survey had entirely disappeared by 1799 and also fail to appear on Drury and Andrews map (HALS 21579). Nevertheless, here too losses in one area were often offset, to an extent at least, by new plantings elsewhere. Almost certainly there was a net reduction in the area of coppice in the two centuries leading up to c.1750, but its scale was probably small. At a reasonable guess, in 1550 no more than 7 percent or Hertfordshire, and 2 or 3 percent of Norfolk, had been occupied by enclosed, managed woodland.

It is important to emphasise that even in c.1780 almost all enclosed, private woodland in the two counties was managed as coppice-with-standards. By this time, it is true, many writers on forestry were advocating the establishment of plantations, without a coppiced understorey. But an earlier generation, including Moses Cook or John Evelyn, appear to assume that coppice-with-standards was the normal form of both existing and newly-created woodland. Post-medieval documents cast

79 much light on the management of such woodland within the two counties. Small areas of woodland might be cut at one go but the larger woods were divided into ‘fells’, usually long-established and with particular names, which were cut in succession. A map of 1815 shows, for example, how Wood in Norfolk was divided into areas called Foxley Corner Fell, The Great Fell, Alder Carr Fell, First, Second, Third, and Fourth Long Fells, Narrow Fell, Woodhouse Fell and Hazle Hill Fell (NRO NRS 4087) (Figure 19). It is often suggested that the average length of coppice rotation increased from medieval through post-medieval times, from around seven years in the thirteenth century to around

Figure 19. Map of , Norfolk, surveyed in 1815, showing its division into named ‘Fells’. fourteen by the nineteenth (Rackham 1976, 64-6, 82-3; Rackham 1986a, 85, 92). But the evidence for this is not, in reality, very clear cut. A survey of Beeston Priory in Norfolk from the fifteenth century thus described how one 23-acre wood was divided up into six compartments: five acres were under 7 years growth, and not valued, but there were ‘3 acres of 7 years growth, 3 acres of 8 years growth, 2 acres of 9 years growth 2 acres of 10 years growth and 8 acres of 14 years growth’ (Jessop 1887). Conversely, at Gillingham in the same county in 1717 the underwood was cut every seven years, as it was at as late as the 1830s (TNA IR 18/6019). When William Cecil leased his woods in Hoddesdon in Hertfordshire in 1595, 46 or 47 acres were to be felled out of a total of 473½ acres each year, according to the terms of the lease, implying a ten year rotation. When these and other woods were surveyed in 1785 the dates at which the underwood was last cut are given for 25 woods divided into 64 fells, of which three had been last cut ten years before and three eleven years

80 before, but none earlier than this, again suggesting little change in rotation length (Austin 2013, 11- 8).

Post-medieval sources in fact suggest a range of rotation periods, presumably related to the species composition of the underwood, the rapidity of its growth and the purpose for which it was being cut. An indenture from 1740 concerning Wood in Norfolk implies a rotation length of 14 years (NRO BER 336 291/7), a figure apparently confirmed by a map of 1805 (NRO 21428 Box F) and a lease agreement for 1828 (NRO BER 336); an agreement for Sporle Wood in the same county from 1745 implies nine years (NRO 20888); another, for Ashwellthorpe in the early eighteenth century, indicates ten years (NRO KNY 571, 372X3); one concerning Wood, from 1801, stipulates ten years (NRO MEA 7/4); while a lease of 1835 for in Wendling specifies a seven-year cycle (NRO EVL 650/6). In Hertfordshire, so far as the evidence goes, there was less variation. Walker in 1795 thought that most of the woods in Hertfordshire were cut on a rotation of ten years; Young in 1804 reported that the woods in the south of the county were cut every nine or ten years, on average (Young 1804, 145); at Hadham in 1611 the cycle was said to be 11 years (HALS 9607); in the woods around Hoddesdon, as noted, rotations of ten or eleven years seems to have been practised throughout the post-medieval period (HALS DE/Bb/E26); while at Knebworth a rotation of eleven years is suggested by estate accounts for 1815 (HALS 9607). On average, in both counties rotations throughout the period from the sixteenth to the nineteenth century seem to have been around ten years, with no very obvious trend towards lengthening. Where much longer rotations are recorded it may often have been because management had lapsed for some reason, as at Wain Wood in Hitchin in Hertfordshire, where it was reported in 1544 that ‘The said wood conteyneth 62½ acres whereof waste 4 acres, 12 of 25 years or 26 years growth valued at 40s the acre, 20 of 18 years growth valued at 26s 8d the acre, And 26½ acres of 10 or 11 years growth valued at 13s 4d the acre which is in the whole £68 6s 8d’ (PRO/TNA E 318/8/297, E 315/458, SC 6/HENVII/ 1238, SC 6/HENVIII/6747, 6764, 6770: Hoyle 1992, 19). Whatever the rotation length, the recommended time for cutting was either from ‘January at the latest, till Mid-March or April; or from Mid-September, till the end of November’ (Austin 1996, 18). A document recording the sale of the ‘crop’ of several woods in Hertingfordbury in Hertfordshire in 1637 stipulated that the coppicing was to be completed by 25 March, and that all the cut wood was to be removed by the end of October (HALS DE/P/T262).

As noted earlier, strong boundaries were essential to protect newly-cut coppices from livestock grazing on adjacent parcels of land, and the lessor here also undertook to ‘make good and stronge hedges ditches and Fences … about the Coppice grounds’ (HALS DE/P/T262). The massive banks which generally enclosed early medieval woods do not appear to have been constructed around new woods established after the sixteenth century, although existing examples were rigorously maintained. Instead, hedges on smaller banks, little different from those forming the boundaries of fields, became normal. Whatever their precise character, woodland boundaries needed to be carefully maintained. In 1545 William Lowyn, custodian of Henry VIII’s woods in the manor of Cheshunt, was paid 20 shillings ‘for kepyng the hedge & beastes & cattell from the spryng in [a wood called] Rough Cattall [and] for the safe inclosure therof’, while Thomas Browne received 13s 4d for ‘the kepyng of the hedges & savegarde of Dyvers Coppices’ in the king’s manor of Bedwell in nearby Essendon (PRO/TNA E315/458). Poorly maintained boundaries, especially when combined with difficult neighbours, could wreck the coppice. The steward responsible for the woods at Castle Rising in Norfolk reported in 1786 how Mr Beck, ‘a very unpleasant tenant’:

81

Will not keep his cattle out of the wood, though I requested him in the most civil terms to do it. I have now ordered the woodman and Browne to impound the cattle if they are found there, otherwise all the underwood and young wood must be destroyed (NRO HOW 757/35). The exclusion of grazing stock was not always total, however. Sometimes cattle were allowed into woodland once coppice re-growth had proceeded for several years, as at Tring in the second half of the sixteenth century, where they were kept out for seven years after coppicing (PRO/TNA E134/43Eliz/East3). But they were carefully controlled, and the need to prevent damage from horses used in forestry operations is often mentioned in leases, one for a Norfolk wood stressing the importance of ‘preserving the shoots and slopps of such wood from being bitt by the horses fetching the same’, while others even refer to the need to keep the horses muzzled.

In the early Middle Ages woods were, like other demesne assets, usually kept ‘in hand’ by manorial lords, and this continued to be the case on many estates right through the post-medieval period. At Wheathampstead in Hertfordshire in 1539, for example, the sale of 115 loads of hardwood (worth 10d per load), 113 loads of bavins (small faggots) (at 7d per load), and oak, beech and ash timber from the woods called Marchalles and Butler brought the owner a gross income of £16 16s 10d, although the costs of felling and dressing the wood and timber, of hedging the wood and for making and hanging a new gate came to 64s 2d, leaving a profit of £13 12s 8d (HALS DE/Lw/E33). On many estates, however, direct management declined from the fifteenth century. Often, as at Hertingfordbury in 1637, the underwood would be sold unfelled to a single purchaser, who would then bear the costs of coppicing. But it was also commonly sold in small portions to numerous different purchasers, each of whom was responsible for felling his own particular allocation. An enquiry into Wayland Wood, near Watton in Norfolk, made in 1594 typically asked whether the wood was sold ‘by the acre or otherwise’ (NRO WLS IV/6/15): portions as small as a rood, or even half a rood, are also mentioned in documents, especially from Hertfordshire. This method of management is recorded at Weston and Knebworth in the fourteenth century; at Walkern and Broxbourne in the seventeenth century; and at Wain Wood near Hitchin, on the Lamer estate in Wheathampstead, and on the Broxbournebury estate in the eighteenth century (PRO/TNA SC6/873/21 and SC6/873/25; HALS K100; HALS 9607; HALS B/86; HALS DE/R/E114/1–10; HALS 27252; HALS DE/Bb/E26). The practice continued into the early nineteenth century in the woods around Hertford on the Marden and Panshanger estates, and also at Bramfield, where the diarist John Carrington regularly recorded wood sales during the spring months, followed by ‘wood feasts’ held towards the end of the year when the purchase money was handed over (Branch Johnson 1973). The feast here was held in October: meat and drink were provided by Lord Cowper, which in 1754 cost him 15s. The Woodward then carried the purchasers’ money to Panshanger, although on one occasion the accountant had to ‘Deduct from the receipts the money which Hale the Woodward was robbed in the night between the 7th and 8th Oct 1813 in returning with it from the Digswell wood feast at Eleston’s at digswell water where he had received it from the Purchasers: £130’ (Austin 2008). Records of a ‘wood feast’ on the Knebworth estate in 1840 suggest that this way of selling the underwood continued at least into the middle of the century (HALS DE/K/E15).

Sometimes the price paid for the different portions of the underwood sold in this way varied significantly, as at Walkern in Hertfordshire in 1612, where the price of the parcels ranged from 22s 6d to 50s per rood, and from 9s 9d to 27s 6d per half rood, presumably reflecting the uneven quality of the underwood. Here the purchasers were all from the local area: fifteen from Much (Great)

82

Munden, twelve from Benington, seven from Little Munden, four each from Walkern, Ardeley and Aspenden and one each from Standon and Westmill (HALS 9607). In some woods, however, the precise parcel of underwood to be purchased was determined by drawing lots; the price per lot was fixed, irrespective of variations in the quality of the underwood being sold. In March 1800 John Carrington recorded in his diary ‘I Drawed No. 1. in Hooks bushes’, a wood on the estate and adjacent to his farm between Tewin and Bramfield; in March 1803 he ‘Drawd No. 1 a Rood …’ also on the Marden Hill estate (Branch Johnson 1973, 46, 84 and 98).

Selling the underwood to a range of individuals in this way could damage a wood’s productivity because the buyers, many with little concern for the next crop, tended to spread the felling and dressing of the wood over a prolonged period of time, thereby hindering the regrowth of the ‘spring’. They might also damage the coppice stools by cutting them too low, or even by digging out the roots altogether, and they might be tempted to fell any young trees on their allotted parcel which should have been preserved for future timber. On the estates of Lord Burghley at Hoddesdon in Hertfordshire in the late sixteenth century it was reported that the 473 acres of underwood had been sold, unfelled, to local people for many years in one-acre parcels, each costing £3. In 1595 Burghley decided to lease the woods to just two men, John Thorowgood of Amwell and William Keeling from London, for the same price per acre, thereby maintaining his income but at the same time reducing the complexities of administration. The productivity of the woods varied from acre to acre, but the leases believed that on average they could clear a profit of £4 5s per year from the 46 acres they were allowed to cut, after deducting rent, tax, and the costs of felling and making faggots, fencing, and wages, as well as those of the liveries and board required for the two men overseeing operations. Inevitably, the cost to local people of purchasing material cut from the wood – faggots at 4s 4d per load and roundwood at 7s per load – was significantly higher than that of buying an acre of underwood uncut, and processing it themselves, and they therefore petitioned Lord Burghley for redress. In response, the lessees agreed to set aside sixty loads of faggots for sale to the local poor at a penny a faggot, but they were adamant that they were not making excessive profits and that the new arrangement would significantly improve the productivity of the woods (Austin 1996).

In other cases one or more individuals rented the entire wood for a term of years, just as they might rent a house or a farm. Usually the timber was reserved for the landlord, but in some cases the lessee contracted to remove a proportion of this as well. The period of the lease might be as short as two years, as at in Norfolk in 1613 (NRO MR 211, 241x 6); as long as 21 years, as in the case of woods at Fransham in the same county in the 1540s and 50s (NRO MS 13253 40B4); or even for 31 years, as at Bressingham in the same county in 1682 (NRO BRA 301/1). The terms of such agreements laid down the frequency of coppicing and the number of timber trees to be left at the end of the lease. That drawn up for the woods at Gressenhall describes, for example, how the lessee could fell two hundred and forty timber oaks, and all the underwood, except ‘the old storrells and standells of young oak and ash which heretofore have been maintained and preserved as former felles of the said underwood’. In addition, twelve new standells or storrels of young oak or ash were to be marked and retained per acre ‘in fit and convenient places to stand still and remain for new storrels or standells on and beside the old and former storrels and standells’. Routes for carrying and carting out the material were to be agreed, and permission was given to ‘dig and make....so many pits called saw pits as shall be .... convenient for the sawing, cutting, contriving and converting of the said timber trees’. A further clause forbade the pasturing or feeding within the wood of any of the horses that were to be used to extract the timber or underwood. In addition, the purchasers

83 were, before operations commenced, to ‘make a sufficient hedge and fence between the wood’ and the adjacent fields (NRO MR 211, 241x 6) – a perennial concern, as we have noted. A lease for Banyards Wood in Bunwell from 1675 (WLS LXIX 25) insisted that ‘the springs thereof are preserved and kepte as they owght’, while at Wayland in 1674 the lessees were allowed £3 towards the costs of fencing. Many leases, like that for Gressenhall, refer to making saw pits yet in Norfolk, certainly, relatively few woods, including Foxley and Hook Wood in Morley, contain unequivocal extant examples. It is possible that a clause in an agreement concerning the felling of timber in woods in Redenhall, drawn up in 1737, explains this, for it allowed the digging of ‘saw pitts on convenient parts of the premises if occasion be …provided they fill up and level the same within the time aforesaid’ (NRO MC 600/L1-6, 780x9). Post-medieval woods in both counties were thus very intensively managed – perhaps not surprisingly, given the relatively small proportion of the land area, even in well-wooded Hertfordshire, devoted to them. Coppices produced firewood but, so far as the evidence goes, much of the cut material was used for specialised purposes, and it is noteworthy that Thomas Hale, who probably came from Norfolk, listed the benefits that coppice brought to the farmer in terms of building repairs and the making of ‘implements’ before those which accrued to ‘his Chimney’ (Hale 1756, 137). Much of the cut material was used to make hurdles, for which hazel and ash were particularly well suited. Hurdles were employed in vast numbers for folding sheep on the arable, especially on light, leached land, of which there were large areas, generally farmed as open fields, in west Norfolk and north Hertfordshire (Allison 1957). John Skayman described in the early sixteenth century how the woods on the Raynham estate in west Norfolk were used to make ‘fencing’ (Moreton and Rutledge (eds) 1997, 115); while in 1730 Edmund Rolfe, the lessee of a large estate in Sedgeford, a short distance away, asked his landlord for permission to plant 40 acres of the worst land on the estate with underwood of sallow, hazel and willow, arguing that such wood was ‘being continually wanted by the occupiers of the said estate for hurdles for sheep of which there are great flocks’ (NRO DCN 59/30/12). Walker in 1795 implies that much of the material from the woods in Hertfordshire was likewise used to make ‘sheep flakes’ or hurdles, while Young in 1804 noted the use of ‘sallow and willow’ cut from the coppices in the south of the county for this purpose, although much of the other wood was made into faggots (Young 1804, 145). Both ash and hazel were also, however, used in hoop-making, another craft mentioned on a number of occasions in relation to woods in both counties; and ash was widely employed for such things as tool handles. Coppice wood was also employed in the construction of timber framed buildings – poles of oak, and sometimes ash and elm, were used as minor elements of the frame, while hazel was a major element in wattle-and- daub infill. In addition, smaller lengths of hazel (and sometimes ash) were used to make ‘broaches’ used in thatching roofs and corn ricks, as was reported at Castle Rising in 1791 (NRO HOW 757/9). In 1796 Nathaniel Kent described how the underwood from coppiced woods in Norfolk was used for sheep hurdles, thatching, hoops and general repairs (Kent 1796, 86).

Although specialised uses thus loom large in our sources, many of the larger poles were also sold as fuelwood, while smaller material, tied together as faggots, was largely used as firing. It produced a short hot blaze, and was thus particularly suitable for household baking and brewing, as well as for a variety of industrial processes such as brick- and tile-making. Faggots from coppiced woods were sold on a large scale throughout the post-medieval period. They appear, together with larger pieces of wood, in many of the probate inventories produced in the sixteenth and seventeenth centuries: in John Johnson’s kitchen in Hertford in 1673, for example, there were fifteen loads of ‘Faggotts, Blocks and Roundwood’ valued at £11 5s; in the woodhouse of William Turner, a gentleman of the same 84 town, in 1683 were three stacks of wood and half a load of faggots worth £2 9s; while in the yard of John Bach, an Alderman and gentleman of Hertford, in 1699 there were ‘one thousand Jack Faggotts, Six Load Roundwood and blocks’ (Adams 1997, 77, 124 and 41).

Bach’s yard also contained twelve bushels of ‘coal’, probably charcoal rather than coal in the modern sense, and in many woods in both counties a proportion of the wood cut from coppices was converted in situ into charcoal. The hornbeam which constituted the principal underwood species in many woods in Hertfordshire and Norfolk was particularly suitable for this purpose, as was the alder which grew in stands in wetter woods, not only on valley floors but in places on the boulder clays. Charcoal, lighter and with a higher calorific value than firewood, could be transported economically over longer distances, and especially to the fuel-hungry cities of Norwich or London. An enquiry into the management of Wayland wood near Watton in Norfolk, made in 1594, records how the underwood was used for hop-poles, poles, hurdles - and charcoal (NRO WLS/IV/6/14); while the lease for the woods at Gressenhall in the same county, drawn up in 1613 allowed the lessees to ‘to make and sett within the said wood so many harthes called coal harthes’ as they needed to convert as much of the underwood as they wished into charcoal (NRO MR 211, 241x 6). In Hertfordshire numerous place-names, such as Colliers End in the parish of Standon, Coles Green in Walkern, Cole Green in Hertingfordbury, Coles Heath in Little Hadham and Coles Hill in Northaw, testify to the prevalence of the practice in medieval times, as do fourteenth-century records from the bishop of Ely’s park at Hatfield and fifteenth-century ones from the parks at Knebworth, Bedwell and at Little Hadham (HHA Manor Papers I, 1396, 265; HHA Court Rolls 10/16; HHA Court Rolls 11/4 fo. 22; HALS K116). In 1610 William Brokett, a gentleman in Essendon, left to his daughter Anne ‘all the benefitt of the cole wood in Busshey Lepes, And Alsoe [the] croppe of wood next adioyninge to yt called Dringe hall grove, to be Felled next yeare’. At the start of the seventeenth century Sir Arthur Capel was paying about £12 18s each year to have charcoal made in his woods at Walkern in order to heat his home at Hadham Hall (HALS 9607). In 1611 his woodward was planning to coppice about 8 acres of wood each year on an eleven year cycle and to make 35 loads of charcoal a year from the cut wood. The costs included felling and cutting the wood, peeling off the bark, making faggots of the brushwood, making the charcoal, carrying water to quench the ‘hearth’, and hedging and ditching the wood.

Although coppices produced much of the income from managed woods, the timber, ‘tops’ and bark from standard trees were also of considerable economic importance and there are few known examples, in either county, of woods comprising only coppiced underwood. The density of standard trees appears to have displayed much variation, however, as particular owners made choices about the relative importance of underwood and timber: an increase in the density of the latter caused shading which reduced the vitality of the former. An act of 1543 stipulated that twelve standils or stores of oak, or other timber, should be left per acre after felling. While lease agreements for Gressenhall Wood in Norfolk in 1613, and for Wayland Wood in the same county in 1674, do indeed insist that this number be left at the end of the lease, in addition to those already growing there (NRO MR 211, 241x 6; NRO WLS LXIX 25), only eight per acre were to be left at Ellingham in 1682 (NRO BRA 301/1). Conversely, some leases stipulate greater quantities – as much as twenty an acre at Honeypot Wood in Wendling in Norfolk in 1835 (NRO EVL 650/6). As none of these agreements gives us any clear idea of how many standard trees already existed in the woods in question it is impossible to ascertain the overall density of standards per acre, but other sources make it clear that this varied greatly. At Sporle in mid Norfolk the 1750s there were as many as 60 standards per acre;

85 but a parliamentary survey of Toft Wood in nearby East , made in 1649, recorded 2,860 timber trees on 143 acres, amounting to only 20 per acre; while a survey of the Duchy of Lancaster estates, carried out in 1609, records a total of 734 trees in 70 acres of woodland in Tunstead, around ten per acre, and in nearby Southwood less than ten (TNA PRO E178/4988 82482). At Honeypot Wood in Wendling in 1777, in contrast, there were no less than 924 trees, all oaks, a density of 92 per hectare – together with 57 pollards, presumably growing on the boundary bank (NRO HNR. 135/3/18).

Oak bark was, as with farmland trees, an important product, stripped from standards or (less commonly in these counties) from coppice poles. Henry Gape paid 20s for bark from oaks felled in Wheathampstead in Hertfordshire in 1539 (HALS DE/Lw/E33): Gape was a prominent industrialist who later became mayor of St Albans, leased properties in St Michaels from St Albans Abbey in 1538, and probably established the tannery excavated in the grounds of St Michaels manor in 1976 (Saunders 1977, 9–12). In the late eighteenth century bark from Earl Cowper’s Digswell estate in Hertfordshire was sold to Thomas Crawley of Welwyn and Daniel Green of Buntingford, both tanners (Austin 2001, 9); while bark from Wain Wood near Ippollitts in the same county was sold, in the years between 1755 and 1767, to George Mugliston, a tanner in Hitchin, in quantities ranging from 26 to 111 yards a year, the price rising steadily from 6d a yard in 1745 to 14d a yard in 1767 (HALS DE/R/E114/1–10).

The overall impression is that, even in the late eighteenth century, coppiced woods were still economically important, especially on poorer soils and where they lay close to major urban centres. A valuation of Lord Monson’s estate around Cheshunt and Broxbourne in south Hertfordshire in 1785 described how the sale of coppice wood – still cut and purchased at this time in numerous small stands – brought in £50 per annum, around 200 acres in all, comparable to the rents for agricultural land in the district, although it should be noted that this figure also includes the value of wood cut from some of the pollards on the estate. Woods were very intensively managed: as in the middle ages, they were, in essence, factories for producing wood and timber, far removed from being in any meaningful sense ‘natural’ environments.

Wood-pastures in Norfolk and Hertfordshire

We noted earlier how, even in well-wooded Hertfordshire, enclosed woodland probably covered around 7 percent of the land area in the post-medieval period, and considerably less than this in Norfolk. These figures are, however, slightly misleading as there were also extensive tracts of grazed woodland – wood-pastures of various kinds - especially in the former county, and these remained locally important until the end of the eighteenth century. Some were to be found on common land, but others were private, and of these the most important were deer parks.

In the Middle Ages most deer parks had been located at a distance from elite residences and had functioned as detached hunting grounds and venison farms, administered from a ‘lodge’ which provided accommodation for the parker responsible for looking after the park, and temporary accommodation for the owner and guests on hunting trips. Parks were invariably bounded by a substantial ‘pale’, or deer-proof fence. Some parks were large, covering a much greater area of land than most coppiced woods. Whereas only four surviving ancient woods in Norfolk, for example, extend over an area of more than 0.75 square kilometres, at least 25 early deer parks did so - the real number may well have been higher, given the problems involved in identifying the likely

86 boundaries of many examples. What is even more striking is the fact that all four of the ancient woods concerned – Foxley, Hockering, Haveringland Great Wood and Horsford – appear themselves to have developed from deer parks in the later Middle Ages, planted with coppice – or filling naturally with underwood – once deer were removed from them. Some medieval parks had been densely wooded: that at Tibenham was described in 1309 as ‘a wood called ‘le Park’, containing in circuit a league and a half’ (Cal IPM vol, 4, 259). But most included a significant proportion of open ground, ‘laundes’, in order to allow space for hunting. Some also contained embanked coppices from which the deer were excluded during the early stages of the rotation growth. In 1391 John Lowyk was granted:

The underwood in the park of Foxle[y], co. Norfolk … on condition that of the said underwood sufficient cover be reserved for the king’s deer within that park, that he suitably enclose at his own expense from time to time that coppice whereof he takes the underwood, and that the underwood is taken in season (CPR 1388-92, 486).

There was a tendency, in both counties, for large, detached parks of this kind to be abandoned in the course of the fifteenth century. Increasingly, parks were placed next to elite mansions, and at the same time they became less densely wooded and, perhaps, more carefully designed in character. Many long-established examples, reclaimed directly from residual areas of ‘waste’ in the twelfth or thirteenth centuries, were also destroyed in the course of the sixteenth and seventeenth centuries due to changes in the wealth of individual families, the demise of many great mansions and the creation of new ones. All of this ensured that there was relatively little continuity between medieval deer parks, and the landscape parks which existed in some numbers in both counties by the end of the eighteenth century. In Hertfordshire probably only around 10 percent of landscape parks had origins as medieval parks, and in Norfolk only 2 percent: even fewer as early medieval parks, enclosed from ancient wood-pastures.

The demise of traditional deer parks was gradual, however: some parks on the old model survived well into the seventeenth century. A map of Lopham Park in south Norfolk, made by Thomas Waterman in 1612 (Arundel Castle Archives P5/1), shows a detached park with a lodge near the centre of the park (its site now marked by Lodge Farm), surrounded by an extensive open laund, sparsely scattered with trees, which was in turn flanked by areas of wood-pasture. There were three large blocks of apparently enclosed and coppiced woodland towards the northern periphery of the park - North Haugh, Hither Haugh and Elme - while other areas of woodland are shown in the south east, named as Little Chimbroke and Poule Chimbroke. Nor should we exaggerate the extent to which the new ‘residential’ parks were necessarily more open and more ornamental landscapes then their predecessors. They continued, into the eighteenth century, to include within their area large tracts of wood-pasture. Blickling Park in north east Norfolk was established in the 1560s and expanded in the 1740s; in 1756 a survey valued the mature trees in the park at no less than £2, 780, a huge sum, constituting around a third of the total value of timber on the estate (NRO MC3/252). A supplementary examination added a further £120 worth of timber in the park, including 70 ‘firs’, probably Scots pines; 75 horse chestnuts; 107 beech; 200 ‘small oaks’; 160 oak pollards and 10 lime pollards. Both valuations explicitly ignored the 3,780 ‘Young planted trees in the Park’ with girths of less than six inches, as well as those in ‘near 30 of new Plantations’. They also apparently omitted the majority of pollards, as they certainly did elsewhere on the estate, including in the survey only those with a perceived timber value. Other, more complete surveys certainly suggest that in most sixteenth and seventeenth-century parks pollards were more numerous than timber trees. One for

87

Hatfield Middle Park south Hertfordshire, made in 1669, recorded 1,613 pollards of oak and beech within the park itself and a further 72 growing on the bank of the park pale: but in addition, 255 values are given for hornbeam pollards, each said to represent 40 trees, suggesting that that there were, in all, 10, 133 pollards growing on the park’s 350 acres, a density of around 70 per hectare (HHA, Hatfield House General 47/28). In contrast, only a handful of timber trees – a mere 24 oaks and 33 ashes – were noted by the surveyor. This said, such a dominance was not found everywhere. In the royal parks, in particular, timber was generally more prominent. In 1553 Hertingfordbury park in Hertfordshire contained 547 oaks ‘being pollards and dotards and valued one with another’ at 4s. each, but these were outnumbered by timber trees, for there were 500 oaks valued at 10s. each, 500 at 8s. and 70 ashes each valued at 10s. A similar picture is presented by a survey of 1608, which described the loads of timber and of firewood that the trees were thought to contain. The ‘diverse trees of oak, ash and elm’ growing on the wider demesne lands of the manor contained 11 loads of timber and 284 loads of firewood, but those growing in the park had 464 loads of timber and 3306½ loads of firewood: the non-parkland trees therefore had about 26 loads of firewood to each load of timber, while those of the park had just 7, clearly suggesting a higher density of pollards in the wider countryside than within the park itself (Austin 2013). Such an emphasis on timber was encouraged by the increasing demand for timber for the navy – particularly for the curved timbers obtained from free-growing parkland oaks (standards in coppiced woods, in response to competition with the underwood, tended to grow in an upright form, without low growing branches). The survey of Theobalds Park in south east Hertfordshire made for the revolutionary government in 1650 recorded no less than 15,608 trees ‘marked for the use of the Navie … besides what is already cutt Downe’, which still left wood and timber worth £7,259 in the 2,500-acre park (PRO/TNA E317/Herts/27). At nearby Cheshunt park 577 trees were considered suitable for the navy, leaving a further 713 trees valued at £77 7s (PRO/TNA E317/Herts/19).

As the taste for more manicured parkland developed through the middle and later decades of the eighteenth century, under the influence of Capability Brown and his contemporaries, parks did finally become less densely treed, and pollarding within them declined. Educated opinion turned overwhelmingly against the practice, which came to be seen as unattractive, ‘barbarous’, and something associated with commons and the poor (Petit and Watkins 2003); in part, perhaps, because of the fact that elites were now burning coal. This said, even in the late eighteenth century wood-pastures often occupied some parts of parks, especially in Hertfordshire. The trees growing in the 320-acre park at Broxbournebury Hertfordshire in 1784 for example included 381 oaks, 172 elms, 115 chestnuts, 20 walnut, 8 ash and 3 beech, plus 1,325 young trees (spars) and 991 pollards (HALS DE/Bb/E27).

At the start of the period studied here common wood-pastures were probably more extensive than private deer parks, in both counties, although they generally degenerated during the following two centuries, due to over-grazing and unregulated felling. In Norfolk, many of the numerous heaths which, by the end of the eighteenth century, were characterised by open expanses of gorse and heather had evidently been quite densely treed two centuries earlier. In the late sixteenth century the inhabitants of Marsham in Norfolk accused James Brampton of having, among other misdemeanors, ‘felleth downe woode growinge uppon the common contrarye to the custome of the mannor’ (Smith et al. 1982, 242-3). What were later open heaths are shown as at least partly wooded on a number of early maps from the same county, including surveys of New Buckenham (1597) (NRO MC 22/11); Haveringland (1600) (NRO MS4521); Castle Rising (early sixteenth century)

88

(NRO BL71); and Appleton (1596); NRO BRA 2524/6). Many of the smaller greens and commons on the heavy clay soils in the south of the county, ringed by farms and cottages, are also shown as grazed woodland on early maps, such as Waterman’s survey of Gressenhall of 1624, where some of the commons have names like ‘Horsefrith’, incorporating Old English terms for woodland (NRO Hayes and Storr 72) (Figure 20).

Figure 20. Detail from Thomas Waterman’s map of Gressenhall, Norfolk, surveyed in 1624, showing a partly wooded common grazed by a variety of livestock.

In Hertfordshire, likewise, there is much evidence for the attrition of common woods in the course of the sixteenth, seventeenth and early eighteenth centuries. Here the largest areas were to be found, in the manner already described, on the Chiltern ridge in the north west of the county, and on the strip of high ground, characterised by particularly poor soils, which straddles the county boundary with Middlesex in the south. ‘Divers parcels of wood’ were felled and sold for £160 from the 200- acre common ‘woodgrounde’ called Mayne Wood in the parishes of Tring and Wigginton in 1584, but by the beginning of the seventeenth century the common was being described as the ‘waste or pasture called ‘The Mayne Wood’ (PRO/TNA E 134/4Jas1/Mich19). By 1650, the wood and trees growing on all the extensive waste of the manor of Tring were valued at just £10, and the customary tenants were entitled to gather, not wood, but small shrubs and bushes from within what was still described as West Wood (PRO/TNA E317/Herts/29). The unauthorised felling of substantial amounts of trees from the 300-acre ‘great wood’, another common wood, are recorded in the court rolls of the manor of Caddington in the 1580s and ‘90s, a process which resulted in the complete clearance of trees from the area by the following century (PRO/TNA E 134/26Eliz/Hil7 and E 134/MISC/2415; Hindle 1998, 48-9). The situation was similar in some places in the south of the county. In the late 89 seventeenth century it was said that Barnet Common ‘was formerly a wood, but had been of recent years laid waste, and used as a common’ (Page 1908, 329– 37). Steady degeneration was encouraged by the fact that, on some of these commons, there was little control on the numbers of livestock which tenants could pasture. The inhabitants of Northaw, Cheshunt, North Mymms and the manor of Brookmans (in North Mymms parish) thus all claimed common of pasture ‘without number’ for all their cattle and beasts, at all times of the year, in the three great wastes of Northaw, Cheshunt and North Mymms; tenants of Enfield in Middlesex also claimed pasture rights in Northaw Common (PRO/TNA STAC 3/1/49; DL43/7/4). In both the Chilterns, and on the high ground straddling the county’s southern boundary, much common woodland was thus gradually replaced by more heathy land, with an abundance of gorse, broom, heather and bracken. By the mid- seventeenth century the southern end of Berkhamsted Frith, wooded in the sixteenth century, had lost its tree cover and was marked on a map of 1638 as ‘Barkhamsteed Heathe with out trees’ (HALS 1985).

We should not, however, exaggerate the speed with which, in either Hertfordshire or Norfolk, common wood-pastures degenerated to open heaths or treeless greens. Close examination indicates that a number of residual fragments of common wood-pasture are still shown on William Faden’s 1797 map of Norfolk, within areas specifically described as ‘heath’, in the form of scattered tree symbols (as on Walsham Heath, Hevingham Heath, Necton Heath, Stock Heath, Edgefield Heath and Cawston Heath); while many of the smaller commons and greens on the heavier soils are also shown as still partly tree-covered, including Stipens Green near Hockering, Podmore Green in Scarning, Hoe Common, Shelton Common, Pulham Common and Fritton Common. As late as 1817 the ‘timber trees and stands’ on Shottesham common were valued at no less than £945.5s (NRO HNR 149/3). Indeed, the remains of heathland wood-pastures still survive in a number of places in Norfolk, in the form of collections of ancient pollards now usually embedded in later plantations, most notably at Bayfield in the north of the county, where more than 70 ancient, pollarded oaks – some pedunculate Q. robur, some sessile Q. petraea - survive on dry, leached soils of the Newport 4 and Barrow Associations; at nearby Pereer’s Hills and in Sand Hill Plantation; at Ken Hill near Snettisham in the west of the county; and at Broom Covert in (TM 025865) and Mere (TL950963), on the edge of Breckland. Traces of former wood-pastures also survive on some of the few clayland commons that escaped enclosure in the early nineteenth century, including Fritton and Old Buckenham. Some of the trees at these places appear to have been pollarded into the twentieth century.

In Hertfordshire some of the commons on the Chiltern ridge are likewise still shown with extensive tree cover on Dury and Andrew’s county map of 1766, as are many of those on the high ground in the south of the county, especially in Elstree, Nothaw and Cheshunt (Figure 21). A few years earlier, in 1748, the Finnish visitor Pehr Kalm described how the great tract of undulating common land which existed ‘between Cheshunt and Bell Bar’ to the south of Hatfield:

Was covered with tufts of ling, between which bracken flourished and swamps abounded. But there was scarcely any grass. Sheep grazed here. In places Carpinus [hornbeam] grew fairly densely to a height of six feet, and the tops of it were cut for fuel (Lucas 1892).

The parliamentary survey of Theobalds, made in 1650, describes the trees growing on a portion of the park which had recently been enclosed from Cheshunt Common by James I: the 149 acres contained ‘4928 lopt pollards and hollow dotrills and horne beame’, an average of 33 an acre, or 82

90 per hectare (PRO/TNA E317/Herts/24, p. 37). As late as 1695 the remaining section of the common, covering 1,186 acres, reportedly contained 24,000 hornbeam pollards, around 20 trees per acre or nearly 50 per hectare (HALS 10996 A/B). An average of around 3,800 of these were lopped each year over the twenty-year period from 1658 to 1678, suggesting a short pollarding cycle of around six or seven years (HALS B/86). On Broxbourne Common, a short distance away, which covered only around 60 acres, there were at least 4,420 pollards in 1682, an apparent density of 73 trees per acre (c.180 per hectare) (HALS B/86).8

Figure 21. Commons in Hertfordshire in the mid-eighteenth century: based on Dury and Andrews’ county map of 1766, with additions from other contemporary maps. (a), commons carrying significant amount of tree cover; (b), largely open commons.

The survival into the late seventeenth and eighteenth centuries of large numbers of pollards on some areas of common land, in both counties, is superficially surprising. Writers on historical ecology often emphasise that wood-pastures were, and are, inherently unstable environments. Their trees were vulnerable to damage from stock, through the stripping of bark for example, or the compaction of the ground above their root systems. When trees were felled, died or were blown down by the wind, moreover, it was difficult to replace them because of the intensity of grazing. Even in private parks, especially those densely stocked with deer, such pressures would have led to a gradual loss of tree cover, but in common wood-pastures the problem would have been acute, for it was difficult to

8 This figure assumes that the extent of Broxbourne common was the same as that shown on a map of the parish of Broxbourne of 1829 and on the tithe map of 1839 [HALS 71014; DSA4/25/2]. 91 establish and protect replacement trees because fencing off portions of a common would have conflicted with the rights of other commoners to freely access and exploit it (Rackham 1986aa, 121- 2). Against this, however, we should note that institutional structures often existed which militated against tree loss. In many places customary mechanisms existed which allowed tenants to protect new planting on commons (Dallas 2010). During a legal dispute in the late sixteenth century concerning the commons at Pulham in south Norfolk, for example, it was stated that ‘The tenantes of the said manor have used to make benefitt of the trees growing upon the common near their houses which were planted by themselves and their predecessors’ (NRO NAS II/17). A survey of the manor of Gressenhall, drawn up in 1579, describes how tenants admitted to holdings received one or more ‘planting’ (NRO MR61 241X1); Waterman’s map of 1624 shows that these were wooded areas, growing on the various commons of the parish, each of which associated by name with the owner close to whose home it was located (NRO Hayes and Storr 72). Francis Blomefield in 1739 similarly reported how the tenants of his home parish of Fersfield had:

Liberty to cut down timber on their copyholds, without licence and also to plant and cut down all manner of wood and timber on all the commons and wastes against their own lands, by the name of an outrun (Blomefield 1805, Vol. 1, 739, 95).

Figure 22. Pollarded oaks on the margins of Fritton Common, Norfolk, clearly planted in lines.

Blomefield describes similar customs at other places in the south of the county, including Kenninghall, Diss and (Blomefield 1805, Vol.1, 220, 263). Once again the trees established by the commoners were close to their ‘own lands’, and in this context it is noteworthy that where old pollards survive on Norfolk commons – as at Fritton – they tend to be concentrated towards their margins. The group growing on the western side of this particular example appear to be arranged in straight lines, clearly indicative of deliberate planting (Figure 22).

92

Not surprisingly, enclosure awards often include claims made by commoners for compensation for trees planted on commons, which would be lost following enclosure. When Pulham St. Mary Magdalen in south Norfolk was enclosed, for example, the Reverend Jeremy Day claimed for ‘thirty eight poplar trees growing on the waste contiguous to the lands of the said estate, and planted, reared and protected by a late proprietor thereof’ (NRO PD510/19). Manorial lords, as we have seen, also had an interest in preserving tree cover, normally owning the timber trees on commons and the trunk or bolling of the pollards. Enclosure awards accordingly often contain claims for trees felled at enclosure, or incorporated within parcel allotted to others. At the enclosure of in mid Norfolk in 1807 the Earl of Leicester, Lord of the Manor, claimed the value of:

All trees, and all bushes and thorns planted or set by him or his predecessors, or his or their tenants, upon the said commons and waste grounds, contiguous or near to any of his said messuages or farms, which have been usually lopped, topped, pruned, or cut by him or his predecessors, or his or their tenants (NRO BR90/14/2, p.17).

While Lucas Strudwick, lord of the three other manors in the parish, similarly claimed compensation for:

All trees (not being timber trees), and all bushes and thorns planted or set by him or his predecessors upon the said commons and waste grounds contiguous or near any of his said messuages or cottages, which have been usually lopped, [etc.]

In some areas, most notably south Hertfordshire, manorial lords appear to have taken a particularly active role in the management, and perhaps the preservation, of trees on commons. Manorial rights to wood and timber were so strong in this district that at a number of places, including Broxbourne and Cheshunt, the lops of the pollards, as well as the bollings, belonged to the lord. When Lord Burghley leased his Hoddesdon woodlands to John Thorowgood and William Keeling in 1595 the grant thus included ‘all those lops of trees and bushes growing … in or upon Goodsgreen and Redhillstreet’ (HHA Deeds 198/38). In 1613 his widow Anne leased nearby Hoddesdonbury, and held the rights to the tops of the pollards on Cowheath Common and Martins Green. By this time Edward Collyn and Edward Payne held a 30 year lease to most of the Cecil family’s woodlands in Hoddesdon and Cheshunt, including ‘… the lops and tops of pollards growing in upon sundry greens, wastes and highways belonging to the manor of Baas, Hoddesdonbury and Geddings (except Goose Green, Redhill Street, Broxbourne Common, Cowheath and Martins Green)’. In 1556 the ‘loppes and shreddinges’ from the ‘common wood’ of Northaw were sufficient to pay the £56 rent owed by William Clarke, the manorial steward (Rowe 2015, 309). Over the twenty years between 1658 and 1678 the lord of Cheshunt manor made an average of £75 per annum from the pollards on Cheshunt Common: the wood was sold uncut at a set price, usually in batches of a hundred or fifty heads, to local residents who lopped them themselves (Rowe 2015, 310-12). It is noteworthy that most of these commons are still shown with tree cover on Drury and Andrew’s county map of 1766.

Given that significant amounts of money could be made from the pollards, presumably because of the proximity of London, the market it provided for wood and charcoal, and thus the inflated prices locally for wood, it is possible that local lords made not only piecemeal attempts to sustain tree cover on commons in this district, but on occasions undertook deliberate replanting schemes. Anne Rowe has pointed out how the 1650 survey of the land taken into Theobalds park from Enfield Chase noted that there were 140 acres ‘Indifferentlie well planted with lopt Pollards hollow Dotrills

93 hornbeame and oake’; while the 132 acres enclosed from Northaw Common were ‘reasonable well planted’ with pollards, and 149 acres taken in from Cheshunt Common were described as ‘beinge planted with wood’ (Rowe 2015, 310). Not too much should be made of this evidence, but deliberate planting as a way of maximising lordly incomes from unenclosed commonland, otherwise of little value to landowners, may be indicated. Active interventions on these lines, to raise the income from manorial wastes, may have occurred elsewhere. The remarkable beech pollards found in and immediately around Great Wood in north Norfolk, on acid gravely soils near the north coast, may not (as has been suggested) represent the survivors of an isolated pocket of native beeches, the most northerly indigenous examples of this species to be found in England (Rackham 1976, 27; Rackham 1986a, 141; 1986b, 326), but instead trees planted on the heaths by William Windham of Felbrigg Hall, who undertook a sustained forestry campaign in and around Felbrigg from c.1676. None of the surviving specimens appear to pre-date the late seventeenth century, and nineteenth- century writers such as James Grigor, while describing many old and large trees at Felbrigg, fail to mention any ancient beeches here (Barnes and Williamson 2011, 110-12; Grigor 1841).

Wood-pastures, especially on common land, were thus a significant features of the landscape of many parts of East Anglia and south east England well into the eighteenth century. The survival of tree cover on common land indicates the complexity of customary arrangements for preserving timber and pollards; but in south Hertfordshire especially it may also reflect the strength of local lordship, and the incentives provided by the value of the market – for fuel especially – stimulated by the proximity of London. As we have seen, this same district of poor soils was characterised by particularly high densities of farmland trees, and a high proportion of pollards. Where there was less of a market for wood, in contrast, or where rights to commons were widely shared and/or manorial control less strong, the degeneration of commons woods to open pasture seems to have proceeded more rapidly.

It is very difficult – partly because of problems of definition – to suggest figures for the area covered by wood-pastures in either county at any time in the past. In Norfolk, it seems likely that they accounted for around 5 percent of the land area in c.1550, perhaps a little more;9 a figure which had declined to less than 0.1 percent by the 1790s. In Hertfordshire, as we have seen, the figure may have been around 5 percent in 1550, declining to around 2 percent by the mid-eighteenth century.10 In other words, in c.1550 around 7 percent of Norfolk, and 12 percent of Hertfordshire, were occupied by woodland of some kind.

Woodland in the ‘Champion’: the forests of Northamptonshire

In the ‘champion’ areas of the Midlands – exemplified in this study by the county of Northamptonshire – both woods and wood-pastures were, as we have seen, strongly concentrated in areas of , with only scattered examples elsewhere. This situation remained true well into the eighteenth century (Figure 23). In Northamptonshire, the three forests of Salcey,

9 Most upland common land in Norfolk – i.e., commons other than fens - present in 1550 still survived at the time that Faden’s map was surveyed in 1790s (MacNair and Williamson 2010), around 12 per cent of the count area: perhaps a third had been wood-pasture. To this we need to add anther c.1 per cent, to account for deer parks. 10 Bsed on an estimate of 1 per cent of the county occupied by wooded deer parks; on an assumption that most of the common land existing in c.1550 still survived in c.1766; and that around half of this had been wooded. 94

Whittlewood and Rockingham appear to have become less rigorously managed in the course of the sixteenth and seventeenth centuries than in the Middle Agea. The Crown gradually ceased to use them for hunting, and the various royal parks were disparked: that at Cliffe disappeared as early as the 1590s, those at Handley, Hanslope, Hartwell and Stoke Bruerne all went in 1629, while Brigstock and Paulerspury were destroyed in 1639 and Grafton in 1644 (Pettit 1968). The Crown woodlands in Rockingham were progressively alienated by sale or gift, so that only those within Salcey and

Figure 23. Reconstructed land use patterns in Northamptonshire, c.1730: for sources and methodology, see Williamson et al. 2013. Woodland remained strongly concentrated within the area of the three royal forests.

Whittlewood remained royal property by the start of the eighteenth century (Pettit 1968). Nevertheless, little of the alienated land was actually disafforested, in legal terms (the exceptions were Pipewell woods and some other coppices in Rockingham Bailiwick, granted to Sir Christopher Hatton in 1628; and Geddington Chase, also in Rockingham Forest, which was disafforested in 1676). The ancient organisational structures of the forests, the various measures to protect the deer, and the cumbersome systems of forest administration, all remained in place. So, too, did the rights enjoyed by the holders of the principal forest offices, the most important of which had long been monopolised by major families. As late as 1853 the Duke of Grafton had to be compensated at the enclosure of the last portion of Whittlewood Forest for his offices of Warden, Master Forester or Ranger and Master of the Game (NHRO G4167).

More importantly, the rights exercised in the forests by the communities living in and around them also survived relatively intact right through the post-medieval period, their complex and overlapping character tending to militate against sustainable management. The most important of these rights

95 was common of pasture, which extended over not only the planes and ridings but also the enclosed coppices, which were by custom thrown open for grazing seven years after cutting. In Rockingham the grazing season ran from 29 April to 11 November; in Salcey from Old May Day to 22 November; and in Whittlewood from 5 April to 11 November in the case of the ‘in-towns’, and from 4 May to 25 September in the case of the ‘out towns’. The latter phrase reflects the fact that rights in all the forests continued to be shared, as they had been in the Middle Ages, between those holding commonable tenements in a large number of townships, some within or immediately beside them but others more distantly located. Geddington Chase in Rockingham Forest, for example, was grazed by the inhabitants of Geddington itself, Brigstock, Stanion and Newton Willows; Whittlewood by Whitfield and (the ‘in-towns’) and by with Crowfield, Wappenham, and Slapton (the ‘out-towns’) (NHRO G3909).11 There was some curtailment of common rights in the course of the post-medieval period, associated with the Crown’s extensive alienations, but they were relatively limited. Even full disafforestation, when it occurred, appears to have left common rights relatively intact. Lord Montague thus obtained Geddington Chase in 1628, for £1,850 and £10 annual rent, but although it was formally disafforested in 1676 local use-rights were not completely extinguished. Forest villages generally continued to enjoy their extensive areas of grazing, and their populations grew at a higher than average rate for the county, in part perhaps because they absorbed people displaced from enclosed and depopulated townships elsewhere in the Midlands. In Whittlewood the freeholders, with common rights, frequently constituted twenty or twenty-five percent of the population. In Rockingham the situation was different, with smaller numbers of freeholders and with most villages controlled by members of the minor gentry or aristocracy. Nevertheless, even here there was a significant influx of squatters (Pettit 1968).

In the Middle Ages the commoners could only turn cattle (in the modern sense) into the forests, but by the sixteenth century sheep, more damaging to trees and underwood, were commonly being pastured there on a large scale. The coppices were inadequately fenced and often grazed out of season, and overstocking was endemic: when Rockingham was enclosed, many commoners claimed the right to graze unlimited numbers of animals ‘all year round’ (NHRO Brooke of Oakley 318/1). The forest officers also claimed rights to graze in the coppices, their horses doing considerable damage to the re-growth: in 1720 Lord Montagu’s steward described how ‘horses were put into copses of seven years old. I have seen them and can find any manner of damage done ... they say it is a privilege that the keepers of this and other chases have’ (Toseland 2013, 42). Deer were poached on a large scale, and were continually disturbed by people entering the woods during ‘fence month’ (fawning season) and, in particular, while nutting. A notice printed in 1819 declared that as ‘the deer … have of late years have been much disturbed by people gathering nuts’, offenders would be prosecuted (NHRO Brooke of Oakley 313/21). At Silverstone, fights between forest officers and ‘offending nutters’ were said to be as fierce as any with poachers (Linnell 1932, 21, 103). There were numerous disputes between, as well as within, villages over grazing and other rights, and the right to collect ‘sere and broken’ wood, theoretically fallen material, was often interpreted as green and growing wood, either from trees or coppices. The Northampton Mercury in 1775 described how ‘the right to carry away the broken wood in ’ was abused by several persons who ‘most audaciously and unlawfully ... cut and carried away the mounds and fences ... in the new enclosures of Denton and in Castle Ashby grounds’. At in 1623 seven trees were stolen, supposedly

11 In-towns had right of common for their cattle from 25th March to 1st November; Out-towns from 23rd April to 25th September. 96 for maypoles (Pettit 1968, 125). The right to cut oaks for ‘coronation poles’ was similarly claimed at Cliffe in 1702 and Whittlewood and Salcey in 1714 and 1727, when troops had to called in to control the situation. But attempts to assert royal or aristocratic control were, in general, hotly disputed by villagers (Pettit 1968, 154-6). From the standpoint of the authorities, and major landowners, the situation was made worse by the fact that nobody had absolute control of particular tracts of land. Instead there was a complex pattern of overlapping ownership and rights – both manorial, and relating to forest offices.

Because of poor management, many of the wood-pasture commons in forest areas – as opposed to enclosed areas of coppiced woodland – appear to have become seriously degraded, and to have grown steadily more open in character through the seventeenth and eighteenth centuries. This was already occurring in some places by the end of the sixteenth century. Weldon, Benefield and Deenethorpe plains were, by the 1580s, devoid of trees but the density of charcoal hearths recovered by archaeological survey shows that it had been wooded in the Middle Ages (NHRO Westmorland 4 xvi 5 and 30; Watson Mun. A.5.22,D2817). Nevertheless, as late as 1565 a survey of the three forests recorded 93,942 oak standards (valued at £46,355) in the plains and lawns, and in the enclosed woodland still in the hands of the Crown, and a further 14,198 (£4,609) in the remaining royal parks: a by no means negligible figure, given that the land in question probably amounted to some 200 square kilometres – a density of around 5 trees per hectare – and that it appears to exclude pollards (TNA/PRO LRRO 5139). But by the end of the eighteenth century the areas of the forest pastures lying outside the embanked coppices had deteriorated considerably. James Donaldson in 1794 described them in highly negative terms, his account of Whittlewood particularly damning: ‘I know of no land in England, of equal staple, worse misapplied than a great part of this forest’; it was managed under a defective system that rendered it ‘worse than a state of nature’ Donaldson 1794 132-3). Pitt in 1813 similarly commented on the poor state of the plains and lawns. ‘In many places there is not an oak upon a dozen avenues, or any other timber-tree; but the land is wholly occupied by bushes of hazel, blackthorn, underwood and pollards, with here and there a beech sapling; in other places a few ash trees and pollards….’ (Pitt 1813, 148). By this stage, it is true, such views were highly coloured by a growing elite antipathy to commons and pollards, but the decline in the numbers of standard trees at least seems very clear. But the forest wood-pastures did not only disappear because they degraded to open pasture. Even in the sixteenth and seventeenth centuries some areas were deliberately cleared. In 1518 the plain at Pipewell was divided between the various townships which shared common rights there – Rushton, Pipewell and Desborough, Pipewell and Rushton – enclosed and probably cleared of trees, and enclosed; the plain lying between Benefield, Deenethorpe and Weldon suffered a similar fate in the late 16th century. When the two Brigstock parks was sold by the Crown to Sir Robert Cecil in 1602 he cleared and sold the timber and turned their area over to pasture for cattle (Petit 1968).

Timber survived best in the forests where area of wood-pasture had been brought into private ownership, even if common rights were still exercised. Geddington Chase in Rockingham was disafforested in the 1670s but as late as 1749 no less than 20,664 trees were recorded there. Only 7 percent were pollards, while 2.9 percent of the trees were described as ‘recently planted’ elms, and a number were exotics and ornamentals which had also, presumably, been established relatively recently by the Duke and his agents – various conifers, sweet chestnuts, horse chestnuts, limes and poplars – which together amounted to over 12 percent of the total recorded trees. In other contexts, where common rights continued to be exercised by large numbers of people and in a

97 poorly-regulated manner, there seems to have been few timber trees, and a declining number of pollards.

The coppiced woods within the forests, all of which were not only enclosed but, in the case of Rockingham after 1665, privately owned, generally survived better, although these too were affected by trespass, theft and illegal grazing. The Crown woods were surveyed in 1564 by John Houghton (British Museum Add. MS 34 214) and by Roger Taverner in 1565 (TNA/PRO LRRO 5139). These surveys show that, in the sixteenth century, the underwood was often left uncut for long periods: indeed, over 60 percent had not been cut for more than twenty years – although how far this represents a management system based on very long rotations, how far it was a consequence of poor management and neglect, is unclear. Certainly, by the seventeenth century rotations of sixteen to eighteen years appear to have been standard in the forest woods, longer than those described in Hertfordshire and Norfolk, although by the eighteenth century they had apparently shortened, usually to around twelve years. Pitt described how, in the early nineteenth century, the faggot wood was principally sold to bakers as fuel; how much of the ‘smooth wood’ was manufactured into sheep-hurdles and fencing; while the ash and sallow poles are used for various ‘useful purposes in husbandry’ (Pitt 1813, 156). Moreton in 1712 described, in addition, how large quantities of charcoal were made in the Rockingham woods, and sent to Peterborough; how the making of spoons and dishes was a speciality of Kings Cliffe, one of the townships within Rockingham Forest; and how smiths journeyed each year from as far away as Birmingham to buy ash to make bellows (Morton 1712, 11-12, 486). As in the districts already described, the coppices were sometimes cut directly by their owner but sometimes sold to, and cut, by one or more individuals, as at Winshaw and Westmore Hedges in Geddington in 1659, where it was parcelled into ‘furlongs’, each with an area of about 40 poles (NHRO Bru O V 229). In the seventeenth century the income from managed woodland within the forests, to judge from accounts of woods in Rockingham Forest, came mainly from timber, but with the underwood coming a close second. The accounts, which cover a total of 2,600 acres and run for thirteen years (1650-1662), reveal that an average of 41 percent of income came from coppice, 51 percent from timber, and a further 8 percent from bark (Pettit 1968, 116). A similar balance of profit is evident elsewhere: the underwood from 110 acres in Hawe in Stanion was sold in 1650 for £173, but the timber fetched £333 and the bark £37: wood from crab trees, destined for the making of cogs, was sold separately for £9 (NHRO Bru iv II). The income from underwood presumably includes, as in other areas, the twigs and branches from felled timber trees. Right through the seventeenth and eighteenth centuries the woods within the forests remained an important source of both wood and timber, and while much of it was sold locally some – especially timber from Rockingham - went further afield, being taken to Peterborough and thence by water to Kings Lynn.

In terms of assessing the proportion of Northamptonshire occupied by woodland at various points in the past we have to deal once again with problems of definition. How many trees need to be lost from an area of grazed woodland before we are obliged to classify it as an open pasture? This problem is compounded by the fact that accounts of the wood-pastures areas of the forests from the eighteenth century tend to be impressionistic. All we can say with reasonable certainty is that at the end of the Middle Age woodland and wood-pasture, mainly located in the three forests, accounted for around 10 percent of the county’s land area. By the middle of the eighteenth century the area of enclosed woodland in the county amounted to around 6 percent of the land area. It is likely that some of the plains and lawns were still at this time recognisably wooded, although

98 perhaps only a small proportion, perhaps adding no more than a further 1 or 2 percent of the total land area.

Woodland in the North: Yorkshire

Relatively little is known of the character of woodland in Yorkshire before the eighteenth century. The pollen records suggest that a period of clearance peaking in the high Middle Ages was succeeded by one of woodland expansion following the Black Death, before pressures increased again during the late fifteenth century. But as Gledhill (1994, 213–7) has noted, there are a number of problems involved in interpreting this data, including the possibility that the supply of pollen may have been influenced by the spread of coppicing. Other researchers, such as McDonnell (1992) – relying mainly on documentary evidence – have suggested a sustained process of woodland decline which continued right through the later Middle Ages. Whatever the reality, by the sixteenth century pressures from population, grazing and industry were growing. Much of the surviving wood-pasture in upland areas appears to have been enclosed and often replaced by open grazing in the course of the sixteenth and seventeenth centuries; and, following a pattern already established in the late Middle Ages, much of what remained was converted to coppice. In the Dales, by the time of the Dissolution, approximately 69 percent (571 acres, or 231ha) of the Fountains Abbey woodlands were already being coppiced, and only 31 percent managed as wood-pasture (Dormor 2002, 54). For the most part, the coppices were cut on long rotations, of 15 to 20 years (Dormor 2002, 56–7). Following the Dissolution, the attrition of wood-pasture, through conversion to coppice, continued on the former abbey estates (Dormor 2002, 95), a development which could be paralleled elsewhere. A number of deer parks in the county were thus converted, in whole or part, into coppice-with- standards, as at Hesley, Cowley and Tinsley Parks near Sheffield (Jones 2005, 50). It is noteworthy that the ancient pollards studied by Muir in Nidderdale almost all represent former hedgerow timber, rather than the remains of early wood-pastures (Muir 2000); and while Fleming in Swaledale found more evidence for wood-pastures, there are few signs that they continued to be intensively managed beyond the late seventeenth century.

Where wood-pastures did survive into the seventeenth and eighteenth centuries they were often small in area, and under private ownership, rather than (as in areas like south Hertfordshire) occupying large tracts of common land. At Brandsby in the Howardian Hills, for example, the combined evidence of a map and survey, made in 1746 and 1733 respectively, indicate a densely timbered area in the far south of the township called ‘Stephen Ware's Pastures’ (Nrth ZQG IV 1/1 and 2/8). A map for Howsham, on the edge of the Howardian Hills, from 1705 similarly shows that much of the eastern side of the township, towards the higher ground, was occupied by areas of enclosed wood-pasture, located in the same general area as the enclosed Howsham Wood (Bev DDX3/15); while at Newburgh, also in the Howardian Hills, a survey of 1605 shows a high concentration of trees within a knot of small closes just to the north of the deer park, with names containing references to haggs and hollins, managed stands of holly (Nrth ZDV). Gledhill (1994, 306) has suggested that the practise of cutting woodland trees for fodder was in sharp decline during the seventeenth century. However, in South Yorkshire, Jones (2012, 31-2) has found numerous historical references dating to this time. For example, a survey of the Manor of Sheffield in 1637 refers to 27 ‘Hollin haggs’, and there is little doubt that these were for fodder: Abraham de la Pryme described in 1696 how ‘in south-west Yorkshire at and about Bradfield and in Derbyshire they feed all their sheep in winter with holly leaves and bark … To every farm there is so many holly trees … care is taken to plant great numbers of them in farms hereabouts’ (Jackson 1870). There is evidence

99 that holly continued to be used for fodder into eighteenth century. In 1725, a party travelling across Birley Moor and Hollinsend near Sheffield rode through ‘the greatest number of wild stunted holly trees that I ever saw together … [They have] their branches lopped off every winter for the support of the sheep which browse upon them, and at the same time are sheltered by the stunted part that is left standing’ (Jones 2012, 32). As late as 1794, the agent of the Bolton estate said that ‘a very great havock was made last winter amongst the Hollies, which … were cutt down for the purpose of fodder for Cattle’, suggesting that holly, at least, continued to be employed as an emergency fodder crop throughout the eighteenth century (Dormor 2002, 197).

We have already noted how, even in the early Middle Ages, a diverse topography, and varied geology and soils, ensured stark contrasts in the extent of woodland and wood-pasture within the three Ridings. Such variations continued into post-medieval times. The abundance of wood-pasture in the Howardian Hills was part of a wider pattern: woodland of all kinds was most extensive on the fringes of the North York Moors and the Pennines, and to an extent within the Dales cutting through the latter. On the Howardian Hills enclosed woodland thus comprised approximately 8 percent, 14 percent and 15 percent of the land area of the townships of Brandsby, Howsham and Newburgh, in 1746, 1705 and 1605 respectively (Nrth ZQG IV/16; Bev DDX3/15; Nrth ZDV). They tended to be located towards the margins of townships but sometimes – as in the case of Howsham and Newburgh – also occupied (by the seventeenth and eighteenth centuries) more central positions, where they formed part of the parks and demesne lands of large estates. The landscape of Beadlam and Pockley, lying at the junction of Vale of Pickering and the North York Moors, was notable for its woodlands, comprising between 5 and 8 percent of the mapped area of townships and estates in 1785, although it was largely concentrated on the steep slopes of the gullies running down the escarpment (Nrth ZEW M 13 and 14a). In the Dales and on the Pennine edge, in contrast, the varied geology of coal measures, gritstone and carboniferous limestone, ensured much local variation, ranging from less than 1 percent in Walburn in 1700-1743 (0.57 percent) and Skeeby in 1779 (0.17 percent), to as much as 27 percent at Downholme in 1738 (Nrth ZAZ (M) 1; Nrth ZMI; ZBO (M) 1/5). Typical perhaps was Leyburn, with only one significant area of woodland, Leyburn Shawl, comprising approximately 4 percent of the area shown on a map of c.1730 (Nrth ZBO (M) 1/2 ), although later maps suggest that there might have been other wooded areas beyond the tract surveyed. John Leland in the mid sixteenth century noted how ‘the river sides of Nidde be welle woddied above Knarresburgh for a 2 or 3 miles, and above that to the hedde all the ground is baren for the most part of wood and come, as forest ground ful of lynge, mores and mosses with stony hills ... The principal wood of the forest is decayed’. In Wensleydale, similarly, he saw ‘very litle wood’ (Woodward 1985, 19).

The sloping and broken ground on the edges of the main uplands were thus the most densely wooded areas: on the moors themselves, and on the settled land of the Vales, woodland was in short supply. Maps of Healaugh and Catteron (1718) and of Warmsworth (1726) suggest that unenclosed townships in the Vales appear to have had, on average, only c.2-3 percent of their area under woodland (Lds WYL 68/63; Don DD/BW/E11/7); while those which had been enclosed apparently had even less, with no examples at all recorded on the four relevant maps, for Harlsey (1762), Birdforth (1770), Sinderby and Pickhill (1778) and Thormanby (undated, seventeenth century) (Nrth ZNS; Nrth ZDS M2/12; Nrth ZIQ; Nrth ZDS M2/1).

While it is impossible, and arguably meaningless, to estimate with any degree of accuracy the proportion of the county’s land area occupied by various kinds of woodland at the start of the period

100 studied here, or even in the eighteenth century, it is clear that Yorkshire was less wooded than Hertfordshire, Norfolk or Northamptonshire. Tuke in 1800 estimated that in the whole of the North Riding, embracing not only the Vale of York but also the North York Moors and the Howardian Hills, there were only 25,500 acres (10,319 hectares) of woodland - less than 2 percent of the total land area; and there is no doubt that there was even less woodland in the Wolds, which even at the time of Domesday probably had only around 4 percent of its surface area under trees. Taking into account an almost complete absence of wooded land both here, and on the moors above c. 250 metres OD; assuming that in champion ‘Vale’ townships, and those in Holderness, woods accounted for around 2 or 3 percent of the surface area; and that even in townships on the moorland fringes only around 10 percent of the land was wooded; then it is likely that even in the mid-eighteenth century only around 3 percent of the land area of Yorkshire was occupied by woodland or wood-pasture. Indeed, even in the late nineteenth century – after a period of sustained planting by large estates – the figure remained below 4 percent.

Evidence for the management of coppiced woodland in Yorkshire is relatively sparse before the eighteenth century. A late sixteenth-century survey of the Manor of Settrington (King & Harris, 1963), located just below the sharp escarpment lying between the Vale of Pickering to the Wolds, shows that the woodland, lying on the steep sides of gullies and other slopes, was divided into eight coppice springs or ‘haggs’ which were felled on an eight-year rotation. The surveyor found that although 980 trees were standing, a further 1,581 had been illegally felled, while in five of the coppices there was ‘in effect noe Tymber at all left’ (King and Harris 1963, 95) and he advised that ‘better provision be made … for the better preservation of the woodes hereafter’ (King and Harris 1963, 31) and that the most badly damaged areas should be clear felled and converted to pasture. He also suggested other factors militating against the preservation of woodland, most notably the rabbits which were ‘destroying the young springs and they [had] altogether destroyed one springe called Peeke springe’ (King and Harris 1963, 32). In 1537 the manor had been ceded to the Crown following the execution of Sir Francis Bigod, its previous owner; it was then granted to the Earl of Lennox before returning to the Crown again; the survey was probably instigated by as part of another plan to dispose of the manor, and indeed it was granted to Ludovic Stuart, Duke of Lennox in 1603. It was perhaps a lack of consistent, attentive lordly oversight during this period that led to the degradation of these particular woods.

From the eighteenth century we have more information about management, particularly in the form of lease agreements. These suggest similar concerns on the part of landowners as we have noted in the southern counties, especially regarding the need to protect the coppice when newly cut, and to ensure a forward supply of timber trees. One lease from the Duncombe estate, from 1708, describes how John Conway, Richard Sawyer and William Bards were to fell fifty acres of coppice on the condition that sufficient ‘garsell’ - thorny brushwood such as holly and crabapple - and underwood were left for fencing the newly felled area. In addition, there appears to have been more concern than in the other counties studied about the effects of climate on recently-cut coppice. Conway and his partners were thus bound to ‘fairly spring fell and cutt down with axes or hatchets for the best advantage of the wet Growth of the said coppice spring of wood to be cutt down’ (Nrth: ZEW IV 7/25); while another lease agreement, from the Arthington Hall estate, describes how the lessees were to cut ‘sloping … so that the water cannot nor do stay upon the stovens (coppice stool) or roots that are left standing but shall & may slide from them & shall not cut down or fell any tree with a

101 saw not make any use of a saw in felling or cutting down any of the trees in either of the said woods’ (Lds WYL160/220/157).

This document, like a number of others, also describes how the lessees were to leave an appropriate number of 'weavers' (or wavers), ‘Lordings’ and ‘Blackbarks’ (standard trees left standing for one, two and three coppice cycles respectively) to ‘stand for the benefit of the succeeding growth’ – i.e., to grow on as timber trees. Again, protection of young coppice from livestock was of paramount importance, with the lease requiring that draught animals be kept under control, and the cut coppice should be securely fenced. The lessees were also instructed to ‘take care to make & sett their charcoal pitts & make all their fires that shall be made in the said woods in the most open & waste places therein … so that they nor any of them shall or do receive and prejudice or damage thereby’. The lease also included a provision that allowed the landowner, the Earl of Cardigan, to exchange a poorly performing reserve tree (a ‘Lording’ or ‘Blackbark’) for another within the population that was to be felled (Lds WYL160/220/157). Some leases went further than those in the other study areas in safeguarding the interests of the lessees. One from the Ripley Estate, from 1737, included provisions by which the owner (Sir John Ingilby) agreed to protect the price of timber he had sold by not to selling any other oaks in the vicinity of Guisecliff and Leadwath for two years following the sale, which involved nearly 400 timber trees of oak and ash trees (Lds WYL230/2828). Another, relating to woodland on the Stansfield estate near Bradford in 1724, provided compensation if the area of woodland failed to measure up to the estimated 60 acres that was detailed in the contract (Lds WYL500/292). In the cases just described, individuals or small groups contracted to lease all the underwood for a period of years, but in other cases, as in the south, the coppice was sometimes divided up into small blocks, allocated by lot, for cutting.

While care was clearly taken to protect Yorkshire coppices from browsing stock, as in some southern woods domestic animals were not necessarily excluded completely: indeed, the fact that rotations here were often longer than in the south (see below) may have encouraged the use of woods for grazing. In the late seventeenth century the tenants of the Arundel estates around Sheffield, for example, could pay for rights of herbage in the local coppices, presumably exercised late in the rotation: many were, however, prosecuted for illegally taking wood (Shf ACM 2/279 and Shf ACM 2/280). On many Yorkshire manors, however, tenants did have the right to collect wood from the coppices, within prescribed limits, although it is striking that this was generally referred to not as firebote, but as housebote – that is, the taking of larger pieces of wood or timber, for the repair of dwellings, rather than just smaller material, for fuel (this right was being exercised as late as 1789 on the Danby estate (Nrth: ZDS IV 14/6). This is part of a wider pattern: fuel from woods is seldom mentioned, even as faggots, in the available sources. The presence of extensive peat deposits on the moors, and the early development of the South Yorkshire coalfield, already discussed, seems to have limited the use of wood as a source of domestic fuel.

Yet if evidence for woodland as a significant source of domestic fuel is limited, references to alternative markets, and alternative uses, for underwood are abundant. Indeed, material from coppices seems to have been employed in a wider (and certainly different) range of ways than in the Midlands or south. In the Dales ash was commonly used for making dairy utensils and butter furkins (Marshall 1788, 214); besom makers and clog makers are also referred to as customers for woodland produce, alongside the kinds of trades already noted further south, including builders and hoop- makers (Dormor 2002, 259). But more striking are uses related to industry. In the mineral-rich uplands, and on the coalfields in the south of the county, both timber and larger coppice poles were

102 in demand for pit-props, while the textile industry needed coppice wood for bobbin mills. Charcoal making was of particular significance, often conducted within the woods themselves. In Ecclesall Woods on the outskirts of Sheffield the remains of nearly 200 pits have been linked to this industry (Gowans and Pouncett 2003). Whitecoal – that is, the rapidly dried wood widely used for smelting lead before the late eighteenth century – was also made in many woods in lead mining districts, with over 100 of the distinctive ‘Q-pits’ associated with its production recorded from the Eccleshall Woods alone. The close connection between industrial production and woodland management is evident from the fact that forges frequently owned, or leased, areas of woodland.

Post-medieval coppice rotations seem to have been more varied, and often longer, than those practised in the other counties studied, perhaps because of the diverse character of these markets, although the harsh climate endured where coppices were located on higher ground especially might also have been a factor. Tuke in 1800 provided a slightly confusing description of management systems in the North Riding. He suggested that most woods were managed by ‘spring-felling’, by which all the trees and underwood were felled on a long rotation of 20 or 30 years; but he also describes a system in which timber was left to grow to maturity, but the underwood was cut more frequently; as well as a method, reported to be in decline, of more indiscriminate felling, leaving only a few wavers (small trees) to continue to grow after each cycle (Tuke 1800, 183-4). Local documents similarly suggest two broad kinds of rotation. On the Blacketts Estate near Bradford in 1750 Garner Wood was found to be of ‘perhaps 12 or 14 years growth & ought to be felled with Gill Wood’ (Brad 23D98/3/5). Similar rotation lengths, of between 13 and 16 years, are recorded on the Armytage estates in West Yorkshire and at Howsham in North Yorkshire in the late eighteenth and early nineteenth centuries (Cald KMA 1208; Nrth ZCG IV 5/2/10). At other places, in contrast, 20 or 21 year cycles were practiced, as in the Earl of Arundel's woodlands around Sheffield in the early nineteenth century, on the Arthington Hall estate in 1708, or on the estates of the Earl of Rockingham in south Yorkshire where, in 1727, a new management system was adopted by which all the woods were to be coppiced on a 21 year rotation, cutting approximately 40 acres of wood per year (Shef WWM A1273; Jones 2005, 52). It is likely that these longer cycles were important where the production of bark for tanning was a major consideration, for much of the underwood in Yorkshire woods was oak. It is also probable that the substantial poles produced by such long rotations had uses as pit timber for mines. Different woods on a property might be managed under different regimes, to supply different markets. On the Duke of Leeds' estate at Kiveton Park, for example, there was a mixture of shorter and longer felling cycles, which produced a wide range of pole sizes, suitable for ’hop poles, scaffold poles, cordwood, pit wood ('puncheons'), heft wood, hazel hoops, hedge bindings’ (Jones 2005, 52).

Where long rotations, of twenty years or more, were employed they tended to blur the traditional distinction between underwood and timber, which was so important in deciding whether material was subject to payments of a tithe or not. At an enquiry held in the mid nineteenth-century ‘advertisements were produced of the sale of the Estates in which woods of 5, 9, 10, 13 and 15 years growth were called coppice’, and ‘it was admitted that the woods are generally cut from 28 to 40 years growth’, while ‘some woods are cut at 23 years growth when the greater part is used for fire wood but it is not usual to fell them at that age but much depends either on the immediate necessity or convenience of the owner’ (Shf FB/CP/25/11). It is also possible that, where long rotations were practised, there was some reduction in the density of timber trees in woodland, and it is noteworthy that Tuke, writing about the North Riding in 1800, was able to describe how ‘within the memory of man, full-grown timber abounded in many of the parts of the district under survey’, but that ‘large 103 full-grown timber is now become very scarce’ (Tuke 1800, 183-4). There are some indications of such a change in valuations and surveys, with smaller trees gradually coming to form a larger proportion of the recorded population as the nineteenth century began: in the 1780s and 1790s, smaller trees tended to comprise between 5-40 percent of the total population, but after 1800 they were either completely absent or comprised between 60 and 100 percent of the recorded specimens. This said, most woods appear to have included timber trees as well as underwood. Leases frequently list trees left ‘to stand’ through felling cycles, and sales contracts commonly stipulate the protection of such reserves, even when the rotation for coppicing was twenty years (see for example, Lds WYL160/220/157, a lease of two coppice woods in 1708).

The local shipyards constituted a major market for mature timber (the Kiveton estate woods supplied, in 1701, 562 straight oak trees to be delivered to the royal dockyards at Chatham (Jones 2005, 52)), albeit one which was progressively eroded by imports from Scandinavia, brought in via Hull. Norway was a particularly big exporter because its saw mills, equipped with particularly fine blades, were thought to produce the best quality deals. During the eighteenth century, however, the development of sawmills in the Hull area meant that there was an increase in imports of unsawn rough wood ‘from nothing at the beginning of the century to 1,135 loads in 1758’ (Jackson, 1975, 16). By 1783, according to Jackson (1975, 22), ‘we can assume that the new mills, foundries, potteries and workshops east of the Pennines were constructed from wood and iron imported through Hull’. It is difficult to assess whether this increasing level of imports had any significant effect on local markets. Both Marshall (1788, 245) and Tuke (1800, 188) comment on the difficulties and costs of transporting native timber far from its source. Nevertheless, local demand – especially from the mines – remained buoyant, and while some bark for the tanning industry, as noted, came from coppices cut on a long rotation, much also came from timber trees. By the early nineteenth century F. Dunkinfield Astley was able to report that ‘the bark of oak [is] now so valuable as to reach in strong timber nearly half the value of the tree’ (Shef MD3520). Even before this, figures of between 20 and 30 percent of total value were common – on the Millford Estate near Leeds in the eighteenth century bark accounted for around 20 percent of the sale value of oak trees but at Hutton Rudby in the 1630s the figure was as high as 33 percent (Lds: WYL500/939), significantly above the values recorded in Midland and southern counties, suggesting both a greater demand for bark in this more industrialised area, and limited supplies in what was generally a poorly-wooded county.

In short, post-medieval woodland management in Yorkshire, while in its essential features comparable to that found in other districts studied, had its own particular character which was associated, for the most part, with early industrialisation. This appears to have lessened the demand for wood as a domestic fuel, just as it limited the number of farmland trees and the proportion which were pollarded. But at the same time it encouraged the management of the coppice in diverse ways, especially in order to produce pit props, bark, charcoal and whitecoal. Although historians like McDonnell (1992) have joined some contemporary commentators like Tuke (1800) in lamenting the poor management of the county’s woods, and a resultant decline in the availability of oak, there is little in the documentary sources to suggest any great neglect. As Dormer’s (2002) study of the Bolton estates has demonstrated, periods of felling usually coincided with the need to pay off debts or capitalise on periods of acute industrial demand; but these could be followed by phases of more careful management, to allow the woods to recover. Above all, as in other parts of England, early modern industry did not ‘use up’ woodland, but rather ensured its profitability and thus its survival (Jones 1993, 1997, 2005; Hammersley 1973). 104

Age of felling

The average size of the timber trees recorded in woodland in all the areas studied displays more variation than that of trees growing on the surrounding farmland. It was also sometimes lower, as at Panshanger in Hertfordshire in 1719 (16 cubic feet, as against 29) (Austin 2008) but usually higher, with a higher proportion being left to attain a volume of 25 cubic feet or more. Such trees were, however, often accompanied by large numbers of very small specimens, resulting in mean average sizes almost identical to those found on farmland, of around 16 cubic feet. Typical was the timber recorded in a valuation of woodland taken in 1774 at Nunburnholme on the East Yorkshire Wolds, which detailed 244 oaks and 324 ashes, a high proportion with volumes of below 10 ft, with the frequency only gradually dropping away to around 50 ft, giving a mean size of around 15 feet (Hull U DDWA/x1/4/3). This pattern is also seen in the woods on the Boughton estate in Northamptonshire in 1749 (Boughton House archives). The proportion of farmland trees containing more than 20 cubic feet seldom exceeded 5 percent, but at The Haw and Ravens Wood in Weekley it was over 9.8 percent, in Boughton Wood 23.8 percent, and in the wood-pasture of Geddington Chase 37.7 percent. These woods also contained many very young trees: in The Haw and Ravens Wood 44 percent of the trees were oaks with less than 10 cubic feet in them and at Boughton Wood 62 percent. It is possible that such rather polarised age structures reflect felling ‘events’, affecting most or all of the wood, and it is noteworthy that a valuation of 861 oaks from Newburgh in Yorkshire, made in 1776 and probably all from woodland, shows a particularly sharp distribution curve, with a large proportion of trees yielding 10-20 ft of timber and very few above 50 ft, giving a mean of 25 ft per tree, perhaps suggesting a substantial felling, and replanting, of trees at some recent point in the past (Nrth ZDV). Recent management patterns of individual woods may also explain the radically different age structures sometimes exhibited by different species: a 1781 valuation of the trees standing in plantations on the same estate lists oak, ash, elm and sycamore with an average size of 42 ft. per tree, but the sycamore and elm have a greater range of sizes than the oak and ash, and include many more larger trees (Nrth ZDV).

Woods may often, therefore, have contained more older trees than farmland, retained until they were eighty years or even a hundred years of age, for particular purposes. But they were, nevertheless, likewise dominated by young timber, and most of the trees within them were evidently felled when still barely mature. The account books of Clement Chevallier, a landowner in Suffolk (outside the main study areas), shows this clearly. Of the 138 timber trees felled in Aspall and Bedingfield Woods between 1728 and 1742, only 17 contained more than 30 cubic feet and only 11 had a quarter-girth of more than a foot, suggesting an age of less than 50 or 60 years. Few comments in letters or memoranda books seem to contradict this general picture: even the note in the Knebworth estate account book for 1815 that ‘no timber in Humbly Wood [Whormerley] would be fit for felling for the next 50 years’ (Beckley 2002, 15) need imply nothing more than that the wood had been recently clear-felled.

Conclusion.

In all the districts examined, enclosed, coppiced woods occupied a relatively small proportion of the total land area in the sixteenth, seventeenth and eighteenth centuries. Even in well-wooded Hertfordshire it probably constituted no more than 7 or 8 percent of the total land area, even at the start of the period studied here. Such woodland produced fuel but also materials for specialised

105 uses, for making hurdles, tools and containers, for roofing and construction, for producing charcoal and – in Yorkshire – for pit-props and other industrial requirements.

While coppiced woods occupied a relatively small proportion of the land area, in the south of England and in East Anglia wood-pastures – grazed woodlands of various kinds - were also extensive, even in the eighteenth century. Indeed, there is little doubt that, in consequence, the overall size of the woodland resource in the early modern period, usually considered only in terms of enclosed woodland, has been under-estimated by many historians. Most wood-pastures were located on common land, where their survival – the extent to which they resisted degeneration to open heath or pasture – appears to have been largely related to institutional and economic factors. They proved most sustainable where management structures were most efficient, and where there was the greatest economic demand for poles cut from pollards. This form of land use thus survived best in highly manorialised districts where there was a high demand for fuel – most notably, perhaps, on the commons of south Hertfordshire, a district also characterised, because of the proximity to London, by large numbers of farmland trees and a high density of hedgerow pollards. All this said, even when the area of grazed woodland is added to that of enclosed coppices, it is unlikely that the two together constituted much more than 10 percent of the land area in any of the counties examined at any time since 1550. Most trees grew elsewhere, on farmland.

Above all, what this brief review of woodland history has emphasised is the essentially artificial character of woodland. Wood and timber were crops, managed in complex ways for use and profit, just as the majority of trees in the wider landscape were deliberately planted or encouraged there, by human choice, and managed in highly artificial ways. As we shall see in the next chapter, this essential artificiality was also a characteristic of the particular species of tree found in particular contexts in the past.

106

4. Tree Species in ‘Traditional’ Landscapes

Farmland trees: the dominance of oak, elm and ash.

The most striking characteristic of tree populations in England, at least in the period since the late seventeenth century, has been the overwhelming dominance of just three kinds of tree – oak, ash and elm. Surveys and maps almost invariably show that these together accounted for between 85 and 100 percent of the trees growing on farmland, and virtually all of the trees (as opposed to underwood) found in enclosed woods. Only wood-pastures, whether private or on common land, and areas of marsh or fen, sometimes boasted a significantly different range of species. What is also striking is the fact that, within particular areas, the relative balance of the three principal trees – especially on farmland - displayed a degree of variation which, to judge from the available sources, remained fairly stable over long periods of time.

In Norfolk, for example, oak (Q. robur and Q. petraea are not distinguished in the documents) was usually the most common farmland tree, and elm the least, but different regions of the county displayed subtle variations on this general theme. On the boulder clay soils in the south and centre elm was generally present in reasonable quantities (usually between 5 and 15 percent of recorded trees), and ash trees were numerous, reaching 51 percent of the timber recorded on a property in Beeston-next-Mileham in 1761 (NRO WIS 138, 166X3), although this was unusual, and oak was normally in the clear majority. At West Bradenham in 1750, for example, oak made up 42 percent of the 3,783 farmland trees valued and elm (presumably Ulmus carpinifolia) and ash between them, undifferentiated in the survey, the remaining 58 percent (NRO MS 9316, 7B9); while oak made up over 80 percent of the timber sold on a property in Whissonsett in 1762 (NRO HIL3/34/1-27, 879x3). On the light loams in the north east of the county, in contrast, the dominance of oak was generally greater and elm was present in negligible quantities. Of the 24,000 trees recorded on the Earle estates in 1722, 59 percent were oak, with 37 percent ash and a tiny 2.5 percent elm (NRO BUL 11/283, 617X2); on land owned by the Blickling estate in 1576, 192 of the trees were oak or ash; only one was elm (together with a single ‘Great Poplar’). On the poor acid soils of Breckland, and on some of the more acidic and better-drained clays in the south east of the county, elm was often better represented. Of the 79 trees recorded growing on the Merton glebe in 1846, 86 percent were oak and 10 elm, with only 4 percent ash (NRO WLS LXX/22/1-41, 481x7). But elm was most prominent on the clay soils on the edge of the Fens, around the town of Downham Market. On the Stradsett estate lands at West Dereham in 1813 and 1814 elm made up 31 percent of the trees, oak 36 percent and ash 33 percent, but lists of timber felled on the same property in 1813 and 1820 list slightly more elm than ash, and no oak at all (NRO BL/BG 5/3/4). Of the 1,826 timber trees valued on the Stow estate at Wimbotsham and Stow Bardolph in 1815 oak was the most common (762, or 42 percent) but elm came second (392, or 22 percent) with ash third (347, or 19 percent) (the remaining 17 percent included alder and trees like horse chestnut and sweet chestnut growing near the mansion itself) (NRO HARE 5500, 223X1); while on the estate properties in Marsham, Fincham and Shouldham 53 percent of the 1,062 trees were ash, 24 percent were elm and only 23 percent oak (NRO BL/BG 5/3/4). These profiles are striking similar to those found in Northamptonshire, described below, and it is noteworthy that this is the only region in Norfolk which, in terms of both soils and geology, and fields and settlement patterns, is more ‘Midland’ than ‘East Anglian’ in character.

Similar contrasts, from district to district, are apparent in other counties. In east Hertfordshire, on boulder clay soils broadly similar to those of south Norfolk, oak, elm or ash can appear in almost any combination. Elm (here presumably Ulmus procera) was occasionally the dominant tree recorded, as at Much Hadham in 1803 (69 percent elm, 7 percent oak 24 percent ash) or on Pearces Farm in Thorley and Sawbridgeworth in 1807 (elm 71 percent, oak 22 percent and ash 8 percent) (HALS Lob B/PC4/4; HALS DE/H/P11); more frequently ash was the most numerous species, as at Meesdun in 1762 ( 39 percent ash, 34 percent elm and 28 percent oak) (HALS D/Ecn:C1). But on most farms and estates oak was again dominant, as at Olive’s Farm in Hunsdon in 1556, where it accounted for 92 percent of the 403 trees recorded, or on a farm in Colliers End in 1794 (TNA PRO E315/391; HALS DE/B1768 P2), where it comprised 93 percent. In the south of the county, in contrast, on the acid soils of the London clays and the pebble gravels, ash was routinely less prominent and elm noticeably more so, its numbers commonly rivalling or even exceeding those of oak. Of 1,782 timber trees recorded in Barnet in 1786, for example, 63 percent were oak, 30 percent elm and only 7 percent ash (HALS DE/B983/E1); while of the trees valued at the enclosure of Aldenham in 1803 most were elm (51 percent), with oak coming a poor second (24 percent cent), followed by ash (14 percent) (HALS DE/X216/B2). The west of the county – the Chiltern dipslope and outwash gravels - appears different again, although the sources from this district are sparser and generally later. Here oak was almost invariably the most common farmland tree, accounting for an average of 64 percent of the trees recorded in eighteenth- and nineteenth-century surveys, and never falling below 45 percent of the listed trees, with elm and ash vying for second place.

Northamptonshire presents a very different picture. As we have seen, much of the county comprised, until the middle or later decades of the eighteenth century, open fields, rather than closes bounded by hedges, and pre-enclosure surveys reveal that most of the trees growing within the village tofts were elm and ash, with willows common in the wider landscape, where they were concentrated away from the arable land, on floodplains. Oak was relatively rare on all soils, accounting for only 19 percent of the trees surveyed at Wilby in 1764, 15 percent of those recorded when Finedon was enclosed in 1806, and less than 1 percent of those noted at the enclosure of Irthlingborough (NHRO B(G)1; NHRO ZA898; NHRO ZA 906). Even when the open fields had been enclosed, and the land lay in hedged fields, oak often remained the minority tree. In a survey of the Duke of Powis' extensive estates in Upper and Nether Heyford, Glassthorpe and Newbold, made in 1758, only elm and ash are specifically noted (NHRO ZB 1837); while oak was the third most common tree in a quarter of the 16 Northamptonshire parishes in which the Montague estate held property in 1749, and the second in all but two of the others (Boughton House archives). In some of these places elm was the most important tree, but usually ash was in a clear majority, as it was on John Darker's properties in Gayton, Tiffield, Kislingbury, Milton, Litchborough and Upper Heyford in 1791, where 59 percent of the trees recorded were ash, 21 percent oak and 20 percent elm (NHRO YZ 2183). This secondary (or tertiary) importance of oak as a farmland tree does not appear to have changed significantly in the course of the nineteenth century. The species accounted for only 20 percent of the total population on properties at Staverton in 1835, 6 percent at Welton in 1839 and around 8 percent at Teeton in 1812 (NHRO ZB 887; NHRO ASL 392; NRO B(H) 461). The importance of ash and elm, over oak, as a farmland tree was apparently more widely shared across the champion Midlands. A survey of the manor of Milcombe in Oxfordshire, for example, made in 1656 includes a list or ‘what wood and timber is growing on the premises’ of the tenants. Almost all the trees mentioned are elm or ash, together with – as in Northamptonshire – significant numbers of willow (NHRO C(A) Box104 4 1656).

108

In Yorkshire the situation was different again, although the issue is complicated by that county’s vast area, mixed underlying geologies and contrasting enclosure histories. Here, timber surveys and valuations from before the mid nineteenth century scarcely mention elm, or suggest that it was present in very low numbers. In some places, oak was overwhelmingly dominant – as at Ellerton in 1768, where there were 1,015 oaks, but only six ash and no elms (Bev DDX 73/51); Thorne in 1818, where oak made up 72 percent of the farmland trees, ash 21 percent and elm 5 percent (Bar EM 291); or Temple Newsam in 1825 where the figures were 50 percent, 38 percent and 2 percent respectively (Lds WYL189/B/61). Overall, oak constituted around 72 percent of farmland trees, ash 25 percent, and elm around 2 percent, in the surveys examined.12 Again, there were regional and local variations, with ash noticeably more numerous on the more calcareous soils, and rarer on the more acidic. Most striking however is the paucity of other species mentioned in valuations and surveys, constituting in all around 2 percent of the total.

The surveys and maps on which the above discussion is based have, it should be emphasised, numerous biases. They often ignore, or fail to name the species of, trees managed by pollarding, which were often (as we have seen) very numerous and in some areas at least more diverse in character. In south Norfolk, for example, hornbeam seldom appears as a farmland tree in the documentary sources, but is present in the landscape today, albeit at low frequencies, as a veteran hedgerow pollard (Figure 24) (Barnes and Williamson 2011, 74-6). This said, when the species of pollards are specified, they usually appear to broadly replicate the balance of species of the timber trees or at least, make little difference to the overall figures, although there was a general tendency for a higher proportion of pollards to be ash trees. On John Darker's Northamptonshire estate in Gayton, Tiffield, Kislingbury, Milton, Litchborough and Upper Heyford in 1791, for example, ash made up 60 percent of the total trees recorded, oak around 20 percent and elm 20 percent; when pollards are discounted, the figure were 48 percent, 25 percent and 27 percent (NHRO YZ 2183). On the Earle estate in north east Norfolk in 1722, 66 percent of the timber trees recorded were oak, 31 percent ash and just under 3 percent elm; the figures for pollards were 70 percent, 28 percent and 1.2 percent ((NRO BUL 11/283, 617X2). The overall patterns that these sources indicate – the overwhelming dominance of oak, ash and elm, and regional variations in their relative importance – thus appear broadly reliable.

This long-lived dominance of these three species in the English countryside is remarkable given that there are at least twenty-five indigenous, or long-naturalised, species capable of growing into reasonably-sized trees, with a height of ten metres or more. The contrast with the balance of species which existed in the wild vegetation, before the advent of farming, within the areas studied is striking. Although pollen diagrams from the four counties studied do reveal that oak, elm and ash were important trees in the ‘climax’ vegetation, ash in particular was generally present at lower frequencies and elm was much more abundant, at least before the ‘elm decline’ of the early

12 See list of sources at the end of this report. 109

Figure 24. Pollarded hornbeam, probably more than three hundred years old, at Great Moulton in Norfolk. There are a number of ‘veteran’ specimens of hornbeam surviving in the county, but the tree is very rarely mentioned in early surveys.

Neolithic. Oak (accompanied by hazel) dominated the vegetation in Yorkshire, but large areas of birch-alder woodland also occurred on the damper ground (Rackham 1990; Gledhill 1994, 88-95); while in the three southern counties small-leafed lime (Tilia cordata) was usually the most important tree and species like pine and yew were locally important, together with alder on wetter ground (Bennett 1983; Bennett 1986; Peglar et al. 1993; Campbell and Robinson 2008). Pollen of pre- Neolithic date from Diss Mere in south Norfolk was typically characterised by c.40 percent lime, 20 percent oak, and by 10 percent alder, ash and elm (Peglar 1993, 7).

110

Oak, elm and ash were not, however, dominant in the farmed landscape because they were in some way well suited to flourish and sustain themselves in such an environment. Most farmland trees were deliberately planted, or at least encouraged, in particular places. This is clear from the many documentary references to tree- planting, especially but not exclusively when new hedges were being established. The survey made in 1758 of the Duke of Powis' properties in Northamptonshire, for example – comprising a number of parishes which had seen much recent enclosure - contains such phrases as ‘Fine young Quick fences of c. 8 yrs growth with great number of thriving elm and ash trees’, and ‘Well fenced with young Quicks of c 8 years growth in which are planted a great many fine thriving trees’ (NHRO ZB 1837). On many estates the planting of hedgerow trees was a condition of tenancy. On the Temple Newsam estate near Leeds, for example, leases making this stipulation are found from as early as 1608, continuing until at least 1768 (Lds WYL100/NE/2; Lds WYL100/NE/110). One from 1709, for example, bound the tenant to plant six oak, ash or elm trees annually for the 21 year duration of the contract (Lds WYL100/NE/83a). In 1700 Lord Fitzwilliam instructed his steward to inspect all the ‘wood growing which his tenants have planted and do plant yearly’ on his estates in Northamptonshire (Hainsworth and Walker 1990, 44).

Indeed, even in the Middle Ages there are a references to the planting of farmland trees. At Hindolveston in Norfolk in 1312, for example, two men were paid two shillings each for six day’s work, ‘pulling ashes to plant at Hyndringham and Gateli’ while one received 25d for thirteen and a half days work ‘planting ashes in the manor’ (Rackham 1986a, 224). In Forncett in 1378 men were paid for ‘pulling plants of thorn and ash to put on 1 ditch from the south of the manor to the churchyard’ (Davenport 1906). It is striking that seventeenth- and eighteenth-century writers on farming and forestry seem to assume that most trees were deliberately planted in hedges and on farmland, rather than arriving there adventitiously. Ellis always refers to elm as ‘planted in a hedge’; Blagrave urged the raising ‘upon each Lordship or Pasture, Fuell and Fire-wood sufficient to maintain many Families, besides the Timber which may be raised in the Hedge-rows, if here and there in every Pearch be but planted an Ash, Oak, Elm…’ (Blagrave 1675, 114). At the end of the eighteenth century William Marshall, writing about Norfolk, described how ‘Upon some estates it is the practise to put in, when a new hedge is planted, a holly at every rod, and an OAK PLANT at every two or three rods, among the whitethorn layer’; while Nathaniel Kent described how small proprietors enclosing open- field land piecemeal in the same county set within the new hedges ‘an oak at every rod distance’ (Marshall 1787, 113; Kent 1796, 72). Tuke, writing about Yorkshire at the end of the eighteenth century, worried about the ‘neglect of planting trees in hedge-rows, and of proper management of those which are now growing there’ (Tuke 1800, 191). Mortimer in 1707 thought that ‘The best way of raising Trees in Hedges, is to plant them with the Quick’: but it is important to note that he also gave advice on how to establish them ‘wheare Hedges are planted already, and Trees are wanting’ (Mortimer in 1707, 309).

More important, perhaps, is the fact that of these three trees, only ash will easily seed itself in hedges. Oak may be present in some hedges as part of a ‘woodland relic’ population, and it will seed there of its own accord, but only very slowly, as is clear from hedgerow surveys, which generally reveal that that hedges planted in the seventeenth or eighteenth centuries, while they frequently contain oak trees, do not normally contain oak as a shrub - suggesting that the former have not developed from self-seeded specimens but have been deliberately established (Hall 1982; Addington 1978; Cameron and Pannett 1980). The principal species of elm (Ulmus procera and carpinifolia), moreover, do not set seed at all and, while elm trees in hedges can sucker from existing stumps,

111 these must usually have come from specimens which had been planted there in the first place (or again, and rarely, were present in ‘woodland relic’ populations). But even where hedgerow trees were self-seeding, lease agreements and other documents make it clear that they had to be carefully protected from grazing stock and other hazards, including damage by tenants. Initial establishment, that is, might be the consequence of natural process, the plant’s survival to maturity was a function of human agency. A lease for a farm in Barnet in Hertfordshire, drawn up in the early eighteenth century, typically instructed the tenant to ‘do every Thing in his Power for the Encouragement, and growth of the young Timber Shoots, under the Penalty of Twenty Shillings for every Shoot or Sapling which shall be wilfully hinder’d from growing’, with a penalty of twenty shillings prescribed if he was to ‘stub up, prune, or injure any Sapling’, and of five pounds for destroying or injuring any young timber tree (HALS DE/B 983 E1).

The dominance of farmland tree populations by oak, ash and elm was thus to a large extent a consequence of deliberate choice. Two main factors seem to have ensured their overwhelming popularity: their ability to thrive in a wide variety of environments; and the high demand for, and wide range of uses for, their timber and wood. Early writers of books on husbandry or forestry always discussed these three trees, sometimes followed by beech, walnut and chestnut, before any other. All agreed that oak made excellent structural timber, good for the ship-builder and the house- builder: ‘Oak hath the preheminence of all others, for its strength and Durableness’; it was ‘The best Timber in the World for building Houses, Shipping, and other Necessary Uses’ (Meager 1697, 110). But it also made good firewood, excellent charcoal, and clefts easily, making it particularly suitable for making floorboards and fencing, especially the stout fencing required for such things as deer park pales. It was also widely used by the joiner, while its bark was used in the process of tanning – an increasingly important consideration, as we have seen, in the eighteenth and nineteenth centuries. In Ellis’ words, oak ‘of all other Trees, claims the Priority of Regard in this Nation for its many transcendent uses’ (Ellis 1741, 18). It was, in addition, catholic in its habits or, as Mortimer put it in 1707, it ‘thrives best on the richest Clay, ‘tho it will grow well on moist Gravel or the coldest Clay’ (Mortimer 1707, 329). ‘It is a Tree that will grow and prosper in any sort of Land, either good or bad, as on Clayey, Sandy, or Gravelly Ground, whether dry or moist, warm or cold…’ (Blome 1686, 250). It will grow ‘in any indifferent Land, good or bad, as Clay, Gravel, Sand, mixed, or unmixed Soils, dry, cold, warm or moist’ (Meager 1697, 110).

Ash was rather different. While less durable and sturdy, so less useful as structural timber, it had a very wide range of uses and was greatly valued as fuel (Meager 1697, 11). It was ‘one of the most universal forms of Timber we have’, according to Mortimer (Mortimer 1707, 336). Timothy Nourse thought it ‘a most useful wood to the Coach-maker, Wheeler, Cooper, and a Number of other Artificers’, and that it had had numerous uses on the farm, for fencing and bins, ‘for Spittle and Spade Trees, for Drocks and Spindles for Ploughs, for Hoops, for Helves, and Staves, for all Tools of Husbandry, as being tough, smooth and light’, thought Timothy Nourse (Nourse 1699, 119). Its excellence as firewood was universally praised, burning well even when green. In Evelyn’s words, ‘the sweetest of our forest fuelling, and the fittest for ladies chambers’ (Evelyn 1664, 40); while to Moses Cook, ‘Of all the wood that I know, there is none burns so well green, as the Ash’ (Cook 1676, 76). Ash leaves and twigs were also avidly consumed by sheep and cattle, and dried well to form leafy hay. Hartlib put it simply: ‘Ash, for a hundred uses’ (Hartlib 1655, 80). And on top of this, ash was also very fast growing and, like oak, was not very choosy about where it grew. Although some early writers believed that it had some preferences, perhaps for ‘light dry Mould’, there was general

112 agreement that it would grow on ‘any sort of land’, provided ‘it be not too stiff, wet and boggy’, although in general it seems to have been a less prominent feature of farmland tree populations on the more acidic soils (Mortimer 1707, 366).

Elm in its various forms also had a multiplicity of uses. It was ‘proper for Water-works, Mills, Soles of Wheels, Pipes, Aquaducts, Ship Keels and Planks beneath the Water Line … Axel trees, Kerbs Coppers, … Chopping-Blocks … Dressers, and for Carvers work’, as well as for making spades, shovels and harrows. But above all it made excellent boards and planks, for floorboards, external weatherboarding and coffins (Nourse 1699, 115). Like oak and ash, moreover, elm could tolerate a wide range of conditions. While Ellis thought it preferred a ‘damp or wettish soil’, and Mortimer believed it ‘thrives best in rich black Mould’, and not so well on sands and gravels, the latter also conceded that it would grow ‘almost on any sort of Land’ (Ellis 1741, 46; Mortimer 1707). Seventeenth- and eighteenth-century writers singled out another advantage: it caused ‘the least offence to Corn, Pasture and Hedges of any Tree’, in part because (unlike ash) its roots did not spread far, but also because it could be rigorously trimmed up as a timber tree, so that it cast limited shade. Ellis, writing perhaps with his local landscape of the Hertfordshire Chilterns in mind, thought it a pity that ‘more Hedge-Rows are not planted with Elms; because … they don’t damage any thing about them, as some other Trees do, whose Heads must not be trimmed up as these may’ (Ellis 1741, 49). Meager however thought that it was a tree ‘chiefly planted in Hedge-rows’ (Meager 1697, 114).

More difficult to explain than the general popularity of these three species over others are the marked regional variations we have noted in their relative importance. Some have clear environmental determinants, such as the tiny quantities of elm recorded from Yorkshire. Most of the elm planted on farmland in England was Ulmus procera or Ulmus carpinifolia; the latter is restricted to the south and east of the country and even the former is rare in the north, apparently close to the edge of its native range there for climatic reasons. Wych elm (Ulmus glabra) will occur naturally much further north but only in restricted locations, shunning the more acid soils. It does not seem to have been favoured in any area as a hedgerow tree, however, probably because its deep shade adversely effected the shrubs growing beneath it, and the crops in adjacent fields. More difficult to understand is the manner in which in Northamptonshire, and apparently other counties in the champion Midlands (including the clayland fringes of west Norfolk), oak was often the second or third most important tree. This preference for ash and elm as a hedgerow tree was not only long lived, but evidently encouraged by landowners, with Northamptonshire leases often stipulating that tenants should ‘plant yearly four ash or elm’ on their holdings, or one ‘good or sufficient plant of ash’ each year (see e.g. NHRO HOLT 638). It is possible that oak did not flourish on the slowlydraining soils, formed in Jurassic clays, which cover much of the Midlands, as these are generally acidic in the plough-zone: elm would have put on better growth and ash, on the whole, would be more tolerant of such conditions. In addition, the popularity of ash and elm may reflect the fact that most of the information we have relates to either open-field parishes, or to ones which had been enclosed comparatively recently, so that much of the timber recorded had originated in an unenclosed environment. We have noted how the majority of farmland trees, other than willows, were concentrated in the tofts and crofts, on greens and along roadsides, within or around nucleated villages. It may well have been that oak, which develops a wide, dense crown at an early stage of growth, cast too much shade in such locations. Ash, with its more open crown, would have thrown rather less, and although many early writers commented on its tendency to rob the soil in the

113 adjacent fields of nutrients, and to spread its roots long distances, such disadvantages were presumably outweighed by its numerous uses, especially as fuel. Early felling would to some extent have mitigated such problems. Elm made a particularly good choice where trees needed to be tightly packed along the hedges surrounding small village closes, for the reasons stated earlier. Ellis commented that he ‘ never saw so many grow so large and flourish in so little room as these will, even almost close together’ (Ellis 1741, 46). On balance the secondary (and on occasions, tertiary) importance of oak was probably a combination of both environmental and economic/social factors: after enclosure oak certainly increased its importance, although never to the levels seen in the other areas studied.

‘Minority’ trees of farmland

Although the overwhelming majority of farmland hedgerow trees, in all areas, thus consisted of oak, ash and elm, a wide range of other species is recorded in seventeenth, eighteenth and early nineteenth-century sources, albeit at low levels. The most important were willows of various kinds, cherry, black poplar, aspen and walnut; but hornbeam, holly, maple, pear, sycamore, apple, beech, sweet chestnut, small-leafed lime and alder also make regular appearances in the records before the late eighteenth century. Of the main types of indigenous and naturalised tree, only wild service tree and birch appear to be absent, together with whitebeam, although this latter species is virtually unknown as a native tree within the areas studied. These presumably had little to recommend themselves to farmers and landowners. Birch for example is not very durable as timber, decays quite rapidly, and it burns quickly as firewood: while it makes reasonable charcoal, provides a good turning material, and grows rapidly, it offered little that other, more adaptable species could not provide – although we might note, perhaps, that its remarkably low frequency in the traditional landscape may in part have cultural determinants, reflecting its weed-like character and its tendency to invade abandoned land, and some early writers seem surprisingly prejudiced against it.

As already emphasised, the low representation of these various ‘minority trees’ in our records may to some extent be misleading, as it is clear that these were mainly concerned with valuing timber. They sometimes fail to record pollards at all and, if they do so, usually omit to detail their species. For the most part, however, the low level of representation is evidently real and must indicate that these trees had fewer uses than oak, ash or elm, grew more slowly, or were more demanding in their requirements. In some cases, however, it also reflects the fact that contemporaries thought that they were better managed as coppiced underwood in woods or hedges, rather than as timber trees or pollards. Maple, for example, was presumably common in the past, as it is today, as a shrub in hedges, where it seeds relatively easily. It is also - usually in combination with hazel and / or ash – a frequent component of the coppiced understorey in woods. Yet it is only sporadically recorded as a farmland tree. Ellis explained that its wood was brittle, light and soft, and that it was:

Not so profitable to burn as some are. They are sometimes made Pollards, but make a slow Return that Way: in standards they seem to do better, because they are not subject to those Evils that the Pollard is: for this being a soft Wood, is apt to let in the Wet after topping (Ellis 1741, 84).

For the most part, early writers thought that it was best managed as underwood, in woods or hedges. Many contemporaries perhaps shared the views of Mose Cook, who believed in addition that:

114

If you let it grow into Trees, it destroys the wood under it; for it leaves a clammy Honey-dew on its Leaves, which when it is washed off by Rains, and falls upon the Buds of those Trees under it, its Clamminess keeps those Buds from opening, and so by degrees it kills all the Wood under it; therefore suffer not high Trees or Pollards to grow in your Hedges, but fell them close to the Ground, and so it will thicken your Hedge, and not Spoil its Neighbours so much (Cook 1676, 99).

Maple does appear as a tree at low levels in accounts and tree surveys, as it had a range of specialised uses, for making dishes, spoons and ‘other curious Turner’s Ware’, and Langley added that it was also employed in making musical instruments, all of which required pieces of wood larger than would be provided by coppice poles. But it is more usually recorded as a pollard, as at Aldenham in Hertfordshire in 1803 or Kelshall in the same county in 1727, where 53 were recorded amongst the many trees on a property of less than 40 acres (16 hectares), 36 of them growing in one hedge of a small (4-acre) field (HALS D/EX 216 B2-4; HALS DE/Ha/B2112). In 1814 seven maple pollards (and one crab pollard) were felled on the Evans Lomb estate in central Norfolk (NRO HNR 149/3). It may have been a more common tree than the documentary sources suggest, as a number of former pollards can be found growing in hedges in Hertfordshire and Norfolk (Barnes and Williamson 2011, 76-7). But farmers and landowners generally preferred to manage it as underwood, where they wanted poles, and as a timber tree its uses and value were limited,

Black poplar and aspen are often noted in early surveys, if generally in small quantities, their distributions to an extent mutually exclusive. Although many early writers classified both as ‘aquaticks’, only the former was (and is) really a tree of damp ground. Both documentary sources, and surviving examples, suggest that it was mainly found on floodplains – it was a tree of meadows and marshes – although it also grew to some extent on clay soils, in places where the drainage was poor (Cooper 2006, 9-12). On the Broxbourne Mill estate in east Hertfordshire in the late eighteenth century poplars made up only 4 percent of the trees recorded in the ‘inclosure land’ on the boulder clays, but 43 percent of those in the ‘enclosed meadows’ on the flood plain of the river Lea (HALS B479). Today black poplars are most commonly found in the far north west of the county, the edge of a wider population centred on the Vale of Aylesbury in Buckinghamshire (Figure 25): quite why so many were planted here, apparently in the seventeenth, eighteenth and even early nineteenth century, to judge from their sizes, remains unclear. Aspen in contrast was most numerous on slightly acid loams and clays. It was thus almost never recorded on the boulder clays of east Hertfordshire or south Norfolk, but was present at low levels on the clay-with-flints of west Hertfordshire, and in places on the brickearths and loams of north east Norfolk. Ellis believed that it did best in ‘low, moist places’ but noted that it also grew reasonably well in dry locations, in his own hedges and ‘on our high Common’, where it seemed to thrive on a ‘foot of Mould’ above clay (Ellis 1741, 96). Both trees grow with remarkable speed, and their wood is light and easily worked but tough, with a similar range of uses, especially for making boards for flooring or for the exterior of buildings: they might also be employed for such things as chair frames (Cooper 2006, 24-8). Aspen in particular was used to make wagon bottoms, arrow shafts, oars and paddles, and clogs, and charcoal for gun powder, while its bark could be used for tanning. Both woods, moreover, had mildly fire-resistant qualities, so that they were often used in mills and other industrial premises – especially as malt house floors. Indeed, a survey of surviving and historically documented examples in Norfolk suggests that over a third stand beside the sites of mills, smithies, kilns and malthouses, where they were presumably panted to provide material for repairing floors, doors and the like (Barnes et al.. 2009). Neither wood

115 thus makes good fuel although both – and especially black poplar – were sometimes managed as pollards, for reasons that remain unclear. It was this feature, and a relatively limited ranges of other uses, more than any environmental constraints that probably ensured that neither tree was widely planted, and even in waterlogged locations both were generally outnumbered by willow.

Figure 25. Pollarded black poplar near Tring in north-west Hertfordshire.

Other ‘minority’ farmland trees display similar characteristics: they had little to offer that oak, ash and elm could not provide more easily or in greater abundance; had more demanding requirements; or were better grown in other contexts, as underwood. While they had speciality uses, these did not necessitate abundant plantings by farmers or landowners. Sycamore, for example, will grow in most situations and once established in a locality will seed freely. Some have argued that, in parts of northern and western Britain, sycamore is a native species (Green 2005) but if so it is surprising that in Yorkshire, where it is often described in documents as ‘plane’, it never exceeds one percent of the total numbers of trees recorded in valuations, surveys and sales documents from before the nineteenth century, and does not seem to appear at all in documents before the middle of the eighteenth. One of the earliest records is from Healaugh in 1757, where eight specimens were recorded on the farm of John Adcock (Lds WYL68/4). In two years (1786-7) of the Duncombe Estates wood accounts, comprising c.400 entries, sycamore is mentioned only once (Nrth ZEW IV 6/4/6); even in the five years of timber accounts for Newby Hall (1829-34), comprising c. 275 entries, it is only mentioned three times (Lds WYL5013/2234). Further south, in Northamptonshire, it is recorded in the timber accounts for the Montagu estate from 1749, but again only at very low levels (less than two percent) (Boughton House Archives). No examples at all are mentioned in a survey of over 3,000 trees on the farmland of the Duke of Powys’ estates in Upper and Nether Heyford, Glassthorpe and Newbold in 1758; nor, more strikingly (for the survey is a detailed one) amongst the 4,500 trees recorded on John Darker’s estates in Gayton, Tiffield, Kislingbury, Milton, Litchborough

116 and Upper Heyford in 1791 (NHRO ZB 1837; NHRO YZ 2183). Only in the course of the nineteenth century do more references appear, although even then mainly to trees growing in woods and plantations, rather than on farmland - and even then, perhaps, mainly in gardens and tofts in the immediate vicinity of villages. Of the 983 trees recorded in the valuations of timber associated with the enclosure of Wilby in 1806, a mere four were sycamores. In Norfolk, sycamore appears as a tree of gardens, parks and plantations in the middle decades of the eighteenth century (at Hunstanton in the north-west of the county, for example: Williamson 1998, 253), but does not figure as a farmland tree before the nineteenth century. Not a single specimen was recorded amongst the 24,000 hedgerow trees on the Earle estates in 1722 (NRO BUL 11/283, 617X2) and the earliest farmland reference is apparently from Reepham in 1813, where it constituted a mere 1.4 percent of the recorded trees (NRO MC 687/29-31 813 x 3). Similarly, in Hertfordshire sycamore only appears as a farmland tree as late as 1817, on a farm near Hemel Hempstead (HALS D/EB 1622 E9). It is noticeable that many seventeenth and eighteenth-century writers on farming and forestry either fail to mention sycamore at all, or else regard it as an ornamental tree, of gardens and parkland. Indeed, while some nineteenth-century writers do recommend that it should be planted as a hedgerow tree they usually added that this was only the case ‘if ornament be studied’ (Nicol 1799, 249). Ellis and Langley both describe how its wood and timber could be used in similar ways to those of maple, but it was of the ‘soft, woody tribe’, and made only modest fuel. (Ellis 1741, 87; Langley 153-5) There was probably little point in planting it in preference, in particular, to ash, to which several writers compared it, finding it inferior, if by no great margin.

Walnut, an early introduction but unable to propagate itself effectively in the wild, was more highly valued by elite writers, especially as a furniture wood, before the large-scale import of mahogany began in the early eighteenth century. But it did not produce very good firewood and was thought by many to damage the hedges it grew in. Ellis considered its drip ‘pernicious’ and a hindrance to the growth of the hedge plants beneath it. He also warned that ‘boys and others’ would cause damage to hedges while trying to steal the fruit (Ellis 1741, 63). Nevertheless, it is regularly recorded as a farmland tree in eighteenth-century surveys, if at low frequencies, especially in Northamptonshire. On John Darker’s estates in the centre of the county in 1791 walnut was the fifth most common species recorded, after ash, oak, elm and poplar, although in total accounting for only 0.3 percent of the trees on the estate (NHRO ZB 1837). On the Boughton estate it was present at similar levels as a hedgerow tree in 1749 (Boughton House archives); but it was occasionally more prominent, as on a farm at Teeton in 1812 where it made up nearly six percent of the trees (Boughton House archives; NHRO B(H) 461). It was also present in Hertfordshire and Norfolk but again at low levels, making up only 44 of the 24,000 trees on the Earle estates in north east Norfolk in 1722, for example (NRO BUL 11/283, 617X2).

Several other ’minority’ trees are recorded, if only in tiny quantities, on farmland in seventeenth and eighteenth-century surveys. Small-leafed lime appears in a handful of records from Norfolk, three pollarded examples being recorded out of the 24,000 trees described on the Earle estates in the north east of the county in 1722. (NRO BUL 11/283, 617X2). Holly is a common feature of Yorkshire place names, principally in the form of the element 'hollins', probably reflecting its importance as a fodder crop prior the nineteenth century (Spray 1981), although it is only rarely mentioned in valuations and sales contracts, perhaps because it was regarded as underwood as opposed to timber. There are far fewer holly-based place names in the south of the country, and on the rare occasions that holly is mentioned in documents, it tends to be confined to the more acid soils. In

117 other counties studied it is present at even lower levels, although again with a clear association with the more acidic soils. It thus made up a mere 0.4 percent of the trees noted on a farm at Redboune in west Hertfordshire, and 3 percent at London Colney in 1813, but is not recorded as a timber tree elsewhere in that county (HALS D/E Cr 55; HALS D/EB 944 E1-2 & P1). In Norfolk it appears sporadically in records from the mildly acidic loams in the north east of the county, with a frequency ranging from 1 percent at Reepham in 1813 to a tiny 0.004 percent (one out of 24,000 trees) on the Earle estates in 1722 (NRO BUL 11/283, 617X2; NRO MC 687/29-31 813 x 3).

Some species were, like maple, more useful and valuable than those we have just described but, for a variety of reasons, were generally considered to be better grown as coppice in woodland, although they are sporadically recorded growing on farmland. Alder, for example, was much valued for its durability in waterlogged conditions, and was thus widely used for jetties and river bank revetments. It was also used for scaffolding poles – when Blickling Hall in Norfolk was constructed in the early seventeenth century, payments for alder poles are amongst the first to appear in the building accounts (NRO MC3/47) – but was perhaps most important as a source of charcoal. In addition, its bark could be used for tanning and its wood was in demand for a range of specialised uses, including clog making. While mainly grown in woods or ‘carrs’ on damp ground, or as a component of the coppice within wet areas in woods on the boulder clays, alder is also often recorded – in Yorkshire and Norfolk especially – as a free-standing tree in meadows, and growing in the hedges around them. On a farm in Scarning in the latter county in 1764 it constituted less than 1 percent of all trees, but at Brick Kiln Farm at Reepham in 1813, much of which lay on flood plain land, it made up 38 percent of the total population (NRO BCH 20; NRO MC 687/29-31 813 x 3). Beech and hornbeam were even more economically important but, while present in hedges at low frequencies (especially in Hertfordshire and Norfolk) were generally considered a tree of wood-pastures, and woodland, and are discussed in these contexts below.

There are only two species, other than oak, elm and ash, which were locally dominant as pollards or timber on farmland within the areas studied. Firstly, as already noted, willows of various kinds – probably mainly white (Salix alba) and crack (Salix fragilis) – were found on alluvial soils near rivers in all the areas studied. Neither documentary sources, nor many contemporary writers, clearly distinguish the various species, which do indeed hybridise to some extent. In Northamptonshire willows were present in vast numbers on the floodplains of the Nene, Ise and Welland and their tributaries, although they were also to be found more widely, scattered through the open fields, on the damper soils: Ellis emphasised that both the ‘willow’, and the ‘withy’ – by which he seems to have meant osier or Salix viminalis, which only grows to shrub height – would tolerate ‘clayey or loamy grounds’, as well as thriving in marshy conditions (Ellis 1741, 105-6). Surveys from Northamptonshire made in association parliamentary enclosures show how numerous they could be in unenclosed parishes. That for the parish of Irthlingborough beside the river Nene, still 90 percent open when enclosed in 1808, records a total of 3,055 trees, no less than 62 percent of which were willows of some kind; another, drawn up on the eve of the enclosure of Finedon in 1806, similarly almost entirely unenclosed, recorded 1,038 trees, no less than 410 of which were ‘willows’; while a valuation of the timber growing at Wilby in 1764, lists 762 trees, nearly a third of which were similarly described (NHRO ZA898; NHRO X1657; NHRO ZA 906). Large numbers of willows appears to have been a characteristic feature of Midland ‘champion’ landscapes more widely. Ellis described how they were common ‘either in standards, pollards or in hedges’ in the open-field districts of north Hertfordshire, around Baldock; and commented that the loppings were ‘of such great use, that

118

I have known it the only Wood they have in some parts of Rutland in their Open Fields’ (Ellis 1741, 106). The poles from pollards were used, into the twentieth century, for hurdles and rough fencing, large amounts of which were required to fold sheep and otherwise control stock in the largely hedgeless landscape of the open fields. Willow poles also, according to Mortimer, made good stakes and tool handles, as well as faggots for burning, while larger pieces were used by turners (Mortimer 1707, 364). Most willows were cut on a short rotation, of three to four years, ‘just before Winter, or in the Spring’ (Ellis 1741, 106).

The second example of a tree species, or range of species, which records show challenging the dominance of oak, ash and elm comes from the Chiltern dipslope in Hertfordshire, where cherry and, to a lesser extent, apple were widely planted as hedgerow trees. An undated late seventeenthcentury map of a c.10-hectare farm in Flaunden details 27 oaks, 18 elms and 9 ash growing in the hedges, but as many as 15 ‘asps’ (aspen), all outnumbered by 59 apple trees and no less than 165 cherries (Figure 26: HALS DE/X905/P1). John Norden, writing in 1608, bemoaned the fact that fruit trees were gradually disappearing from the hedges of Hertfordshire, as the modern generation failed to replace those which had grown old and died, although Moses Cook, writing of the same county, was still advocating the planting of apple in hedges in 1676 (Norden 1608, 201; Cook 1676, 138). William Ellis, who lived at Little Gaddesden in west Hertfordshire, wrote as if hedgerow cherry trees were a normal feature of the local landscape. Most references are presumably to the native Gean, but Ellis and others make it clear that cultivated varieties were sometimes grafted onto field trees. The situation is slightly confused, however, because both Cook and Ellis describe the wild Hertfordshire cherry as the ‘Black cherry’, although that term is now applied to Prunus serotina, introduced from north America in 1629. Fruit trees, and especially cherries, were an important feature of the hedges of other districts: Ellis noted that they were common in those areas of the Buckinghamshire Chilterns lying across the county boundary from Hertfordshire, and also in Kent; while Norden commented on the abundance of fruit trees more generally in the hedges in Devon, Gloucestershire, Kent, Shropshire, Somerset and Worcestershire, as well as in many parts of Wales.

Within the areas studied both cherry and apple appear sporadically in records from Yorkshire to Norfolk. Cherry was apparently planted for its wood as well as its fruit. Ellis thought that, as well as the money that farmers could derive from cherries, which was usefully harvested between the hay and grain harvests, they could also benefit from the timber, which he considered to be comparable in durability and strength to that of oak, although modern foresters note its variable quality in this regard (Ellis 1741, 65-6). Moses Cook – who lived, like Ellis, in west Hertfordshire - suggested that:

Many will say, that it is not proper to rank this among Forest-Trees; but if such did but see the fine stately Trees that we have growing in the Woods at Cashiobury, they would then conclude it proper for Wood; and if for Woods, then for Forests (Cook 1676, 92).

Documentary sources confirm this dual importance. In 1805, for example, there was a dispute at Hackford in Norfolk concerning the illegal felling on copyhold land of ‘969 Timber trees of different descriptions consisting of oak, ash, elm, walnut and cherry trees …..’ (NRO AYL 900); while the valuation of trees made at the enclosure of Aldenham in south Hertfordshire (D/EX 216 B2-4) makes an explicit distinction between ‘cherry’ and ‘cherry timber’. It does not appear in records as a pollard, probably because its resinous character ensured that it made poor firewood.

119

Figure 26. Detail from an undated, probably late seventeenth-century map of a farm in Flaunden, west Hertfordshire. As well as containing oak and ash, the hedges are liberally filled with fruit trees and ‘asps’ (aspen).

To an extent, the prominence of cherry as a hedge tree in districts like the Hertfordshire Chilterns was a function of natural factors. Moses Cook noted in 1676 how it was common in the woods of west Hertfordshire, and seeded spontaneously in hedges, while Ellis noted that while it grew well in

120 the ‘loamy ground’ of the Chilterns, on soils formed in fairly porous clay over chalk, it was not found on the heavy clay soils of the Vale of Aylesbury in Buckinghamshire, immediately to the north. Tree surveys show that it was similarly absent from the boulder clay soils of east Hertfordshire, and in Norfolk there was a similar contrast, with cherry recorded, albeit at very low levels, on the slightly acid loams in the north east of the county, but hardly ever on the boulder clay soils in the south and centre. The importance of cherry trees in Chilterns hedges may thus result in part from the management of naturally seeding specimens, although as noted Ellis suggests that many were then grafted, and some of the documentary evidence certainly implies that many trees were naturally recruited. In Aldenham in 1805, beside the ‘cherry’ and ‘timber cherry’ recorded, there were also ‘cherry saplings’ (HALS D/EX 216 B2-4). This said, naturally seeding plants could simply have been laid, with the other hedges shrubs – treated in the same way that maple generally was – so in the Chilterns, as in Kent, their presence as trees was a conscious choice reflecting the proximity of London, with its insatiable appetite for fresh fruit. By the nineteenth both areas boasted numerous commercial orchards, supplying the capital with both apples and cherries.

The majority of the evidence on which this discussion is based, it must be emphasised, comes from documents, maps and published sources dating from the late seventeenth century onwards, and it is possible that the various ‘minority’ trees were more common in earlier periods, and that the dominance of oak, ash and elm on farmland correspondingly less. Rackham has pointed out that black poplar is more frequently mentioned in sixteenth-century and earlier sources than in those from the seventeenth century and later, and that its functions (as a source of light board wood in particular) may have been eroded by the increasing availability of pine (Rackham 1986, 207). Fruit trees, especially cherry, may also have been more widely planted in earlier centuries, to judge from the comments of Norden about their loss from Hertfordshire and Middlesex, noted above. Either way, what is particularly striking is the way in which the character of these ‘minority’ farmland trees displayed a significant degree of regional variation. In Hertfordshire, for example, there were significant differences between the west of the county – the Chiltern dipslope, with soils largely formed in clay-with-flints and outwash gravels; and the east, with soils mainly formed in boulder clay. Cherry was found as a hedgerow tree in the west, together with aspen and some beech. All three were virtually unknown on farmland in the east, where instead maple is recorded, with small quantities of hornbeam, and accompanied by black poplar in damper areas. In Norfolk likewise there were subtle differences between the boulder clays in the south and the lighter, slightly acid loams of the north-east. Hornbeam and maple were the most important ‘minority trees in the former district, to judge from both documentary sources and surviving examples, but in the latter hornbeam was unknown, maple was rare and instead aspen, apple, walnut and cherry were the characteristic ‘minority’ trees. Beech were also found on some farms in north east Norfolk, and is arguably indigenous here – areas of relict wood-pasture at Felbrigg boast large old examples of beech trees which have been interpreted as remains of an isolated natural population although, as we shall see, there are other interpretations. Either way, a survey of the timber on the Blickling estate, made in 1756, explicitly excluded ‘all trees under 6 inches Girth and also all Firrs Beech Pollards …; while a sale of trees at and Hempstead in 1789 included, besides 618 oak, 174 ash, 4 elm and a chestnut, 25 beech trees (NRO MC 3/252; NRO MC 688/11813). Similar regional variations in the composition of ‘minority’ populations are apparent in Yorkshire, with beech, maple, alder and sycamore recorded at low levels in the south of the county, but with the latter only regularly appearing elsewhere; while, as we have seen, large numbers of willows appear to have been a

121 characteristic feature of the champion Midlands. Such patterns were the consequence of soils and climate but also, as we noted in the case of the Chiltern cherries, of economic and social factors.

Trees in woods and wood-pastures

There were often significant differences between the balance of tree species grown as standards (as opposed to those cut as coppiced underwood) in woodland, and those found as pollards or timber on neighbouring farmland. Although in all areas elm and ash could be found as woodland timber these were – certainly by the seventeenth and eighteenth centuries – much less important than oak. Oaks were the only trees sold out of Melton Wood in in Norfolk in 1654 and 1656, and the only ones recorded growing in Honeypot Wood, Wendling, in 1777 (924 in all, or 92 per hectare, a surprising density – together with 57 pollards, presumably growing on the boundary bank) (NRO HNR 135/3/18; NRO NAS1/1/14/ 7-9). In Helmsley in Yorkshire in 1642 the woods were composed almost exclusively of oak, with only 5 percent of the recorded timber being ash and with no elm (Nrth ZEW IV 1/6). In Northamptonshire the contrast in this respect with the surrounding farmland, dominated by ash and elm, was particularly striking. Documentary records confirm the comments of eighteenth-century writers like Pitt that woods in the county were overwhelmingly stocked with oak timber (Pitt 1813). On the Montagu estates in 1749, for example, no less than 99.8 percent of the 5,629 timber trees within Boughton Wood were oaks; at nearby Prist Coppice the figure was 99.9 percent , as it was at Coppingford Wood (Boughton House archives). The essentially unnatural dominance of oak in the local woods is clear from contemporary comments about how it should be maintained. Thus in 1726 Daniel Eaton informed the Earl of Cardigan that if he did not take his advice for the management of woods at Corby in Northamptonshire, ‘of planting acorns in the vacancies, in two sales more there will not be an oak left’ (Wake and Webster 1971, 114).

In some woods, of course, the dominance of oaks was less. The trees felled from Langley Wood in Norfolk in 1671 consisted on 126 oaks and 54 ash (NRO NRS 1126, 25E5); while those sold out of Curples Wood in 1771 comprised 300 oak, but only 40 each of ash and elm (DN/EST 52/22). Amongst the timber trees valued in the woods on the Stow Bardolph estate in west Norfolk, surveyed at various times between 1811 and 1815, 1,420 were oak but ash came a close second (1,291) and elm was surprisingly prominent (316) (NRO HARE 5499, 223 x 1). In the Hertfordshire Chilterns, and on the Chiltern dipslope, oak was again usually dominant, although ash and elm were usually present and, in some woods, small amounts of cherry and aspen; but beech appears to have increased its importance over time, often leading by the nineteenth century to the suppression of the understorey and the management of mixed oak-beech woods on a ‘selection forestry’ system, largely to supply the local furniture industry, or to wood composed of beech alone. No less than 89 percent of the timber sold from a wood in Tring in c.1880, and 74 percent from its neighbour, were beech, the remaining trees comprising ash alone in the latter case and ash and cherry in the former (HALS DE/Bn (Add) B13). Similarly, the trees felled from Oaken Grove in Rickmansworth in 1859 comprised 29 beech, two ash but no oak. Even in the nineteenth century, however, there was much variety (DE/Bn (Add) B19-20). Pancake Wood, on the county boundary near Ashley Green, contained 74 percent oak, 15 percent elm, 4 percent ash but only 3.5 percent beech timber in 1838, together with small quantities of cherry and aspen (HALS D/E Cr 55). It is true that all these references to beech as a prominent timber are from nineteenth-century documents, but Ellis in 1741 was already noting the way that beech was replacing oak in woods where both were present. After felling, ‘a Wood of Beech has spontaneously succeeded; and when this has once got Dominion, it will be sure always to remain Master’ (Ellis 1741, 26).

122

The available documents seldom describe the composition of the underwood in coppiced woodland, but they often provide some hints to its character when detailing the uses to which it was put. It was considerably more diverse than that of the timber, displaying clear local and regional patterns of variation which seem to have been only partly related to soils, climate and topography. In Norfolk the majority of woods were found on boulder clay soils, as were many in east Hertfordshire, while the vast majority of those in Northamptonshire, lying within the three forests of Rockingham, Salcey and Whittlewood, were likewise located on boulder clay uplands. Where these woods still survive, and have not been replanted with conifers, a high proportion of the coppice (now usually neglected and outgrown) is made up of a mixture of ash, hazel and maple, with smaller amounts of other species like elm, although in parts of Hertfordshire and Norfolk hornbeam is also prominent. On more acid or loamy soils rather different species are found. In the west of Hertfordshire, on the dipslope of the Chilterns, hornbeam, ash and maple are again present but hazel is usually more important and is accompanied in places by cherry and, more rarely, aspen. On the poor acid soils in the south of the same county, formed in London clay and pebble gravel, hornbeam is normally the dominant species, most notably in the great woods which still cover many hectares in the Wormley, Hoddesdon and Broxbourne areas. In parts of north Norfolk, on particularly sandy and gravelly soils, the coppice is sometimes dominated by oak, as in Edgefield Little Wood or Great Wood.

In Yorkshire coppices there tends to be a greater prominence of oak and ash. Generally speaking, birch is found alongside oak, with both favouring drier, more acidic soils. Ash, more dominant on calcareous soils and therefore the limestone areas, is frequently accompanied by alder, particularly on wetter ground. Hazel is found in woods throughout the county, and species such as hawthorn, crab apple, rowan and cherry are all locally significant, together with alder, the latter especially important on the millstone grits of South Yorkshire. Sycamore is also found as underwood in places, in contrast to the situation in the south, and this appears to be a long-established characteristic: Gledhill (1994, 245) has noted that that oak, ash, elm and sycamore were all found in the woods at Downholme in the early nineteenth century, while 35 sycamore poles were recorded in a timber sale at Shipley in 1835 (Lds WYL639/339). All these were, and are, very broad patterns, however, and there was and is much local variation. In Norfolk for example a handful of woods are known in which lime is a significant component of the underwood, mainly in the centre and north of the county, such as and Swanton Novers Great Wood; while Wayland Wood in Watton, and a few other woods in the centre-west, are characterised by coppice featuring large amounts of bird cherry.

Two questions arise about these patterns: how far is the present composition of the coppice a good guide to its character in the past (i.e., how stable has the character of the underwood been over time)?; and, connected with this, are the species present in any particular wood, or group of woods, largely a function of environmental factors, or has the underwood been significantly modified by deliberate choice, or as a side-effect of management? In this context, it is important to emphasise that the three species most frequently found as coppice– ash, oak and hazel – produced material of particular practical and economic value. Hazel made excellent hurdles, provided wattle-and-daub for timber framed buildings, thatching broaches, hedging stakes, and firewood in the form of faggots; ash poles were extremely useful, as we have noted, for fencing, hurdles, tool handles, minor building timber, as well as making first rate firewood; while oak also made good firewood, as well as being the best fencing and building material. In addition, the bark of the larger poles from oak coppice could be stripped and sold to tanners. Other kinds of underwood species, of more local importance,

123 are also distinguished by their utility. Hornbeam, for example, produces excellent charcoal and first rate firewood. It would not be surprising if foresters had managed the understorey to encourage these particular species, perhaps removing less valuable shrubs which might compete with them, such as guilder rose or dogwood.

Many coppiced woods were newly established or significantly extended in the course of the post- medieval period – one on the Boughton estate in Northamptonshire was said in the 1760s to have been ‘lately added to .. and is not yet come into cutting’ – and in such cases the coppice must have been deliberately planted with appropriate species (Boughton House archives). Where the archaeological evidence indicates that particular woods have expanded at the expense of farmland or common land, the understorey of the extension is sometimes still significantly different to that of the original ‘core’. The ‘primary’ portion of Wayland Wood in Norfolk, for example, has an understorey characterised by a mixture of ash, hazel and bird cherry. That within the later ‘extension’ to the east, in contrast, probably added in the fifteenth century, mainly comprises hornbeam and ash (Barnes and Williamson 2015, 240-3). The narrow addition made at some unknown date to the western edge of Shotesham Little Wood in the same county has an understorey almost exclusively composed of coppiced hazel, unlike the rest of the wood, which mainly comprises hornbeam, mixed with varying amounts of ash and hazel (Barnes and Williamson 2015, 226-8). Even where a coppiced wood is entirely of ‘primary’ character, moreover, we should not assume that the underwood represents, in any simple and direct manner, a modification of the vegetation which was present on the site when the wood was first enclosed from the wider wooded wastes in the early Middle Ages – not least because, being intensively grazed, these must often have lacked a significant understorey. Nor should we assume that it remained unaltered thereafter. A lease for South Haw wood in Wood Dalling in Norfolk, drawn up in 1612, bound the lessee to plant sallows in cleared spaces following felling (NRO BUL 2/3, 604X7); the tithe files of 1836 describe how there were 35 acres of coppice wood in Buckenham in the same county, ‘part of which has been newly planted with hazel’ (IR 29/5816); while ash and sallow were planted in what appears to have been an existing wood near Boughton in 1752 (Boughton House archives). It is hard to believe that the prominence, in many woods in south Yorkshire, of oak as underwood was entirely the consequence of natural factors, and unrelated to the scale of local demand for bark and charcoal. As explained below, the prominence of hornbeam in many woods in Hertfordshire and Norfolk woods does not reflect its status in the ‘natural’ environment and is probably related, primarily, to economic factors.

The composition of the underwood may also have been changed as a side effect of management. The Crown woods were surveyed in 1564 by John Houghton (British Museum Add. MS 34 214) and by Roger Taverner in 1565 (TNA/PRO LRRO 5139). These surveys show that at this stage almost all the standards within Rockingham Forest were oak and that the underwood was dominated by 'thorn' (presumably hawthorn and blackthorn), with significant quantities of maple and hazel, and some ash, sallow and oak (Peterken 1976). Just over two centuries later the description provided by Donaldson in 1794 suggests that any changes had been marginal, for he notes that the underwood ‘principally consists of black and white thorn, ash, sallow, maple, and a small proportion of hazel’. Today, in surviving coppices, it is generally characterised by a mixture of hazel, ash and maple, similar to that found in woods on the boulder clay soils of Hertfordshire and Norfolk, with localised dominance of elm, oak, sycamore and willow (Salix caprea), and a scatter of other species, including alder, hornbeam and small-leafed lime. There is little thorn. Over the last two centuries hazel has

124 thus increased in importance and thorn declined and this may be the result of a decline in the intensity of grazing (Peterken (1976). Even before enclosure and disafforestation, the alienation of many of the coppices into the hands of local landowners seems to have reduced the extent to which they were grazed by commoners out of season; but it was only following disafforestation at the start of the nineteenth century that grazing was effectively curtailed.

Economics, choice and management may have had an even greater impact on the composition of the wood-pastures which, as we have described, remained extensive in many districts well into the eighteenth century. In Midland forests most of the trees growing on the plains and ridings were oak and ash, a subtle contrast to the situation on the surrounding farmland, where elm was more prominent. But more striking differences could be found in Norfolk and, in particular, Hertfordshire. In the Chiltern Hills in west Hertfordshire, in parks and to some extent on common land, beech appears to have been a prominent tree, although it formed only a small proportion of the timber growing in hedges in the surrounding countryside, where oak was the most common tree. Beech features prominently in the medieval timber records for Berkhamsted park, and in 1353, in order to obtain material for the pale, the Constable of the manor was obliged to sell beech trees to the value of £20, using the money to buy oak for the purpose – clearly suggesting that there were few examples of the latter species growing within the park (Whynbrow 1934). The manor of Caddington contained a ‘great beechwood of 300 acres’, probably a wood-pasture, in the early thirteenth century. William Ellis reported that beech was common along the whole length of the Chiltern Hills and their south western continuation, from Dunstable as far as Wallingford (Ellis 1741, 25-6). But it was also abundant in wood-pastures on the long dipslope of the hills, especially in deer parks. There were 1,420 growing in the park at King’s Langley in the mid-sixteenth century, far outnumbering the 300 ash and 180 oaks recorded (TNA E315/391); while large quantities were sold from the deer park at Knebworth in the early fifteenth century (HALS K117 and 119). It has been suggested that beech became more prominent in the district in the course of the eighteenth century, and that earlier the local woodlands had contained significant quantities of oak (Rackham 2006, 363-4): but as these examples indicate, beech appears to have dominated many of the grazed woodlands of the district since the fourteenth century: it was in enclosed coppices that beech really increased its representation in the course of the post-medieval period.

Beech was evidently well suited to the local soils but its dominance of local wood-pastures was probably at least in part a consequence of deliberate planting and systems of management. Ellis certainly believed that its importance in the district was due to planting. Its suitability for the chalk soils had ‘obliged our Fore-fathers, as well as those of the present Age, to set the Sides of their chalky Hills &c. with Beech-mast, where this tree will run up to a vast height with great Expedition ‘ (Ellis 1741, 26). The vast numbers present in the park at King’s Langley, certainly, cannot have developed from a wild population. The park was not cut out of wooded ‘wastes’ but was created around 1290 at the expense of ‘eight acres of meadow which used to be mowed before deer were placed therein … And one hundred and twenty acres of land which were arable’, to which a further 160 acres ‘of arable land’ were added in 1397 and more in subsequent decades (Munby 1977, 152). It is noteworthy in this context that, by late medieval times, beech was also an important tree of wood-pastures – especially parks - elsewhere in Hertfordshire, in areas in which it is now virtually unknown, most notably on the poor, acidic ground on the London Clay uplands in the south of the county. Beech timber worth 68 shillings was thus sold from the Great and Little parks at Hatfield in 1428, and a combined total of 12,000 oak and beech trees was recorded there in a survey of 1538

125

(HHA Manor Papers: Summaries I, 339; Page 1902, 99). A further survey, in 1626, recorded twice as many beech (520 trees) as oak (227) in Hatfield Great Wood, the former Great Park; while large amounts of beech timber were sold from the New Park (within the Great Wood), including four hundred in 1629 alone (Austin 1995, 6–7; HHA Sals. General 44/1 and Sals. Legal 66/7). To the south of Hatfield, beech comprised about 20 percent of the trees growing in the woods, probably wood- pastures, on the manor of North Mymms in the early fifteenth century; and it was a major component (together with oak and hornbeam) in the woods in Broxbourne and Hoddesdon (HALS 80226; Page 1912, 430). It is unclear when, or why, beech declined in this district but it can evidently still thrive there, for recent surveys show that ‘on acidic gravels in the south of the county it has been until recently increasing steadily at the expense of oak’ (James 2009, 112).

Beech was generally accounted a useful timber by seventeenth and eighteenth-century writers. It was used to make floor boards and external weather-boards, furniture and threshing floors, while its resistance to decay in waterlogged conditions meant it was much sought after for the construction of mills and ships. Mortimer thought that it was good for making trenchers, shovels and spades, and that it made excellent firewood (Mortimer 1707, 338-9). None of this explains its particular association with wood-pastures, especially parks. Ellis, however, noted that ‘The Beech excels all other Trees in Parks &c. for the Returns it makes of prodigious Quantities of sweet, healthful Mast, which greatly helps to subsist the red and fallow Deer’ (Ellis 1738, 41). Moreover, it is equal only to hornbeam and oak in its ability to tolerate intensive browsing (Rackham 2006). This, above all, must explain why it was so important, in parks and on commons.

The other great wood-pasture tree of Hertfordshire was the hornbeam. This was usually the dominant tree on the commons found on the London clays and pebble gravels in the south of the county, and in many deer parks not only in this district but more widely across the east. It was also important on commons and in forests in the adjacent parts of Essex. Yet it appears to have been poorly represented in the prehistoric landscape, to judge from pollen evidence, and does not seem to feature much in documents before the second half of the fourteenth century, when it is referred to in the accounts for a number of parks (Rowe and Williamson 2013, 156-9). In the fifteenth century the bailiffs of the parks at Great Munden and Knebworth recorded receipts for ‘trees called hornbeam’, as though the name was unfamiliar to them (Rowe 2009, 34). While oak, ash, elm, hazel, beech and maple all occur as elements in pre-Conquest placenames, hornbeam does not: probably the earliest example of its use in this way is Hornbeamgate, a hamlet in Hatfield in Hertfordshire, first recorded as late as 1366 (Mawer and Stenton 1938, 224). Hornbeam thus only appears to have become widely established in the wood-pastures of the district during the Middle Ages, and there are indications that its importance increased still further thereafter. In 1538 Hatfield Great Park in Hertfordshire was said to contain 2,000 oak and beech, but no hornbeams are mentioned. By 1626, however, the Great Park – now described as the Great Wood - contained 5,227 hornbeams ‘of all sorts’, ten times the number of beech, the next most numerous species (Austin 2013, 143). By this time, huge numbers were present on the local commons, such as the 24,000 recorded on Cheshunt Common in 1695 (Rowe and Williamson 2013, 156-7). As already noted, hornbeam is a also major component of the coppice understorey of the woods of south Hertfordshire, and also of those of south Norfolk and north Suffolk, where the contrast with its representation in the prehistoric pollen record is again stark. The pollen sequence from Diss Mere, close to the centre of the tree’s East Anglian distribution, indicates that it was only present in the area from the late Iron Age or Roman periods, and even then only at very low frequencies. It only became a significant component of the

126 local vegetation some time in the medieval period, the precise date being uncertain as the pollen sequence is itself not radiometrically dated (Peglar et al 1989).

It is hard to account for hornbeam’s rather late rise to prominence unless it was actively planted or otherwise encouraged, in both woods and wood-pastures, and Anne Rowe has recently argued that it was being deliberately established on common land as late as the seventeenth century, in vast numbers, by manorial lords (Rowe 2015, 310-14). In fact, as late as 1742 William Ellis commented on how, in Hertfordshire, ‘This (the hornbeam) is in great reputation for both copsehedge and wood and is planted in many parts, but more abundantly about Whethamsted in this County’ (Ellis 1742, 71). It was presumably chosen in part because of its resistance to grazing pressure, which is considerable. Its ubiquity may thus have owed less to ‘natural’ factors than to economic ones. Hornbeam had a range of specialised uses which reflect its name – the ‘hard beam’. Langley thought that it was of ‘great Use to the Mill-wright, for Coggs to his Wheels, as well as to the Turner, Carpenter and Joiner; and for Fire-wood there’s none better’ (Langley 1728, 161-2). Ellis similarly considered that it ‘so far excels most other Fire-woods, that when it is burnt enough, the Coals will hold a bright Fire like Charcoal for a long time’ (Ellis 1741, 72). We have already noted the importance of the charcoal industry in Hertfordshire. It is noteworthy that the two main concentrations of hornbeam in England – in south Norfolk/north Suffolk, and south Hertfordshire and south Essex - are close to the country’s two largest medieval cities, Norwich and London, both hungry for firewood and charcoal (Barnes and Williamson 2015, 80-85). The late spread of hornbeam into the woods and wood-pastures of these areas may have been the consequence of ‘natural’ factors and processes, but it is hard to believe that it was not, to an extent at least, actively encouraged, and as a wood-pasture tree in south Hertfordshire it was almost certainly, in many cases, deliberately planted.

Lastly, we should note the special wood-pasture groves of holly, the ‘Hollins’ which are sporadically recorded in Yorkshire, and which were intended – in the period before the eighteenth century – as a source of winter fodder for livestock. These, too, were presumably deliberately planted.

Conclusion

There seems little doubt that, while the consequence in part of edaphic and climatic factors, the particular balance of tree species found in post-medieval period on farmland, and to a lesser extent in woods and wood-pastures, was largely the result of human agency, and was structured by economic and practical influences. Although as we have emphasised the available sources need to be interpreted with caution, there is no real doubt that oak, ash and elm were the dominant trees in most contexts, their relative importance varying from region to region, because they were undemanding in their requirements and useful in a wide variety of ways. Other trees were less useful, more demanding, or were better grown as coppice. They accordingly occur at low levels in our records, likewise with regional preferences. The artificial character of rural tree populations is perhaps particularly clear in the case of maple, which seeds freely, is widespread in hedges and in the understorey of woods, but which was seldom allowed to grow as a standard or pollard, although it makes a fine tree. Only in particular circumstances did trees other than oak, ash and elm dominate the landscape, such as willows in alluvial wetlands and meadows, especially in the Midlands; and beech and hornbeam in some wood-pastures, where grazing pressure was intense. Human choices about what to plant, preserve or encourage in particular contexts were certainly contingent on

127 environmental factors – on soils, climate and drainage. But tree populations were nevertheless highly unnatural in character, if by ‘natural’ we mean largely uninfluenced by human actions.

Perhaps the clearest indication of this is the loss from the landscape of small-leafed lime, which pollen evidence indicates was the dominant tree in the pre-Neolithic landscape in Norfolk, Hertfordshire and Northamptonshire. By the end of the seventeenth century writers like Moses Cook, when writing of lime, suggest that it was principally a tree of gardens and parks, and they generally fail to distinguish clearly between cordata and the rarer (and possibly introduced)Tilia platyphyllos or the hybrid Common lime (Tilia X europea). Many early documentary references are probably to these latter species. On the Boughton estate in Northamptonshire in 1749 lime (of whatever kind) was only recorded in the gardens and park of Boughton house and the nearby woods, where it was listed with ornamentals like horse chestnut. Elsewhere in the county it fails to feature in surveys before the early nineteenth century, when a single example was noted amongst the 1,038 trees valued at the enclosure of Finedon in 1806, and four listed amongst the 888 surveyed at Staverton in 1835, all again probably ornamentals. In Hertfordshire the single lime listed at Broxbourne in 1784 was probably cordata, but it formed a tiny proportion of the 6,000 or more trees described in the survey. Nine examples were recorded at Barnet in 1787, amongst a total of 1,228 trees listed on the copyhold lands, but – growing on the common, and accompanied by chestnuts and firs – these were again probably ornamental plantings, of Tilia X europea. In Norfolk, similarly, lime is but rarely recorded as a farmland tree, accounting for only three of the 24,000 trees on the Earle estate in the north east of the county in 1722, all probably cordata, which also occurs as a significant component of the underwood in three surviving ancient woods (Hockering Wood, Park Wood in Hockering, Swanton Novers Great Wood). In Northamptonshire it is present in only low frequencies in woodland, as it is in Hertfordshire. So far as the evidence goes, small-leafed lime had largely disappeared from the landscape by the time of the Norman Conquest, perhaps because of its poor resistance to grazing. It had few features to recommend it to the forester. Its wood was soft and not very durable and, while it was good for decorative carving, this did not constitute a major market and the use of its bark for rope production had effectively ceased by the start of the period studied here. The almost complete absence of lime from farmed landscapes, woods and wood- pastures is a striking testimony to the true character of rural tree populations in pre-industrial England.

128

5. The Development of Trees and Woodlands, c.1780 – 1880.

Context

We have employed the term ‘traditional’ to describe early tree populations on a number of occasions over the preceding pages, as a useful shorthand. In reality, as we have also intimated, the phrase probably conveys a misleading impression of long-term stability. The evidence for the character of farmland trees, woods and wood-pastures before the later seventeenth century is patchy, and what we have described may be specific to a particular period of time – that is, the seventeenth and early/mid eighteenth centuries. In earlier periods, social and tenurial structures, economic demands, and technological constraints may have engendered rather different forms of exploitation, leading to rather different strategies of tree and woodland management, and encouraging a different range of tree species in farmland and woods. All this said, it seems likely that the period from the later eighteenth century, to the end of the nineteenth, saw more radical changes to the character of England’s trees than anything which had occurred for many centuries. The development of novel forms of management, the complete demise of many long-established practices and environments, and the widespread planting of new species, introduced from abroad, all transformed the character of England’s woodlands, and of its farmland trees. It is also true, as we shall see, that in some ways there was more continuity in management systems – or at least, a slower rate of change in some forms of management – than is often suggested. But it is the sheer scale of transformation that is the most striking feature of the period.

The ultimate causes of these profound changes are complex and cannot be dealt with in any detail here: in the last analysis the transformation of England’s treescapes reflects the fact that the period between c.1750 and c.1860 saw, in many senses, the birth of the modern world. The early eighteenth century had seen the development – gradually – of a truly global system of trade, a development with which Britain, with its new colonies in the Americas, was centrally involved (Colley 2009, 67-71). But from 1750, economic growth stepped up a gear. The population of England began to increase rapidly, rising with unprecedented speed from around five and half million in 1750 to some 9 million by 1800, and to nearly 19 million by 1861 (Mitchell and Deane 1962, 5-6). This engendered a series of profound agricultural changes – this was the period of the classic ‘agricultural revolution’, in which the widespread adoption of a range of new crops and rotations, and novel techniques, boosted production. Even in 1851 imports only accounted for some 16 percent of foodstuffs consumed in England and Wales. To feed an expanding population the volume of wheat produced more than doubled, while that of barley may have increased by over two thirds, in the course of the later eighteenth and early nineteenth centuries; increases in the production of other foodstuffs were probably of a similar order (Holderness 1989; J.V. Beckett 1990, 9). There was, moreover, a dramatic improvement in what economic historians call ‘labour productivity’, that is, the number of individuals required to produce a given amount of food (Overton 1996, 121-8). In 1760 the output of each agricultural worker could feed around one other person, but by 1841 it could feed another 2.7. This was vital because it was in this period that England also, and perhaps more importantly, crossed the threshold to an economy powered by fossil fuel, and experienced an industrial revolution, so that a higher and higher proportion of the workforce was employed fulltime in mines, mills and factories. Growth was based on three key sectors: coal, iron and textiles. Coal production rose from perhaps 2.5 million tons per annum in c.1700 to over 11 million by 1800, reaching more than 22 million by 1830 (Flinn 1984,27). There were less than twenty blast furnaces in England in 1700; by 1805 there were 177 and by 1852 no less than 665, all now fired with coke

rather than charcoal. In 1700 around 30,000 tons of pig iron were produced in England each year. By 1850 the figure was two million. The expansion of textiles was if anything even more dramatic, although it came a little later, following the improvements in spinning machinery made by Hargreaves, Arkwright and Compton in the 1770s (Palmer and Neaverson 1994, 94-119). The two ‘revolutions’, industrial and agricultural, while often considered separately by historians, were intimately interconnected, in innumerable ways. Without a step-change in agricultural production, large-scale industrialisation could never have taken place. But industrial development also allowed an increase in agricultural production by, for example, allowing coal to be burned instead of traditional fuels, thus freeing up land for crops and grazing.

The period also saw the growth and consolidation of large landed estates, a development probably exacerbated by the emergence of a global economy, and the subsequent rapid expansion of industrial and agricultural production, but with older roots. Landed estates may, for convenience, be defined as extensive and continuous areas owned as absolute private property. At their heart lay a mansion and its grounds, usually accompanied by a ‘home farm’ which was retained ‘in hand’. Beyond lay farms which were leased to tenants, together with a scatter of plantations and game covers which were usually retained under the owner’s direct control (Clemenson 1982; Williamson 2007; Beckett 1984; Beckett 1986). Aristocratic properties might extend over 10,000 acres (c.4,000 hectares) or more; the estates of the local gentry embraced a parish or two and ranged from perhaps 1,000 to 10,000 acres (400 – 4000 hectares) (Clemenson 1982, 7-9). Whatever the scale of their possessions, the emergence of large, continuous properties of this kind from the seventeenth century, and at an accelerating rate in the eighteenth century, had profound effects on tree populations, especially because it was associated with the development of an increasingly market orientated rural economy, and with the decline of peasant production and the demise of many of the ‘traditional’ forms of land management with which this had been associated. Of particular importance in all this, and in the ‘agricultural revolution’ itself, was the large-scale enclosure of common land, something which reached a peak during the decades either side of 1800, and which had profound implications for the management of wood and timber. Areas of heathland, and of upland moor, where they could not be converted to farmland, were often seen as good places to plant trees. In short, a complex range of interconnected developments in the later eighteenth and nineteenth centuries ensured radical changes in the character of England’s tress and woodlands.

Changes in woodland area.

Many writers have emphasised the way in which, in the period after c.1660, there was a marked increase in the establishment of new areas of woodland in England, mostly carried out by large landed estates, with Thomas Coke of , for example, planting around two million trees between 1782 and 1805 (Prince 1987). Their enthusiasm was fired up by the writings of men like John Evelyn, whose book Sylva, or a Discourse on Forest Trees of 1664 was followed by a rash of similar texts, including Moses Cook’s The Manner of Raising, Ordering and Improving Forest Trees of 1676 and Stephen Switzer’s Ichnographica Rustica of 1718. Planting was a patriotic duty, for there was widespread concern that there was a general timber shortage which had implications for the nation’s naval power, as writers like Phillip Miller (1731), James Wheeler (1747), Edmund Wade (1755) and William Hanbury (1758) all warned. In a more general sense the planting of trees demonstrated confidence in the new political dispensation brought about by the Restoration of the Monarchy, and by the Glorious Revolution of 1688 (Daniels 1988). It likewise expressed confidence

130 in the continuity of ownership on the part of local dynasties. Patriotic considerations, and a spirit of ‘improvement’, continued to inspire into the nineteenth century. One Yorkshire commentator in 1823 thought that it was ‘to be regretted that so much barren land should be suffered to remain unplanted … and ought not to lie useless to the Community when the soil could be advantageously employed’ (Hatfield Mercury 16 August 1823).

But planting was also carried out to beautify estates, especially as the fashion for ‘landscape parks’ took hold from the 1750s and 60s, under the influence of Lancelot Brown and his contemporaries: the number of parks proliferated steadily in the second half of the eighteenth century, and woodland clumps and belts were a key feature of their design. New plantations were also established to provide cover for game. In the course of the eighteenth and nineteenth centuries progressive improvements in gun technology – culminating in the 1850s with the development of the breach- loading shotgun - made it easier to shoot game in the air, rather than on the ground. At the same time, enclosure and the consolidation of land into larger and more continuous properties allowed game to be more carefully preserved (Munsche 1981, 8-27). Shooting steadily became more organised and competitive, with shoots involving larger numbers of participants; and this in turn led to the more systematic management and encouragement of game, and to an increasing focus on the pheasant as the principal quarry of sportsmen in lowland areas. Unlike alternative kinds of quarry, the pheasant occupied relatively small territories, could thus be raised in large numbers; and was easily scared into flight, making it an excellent target (Hopkin 1985, 68). But it is a woodland bird, and so large areas of new woods needed to be planted to provide it with a congenial environment (Delabere Blaine 1838, 854; Hill and Robertson 1988, 38-45). While tree-planting schemes were thus often partly ornamental or recreational in character, and represented flamboyant and ostensibly public-spirited attempts at improvement, in the final analysis, many landowners were motivated by hard-nosed economic considerations. As Strickland, commenting on the progress of afforestation on the Bolton estate in the West Riding of Yorkshire, commented: ‘others, encouraged by their success, and the increasing demand and value of timber, continue to plant, and some even with the sole object of profit’ (Crowther 1992, 62).

In the final analysis the upsurge in planting was determined by changes in patterns of ownership and tenure – the growth of large estates and the enclosure of common land. As we have noted, there had always been a close association between large landowners and managed woods: in the Middle Ages intensively managed, coppice-with-standard woods had invariably been the property of manorial lords, lay or ecclesiastical. The same connection is evident in the eighteenth and nineteenth centuries. Tree planting was intimately connected with the rise of great estates because only those owning extensive properties could afford to put hundreds of acres out of agricultural production, foregoing immediate for medium- or long-term financial benefit. The enclosure of heaths and moors was a further encouragement, for forestry made good economic use of more marginal land. The Fairbanks family, surveyors in the Sheffield area, claimed in the early nineteenth century that ‘the proprietors in many instances derive more profit from the sale of wood, than could be obtained from the rent of land if cultivated’ (Shf FB/CP/25/1-2).

The impression given in many published sources is that the ‘great replanting’ of the period after c.1700 ensured a very significant expansion in the area of woodland in England (Hadfield 1967, 79174), but in reality the situation is more complicated. The extent to which new woods were planted in different districts varied according to a range of influences. These included, crucially, the extent to which common land was enclosed, and the quality of the land enclosed – at a time of rising

131 agricultural prices, landowners were unlikely to use land better suited to arable, or even pasture, for tree-planting schemes; the extent to which land in any area was, indeed, owned by large estates, for in many districts small proprietors (or minor gentry) continued to occupy a substantial proportion of the land area; and the amount of woodland already present in the landscape - for where woods were already abundant, on the whole, the extent of new planting was less. More importantly, new planting was also balanced by the loss of existing woodland, to varying degrees in different districts. In this age of agricultural ‘improvement’, many small copses, and some larger woods, growing on the more fertile clays especially, were grubbed out by owners. Methods of improving soil quality through marling and, in particular, novel forms of land drainage, now made possible the cultivation of the sites they occupied, at a time of steadily rising agricultural prices (Wade Martins and Williamson 1999, 61-7). In addition, by the late eighteenth century grazed woods – wood-pastures – became increasingly redundant in economic terms due to the spread of coal use, and they were regarded as inefficient and old-fashioned features of the landscape by the social elite. Surviving tracts – especially in forest districts – were systematically destroyed following enclosure. The interaction of these varied factors assured that the area of woodland in different districts, and regions, increased to very varying degrees, although our sources only allow us to estimate such changes in very broad terms.

In Norfolk, as we have seen, the extent of woodland in the later eighteenth century can be estimated, to an extent, using William Faden’s county map of 1797 (surveyed 1794-5), although as we have seen only within broad limits (MacNair and Williamson 2010, 119-25). A total of 13, 417 hectares of enclosed woodland – ancient woodland and plantations – are depicted on the map as a whole, amounting to 2.4 percent of the county’s surface area. To this we need to add surviving areas of wood-pasture commons, several examples of which are depicted on the map, although with what degree of accuracy, and how comprehensively, is uncertain. There are also the familiar problems of definition - how closely-spaced must trees (and tree symbols) be across an area of heath for us to classify it as ‘grazed woodland’? At most, however, the wooded commons shown by Faden account for only a further 670 hectares, bringing the total area of woodland in the county to around 14,000 hectares, or around 2.5 percent of its land area.

A significant proportion of the enclosed woodland shown by Faden – around a third - appears to have been established over the previous half-century or so, comprising as it does belts and clumps in parks or – in some cases – examples specifically labelled ‘plantation’. The marked concentration of new woodland in and around the landscape parks laid out around elite residences to some extent mirrors the comments made by William Kent in 1796, that while ‘gentlemen of fortune’ in Norfolk had carried out much tree-planting ‘in their parks and grounds’, the planting of ‘pits, angles, and great screens upon the distant parts of their estates, which I conceive to be the greatest object of improvement, has been but little attended to’ (Kent 1796, 87). In fact, he appears to have exaggerated to some extent: around half of the probable recent woodland shown by Faden was located away from estate centres. Either way, the majority of the new plantings were located on the light soils to the west and north of the county. This was former sheep-corn, ‘champion’ land, largely devoid of significant areas of enclosed woodland before the enclosure of open-fields and heaths in the course of the eighteenth century. It was the part of the county in which the largest landed estates, and the largest areas of parkland, developed, and also one with large tracts of former common land, ripe for planting.

132

If the area of enclosed woodland was already increasing by c.1795, expansion continued over the following century, and by the end of the nineteenth century, to judge from the First Edition Ordnance Survey 25-inch maps, and the government woodland census of 1895 (Board of Ag Returns 1896, 36), woodland cover had reached around 4 percent of the county’s land area. In rough terms, we might say that the area of woodland increased by a factor of around 60 percent in the century after Faden’s map was surveyed. While the planting of belts and clumps in parks – and the creation of many new parks and the extension of many existing examples – accounts for much of this, the large-scale afforestation of areas of marginal land, mainly former heathland enclosed in the decades either side of 1800, was also important: especially on the Holt-Cromer ridge, on the light, sandy land immediately to the north of Norwich, and, to a more limited extent, in Breckland. It needs to be emphasised, however, that the bald figures of woodland area noted above to some extent under- represent the scale of aristocratic planting, because as well as major additions to the woodland area the period saw a number of reductions. In particular, most surviving wood-pastures on common land in the county were destroyed following enclosure; while, as we have already intimated, some areas of long-established coppiced woodland were grubbed out and converted to farmland. Smaller areas of coppiced woodland – and some larger ones, like Wood in the east of the county - were removed altogether, others were drastically reduced in size. for example lost around a third of its area, Horningtoft Great Wood more than two thirds and Rawhall Wood over a quarter in the first three decades of the nineteenth century; while between c.1840 and 1880 Ashwellthorpe Wood, Banyards Wood in Bunwell, Billingford Wood, Horningtoft Great and Horningtoft Little Woods, Old Pollard Wood in Holt, Great Wood and Grove were all significantly truncated (Barnes and Williamson 2015, 113). Most ancient woodland was located in the old-enclosed clayland districts in the south and east of the county, and it was here, accordingly, most of the destruction occurred. Most of the new plantations, in contrast, were established (and noted) on the lighter land in the north and west. The overall affect of new planting, amplified by much grubbing out of woods, was thus to even out the distribution of woodland between different districts. Only in the drained wetlands of the Fens, in the far west, did woods and plantations remain rare.

In Hertfordshire there were, overall, less dramatic changes in woodland area in this period. The county’s equivalent to Faden’s county map, Dury and Andrews’ survey of 1766, unfortunately provides only a partial and inaccurate indication of woodland in the county, as we have seen. As well as omitting many examples which unquestionably existed at the time it was surveyed, the map also includes a number of what can only be described as ‘fantasy woods’ – areas depicted at woodland which were either entirely imaginary, or which were in fact other kinds of well-treed area, including even the famous nursery run by the Rivers family near Sawbridgeworth (MacNair et al. 2016, 106-7). Comparison with later maps, especially the draft Ordnance Survey maps of the early nineteenth century (although themselves of only limited reliability) allows us to remove most of these erroneous inclusions, however, and – within broad estimates – it appears likely that in 1766 enclosed woodland covered around 10,000 hectares, or 6.1 percent, of the county’s land area, a much higher proportion than in Norfolk. To this, moreover, we again need to add a number of surviving areas of wood-pasture, which were also more extensive than in Norfolk, perhaps amounting to a further 2,400 hectares of wooded common and – somewhat less certainly, because of problems of definition – around 700 hectares in parks (many more of which in this county had been established as true deer parks over the previous three or four centuries, and continued to maintain, to an extent at

133 least, a well-wooded character). This would bring the total of woodland to 13,100 hectares, or c. 8 percent of the county’s land area.

In marked contrast to the situation in Norfolk – and demonstrating again, and more sharply, the principal that the most wooded areas saw the least new planting in this period - the period from 1766, to the end of the nineteenth century, did not see a significant increase in woodland area. Local maps, and smaller scale surveys like the draft Ordnance survey drawings of the early nineteenth century and the 6-inch First Edition of the 1890s, suggest numerous localised losses, mainly involving small areas of woodland (especially narrow ‘hedge rows’) but including some larger examples. These were perhaps most marked on the boulder clay soils in the east of the county, where for example no less than 11 small and medium-sized woods, with a combined area of 27 hectares, disappeared in the adjacent parishes of Much and Little Hadham between the early nineteenth century and the 1880s (HALS DE/Cn/P4; HALS DE/X713/P1). The remaining wood-pasture commons in the south of the county, and most of those in the Chilterns, were also destroyed following enclosure around 1800. Northaw Common is shown as still largely tree-covered on the Ordnance Survey draft drawings of 1805 but the words ‘clearing for inclosure’ have been written across it, in anticipation of its expected fate. Yet at the same time, there were many piecemeal additions of woods and plantations especially within, and on the margins of, the landscape parks which were particularly numerous in this county, on account of its proximity to the metropolis. Such woods had only accounted for a small proportion of the county’s woodland in 1766, probably less than 800 hectares in all. By the later nineteenth century they were much more extensive, in part because the number of parks increased considerably, rising from around 5.4 percent of the county’s land area in 1766 to more than 8 percent by the 1880s. In sharp contrast to Norfolk, the overall area covered by woodlands of various kinds thus shrank, although not perhaps to a dramatic extent, in the period between the late eighteenth and the late nineteenth centuries. In 1895 the total area of woodland in Hertfordshire was estimated at 10,338 hectares, or 6.3 percent of the land area (Board of Ag Returns 1896, 36). This was less than the 13, 100 hectares (8 percent of land area) shown on Dury and Andrews, but - given the problems with this source already noted, and the problems of estimating the extent of grazed woodland shown on this or any other map - it is possible that the real scale of decline was rather less.

If Norfolk saw a significant increase in the area under trees, and Hertfordshire some reduction, then Northamptonshire seems to have experienced a more substantial contraction in woodland area. There are no reliable county maps for Northamptonshire which can be used to estimate the extent of woodland in the middle or later decades of the eighteenth century but a comprehensive examination of the earliest available map of every individual township in the county suggests that there were around 15,100 hectares of woodland within its boundaries in c.1750, c.6 percent of total land area, together with an uncertain amount of wood-pasture, possible equivalent to anther 2 percent.13 The vast majority of both continued, as in earlier centuries, to be concentrated in the three

13 This was carried out as part of an earlier AHRC-funded investigation of changing pattern of land use in Northamptonshire. See Williamson et al. 2013; and http://archaeologydataservice.ac.uk/archives/view/midlandgis_ahrc_2010/

134 royal forests of Salcey, Whittlewood and Rockingham, all occupying boulder clay uplands between the principal river valleys. The enclosure and destruction of the forests was the single greatest influence on the changing extent of woodland in the county in the course of the nineteenth centuries.

In part because of the complicated patterns of overlapping rights and ownerships discussed in Chapter 3, enclosure of the forests came relatively late. Brigstock Bailiwick in Rockingham Forest was enclosed in 1805 under an act passed in 1795, and Cliffe Bailiwick was enclosed in 1806 under an act of 1796. But the act for enclosing Salcey was only passed in 1825, with the award coming the following year; Rockingham’s act came in 1832, with the award in 1837; while Whittlewood was enclosed in two stages, with awards in 1826 and 1856 (Pettit 1968; Williamson et al. 2013, 144-48). Each award made extensive allotments to the Crown (for forestial rights, manorial rights and ownership of land), to the holders of forest offices, and to local commoners. Provision was also generally made for the poor by an allocation to the churchwardens or overseers for each parish, in recognition of the fact that the local poor had formerly exercised the right to collect wood for fuel. The acts in all cases emphasised the ‘injurious’ effects of common rights on the value of the underwood in the enclosed coppices and on timber, as well as a more general level of mismanagement undermining the economic potential of the forest areas. The Salcey award specifically mentions that the allotments to the Crown and the Duke of Grafton were to be selected on the basis that they would be the best areas for growing trees; in Haselborough in Whittlewood the Crown allotment was likewise ‘for the growth and preservation of timber’ (NHRO X1693 Bundle 21; G3909). But it was also envisaged that many of the allotted areas could be cleared and cultivated and, while the act for Rockingham states that the woods ‘would produce a large quantity of timber’, it notes that other areas ‘might be profitably converted into and used as farms’ (NHRO YZ 6685).

The wood-pasture plains were generally destroyed immediately after enclosure: any trees which still remained on them were cleared and all the acts made provision for the removal or destruction of the deer, except those for Geddington Chase (already disafforested) and for Brigstock and Cliffe bailiwicks in Rockingham Forest. The coppiced woods suffered diverse fates. Many of those in Rockingham were destroyed during the following decades: in all, around 3,600 hectares of the enclosed woodland there appears to have been converted to farmland (or occasionally to parkland) by the time the Ordnance Survey 25-inch maps were surveyed in the 1880s. In Whittlewood around 1,600 hectares were destroyed on the Northamptonshire side of the county boundary; in Salcey, a further 480 hectares of coppice was lost.

In the course of the nineteenth century Northamptonshire thus lost nearly 5,680 hectares of forest coppice, and to this we should perhaps add an unknown area of residual wood-pasture surviving on the 3,100 hectares of woodland plains and ridings. To some extent, as in the two counties just discussed, such losses were offset by the appearance of new woods, in this county including many fox coverts, as well as the belts and clumps around great houses and the occasional larger areas of woodland established by major landowners, throughout the county. On the other hand, some of the relatively few ancient, coppiced woods lying away from the forests also disappeared in this period, including large areas around Stow Nine Churches and in the west of the county. Overall, the area of woodland in the county (including wood-pasture) probably fell from around 16,500 hectares in the later eighteenth century, to around 11, 338 in 1895 (Board of Ag Returns 1896, 36): that is, from around 7 or 8 percent to around 4.5 percent of the land area. In spite of disafforestation, however, woodland still remained concentrated in the old ‘forest’ areas.

135

Yorkshire, as ever, presents a more complex picture, in part because of its large size and varied geology and terrain, and in part because many descriptions, and much of our data, is framed within the three administrative divisions of the county, which are poorly related to patterns of topographical variation. The county map surveyed by Jefferys and published in 1771 is too schematic in its treatment of woodland to provide a reliable impression of its extent, and no Ordnance Survey drawings from the early nineteenth century exist for the north of England, which might be used to remedy this. It is thus difficult to obtain a reliable impression of the pattern of change in woodland area during the period between the mid-eighteenth and late nineteenth century, although the broad outlines are reasonably clear.

In the eighteenth century the East Riding was the most sparsely-wooded of the three ancient divisions of the county. Much of its area was made up of the chalk Wolds, a ‘champion’ area of open- field arable and sheep-walks where, Leatham believed in 1784, the soil and climate were ‘not generally adapted to the growth of wood, and very little is to be found’. In the area immediately to the north, in the lowlands extending beside the rivers Hertford and Derwent, there was likewise he believed ‘but little wood’ (Crowther 1992, 35). Even in 1812 Henry Strickland was able to claim that the woods and hedgerows on one estate in Escricke were together worth more than all the full grown timber in the East Riding put together (Crowther 1992, 61), while the district of Holderness lying to the east of the Wolds ‘never contained any woods … and very little hedge-row timber, and that of inconsiderable value’ (Crowther 1992, 62). Only in a few parts of the Riding was the situation any different: in the south of the district, in the strip of land between the rivers Ouse and Humber and the Wolds, where Leatham noted that ‘a considerable part is well wooded (chiefly with oak)’; and to the south of York, in the Ainsty of the City, there were also ‘some extensive woods, many small patches, and a large quantity of trees in the hedgerows’ (Crowther 1992, 34–6). But it was the great mass of the treeless Wolds which was the main feature of the East Riding, and Jefferys’ map of 1771 confirms that, except for a few areas in the vicinity of country houses, remarkably little woodland existed in the district.

In the course of the later eighteenth and nineteenth centuries, as large estates acquired more and more of the land area and as enclosure proceded, this situation changed. Sir Tatton Sykes, whose estates in the Wolds, centred on Sledmere, covered some 34,000 acres (13,760 hectares), famously planted some 1,000 acres (400 hectares) with trees, a memorial tablet in West Heslerton church proudly proclaiming: ‘Whoever now traverses the Wolds of Yorkshire and contrasts their present appearance with what they were cannot but extol the name of Sykes’ (Ward 1967, 13). Neighbours with less extensive properties did likewise, significantly changing the appearance of the landscape, although the areas involved were, individually, often small: on a journey through the East Riding in 1845, George Head observed that the farms and outbuildings were often ‘surrounded by a belt of thriving plantations, the whole together, seen from a distance, like a little village’ (Crowther 1992, 75). By 1895 the total area under woods and plantations in the East Riding had reached 17, 181 acres (6,953 hectares), some 2.3 percent of the surface area (Board of Ag Returns 1896, 36). The vast majority of this woodland must have been planted since the early or middle decades of the eighteenth century. In 1924, only a few acres were considered to be coppice-with-standards, and the vast majority ‘high forest’, and there are remarkably few woods in the former area of the Riding are designated by Natural England as ‘ancient woodland’. There are exceptions, mainly in the area away from the Wolds, to the south of York, such as Hollicarrs Wood is Escrick – but they are few. At a conservative estimate, around two thirds of the wooded areas in the East Riding present in 1895 had

136 probably been established since c.1750 and, while some existing areas of woodland were doubtless lost in the intervening period, it seems likely that the percentage of the land area occupied by woods and plantations rose from around 1.5 percent to 2.3 percent in the century and a half after 1750.

The other ancient subdivisions of the county, as we have already noted, had long been rather better endowed with woodland. By 1895 the North Riding boasted 52, 816 acres (21,374 hectares) of woodland – 3.9 percent of its surface area (Board of Ag Returns 1896, 36). Again, only a small proportion (1.2 percent) was classified as coppice-with-standards in 1924, the majority – 58 percent - considered as ‘high forest’ (the rest was scrub (10 percent), felled or devastated (22 percent) or ‘uneconomic’ (8 percent)). But many of these woods had probably been converted from old areas of coppice woodland in the course of the nineteenth century, as we shall see; while Jefferys’ map of 1771 shows, albeit schematically, numerous areas of woodland, mainly growing on the steep slopes of the deeply incised valleys draining from the North York Moors to the Vale of Pickering, and in the Hambleton and Howardian Hills. Some old woods also existed in the Vale of York, which in 1945 were described as the remains of carr woodland, partly colonised by birch. Indeed, in this Riding there are considerably more areas of ancient woodland recorded on the Ancient Woodland Inventory than in the East Riding and fewer which – by name, or location – appear to be of eighteenth- or nineteenth-century origin. Tuke in 1800 estimated that in the whole of the North Riding, embracing not only the Vale of York but also the North York Moors and the Howardian Hills, there were only 25,500 acres (10,319 hectares) of woodland – a little under 2 percent of the total land area – and this may have been roughly correct. At a rough estimate, around half of the woods shown on the late nineteenth-century Ordnance Survey maps had originated before the late eighteenth century, and around half were the consequence of estate planting, mostly on the poor sandy soils in the Vale of York, to the north of York, and on the Howardian Hills.

In the West Riding there was a similar amount of woodland by 1895 – 69,592 acres (28, 163 hectares), or 3.9 percent of the land area (Board of Ag Returns 1896, 36). Much of this again predated the mid-eighteenth century, most notably the coppices or shrubby thickets of hazel and ash on steep slopes on the Carboniferous limestone, and the thickets of birch and rowan, and scattered larger woods with ash timber on the lower slopes, on the Millstone Grit. But there was, nevertheless, much estate planting in the later eighteenth and nineteenth centuries, often following hard on the heels on the parliamentary enclosure of upland moors (Dormor (2002, 232). On the Bolton estate near Wensleydale, for example, Lady Bolton and Sir John Orde established some 30 new plantations, containing around 55,000 trees, in the early years of the nineteenth century (Dormor 2002, 232). We might tentatively estimate that overall the woodland area of the Riding similarly doubled, from around 2 to 3.9 percent of the surface area. Combining the figures from the three Ridings – and allowing for the fact that woods were, as in other areas studied, lost as well as gained - it seems likely that the area of woodland in Yorkshire as a whole increased from around 43,000 hectares – just under 3 percent of the land area - to around 56,500 – 3.6 percent – in the century after c.1780. These figures, it must be admitted, are more speculative than those proposed for the other three counties studied, but are probably broadly correct.

137

Figure 27. Generalised ‘heat’ maps, providing a broad indication of the distribution of woodland in two sample counties on the late nineteenth century OS 6” maps. Top: Northamptonshire. Below: Hertfordshire.

138

None of the four counties studied was thus well endowed with woodland in the eighteenth and nineteenth centuries: in 1895, when we have the first reasonably reliable figures, none had more than 6.3 percent of their surface area under trees and, even well-wooded Hertfordshire it is unlikely that the figure even in c.1780 had much exceeded 8 percent. Moreover, although attention has often been drawn to the upsurge in tree planting which occurred in the period after the Restoration, not everywhere saw an increase in the wooded area in the course of the eighteenth and nineteenth centuries. Hertfordshire and, in particular, Northamptonshire appear to have experienced a reduction and, while both Norfolk and Yorkshire saw an overall increase, this was concentrated in particular areas: on lighter land, especially on thin chalk and sandy soils, where planting often followed the enclosure of heaths and sheepwalks; and on upland moors. The greatest losses of woodland in all of the counties studied, conversely, seems to have occurred on fertile clay soils, and especially on those formed in boulder clay – most notably in Northamptonshire, where the three royal forests all occupied clay uplands, but also in south Norfolk and, albeit to a lesser extent, east Hertfordshire. Woods were, in all the areas studied, lost as well as gained, and the overall increase or decrease was a function of whether one process or the other predominated.

Accompanying these changes in the extent and distribution of woodland, driven as we have noted by changes in agricultural markets, agricultural technology and industrialisation, there were important alterations in woodland character, and in how woods were used and perceived by contemporaries. While most wood-pastures, especially in royal forests and on commons, disappeared, woodland with a quasi-ornamental function – clumps and belts in parks, games cover and the like – increased significantly; and while larger schemes of aristocratic afforestation, especially on former heaths, appear to have been less aesthetic or recreational in character these also were, in many cases, an expression of a fashionable interest in ‘improvement’. Indeed, to some extent any distinction between ‘aesthetic’ and ‘economic’ forestry can be over-drawn. The regular fellings made as plantations were thinned, as described below, ensured that even the clumps and belts in parks were regarded as a source of income, and all estate planting represented, to some extent, money in the bank. The great belt planted around the perimeter of the park at in south west Norfolk was put up for sale along with the rest of the estate in 1788, and the sales particulars described at some length the ‘most agreeable ride’ which ran along its centre. Yet it also noted how:

The number of trees that will remain in the Plantations, after they are thinned so as to leave them at a proper distance, to facilitate their Growth, will be about Six Hundred Thousand: which in the Course of a few years, will at least be worth a shilling a Tree, and consequently amount to Thirty Thousand Pounds (NRO MC 77/1/521/7).

- a sum equivalent to tens of millions of pounds today. Estate forestry was not only about balancing aesthetics and profit, however. As we have noted, this was a period in which game shooting became increasingly important, and most plantations also served as pheasant cover, the trees being under- planted with shrubs like snowberry and rhododendron as the trees grew mature and were progressively thinned. As Marshall said of Norfolk in 1787:

Ornamental plantations, about the residences of men of fortune, are here, as in other districts fashionable: not, however, as objects of ornament merely, but likewise as nurseries of game (Marshall 1787, I, 120-1).

Whatever the balance of motives in aristocratic forestry schemes, we must be careful not to exaggerate their overall significance. Although in west and north west Norfolk, and on the Wolds and

139 moors of Yorkshire, new planting transformed the appearance of the landscape, overall it was on a more muted scale, and was accompanied by significant losses of existing woodland. Averaged across the four counties, the area of woodland probably rose from around 3.7 percent of the land area in c.1780, to around 4.2 percent at the start of the twentieth century, a rather less dramatic increase than sometimes suggested by historians (Hadfield 1967; Daniels 1988).

Plantation forestry

Most of the new woods established in this period were plantations, planted only with timber trees and lacking any coppiced understorey. In most cases, deciduous species – old-established ones like oak and elm, but also sweet chestnut, beech and sycamore - were mixed with a rather larger number of ‘nurses’, most of which were conifers, especially larch and Scots pine. In the words of the agricultural writer Nathaniel Kent, plantations in Norfolk consisted of ‘Great bodies of firs, intermixed with a lesser number of forest trees’ (Kent 1796, 87). The bills for the trees planted in the 1790s at Heydon in Norfolk typically suggest that the belts and adjoining plantations contained around four times as many conifers as hardwoods. On the Ingliby estates in north Yorkshire the entries in Sir John Ingilby’s planting book for 1781 likewise records the planting of a 'great quantity' of oak, ash, beech, birch and holly, together 1000 sycamore, but also 600 larch, 700 spruce and as many as 3,500 Scots pines (there were also smaller quantities of alder, willow, beech, Dutch elm, Weymouth pine, Black cherry, lime, horse chestnut, hornbeam and ash, most probably planted for aesthetic reasons). In 1783, 2,290 oaks were planted in the estate woods, but 5,470 Scots pines and 3,160 larches, together with 1,400 silver firs, 150 horse chestnuts, 206 Weymouth pines and a mere 70 ashes (Dormor 2002, 106-7). The Fairbanks family were likewise advised that on moorland in south Yorkshire ‘the larch will probably pay best in both thinnings & timber’, although ‘experienced planters aware of the great uncertainty in regard to any criterion for practice will always plant a variety of sorts some of which will scarcely fail to make timber in any situation’: they should thus plant ‘½ larch & the other ½ scotch fir with spruce fir in moister parts, for the noble tree the spruce will not succeed in dry ground; oaks, elms, beeches, birches, in equal proportions’’. The whole plantation would therefore consist of ‘1550 larches, 300 of each of the other makes abt [sic] 3000 plants on an acre. Every other row as wholly larch, the intervening rows mixed’ (Shf FB/CP/25/1-2).

Some of the new plantations established in this period, however, appear to have comprised conifers alone, usually Scots pine, spruce and larch but occasionally other species, as at Westwick in Norfolk where the plantations to the north west of the park established by John Berney Petre were reported, in the 1790s, to consist entirely of ‘pinasters [Maritime pine or Pinus pinaster], except in the valley where other trees grow’. A valuation of the timber growing at Mundford, in Norfolk’s Breckland, made in 1806 described one plantation – ‘Long Plantation’ – which consisted, in the manner just described, of a mixture of larch, pine, fir, elm, oak and birch. But two others were planted with conifers alone: ‘Square Plantation’ contained 9,827 larch, spruce Scots and Weymouth pine, while ‘Round Plantation’ consisted of 4,360 larch, spruce and Scots pine (NRO MS 13751, 40E3). On upland moors in Yorkshire there was a similar tendency to plant a high proportion of conifers, usually dominated by larch.

Whatever the precise composition of new plantations, the trees were usually planted more closely than would be usual today, at a density of around one per square metre. It was reported that on the Longshaw Estate in the Peak District and trees were planted in ‘rows 4ft from row to row & 3ft 6 inches from plant to plant. This is much better for cutting out the thinnings, which after all is the

140 main thing in this business & they make clean timber with less pruning’ (Shf FB/CP/25/1-2). Tuke in 1800 similarly described how Yorkshire plantations were composed of ‘forest-trees of two years old, and planted at about four feet distance’ (Tuke 1800, 181). Such dense planting was necessary because it was difficult to deal with the weeds which competed with the young trees, and also to some extent because significant losses were anticipated from drought and the depredations of rabbits and other animals. Indeed, the journals of eighteenth-century landowners often describe their activities almost as heroic striving against the odds of nature, William Windham of Felbrigg in north Norfolk for example bemoaning in 1826 how ‘The hares and rabbits have destroyed the plantations on Harrisons Brake entirely, not withstanding the great cost of the fencing’ NRO WKC 7/134). Secure fencing was, therefore, a sine qua non of successful forestry, but in upland areas especially drainage was also an important consideration: when a new plantation was established by the Sheffield Planting Company on the Longshaw Estate in the Peak District it was reported that ‘draining should be done as much as possible by open survey cuts on drains’ (Shf FB/C/25/1-2).

Having been planted densely, the trees were then progressively thinned, beginning at around five years and continuing at frequent intervals thereafter. The fact that the extracted material is often referred to in estate accounts as ‘poles’ signifies, clearly enough, that it was used in a similar way to the produce of coppices (Williamson 1998, 183-6). Nathaniel Kent typically described the great plantation belt around the park at Holkham as comprising ‘four hundred and eighty acres of different kinds of plant, two thirds of which are meant to be thinned and cut down for underwood, so as to leave the oak, Spanish chestnut, and beech, only as timber’ (Kent 1796, 90: our italics). In many cases, the final timber crop was itself only thinned, leaving the plantations to provide shelter, cover for game, or to beautify the countryside. Careful attention needed to be paid to thinning at the appropriate time, otherwise the trees grew too cramped, and the deciduous species might be over- topped by the faster-growing conifers. Marshall in 1787 described the poor state of a plantation containing hardwoods and conifers which had been established 25 years earlier on the Gunton estate in north Norfolk:

The Scotch fir has outgrown any other species; and the plants, though few, have become a burden to the grove. The wood being of quick growth, the plants have not only out-topped the rest, but have, in general, had time enough to furnish themselves with boughs on every side … The larches, too, where they stand free from the Scotch firs, are of considerable size (Marshall 1787, II, 95).

It is important to emphasise that the new forestry now only involved planting novel coniferous species, introduced from abroad – notably larch and spruce. It also saw the spread of indigenous trees beyond their natural range into new parts of the country. Beech, for example, appears to have been common enough in the south of England in the seventeenth century but not in the Midlands or East Anglia, while in the north it was unknown. In the later eighteenth century it was planted extensively in all areas. Sweet chestnut, naturalised at an early date in the south of the country, only now spread more widely. Scots pine similarly extended its range from Scotland across all parts of England in the course of the later seventeenth, eighteenth and nineteenth centuries. It is not entirely clear how far the enthusiasm for species like beech and sweet chestnut was motivated by economics, and how far by aesthetic concerns and the dictates of fashion, but beech as we have seen was a useful wood, widely used for boards and furniture making, and while it appears to have fetched prices around half those for oak, it was a faster-growing tree. This said, much appears to have been extracted and sold at pole stage: in many contexts the final ‘crop’ was perpetuated as an

141 adornment of the landscape. Sweet chestnut was likewise a useful wood, especially for fencing, but again its aesthetic qualities may have been a major reason for the enthusiasm with which it was planted by landowners.

The impact of the new species on estate planting should not be exaggerated, however. A survey of the timber in the plantations on the Houghton estate in Norfolk, growing outside the park, made in 1816 listed 13, 242 trees, no less than 18 different varieties (alder, ash, beech, birch, cherry, cedar of Lebanon, elm, ‘firs’, hornbeam, horse chestnut, larch, lime, oak, Scots pine, spruce, sweet chestnut, sycamore and white poplar) (Houghton Hall archives). But oak, ash and elm made up two thirds of the total. After oak (50 percent) and ash (13 percent), Scots pine (10 percent) was the next most important tree; no other species accounted for more than 5 percent, and all the conifers combined only made up 15 percent of the trees listed. In Yorkshire, similarly, oak continued to be the dominant tree recorded in nineteenth-century surveys, followed by ash. Out of c. 73,600 trees (including poles, cyphers and the like) whose species is described in valuations dating to the period after 1800, 48% were oak, 21% ash. Larch made up the third largest category. As already noted, John Ingilby's planting book shows that 600 larch were planted in 1781, with more throughout the early 1780s Dormor (2002, 106-7), although in general the tree only appears in any numbers – at least by name – from c.1830 (Scotch ‘fir’ is likewise only mentioned frequently in records from the 1830s, as is the rather vaguer ‘fir’, although the latter first appears from the 1750s)(Brad Tong MSS 4f/11; Brad 23D98/3/5).

Even in the early twentieth century oak, ash and elm continued to be the most important trees in the woods and plantations on most estates. At Heydon in north Norfolk in 1921, for example, oak and ash made up 73 percent of the 728 trees sold: the next most numerous species was beech (6 percent), followed by sycamore (2.6 percent) (NRO BUL 4/118, 615x9). This said, on the poorer, more acid soils in particular, conifers, and species like beech, sycamore and sweet chestnut, were widely planted. A list of trees uprooted in a storm in 1916 on the Bagge family’s properties on the poor, sandy soils to the east of King’s Lynn, around Bawsey, listed 12 oaks and 9 elm, together with seven beech, six poplar, a lime and a sweet chestnut, all outnumbered by 77 conifers of various kinds, in addition to the ‘many Pine, Spruce and Larch’ noted as lost in one of the plantations (NRO BL/BG5 /3/18).

Coppiced woodland

It might be expected that the fashionable interest in the ‘new forestry’ would have led to the neglect of traditional forms of woodland, and it is true that many forestry writers were, by the end of the century, paying comparatively little attention to them. Nathaniel Kent’s book on the agriculture of Norfolk, published in 1796, for example, devotes three hundred and nineteen lines to plantations (with a further appendix devoted to those composed of sweet chestnut) but just nineteen to woodland in the traditional sense (Kent 1796). Moreover, as already noted, many areas of existing, mainly ancient woodland were lost or truncated in the course of the eighteenth and nineteenth century. Yet closer inspection suggests a more complex picture. In Yorkshire, it is true - where labour costs were high and coal and industrial substitution reduced demand for underwood - coppicing unquestionably declined from the middle decades of the nineteenth century, a development also encouraged by a decline in the price of bark as new methods of industrial tanning, utilising chromium salts, were adopted. Of particular importance were changes in furnace technology, as charcoal and whitecoal were gradually replaced by coal and coke. Over time, regular coppicing was

142 replaced with long-term management for larger timber, and big swathes of woodland were converted to high forest. Jones (2005) has examined this process in South Yorkshire, where former coppice-with-standards woodlands were gradually filled with new planting, and then left to mature: ’By the end of the nineteenth century coppicing, and its related trades and industries had all but disappeared’ (Jones 2005, 55). Industrialisation affected woodland management in other areas, but evidently to a lesser extent. Even in Northamptonshire, faggots, fencing poles, headstakes, thatching broaches and bobbin wood were all being produced from the woods on the Boughton estate in the 1880s, and as late as 1924 12, 517 acres (5,056 hectares) of woodland were still being actively coppiced in the county, compared with 9,916 described as high forest (Forestry Commission 4th Annual Report, 1924, 25; Boughton House archives). In Hertfordshire and Norfolk, similarly, most areas of coppice continued to be managed along traditional lines right through the nineteenth century and even beyond, in spite of the increasing availability of coal. Timbers accounts from the Stow estate in west Norfolk, for example, show that the estate woods continued to be regularly coppiced long after 1850: there are records of sales of pea sticks, stakes, poles, faggots and above all hurdles from the two main coppiced woods on the estate, Spring Wood and Toombers, the latter fetching £3.16s a dozen (NRO HARE 5501, 223 x 1). Similarly, the accounts for Hickling Wood on the Evans-Lomb estate in north east Norfolk in the 1880s show how faggoted wood continued to be sold, presumably for firing, to a large number of separate buyers, and how larger poles were also produced and sold (NRO HNR 736/9). Of particular interest is the description provided in 1851 by the agent for the Merton estate, Henry Wood, of the management of Wayland Wood in Watton. The larger poles were used to make hurdles, fencing or bins for storing hay or straw on the home farm; other material was cut as splints, about 6 feet in length, used for building repairs; and the smaller material went for thatching broaches and sways, and for pea sticks for gardens. Only the residue appears to have been destined for fuel, the smaller brush faggots being sold to bakers and cottagers for ‘oven wood’, and off-cuts of all products being sold for ‘cottage firing’ (NRO WLS XVIII/7/1). On other estates, throughout the nineteenth century, poles continued to be allocated to tenants for farm repairs, gates and fencing (as for example on the Evans Lomb estate in mid Norfolk: NRO HNR 112/19). There were also specialised local markets. In west Norfolk, many poles and faggots were used to reinforce the banks of the larger watercourses in the nearby Fens, as at Mintlyn and Gaywood in 1835 (BL/BG5/3/11). Some writers have suggested that there was an increase in the length of coppice rotations in the post-medieval period, due to the declining value of underwood relative to timber (Rackham 1986a, 92). As we have seen, there is no clear evidence for this, the available sources instead suggesting continued variety, with those recorded in the tithe files from the 1830s, and other sources, ranging from seven years to fourteen years in Norfolk, Hertfordshire and Northamptonshire. This said, it is true that the former sources does comment on a number of occasions on the poor condition of coppices - no less than 52 percent of comments on the quality of those in Norfolk describe them as ‘poor’, compared with 30 percent as ‘good’. In some places this was explicitly because of the density of timber trees had been increased. But in general the sources make clear, in the southern and Midland counties at least, the continuing importance of coppicing up to and beyond the end of the nineteenth century, mirroring the suggestions made by Collins about the area to the south of the Thames (Collins 1989).

Indeed, in the course of the nineteenth century attempts were often made to improve long- established coppiced woods by, in particular, installing networks of land drains within many examples, usually taking the form of ditches around 0.4 metres deep and between 1 and 2 metres in width and often forming dense and regular networks. This was carried out in the largely erroneous

143 belief that, like the under-drains now widely employed on agricultural land, they would improve productivity. Brown for example urged that ‘a properly executed system of drainage will result in heavier and more profitable tree-crops being yielded’ (Brown 1847, 531) (the fact that the drains were usually dug ruler-straight – no easy task in a wood containing numerous coppice stools and sporadic timber trees with extensive root systems –betrays the hand of faddish, ‘scientific’ improvement).

On large estates an additional reason why traditional management was continued in coppiced woods was that the underwood, regularly cut, provided good game cover. Whereas plantations, as they thinned and matured, needed to be underplanted with rhododendron or other shrubs, the larger coppice woods, felled on rotation, always had a significant part of their area suitable for the roosting birds. Indeed, the owner of Wayland Wood in Norfolk, Lord , described in 1893 in some detail how the wood was beaten, commenting that ‘in about four years after felling, the condition of the undergrowth will be most suitable for this purpose … In some portions of a large wood, the undergrowth will be from five to ten years old, and much of this will be difficult to beat through and more difficult to shoot in where paths or rides are cut’ (Walsingham and Payne-Gallwey 1893, 220). Miche suggested in 1888 that ‘the principal reason for maintaining old woods near the mansion’ was that there was ‘never at any one time … so much wood cut down as to cause serious blanks or openings’ (Miche 1888, 105). In part because of their value as coverts, long-established woods were increasingly taken in hand by their owners from the middle of the eighteenth century. Where they were not, there was obvious potential for conflict, as at Helhoughton in Norfolk in 1813 where it was alleged that ‘William Withers… caused the whole of the underwood which was growing …to be cut…which was very predjudicial to the game’ (NRO BUL 11/436).

In addition to being used as game reserves ancient woods were - like existing farmland trees (Rackham 2004) – also often incorporated into the designed landscapes laid out around great houses. In the early part of the eighteenth century, when parks and gardens were designed in a geometric manner, they might be used to create formal ‘’ or forest gardens, as at St Paul’s Waldenbury and Tring in Hertfordshire, or at Hethel in Norfolk (Barnes and Williamson 2015, 196-99). From the 1750s, as the taste for more ‘naturalistic’ landscaping developed under the influence of Capability Brown and his various rivals, existing woods might be incorporated into the planting, often coming to form part of the perimeter belts. The use of woods in these ways tended to encourage the perpetuation of regular coppicing, in order to provide a solid mass of vegetation to close a view or to provide a sharp and continuous outline.

Such was the enthusiasm for traditional forms of woodland management that, well into the nineteenth century, completely new areas of coppice-with-standards were sometimes planted by large landowners, especially in Norfolk. Moreover, where these were established in long-enclosed areas, where many of the ‘ancient woodland indicator’ herbs were present in adjacent hedgerows, they can often contain abundant quantities of plants like dogs mercury and bluebells, and even rarer species like wood anemone. As a result, some woods planted in this period can be hard to distinguish from ‘genuinely’ ancient woods (Stone and Williamson 2013; Barnes and Williamson 2015, 122-33). Lopham Grove in south Norfolk comprises coppice of hazel, ash, maple and hornbeam, under standards of oak (Q. robur) with some ash and the occasional sweet

144

Figure 28. Primrose Grove and Old Grove in Gillingham, Norfolk, did not exist when the tithe award was surveyed in 1840, but they are nevertheless included in the Ancient Woodland Inventory – not unreasonably, given their coppice structure and the range of plants they contain. They are among many examples of ‘pseudo-ancient woodland’ which have been identified in Norfolk. chestnut. The ground flora contains no fewer than six ‘ancient woodland indicators’: dog's mercury, wood spurge (Euphorbia amygdaloides), wood sedge (Carex sylvatica), primrose (Primula vulgaris), bluebells (Hyacinthoides non-scripta), and early purple orchid (Orchis mascula). Yet a map of 1720 shows the entire site as completely unwooded (Arundel Castle Archives 5H3 18). Not all such woods display quite such a range of ‘indicators’ - Blake’s Grove and Ladies Grove at Gawdy, planted between 1734 and 1783 (NRO Ms 4567 and 4568), comprise oak standards over mixed coppice of ash, hornbeam and maple, and have a dense ground cover of dogs mercury but few other ‘indicator’ herbs. Nevertheless, a number are surprisingly rich in such plants. Primrose Grove and Old Grove in Gillingham – conjoined areas of ancient woodland which were planted between 1840 and the 1880s – comprise ash, hazel and hornbeam coppice and a ground flora includes dog’s mercury, primrose, dog-violet (Viola reichenbachiana), hard shield-fern (Polystichum aculeatum) and wood speedwell (Veronica montana) (Figure 28) (Barnes and Williamson 2015, 218-9).

Changes in the density of farmland trees

The fashion for planting new species – other than oak, ash and elm – in woods and plantations is also apparent, although to much more limited extent, on farmland. On the Ingilby Estates near Harrogate in Yorkshire, for example, while oak and ash remained by far the most common species, beech, ‘Dutch’ elm, pine, black cherry, lime, horse chestnut and hornbeam were, from the 1780s, also

145 planted as hedgerow standards, in part for their ornamental qualities (Dormor 2002, 109). In 1827 a sale of timber on a farm in Coltishall in Norfolk – described as ‘principally Hedge Row Trees’ – included beech, ‘abele’, ‘red’ poplar and sycamore, albeit much outnumbered by oak and ash (NRO SPE 235, 316 x 1); while the ‘large quantity of superior Hedgerow Timber Trees’ advertised for sale at Bacton in Norfolk in 1868 included, in addition to 100 oak and 100 ash, 150 ‘sycamore, elm, beech and spruce’ (Norfolk Chronicle 14 March 1868). Nevertheless, oak, ash and elm continued to be the principal farmland trees, in all areas. In the late 1840s a survey of 1,041 trees (excluding stands and poles) on the Evans-Lombe estate in mid Norfolk, for example, shows that they together made up around 84 percent of the total, although this figure included some old pollards: the remainder mainly comprised unspecified ‘pollards’, and willow and alder. Sycamore, beech, walnut, birch and aspen and poplar, were also present but together contributed only 3 percent of the total: pines and firs were present but only as ‘poles’, presumably nurses in copses or small plantations (NRO HNR 112/19). Oak and ash similarly made up virtually all the hedgerow timber sold at Wallington in the same county in the 1880s and 90s (NRO MC 111/48, 581 x 5):13 at Park Farm Saham in 1906 (NRO RQG 226 489 x 4); at in 1905; (NRO RQG 178); and at Sandringham, Castle Rising, Weasenham and Hargham in 1926 (NRO BL/BG 5/3/22). A survey of 478 trees made on the Heydon estate in 1911, probably a list of timber to be felled, comprised 39 percent oak, 49 percent ash and 6 percent elm; the remaining 6 percent included maple, wetland species like poplar, alder and willow, but only small quantities of more exotic planting – six sycamore, five beech, three horse chestnut and a single larch (NRO BUL 11/118/1).

More striking than any alterations in the composition of farmland timber were the complex changes in tree densities which occurred in the course of the late eighteenth and nineteenth centuries. We might expect that the rapid progress of enclosure in the period after 1760, with the advent of enclosure acts, would have served to significantly increase the numbers of hedges and farmland trees in former open-field townships, as well as in areas previously occupied by open common land. In fact the situation is more complicated: enclosure may in the short term have added trees to the landscape, but other influences ensured that, over the course of the period, there was often little increase and sometimes a net decline. The evidence from Northamptonshire, where large amounts of open field were enclosed in this period, makes the point well. The parish of Luddington for example lay almost entirely open until its enclosure by parliamentary act in 1807, and in 1749 contained around 517 trees (Boughton House archives): the First Edition Ordnance Survey map from the 1880s depicts only 634. The parish of Woodforde in the same county was enclosed in 1764 and a particularly detailed map of 1731 shows 1,040 trees (Boughton House archives): just over a century later the Ordnance Survey recorded around 1,530 – an increase, it is true, but only from 1.1 to 1.7 trees per hectare. Broughton parish had only 529 trees when mapped in 1729, then almost entirely open: it was enclosed in 1786, but the Ordnance Survey shows only slightly more, around 683 (Boughton House archives). A broadly similar pattern of long-term decline is apparent in many early- enclosed townships in the county, already entirely without common land and open fields by c.1760. Of the fourteen places in the county with seventeenth- eighteenth- or early nineteenth-century maps which appear to depict, with at least an attempt at accuracy, the disposition of farmland trees, two (Quinton and Rushton) show a similar density to that depicted on the Ordnance Survey 6-inch maps from the 1880s (NHRO Map 1349; private collection) but seven (Newton le Willows, Elkington, Nobottle, Papley, Cranford, Halse and Loddington) suggest an overall decline, ranging from 24

13 A handful of elm, beech, sweet chestnut and Lombardy poplar are also mentioned in the accounts. 146 percent to 82 percent, and averaging 48 percent (Boughton House archives; British Library AddMss 78136A; British Library AddMss 78143; NHRO Map 2221; NHRO Map 1388; NHRO 4320; NHRO T154). Sometimes this was mainly a consequence of the removal of free-standing timber from pasture closes (as at Papley and ); more usually it was related to the reduction in the length and numbers of field boundaries; but sometimes (as at Elkington and Nobottle) boundary patterns remained largely unchanged, and the loss was a direct consequence of a reduction in the numbers of trees present in hedges. Such declines in tree numbers were not universal. In six places – Overstone, Benefield, Hannington, Grafton, Brockhall, Deene and Deenethorpe, and Fawsley – the numbers of farmland trees appear to have increased in the period between the earliest surviving post-medieval map, and the late nineteenth, mainly though a rise in the number growing in existing hedges (NHRO 5264; NHRO Inclosure Plan 4; Boughton House archives; NHRO map 5704; NHRO Bru Map 8; NHRO Nap 853). What is striking is that five of these places were the heartlands of large estates, containing large parks and mansions - Overstone, Grafton, Brockhall, Deene and Fawsley. Here the number of farmland trees increased by 84 percent, 158 percent, 195 percent, 81 percent and 10 percent respectively (only one out of the nine places where decreases, or little change, is indicated - Cranford St Andrew – lay close to a major country house). The overall density of trees in the countryside in these five places was further increased by the presence of avenues and parklands. At Deene and Deenethorpe, for example, the number of hedgerow trees within the mapped area appears to have increased from 274 to 497 between 1738 and the 1880s, but the overall increase in tree numbers – including trees in parkland and avenues – was from 274 to over 1,800. The cartographic evidence thus suggests that in general tree numbers saw either little increase, or a significant decline, in the former ‘champion’ areas of Northamptonshire in the period between the mid eighteenth century and the late nineteenth, except in some of the heartlands of large estates, where there was often a significant rise. Written surveys similarly suggest a similar picture. The density of 10.6 trees per hectare on John Darker’s scattered estates in Gayton, Tiffield, Kislingbury, Milton, Litchborough and Upper Heyford in 1791 had fallen to around 2.5 by the 1880s; that of more than 9.6 on the Duke of Powis' properties in Upper and Nether Heyford, Glassthorpe and Newbold in 1758 to around 2.7 (NHRO ZB 1837; NHRO YZ 2183). Armston lay towards the margins of the Boughton House estate and was already fully enclosed in 1749, when a survey recorded 1,721 trees (Boughton House archives). By the 1880s there were 738 – a decline of 57 percent. Like the maps, moreover, written surveys suggest that the numbers of hedgerow trees held up better in parishes closer to the centre of an estate. In Warkton, next to Boughton Park itself, the same survey lists 1,350 trees, whereas the Ordnance Survey shows nearly 2,000, although a proportion of these stood in avenues.

The number of trees recorded in Northamptonshire by the Ordnance Survey 6-inch maps is, in fact, not significantly greater in some parishes than the numbers indicated by pre-enclosure sources from the seventeenth and eighteenth centuries, ranging from around 1 tree per hectare, averaged across a parish, to around 4, these figures only being exceeded in the vicinity of some great houses. It is true that, for the reasons we have noted, such sources probably underestimate, possibly by a factor of 20 percent, the numbers of trees in the landscape: but they certainly imply that, following a rise in the seventeenth and eighteenth centuries associated with the early enclosure of open fields, tree numbers in the Northamptonshire countryside had fallen back again by the end of the nineteenth century, at least in places lying at a distance from a major country house. By the late nineteenth century the key factor structuring variations in farmland timber densities within the county, other than the presence or absence of a great landowner, appears to have been land use, with the more

147 arable east of the county exhibiting average densities on the First Edition 25-inch maps of between 1 and 2.5 trees per hectare, the pastoral areas of the far west usually between 3 and 4 per hectare (Figure 30). The date at which a parish was enclosed may have been a contributory factor: there are indications that the later the enclosure date, the lower the number of farmland trees depicted on the late nineteenth-century Ordnance Survey maps. Unfortunately, a high proportion of lateenclosed places were located in the more arable south and east of the county, ensuring that date of enclosure is hard to isolate, as a factor, from land use. This said, when adjacent parishes are examined in the more pastoral north west, a pattern is often clear. Potcote, Litchborough and Cold Heigham were enclosed in the early sixteenth century, c.1711 and 1812 respectively, and according

Figure 29. ‘Heat’ map showing variations in the density of farmland trees in Northamptonshire in the late nineteenth century, as depicted on the 1: 10,560 OS maps. to the Ordnance Survey maps had densities of c.3.3, 2.5 and 2.3 trees per hectare. It should be emphasised that Figure 29, and the three other ‘heat’ maps presented here, should not be taken entirely at face value. Variations in farmland tree densities are not merely a function of how thickly trees are crowded in hedgerows, but also a consequence of field size, the quantity of woodland in sample squares, and the presence or absence of ornamental parks. Nevertheless, they do provide a reasonable guide to variations in farmland tree densities across the counties in question.

In Yorkshire the First Edition 6-inch Ordnance Survey maps were surveyed earlier than in the Midlands and south, in the 1850s, and this allows us to compare the densities shown on eighteenth or early nineteenth-century maps, with the situation in both the 1850s and in the 1890s, when the second edition 6-inch map, and First edition 25-inch, were surveyed. When Warmsworth in the Vale of York was mapped in 1726 around two-thirds of its area still lay largely open, with three great fields. The enclosed third of its area was partly located in the vicinity of the village and partly in the

148 south of the township, where there was a network of hedged pasture fields (Don DD/BW/E11/7). By the 1850s all the open arable had disappeared through piecemeal enclosure, and the number of trees had increased, rising from 687 (1.8 per hectare) shown in 1726 to just over 1150 (3.1 trees per hectare) in the 1850s. The increase is related, to a large extent, to the proliferation of hedges although it is noteworthy that most of the trees present in the 1850s were growing within areas already enclosed by 1726; there were rather fewer trees in the boundaries created after this date. But more importantly, by the end of the nineteenth century – when the second edition 6”, and the 25”, were surveyed – the number of trees had fallen back again, by around 10%, to c. 1050, or c.2.8 per hectare.

A similar pattern – of a modest rise through the eighteenth and early nineteenth centuries, followed by a decline – also occurred in Vale townships which were already largely or entirely enclosed when first mapped. At Sinderby in the Vale of Mowbray a map of 1778 suggests a density of around 0.6 trees per hectare, which increased significantly, to 1.7 per hectare, by the 1850s; but by the end of the century it had fallen back to a mere 1 per hectare (Nrth ZIQ). At Thormanby, also in the Vale of Mowbray, apparent densities of 2.5 per hectare shown on an undated seventeenth-century map had risen to 3.4 in the 1850s, but then declined, almost to their previous level, by the 1890s (Nrth ZDS M 2/1). In both cases, the main concentrations of trees remained in the vicinity of the villages and along the principal routeways. At Skeeby, a third Vale of Mowbray township, the 0.5 trees per hectare apparently recorded on a map of 1779 had risen to around 2.9 by the 1850s, but had fallen back to 1.2 per hectare by the 1890s (Nrth ZMI). In some Vale townships late nineteenth-century losses were on a huge scale: at Birdforth, 75 percent of the trees shown on the 6-inch map of the 1850s had disappeared by the end of the century, although even here changes in overall numbers largely continued, or perhaps accentuated, an existing pattern, with concentrations of trees near settlements and along the main routeways. To some extent, but only to some extent, the late nineteenth-century decline was associated with significant changes in the pattern of field boundaries.

Similar patterns can be seen in areas outside the Vales. The density of c.2.2 per hectare in the Dales township of Leyburn in 1730 (Nrth ZBO (M) 1/2) had risen to 4.5 in the 1850s, before apparently declining to 3.7 per hectare by the 1890s, again with little overall change in distribution but, once again, some reduction in the density of field boundaries; a similar pattern can be seen in other Dales townships, like Walburn and Downholme (Nrth ZAZ (M) 1; Nrth ZBO (M) 1/5). In Newburgh in the Howardian Hills the 0.6 trees per hectare suggested by a map of 1605 (Nrth ZDV) had increased to around 2 per hectare by the 1860s, before falling back to around 1 per hectare by the end of the century; in nearby Brandsby, the 0.8 per hectare indicated by a map of 1746 (Nrth ZQG IV 1/1) had risen to over 2 by the mid nineteenth century, before declining to c.1.5. At Beadlam, on the margins of the North York Moors, a map of 1785 (Nrth ZEW M 13) suggests a density of just under 0.9 trees per hectare, which rose to around 2.5 per hectare in the 1850s, before declining to around 1.6 by the end of the century; while at nearby Pockley, a seventeenth-century survey (Nrth ZEW IV 6/4/6) suggests a density of 0.6 trees per hectare while a map of the same township made in 1785 (Nrth ZEW M 14a) suggests 1.1 per hectare. This had risen to 3 per hectare by the 1850s before falling back, in the now familiar manner, to 2.2 by the 1890s.

149

Figure 30. The density of farmland trees in Yorkshire in the late nineteenth century, as depicted on the 1: 10,560 OS maps.

Not every Yorkshire township for which we have information shared in this pattern. Healaugh and Catterton, in Tadcaster, for example, were already largely (c.84 percent) enclosed when mapped in 1718, at which time around 2,344 trees are shown, or 1.7 trees per hectare (Lds WYL68/63). A timber survey of 1751 suggests a rather higher density, of 2.8 per hectare (Lds WYL68/33) but by the time the Ordnance Survey first edition 6” was surveyed in the 1850s this had already fallen back to 2,929, or 2.1 per hectare: and in this case, the 25” map from the 1890s actually shows an increase, to 2.7 per hectare. On the whole, however, the evidence suggests a modest rise in timber densities which continued into the early nineteenth century, followed by a decline. It must be emphasised, moreover, that although in terms of percentage gains and losses these changes might appear significant, they were less dramatic than was general in Northamptonshire. In most parts of the Yorkshire numbers simply rose from around 1.5 trees per hectare to around 2.5 per hectare in the period leading up to the mid nineteenth century, before falling back once again to around 2 per hectare by the end of the century.

The rise in the numbers of farmland trees in Yorkshire in the course of the seventeenth and eighteenth centuries was presumably caused by the proliferation of hedgerows together with an increased attention to planting on the part of landowners, as the value of timber in this poorlywooded county rose. But by the end of the eighteenth century, and at an increasing rate as the nineteenth century progressed, the rate of planting tended to fall off. Tuke in 1800 observed that ‘a neglect of planting trees in hedgerows, and of a proper management of those which are now growing there, is very evident’, and that ‘the axe is often heard, but the planter is seldom seen’ (Tuke 1800, 191). In part this may have been because – in arable areas especially – there was an increasing emphasis on agricultural production, in an age obsessed with agricultural improvement. In addition, it may reflect the fact that timber imports through Hull rose from 1135 loads (of 50 cubic feet) in 1758 to 8260 loads in 1768, and to nearly 21000 loads in 1790 (Jackson 1975, 21) – although 150 imports of worked wood – deals for example – began to stagnate, perhaps as result of the establishment of new sawmills in Hull.

Whether the chronology of change implied in Yorkshire would also be evident in Northamptonshire, if this was blessed with a similar range of Ordnance Survey maps, is uncertain. What is clear is that, at similar if not identical rates, the numbers of farmland trees in both counties declined in the later eighteenth and nineteenth centuries. Such reductions pale into significance, however, when

Figure 31. The density of farmland trees in Hertfordshire in the late nineteenth century, as depicted on the 1: 10,560 OS maps. The south of the county, towards London, was still better endowed with farmland trees than districts further north. compared with those experienced in old-enclosed areas in the south of England and East Anglia, to judge from the evidence from Hertfordshire and Norfolk. The average density of farmland trees on the boulder clays in the north east of Hertfordshire thus appears to have fallen from around 15 per hectare in the later eighteenth and early nineteenth century, to around 7 per hectare by c.1880; and on the London clays in the south from 25 to around 10 (Figure 31). These figures, as we have noted before, need to be treated with some caution, not least because trees in settlements are almost certainly under-represented by the Ordnance Survey. Nevertheless, there can be no doubt that the density of farmland trees fell to a much greater extent than in the counties just discussed. Similar changes occurred in Norfolk. On the boulder clays in particular tree densities appear to have collapsed in the course of the later eighteenth and nineteenth centuries, from around 16 per hectare to around 8, and in the north east of the county from 25 to 7. At a local level, reductions were on an even greater scale. The 531 trees mapped by Henry Keymer on a property at next Mileham in 1764 (NRO WIS 138) had been reduced, by the 1880s, to around 80; the 1,096 trees recorded on the farm in Scarning by the same surveyor in 1761 (NRO BCH 20) had by 1884 fallen to a mere 63. To some

151 extent this reduction was associated with an increase in the size of fields and a concomitant reduction in the length of field boundaries. On the Beeston property mapped by Keymer, for example (NRO WIS 138), average field size rose from c.1.4 to 3.6 hectares: indeed, the field pattern was largely redrawn between the two dates, an irregular field pattern partly replaced by a new, rectilinear arrangement reminiscent of the landscapes created by parliamentary enclosure. This said,

Figure 32. The density of farmland trees in Norfolk in the late nineteenth century, as depicted on the 1: 10,560 OS maps. even where changes in boundary patterns were more limited, and where there was less increase in field size, reductions could be massive. At Scarning for example there was only a modest increase in average enclosure size, from 1.4 to 2 hectares, yet tree densities fell by nearly 85 percent.

There are many eighteenth- and early nineteenth-century references to field amalgamation, the realignment of boundaries and the removal of hedgerow trees, in the old-enclosed areas of East Anglia and the south of England, a process which was part and parcel of the ‘agricultural revolution’ and associated, in particular, to the conversion of pastures to arable (Williamson 2002, 91-7). Small, oddly-shaped fields were inconvenient for ploughing, tall hedges and hedgerow trees shaded out the crop and robbed the soil of nutrients, while wide outgrown hedges took up large areas of potentially productive land. Hedges were often drastically cut back, or grubbed out and replanted with hawthorn, and the number of trees within them was often substantially reduced. The rector of Rayne in Essex typically observed in his tithe accounts of the 1790s how on one farm in his parish ‘the fields were overrun with wood’, but ‘since Mr Rolfe has purchased them, he has improved them by grubbing up the hedgerows and laying the fields together’ (Williamson 2002, 95). Randall Burroughes, a gentleman farmer in Wymondham on the Norfolk boulder clays, was at the forefront

152 of this war on the ‘traditional’ landscape, the details of which are carefully recorded in his farming journal from the 1790s. Every winter his men were busy stubbing out hedges and taking down trees. In the last two weeks of 1794, for example, Burroughes described how ‘Elmer & Meadows began to through down & level an old bank in part of the pasture between little Bones & Maids Yards’; reported that ‘the men were employ’d in stubbing a tree or two for firing & other odd jobs’; described how ‘some ash timber’ was cut down; and how ‘the frost continued very severe so much so that ... the men employed in throwing down old hedgerows found the greatest difficulty in penetrating the ground with pick axes’ (Wade Martins and Williamson 1995, 48-9). Thick, outgrown hedges were hacked out and replaced with others, straighter and neater in form, or else removed entirely, as small pasture fields were converted into large arable closes.

A government enquiry into the state of the nation’s timber supplies was initiated in 1791 and included the question: ‘Whether the Growth of Oak Timber in Hedge Rows is generally encouraged, or whether the grubbing up of Hedge Rows for the enlarging of fields, and improving Arable Ground, is become common in those Counties?’ (House of Commons Papers 1792). The answers received from the principal groups questioned - purveyors of His Majesty’s dockyards, selected timber merchants, land surveyors and chairmen of Quarter Sessions - were to some extent contradictory. It was reported that the removal of timber and hedges was limited or non-existent in some districts - Cheshire, Durham, parts of Yorkshire. But in eastern and southern counties especially they were generally thought to be ‘frequent’, ‘becoming common’, or the ‘general practice’. Respondents (and indeed, the question itself) associated high rates of removal, of both hedges and the trees within them, with arable as opposed to pastoral husbandry. One typically noted that ‘The county of Hertfordshire consists chiefly of Land in Tillage, and by clearing the Hedges of all kinds of Trees they admit of plowing to the utmost Bounds of their Land’. More emphatically, a Suffolk respondent stated that ‘Much Timber and the Improvement of Arable Land are incompatible. Arable land in Suffolk is improved, and therefore timber is lessened’ (Lambert 1977, 749).

Underwood, particularly Blackthorn Bushes, in Hedge Rows that spread Two or Three Rods wide, is the true nursery of Oak Timber, but such Rows are a dead Loss and Nuisance in a well cultivated Country. England possessed in the past Age a great Plenty of Oak. Why? Because Cultivation was in a barbarous State. It is the Improvement of the Kingdom ... that has brought about the very good and proper Diminution of Oaks; and it is to be hoped that the Diminution will continue, for if it does not, the Improvement of our Soil will not advance (House of Commons Pappers 1792).

Not surprisiginly, new hedges planted after the 1790s – such as those establihsed by the great wave of parliamentary enclosure that removed almost all remaining open fields, and most commons, during the Napoleonic Wars – are generally sparsely-timbered, as well as surrounding fields that were often larger than those created by medieval or early post-medieval enclosure (Figure 33).

153

Figure 33. Contrasting tree densities in Hertfordshire in the 1880s. Left: south Hertfordshire, a landscape of small fields, early enclosure, and numerous farmland trees. Right: north Hertfordshire, with large fields created by late enclosure, and patches of surviving open-fields.

The management of trees: the decline of pollarding and changes in felling age

While the government may have worried about the loss of timber in Hertfordshire, Norfolk and other old-enclosed counties, because of the implications it had for the construction of ships and thus national defence, the real targets of the agricultural ’improvers’ were clearly pollarded trees. Indeed, it is noteworthy that in most of the districts studied, in the Midlands and south at least, the reduction in overall tree numbers is roughly comparable, in percentage terms, with the proportion of farmland trees which had formerly been managed as pollards: that is, most although by no means all of the reduction is explicable in terms of the wholesale removal of the latter. By the end of the eighteenth century agricultural writers seem to have shared a universal hatred of pollarding (Pettit and Watkins 2003). The dense, spreading heads of pollarded trees spread a particularly deep pool of shade, and they were looked on by improvers as relics of backward peasant agriculture, although it has also been argued that they additionally offended ‘polite’ taste because they were as ‘unnatural’ as the topiaried trees in formal gardens, which were now out of fashion amongst the rich and educated (Thomas 1983, 220-21). ‘Let the axe fall with undistinguished severity on all these mutilated heads’, urged Thomas Ruggles in 1796 in the Annals of Agriculture (Ruggles 1786, 180). As William Marshall put it: ‘We declare ourselves enemies to Pollards; they are unsightly; they encumber and destroy the Hedge they stand in (especially those whose stems are short), and occupy spaces which might, in general, be better filled by timber trees; and, at present, it seems to be the prevailing fashion to clear them away’ (Marshall 1796, 100-101). Anthony Collet described how Suffolk landlords were, by the 1790s, giving their clayland tenants leave to ‘take down every pollard tree that stands in the way of the plough’ (Young 1797, 57). We have seen how eighteenth- and early nineteenth-century surveys from these old enclosed areas commonly show more than two thirds of the trees as pollards. By the middle of the century figures were significantly lower. On an unnamed estate in north Norfolk in 1835 only 38 percent of the trees were pollards; of the 1, 041

154 trees described in a survey of Evans Lombe estates in mid Norfolk in 1850, only 20 percent were so described (NRO HNR 112/19 Timber book). Estate accounts from Earsham, and elsewhere suggest the sale of many hundreds of pollards, ruthlessly culled from hedges, in the early decades of the nineteenth century (NRO HNR 149/2; MEA 3/549, 658). By the end of the century pollards are hardly mentioned in documents at all, in any context. It is true that pollarding continued, on a casual basis, well into the twentieth century. The diminutive size of some oak and ash pollards found in south Norfolk suggests that on some farms new pollards may have been established as late, perhaps, as the 1920s (Barnes and Williamson 2011, 82-6). But for the most part, this aspect of traditional management had disappeared by the end of the nineteenth century.

There is little doubt that this change was closely connected not only with agricultural improvement, and rising food prices, which made farmers and landowners less tolerant than they had been on hedgerow trees and the problems they brought in terms of shading and nutrient loss. It was also the consequence of changes in regional fuel economies. We noted in Chapter 2 the wide range of organic fuels, including peat and various forms of vegetation cut from heaths and other ‘wastes’, which were employed in early modern England: the use of coal use was intensive close to coalfields, but was otherwise only widely employed on the coast, or in towns on navigable rivers. The spread of coal use across the whole of England was in part a consequence of the steady expansion of output from the coalfields, but also of an extension of navigable waterways, with the development of the canal network. This began in earnest with the completion of the Bridgewater canal (designed to serve the coal mines at Worsley) in 1761, ‘the movement reaching a crescendo in the ‘mania’ of 1789-93’ (Cossons 1987, 256). The canal mania was directly paralleled by the expansion in the numbers of turnpike trusts, which was likewise most rapid in the period between 1750 and 1850 (Bogart 2005; Albert 1972). All this gradually allowed the use of coal as both a domestic and an industrial fuel to spread throughout the country, although in a few districts only the arrival of railways in the first half of the nineteenth century sounded the final death knell for traditional forms of firing. Coal only began to be used on the Bedfordshire brickfields, for example, following the arrival of the local rail lines: the Duke of Bedford’s Crawley Kiln for example was only fired with coal from around 1845, the Duke’s steward, Thomas Bennett, recalling in 1869 how ‘Furze used to be grown for a demand for brickmaking, but this fell off some years ago’ (Cox 1979, 44).

In both Hertfordshire, and Norfolk, larger farmers as well as landowners had access to coal at a reasonable price by the early decades of the nineteenth century, and pollards were removed accordingly: a development in part motivated by practical economic and agrarian considerations, and in part, perhaps, by fashion and social emulation. In this age of agricultural revolution, farmland was increasingly viewed simply as a place to grow food, not as a source of fuel. In a similar way, the spread of coal use may have acted as a further encouragement to the enclosure of and destruction of heath land – formerly a major source of firing - and of the remaining areas of common wood- pasture, in places like south Hertfordshire.

155

Figure 34. Average size (in Hoppus feet) of oak (above) and ash (below) of groups of timber trees either felled, or selected for felling, recorded in timber accounts, sales catalogues and similar documents, from early 1700s to early 2000s. Both species show a sharp rise in average felling age from the mid nineteenth century.

But it must be emphasised that while the wholesale loss of hedgerow pollards explains most of the overall reduction of tree densities in England, it does not explain all of it. In Yorkshire, as we have noted, the numbers of farmland tree also fell, even though there had been few pollards there. To some extent, farmland trees of all kinds were being seen, by the start of the nineteenth century, as ‘incumbrances’ that needed to be removed: ‘the farmers say 'knock them up, no corn will grow near them' ‘ (Tuke 1800, 196). Timber should be grown in woods or plantations, or imported from the vast forests of the Baltic. Farmland was for growing food. This in large measure explains why the decline in pollarding was not, as we have seen, matched by a concomitant decline in coppicing. There was sufficient demand in many districts to ensure that coppiced woods were still actively managed into the 1890s and beyond, long after pollards had been largely stripped from the landscape around them. Indeed, some of the agricultural improvements of the late eighteenth and earlier nineteenth centuries may in fact have provided new outlets for coppice poles: much of the material employed in the ‘bush drains’ which were installed in such numbers on heavy land in this period came from woods, Randall Burroughs of Wymondham in Norfolk, for example, regularly

156 sending to nearby Ashwellthorpe Wood for ‘faggot-wood for draining’ (Wade Martins and Williamson 1995, 75-8).

As well as being characterised by a rapid decline in pollarding, the nineteenth century also saw – albeit, at a slightly later date – important changes in the management of timber trees. The ages at they were felled, in both woods and on farmland, began to increase from the middle decades of the nineteenth century, in all districts, leading in turn to a gradual rise in the average age of timber. Up until the middle of the nineteenth century trees had continued, as in earlier periods, to be felled young. Most oak and ash trees felled before the 1840s or 50s contained less than 10 cubic feet, and the mean average was below 20 cubic feet – few trees other than pollards lived much beyond their sixtieth year. Some of these trees may have been thinnings from plantations but the majority were not, and it is striking that in 1787 Coggeshall in Essex, just outside the study area, oaks averaging less than 13 cubic feet were described as having ‘done growing’ (ERO D/DHt E13). By the 1870s and 80s, in contrast, the average volume of trees felled both in woods, and on farmland, was between 40 and 50 cubic feet (Figure 34). Many more timber trees were being allowed to grow up to, and beyond, maturity. To some extent this change may reflect new aesthetic preferences on the part of large landowners, for a landscape dominated by full-grown standard trees. But for the most part it probably indicates the fact that imports of softwoods were now on a large scale, allowing pine, larch and the like to be used as minor structural timbers, reducing the demand for small-diametre oak, elm and ash. More importantly, perhaps, it reflects the development of industrial sawmills, powered by water or steam, which allowed larger timber to be processed with relative ease; combined with the development of better roads, canals and ultimately railways, which allowed larger timber to be transported more easily than before. The declining importance of bark after c.1850, as new forms of tanning using chromium salts replaced traditional methods, may also have been a factor, for small trees produce better quality bark than larger ones and a greater quantity in proportion to their volume. Whatever the precise combination of causes, the change was a dramatic one, and one which had a profound impact on the age structure of tree populations.

Conclusion

The century or so after c.1780 saw profound changes in the character of England’s trees and woodlands. Their causes were complex and multiple, but were ultimately related to three key developments: the rise of large landed estates, the agricultural revolution and large-scale industrialisation. Firstly, there were significant alterations in both the character and the distribution of woodland. Woods tended to decline in districts of fertile clay soil, but at the same time to increase in areas of marginal, acid land – formerly occupied by common heaths and moors, prior to enclosure – as well as in formerly ‘champion’ districts, such as the Wold of east Yorkshire. In the latter areas new woodland was planted, in particular, in and around the parks laid out around the homes of large landowners, although this was an association which could be found, albeit to a lesser extent, in all parts of the country. Surviving areas of wood-pasture, mainly located on common land, were enclosed and destroyed on a large scale within the counties studied, so that only fragments remained by the end of the nineteenth century. Above all, there was a noticeable tendency for woodland to decline where it was most abundant, and to increase in districts where it had formerly been rare. The overall impact of the changes of the late eighteenth and nineteenth century was thus to spread woodland more evenly across the country. Indeed, this development was perhaps more important than any overall increase in woodland area in this age of the ‘great replanting’: for averaged across the four areas studied, the area covered by woods and plantations rose from

157 perhaps just under 3 percent to just under 4 percent of total land area. England, in short, continued to be a sparsely-wooded country.

Secondly, there were important changes in the management and species composition of woodland. Many of the new woods established in this period took the form of plantations, without a coppiced understorey; and, while long-established timber trees like oak continued to be prominent in estate planting schemes, they were now accompanied by new species. Some of these, like larch and spruce, were introductions from abroad; others, while native or long-naturalised, were now planted across a much wider geographical range. England’s woodland thus grew noticeably more diverse, although it is also important to emphaise that, in most areas, coppice-with-standards woodland continued to be managed on traditional lines well into the second half of the nineteenth century. There was continuity, as well as change, in management.

Thirdly, there were significant alterations in the character of farmland tree populations. They continued to be characterised by a similar range of species, but there were major changes in management, with an almost complete cessation – by the end of the nineteenth century – of the practice of pollarding, and, after c.1850, a sharp rise in the age at which timber trees were felled, something which also affected trees growing in woods and plantations. Above all, in all districts there was a marked reduction in the density of farmland trees, a decline which, in the old-enclosed areas of Hertfordshire and Norfolk especially, reached catastrophic proportions.

That the character of England’s trees and woodland experienced such a far-reaching and rapid transformation in the later eighteenth and nineteenth centuries should occasion no surprise. As we have repeatedly emphasised, in all periods trees and woodlands, far from being elements of a ‘natural’ landscape, were intensively exploited for practical agrarian and economic ends, and their character and composition shaped accordingly. As England was transformed by industrialisation, the large-scale modernisation of farming, and unprecedented levels of population growth, it was inevitable that its trees and woods would experience corresponding levels of change. Even more dramatic developments were to occur in the course of the twentieth century, as social, economic and technological systems were further transformed.

158

6: Trees and Woodland since c.1880

Contexts

In the later eighteenth century, and the early and middle decades of the nineteenth, the progressive modernisation of agriculture, changes in patterns of land ownership and industrialisation had been major influences on the numbers and management of farmland trees. Now, from the late 1870s, both the profitability of agriculture and the viability of large landed estates began to be challenged. Farming began to slide into a long period of depression, principally caused by the expansion of the American railway network into the prairies of the mid-west, so that European markets were flooded with cheap grain. No longer kept artificially high by the operation of the Corn Law, repealed in 1846, wheat prices were halved between 1873 and 1893, while those for barley and oats fell by a third. Following a brief period of stabilisation, a further intense depression occurred after 1896, this time affecting not only arable farmers but also livestock producers, as cheap meat began to be imported on refrigerated ships from the New World and Australia (Perren 1995; Perry 1974). In an increasingly globalised world, British agriculture could not compete. Fortunes recovered during the First World War but there was then a further slump, with only a partial and patchy recovery through the 1930s. Only in the period following the outbreak of war in 1939 did British agriculture return to long-term profitability, ushering in a period of intensive farming which has continued, more or less, to this day.

One important consequence of the agricultural depression on rural tree populations was the impact that it had on the fortune of large and medium-sized landed estates. Agricultural rents plummeted through the 1880s and 90s, especially in arable districts, and with some fluctuations remained low until the Second World War. At the same time, large estates were financially challenged by a raft of other factors, most notably Death Duties, introduced in 1894 and raised to 15 percent by Lloyd George and subsequently, in 1919, to 40 percent on estates valued at more than £200,000 (Thompson 1963, 325-30). Apart from the opportunistic fellings of trees in woods and on farmland that these difficulties often engendered, through the first four decades of the century many established landowners were gradually obliged to sell their ancient properties: great mansions were demolished or converted to new uses, parklands ploughed, and the wider estate land fell into the hands of a new class of owner-occupier – often former tenant farmers – with a radically different range of land-use priorities. In particular, traditional management of woodland usually came to an end and integrated systems of forestry, embracing farmland trees as much as woods and plantations, went into steep decline.

The agricultural depression, and the financially challenged position of large landowners, had other effects, as we shall see. But not all changes in rural tree populations in the first half of the century can be attributed, directly or indirectly, to problems in the agricultural economy. Social, economic and technological changes all ensured that much marginal land – surviving commons, or poors’ allotments (that is, land allotted at the time of enclosure either to be used directly by the poor, for fuel cutting or grazing, or leased to raise money to provide them with coals (Birtles 2003, 194-6) - was no longer used intensively for grazing, or as a source of fuel and materials, and began to regenerate to woodland. More importantly, the state’s role in land management increased as that of large estates waned, a phenomenon most dramatically manifested in 1919 in the establishment of a state forestry authority, the Forestry Commission, empowered to purchase poor land for tree planting and responsible in many districts for a massive increase in the area of conifer plantations.

For many decades there had been mounting concern in government circles about the limited area and poor condition of the nation’s woodlands, and these were greatly intensified by the effects of wartime blockade when the shortage of pit props for coal mines had threatened the war effort. In 1916 the Prime Minister, H. H. Asquith, appointed the Forestry Sub-Committee of the Ministry of Reconstruction:

To consider and report upon the best means of conserving and developing the woodland and forestry resources of the , having regard to the experience gained during the War (Ryle 1969, 25).

The committee proposed that, over the following eighty years, no less than 1,770,000 acres (c.72,000 hectares) of land should be planted with trees. One and half million acres of this was to be by direct state purchase and planting, the rest through private enterprise, or by joint public/private schemes. The committee’s recommendations were accepted by the cabinet, in 1918 an interim Forestry Authority was established and in 1919 the Forestry Act authorised the appointment of eight Forestry Commissioners, empowered to acquire land (Ryle 1969, 25-39). The areas targeted for planting were principally former heaths (or recently abandoned farmland earlier reclaimed from heath) and upland moors. This in turn ensured that the greatest changes in woodland area in the course of the twentieth century occurred in districts which possessed the largest amounts of such marginal land. Of the four counties studied here, Forestry Commission planting thus had the greatest impact in Norfolk, and especially in the arid, sandy in the south-west of the county (and extending into Suffolk); and in the North Riding of Yorkshire. Not all of the new woods which appeared in the first half of the twentieth century were planted by the Forestry Commission, however. Where landed estates survived intact, as they often did, they likewise displayed an enthusiasm for new forms of forestry, at the same time as their enthusiasm for perpetuating traditional modes of woodland management declined. Plantations represented a form of diversification, on land no longer profitably farmed, while it also provided cover for game, itself in many cases a form of economic diversification.

The Second World War and the decades following the peace saw agriculture return to profitability. First the national government, and latterly the European Economic Community, introduced a range of subsidies which were aimed at increasing levels of production. But in addition, increased levels of state support were associated with other changes in the practice and organisation of farming, many of which had an important impact on tree populations. Increases in farm size, the widespread adoption of tractors and combine harvesters, and a shift in eastern districts away from mixed farming to purely arable husbandry led to the wholesale loss of hedges and hedgerow trees, and the grubbing out of many areas of ancient, semi-natural woodland. Between 1946 and 1970 around 4,500 miles of hedge were destroyed each year in England and Wales, with the greatest losses occurring in the eastern counties (Dowdeswell 1987, 118; Muir and Muir 1997, 225-6). Agricultural intensification, combined as we shall see with other factors, also saw a further shift away from the economic management of farmland trees, and a further concentration of forestry in woods and plantations. All this was compounded by the arrival in Britain, in the late 1960s, of Dutch Elm disease. From a small number of initial points of infection this spread at a rate of around eight miles a year, and by the late 1980s the disease was well-entrenched in all the areas studied, leading to the almost complete destruction of elm as a tree although not, of course, as a suckering shrub (Rackham 1986, 240-47; Brasier and Gibbs 1973). Elm had, as we have seen, long been one of the three most

160 numerically important of farmland trees, in some districts rivalling oak. Its loss had a catastrophic impact on the landscape. It was not the first outbreak of the disease, as we shall see in the following chapter, but it was much more severe than any tree disease seen in England before or – at the time of writing – since.

Although farmland timber thus declined sharply through the 1950s, 60s and 70s, there was at the same time a continued increase in the area of woodland in most parts of England, as scrub developed into woodland on unmanaged commons and other marginal land and, increasingly, on derelict industrial sites. The period from the 1980s saw further changes. Government policy began to increasingly prioritise conservation and amenity use of the countryside, with grants for planting trees and hedges under Countryside Stewardship and other schemes. The grubbing out of hedges was effectively curtailed with the new Hedgerow Regulations of 1997 (later amended in 2002), which finally brought hedge removal within the planning process and laid down a range of criteria on which permission for removal could be denied. From the early 1980s, moreover, the policies of the Forestry Commission turned decisively in the same direction, leading to a renewed emphasis on deciduous planting, and on multi-use forestry with an important recreation and conservation element. In short, the last century has seen a raft of complex and interconnected social, economic, technological and ideological changes which have had far-reaching and often surprising impacts on rural tree populations. Changes in the character and extent of woodland

Norfolk

Looking at the period since c.1880 as a whole, the most striking and in some ways surprising development has been the marked increase in the area of woodland in England, a fact which is perhaps insufficiently appreciated even by the informed public. To a significant extent this has been a consequence of the large-scale plantings made by the Forestry Commission, especially in Yorkshire and Norfolk, which served to intensify the concentration of woodland in areas of marginal land already begun in the eighteenth and nineteenth centuries. In Norfolk, the Commission began acquisitions in Breckland in 1922, with the purchase of a small area near Swaffham, followed soon afterwards by that of 3,149 acres (1,275 hectares) of the Elveden estate, just across the county boundary in Suffolk. The Downham Hall estate (4,944 acres) was bought in 1923; (6,208 acres, c.2,500 hectares) and part of Beechamwell (822 acres, c.330 hectares) in the following year (PRO/TNA: FC 54386/2; FC 374/24). In 1925 the estate and parts of the Croxton and estates were purchaed, followed by the Weeting estate in 1926, further portions of the Didlington and Croxton estates in 1927, together with 4,299 hectares of the Croxton Hall estate (much of it already leased by the Commission). In 1930 the West Harling estate (3,077 acres, 1,245 hectares) was acquired, as well as a further portion of Lynford and 2,025 acres (820 hectares) of the Hockham Hall estate (FC archives: acquisition reports; FC Annual Reports: Edlin 1972, 7; Dannat 1996, 21-5)). After this there was a lull in acquisitions. In 1931, in the face of mounting economic crisis, the Commission’s budget was reduced, land purchases were concentrated elsewhere in the country, and only small quantities of land (such as a property of 176 acres at Feltwell, acquired in 1933) were purchased in Breckland (FC archives, acquisition report, Feltwell; PRO/TNA FC L3/3, Vol.1; L3/3/15; L3/3/9; L 3/1/1; Barnes 1984, 51-7). Not all the land planted by the Commission in the 1920s and early 30s was purchased freehold, it should be noted. Some was leased for periods of between 120 (enough time to obtain two successive crops of timber) and 999 years.

161

Most of the land acquired by the Forestry Commission in Breckland thus came in large blocks. Owners of estates on this poor land were eager to sell, given the impoverished state of their finances as agricultural rents plummeted on this marginal land. And the sums paid for the land were low, with £2 4s 8d per acre being paid for the Downham Hall estate; £3 10s for Weeting in 1926; and just over £3 per acre for Croxton in 1929 (FC archives, Acquisition Files; PRO/TNA FC L3/3, Vol.1; L3/3/15; L3/1/1). Such prices reflected the derelict or semi-derelict condition of many properties, something which is very clear in the various ‘acquisition reports’ drawn up by Commission officers prior to purchase. That for Croxton Park, for example, made in 1929, described the estate as ‘partly heath and partly low grade light arable or pasture land which has passed, or is about to pass, out of cultivation’ (FC archives, acquisition file). By the outbreak of the Second World War the Commission's holdings in Breckland, including those on the Suffolk side of the county boundary, amounted to 59,000 acres (23,000 hectares) (Skipper and Williamson 1997, 23. In addition, it undertook some planting in other parts of the county, especially on the outwash gravels to the north of Norwich, and in the area around Swanton Novers in the centre-north of the county.

Figure 35. The distribution of woodland in Norfolk in c.1940, as recorded by the Land Utilisation Survey (after Mosby 1941). The large areas in the south west of the county are the plantations established over the previous two decades by the Forestry Commission.

The numbers of trees planted by the Commission grew steadily through the 1920s, peaking in 1927, when no less than eight million trees were established on 3,700 acres in the main area of Breckland (including land in Suffolk), and a further 700 acres planted on the Commission’s land further north, around Swaffham (FC Annual Reports; FC Working Plan, 1959). Between 1924 and 1929, an average

162 of 2,226 acres (909 hectares) of plantation was established annually in Breckland. But just as striking as the scale of the new planting was the range of species employed. The advent of the Forestry Commission saw the final triumph of ‘plantation’ silviculture in England. The new plantations invariably lacked a coppiced understorey, and the emphasis had now shifted decisively towards conifers. Even in the eighteenth century, as we have seen, some plantations composed entirely of pine, spruce or larch had been established by landowners, especially on the more marginal land, and this increased in the course of the nineteenth century. Now, influenced in part by Continental practise, in part by the poor quality of the soils being planted, and in part by a desire for rapid returns, the emphasis was firmly on the Scots pine, although from the start the Corsican pine was also employed in certain locations, and began to be used in preference to Scots from the 1930s. Small areas of Douglas fir, European larch, silver fir, western red cedar, western hemlock, maritime and lodgepole pine were also established. This said, to begin with substantial number of indigenous hardwood trees – principally oak and beech - were also planted in Breckland by the Commission. In 1935, for example, 1,186 acres (480 hectares) of conifers were planted in Breckland, but as many as 428 acres of hardwood trees, principally oak and beech (Skipper and Williamson 1997, 30-31). In this sense, there was rather more continuity with established estate forestry practice than is sometimes assumed. But on poor, sandy soils like those of Breckland oak and beech grew more slowly than pine. They were also vulnerable to the sharp spring frosts to which the district was prone, and to the browsing of deer, the numbers of which rose sharply in the afforested areas as planting progressed. The numbers of deciduous trees being planted thus declined after 1935 (FC Annual Reports). Most of those that were established were placed along the sides of the principal roads, in part as a fire prevention measure and in part to assuage growing public opposition to the ‘serried rows of conifers’ which were transforming the formerly open landscape of heaths and ‘brecks’ (Skipper and Williamson 1997, 65-7; Chard 1959; Tennyson 1939, 76-7).

Although the new plantations established by the Commission in Breckland and elsewhere were the most significant addition to the area of woodland in Norfolk, some contribution was also made by private landowners. Where large estates remained intact their owners often continued – sometimes on a large scale – the tree-planting schemes of their predecessors. Private owners were encouraged, from 1923, by the grants which were made available by the Commission of £2 per acre for conifers and £4 for hardwoods (Ryle 1969, 265-6). But even before this they planted on some scale. The Bayfield estate, for example, in the north of the county saw a sustained forestry campaign which began with its purchase by Alfred Jodrell in 1882 and continued until his death in 1929, all funded largely with income from his Lancashire cotton mills. Existing woods were restocked and linked with new planting, so that the park was increasingly enclosed by woodland to the north and east, with further plantations framing or closing the more distant views from the house (Wade Martins and Williamson 2008, 84-5). The prominence of box and rhododendron in the planting, here as on other estates, shows that one of the main intentions was to provide cover for game. Indeed, much planting in the period took forms which clearly indicates its primary role as game cover: small blocks, clumps, and narrow strips predominated, as on the Sennowe estate in mid Norfolk where Thomas Cook in the 1920s ‘added a large number of plantations to his estate, mainly ‘roundabouts’ in the centre of large fields, narrow belts on the sides, and triangular woods on the corners’ (Mosby 1938, 178). On light land especially thin strips of conifer plantation were widely established along field margins. In Breckland, the single lines of pines – originally managed as hedges – bordering the fields were increasingly augmented with further planting, to create the strips of pines now so characteristic of the district.

163

From the late nineteenth century private landowners seem to have concentrated more and more on the planting of conifers. Butcher in 1941, writing of nearby Suffolk, contrasted twentieth-century estate plantations, which were ‘mostly spruce, Douglas fir, Scots pine, and very mixed plantings with conifers in the majority’, with the ‘older woods’, which were mainly of ‘Scots pine, elm, poplar, beech … chestnut and oak’ (Butcher 1941, 332). Even on the heavier and more fertile land conifers were extensively planted. Burrell in 1914 noted in Norfolk that ‘The occasional planting of pure coniferous woods on heavy land, naturally adapted to oak and ash, is probably due to the need of raising fresh cover for game in the shortest time possible’ (Burrell 1914, 25). Nevertheless, given that extensive areas of woodland already existed, most estates – and certainly those located on the heavier or more fertile soils in the county - continued to be dominated by deciduous planting, and usually by indigenous species. Even on the Gayton estate in west Norfolk, on the poor sandy soils of the Isleham and Newport 4 Association, although large quantities of conifers were purchased at the start of the century (including 550 spruce, 550 Scots pines, and no less than 5,550 larch in 1901) a survey by the newly-formed Forestry Commission in 1924 recorded that 84 percent of the trees in the estate woods were oak and ash, and 13 percent alder, with only small quantities of beech and pine. As time went on more and more landowners increased the proportion of conifers in their woodlands, but seldom to the extent seen in the Commissions plantations (NRO MC 547/81-104, 768x8).

To these deliberately planted woods and plantations, we need to add the many areas of woodland in the county which developed during the twentieth century through natural regeneration on neglected marginal land. Some of this was private heath, but a significant proportion comprised surviving areas of common land or ‘poors’ allotments’. Clarke in 1918 believed that ‘land accessible to the public’, mainly in the latter two categories, amounted to some 11,324 acres (4,582 hectares) in the county (Clarke 1918, 295). The intensity with which such land was exploited had begun to decline even before the start of the agricultural depression. The new ‘improved’ breeds of sheep widely adopted in the nineteenth century were less well adapted to rough grazing than the old, traditional variety, the Norfolk Horn, and other and better kinds of feed were now available for livestock. More importantly, with the widespread adoption of coal as a domestic fuel, turf, heather and peat were no longer regularly cut for firing. As early as the 1870s it was reported that the local population has ceased to cut turfs on Whitwell Common because ‘the houses and fireplaces of the commoners are unsuitable for the burning of turf’ (Birtles 2003, 209). Bird, describing the use of ‘Common’ (really a fuel allotment made to the poor at enclosure) in 1909, noted that the cutting of flags, peat and furze for firing had largely ceased, and while fodder cutting was still continued the use of the common for grazing had much declined (Bird (1909). Here, as in many other places, ‘The poor are too poor to buy a cow’. Indeed, there were insufficient animals to prevent the colonisation of much of the area by gorse. Clarke in 1918 noted a number of places where traditional forms of cutting and mowing, for peat, reeds, sedge or litter, persisted. The Oxburgh Fuel Allotment, for example, was mown each year on a date fixed by the villagers: ‘the man employed puts the litter in heaps and receives about 1sh. for each heap; these are numbered and the ownership for each is settled by drawing lots’ (Clarke 1918, 298). On Marsham Heath and both Calluna and Erica were cut ‘for besoms or sink brushes’. Even bracken was still mown in some places, as at Congham, North Wooton, and Great Bircham (Clarke 1918, 299). But in general, he reported, traditional uses were in decline.

164

By the time of the First World War many of these areas were being invaded by scrub, ‘dominated by whitethorn, blackthorn, furze, and rubus’ Clarke 1918, 305). Birch and pine, and sometimes oak, were also invading by this stage: their seedlings were ‘common on many heaths’, often coming in from neighbouring plantations established by local landowners (Clarke 1918, 308). By the outbreak of the Second World War many such areas had begun to be occupied by woodland, and regeneration continued apace in the post-War period. Wetland areas in the county were also affected by the decline in traditional forms of management. The fens in the Broadland valleys had traditionally been cut for peat or mown, for rough hay and litter, but by 1939 Boardman was able to describe how:

A great change has taken place over a large portion of the river valleys about here. Thirty years ago all the large acreage of rough marsh covered with sedge rush and grass was mown for what was called marsh litter. The best was made into hay, or chaffed and used for feed locally or sent to London for bus horses and cows, while the rougher stuff was put into the bullock yards to be trodden into manure. Now, since acres upon acres of this material remain uncut and the vegetation gets into such a terrible tangle, the marshes have to be burned. Alders, birch and sallows are taking possession … When the marshes were mown regularly all the young trees were kept under... (Boardman 1939, 14).

Dereliction of this kind contributed significantly to the steady expansion of woodland in Norfolk in the course of the twentieth century, although nothing like as much as deliberate planting, especially by the Forestry Commission. That expansion was inexorable. In 1895 it was thought that there were 21,000 hectares of woodland in Norfolk, around 4 percent of its land area (Board of Ag Returns 1896, 36). In 1905 the figure was estimated at 23,926 hectares, or 4.5 percent (Board of Ag. Statistics, 1910, 56); in 1913 it was thought to be 23,523 hectares (Board of Ag. Stats, 1914, 92); and in 1924 the Forestry Commission reckoned it as 19, 798 hectares, or 3.7 percent. (FC 4th Annual Report, 1924, p.25) These differences probably reflect variations in survey methods and classification, and it was only during subsequent years that the woodland appears to have risen significantly, the Commission estimating a total of 27, 256 hectares (5.2 percent of county land area) in 1932-5, and the Land Utilisation Survey in 1938 no less than 33,144 hectares (6.3 percent) (Mosby 1938; FC 15th Annual Rept, 1935). In 1980 42, 697 hectares of woodland were recorded in the county (8 percent of land area); by 1997 the figure had reached 52,046 (9.7 percent); and by 2015 it was 54,410 hectares, or 10.3 percent (Forestry Commission 2002). These post-War increases reflect, not the establishment of more Forestry Commission plantations, but the planting of numerous small woods by landowners for conservation and sporting purposes, amenity woodland, and the continuing process of natural regeneration on marginal land and abandoned industrial sites. Overall, the increase in woodland in the county has been remarkable, more than doubling in area during the course of the twentieth century, and arguably returning it to the levels of the late Middle Ages.

Yorkshire

As noted above, the Forestry Commission did not only target areas of heath, or of redundant arable land reclaimed from it, for their afforestation schemes. Upland moors were also extensively planted. In Yorkshire, the largest plantations were established in the North Riding, mainly on the edges of the North York Moors. Planting began in 1924 and by 1938 no less than 9, 010 acres (3, 246 hectares) of new woodland had been established, with the largest concentration in Allerston Forest, now known as Dalby Forest (2,209 hectares), in the Hambledon Hills. There were other extensive tracts of afforested land, near Ampleforth (1,077 acres, 436 hectares); Rosedale (1,697 acres, 687 hectares);

165 and Langdale (129 acres, 52 hectares); as well as 658 acres (266 hectares) at Arkengarthdale in the Pennines (Wooldridge 1945, 383-4). Elsewhere in Yorkshire very little land was acquired (none at all in the East Riding), and newly-planted, by the Commission in the inter-War years, although a number of ancient woods were transformed by re-planting, mainly with conifers, either by private owners or by the Commission itself, including the 850-acre (343-hectare) Bishop Wood near Selby in the Humberhead Levels, which occupies a relatively low-lying site and was stocked with Scots and Corsican pine (Beaver 1941, 121).

Since the Second World War, in sharp contrast to the situation in Norfolk, there has been a further significant expansion in the Forestry Commission holdings, to more than 19,000 hectares (Forestry Commission 2002). The overwhelming majority of the effected land, however, is likewise concentrated around the margins of the North York Moors, with further expansion in the Allerston/Dalby and Langdale areas, the planting of Compton Forest a little to the west, and much smaller-scale afforestation (Boltby Forest, Silton Moor) on the western margins of the moors. Whether established before or after the War, all the Commission’s plantations here were dominated by Scots pine, Sitka spruce and Japanese larch, with smaller quantities of other coniferous species such as Lodgepole pine, Douglas fir and Corsican pine. Only in south Yorkshire do significant numbers of decisuous trees appear to have been planted in this period.

Figure 36. The distribution of woodland in the North Riding of Yorkshire, c.1940 (after Wooldridge 1945). Note the concentration of woodland on the southern flanks of the North York Moors.

Although, as in Norfolk, the spontaneous development of scrub woodland (especially on abandoned industrial sites) and planting by private landowners may have contributed to the increase in woodland area in the course of the twentieth century, the main factor in Yorkshire was unquestionably this afforestation of upland moors by the Forestry Commission. Before the commencement of planting in the mid 1920s the overall woodland total for the county appears to

166 have remained relatively stable, variously estimated at 3.6 of land area in 1895, 3.8 in 1905, 3.7 in 1913 and 3.6 in 1924; but it then rose significantly, to 4.2 percent, by the mid-1930s, an increase mainly accounted for by an expansion of Commission plantations in the North Riding (Board of Ag Returns 1896, 36; Board of Ag. Statistics, 1910, 56; Board of Ag. Stats, 1914, 92; FC 4th Annual Report, 1924, 25; FC 15th Annnual Rept, 1935). Subsequent comparisons of woodland areas are difficult to make, due to changes in administrative boundaries, but woodland expansion was clearly on a significant scale; in part a consequence of further planting, as described, by the Forestry Commission, although in part also the result of the bushing-over of marginal and industrially derelict land, and of private planting. By 1980 the figure had probably risen to above 6 percent of the old area of the three Ridings; by 1999 it was around 7 percent; and it is now 7.2 percent, a total of 113,328 hectares (Forestry Commission 2002). Although boundary changes, differing definitions of ‘woodland’, and other problems preclude any very accurate comparison over time, it seems likely that the area of woodland in Yorkshire has more than doubled over the last century, a similar scale of increase to that seen in Norfolk.

Northamptonshire and Hertfordshire

Even in counties in which Forestry Commission activities have been more limited, and directed more towards the replanting of existing woods than to the establishment of new ones, the increase in woodland area in the course of the twentieth century has generally been on a significant scale, although sometimes more delayed than in the cases of Norfolk and Yorkshire. In Northamptonshire, few new plantations were established by the Commission, although many existing woods were restocked with conifers following the transfer of the Crown woodlands into Commission ownership in 1923 (the proportion of woodland classified as ‘coniferous’ in the county rising from 8.6 percent in 1924 to 38 percent by 1980) (Ryle 1969, 145; FC 4th Annual Report, 1924; Forestry Commission 2002). The area of woodland in fact remained largely unchanged throughout the first half of the century, estimates for 1895, 1905, 1913, 1924, 1939 and 1947 all suggesting that between 3.9 and 4.5 percent of the county area was occupied by woods and plantations (Board of Ag Returns 1896, 36; Board of Ag. Statistics, 1910, 56; Board of Ag. Stats, 1914, 92; FC 4th Annual Report, 1924, 25; FC 15th Annnual Rept, 1935; Beaver 1943, 362-5). Some new areas of woodland were certainly planted by large estates in this period - the records of the Boughton estate, for example, show an active interest in forestry in the early years of the twentieth century, with 1,000 oak, 100 ash, 200 larch, 1,150 beech and 200 sycamore being planted in 1908 alone (Boughton House archives). But a number of areas of woodland were also lost, not least as a consequence of the development of an ironstone quarrying industry in the county from the late nineteenth century, much of which was concentrated in well-wooded areas – formerly part of Rockingham Forest – around Kettering and Corby - Nicola Orchard has shown (Orchard 2007). New planting, and the destruction of existing woods, more or less cancelled each other out in the early and middle decades of the century, and in 1979, according to another Forestry Commission survey, the total area of woodland within the county was still virtually unchanged, at 11,600 hectares – 4.5 percent of land area. By 2000, however, it had increased significantly, to 14,500 hectares, around 5.7 percent of the land area; and by 2015 it had reached 7.1 percent (Forestry Commission 2002). This expansion was a result of grant schemes, countryside stewardship agreements, and game and amenity planting, together with much replanting and regeneration over abandoned industrial sites, especially derelict ironstone quarries.

The area under woodland has thus returned, more or less, to that of coppiced woods and wood- pastures combined in c.1750, before the enclosure of the county’s forests - a remarkable recovery.

167

Figure 37. The distribution of woodland in Northamptonshire in c.1940, as recorded by the Land Utilisation Survey (after Beaver 1941).

Even in well-wooded Hertfordshire there has been a significant expansion in woodland area in the period since the late nineteenth century, in spite of the considerable levels of urbanisation and suburbanisation which the county has experienced. Even where development was most intense in the first half of the century, woodland often survived while farmland was lost: Oxhey Woods, close to the county boundary with Middlesex, is now a local nature reserve featuring a mixture of ancient and secondary woodland, with areas of remnant heath and wood-pasture, and noted for its wild service trees and bluebells. Yet it is hemmed in by housing estates and suburbs on all sides except the south, where there is a large golf course. Other examples of such survival include Wick Wood in St Albans and Albans Wood in North Watford. By the second half of the century, established woods were being consciously incorporated into the designs of New Towns like Stevenage, where Whomerley Wood, Monk Wood, Shackleton Spring, Abbots Grove, Ashtree Wood, Great Collens Wood, Collens Leg Wood, Pestcotts Spring and Pestcotts Wood were all utilised as public open spaces (Munby 1977, 248-9). More importantly, most of Hertfordshire, even today, remains unaffected by suburbanisation, and many of its more rural parts are owned and occupied by wealthy exiles from London who have sought to preserve woodland for its sporting and amenity value to a

168 greater extent, perhaps, than landowners in some more truly rural districts. To such established areas of woodland, however, more were added in the course of the twentieth century.

Figure 38. The distribution of woodland in Hertfordshire in c.1940, as recorded by the Land Utilisation Survey (after Beaver 1943).

Estimates of the proportion of the county’s land area occupied by woodland made in the late nineteenth century display some variation, beginning with 6.1 percent In 1895, rising to 6.5 percent in 1905, falling back to 5.2 percent in 1913 5.2, but recovering to 6.4 percent by 1924 (Board of Ag Returns 1896, 36; Board of Ag. Statistics, 1910, 56; Board of Ag. Stats, 1914, 92; FC 4th Annual Report, 1924, 25). Almost certainly, such variations reflect, once again, differences in recording and definition more than true changes in woodland area. By 1949 the figure had risen to 6.6 percent, perhaps suggesting some increase. By 1959 a more decisive rise had occurred, to 7.5 percent; in 1980, the figure was 7.7 percent; and by 1998 it had reached 9.3 percent, although this includes the category of ‘open land scattered with trees’ (such as public parks) and the real figure is probably nearer 9 percent (Forestry Commission 2002). This, it should be noted, is well above the 13,100 hectares, or c. 8 percent of land area, for woodland and wood-pasture combined which we have suggested for the middle decades of the eighteenth century. The most recent woodland census, for 2015, estimates that there are 17,770 hectares of woodland in the county, around 10.8 percent of the land area. As always, we need to add a note of caution: definitions of woodland are shifting and subjective, and there have been a number of minor changes to the boundaries of the county. Nevertheless, there seems little doubt that the area of woodland in Hertfordshire almost doubled in

169 the course of the twentieth century, a surprising development in a county already, in relative terms, well-endowed with woodland.

This marked increase is partly explained, as in Northamptonshire, by the establishment of woods by landowners, often benefiting from Countryside Stewardship or Small Woodland Grant scheme. But it was also a consequence of the fact that, compared with Northamptonshire, many large commons in the county escaped the impact of parliamentary enclosure. Increasingly redundant in agricultural terms, and primarily valued for their recreational potential in this suburbanised and densely-settled county, many became covered with scrub in the middle decades of the century and, by the 1960s, were regenerating to secondary woodland. On the high Chiltern ridge, Berkhamstead Common, Pitstone Common and Aldbury Common are still shown as open ground on Ordnance Survey maps from the 1960s, but are now wooded; the same is true of most of the smaller commons scattered across the Chiltern dipslope, such as Chipperfield Common, Bernards Heath in St Albans or, to a lesser extent, Nomansland Common on the boundary between St Albans and Wheathampstead. These examples alone would have served to increase the amount of woodland in the county by some 600 hectares, nearly 0.4 percent of its total land area.

All the counties studied thus saw significant increases in woodland area in the course of the twentieth century. Before the Second World War this was largely the consequence of Forestry Commission planting, leading to a massive increase in the proportion of woodland entirely coniferous in character; in the middle and later years of the century, it was augmented in some districts by the gradual bushing over of marginal (usually common) land, or derelict industrial land, mainly with deciduous trees. Over recent decades, woodland planting has been carried out mainly by private owners, or as part of local government schemes of industrial reclamation, and has increasingly involved deciduous trees, as has – since the 1982 ‘Broadleaves in Britain’ conference and the 1985 policy statement – that within woodland owned or managed by the Forestry Commission. In Hertfordshire the proportion of woodland occupied by conifers thus fell from 20 percent in 1980 to 18 percent in 1997; for Northants the figures are 38 percent and 22 percent; while in Norfolk, where Commission plantations were most extensive, it declined from 52 percent to 36 percent (Forestry Commission 2002).

One other variety of woodland established in the course of the twentieth century, locally important in both Norfolk and Hertfordshire especially, is worth noting briefly, because of its visual impact on the landscape: small-scale plantations of willows and poplars of various kinds. In the early years of the century, inspired in part by an article in the Journal of the Royal Agricultural Society, large numbers of ‘cricket bat willows’ (Salix alba var. caerulea) were planted, either in lines around field margins or as small plantations. A survey in 1927 noted that most of these fast-growing trees were already approaching maturity, but warned that few had any real commercial value: 30.5 percent were the ‘wrong type of willow or otherwise faulty set’; 31 percent had been planted too close to produce good bat timber; most of the rest suffered from a range of diseases; and only 5 percent were judged likely ‘to make good quality bats’ (Pratt 1927). The failure of the willows to come up to scratch became something of a joke amongst landowners, Lilias Rider Haggard for example consoling herself over the loss of some willow sets with the thought: ‘My only consolation is that when it comes to the point no willow is ever found to be quite the right sort to make bats! (Haggard and Williamson 1943, 17). Nevertheless, small plantations of caerulea continued to be established throughout the twentieth century, mainly in East Anglia and parts of the Home Counties, not least because – as a Forestry Commission Bulletin explained in 1936 – ‘When properly planted and tended

170 this tree may attain the dimensions required by bat manufacturers in twelve to eighteen years’ (Forestry Commission 1936). In addition, the planting of small areas of hybrid willows began in the middle decades of the century, again principally in the south east and East Anglia, including Hertfordshire and Norfolk. This was principally for the production of matches, but also to supply basket-making companies, with Bryant and May Ltd taking the lead from the late 1940s, growing and selling planting stock to private growers, providing free advice on establishment and cultivation, and guaranteeing to purchase all timber of acceptable quality. The company also began its own planting programme in the early 1950s. By the late 1950s around 400 hectares were being planted with hybrid poplars (Populus serotina) by private owners each year across the country as a whole, mostly in the south east and East Anglia. From the 1960s, however, declining use of matches, and the increasing use of plastics to make baskets, led to a steady decline in planting. Nevertheless, there were by this stage more than 4,000 hectares of private poplar plantations, and 1,500 hectares directly established by the industry, together with some 1000 hectares planted by the Forestry Commission (Tabbush and Beaton 1998). Some planting continued, however, and many stands have been left, unfelled, to grow into senescence. They form a distinctive feature of the landscape in many areas, especially on floodplains.

The fate of ancient woodland

The management of coppiced woodland, most of it of ancient, semi-natural character, declined steadily from the later nineteenth century, albeit at different rates in different areas. Coal was now the standard domestic fuel everywhere, and industrial substitution and large-scale imports undermined the market for underwood products. In addition, as we have suggested coppicing had been sustained, in many cases, by a desire on the part of large landowners to provide cover for game, and was thus adversely affected by the declining fortunes of large estates. So far as the evidence goes, coppicing had already been largely abandoned in Yorkshire by the outbreak of the First World War. Less than 4 percent of the woodland was thought to be coppiced in 1913, and by 1924 only 2.4 percent; virtually no coppiced woodland was recorded in the whole of the East Riding, although this mainly reflects the fact that, largely devoid of woodland before the eighteenth century, relatively little woodland of coppice-with-standards type had ever existed there (Board of Ag. Stats, 1914, 92; FC 4th Annual Report, 1924, 25). In Midland and southern counties, in contrast, coppicing continued on a larger scale well into the inter-War years. In Hertfordshire, for example, it was thought that around 30 percent of woods were still being regularly coppiced in 1905, and it is noteworthy that woodland-related trades such as coopmakers and stavemaker widely appear as occupations in census returns in the early twentieth century (Cox 1914, 280). In 1913 the Board of Agriculture estimated that only 20 percent of the county’s woodland was still coppiced, but this may have been a conservative estimate because the same body in 1924 put the figure as high as 46 percent (Board of Ag. Stats, 1914, 92; FC 4th Annual Report, 1924, 25). This had again fallen to 27 percent by 1937-8, largely due to an increase in the proportion of woodland in the county categorised as ‘scrub’, ‘felled’ or ‘devastated’ (Cameron 1941, 318-9). Even this rather lower area was no longer, for the most part, managed intensively: ‘With a few notable exceptions … coppice is not cut regularly, mature trees are allowed to deteriorate; where cutting takes place the whole wood is cleared and no provision made for re-stocking whilst natural vegetation is almost absent due to the ravages of rabbits’. ‘Where such woods are ‘managed’ at all they are managed primarily to supply foxes for the local hunts’ (Cameron 1941, 318). As late as 1967 it was estimated that coppice still accounted for 23 percent of the total woodland area, but it was cut on a casual basis for ‘small

171 items such as posts for making or repairing fences, logs for household fires, pea and bean sticks etc.’ (Gardener 1967, 166). In addition, some of the industrial malthouses in the east of the county continued to be fuelled with hornbeam cut from the local woods, rather than coke. In 1970 it was reported that Taylor’s of Sawbridgeworth still used ‘hornbeam faggots exclusively, of which immense stacks like haystacks are built close by’ (Branch Johnson 1970, 33).

In Northamptonshire coppicing continued on an even more substantial scale, with just under half the woodland described as ‘coppice’ or ‘coppice-with-standards’ in the census of 1924. In 1902 the eminent forester Wilhelm Schlich was commissioned to prepare a report on the woodland on the Boughton estate and suggested that much could be converted to high forest (Boughton House archives): it is noticeable, however, that he also set out an alternative method of managing the woods as coppice-with-standards. Schlich noted the advantages to game and foxes of dense underwood, and it is possible that the late continuance of traditional management in the county owes much to the fact that many of the estates with major holdings in the forest areas – including Boughton, and Deane – survived the various threats to landed property intact, and continued to devote much of the energy towards field sports. Even at the outbreak of the Second World War it is possible that well over a third of woodland in the county comprised managed coppice, but steady decline set in thereafter and by the end of the century the figure had fallen to around 2 percent (Forestry Commission 2002).

In Norfolk the Board of Agriculture census of 1913 suggested that only 11 percent of the county’s woodland was still actively managed as coppice (Board of Ag. Stats, 1914, 92), but there are grounds for believing that this was an under-estimate, and on some estates, such as Wolterton in north Norfolk, woods were still actively coppiced on a large scale into the 1940s (Wolterton archives, WOLT 3/1/16-19). As late as 1923 hurdles and wattles were still being made at Melton Constable, and other local industries still provided a market, especially brush-makers like S.D.Page and Britons, both based at Wymondham – although even they, in the inter-War period, gradually came to rely more and more on beech imported from France and Belgium and birch brought from Norway (Clark 1996, 14; Briton Brush Co. 1935, 6-7). But by the 1960s, certainly, coppicing had effectively come to an end in the county.

By this time, moreover, many areas of ancient woodland had been replanted with conifers, sometimes by the Forestry Commission - as with Bishop Wood in Yorkshire or Hevingham Wood in Norfolk (purchased in 1941) (NRO HNR 64/1) – and sometimes by private owners, as with Hockering Wood in Norfolk where, the oak timber having been largely removed following the break-up of the Morton Hall estate in 1923, large areas were restocked soon after 1930 with Douglas fir, larch, Scots pine, red oak and Turkey oak (Natural England archives, Peterborough). The scale of replanting intensified from the late 1940s with the development of the Commission’s Dedication Scheme. The conversion of ancient woods to ‘plantations on ancient woodland sites’ affected not only small and medium sized woods but also some of the largest examples, such as Brooke Wood (60 hectares), Haveringland Great Wood (92 hectares) and Hevingham Park (63 hectares) in Norfolk, or Scales Park Wood in north-east Hertfordshire. All in all, between the 1930s and 1973 nearly half the ancient woods that remained in Norfolk were either coniferised or grubbed out completely (English Nature 1992). In Northamptonshire, 11 percent was grubbed out and 65 percent coniferised, while in Hertfordshire the figures were 9 percent grubbed out and 41 percent coniferised. In Yorkshire

172 statistics are only available for the boundaries of the ‘new’, post-1974 , but were in total 6 percent and 38 percent respectively (Spencer and Kirby 1992).14

Where replanting did not occur, coppiced woods usually slid into a state of neglect. They were still often used as game cover but otherwise left largely unmanaged. The numbers of timber trees was reduced to varying degrees by opportunistic fellings, and the uncut coppice allowed to grow to canopy height. The consequent shading, denser than in traditionally managed woods and now uninterrupted by regular cutting, steadily reduced floristic diversity, significantly affecting the characteristic woodland flora in a manner which has been described and documented by a number of researchers (Rackham 2006, 532). In heavily shaded woods dog’s mercury often continued to flourish, but to the exclusion of other ‘woodland indicator species’; in the most heavily shaded examples the ground flora disappeared altogether across large areas of the woodland floor. Neglect may also have had an effect upon the shrubs and trees in the understorey, creating stand types which were rather different from those which had existed while management continued. We noted in Chapter 3 how hornbeam appears to have increased its importance both Hertfordshire and Norfolk during the later Middle Ages. In those ancient woods in which the species formed a component of the understorey, its prominence appears to have increased still further, so that many woods once characterised by mixed coppice in which hornbeam was a significant component, now consist almost entirely of outgrown hornbeam stools. Their dense shade has suppressed the growth of other underwood species, so that hazel, maple and ash only survive well on the margins of the wood, or beside the principal rides. The pollen sequence from Diss Mere, which lies within the area of south-eastern Norfolk characterised by hornbeam woods, thus suggests a substantial increase in the proportion of hornbeam in the locality in the course of the nineteenth and twentieth centuries, although the sequence is not closely dated (Peglar et al 1989). Documentary records also hint of such a change. The tithe files for Denton in Norfolk, drawn up in 1836, thus describe how the main wood in the parish, which can hardly be other than East Wood, ‘cuts into hurdles’, strongly suggesting that hazel was then more important than it is today (PRO/TNA IR 18/5878): the hornbeam which is now the wood’s main component is almost impossible to use for this purpose because it does not split or ‘reave’ easily. In a similar way, Hedenham Wood, another south Norfolk hornbeam wood, was described in the same source as producing ‘hurdles and thatching stuff’ (PRO/TNA IR 18/5861), and the late nineteenth-century owner described how a period of neglect had ensured that the coppice could not be used for ‘hurdle wood’, likewise implying that hazel was then a more prominent component than hornbeam (NRO MC 166/238). In this particular wood the dominance of hornbeam has continued to increase through recent decades. Oliver Rackham’s plot of the tree communities present here, made in the 1970s, appears to show a much more diverse wood than exists today (Rackham 1986b, 174). Some of the areas which he then mapped as ‘mixed hazel’ now seem to contain pure stands of hornbeam. Where – as here – portions of coppice have been brought back into management during the last few decades they already have a more diverse composition. Hornbeam woods are probably an extreme case of a wider phenomenon: the stand types in other kinds of ancient woodland have probably also been modified by a century or more of neglect.

14 Made up of North Yorkshire, 2 per cent destroyed, 33 per cent replanted; South Yorkshire, 18 per cent and 46 per cent; West Yorkshire, 6 per cent and 47 per cent:

173

Eighteenth- and nineteenth-century plantations have suffered similar fates to ancient, semi-natural woods. Many are either neglected and over-mature; have suffered clear-felling, often during the Second World War, frequently followed by invasion by sycamore; or have been replanted with conifers. Nevertheless, the extent to which conifers have triumphed over hardwoods in woods on plantations in the course of the twentieth century can be exaggerated. Only around 28 percent of all woodland in the four counties now consists of pure conifer plantations, and a further 14 percent comprises mixed conifers and broadleafs, although this represents a decline in the proportion of conifers from the peak in the late 1970s, when around 48 percent comprised some form of coniferous planting (Forestry Commission 2002). Where Forestry Commission holdings are limited, as in Hertfordshire, the proportion of coniferous woods is correspondingly less.

Farmland trees

It has been suggested that, because farming was a state of depression in the period between the late 1870s and the outbreak of the Second World War, the numbers of farmland trees in England increased once again, after an extended period of decline. Oliver Rackham in particular argued that ‘The period 1870-1951 was, on the whole, an age of agricultural adversity, in which there was less money to spend on either maintaining or destroying hedges. Neglect gave innumerable saplings an opportunity to grow into trees’ (Rackham 1986, 223). In fact, there are good grounds for believing that the opposite was the case, and that the decline in tree numbers, already underway in the nineteenth century, continued into the twentieth. There is certainly little reason to believe that saplings now had a greater chance of growing into timber, in the manner suggested. Hedges were still actively maintained in most districts (not least because outgrown hedges provided cover for the exploding population of rabbits) albeit now often by trimming rather than laying or coppicing, and Lilias Rider Haggard bemoaned in the 1940s how it was ‘difficult to get the hedgers to remember to leave oak and ash saplings unless they have gained sufficient height to be obvious. Down the hedge goes the sickle, taking in its deadly sweep many a hopeful little tree which would have graced the property in twenty year’s time…’ (Haggard and Williamson 1943, 21). More importantly, for a variety of reasons many existing trees were deliberately felled during the agricultural depression years. On large estates hedgerow trees were the landowner’s property, retained for profit or for the beauty they conveyed, and not the tenant farmer’s, for whom they were nothing but a nuisance. As agricultural profitability declined it was often difficult to find tenants when leases ran out, and landlords were now more sensitive than they had formerly been to complaints concerning the problems caused by mature trees in terms of the shade they cast, or the nutrients they robbed from crops. Lilias Rider Haggard typically described: A consultation about the always difficult question of tree cutting on the farm. This particularly affects the arable fields, where the farming tenant has cause for some complaint. Decided somewhat sadly that some dozen small oaks must come out before the sap rises, or next autumn when the crops are off (Haggard 1946, 73). In addition, the escalating financial difficulties experienced by landed families often led them to capitalise on their standing timber. In 1902 Rider Haggard noted the large-scale felling of hedgerow oaks in the area between Whissonsett and Wendling in Norfolk, commenting: ‘I think that ‘ere long this timber will be scarce in England’ (Haggard 1902, 506). Lilias Rider Haggard similarly described in the 1930s how ‘the wholesale cutting of timber all over the country is a sad sight, but often the owner’s last desperate bid to enable him to cling to the family acres…’, noting elsewhere: ‘The other

174 day, going past a well-known and well-loved place, I was hit like a blow in the face by a scene of complete desolation – every tree gone’ (Haggard and Williamson 1943, 97). When, as was often the case, large estates were broken up in the first half of the twentieth century, the purchasers of particular farms – often their former tenants – were particularly keen to sell much of the timber in order to help recoup the purchase price, as well as to improve the yields in the adjacent fields. To a very limited extent the character of changes in farmland tree populations in the Depression years can be ascertained by comparing the numbers of farmland trees shown on the Ordnance Survey 1:10,560 maps, surveyed in the 1880s and 90s, and on the RAF vertical air photographs of 1946: although given the problems with the former source, already discussed, and the difficulty of counting individual trees on the latter, this is not a particularly reliable procedure.15 On the boulder clays of south Norfolk an examination of sample grid squares seems to suggest that average densities of around 8 trees per hectare had fallen, by 1946, to around 4; while on the fertile loams in the north east of the county they fell from c.7 to around 3. In west Hertfordshire, on the dipslope of the Chiltern Hills, average densities of around 4.5 trees per hectare in the 1880s had fallen, by 1946, to less than 3; while on the heavy clays in the south of the county the decline was from 10 to under 5. Declines elsewhere in the two counties were generally on a comparable scale. While we should not make too much of these figures, which are also complicated by changes in the area occupied by small woods and copses, they do not suggest a major increase in farmland tree densities but instead, as we might expect, a continuing decline. Only in one district studied does the neglect of hedges appear to have led to a significant rise in the numbers of farmland trees, in the manner suggested by Rackham. In Breckland, in south-west Norfolk, enclosures made from heaths and open fields in the early decades of the nineteenth century were often defined by hedges of Scots pine. Some of these were probably never managed as true hedges, for they had already grown into lines of trees by the time the First Edition 6-inch map was surveyed in the 1880s. But most were still regularly trimmed and are depicted as boundary lines, rather than as lines of conifer symbols, on these maps. In many cases management probably ceased during the First World War, although Clarke in 1925 was still able to describe the importance of Scots pine in the district, ‘planted either in rows known as ‘belts’, or artificially dwarfed as hedges’ (Clarke 1925, 17). By 1946 most had become the lines of twisted trees, locally known as ‘rows’, which remain an icon of Breckland, adding many hundreds of farmland trees to the landscape (Barnes and Williamson 2011, 138-52) (Figure 39). With this notable exception, the decline in the number of farmland trees appears to have continued with little interruption through the first half of the twentieth century. It then accelerated in the 1950s and 60s, in large measure as a consequence of the widespread grubbing out of hedges. This is best documented in Norfolk, due to the work of Baird and Tarrant in the 1970s. Around 500 miles (800 kilometres) of hedgerow were destroyed in the county each year from 1946 to 1955, rising to around 2,400 miles (3840 kilometres) per year by 1962, and reaching 3,500 miles (5,600 hectares) over the next four years. By 1970 the rate had dropped back to around 2,000 mile a year (Baird and Tarrant 1970), and it declined more gradually thereafter. In all, between 1946 and 1970 around 13,000 kilometres (20,800 kilometres) of hedges in the county were lost, around 45 percent of the 1946 length. In Hertfordshire, especially on the more arable boulder clays in the east of the county, losses occurred on a similar scale, and by the 1970s Munby was bemoaning the emergence of

15 The Second and Provisional editions of the Ordnance Survey 1: 10,560 did not record farmland trees. The RAF 1946 vertical air photographs are kept at the National Monuments Record, Swindon. 175

Figure 39. Typical Breckland ‘row’ of contorted Scots pines, formerly a hedge, at Cockley Cley in Norfolk. ‘substantial areas of open landscape where the hedges have gone’ (Munby 1977, 168). Today this particular district – still largely rural, in contrast to the now much suburbanised south and west of the county - retains just over a third of the hedges that it had in the late nineteenth century, although the precise figure varies greatly from area to area.16 While in some places there are still around half the number in others the reduction has been on a substantial scale, as for example in parts of Aspenden where less than 15 percent of the hedges shown on the First Edition Ordnance Survey 6” map now remain. Developments seem to have been similar in the eastern, more arable parts of Northamptonshire, although in the more pastoral west losses have been more muted. In Yorkshire, likewise, the scale of loss seems to have been greater in arable districts, especially in the Vale of York. It is true that, in all areas, trees were sometimes retained when hedges were grubbed out, left to grow free-standing in the middle of fields. But this needs to be balanced against continuing losses of trees even where hedges survived. In the single-minded drive to increase agricultural production during the Second World War, and in the first three decades of the ensuing peace, hedgerow trees that died or were felled were seldom replaced, even on large landed estates

16 Based on a random sample of 10 1 kilometre by 1 kilometre squares, and on a generous definition of what constitutes a hedge. The figures are: 8.8 kilometres reduced to 3.8 (TL3429); 6.4 reduced to 0.7 (3427); 7.8 reduced to 2 (3826); 8.5 reduced to 0.46 (4125); 9.9 reduced to 5.6 (4319); 16.4 reduced to 5.3 (4418); 10.8 to 6.1 (4223); 6.1 to 1.3 (3718); 7.1 to 4.1 (3334); and 9.1 to 4.4 (3232). 176 running their own forestry departments. Timber was now to be grown in woods and plantations, not on farmland. The impact of these post-War changes ought to be reasonably clear from the survey of farmland trees carried out by the Forestry Commission in 1972, which included the three southern counties studied here, although not Yorkshire (Forestry Commission 1972). However, this survey was based on a two percent sample, comprising half-mile square plots distributed at five mile intervals along major roads – a method candidly described as ‘not sophisticated statistically’ even by the authors of the report, and which was robustly criticised by members of the Commission. The results for Norfolk, suggesting that average densities had now fallen to just under 2.5 trees per hectare, can fortunately be compared with the survey made by Norfolk County Council in 1976, which estimated that there was an average density of 1.52 per hectare (NCC 1976). This figure also has to be treated with caution, however, as it embraces wetland and former wetland areas which, even in 1946, had been largely devoid of trees – the Fens and the ‘Broads’. Excluding these, the average figure would be around 1.75 trees per hectare. Even these trees were generally in poor condition, the survey estimating that 48 percent of hedgerow oaks were mature and stag-headed, although the figure for ash was 19 percent and for elm only 9 percent. These findings are broadly in line with those in the 1972 Forestry Commission census, which estimated that 45 percent of oaks in Norfolk were stagheaded, the highest percentage recorded in the eastern counties. The average farmland tree densities suggested by the 1976 Norfolk survey are difficult to compare with those derived from the RAF 1946 aerial photographs, but suggest a less marked decline than might be expected, given the scale of hedgerow removal, but a significant one nevertheless, from perhaps 3.5 to 1.75 trees per hectare within the non-wetland parts of the county. Further losses were to follow, however, for the 1976 survey was carried out at around the time that Dutch elm disease had begun to affect the county, but before major losses of trees had occurred: indeed, 27 percent of the trees recorded in the county by the survey were elms. The next comprehensive survey, in 1980, by the Forestry Commission, not surprisingly indicates a further significant decline, for the average density was now 0.78 per hectare across the county as a whole. By 1998, when the next survey was carried out, it had fallen still further, to a mere 0.62 per hectare (Forestry Commission 2002). In Northamptonshire, the Forestry Commission 1972 figure, for what it is worth, indicates a rather higher density of hedgerow timber than in Norfolk, at around 3 per hectare, while in Hertfordshire the figure was roughly the same: the proportion of stag-headed trees recorded in both counties was less, although significant nevertheless. If we correct these figures by the kind of factor suggested by the 1976 Norfolk Survey they indicate average densities of only around 1.8 trees per hectare, a relatively small reduction on the figures suggested by the 1946 aerial photographs. In these counties, too, Dutch elm disease – combined with continuing ageing of trees, increased drainage, intensive arable land use and some further hedge removal – saw further attrition in subsequent years, the density of free-standing trees in Northamptonshire falling to 0.57 per hectare in 1980 and a mere 0.27 in 1997 (Forestry Commission 2002). In Hertfordshire comparisons across time are more problematic, due to the extent of suburbanisation, but in the still rural areas of the county (mainly in the east) the reduction appears to have been on a similar scale.

In Yorkshire the pattern of decline is even less easy to trace, due to changes in administrative boundaries, as well as by urbanisation. Nevertheless, there is little doubt that loss of farmland trees has occurred on a lesser scale than in the Midlands or south. As we have seen, for a variety of

177 reasons farmland tree densities in the three Ridings were already, by the late nineteenth century, considerably below those of the southern counties, averaging around 2 trees per hectare, albeit with much local variation. Even in 1997, they remained at or above 1 per hectare, on average, across the county.

Although there were thus, in all of the counties examined, significant declines in the numbers of farmland trees in the period after the late nineteenth century, what is striking is that before the advent of Dutch elm disease in the mid-1970s the overall ratio of the principal species within each county remained much as recorded in eighteenth- and nineteenth-century documents. In particular, the Forestry Commission unpublished survey of 1972 suggests that whereas oak continued to be the commonest farmland tree in Norfolk (52 percent), followed by roughly equal numbers of elm and ash, it was still the third most common in Northamptonshire, at 7 percent, after elm (41 percent) and ash (35 percent). In Hertfordshire, as in earlier centuries, the three were more evenly balanced, with 13 percent ash, 32 percent elm and 21 percent oak. The main change from earlier centuries, reflecting the developments in the nineteenth century already discussed, was an increase in the percentages of other farmland trees, which together accounted for 17 percent of the trees in Northamptonshire, 34 percent in Hertfordshire and 24 percent in Norfolk, this latter figure in part inflated by the development of the ‘pine rows’ in Breckland. Beech and sycamore were locally significant, although only in Norfolk did their combined number exceed ten percent of the total population. Scots pine, willow and poplars of various kinds account for most of the remainder. Although Yorkshire was not included in the 1972 survey, examination of trees present in the landscape today shows that – in the Vale of York especially – beech and sycamore became increasingly common as farmland trees in the course of the later nineteenth and twentieth century, especially on estates like Bramham. In terms of species composition, in other words, the later nineteenth and twentieth centuries were a period of both stability and change. The local ratios of the main farmland species remained as before, but the proportion of ‘minority’ trees increased. Farmland trees were becoming more diverse.

It must be emphasised, however, that for a number of reasons the statistics presented in the last few paragraphs, especially those concerning densities of farmland trees, need to be treated with considerable caution. One problem has already been noted: the difficulties involved in comparing tree densities at different periods when areas are urbanising or suburbanising. More importantly, the problem of what we actually mean by a ‘tree’, and how this varies between different sources, cannot be over-emphasised. Changes in the management of hedges in the course of the twentieth century create particular difficulties in this respect. When the Ordnance Survey First Edition 6-inch maps were surveyed in the late nineteenth century there was a fairly clear distinction between a hedgerow tree and the hedge that it grew in, even if the tree was fairly young, because hedges were generally kept low by plashing or cutting. In the twentieth century, and particularly since the Second World War, this distinction has become blurred, especially in arable areas. Many hedges, or sections of them, have been allowed to grow, unmanaged, into lines of bushes and trees, sometimes simply through neglect but often in well-intentioned attempts to increase tree numbers. Even where hedges are still regularly cut, now usually with a mechanical flail, a proportion of coppice stools of ash and, to an extent, other species like maple have often been allowed to grow – sometimes into full trees, but often into something intermediate in status. In addition, since the publication of the Countryside Commission’s New Agricultural Landscapes report in 1974 (Westmacott, R. and Worthington 1974) new forms of rural planting, intermediate between woods and hedgerows, have

178 been adopted in many places, especially small groups of trees in field corners and other situations where farming operations were difficult. These were widely established in the 1970s and 80s especially.

It is noteworthy that the more recent Forestry Commission surveys of farmland trees and woods have begun to include new categories of non-woodland tree - ‘groups’ (covering less than 0.1 of a hectare) and ‘linear plantings’ (i.e., strips of trees less than 50 metres wide). These have varied origins and character, the latter for example embracing outgrown hedges, linear copses established for game cover, and wide strips of trees planted besides modern roads or road improvements or naturally regenerating on abandoned railway lines. In many areas the numbers of trees present in these features is considerable, and serves to offset to varying extents the loss of individual farmland trees in the traditional sense. In 1998 the Forestry Commission estimated that there were 33,2600 individual trees growing in the rural parts of Norfolk, but no less than 1,341,800 in ‘narrow linear features’, and a further 894,000 in small groups, covering less than 0.1 hectare (Forestry Commission 2002). If these are included in calculations of farmland tree densities, the number rises from around 0.6 per hectare to 4.7 per hectare, actually above the average density apparently shown on the late nineteenth-century Ordnance Survey maps. In Northamptonshire the figure increases from 0.27 to 2.14, in this case below the late nineteenth century figure, although not by very much. Such calculations are difficult to make for Yorkshire and Hertfordshire, due to the extent of urbanised and suburbanised land, but are of a similar order. It should also be noted that the current trend is towards a further reduction in free-standing trees, and a concomitant increase in the numbers in ‘groups’ and ‘linear features’. In Norfolk, free-standing trees declined by 21 percent between 1980 and 1997, while trees growing in ‘groups’ increased by 72 percent and ‘linear features’ by 132 percent. The figures for Northamptonshire are minus 53 percent, zero change, and 108 percent.

Some of these complexities, and the conceptual difficulties they create, can be illustrated by looking in detail at some smaller areas. Henry Keymer surveyed a farm of 80 acres (32 hectares) in the Norfolk parish of Scarning in 1764, and recorded a total of 1,095 trees, or 12.4 per acre (31 per hectare) (NRO BCH 20). The First Edition Ordnance Survey 25-inch map of 1886 marks only 72 freestanding trees, or 2 per hectare, a total which had fallen to around 55 by 1946. Today there are only 32 hedgerow trees within the area, one per hectare, many of which are less than thirty years old. However, when the abandoned railway line that runs through the middle of the area was turned into a new route for the A47 in the 1970s, narrow strips of trees were planted to either side, the widths of which – c.13 metres – would allow them to be classified as ‘linear features’ in Forestry Commission terms. These contain, within the area mapped by Keymer, a further 514 trees; while a small clump of trees, developed around a pond, would fall into the Commission’s ‘group’ category, adding another 8. Looked at in one way – considering only free-standing hedgerow trees – the area thus saw a catastrophic decline of 93 percent in the course of the late eighteenth and nineteenth century, followed by a more muted decline, of 55 percent, over the following 130 years. But looked at in another way, and including within the calculation trees growing in ‘linear features’ and the like, the later twentieth century saw a significant recovery within the areas studied, so that tree numbers today are actually far greater than those present in the late nineteenth century, at 554 as against 72, although this is still under half the 1764 level.

What is striking about such very local studies, however, is the great variation in developments they reveal. Another Norfolk property surveyed by Keymer, in Beeston-Next -Mileham, although lying only some seven kilometres to the north west, exhibits a very different pattern of change to that

179 seen on the Scarning holding (NRO WIS 138). On the enclosed portion of the farm, covering 63 acres (25.5 hectares), there were 531 trees in 1764, or 21 per hectare. By the 1880s this had been reduced to around 100, although to an extent this reflects a significant degree of boundary reorganisation, so that few of the hedge lines shown by Keymer still existed; and by 1946 to a mere 60. Today there are only 7 free-standing trees within the plots mapped by Keymer, Here, moreover, ‘small groups’ around ponds and the like only raise the overall total to around 20, and there are no ‘linear features’.

Such marked variations, even over short distances, are evident elsewhere, and are often a function of how far landowners have embraced agro-environment schemes over the last few decades, or have chosen to manage their land for game and wildlife, rather than for agriculture alone. In north east Hertfordshire an estate at Whempstead in Little Munden had 766 trees on 98 acres when mapped in 1808, including examples growing in well-timbered fields – diminutive wood-pastures - as well as hedgerow specimens, although this figure includes ‘spires’ or immature examples. Excluding these the figure falls to 653, or c.7 per acre (16 per hectare) (HALS 81750). By the 1890s this had been reduced, within the bounds of the same area, to around 60. Today there are still 55 free-standing trees, several of them only just mature; but if we include examples growing in ‘groups’ and ‘linear features’, especially those in a magnificent outgrown hedgerow, this rises to 130, significantly above the late nineteenth-century figure. Some four kilometres to the south east, the central portion of a farm at Green End, Standon – covering some 40 hectares - had around 510 trees in 1774 (HALS E/2833). By the 1890s this had fallen to 120. But today, within the same area, there are actually more free-standing trees – around 130 – a high proportion of which, nearly a third, are less then c.35 years old. Adding to these ‘groups’ and ‘linear features’, the total number of trees rises to over 150. Yet just two kilometres to the east, on 32 hectares forming what was part of the demesne land of the manor of Barwick, the 75 hedgerow trees shown (with evident accuracy) on a map of 1774 (HALS E/2832) had declined to 56 by the 1880s, and today there are only 32 trees here. In this case, no linear features or groups exist, and less than a quarter of the trees are less than thirty years old. The contrast here is between a farm run, over several decades, with a keen interest in conservation; one on which arable production appears the main or only concern; and one lying in an intermediate position on this notional spectrum.

One effect of conservation management – including here both the practice of allowing hedgerow shrubs to grow into trees, and the deliberate planting of individual trees, ‘linear features’ and ‘small groups’ - has been to increase the numbers of ‘minority’ trees in the landscape, especially maple, which accounts for 9 percent of the trees recorded today at Green Farm, Standon; 36 percent of those at Whempstead; and – because of deliberate planting of the roadside belts some 40 years ago, beside the A47 – no less than 48 percent of the trees at Scarning. Indeed, oak now forms only a small minority of the trees in the latter area, outnumbered by sycamore and even by Lombardy poplar. At both Standon and Whempstead, in contrast, oak continues to be the majority tree, comprising around 50 percent of the population. Where the degree of new planting (and conservation management) has been less, as at Barwick and Beeston, oak tends to be more dominant, usually comprising more than half of farmland trees, although these are mostly overmature specimens.

Perhaps the most important point to emphasise is that these five sample areas developed along broadly similar lines in the period up to the late nineteenth century, and in most cases up to 1946. In all, the number of farmland trees declined sharply between the late eighteenth century and the 1880s/90s, normally falling by between 76 and 93 percent, although with only a 25 percent decline

180 at Barwick: the fact that there were already relatively few trees on this holding in the eighteenth century should be noted. There was then a slower decline to c.1946, usually of between 20 and 40 percent. But after this, there was much diversity of experience. Sometimes, as at Beeston and Barwick, decline continued at a steady rate to the present, whatever form of measure we employ. But sometimes it was reversed, even if only free-standing hedgerow trees are counted – as at Green Farm, Standon. More usually, as at Little Munden and Scarning, the number of farmland trees continued to decline if we count only free-standing example, in hedgerows or in pastures; but rose significantly if ‘linear features’ such as outgrown hedges, and ‘groups’, are included in the calculation.

Conclusion

Many of the statistics and other sources of information on which the above discussion is based need to be treated with extreme caution; successive kinds of data set are not always closely comparable, early estimates of woodland area are based on uncertain sample size, and many of the types and categories being counted have indistinct boundaries: when does an overgrown common become a wood, when does an outgrown hedge shrub become a tree? This said, the broad outlines of changes in trees and woodlands in the four sample counties since the late nineteenth century, set out above, are tolerably clear. They are also in line with what has been noted by others, such as the Forestry Commission in their successive inventories of woodland and trees, although we believe that the information we have presented both helps to better explain recent changes and to put them into a better historical perspective.

In the late nineteenth and first half of the twentieth centuries the management of both woodlands, and farmland trees, was primarily structured by economic factors, although on large estates especially recreational and aesthetic considerations were also important. The twentieth century witnessed a growing separation, with commercial forestry increasingly focused solely on woods and plantations. By the 1890s few hedgerow trees were pollarded, although most were still economically managed, as timber; but by the 1950s management for timber had also largely come to an end. Farmland trees were now primarily regarded, depending on the individual’s point of view, as a natural adornment of the landscape, or as an irrelevant encumbrance to farming, except perhaps in pastoral districts where some advantage might be derived from the shade they provided for livestock. To varying degrees, in all areas studied the numbers of farmland trees dwindled accordingly, for as trees died they were often not replaced – a situation made much worse by the onset of Dutch elm disease in the 1970s. During the last three of four decades this situation has to some extent been reversed, due to the impact of agro-environment schemes and a growing awareness of conservation issues on the part of many private landowners. The extent of this reversal varies greatly, however, from holding to holding, and also in terms of how it is measured. This is because the relatively clear distinction between hedges and hedgerow trees, and between woods and hedges, present in the nineteenth century has become increasingly blurred, the former by a decline in the practical function and thus management of hedges, and a consequent growth of some examples into lines of trees; the latter by the emergence of new kinds of farmland planting, in the form of ‘linear features’ and ‘small groups’ .

It is popularly assumed that the area of woodland in England has contracted markedly over the last century but this is clearly untrue. Instead it has grown from around four percent of the combined area of the four counties, to something approaching seven percent: they now collectively boast

181 more woodland now than in the sixteenth century, and quite probably more than in the twelfth. This expansion has, however, been accompanied by marked changes in composition, location and management. Traditionally managed woods, with few exceptions, slid gradually into a condition of neglect in the course of the twentieth century, and a significant number were grubbed out and converted to farmland or replanted with commercial conifers. At the same time, vast areas of new conifer plantations were established by the Forestry Commission, and a significant area of adventitious woodland regenerated across neglect commons, abandoned industrial land and the like. These developments intensified trends, already well-established in the nineteenth century, for a higher proportion of woodland to be concentrated in areas of marginal, acid land, former heath and moor. More recently, much deliberate planting of smaller woods, (and some larger ones) has been undertaken, for wildlife and amenity, by private landowners or local government bodies, and the forestry commission plantings have become more diverse, in terms of age structure and species composition, as that body has come to adopt more complicated management objectives.

In conservation terms, the issue is thus not so much an absolute shortage of woodland in the country; indeed, in some areas, it is not even an absolute shortage of farmland trees, if these are counted in such a way as to include ‘linear features’ like outgrown hedges. It is rather the character of our woodland, the mode in which it is planted and managed; and, crucially, the disposition of hedgerow trees, which now displays a more clustered character, with a much lower degree of ‘connectivity’ than in the eighteenth or nineteenth centuries. Indeed, levels of connectivity are in some areas at levels as low as those which pertained in unenclosed Midland open-field parishes. Yet even judgements about the environmental benefits, or disbenefits, of changing patterns of planting depend to an extent on what we measure and how we measure it. The establishment of extensive plantations of pure conifers in the course of the twentieth century, while rightly criticised by conservationists, in fact increased the numbers and distribution of a wide range of species, from beetles such as Phloeostiba lapponica and Plegaderus vulneratus (Welch and Hammond 1996, 97) to the hen harrier and most species of deer. But it is the task of biologists, not historians, to assess the relative value of different kinds of woodland and different modes of their management.

The current trend appears to be for a continued decline in the numbers of free-standing farmland trees; an increase in new forms of planting, such as ‘linear features’; little change in the proportion of ancient, coppiced woodland which is actively managed; and, on Forestry Commission holdings especially, a decline in the area occupied by pure stands of conifers and a rise in deciduous planting. Looked at in historical perspective, all the varied developments in trees and woodland over the last century, including those which have occurred over the last few decades, have served create treed landscapes radically different from anything which existed in previous centuries. Rather than see recent developments as marking a radical break with ‘tradition’ or the ‘natural’, however, it is probably more useful to view them as yet another phase of change, albeit perhaps more rapid then in earlier periods.

182

7. Health, Age and Management

The threats to tree health

In previous chapters we discussed the various ways in which the numbers and character of woods and trees have changed in England over the last four centuries, emphasising throughout the essentially ‘unnatural’ character of tree populations. Although in the course of the eighteenth and nineteenth centuries an increasing amount of tree-planting was carried out for reasons that were partly or even wholly aesthetic and recreational in character, farmland trees and woodland were, before the twentieth century, largely regarded in practical and economic terms. Only since the 1960s have both planting, and management, begun to be motivated on a significant scale by conservation and amenity considerations, which now influence the policies even of bodies like the Forestry Commission. A far more important influence on tree populations in this same period, however, has been an apparent decline in overall tree health and the appearance of a range of epidemic diseases. These are of varied character but many are associated with invasive bacteria, insects or fungi – alone or in combination – the appearance of which is generally accepted to be a consequence of increases in international trade and communication, climate change, or both (Brasier 2008).

Such threats began to be noted in the first decades of the twentieth century. Oak mildew was first recorded in England in 1908 (Mougou et al. 2008; Takamatsu et al. 2007; Marcais et al. 2014). Caused by the fungus Erysiphe alphitoides, probably of Asian origin, it affects younger leaves and soft shoots (especially the second growth of leaves in summer) and while it does not kill effected trees it can weaken them, leaving them susceptible to other problems. Dutch elm disease, caused by the fungus Ophiostoma ulmi disseminated by elm bark beetle in the genus Scolytus, may have infected tree populations in England in the prehistoric past but was likewise first clearly recognised in the 1920s. But the scale of such invasive infections escalated significantly in the second half of the century. The 1920s outbreak of Dutch elm disease had limited effects, but a second epidemic, beginning in the late 1970 and this time caused by a combination of Scolytus and O. novo-ulmi (a fungus probably originating in Japan) was far more serious, and within a decade it had effectively wiped out elm as a tree (although not of course as a shrub) across England (Brasier 1991; Gibbs et al. 1994; Clouston and Stansfield 1979). The disease was particularly serious because, as we have noted, elm was, in most southern areas of the country, the second or third most common farmland tree, and was especially prominent in designed landscapes, in the form of avenues and the like. Not long afterwards, in 2000, bleeding canker in horse chestnuts was reported in England, having been recognised in the USA as early as the 1930s. Originally associated with two Phytophthora pathogens, the causal agent is now most often due to a bacterial pathogen, Pseudomonas syringae v aesculi: by 2007 around half the horse chestnut trees in Britain showed some degree of symptoms. Further pests and diseases have arrived in the course of the twenty-first century. Horse chestnut leaf miner (Cameraria ohridella), the larvae of which mine within the leaves, destroying the tissues, was first observed in Macedonia in 1985, and was described as a new species in 1986, spreading thereafter steadily through Europe and acting, in association with canker, to pose serious threats to the future of chestnut (Straw and Williams 2013). It was first recognised in London in 2002 (Tilbury and Evans 2003). Oak processionary moth, a major defoliant of oak across Europe, arrived in England in 2005; but most serious of all is Chalara or ash dieback, caused by the fungus Hymenoscyphus fraxineu (Cheffings and Lawrence 2014). It causes leaf loss, crown dieback and bark lesions in affected trees, with high mortality rates in young specimens, and slower decline in mature examples: the scale of its

183 impact is unknown, in part because while first recognised as recently as 2012 in East Anglia – its spores apparently brought by wind from Europe, although imported plant material was a contributor to outbreaks elsewhere – there are indications that it may have been present in the country for over a decade. A range of other alien pathogens, present in Europe or Asia, have either appeared briefly in this country (such as sweet chestnut blight, Cryphonectria parasitica); or are a potential threat in the future (such as emerald ash borer Agrilus planipennis).

In addition to these clearly defined tree illnesses, with identifiable causes in invasive pathogens, a number of less specific ailments have caused increasing concern over the last few decades. The most important of these is ‘oak decline’, a syndrome currently thought to have multiple and complex, although debated, causes. Effected trees suffer deterioration in the appearance of foliage followed by progressive thinning of the crown and death of branches: in some cases the decline is terminal and leads to the death of the whole tree. The first documented cases in England came in the 1920s, and were initially interpreted as a consequence of defoliation by caterpillars of the indigenous oak leaf roller moth (Tortrix viridana), combined with the effects of the newly-introdcued oak mildew fungus. But the syndrome was increasingly noticed during the middle and later decades of the century, and is now variously attributed to factors like disease and drought (Thomas et al. 2002). During the 1980s a more extreme form of the syndrome, ‘acute oak decline’, was identified. This is associated with vertical, weeping fissures in the bark of the tree and often leads to death within six years (Denman et al. 2014 Brown et al. 2014). The larval galleries of the native beetle Agrilus biguttatus are frequently found when infected specimens are felled and it is probable that the insect acts as a vector for a range of bacterial diseases (Morall et. al 2000), although it is also possible, as some have suggested, that the presence of the insect is more a symptom, than a cause, of the disease (Wargo 1996).

The rather non-specific character and multiple proposed causation of oak decline (at least in its nonacute forms) are paralleled in the condition known as ‘ash dieback’, recognised well over two decades before the identification of chalara in Britain. It, too, is characterised by severe crown dieback and general poor condition. Recent research suggests that the syndrome may be essentially physical rather than biological in character – the consequence of such things as root damage through ploughing, or a falling water table - but its effects can now be confused with those of chalara and its incidence may have masked the first onset of the latter disease.

Writing about tree health before the late nineteenth century

The publicity afforded to recent diseases of trees, especially those associated with invasive organisms, coupled in many cases with their epidemic scale, should not be allowed to obscure the fact that English trees have always been subject to ill-health – it is a normal part of their existence, as it is of ours. Oak, for example, has always been prone to fungal attacks from Aetiporus sulphureus (cuboidal brown rot of heartwood), Stereum gausapatum (pipe rot), and Fistulina hepatica. The caterpillars of the indigenous micromoth Tortrix viridana have long fed on young oak leaves and in some periods have caused successive years of severe defoliation, severely weakening trees and arguably contributing to, although never directly causing, their death. More seriously, the condition known as ‘shake’, which affects oak but also sweet chestnut and some other timber trees – that is, internal (longtitudinal) splitting of the timber, either across or between the growth rings – has for a long time been of particular concern to foresters because it is hard to identify before the tree is

184 felled. Its precise causes continue to be debated, although they appear to be mainly physical rather than biological in character, with trees on well-drained sites, especially on acid soils, being particularly prone. As many as 20 percent of English oaks, according to some authorities, may currently be affected (Price 2015; Mather et al. 1994). The symptoms of such long-established native illnesses and pathogens can be confused with those of recent arrivals. Some of the symptoms of chalara, for example, are comparable with those caused by the common fungal pathogen Nectria galligena, the bacterial pathogen Pseudomonas savastonoi pv fraxini, the ash bark beetle Leperisinus varius or the ash bud moth Prays fraxinella.

While sick trees have always been with us, changes in the scale of morbidity and mortality over time are difficult to chart through documentary sources and published literature. There is, it is true, an increase in references to tree diseases in forestry texts as we move through the second half of the nineteenth and into the twentieth century, but it is difficult to assess the degree to which this reflects the growing sophistication of forestry as a discipline, its increasingly scientific character, and a greater optimism about what could be done to prevent or limit disease, rather than any real deterioration in tree health. In earlier times, a more fatalistic or at least accepting attitude seems to have been the norm. Serious writing on trees and woodlands began in the later seventeenth century, with books such as John Evelyn’s Sylva (1664) Moses Cook’s The Manner of Raising, Ordering and Improving Forest Trees (1676) and William Ellis’ The Timber Tree Improv’s (1741). These were followed through the eighteenth century by a rash of texts which added to, borrowed from, and at times simply plagiarised these and other works, including Stephen Switzer’s Ichnographica Rustica (1718), Batty Langley’s Sure Method of Improving Estates (1728), Phillip Miller’s The Gardener’s Dictionary (1731), James Wheeler’s The Modern Druid (1747) and William Hanbury’s, An Essay on Planting (1758). All combined, in differing proportions, practical advice, aesthetic comment and polemic. Equally important, although less often cited by writers on woodland history, are a series of texts from the seventeenth and eighteenth centuries whose main focus was not in fact forestry, but farming life and the practice of agriculture, but which also deal extensively with the planting and management of farmland trees and woods. These include Samuel Hartlib’s Enlargement of the Discourse of Husbandry (1651), John Mortimer’s Whole Art of Husbandry (1707), John Worlidge’s A Compleat System of Husbandry and Gardening (1660) and Leonard Meager’s The Mystery of Husbandry (1697). Other references to the practice of silviculture occur in such things as the early county histories, which appeared from the late seventeenth century.

While seventeenth and early eighteenth-century writers do refer to disease in trees, the details they provide are generally vague, or even completely mysterious, such as the observations by the Northamptonshire historian Morton that ash trees between Thorpe and Peterborough in Northamptonshire had been badly infected by exudations of ‘manna’ in 1684: ‘it issued forth in such plenty that it dropped upon the ground’ (Morton 1712, 403). Most of the literature suggests that contemporaries simply considered ill health as normal. Moses Cook thus advised that coppices should not be allowed to grow too high, for too long, because ‘you cannot come to survey your Timber-trees, to see which are decaying … Why should any reasonable Man let his Trees stand in his Woods, or elsewhere, with dead Tops, hollow Trunks, Limbs falling down upon others and spoiling them, dropping upon young Seedlings under it, and killing them?’ (Cook 1676, 163). He devoted a short chapter to ‘Diseases of Trees’, which he divided into a number of categories. Firstly, he thought that trees were often adversely effected by being planted on unsuitable ground: ‘The Roots may be the decay of the Tree; as, if they stand on Ground contrary to its kind, as the Beech on cold Clay, and 185 wet ground’ (Cook 1676, 160-1). Secondly, he thought that their growth could be impeded by competition from weeds growing around them, robbing them of nutrients; by climbers like ivy and traveller’s joy; or – in the case of species like cherry and elm – by competition from suckers. Thirdly, trees might be damaged (especially when young) by the browsing of deer, hares and rabbits, or by the activities of moles, mice and rooks. Cook paid particular attention, however – as did other writers – to the effects of rot, canker and ‘worms’ (Cook 1676, 162-3). These could be caused by natural damage and poor pruning, but also remedied by astute, surgical pruning. ‘If Worms are got between the Bark and the Body of your Tree, they must be cut out, and the Place done over with Loom’. ‘Cankered, or galled Places, or Boughs broke, are to be cut smooth, and covered with Loom; the Canker must be cast clean out. If a great Bough be broke, and the Tree old, cut it off at a distance from the Body; but little Boughs close’. ‘If a Tree be blasted in part, or the whole Head, cut all that is blasted or dead, close off to the quick, and take out all dead Boughs’. Other writers, such as Mortimer, refer to the effects of mildew in oak ‘and other trees, whose leaves are smooth, and do not easily admit the moisture in them, as the Elme and other rougher Leaves do’; they also discuss honeydew on oak, maple, hazel and lime, and note the role of such things as blackfly as a cause of decay.

In second half of the eighteenth century writers begin to provide more details of the various threats to tree health, although their main concerns were as before, and couched in similar terms. Galls, cankers and bleeding were a particular concern, with Forsyth in 1791 describing the importance of applying plasters to tree wounds (Forsyth 1791, 321). The effect of dampness and ‘vapours’ were a perennial concern, as they were amongst writers on human health, and by 1775 William Boutcher was able to suggest that dense planting of trees in the newly-fashionable plantations could cause problems, referring to the ‘many mortal diseases incident to large and crowded plantations’ resulting from ‘the damp vapours’, a concern which may in part explain the greater attention to the drainage schemes implemented in woods and plantations from the late eighteenth century which we alluded to earlier (Boutcher 1775, 298). There were now clear descriptions of diseases like ‘shake’, which Walter Nicol in The Practical Planter of 1799 believed could be prevented by careful pruning: interestingly, he described it primarily as a condition of elm, although noting that it affected other ‘timber’ (Nicol 1799, 159-60). Robert Monteath, in The Foresters Guide (1824), likewise described it as particularly a disease of elms, noting amongst its symptoms stag headedness and ‘swelling veins’ above the bark. His description of how ‘severe frosts are often fatal to trees even of a large size, and oftentimes rend the trunks in a manner similar to thunder, and kill the tree’ (Monteath 1824, 420) may also be a reference to the effects of ‘shake’, which can lead to high levels of mortality in extreme cold weather (Cinotti 1989). References to stag-headedness, and to what sound like the symptoms of ‘decline’ in oaks, are frequant: ‘That dead-topped oaks are very common, cannot be disputed’ (Pontey 1805, 130). Yet while late eighteenth- and early nineteenthcentury texts show a greater awareness of different threats to trees, the remedies proposed still often appear crude to modern eyes, Monteath for example devoting space to such dubious ‘cures’ as pouring salt water over the roots of larch trees attacked by insects. The need to prune dead branches from oaks and other timber continued to be emphasised, William Corbett in 1825 urging that this was necessary to ‘prevent the mortal disease reaching the trunk’, and Steuart in The Planters Guide of 1828 similarly emphasising how ‘having displaced the proper branches, you should also cut off all such parts of branches as have by accident been broken or wounded for these will remain a disagreeable sight, and often occasion disease in the tree’ (Cobbett 1825, para. 437; Steuart 1828, 431).

186

Only from the middle decades of the nineteenth century do comments about diseases, and how to deal with them, begin to become more detailed, and to acquire a more modern tone. Prideaux in 1842 discussed the severe damage caused to oaks by the oak leaf roller moth Tortrix viridana, as well as other insects; noted the prevalence of canker in larch; and explained the impact of several insects on elm trees, of which Scolytus:

Was considered, some years ago, as by far the most formidable and destructive, as to its ravages in the larva state were attributed the decay and subsequent death of the finest Elms in the vicinity of London, particularly those in St James’ and Hyde Parks. Subsequent observations on the economy and habits of the insect have, however, shown that the Scolytus is not always, indeed perhaps but seldom, the proximate cause of decay, and that the trees were not attacked by the pregnant females for the purpose of depositing their eggs beneath the bark, until they have become injured and diseased from some other cause… (Selby 1842, 114).

Coleman in 1859 likewise records several insect pests of the elm, including Scolytus and goat moth, while noting how ‘the foliage of ash gives food to several fine insects of the moth tribe’ (Coleman 1859, 23).

The rise of scientific forestry

By the 1880s and 90s the discussions of tree diseases in texts become more detailed and extensive. In The Oak, published in 1892, H. Marshall Ward listed numerous fungal and insect pests and their effects; while Charles Curtis published in 1892 the first book (49 pages in all) entirely devoted to tree health. In it he discussed the poor state of many woods and plantations: ‘If we walk through areas of woodland we see it all round us, even in well-managed tracts; we see it in ground rot, in broken limbs, in holes in the stems full of water, in over- crowding, in the presence of lichen, in partially uprooted trees, and in the creeping ivy, so dear to those who judge beauty from the artistic standpoint’ (Curtis 1892, 7). He highlighted signs of incipient disease: stag-headedness, holes filling with water, slowness of growth, premature seeding in young trees, the condition known as ‘barkboundness’. He reviewed the problems associated with particular trees, such as the larch (planting on unsuitable sites, overcrowded plantations leading to infestation by the fungus Peziza Wilkommii and the development of canker (although he acknowledged that this could occur in other contexts, as could damage from the larch bug Chermes laricis)) (Curtis 1892, 21). He paid particular attention to stagheadedness, which he discussed mainly in the context of oak but also lime; and to ‘diseases of the bark’, which arose:

From manifold causes, from the growth of epicormic branches, induced by injudicious thinning, wounds, ivy, wet and cold sub-soil, frost, and insect attack; the result may be excrescences, a bark-bound condition, the separation of the bark from the wood, and various other manifestations (Curtis 1892, 30).

He also listed the various insect pests of trees;

Among these are the ash bark beetle (Hylesinus fraxini), ash bark scale (Chionaspis fraxini), elm bark beetle (Scolytus destructor), goat moth (Cossus ligniperda), buff tip moth (Pygera bucephala), cockchafer (Melolontha vulgaris), marble gall fly (Cynips Kollari), oak leaf roller

187

moth (Tortrix viridana), pine beetle (Hylurgus piniperda), pine saw fly (Lophyrus pini), pine weevil (Hylobius abietis) and spruce gall aphis (Chermes abietis) (Curtis 1892, 33).

Such pests caused a host of major and minor problems, including ‘the destruction of the leaves of the oak, the black fungoid spots upon the leaves of the sycamores and maples’. In most cases, he suggested, insect attack was best dealt with by clear felling the affected areas of woods or plantations, especially in the case of Hylesinus fraxini, Chionaspis fraxini, and Scolytus. Curtis singled out the attacks of Tortrix as particularly damaging: ‘The Oak Leaf Roller Moth is an insect which causes great injury to our oak woods and plantations by depriving the trees of their leaves in the early season, thereby materially affecting the health of the tree and the development of timber … Whole districts and whole woods are sometimes stripped bare of leaves, so as to give an appearance of winter’ (Curtis 1892, 41).

By this time, information about disease had become an indispensable part of the various texts on scientific forestry which were appearing in growing numbers. Notable examples include W.R.Fisher’s volume on ‘Forest Protection’ in Schlich’s classic forestry text book of 1895, and the relevant chapters in John Simpson’s The New Forestry of 1900 and Percival Maw’s The Practice of Forestry of 1912 (Fisher 1895; Simpson 1900; Maw 1912). The early volumes of the Quarterly Journal of Forestry not surprisingly devoted much space to tree disease. The first, which appeared in 1907, discussed diseases of Larch (Anon 1907); the second included a report on the ‘Injurous insect etc of the Midland Counties’, with decriptions of control experiments using insecticides and fungicides, including against Scolytus on elm (anon, 1908); while volume 4, published in 1910, included an article on larch sawfly and a report on a newly-recognised disease of alder (Annard 1910; Jones 1910). By this time, regional scientific societies, as much as forestry experts and professionals, were taking a keen interest in the subject, The Transactions of the Manchester Microscopical Society for example issuing a special report on insect pests in 1896.

The more sophisticated approach to tree disease manifest by the late nineteenth century was in part the consequence of incremental developments in indigenous forestry, but in part it was the result of the influence of the new scientific approach to silviculture which had developed over the previous decades in Germany. In 1893 Herman Furst’s book on The Protection of Woodlands Against Dangers Arising from Organic and Inorganic Causes was published in English (in Edinburgh) (Furst 1893). This dealt in detail, in particular, with the predisposition of trees to disease through mechanical injury, poor soil, atmospheric pollution (especially near cities) and weather, especially severe frosts. But it also discussed a wide range of fungal and insect pests of trees, as did his fellow-countryman R. Harting’s Diseases of Trees, which was published in English the following year (Harting 1894).

Early documentary evidence for tree health

These more detailed discussions of tree disease do not, of course, necessarily reflect any greater incidence of ill-health from the middle or later decades of the nineteenth century. Not only do earlier texts, as noted, make it clear that poor health was commonplace; some of the forms it took can be identified with modern diseases in trees. Early references to caterpillars affecting oaks, such as those made by Moses Cook in the late seventeenth century, must almost certainly relate to Tortrix, while the obsession with galls and rots probably reflect the endemic character of Aetiporus sulphureus and Stereum gausapatum. Documentary evidence, moreover – estate surveys and the

188 like – similarly suggest that ill-health was normal and expected in the period before the middle of the nineteenth century. The terms used to categorise trees in a survey of Staverton in Northamptonshire in 1835, for example, included ‘decayed’, ‘damaged’, ‘small and very bad’, ‘very bad’, ‘decayed very bad’ and ‘dead’. In the woods at Geddington Chase in Northamptonshire in 1760 it was said that ‘a great deal of the timber’ was decayed (Boughton House archives); while at Mundford in Norfolk in 1805 the surveyor commented that he was ‘disappointed in the quantity of trees in the Square Plantation, finding such a quantity of dead ones’ (NRO MS 13751, 40E3). Felling and sales records often refer to ill or dead trees and - given that most diseased specimens appear to have been taken down long before they actually died – they probably convey an unreliable impression of healthy populations. In the 1760s it was reported on the Broughton estate in Northamptonshire that ‘a great deal of it [the timber] in some of the sales falls very faulty and rotten and makes but very little money…’; while on another occasion it was said that 14 out of 101 trees inspected in one estate wood were ‘so very red and defected that the body will by no means do for shipping’, although it was thought that ‘some parts of the Heads may produce Crooks and Knee timber’ (Boughton House archives). At Prior Roydlwood near Sheffield it was reported in the late eighteenth century that no less than a quarter of the 21,000 wavers were ‘very sickly’ (Shef ACM/MAPS/Shef/169); while on the Bolton Estate in Wensleydale in 1809 it was said that “the timber at Capplebank will be insufficient for the fencing’ on account of ‘many of the trees being much decayed in the middle’ (Nrth ZBO 4404).

It is noteworthy that references to dead ash seem to be more common in estate records than dead oak, although it is unclear whether trees of this species were more prone to sudden death or, being less valuable than other timber, were less carefully monitored, and thus less likely to be felled before the problem became acute. A timber account for the Heydon estate in Norfolk, drawn up in 1911, had a separate category for ‘dead ash’, but not for any other tree (NRO BUL 11/118/1); the trees sold from the Evan-Lomb estate in 1835 included ‘Spruce, Hornbeam, dead ash, and elms’, while estate memorandum from Bradenham Hall estate in 1904 refer to several examples, including a ‘dead ash in Great Wood’ to be used for firewood and a ‘Dead Ash next Vont’s shop’ (NRO HNR 465/3/1). At Bradenham in Norfolk in 1750 the largest ash were expected to fetch a good price but ‘with the smallest that are decaying twill be best to advantage and a better price to sell them mixed together Great and Small that are past improvement’ (NRO MS 9316, 789). Not surprisingly, given problems with Scolytus, ‘dead elm’ is another category sporadically encountered in records, as on the Stradsett estate in west Norfolk in 1815 (NRO BL/BG 5/3/4). Attacks by the beetle could certainly be severe: ‘the ravages committed by this small creature would scarcely be credible‘ (Johns 1899). Trees exhibiting what we would probably now describe as ’oak decline’ are also often referred to in documents. At Prior Roydlwood on the far northern outskirts of Sheffield a note on felling drawn up in the late eighteenth century reported that 206 ‘reserve’ trees were ‘nearly all dead top'd’ (Shf ACM/MAPS/Shef/169); while the owners of the Bolton Estate in the eighteenth and nineteenth centuries ‘generally used the tops of old trees for mine timber, an indication that many of the trees were affected by crown dieback’ (Dormor 2002, 222). In short, poor health in trees, clearly featuring many of the ailments still common, was not unusual in the past: but, before the advent of scientific forestry, it was simply something to be tolerated, with diseased trees being felled and sold for the best price possible.

In some cases, of course, it was drought or waterlogging, or attacks by animals, rather than infestations by insects, fungus or bacteria, that were the cause of illness and death. While in general

189 deer are not recorded as a major problem in the period before the twentieth century, their numbers in the countryside generally being low, they did cause severe difficulties in the royal forests of Northamptonshire, where they had long been carefully protected. In the 1760s it was said of the woods around Geddington in Northamptonshire that ‘there can be not succession but when the old stuff is cut down it will be quite naked of everything but underwood, nor will there be any young oaks raised from acorns unless the coppices are better preserved from the deer…’ (Boughton House archives). Rabbits too were a problem, although probably only on a localised basis before the eighteenth century. At Settringham in Yorkshire in the 1590s it was reported that ‘after the cony warren was moved out of the low commons to be nearer to the Manor house, they have multiplied and do great damage to the poor tenants, and do great hurt in the wood by destroying the young springs and they have altogether destroyed one spinge called Peeke spring’ (King and Harris, 1963, 32). Rabbits will have become a more serious problem during the eighteenth and nineteenth centuries as - with the enclosure of the remaining open fields, the adoption of fodder crops and the decline in bare fallows – they increased in numbers and colonised new areas (Sheial 1971, 133-7). In 1855 the agent for Sir Henry Meux of Theobalds in south Hertfordshire wrote to his employer with ‘the account of the underwood now felled in Hoddesdon Park, Cow Heath, Highfields … I am sorry to add that your woodman complains bitterly of the damage done by rabbits to your young wood’ (Page 1912, 450).

Long before the development of scientific forestry in the later nineteenth century the sheer scale and diversity of the problems facing the forester were being emphasised by many observers. Hamilton, writing in a Scottish context in 1816, believed that ‘forest-trees of all descriptions, are subject to such a variety of mischiefs and diseases, even under the most judicious management, that it is to be ranked among the favours of providence that the sustenance of man does not depend upon them’ (Hamilton 1816, 446). He listed the effects of insect infestations, especially on larch; frost; the depredations of hares and rabbits; mildew; lightening; canker; and the tendency of trees to become ‘bark-bound’, a condition affecting oak, ash and elm. He also described in some detail the ‘resinous distemper of the Scots fir’, although he thought this a disease restricted to highland areas of Scotland (Hamilton 1816, 447).

All this said, some increases in the incidence of disease in the period between the late eighteenth and the mid twentieth century can perhaps be discerned in documents and texts. These were encouraged, in particular, by the emergence of the new plantation silviculture, which involved the establishment of large numbers of closely-spaced trees of a single species, often in places where they were not normally found, and frequently on marginal land, newly enclosed from the commons. The tendency for many to be planted on acid soils, former moors or heathland, may explain an apparent increase in references to ‘shake’ in texts and documents in the early nineteenth century. In 1803, for example, the agent of the Bolton estate in Yorkshire complained that many trees were affected by 'shakes' (Dormor 2002, 223); while in 1809 there was concern that ‘the timber at Capplebank will be insufficient for the fencing and that it cannot be fully ascertained until it is cut up, on account of many of the trees being much decayed in the middle’, possibly again as a consequence of shake (Nrth ZBO 4404). By the end of the century Nisbet, in his revision of James Brown’s The Forester, could note how ‘in artificial forests and plantations…. Trees grow under conditions that are not so favourable to normal healthy development as wild growth in natural woodlands…. For the plants are not indigenous to the soil, and more or less of one age and are nearly all at equal distances apart from each other’, going on to describe a range of fungi and insects that might infect

190 such trees (Nisbet 1894, II, 285). Larch, widely planted in this period, was especially susceptible, but other conifer species also exhibited signs of poor health in plantations. Loudon in 1838 noted how ‘The highland Society in their list of premiums for 1838, offer a medal for the best account of the disease which has of late years attacked the stem, larger branches, and occasionally the twigs, of the silver fir somewhat resembling the well known rot of the larch’ (Loudon 1838, 2601). The even more extensive stands of conifers established by the Forestry Commission from the 1920s were particularly susceptible to epidemic disease. Not long after the planting of Scots pine began in the East Anglian Breckland large areas of the new forests were badly affected by attacks by the pine shoot moth Evetria buoliana (Ross 1935; Gibbs et al. 1996). Far more serious was the emergence, during preliminary stages of thinning in the late 1930s, of the fungus Fomes annosus (since renamed Hererobasidion annosum), which had been first described by Elias Fries as early as 1821 (Day 1948; Rishbeth 1963; Wass 1956; FC archive, Santon Downham, ‘Fomes’ file). The airborne spores of the fungus took hold in the stumps of felled trees, removed during the first thinning operations, and then spread through the root system to infect unfelled specimens. The deaths were initially attributed to drought or waterlogging, and the fungus viewed as a manifestation of stress. By 1946, however, the Annual Report of the Commission was able to describe how ‘Die back resulting from the attack by the root fungus Fomes annosus in 20 to 25 year old plantations of Scots and Corsican pines in East Anglia has caused appreciable losses in certain compartments’. Infected stumps were initially isolated, by being surrounded with a trench; then inoculated using the fungus Peniophera gigantean; but the disease continued to effect new planting, especially on the more calcareous soils, and since the 1970s stumps in such areas have been routinely bulldozed into long rows, to prevent the spread of infection through the roots (Skipper and Williamson 1997, 72-4).

The advent of new pathogens

By this stage, as already noted, more serious threats to tree health were appearing in England in the form of invasive pathogens, a consequence of long-distance trade both in wood and timber, and in live plant materials. Neither of these things, we would emphasise, was new. Even in the Middle Ages Britain was less wooded than most of its European neighbours, and thus obliged to import some timber, while certain types and qualities of timber were only available overseas. Initially, most imports were sourced from the vast forests of the Baltic region. As early as 1273 an entry in the Account Rolls of Norwich Cathedral Chapter recorded the journey of ‘John the carpenter’ to Hamburg to buy timber, and the carriage of boards from Hamburg to Yarmouth (Latham 1957, 28). Throughout the late medieval and post-medieval periods there are references to Baltic wood and timber in all the districts studied. Baltic oak was brought to from Boston and Peterborough in 1500, for example, while in 1550 John Johnson was bringing in oak from Antwerp to King’s Lynn, and then transporting it by water to Yaxley, and on by land to Glapthorn in Northamptonshire (St John’s College . MS 102.9). In 1727 David Eaton purchased ‘Norway oak’ for flooring the dining room at Deane House in the same county (Wake and Webster 1971, 105). But the eighteenth century saw a rapid increase in the scale of imports, especially into east coast ports. Studies of the trade and shipping of Hull have shown that during the seventeenth century, wood formed a large component of the flourishing trade with the Scandinavian countries (Davies, 1964). Norway was a particularly big exporter because its saw mills, equipped with fine blades, were thought to produce the best quality deals. During the eighteenth century, however, the development of sawmills in the Hull area meant that there was also a substantial rise in the volume of imports of unsawn rough wood. The same period saw an increased dependence on the New

191

World as a source - initially for specialised woods like mahogany, but by the nineteenth century for a wide range of general hardwoods and conifers, although the importance of the Baltic as a source increased again in the later nineteenth century, as Canadian timber exports were increasingly absorbed by markets in the USA. By 1905 oak timber was also being imported from Japan (Latham 1957). The quantity of timber being imported into Britain increased rapidly with the onset of industrialisation, and especially in the second half of the nineteenth century, reaching 2.6 million cubic metres by 1851, 5.9 million by 1871, 13.4 million by 1900 and 16.4 by the outbreak of the First World War (Fitzgerald and Grenier 1992, 18). Following the war there was a decline, with only 6.8 million square metres being imported in 1921, but this was followed by a second peak, of 16.2 million, in 1936 (Fitzgerald and Grenier 1992, 144). Since then the volume of imports has followed an erratic course, but currently stands at around 6.3 million cubic metres of sawn wood per annum, in addition to 3.2 million of wood-based panels and 9.7 million of wood pellets (Fitzgerald and Grenier 1992, 144; http://www.forestry.gov.uk/forestry/infd-9nrkgg). It is important to emphasise not only the rising scale of imports in the second half of the nineteenth century but also the increased speed with which material could be brought across the sea. By 1870, a number of inventions, including the screw propeller, the compound engine and the triple-expansion engine, made trans-oceanic shipping of bulk cargoes (as opposed to passengers) by steam rather than wind economically feasible, and goods thus travelled much faster (Carlton 2012).

Live tree materials were also being moved long distances from an early date. This initially involved fruit trees, and ornamentals for gardens – in November 1696 the Norfolk landowner Richard Godfrey lamented that frost was preventing the delivery of fruit trees he had ordered from Holland (NRO Y/C 36/15/18) - but by the end of the century forest trees were also being brought from overseas, and in 1700 John Bridges of Barton Seagrave in Northamptonshire reported that he had recently planted ‘500 limes from Holland’ (Morton 1712, 486). By the end of the nineteenth century large quantities of commercial conifer seedlings were being regularly imported from both France and Germany. There was also much movement of plant material within the British Isles, as large-scale estate forestry got under way from the later decades of the eighteenth century. By 1803 seedlings of larch, spruce, silver fir, Scots pine sweet chestnut, wych elm, birch, sycamore and even ash – 72,000 trees in all - were brought from Scotland and planted on the Bagge family properties at Mintlyn in west Norfolk (NRO BL/BG 5/3/3). Such movements continued on a large scale right through the nineteenth century and in 1910 the same estate was still purchasing spruce, larch, Scots pine, and even hazel, birch, privet and oak, from Scottish nurseries (NRO MC 547/81-104).

By this time, the scale of imports from abroad was causing concern in some quarters. In 1910 a letter in the Quarterly Journal of Forestry from the noted forester P.T. Maw urged the society’s governing council to:

Put forth every effort in their power to induce the Board of Agriculture to…..to prohibit the importation of seedlings raised abroad….a most dangerous practice and one calculated to destroy the comparative immunity from fungus attacks which at present our young plantations enjoy…..I am quite aware that many nurseries are stocked with continental seedlings but I can assure those who adopt this course that they are only “buying Trouble (Maw 1910).

A number of replies contested Maw’s suggestion. One, while stating a preference for English seedlings, denied that there was any hard evidence that any disease had been caused by imported

192 plant materials, and pointed out that imports were likely to continue for as long as the price of imported plants remained lower than those for home-grown stock. German sitka spruce was at this time around half the price of English, and even beech was available from Germany at 13s 6d a plant, as opposed to the 25s for an English seedling. Another contributor similarly denied that there was any evidence for the spread of foreign diseases through imports.

The bulk of our imports of plants come from the Continent of Europe and no tree disease is troublesome there that we have not already got in this country. Even assuming that a small proportion of imported young trees are not altogether sound it can make no practical difference to the condition of things here if the quantity of disease already present in this country as augmented to a fractional extent. In the great majority of cases what determines a disease attack is not the number of fungal spores but the presence of conditions favouring or limiting the attack (Knutchbull 1911).

In spite of such denials it is clear that the rapid movement of timber and plant materials around the globe had, by this stage, led to the arrival of oak mildew in England, in 1908; and this was soon followed by the appearance of Dutch elm disease (assuming that previous incidences of disease in elm were directly caused by Scolytus itself, rather than by Ophiostoma ulmi). In a similar way, European diseases were also now spreading to other parts of the world. Beech bark disease, common in Europe, is reported as having arrived in Halifax Nova Scotia in 1890, while the Journal of the Arnold Arboretum reported in 1930 the appearance of a European phytopthera in Maine (Ehrlich 1932; Chester 1930). In Britain, as already described, the late twentieth and early twenty-first century witnessed the arrival of further alien pathogens: a more virulent form of Dutch elm disease in the 1970s; horse chestnut canker in 2000; horse chestnut leaf miner in 2002; oak processionary moth in 2005; and, most recently, ash chalara. While it is possible that foreign tree pathogens had already arrived in this country before the twentieth century, there is certainly no evidence in the documentary record for the outbreak, at least over the last four centuries, of anything as serious as Dutch elm disease, in its 1970s form.

Changes in the age structure of tree populations.

While the increasing scale and speed of transport systems, and the resultant spread on invasive pathogens, are evidently the principal cause of declining tree health in England, changes in the character of tree populations, and in forms of management, over the last century and a half may have been contributory factors. In particular, we should note the steady increase in the number of old trees in the countryside which has occurred since the middle of the nineteenth century. While there were always some old timber trees in the countryside, especially in parks, most were felled before they were sixty or seventy years old, many long before this. But as we have seen, from the middle of the nineteenth century the age at which timber trees were taken down rose steeply, as a consequence of both changes in technology and of the declining value of bark as a commodity. This important development was followed by another, for from the early twentieth century the average age of farmland trees increased still further, as the intensity with which they were managed declined. The numbers of new trees being planted dwindled, and existing timber, where it survived opportunistic felling, often grew to an advanced age. These developments were the consequence of the complex interplay of a number of factors.

193

Firstly, the scale of timber imports, especially of oak from eastern Europe and of hardwoods from the tropics, made it difficult to justify, in financial terms, the felling and transportation of single trees growing in isolated locations, often difficult to access. Forestry was accordingly concentrated more and more in woods and plantations, and farmland trees were retained principally for aesthetic reasons, on large estates at least. Secondly, where – as was so often the case in the first half of the twentieth century – landed estates were broken up, their constituent portions mainly passed into the hands of a class of ‘farmers who know little and care less for forestry’, and whose main concern was the growing of crops and the raising of livestock (Butcher 1941, 361). This was in marked contrast to the situation which had pertained when large landed estates operated integrated forestry programmes, embracing both woods and plantations and timber growing in hedges and on pastures. Even where hedgerow trees were tolerated by new owners, new ones were seldom planted. Where large estates remained intact hedgerow timber was often more commercially and intensively managed, although at a declining rate, especially after the Second World War, when priority was given to food production over all other kinds of activity in the countryside. To all this we should add that, as farming became more industrialised in character in the course of the twentieth century, many of the features or equipment formerly constructed with timber grown on the farm – fencing, gateposts, minor buildings – were now supplied by manufacturers.

In addition to all these essentially economic and silvicultural factors, there were important changes in social attitudes. The late nineteenth and early twentieth century saw the establishment of a number of organisations dedicated to the conservation of rural landscapes, open spaces and wildlife. These included the Commons Preservation Society, founded in 1865; the Society for the Protection of Birds, later the RSPB, in 1889; the National Trust, founded in 1895 to preserve ‘for the benefit of the Nation …lands and tenements (including buildings) of beauty or historic interest’, principally by purchasing them; and the Society for the Promotion of Nature Reserves, set up in 1912 (Evans 1992; Cowell 2002; Waterson 1994; Sheial 2002). Of particular significance was the establishment, in 1926, of the Council for the Protection of Rural England, with the aim of preserving ‘all things of true value and beauty’ (Sheial 2002, 106). The appearance of these various organisations was associated with a growing divorce of the majority of the population from the realities of rural life. By 1851 half the population of England already lived in towns and cities but by 1911 this had risen to 80 percent (Sheial 2002, 12). Moreover, in the first half of the twentieth century, as a result of significant changes in patterns of disposable income, and of improvements in personal transport, an urbanbased and increasingly mobile group visited the countryside on a larger and larger scale, in some cases began to settle in it, or in suburbs on its margins, and to take an active interest in preserving it from what were believed to be the insidious effects of modernity and insensitive exploitation. The idea that the countryside was essentially ‘natural’, which had been developing (alongside urbanisation and industrialisation) since the eighteenth century, now triumphed, not only amongst the new urban and suburban middle class but also amongst many landowners. In such circumstances, felling prominent hedgerow trees gradually came to be regarded not as a normal part of land management, but rather as a desecration. Such ideas were manifested with particular clarity, somewhat paradoxically, where countryside was being lost through large-scale urban or suburban development. When Letchworth Garden City, the first garden city in the country, was established in north Hertfordshire in 1902 the chief architect Raymond Unwin claimed that the alignment of the main this axis of the town, running from the railway station to the Town Square, had been decided by the positions of three old oak trees which were retained in the design: he also boasted that only a single tree had been felled during the town’s construction (Rowe and Williamson 2013, 274). Sales

194 particulars for one suburban development in the south of the same county, on what was effectively becoming the northern edge of London, emphasised in 1912 that ‘It is desired to preserve the rural characteristics of the locality as much as possible, and with that object in view the natural hedges and as many of the trees will be retained as is consistent with convenient development’ (Bushey (Herts) Museum archive). By 1930, when Hertfordshire’s second Garden City was being constructed, Sherrards Wood – an area of ancient woodland which also formed part of the Panshanger estate sold for development by Lord Desborough in 1919 - became the focus for a spirited ‘Save the Woods’ campaign which resulted in the preservation, for the enjoyment of the public, of 100 of its 165 acres (Austin 2001, 10).

By the time of the Second World War the idea – long promulgated by land use planners like Patrick Abercrombie and campaigners like Clough Williams Ellis (Abercrombie 1943; Williams Ellis 1928) – that state intervention was required to preserve the rural landscape from large-scale development and unsympathetic management was widely accepted, culminating in the Town and Country Planning Act of 1947 (Rowley 2006, 112-15). As well as introducing, for the first time, the idea of spatial planning, this also established Tree Preservation Orders, which allowed trees deemed to be of particular value to be preserved from felling. Although largely applied in urban areas, TPOs represented the triumph of the new attitude to trees. Those examples growing in prominent locations would now be valued for their appearance and the amenity benefits they brought. Rather than being felled when economically mature, they should be left to age, unhindered by the axe.

As a consequence of a complex range of factors, although all ultimately related to industrialisation and urbanisation, rural tree populations thus grew significantly older in the period after c.1850, a development which arguably had implications for their overall health. However, there is a potential flaw in this argument: while what we have said may be true of timber trees, it does not – superficially at least - apply to pollards, which had always been allowed to grow to a venerable age because they continued to be economically productive (albeit to an extent which depended on species) for several hundred years. Given that very large numbers of pollarded trees existed in the pre-industrial landscape, and that their numbers declined rapidly in the first half of the nineteenth century, then it might reasonably be argued that average tree age was actually much higher before 1850, and certainly before 1800, than it was to be by 1950: that is, looking at all trees (and not just at those managed as timber) average age, and the numbers of old trees, may actually have increased over time. However, it has been suggested by a number of researchers that active pollarding ensures that, in biological terms, the effected trees exhibit many characteristics of much younger trees. Periodic removal of the crown, that is, stimulates renewed vigour (Read 2008, 251). In Lennon’s words: Because the tree is regularly being cut back and the crown is constantly having to reform, pollarding can delay the emergence of the tree from the formative growth period. Where trees are continually pollarded the ring width will remain trapped in the formative cycle. This can extend the natural lifespan of the tree significantly and some of the oldest and largest trees in the country have been managed under this system for centuries. (Lennon 2009, 173) Such advantages would have been lost, and biological age would have increased, once regular management came to an end. Indeed, there are signs that , while old pollards was being celebrated from at least the early eighteenth century, their size generally increased through the nineteenth and twentieth century, as examples which were not taken down as redundant grew steadily older. James Grigor, writing in 1841, thus described an oak growing at Thorpe next Norwich as ‘One which we 195 may class with the most noted in our county’, but it had a circumference of only 16’ 3” (4.95 metres); another, at Taverham, described as ‘amongst the largest in Norfolk’, had a circumference of only 19’6” (5.9 metres); the ‘Great Oak’ at Thorpe near Gunton, ‘the oak of the county, one of the most extraordinary trees in this land of trees’ – was no more than 6.6 metres in girth; while the ‘Bixley Oak’ – ‘a tree that has been conceded to be of no common pretensions’ –only around 5.3 metres (Grigor 1841, 53, 57-8, 134, 270). He listed only three old oaks girths greater than seven metres, suggesting that Norfolk in the mid- nineteenth century, while it contained many more large old oaks than today, did not boast many which were larger than (or even as large as) those recorded in modern surveys: around thirty trees in the county are known today with girths greater than 7 metres, and many more with girths in the 6 – 7 metre range (Barnes and Williamson 2011). In the pre-industrial countryside pollards were primarily regarded in economic terms, and often turned into firewood when their productivity fell off in later middle age: writers like Thomas Hale were insistent that pollards should be replaced before they became old and hollow (Hale 1756, 41). But like timber trees, they now grew steadily older as intensive management declined, and as more romantic attitudes to nature and the countryside burgeoned. Such attitudes to trees had formerly been largely, although not entirely, restricted to specimens growing in gardens and parks: and today, in all the areas studied, a high proportion of ‘veteran’ trees are still to be found in parkland locations (Barnes and Williamson 2011, 132-7). The Panshanger oak (unquestionably a standard for it still survives) grew in woodland on the edge of Panshanger Park in Hertfordshire and was described in 1757 as ‘The finest oak in all this country … & not the least decayed’ (although its circumference, around five metre, suggests that it was less than three centuries old). Its owner had ‘made a grand Walk thro the coppice to it’ (Hampshire Record Office 9M73/958). In the grounds of the wealthy timber trees might be retained well beyond economic maturity and redundant old pollards, often incorporated from earlier hedgerows when parks were established, would be protected from conversion to firewood, and allowed to grow uncropped, because of their aesthetic value. In other places the life span of all trees was usually determined principally by economic considerations. The key change from the late nineteenth century was the gradual extension of such attitudes to a far wider population of trees.

The implications of senescence

The increasing age of farmland trees was manifested, in particular, in the growing incidence of ‘stagheadedness’ or dieback. While this development is difficult to quantify, photographs of the countryside dating from the late nineteenth and early twentieth century unquestionably show, by modern standards, remarkably few stag-headed trees. While it is true that there are numerous difficulties involved in interpreting images which, by their very nature, are to a degree selected and composed, and while for the most part trees form an incidental backdrop to a view, the relative youth of the trees recorded is striking: while a number of images show veteran trees, the majority of trees depicted appear not only healthy but also relatively young. By the post-War period, in contrast, photographs show far more over-mature trees, and the ageing character of trees in the countryside was becoming a matter of concern in a number of quarters. The Report of the Committee on Hedgerow and Farm Timber of 1955, using the statistics provided in the Forestry Commission’s Census of Hedgerow and Park Timber and Wood under five acres of 1951, emphasised the way in which trees felled or dying of age or disease were not being replaced by sufficient amounts of new planting, or natural regeneration. Dividing trees into four main size categories, ‘saplings and three separate quarter-girth classes’, they noted that ‘taking as a whole

196 there was an approximate distribution of 2: 1 : 1 : 1 in these four classes. An expert of repute has suggested to us that these ratios are insufficient to provide a continuous succession, and that a safer distribution would be 6 : 3 : 2 : 1’ (Forestry Commission 1955, 25-6). The report commented that the establishment of new trees was not helped by the fact that hedges were now more likely to be maintained by mechanical trimming than by hand.

The unpublished Forestry Commission Survey of 1972 suggests that the situation had by then significantly worsened (Forestry Commission 1972). It recorded that 43 percent of oaks in Northamptonshire were stag-headed and 2.5 percent were actually dead; in Hertfordshire, the figures were 34 percent and 5 percent; while in Norfolk they were as high as 65 percent and 7.5 percent. The figures for the other principal species of farmland tree, it is true, were not quite so alarming - elm was not so badly affected in Northamptonshire, where only 4.5 percent of trees were thought to be stag headed and a mere 0.35 percent dead, or in Hertfordshire, with 6.8 percent stag headed and 3.8 percent dead, although in Norfolk 41 percent were stag headed and 5 percent were recorded as dead. For ash, the figure for Northamptonshire were 24 percent and 1 percent dead; for Hertfordshire, 29 percent and 1 percent; and for Norfolk, 27 percent and 2.4 percent. The overall figures for all farmland trees combined were 17 percent stag-headed and 1 percent in Northamptonshire, 15 percent and 3 percent in Hertfordshire, and 45 percent and 6 percent in Norfolk: the reliability of the latter figures are perhaps confirmed by the Norfolk County Council survey of 1975, which identified 48 percent of oaks in the county as ‘Mature, stagheaded’, 9 percent of elm and 19 percent of ash, and 25 percent of all trees (dead trees were not listed: the other categories employed were ‘semi-mature’ and ‘mature, healthy’) (NCC 1975). The evident variations between the different counties may suggest that factors other than age were also important in the observed deterioration in tree condition, especially land use, with Norfolk having the greatest proportion of its area under arable cultivation and Northamptonshire the least. On the other hand, we should note that Norfolk (precisely because it was primarily an arable county) experienced a higher degree of estate and country house demolitions in the first half of the century than either Hertfordshire or Northamptonshire, and thus perhaps a greater decline, and at an earlier date, in the systematic management of its farmland timber. The Forestry Commission survey of 1972 suggested ‘general old age’ as a major reason for the poor condition of farmland trees, although it also acknowledged the importance of other factors, such as increased field drainage, ploughing too close to hedges, pollution and stubble burning.

It is within this context, of an ageing and under-managed rural tree population, that we need to consider some of the current concerns about tree health, and especially the conditions described as oak decline and ash dieback. These are principally, if not exclusively, restricted to trees a century or more in age and it is arguable that they are, to a significant extent, simply symptoms of normal old age, transformed into a ‘condition’ by modern and unrealistic expectations of perpetual arboreal health. In historical terms, such trees are certainly over-mature as timber; and while in an earlier age pollards were allowed to attain a greater age it is likely, as we have noted, that repeated cropping served to re-set their biological clock to a significant extent. More important, perhaps, is the fact that when tree populations were rigorously managed very few specimens would have exhibited symptoms of dieback or stag-heading for long because these things were taken as a sure sign that a tree was ready for felling. A tree exhibiting ‘decline’ was one whose useful growth was over. It was better in economic terms to have it down before the timber value was reduced by the onset of decay, and get another growing in its place. Early nineteenth-century correspondence relating to

197 the management of timber on the Bolton Estate in Yorkshire, for example, makes it clear that it was the ‘bad topp’d’ and ailing trees that were selected for felling (Dormer 2002, 197; 217-8; 223).As Moses Cook put it in 1676:

When a Tree is at its full Growth , there are several signs of its decay, which give you warning to fell it before it can be quite decay’d; as in an Oak, when the top boughs begin to die, then it begins to decay; in an Elm or Ash, if their Head dies, or if you see wet at any great Knot, which you may know by the side of the Tree being discolour’d below that place before it grows hollow …these are certain Signs the Tree begins to decay; but before it decays much, down with it, and hinder not your self (Cook 1676, 171).

Although ‘oak decline’ was only formally named and characterised in the 1920s, trees exhibiting the appropriate symptoms are often referred to in seventeenth- and eighteenth-century texts. Curtis in 1892 described how ‘dead upper branches or "stag-horn top," as it is usually called, is often met with … The manifestation needs but little remark, for it is apparent to all. The top branches die, the yearly growth is meagre, and the whole tree presents an enfeebled condition‘ (Curtis 1892, 25). It is noteworthy, however, that he drew attention to the prevalence of the condition, not on farmland or in woods, but ‘on lawns and pleasure grounds … and park lands’ – that is, in locations where trees were already, by the late nineteenth century, being retained beyond economic maturity on a large scale. The spread of the condition more widely, in other words, reflects the decline in intensive management of rural trees more generally, and an increase in the proportion of over-mature examples in the countryside as a whole.

Other environmental changes contributing to poor tree health As already noted, a number of environmental factors, in addition to age, have been suggested as contributing to the rise of oak and ash ‘decline’, and to poor health in trees more generally. These include falling water tables resulting from high levels of abstraction; deep ploughing on adjacent land, leading to root damage; and changes in cropping, with a shift to late summer cultivations and continuous courses of cereals leading to higher rates of moisture loss from the soil. Looked at over the medium to long term, two further environmental changes with possible impacts on tree health might be highlighted. The first is the overall increase in soil nitrogen levels which has certainly occurred since the middle decades of the eighteenth century, even if its precise scale cannot be quantified. The adoption of crops like turnips and clover was specifically intended to increase stocking levels, and thus manure supplies, by allowing more stock to be over-wintered; while clover also served to fix Nitrogen in the soil directly from the atmosphere (Beckett 1990, 11-18). From the 1820s, the use of manufactured oil cake as livestock feed rose steeply, further raising livestock numbers and thus soil nitrogen, national consumption rising from around 24,000 tons per annum in 1825 to 160,000 in 1870 (Thompson 1968, 73-4). From around the same time natural fertilizers began to be imported from abroad, in the form of guano from South America, while bone dust began to be applied to arable fields from the 1830s, and then, from the 1840s, a range of manufactured fertilizers, the use of which has continued with varying degrees of intensity until today. Nitrogen levels are thus far higher than they were a century ago not only on arable land but also more widely, due to the movement of water in the environment: indeed, as early as the 1850s there is clear evidence, in arable Norfolk, for serious eutrophication in the Norfolk Broads, to the consternation of contemporary naturalists (George 1992, 105-112). Increased Nitrogen levels may have implications for tree health given that accumulating research has demonstrated that inorganic

198 fertilisers can suppress the development of mycorrhizal fungi, damaging their reproductive parts (e.g. Ryden et al. 2003). A second development which may be relevant is the marked rise in the quantity of dead wood in the environment over the last century and a half. Before coal became the principal domestic fuel throughout England, every possible source of firing appears to have been exploited, especially by the poor. Thorns and shrubs growing on commons, furze, heather, turf, peat and even bracken were regularly cut or dug (Warde and Williamson 2014). Thefts of firewood, even of stacked faggots, continued to be reported in the county newspapers into the late nineteenth century, and laws against ‘hedge-breaking’ may have been directed at individuals attempting to acquire fuel as well as those trying to enter property illegally, or protesting against enclosures. In such circumstances, fallen branches and even twigs would have been rapidly removed from the countryside on a massive scale, and local courts frequently attempted to curtail the activities of the poor in this respect: in 1707 the manor court of Healaugh and Catterton in Yorkshire typically ordered tenants not to ‘carry away any wood green or dry out of my Lord's Park or any of his demains’ (Lds WYL68/33). Disputes over the rights to collect dead wood on common land and, in some cases, from enclosed woodland were often intense, as in the Northamptonshire forests, where the commoners’ right to collect only ‘sere and broken’ wood was very loosely interpreted (Pettit 1968). The way in which parliamentary enclosure awards both here, and in other areas, allocated ‘fuel allotments’ to the local poor attests the extent to which commons had been used as a source of firing (Birtles 2003, 304-9). Even on land where such rights did not exist, or were restricted to those with commonable tenements, landowners often turned a blind eye to the wood-collecting activities of the poor: Humphry Repton, in his Red Book for park in north Norfolk of 1811, suggested that the poor should be admitted into the estate woods once a month to gather fallen boughs, under the watchful eye of the gamekeeper (Repton 1816, 204). Although modern conservation practice rightly encourages the preservation of dead wood, especially when still part of live trees, the steady accumulation of large quantities of such material in woodland and hedges is, in historical terms, a relatively recent development with possibility for tree pathogens. The native buprestid beetle Agrilus biguttatus, implicated as a factor in acute oak decline, was until recently considered a ‘red book’ species, to be encouraged by the retention of fallen wood. An earlier generation of foresters, such as Curtis, writing in 1892, were clear about the potential threat posed by large accumulations of decaying wood to tree health: ‘At the risk of repetition I would impress upon all foresters the necessity of cleaning up after every fall of timber, and the total destruction by fire of all dead organic matter’ (Curtis 1892, 46).

It is unlikely that any of the changes discussed over the last few pages, including the marked rise in average age of trees, can have been significant influences on the scale or incidence of the various epidemics, caused by invasive organisms, which have appeared in this country over the last half century. Dutch elm disease and ash chalara, in particular, appear to infect young specimens at least as much as senescent ones. Indeed, the current impact of the latter disease has mainly been on immature trees. This said, it might be noted that if trees displaying signs of such illnesses were immediately cut down, this might serve to retard the spread of some pathogens, and such rigorously managed tree populations would at the very least appear to be healthier than under-managed or wild ones. Indeed, in modern plantations ‘quarantine felling’ is generally less regularly practised than in earlier periods, when foresters like Curtis recommended it as the best way of dealing with problems like Hylesinus fraxini, Chionaspis fraxini, and Scolytuss. The message from history may be,

199 not simply that disease is a natural condition of trees, but that the most unnatural and most rigorously managed tree populations are also the most healthy ones.

Conclusion

There seems little doubt that major epidemics caused by invasive pathogens did not occur in England before the twentieth century, or at least, not within the three and a half centuries studied here. Nevertheless, there have always been localised episodes of severe mortality, and in a more general sense poor health in tree populations appears to have been endemic, and accepted as normal in the past – more normal, perhaps, than we find it today. Long-distance movement of plant materials and timber seem to be the key factors in the arrival of new foreign pathogens, as many have argued before us. Such movements have taken place for centuries, but the increased speed, and perhaps volume, of traffic in the later nineteenth and twentieth centuries appears to have been critical in the escalating incidence of such events. This said, other factors may have encouraged a decline in arboreal health in England. A reduction in rigorous and commercial management, especially of farmland trees, in the period since the late nineteenth century, and a resultant increase in the age of trees, appears to have been especially important. In addition, levels of disease may have risen with the planting, from the late eighteenth century, of extensive plantations composed of single species like larch and Scots pine; with increases in soil nutrient levels which have occurred over the same period; and perhaps with the rise in the quantity of dead wood present in the environment. All this is in addition, of course, to the range of causes more commonly cited for such things as oak decline, including excessive abstraction of water, continuous cultivation for cereals with late summer ploughing, deep ploughing and the use of heavy machinery.

An historical perspective also, however, suggests that an equally important factor behind the scale of recent epidemics is the extremely limited composition of English tree populations, in terms of species. Surveys and maps from the late seventeenth, eighteenth and nineteenth centuries from the four counties studied almost invariably show that oak, ash and elm together made up between 85 and 100 percent of the trees growing on farmland, and virtually all of the trees (as opposed to coppiced underwood) found in enclosed woods. This dominance has to some extent been reduced in the period since the 1960s, and not just as a result of the collapse of elm populations: there has also been a rise in the numbers of trees like maple which have been planted or allowed to grow in particular locations. But in most places, ash and oak remain massively dominant, representing a lack of diversity which renders our treed landscapes extremely vulnerable. If some serious new disease was to decimate our wild service trees or our black poplars, it is probable that few people would notice; but so scattered are examples of these species that it is unlikely that any such disease would spread very rapidly. Neither of these observations could be applied to oak or ash. The dominance of the farmed landscape, over vast tracts of the country, by these two species arguably poses as much of a problem for the future as the arrival of any new pathogens.

200

Figure 40. Landscape in north Norfolk, almost entirely dominated by oak trees.

201

8. Conclusion

The character of the trees growing in our countryside is often considered to be the outcome of largely ‘natural’ factors but most were deliberately planted, and even where they were not their continued presence was a consequence of human choice, for in many contexts trees interfered with food-production or other necessary activities. Pre-industrial tree populations in woods, on commons or on farmland were very intensively managed, for wood or timber, and in most cases trees were regarded exclusively or at least primarily in economic terms. The ways in which trees were managed, the species planted and the numbers tolerated, were the consequence of the interplay of a complex range of influences. One major factor was the patterns of land use, tenure and landscape inherited from the medieval past. In some areas woods were thus frequent at the start of the period studied here, and in others rare. In some districts much of the land lay in enclosed fields and hedges were abundant, while in others champion, ‘open field’ land, with far few hedgerows, dominated the landscape, providing fewer opportunities for trees to grow. In some districts there were few commons, but in others they were extensive and often to varying degrees tree-covered. In the course of the sixteenth, seventeenth and eighteenth centuries changes in these basic landscape patterns, and especially the increase in hedgerows with the progress of enclosure, had a major impact on the character of tree populations. But so too did a range of other social, economic, technological and tenurial changes.

The density of farmland trees in the preindustrial landscape displayed a considerable degree of variation. In some areas, such as the old-enclosed districts of Norfolk and Hertfordshire studied here, very large numbers of trees generally existed in the late seventeenth and eighteenth centuries, when our records first become abundant; usually more than 15 per hectare, often above 20 and in some districts in excess of 25 or even 50. Such trees were, moreover, generally scattered fairly evenly across the landscape, mainly growing within a relatively tight mesh of hedges but sometimes in ‘rows’ around field margins or densely scattered across pasture fields. In ‘champion’ open-field districts, in contrast, there were far fewer hedges and thus fewer trees on farmland although, as we have emphasised, rather more than is normally assumed. Large numbers existed in the tofts and yards of the village ‘envelopes’, and others could be found in the wider landscape, on wastes and roadsides, in hedges planted around the great fields and on parish boundaries, and in particular on floodplain meadows, where vast numbers of willows often grew. Averaged across the area of individual parishes tree densities in unenclosed landscapes could reach 4 or 5 per hectare but for the most part, in the Midland ‘core’ of the champion – exemplified in this study by the county of Northamptonshire - they were generally between 1 and 2.5 per hectare. But these trees were also, it should be noted, strongly concentrated in particular areas of the landscape, providing much lower levels of ‘connectivity’ than in old-enclosed districts.

In the course of the sixteenth, seventeenth and eighteenth centuries open fields were gradually enclosed, but this did not mean that the density of trees in these areas necessarily increased to the kinds of levels described in districts of ‘ancient countryside’ like Hertfordshire. Some early enclosures – because they led to the destruction of settlements, and thus to the removal or neglect of the hedges around tofts and crofts - may actually have led to a net reduction in the numbers of trees. While this was less true where parishes were enclosed after the mid seventeenth century, trees numbers seldom climbed, within the areas studied, above 10 per hectare; were sometimes less than 5; and, in northern ‘champion’ areas such as the Vales of York or Mowbray, generally less than 2.5 trees per hectare. This was part of a wider pattern, and one which was not primarily related to any

distinction between ‘woodland’ and ‘champion’, for even long-enclosed districts in Yorkshire, on the fringes of the uplands or in the Dales, seldom boasted tree densities exceeding 5 per hectare. In other words, by the middle of the eighteenth century most parts of Norfolk and Hertfordshire boasted very high numbers of farmland trees, newly-enclosed areas of Northamptonshire medium numbers, and Yorkshire very few.

The management of farmland trees displayed both broad similarities across the various areas studied, and also significant regional variations which, quite closely, mirrored those in overall tree densities. The broad similarities relate to timber trees. In all areas these were, before the nineteenth century, almost invariably felled young – generally before they were seventy years old, and often long before. This form of management seems to have been dictated by a desire to minimise sawing, by harvesting trees at sizes appropriate to the dimensions of the timber required; by a reluctance to tolerate very large, spreading trees on farmland or in village closes; by an awareness that, after eight decades, growth rates of most species begin to decline; and, in the case of oak, by a desire to maximise the production of bark. The variations in management relate to pollarding. In the oldenclosed districts of Hertfordshire and Norfolk over 60 percent of trees were generally managed in this way, and in most areas at least 70 percent. In some places, such as south Hertfordshire, the numbers were higher still, reaching 80 percent or more on some properties. In Midland open-field areas, in contrast, even after enclosure, the proportion of pollards was significantly less, generally around 25 percent and only rarely exceeding 50 percent. In Yorkshire the situation is more complex and the data less abundant: there are signs that in the sixteenth and seventeenth centuries there may have been quite high numbers of pollards in some districts, most notably the Dales, but in the ‘champion’ Vales the situation was probably comparable to that in Midland areas like Northamptonshire. Either way, by the end of the eighteenth century pollards were a rare feature of the Yorkshire landscape, scarcely mentioned in documents.

To a significant extent, differences in the overall density of farmland trees were simply a function of how many pollards were present. The numbers of timber trees, that is, were broadly comparable everywhere, at between 2 and 6 per hectare, but to varying extents in different areas the population added pollards to this base level, sometimes to a degree that must surely have impacted on the yields of crops and grass in the adjacent fields. Their main reason for so doing was to produce fuel. Where other sources of fuel were abundant, the need for pollards was less. Yorkshire, with abundant supplies of peat and moorland vegetation, and the early development of coalfields, had little need for firewood, and thus few pollards, at least by the later eighteenth century. In Northamptonshire communities had some access to coal, from the Warwickshire fields, at an early date, and some could exploit peat in the great Fens to the east, or the woods and wood-pastures of the forests. Nevertheless, a significant proportion of firing was supplied by farmland trees. In the old-enclosed areas Norfolk and Hertfordshire there was limited access to coal and little in the way of peat or other alternative sources of firing. Not surprisingly, pollard numbers were high. There are signs of more local variations in pollard numbers, within these counties, which can also be explained in these terms. North-east Norfolk was reasonably well endowed with heaths and fens in the seventeenth and eighteenth centuries and accordingly had noticeably lower numbers of pollards than south Norfolk, where no such alternative fuel sources existed. In Hertfordshire the highest numbers of farmland pollards were in the south of the county, reflecting the distorting effect on the market for wood and charcoal provided by demand from London, only 15 kilometres or so to the south. Factors other than fuel supply may, however, also have contributed to these patterns,

203 including local styles of vernacular architecture. Small pieces of oak wood, as opposed to timber, were used in the construction of timber-framed houses – for scantling and studs – and it is noteworthy that areas in which pollards were relatively rare also tended to be ones with a high proportion of stone buildings by the seventeenth century at vernacular level.

The various counties, and their geographical subdivisions, also displayed much variation in the extent of woodland and wood-pasture. These were largely a function of complex social and economic developments during the millennium leading up to 1550. It is generally assumed that ‘champion’ areas were poorly-wooded and that old-enclosed, ‘woodland’ districts were extensively so, but to some extent woodland of various kinds was simply more clustered in champion areas, especially in royal forests. In ‘woodland’ districts, in contrast, such as Hertfordshire, it was more evenly distributed. Although estimates are difficult to make, given the nature of the evidence, ‘woodland’ Hertfordshire and ‘champion’ Northamptonshire may thus have had a broadly similar proportion of their land area – ten and eight percent respectively - occupied by woodland of various kinds in c.1550. Norfolk, for reasons largely lost in the medieval or earlier past, had slightly less, around 7 percent; Yorkshire, perhaps in part once again because of an abundance of alternative fuels, probably around half this density, although the different Ridings displayed significant contrasts in this respect, with the East Riding having very little woodland and the North and West Ridings rather more. Indeed, in all counties there were marked contrasts in the density of woodland from district to district, although overall these followed broadly similar patterns. Areas of light chalk, or sands, deforested for the most part before the Middle Ages, had very little woodland of any kind by 1550 (the ‘champion’ areas of north Hertfordshire; the Yorkshire Wolds; west Norfolk). Woodland was also sparse on the highest moorlands in Yorkshire, clustering instead on the more sloping ground on the margins of the main upland masses. In the Midlands and the south woodland was especially abundant on heavy boulder clays, although also - usually in the form of wood-pasture – on acid, gravelly soils, especially where these occurred on high, remote watersheds. It is almost impossible to estimate the extent of woodland and wood-pasture in c.1550, but it may, averaged across the four counties, have amounted to no more than c.4 percent of their combined land area.

The distinction between enclosed woodland, managed as coppice, and wood-pasture or grazed woodland, was well-established by 1550 in most of the areas studied, and in many cases the two varieties of woodland displayed markedly different distributions, with the greatest concentrations of wood-pasture generally occurring on high watersheds, especially in Northamptonshire (in the three royal forests) and Hertfordshire (on the high round in the south of the county, and the crest of the Chiltern Hills), but with enclosed woods generally located in more convenient areas, closer to the principal settlements. Some wood-pastures were found within private deer parks but most were on common land. In the period after 1550 a significant proportion of the former were converted to agricultural land or to more open, ornamental grounds; while much of the latter gradually degenerated to open pasture, due to poorly controlled felling and grazing. Common wood-pastures survived best into the eighteenth century where strong organisational structures existed, either under lordly direction or controlled by commoners; and also where there was a good market for fuel. South Hertfordshire, where both these circumstances applied, thus retained large tracts of wooded common into the eighteenth century. The Northamptonshire forests in contrast, poorly if not chaotically managed and regulated, became increasingly denuded of trees.

Private coppiced woodland was more intensively and commercially managed than wood-pastures, either by owners or by their lessees. Livestock were rigorously excluded for much or all of the time

204 and this, as much as regular coppicing per se, crucially shaped aspects of their ecology, especially the presence within them of a particular range of vascular plants, the so-called ‘ancient woodland indicators’. Here, as in the wider countryside, timber trees were usually felled when scarcely mature but in some woods, at least, a proportion were allowed to reach a greater age – 80 or even a hundred years – to be used in specialised ways. Some of the cut coppice provided domestic fuel but much had specialist uses and this, as much as differences in species composition, may explain the variations in the length of rotations from region to region. Rotations in Yorkshire were commonly twice the length of those practised in the southern counties: larger poles could be used for props in the local mines, and also produced better charcoal and whitecoal, and (in the case of oak) more bark, all in high demand in this industrialising area.

In many districts there are signs that a reduction in the area of coppiced woodland took place over the two centuries following 1550, possibly amounting to as much as 20 percent of woodland area in Norfolk and Hertfordshire. Because woods were private property they could be grubbed out and converted to other uses by their owners, in whole or part, in response to perceived short- or medium-term economic advantage. Wood-pastures likewise declined, through gradual degeneration or deliberate destruction, so that the proportion of the four counties combined occupied by woodland may have fallen from around 5 percent in 1550 to around 3.7 percent by the second half of the eighteenth century. It must be emphasised however that, given the nature of the sources, the former figure in particular is highly speculative.

Coppiced woods and most wood-pastures were thus intensively managed environments, the former exclusively to produce timber and wood, the latter also for grazing and other produce. But the wider countryside was also, in enclosed districts, a major source of timber – from trees in hedges and freestanding in pasture fields – and of wood – from pollards, and also from hedges, which were managed as a major fuel source in most districts. It is not surprising, then, that rural trees populations, both in woods and on farmland, were very different from those found in the ‘natural’ woodlands which had existed before the impact of farming. Indeed, one of their most striking features, considered in the long term, is the overwhelming dominance of just three species, oak, ash and elm. In most districts, at least since the seventeenth century, these constituted between 85 and 100 percent of the trees recorded on farmland, ignoring commons and meadows, and almost all of the timber grown in managed woods. On farmland, the relative importance of the three species displayed a measure of regional variation: in particular, elm was rare in the north, and while oak was in most districts the most important, or the second most important, species, in Northamptonshire (and probably other Midland areas) it was routinely the second or third most common. In all districts, however, oak was - by a wide margin - the dominant timber tree in managed woodland. As we have argued, the importance of these three species was not the consequence of ‘natural’ factors, of their ability to thrive in cultural landscapes. Most trees were deliberately planted, or permitted to grow in a particular location through conscious choice. The example of maple is instructive here. It seeds freely in hedges but it was seldom allowed to grow into a tree: instead it was plashed or coppiced with the other shrubs in the hedge. Oak, ash and elm (the two main species of oak, and of elm, not distinguished in the documents) were selected because they had the greatest range of practical uses, and could tolerate a wide range of conditions, in terms in particular of soils and drainage. Other species were more choosey in their requirements, had fewer uses, and could offer little that these could not provide, or were (like maple) thought to be better managed as coppice, in woods or in hedges.

205

The dominance of oak, ash and elm was not total. On farmland, in almost all areas, other species were present in small numbers, either self-seeding in hedges and somehow surviving there or, more usually, deliberately planted or maintained because they provided some particular product required by landowners or farmers. The character of these ‘minority’ trees displayed a measure of regional variation: in Hertfordshire, for example, the most important species in the west of the county, on the Chiltern dipslope, were aspen and cherry; but in the east, on boulder clay soils, maple, black poplar and hornbeam were more typical. In addition, low-lying wetlands were characterised by large numbers of willows (especially numerous in the open-field Midlands) and alders (where soils were more peaty and acidic); while wood-pastures sometimes had their own range of trees, selected and in some cases deliberately planted because they had good resistance to grazing and browsing, or because – in deer parks – they provided feed for deer. Hornbeam and beech, usually rare as trees in the wider countryside and as a timber trees (as opposed to coppiced underwood) in enclosed woods, were classic species of wood-pastures.

We would not wish to exaggerate the anthropogenic character of tree populations: soils and climate, as we have emphasised, also played a part in deciding which trees grew where. But in large measure the species of trees found in woods, wood-pastures and on farmland owed more to human choice and human management than to any other factors. Even the underwood in coppiced woodland may have been deliberately modified by human intervention, including replanting, or as a by-product of management. Given all this, and the close connections between the character of trees and social and economic systems, it is quite possible that the particular balance of species, and the particular modes of management, which we have derived from maps and documents dating mainly from after the mid-seventeenth century, may not have pertained in earlier periods, in different economic circumstances. The gradual expansion of coal production, in particular, may well have engendered greater regional variations in the extent of pollarding than in earlier centuries. This said, many of the patterns and processes we have highlighted, such as the distinctive character of wood-pasture trees, do – to judge from the available evidence – appear to have been true of earlier periods.

The character of English tree populations was changed radically in the period between the late eighteenth and the late nineteenth centuries. There was, from the start of the eighteenth century, a steady expansion of the area under woodland, although to some extent this was cancelled out by the continued degeneration or deliberate destruction of wood-pastures, and the continued grubbing out of long-established coppiced woods, most dramatically following the enclosure of the Northamptonshire forests. The overall result, averaged across the four counties, was a modest increase in the proportion of the land area devoted to woodland of various kinds, from around 3.7 percent to around 4.2 percent between the 1780s and the 1880s. Woods tended to be lost from areas of fertile boulder clay, and they tended to proliferate on the kinds of light, chalky and sandy soils, and in upland districts, where they had formerly been rare; wood-pastures of all kinds went into sharp decline. Overall, the period thus saw – to some extent - an evening-out of woodland across the landscape, with losses in forests and old-enclosed areas, and an increase in ‘champion’ districts and in areas of marginal land formerly occupied by common heaths and moors. While woods of traditional, coppice-with-standards type continued to be established, most of the new planting consisted of plantations, without a coppiced understorey. These, moreover, were planted with a more diverse range of species than was found in old-established woods, including both nonenative conifers like spruce and larch and indigenous or naturalised species which now spread being beyond their former ranges, especially beech, sweet chestnut and Scots pine. In some cases,

206 stands of pure conifers were established, especially on the most marginal land. Much of this new planting was essentially commercial in character but in many cases it was motivated, in whole or part, by considerations of status, aesthetic and amenity. Most was carried out by large estates and ultimately reflected the progress of enclosure and the consolidation of landownership in larger units.

The second key change of the late eighteenth and nineteenth centuries was the consequence of industrialisation, and of a growing enthusiasm for agricultural ‘improvement’. The numbers of pollards on farmland declined dramatically, as improvements in transport ensured that coal became the principal domestic and industrial fuel in all districts. The overall numbers of farmland trees also fell, partly although not entirely as a consequence of this change, in many areas by more than 70 percent in the period between the late eighteenth and the late nineteenth century, although perhaps with different chronologies in different regions. The consequence was a much less densely timbered countryside. Few districts could, by the 1880s boast more than 8 farmland trees per hectare, and in most the figure was between 3 and 7.

Thirdly, there were changes in the average age of timber trees, both in woods and on farmland, which rose markedly from the 1850s. The development of industrial sawmills and better transport systems, together with the declining value of bark, all discouraged landowners from felling trees when they were young, instead allowing most to grow until they were eighty years old or more.

If the hundred years between the late eighteenth and the late nineteenth centuries saw radical changes in the character of England’s trees, developments in the course of the following hundred years were, if anything, more far-reaching. They were the consequence, once again, of complex social and economic factors, including the increasing scale of timber imports; the break-up of large landed estates and changing land-use priorities; growing urbanisation and an associated divorce from, and romanticisation of, the ‘rural’ and the ‘natural’ on the part of opinion-formers; and a steady expansion in the involvement of the state in land use and land management. Perhaps the most important development was an acceleration in the establishment of new areas of woodland, initially as a consequence of the activities of the Forestry Commission, subsequently as a result of the bushing over of surviving areas of common land and other rough or derelict land and, in the last decades of the century as the outcome of government-funded agro-environment schemes and private initiatives, intended to enhance the environment or provide cover for game. Although in the popular mind England’s woodlands may be shrinking, in reality the area of land across the four sample counties devoted to woods and plantations of various kinds has risen steadily since the late nineteenth century, from just over 4 percent of the combined land area of the four counties to around 5 percent by the middle of the twentieth century, and reaching over 8 percent today. Of course, it can be legitimately argued that there has been a deterioration in the character of this woodland, in environmental terms. Much of the new woodland (although to a declining extent, over time) comprises pure conifer plantations, while the active management of coppices in ancient woodland had largely ceased by the 1960s. This said, it is clear that, considered in historical terms, that woodland per se is not in short supply: what may be lacking, arguably, are forms of management which could best promote biodiversity, and provide other social benefits.

The fate of farmland trees over the last century or so has been more complex, and our interpretations are, to a large degree, dependent on definitions. The impact of such things as the opportunistic fellings following the break-up of large estates in the first half of the century, and during two world wars; or of Dutch elm disease; were compounded in the first half of the century by

207 a general reduction in the planting of free-standing timber trees in hedges, as forestry became increasingly concentrated in woods and plantations. By the 1960s and 70s farmland trees were regarded – depending on perspective – as useless encumbrances, unnecessary hindrances to the real business of the countryside; or as timeless, natural adornments of the landscape. As a result of all this, not only did the numbers of farmland trees decline to very low levels – generally around 1 or 2 per hectare in most districts by the 1970s – but a high proportion comprised over-mature specimens. The complication is that on some measures, but not others, this situation was reversed during the following four decades. While individual, free-standing trees in managed hedges have continued to decline, or have at least displayed no marked recover, a reduction in or cessation of management has allowed some hedges to develop into lines of close-set trees; while at the same time new forms of planting – ‘linear features’ and ‘groups’, to use Forestry Commission terminology, clusters of trees too small to be counted as woods -– have appeared in many areas. If the individual trees in these various features are included in the calculation, then the numbers of farmland trees have recovered sharply, returning in some places to their late nineteenth-century levels. It must be emphasised, however, that these generally have a much more clustered distribution, ensuring low levels of connectivity in the landscape; and that, on the ground, the character of recent changes has displayed much local variation, relating to patterns and types of ownership, so that farms now wellendowed with trees, many of them recently planted, can exist beside holdings sparsely populated by over-mature specimens, mainly oaks.

Close examination of both documentary sources, and early texts on forestry, make it clear that, within the period studied here, no serious, national epidemic disease of trees occurred in England prior to the outbreak of Dutch elm disease in the late 1960s and 70s. This said, it is also clear that poor health in trees was common and considered more normal in the past than it often is today. Scolytus, Tortrix, and various forms of rot have always affected trees: but when populations were vigorously managed, diseased specimens were simply felled and sold with the first onset of serious symptoms. There are some signs that the development of plantation forestry – the planting of large numbers of trees of one species on often marginal, acidic land – may have increased the incidence of disease during the eighteenth and nineteenth centuries, but it is difficult to dissociate any real deterioration in tree health from the burgeoning interest in and knowledge of pathology associated with the rise of scientific forestry, and from the increasing optimism about what this could achieve. The twentieth century, however, did unquestionably see a steady decline in arboreal health. The onset of globalisation and the increasing speed with which plant materials and timber could be transported ensured the arrival of new pathogens, starting with oak mildew in 1908 and the first serious outbreak of Dutch elm disease in the 1920s. The rising scale, and speed, of international trade – possibly compounded by the effects of global warming – have, as we have seen, led to further epidemics in the later twentieth and early twenty-first century. Equally important, however, has been a more general decline in the health of trees, with the appearance of stag-headedness, oak and ash dieback, and acute oak decline, none of which have yet been plausibly related to invasive pathogens.

An historical perspective on tree health, besides confirming that the current and recent scale of epidemics is unprecedented (at least in the period studied), can contribute to our understanding of these issues in three key ways. Firstly, it seems likely that some of the more general signs of ill health in trees are a function of management and perception, rather than of new pathogens. Most of the chronic, ill-defined conditions like ‘decline’ affect trees which are more than 100 years old. Such

208 trees would have been rare before the late nineteenth century, when trees were either pollarded – and thus, in many ways, kept in a state of perpetual juvenescence – or were taken down when still relatively young. It is also clear, moreover, that when farmland trees were intensively managed, specimens suffering from ill health were simply felled and sold with the first onset of serious symptoms. Indeed, stag-heading and ‘bleeding’ were taken to be signs that it was time to fell, and what we perceive as symptoms of illness may, to an extent, be the normal signs of ageing. In the ‘wildwood’, whatever precise form this may have taken, a very high proportion of trees may have exhibited such characteristics.

An historical perspective also serves to highlight some neglected changes in the wider environment which may have contributed to the susceptibility of trees to disease, most importantly a marked increase in the quantity of dead wood in the countryside, and rising levels of soil nitrogen. Above all, it serves to reveal the highly artificial character of rural populations and their highly restricted character, massively dominated by only three species in the period up to the 1970s, and now by just two – for in spite of some increase in diversity over the last half-century, farmland trees are still overwhelmingly, in most districts, dominated by oak and ash. This has ensured that our tree populations have very little potential resilience in the face of future invasive pests and pathogens, such as emerald ash borer. It is true that the artificial dominance of these trees in the landscape – the consequence, primarily, of past economic factors – has been reduced to some extent since the nineteenth century, with more diverse new plantings, and more recently with natural regeneration of species like maple where hedges have been neglected. But more radical attempts to diversify the character of farmland trees are now urgently required, not least in the face of chalara which might, in the worst case, leave oak as the sole mature hedgerow tree across large parts of the landscape.

Recommendations

Arising from these findings, we would suggest that the following actions might be considered by the appropriate agencies in the future, as resources become available.

• Although raw figures for the numbers of farmland trees present in particular areas often suggest a marked recovery since the 1970s, these are in some ways misleading. Increases are a consequence of small group planting and the outgrowth and neglect of hedges. More planting is urgently needed, of trees which are scattered more evenly across the landscape: this would not only increase biological ‘connectivity’, and would be more in keeping with the long-term character of the English countryside – captured, for example, in Constable’s iconic paintings. • Future plantings should be made more diverse, to provide greater resilience in the face of likely pathogens. In the past, tree populations were largely shaped by economic and practical considerations. We are now obliged to make radical changes to the composition of our farmland tree populations, due to increases in the incidence of invasive pathogens, but we are no longer constrained by such considerations. Amenity, biodiversity, and aesthetics can be the main drivers.

• While future planting should continue to prioritise oak; and while we believe that we should also persevere with the planting of ash, at least until the scale of the impact of chalara becomes clear (using the more resistant strains as these become available); these trees should together make up no more than a half or perhaps two thirds of new planting. The

209

remainder should comprise other species, in order to provide a measure of diversification and resilience. • Some authorities currently advocate that diversification should be achieved by the widespread establishment of non-native species, in anticipation of the likely effects of climate change. We would urge caution here: it seems unwise to make such a radical departure, with uncertain ecological effects and clear cultural disbenefits, given that many indigenous and long-naturalised species have ranges which extend far beyond the UK, and that the speed of warming remains uncertain. • Nor should a standard palette of native species be introduced in all contexts, including trees quite alien to the areas in question. We would suggest instead that emphasis should be placed on the planting of the ‘minority’ trees which have long been characteristic of particular areas and districts. Thus in Yorkshire sycamore and holly might be widely planted, and maple, willow and alder more locally: beech, although a relatively recent arrival, is widespread in many districts as a hedgerow tree and might also be more widely established. In Hertfordshire, aspen, cherry, beech and perhaps apple might be appropriate trees in hedgerows in the west; on the boulder clays in the east, black poplar, hornbeam and maple (the latter anyway widespread as a hedge shrub). In other areas, different ranges of species might be planted, following a limited amount of documentary research. Such species are ‘tried and true’, and evidently well suited to the localities in question. Planting these, rather than an indiscriminate ‘diversity mix’, would ensure a continuing measure of local and regional distinctiveness in the character of farmland trees, so important for sustaining a ‘sense of place’. • In addition, attempts might be made to enhance or recreate some other long-term characteristics and features of ‘traditional’ treed landscapes. Thus the willows which have long been a major feature of the great Midland floodplains should be replanted on a large scale, and attempts might be made to restore, in some form, the long-lost stands of hornbeam trees which once existed on the uplands of south Hertfordshire. There are many other examples. • In general, our advice is thus that planting should be informed by history, but without simply replicating the character of past tree populations which, as noted, has long been dominated by a narrow range of species. This said, in particular circumstances trees which have never featured widely in farmland might also be increased, in areas where they are likely to flourish, such as small-leafed lime and the magnificent wild service tree. • Scientific research might usefully be directed towards assessing whether, as we suggest in the main report, ill-defined conditions such as ‘oak decline’ may in part be artefacts of management, and of perception, rather than simply symptoms of pathogens or the consequence of specific environmental changes. Research might also explore the possible impacts of two other key developments noted in this report: the general rise in soil nitrogen levels over the last two centuries, and the increase in the amount of dead wood present in the environment. • We suggest that commercially-managed populations of farmland trees are more rigorous and resilient than un-managed ‘natural’ ones. Some attention might thus be paid not simply to providing incentives to landowners to plant and manage for conservation and amenity, but (at least in part) for reasons of commercial forestry. In some ways, conservationist sentiment may achieve less in the long term than commercial management. Looked at in

210 one way, devices like Tree Preservation Orders may be part of the problems we face, rather than part of their solution.

211

References and Bibliography

Secondary Sources

Abercrombie, P. (1943). Town and Country Planning. London: Home University Library.

Adams, B. (ed.) (1997). Lifestyle and culture in Hertford … Wills and inventories for the parishes of All Saints and St Andrew, 1660–1725. Hertford: Hertfordshire Record Society.

Addington, S. (1978). ‘The Hedgerows of Tasburgh’. Norfolk Archaeology 37, 70 – 83.

Albert, W. (1972). The Turnpike Road System in England 1663–1840. Cambridge: Cambridge University Press.

Allison, K.J. (1957). ‘The Sheep-Corn Husbandry of Norfolk in the Sixteenth and Seventeenth Centuries’, Agricultural History Review 5, 12-30.

Annard, J. F. (1910). ‘Observations on the Larch Sawfly with suggestions for remedial and preventive treatment in infested larch woods’, Quarterly Journal of Forestry 4, 203-221.

Anon. (1908). ‘Report on injurious insects etc in the midland counties, 1907’, Quarterly Journal of Forestry 2, 199-200.

Anon. (1907). ‘The Larch disease’, Quarterly Journal of Forestry 1, 38-46.

Ashby, M. and Cudmore, A. (2011). ‘New Towns’, in D. Short (ed.), An Historical Atlas of Hertfordshire. Hatfield: University of Hertfordshire Press, 166-7.

Aveline, W. T. and Trench, R. (1860). The Geology of Part of Northamptonshire. London.

Austin, P. (1996). ‘The Leasing of Lord Burghley’s Hoddesdon Woods in 1595. An insight into woodmanship’, Hertfordshire’s Past 41, 11-21.

Austin, P. ( 2001). ‘Traditional Woodland Management. The woods of the Manor of Digswell 1776– 1843’, Hertfordshire’s Past 51, 2-12.

Austin, P. (2008). ‘Traditional Woodland Management: the timber trees of Panshanger in 1719’, Hertfordshire Past and Present 12, 3-9.

Austin, P. (2013). ‘Traditional woodland management in south-east Hertfordshire. Herts Past and Present. 51, 2-12.

212

Bailey, M. (1998). ‘Introduction’, in J. Brooker and S. Flood (eds), Hertfordshire Lay Subsidy Rolls 1307 and 1334. Hertford: Hertfordshire Record Society.

Baird, W. and Tarrant, J. (1970). Hedgerow destruction in Norfolk 1946-1970. Norwich: Centre of East Anglian Studies.

Bailey, B. (ed.) (1996). Northamptonshire in the Eighteenth Century: the Drawings of Peter Tillemans and Others. Northampton: Northamptonshire Record Society Volume 39.

Banham, D. (20100. ‘In the Sweat of Thy Brow Thou Shalt Eat Bread: Cereals and Cereal Production in the Anglo-Saxon Landscape’, in N. J. Higham and M. J. Ryan (eds) The Landscape Archaeology of Anglo-Saxon England. Woodbridge: Boydell, 175-92.

Barnes, P. (1984). ‘The Economic History of Landed Estates in Norfolk since 1880’. Unpublished PhD thesis, University of East Anglia.

Barnes, G. and Williamson, T. (2006). Hedgerow History: ecology, history and landscape character. Macclesfield: Windgather Press.

Barnes, G. and Williamson, T. (2011). Ancient Trees in the Landscape: Norfolk’s arboreal heritage. Oxford: Windgather.

Barnes, G. and Williamson. T. (2015). Rethinking Ancient Woodland: the archaeology and history of woods in Norfolk. Hatfield: University of Hertfordshire Press.

Barnes, G., Dallas, P. and Williamson, T. (2009). ‘The black poplar in Norfolk’, Quarterly Journal of Forestry, 103, 31-8.

Barrett-Lennard, T. (1921). ‘Two hundred years of estate management at Horsford during the 17th and 18th centuries’, Norfolk Archaeology 20, 57-139.

Beckett, J.V. (1990). The Agricultural Revolution. Oxford: Wiley-Blackwell.

Beaver, S.H. (1941). The Land of Britain, Part 46: Yorkshire, West Riding. London: Geographical Publications.

Beaver , S.H. (1943). The Land of Britain Part 58: Northamptonshire; Part 59, the Soke of Peterborough. London: Geographical Publications.

Beckett, J.V. (1984). ‘The Pattern of Landownership in England and Wales 1660-1880’, Economic History Review 38, 1-22.

Beckett, J.V. (1986). The Aristocracy in England 1660-1914. Oxford: Wiley-Blackwell.

Beckley, D. (2002). ‘ and Whomerley Wood: an historical and field survey in Stevenage’, Hertfordshire's Past 53, 15-21.

213

Bennett, K.D. (1983). ‘Devensian Late Glacial and Flandrian vegetational history at Hockham Mere, Norfolk, England 1. Pollen percentages and concentrations’, New Phytologist 95, 457-87.

Bennett, K.D. (1986). ’Competetive interactions among forest tree populations in Norfolk, England, during the last 10,000 years’, New Phytologist 103, 603-20.

Bird M.C.H. (1909). ‘The rural economy, sport, and natural history of ’, Transactions of the Norfolk and Norwich Naturalists Society, 8, 631-670.

Birtles, S. (2003). ‘”A Green Space Beyond Self Interest”: the evolution of common land in Norfolk’, unpublished PhD thesis, University of East Anglia.

Blagrave, J. (1675). The Epitome of the Art of Husbandry. London.

Blith, W. (1649). The English improver improve. London.

Blome, R. (1686). The Gentleman’s Recreation. London.

Blomefield, F. (1805). An Essay towards a Topographical History of the County of Norfolk. London.

Blomley, N. (2007). ‘Making Private Property: Enclosure, Common Right and the Work of Hedges’, Rural History 18 (1), 1–21.

Boardman, E.T. (1939). ‘The development of a Broadland estate at How Hill , Norfolk’, Transactions of the Norfolk and Norwich Naturalists’ Society 15, 14.

Bogart, D. (2005). ‘Did turnpike trusts increase transportation investment in eighteenth-century England?’, The Journal of Economic History 65, 2, 439-468.

Boutcher, W. (1775. A Treatise on Forest-Trees. Edinburgh.

Branch Johnson, W. (1970). The Industrial Archaeology of Hertfordshire. Newton Abbot: David and Charles.

Branch Johnson, W. (ed.) (1973). Memorandoms for … The Diary between 1798 and 1810 of John Carrington. London and Chichester: Phillimore.

Brasier, C.M. (1991). ‘Ophiostoma novo-ulmi sp. nov., causative agent of current Dutch elm disease pandemcs’, Mycopathologia 115, 151-161.

Brasier, C.M. (2008). ‘The biosecurity threat to the UK and global environment from international trade in plants’, Plant Pathology 57, 792–808.

Brasier , C.M.and Gibbs, J.M. (1973). ’Origin of the Dutch elm disease epidemic in Britain’, Nature 242, 607-9.

Briton Brush Co Ltd. (1935). Brochure: Wymondham.

214

Brown, A.E. and Taylor, C.C. (1989). ‘The Origins of Dispersed Settlement: Some Results from Bedfordshire’, Landscape History 11, 61-82.

Brown,N., Daegan, J.G. Inward, Jegar, M. and Denman, S. (2014). ‘A review of Agrilus biguttatus in UK Forests and its relationship with Acute Oak Decline’, Forestry 2014 88 (1), 53-63.

Brown, T. and Foard, G. (1998). ‘The Saxon Landscape: a Regional Perspective’, in P. Everson and T. Williamson (eds) The Archaeology of Landscape. Manchester: Manchester University Press, 67-94.

Brown, J. (1847). The Forester. London.

Burrell, W.H. (1914). ‘Physiography and Plant Distribution’, in W.A.Nicholson (ed.) A Flora of Norfolk. London: West, Newman and Son, 13-33.

Butcher, R.W. (1941). The Land of Britain: Suffolk. London: Geographical Publications.

Cameron, L.G. (1941). The Land of Britain, Part 80: Hertfordshire. London: Geographical Publications.

Cameron, L.G. (1941). The Land of Britain, Part 80: Hertfordshire. London: Geographical Publications.

Cameron, R.A.D. and Pannett, D.J. (1980). ‘Hedgerow Shrubs and Landscape History in the West Midlands’. Arboricultural Journal 4, 147-52.

Campbell, B.M.S. (1981). Commonfields origins: the regional dimension’. In T.Rowley (ed.) The Origins of Open Field Agriculture. Croom Helm, London, 112-129.

Campbell, G. and Robinson, M. (2011). ‘Environment and land use in the valley bottom’, in Harding, J. and Healy, F. (eds) A Neolithic and Bronze Age Landscape in Northamptonshire: Volume 1 – The Raunds Area Project. London: English Heritage, 18-36.

Carlton,J. (2012). Marine Propellers and Propulsion. London: Butterworth-Heinemann.

Carthew, G.A. (1877). A History of the Hundred of Launditch and the Deanery of . Norwich.

Catt, J. and Doyle, J. (2010). ‘The Palaeogene period’, in J. Catt (ed.) Hertfordshire Geology and Landscape. Welwyn Garden City: Hertfordshire Naturalists Society.

Chard, R. (1959). ‘The Fire Plan’. Journal of the Forestry Commission 28, 154-189.

Chatwin, C.P. (1961). British Regional Geology: East Anglia and Adjoining Areas. London: British Geological Survey.

Cheffings, R. and Lawrence, C.M. (2014). Chalara. A summary of the impacts of ash dieback on UK biodiversity, including the potential for long term monitoring and future research on management scenarios. Peterborough: JNCC report No 501.

215

Chester, K.R. (1930). ‘The phytopthora disease of the Culla in America’, Quarterly Journal of Forestry 11, 169-171.

Cinotti, B. (1989). ‘Winter moisture content and frost-crack occurrence in oak trees (Quercus petraea Liebl. and Q. robur L.)’, Annales des Sciences Forestierès 46, 614–16.

Clark, C. (1996). A Brush with Heritage: the history of Hamilton Acorn, Norfolk brushmakers since 1746. Norwich: Centre of East Anglian Studies.

Clarke, W.G. (1918). ‘The natural commons’, Transactions of the Norfolk and Norwich Naturlists’ Society 10, 294-318.

Clarke, W.G. (1925). In Breckland Wilds. Cambridge: Heffer.

Clarkson, L.A. (1974). ‘The English Bark Trade, 1660-1830’, Agricultural History Review 22, 136-52.

Clemenson, H. (1982). English Country Houses and Landed Estates. London: Croom Helm.

Clouston, B. and Stansfield, K. (1979). After the Elm. London: Heinemann.

Coleman, W. S. (1859). Our Woods, Land, Heaths and Hedges. London.

Colley, L. (2009). Britons: forging the nation 1707-1837 . London: Yale University Press.

Collins, E.J.T. (1989). ‘The coppice and underwood trades’, in G.E. Mingay (ed.) The Agrarian History of England and Wales Vol. 6, 1750–1850. Cambridge: Cambridge University Press, 484–501.

Cook, M. (1676). On the Manner of Raising, Orderting and Improving Forest-Trees. London.

Cooper, F. (2006). The Black Poplar: history, ecology and conservation. Oxford: Windgather.

Corbett, W. (1825). The Woodlands. London.

Cossons, N. (1987). The BP Book of Industrial Archaeology. Newton Abbot: David and Charles.

Cowell, B. (2002). ‘The commons preservation society and the campaign for Berkhamsted common, 1866-70’, Rural History 13 (2), 145-162.

Cox, A. (1979). A Survey of Bedfordshire Brickmaking: a history and gazetteer. Bedford: RCHME and Bedforshire County Council.

Cox, J.C. (1914). ‘Forestry’, in W.Page (ed.) Victoria County of the County of Hertfordshire. London: Constable and Co.

Crowther, J. (1992). Descriptions of East Yorkshire: De La Pryme to Head. York: East Yorkshire Local History Society.

216

Curtis, C.E. (1892). The Manifestation of Disease in Forest trees: The causes and remedies. London.

Dallas, P. (2010). ‘Sustainable Environments: common wood-pastures in Norfolk’, Landscape History 31, 1, 23-36.

Daniels, S. (1988). ‘The political iconography of woodland in later eighteenth-century England’, in D. Cosgrove and S.Daniels (eds) The Iconography of Landscape. Cambridge: Cambridge University Press, 51-72.

Dannatt, N. (1996). ‘: its history and development’, in P.Ratcliffe and J.Claridge (eds) Thetford Forest Park: the Ecology of a Pine Forest. Edinburgh: Forestry Commission.

Davies, R. (1964). The Trade and Shipping of Hull 1500-1700. York: East Yorkshire Local History Society.

Davison, A. (1990). The Evolution of Settlement in Three Parishes in South East Norfolk. Dererham: East Anglian Archaeology Vol. 49.

Day,W.R. (1948). ‘The penetration of conifer roots by Fomes annosus’, Quarterly Journal of Forestry 42, 99-101.

Delabere Blaine, P. (1838). An Encyclopaedia of Rural Sports: or a complete account of … Hunting, Shooting, Racing &c . London.

Denman, S., Brady, C., Kirk, S., Cleenwerck, I., Venter, S., Coutinho, T.A. and De Vos, P. (2012). ‘Brenneria goodwinii sp. nov., associated with acute oak decline in the UK’, International Journal of Systematic and Evolutionary Microbiology 62, 2451-2456.

Denman, S., Brown, N., Kirk. S.A., Jeger.M. and Webber. J.F. (2014). ‘A description of the symptoms of acute oak decline in Britain and a comparative review on causes of similar decline diseases of oak in Europe’, Forestry 10, 1 - 17.

Denman, S. and Webber, J.F. (2009) ‘Oak declines - new definitions and new episodes in Britain’, Quarterly Journal of Forestry 103 (4), 285-290.

Donaldson, J. (1794) General View of the Agriculture of the County of Northampton. London.

Dormor, I. (2002). Woodland Management in Two Yorkshire Dales Since the Fifteenth Century. Unpub PhD. Thesis, University of Leeds.

Dowdeswell, W.H. (1987). Hedgerows and Verges. Lonndon: Allen and Unwin.

Ducumb, J.M. (1805). General View of the Agriculture of the County of Hereford. London.

Edelen, G. (ed.) (1994). The Description of England: the Classic Contemporary Account of Tudor Social Life, by William Harrison. London: Folger Shakespeare Library.

217

Eden, F.M. (1797). The state of the poor: or, an history of the labouring classes in England from the Conquest to the present period. London.

Edlin, H.L. (1972) East Anglian Forests. London: HMSO.

Edmonds, E. A., Poole, E. G., and Wilson, V. (1965). The Geology of the Country around Banbury and Edge Hill. London: British Geological Survey.

Ehrlich, J. (1932). ‘The occurance in the United States of Cryptococcus fagi (Baer) Dougl. The insect factor in a menacing disease of beech’, Quarterly Journal of Forestry 13, 73-80.

Ellis, W. (1732). The Practical Farmer. London.

Ellis, W. (1741). The Timber Tree Improv’d: or, the best practical methods of improving different lands with proper timber. London.

Evans, D. (1992). A History of Nature Conservation in Britain. London: Routledge.

Evelyn, J. (1664). Sylva, or A Discourse of Forest-Trees. London.

Eyre, S. R. (1955). ‘The Curving Ploughland Strip and its Historical Implications’, Agricultural History Review 3, 80-94.

Faden, W. (1797). A Topographical Map of the County of Norfolk. London.

Fisher, W.R. (1895). Forest Protection, forming volume 4 of W. Schlich, Manual of Forestry. 4 vols, London.

Fitzgerald, R. and Grenier, J. (1992). A History of the Timber Trade Federation. London: Batsford.Fleming, A. (1998). Swaledale: the valley of the wilNd river. Edinburgh: Edinburgh University Press.

Fitzherbert, J. (1533). The Boke of Husbandry. London.

Flinn, M.W. (1984). The History of the British Coal Industry Vol. 2, 1700-1830. Oxford: Oxford University Press.

Forestry Commission (1936). Bulletin No. 17: The Cultivation of the Cricket Bat Willow. London: Forestry Commission.

Forestry Commission (1955). Report of the Committee on Hedgerow and Farm Timber. London: HMSO.

Forestry Commission (1972). Unpublished hedgerow tees survey in East England Conservancy. Now in private collection.

218

Forestry Commission (2002).National Inventory of Woodland and Trees. Edinburgh: Forestry Commission.

Forsyth, W. (1791). Observation on the Diseases, Defects, and Injuries in all kinds of Fruit and Forest Trees. London.

Fowler, P. (2002). Farming in the First Millennium AD. Cambridge:, Cambridge University Press.

Fryer, V., Murphy, P. and Wiltshire, P. (2005). ‘Vegetational history and early farming’, in T.Ashwin and A.Davison (eds) An Historical Atlas of Norfolk, 3rd edn. Chichester: Phillimore, 10-11.

Funnell, B. (2005). ‘Geological background’, in T.Ashwin and A.Davison (eds) An Historical Atlas of Norfolk, 3rd edn. Chichester: Phillimore, 4-5.

Furst, H. (1893). The Protection of Woodlands Against Dangers Arising from Organic and Inorganic Causes. Edinburgh.

Gardener, H.W. (1967). A Survey of the Agriculture of Hertfordshire. London: Royal Agricultural Society of England.

Gardiner, H.W. (1976). A survey of the agriculture of Hertfordshire. London: Royal Agricultural Society of England County Agricultural Surveys no. 5.

George, M. (1992). The Land Use, Ecology and Conservation of Broadland . Chichester: Packard Publishing.

Gibbs J.N., Brasier C.M. and Webber J.F. (1994). The Biology of Dutch elm disease. Edinburgh: Forestry Commission, Research Information Note. 252.

Gibbs, J., Greig, B., and Rishbeth, J. (1996). ‘Tree diseases of Thetford Forest and their influence on its ecology and management’, in P.Ratcliffe and J.Claridge (eds) Thetford Forest Park: the ecology of a pine forest. Edinburgh: Forestry Commission.

Gledhill, T. (1994). A Woodland History of North Yorkshire. Unpub. PhD thesis, University of Sheffield.

Godwin, H. (1968). ‘Studies in the post-glacial history of British vegetation 15. Organic deposits of Old Buckenham Mere, Norfolk’, New Phytologist 67, 95-107.

Gowans, E.,and Pouncett, J. (2003). Rapid Assessment of Q-Pits, Ecclesall Wood, Sheffield: archaeological survey. Oxford: Archaeological Survey and Evaluation Ltd.

Green, T. (2005). ‘Is there a case for the Celtic Maple or the Scots Plane?’, British Wildlife 16, 184-88. Grigor, J. (1841). The Eastern Arboretum: or a Register of Remarkable Trees, Seats, Gardens &c in the County of Norfolk. London.

219

Gurney, D. (2005). ‘Roman Norfolk (c. AD 43-410)’, in T.Ashwin and A.Davison (eds) An Historical Atlas of Norfolk, 3rd edn. Chichester: Phillimore, 28-9.

Hadfield, M. (1967). Landscape with Trees. London: Country Life.

Haggard, H.R. (1902) Rural England. Two volumes, London: Longmans.

Haggard, L.R. (1946) Norfolk Notebook. London: Faber.

Haggard, L.R & Williamson, H. (1943) Norfolk Life. London: Faber.

Hains, B. A. and Horton, A. (1969). British Regional Geology: central England. London: HMSO.

Hainsworth, D.R. and Walker, C. (1988). The Correspondence of Lord Fitzwilliam of Milton and Francis Guybon his Steward, 1697-1709. Northampton: Northamptonshire Record Society 36.

Hale, T. (1756). A Compleat Body of Husbandry. London.

Hall, D. (1982). Medieval Fields. Aylesbury: Shire.

Hall, D. (1995). The Open Fields of Northamptonshire. Northampton: Northamptonshire Record Society.

Hall, D. (2014). The Open Fields of England. Oxford: Oxford University Press.

Hall, D. (2001). ‘The Woodland Landscapes of Southern Northamptonshire’, Northamptonshire Past and Present 54, 33-46.

Hall, J. (1982). ‘Hedgerows in West Yorkshire: the Hooper method examined’, Yorkshire Archeological Journal 54, 103-109.

Hammersley, G. (1973). ‘The charcoal iron industry and its fuel’, Economic History Review 26 (4), 593–613.

Hanbury, W. (1758). An Essay on Planting. London.

Harris, A. (1961). The Rural Landscape of the East Riding of Yorkshire, 1700-1850: a study in historical geography. Wakefield: S. R. Publishers.

Hartig, R.H. (1894). Textbook of Diseases of Trees. London: MacMillan.

Hartlib, S. (1651). An Enlargement of the Discourse of Husbandry. London.Langley, B. (1728). A Sure Method of Improving Estates. London.

Hatcher, J. (1993). The History of the British Coal Industry. Volume 1. Before 1700: Towards the Age of Coal. Oxford: Oxford University Press.

220

Hatley, V.A. (1980). ‘Locks, lords and coal: a study in eighteenth-century Northamptonshire history’, Northamptonshire Past and Present 6, 207-218.

Hey, D. (1984). ‘Yorkshire and Lancashire’, in J. Thirsk (ed.) The Agrarian History of England and Wales, Volume V.2. Cambridge: Cambridge University Press, 59-88.

Hill, D. (2000). ‘Sulh – the Anglo-Saxon Plough c.1000 AD’, Landscape History 22, 7-19.

Hill, D. and Robertson, P. (1988). The Pheasant: ecology, management and conservation. Oxford: BSP Professional.

Hindle, S. (1998). ‘Persuasion and protest in the Caddington common enclosure dispute 1635–1639’, Past & Present 158, 37-78.

Hodder, K.H., Buckland, P.C., Kirby, K.J. & Bullock, J.M. (2009). ’Can the pre-Neolithic provide suitable models for re-wilding the landscape in Britain?’, British Wildlife, 20, 5 (special supplement), 4-14.

Hodge, C. A. H., Burton, R. G. O. Burton, Corbett, W.M. , Evans, R. and Searle, R.S. (1984). Soils and their Use in Eastern England. Harpenden: Soil Survey of England and Wales.

Holderness, B.A. (1989). ‘Prices, Productivity and Output', in G.E. Mingay (ed.) The Agrarian History of England and Wales Volume VI. Cambridge: Cambridge University Press, 84-189.

Hollingworth, S. E. and Taylor, J. H. (1946). ‘An Outline of the Geology of the Kettering District’ Proceedings of the Geologists’ Association 57, 204-33.

Hopkins, H. (1985). The Long Affray. London: Secker & Warburg.

Hunter, J. (1999). The Essex Landscape: a study of its form and history. Chelmsford: Essex Record Office.

Jackson, C. (ed.)( 1870). The Diary of Abraham de la Pryme, the Yorkshire Antiquary. Durham: Publications of the Surtees Society 54, Andrews and Company.

Jackson, G. (1975). The Trade and Shipping of Eighteenth Century Hull. York: East Yorkshire Local History Society.

James, T. (2009). Flora of Hertfordshire. Welwyn Garden City: Hertfordshire Natural History Society.

Jeffery, T. (1771). The County of York Survey'd in MDCCLXVII, VIII, IX and MDCCLXX. London.

Johns, C.A. (1899). The Forest Trees of Britain. London.

Johnson, W. (1978). ‘Hedges: a review of some early literature’, Local Historian 13, 195 – 204.

221

Jones, M. (1993). ‘South Yorkshire’s Ancient Woodland: the historical evidence’, in I. Rotherham and P. Beswick (eds) Ancient Woodlands, their Archaeology and Ecology: a coincidence of interest. Sheffield: Landscape Conservation Forum, 26–48.

Jones, M. (1997). ‘Woodland Management on the Duke of Norfolk’s Sheffield Estate in the Early Eighteenth Century’, in M.Jones (ed.) Aspects of Sheffield: discovering local history. Trowbridge: Wharncliffe, 48–69.

Jones, M. (1999). ‘Deer in South Yorkshire: an historical perspective’, Journal of Practical Ecology and Conservation (Special Publication No.1), 11–26.

Jones, M. (2012). Gapping, Raddling and Snagging: An illustrated glossary of terms associated with woodland management in South Yorkshire 1086- c.1900. Sheffield: Wildtrack Publishing.

Jones, M. (2005). ‘Ancient Woodland Destruction, Survival and Restoration: a South Yorkshire perspective’, Landscape Archaeology and Ecology 5, 46–58.

Jones, W.S. (1910). ‘A disease of the alder’, Quarterly Journal of Forestry 4, 221-224.

Kent, N. ( 1796). General View of the Agriculture of the County of Norfolk. London.

Kerridge, E. (1967). The Agricultural Revolution. London: Routledge.

Kerridge, E. (1973). The Farmers of Old England. London: Allen and Unwin.

Kerridge, E. (1992). The Common Fields of England. Manchester: Manchester University Press.

King, H., and Harris, A. (eds.) (1963). A Survey of the Manor of Settrington. Leeds: Yorkshire Archaeological Society Record Series Vol. 126.

Kirby, J. and Baker, A (2013). ‘The dynamics of pre-Neolithic European landscapes and their relevance to modern conservation’, in I.D.Rotherham ((ed.) Trees, Forested Landscapes and Grazing Animals. London: Routledges, 87-98.

Knutchbull, W.T. (1911). ‘The importation of foreign plants… and a reply by the editor’, Quarterly Journal of Forestry 5, 174-175.

Langton, J. (1979). Geographical Change and Industrial Revolution. Coa Mmining in south-west Lancashire 1590-1799. Cambridge: Cambridge University Press.

Latham, B. (1957). Timber. Its Development and Distribution: an historical survey. London: George Harrap and Co.

Lennard, R. (1959). Rural England, 1086-1135: a study of social and agrarian conditions. Oxford: the Clarendon Press.

222

Lennon, B. (2009). ‘Estimating the age of groups of trees in historic landscapes’, Arboricultural Journal 32, 167-88.

Linnell, J. E. (1932). Old Oak: The Story of a Forest Village. London: Constable and Co.

Loudon J.C. (1838). Arboretum et Fruticetum Britannicum; or, the trees and shrubs of Britain. London.

Lucas,J (ed.) (1892). Kalm’s Account of his Visit to England on his Way to America in 1748. London.

McDonagh, B.A.K. (2009). ’Subverting the Ground: Private Property and Public Protest in the Sixteenth-Century Yorkshire Wolds’, Agricultural History Review 57 (2), 191–206.

McDonnell, J. (1992). ‘Pressures on Yorkshire woodland in the later Middle Ages’, Northern History 28 (1), 110–125.

MacNair, A. and Williamson, T. (2010). William Faden and the Norfolk Eighteenth-Century Landscape. Oxford: Windgather.

MacNair, A.D.M, Rowe, A. and Williamson, T. (2015). Dury and Andrews’ Map of Hertfordhsire: society and landscape in the eighteenth century. Oxford: Windgather Press.

Marcais, B. and Desprez-Loustau, M. (2014). ‘European oak powdery mildew: impact on trees, effects of environmental factors, and potential effects of climate change’, Annals of Forest Science 71, 633-642.

Marshall, W. (1787). The Rural Economy of Norfolk, 2 Volumes, London.

Marshall, W. (1788). The Rural Economy of Yorkshire. London.

Marshall, W. (1796). Planting and Rural Ornament. London.

Martin, E. and Satchell, M. (2008). Wheare Most Inclosures Be. East Anglian Fields, History, Morphology and Management. Ipswich: East Anglian Archaeology vol. 124.

Mather, R.A. and Savile, P.S. (1994). ‘The commercial impact of oak shake in Great Britain’, Forestry 67(2), 119-131.

Maw, P.T. (1910). ‘The importation of foreign grown seedlings’, Quarterly Journal of Forestry 4, 267268.

Maw, P.T. (1912). The Practice of Forestry. London: Fisher Unwin.

Mawer, A. and Stenton, F. M. (1938). The Place-Names of Hertfordshire. Cambridge: Cambridge University Press.

Meager, L. (1697). The Mystery of Husbandry: or, arable pasture and wood-land improved. London.

223

Miche, C.Y. (1888). The Practice of Forestry. London.

Miller, P. (1731). The Gardener’s Dictionary. London.

Mitchell, B.R. and Deane, P. (1962). Abstract of British Historical Statistics. Cambridge: Cambridge University Press.

Monteath, R. (1824). The Forester's Guide and Profitable Planter. London.

Moraal, L.G. and Hilszczanski, J. (2000). ‘The oak buprestid beetle, Agrilus biguttatus (F.) (Col. Buprestidae), a recent factor in oak decline in Europe’, Journal of Pest Science 73 (5), 134-138.

Moreton, C. and Rutledge, P. (eds) (1997). ‘Skayman’s Book. 1516-1518’, in C.Noble. C.Moreton and P.Rutledge (eds) Farming and Gardening in Late Medieval Norfolk. Norwich: Norfolk Record Society, 95-155.

Mortimer, J. (1707). The Whole Art of Husbandry: or, the way of managing and improving of land. London.

Morton, J. (1712) The Natural History of Northamptonshire. London.

Mosby, J.E.G. (1938). The Land of Britain: Norfolk. London: Geographical Publications.

Mougou, A., Dutech, C. and Desprez-Loustau, M-L. (2009). ‘New insights into the identity and origin of the causal agent of oak powdery mildew in Europe’, Forest Pathology 38 (4), 275–287.

Muir, R. (2006). ‘Pollards in Nidderdale: a landscape history’, Rural History: economy, society and culture 11,1, 95-111.

Muir, R. (2006). Ancient Trees, Living Landscapes. Stroud: The History Press.

Muir, R. and Muir, N. (1997). Hedgerows: their history and wildlife. London: Michael Joseph.

Munby, L. (1977). The Hertfordshire Landscape. London: Hodder and Stoughton.

Munsche, P.B. (1981). Gentlemen and Poachers: the English Game Laws 1671-1831. Cambridge: Cambridge University Press.

Murphy, P. (1993). ‘Pre-Norman vegetational change and woodland clearance’, in Wade-Martins (ed.) An Historical Atlas of Norfolk. Norwich: Norfolk Museums Service, 20-21.

Nicol, W. (1799). The Practical Planter, or, a Treatise on Forest Planting. London.

Nicol,W. (1820). The Planters Kalendar. Edinburgh.

Nisbet, J. (1894). John Brown, The Forester: enlarged and revised in 2 vols. London.

224

Norden, J. (1608). The Surveyor’s Dialogue. London.

Norfolk County Council (1976). Farmland Tree Survey of Norfolk. Norwich: Norfolk County Council.

Nourse, T. (1699). Campania Felix: Or, a Discourse of the Benefits and Improvements of Husbandry. London.

Orchard, N. (2007). ‘Further Changes in Rockingham Forest Woodlands’. Unpublished MSc Dissertation. Birkbeck College, London.

Overton, M. (1996). Agricultural Revolution in England: the transformation of the agrarian economy, 1500-1850. Cambridge: Cambridge University Press.

Page, M. (2003). ‘The Extent of Whittlewood Forest and the Impact of Disafforestation in the Later Middle Ages’, Northamptonshire Past and Present 56, 22-34.

Page, W. (ed.) (1902). The Victoria History of the County of Hertford, Vol. 1. London: Archibald Constable.

Page, W. (ed.) (1908). The Victoria History of the County of Hertford, Vol. 2. London: Archibald Constable.

Page, W. (ed.) (1912). The Victoria History of the County of Hertford, Vol. 3. London: Archibald Constable.

Palmer, M. and Neaverson, P. (1994). Industry in the Landscape 1700-1900. London: Routledge.

Partida, T. (2014. ‘Drawing the Lines: a GIS study of enclosure and landscape in Northamptonshire’. Unpubished PhD thesis, University of Huddersfield.

Pearce, W. (1794). General View of the Agriculture in Berkshire. London.

Peglar, S., Fritz, S. and Birks, H. (1989). ‘Vegetation and land use history at Diss, Norfolk’, Journal of Ecology 77, 203-22.

Perren, R. (1995). Agriculture in Depression 1870-1940. Cambridge: Cambridge University Press.

Perry, P.J. (1974). British Farming in the Great Depression. Newton Abbot: David and Charles.

Peterken, G.F. (1976). ‘Long-term changes in the woodlands of Rockingham Forest and other areas’, Journal of Ecology 64 (1), 123-146.

Petit, S. and Watkins, C. (2003). ‘Pollarding trees; changing attitudes to a traditional land management practise in Britain, 1600-1900’, Rural History 14, 157-76.

Pettit, P. A. (1968). The Royal Forests of Northamptonshire: A Study of their Economy 1558-1714. Northampton: Northamptonshire Record Society.

225

Pitt, W. (1813). General View of the County of Northampton. London.

Pollard, E. Hooper, M. D. and Moore, N.W. (1974). Hedges. London: Collins.

Pontey, W. (1805). The Forest Pruner, or Timber-owner’s Assistant. London.

Pratt, E.K. (1927). ‘The East Anglian Timber Willow’, Journal of the Forestry Commission 88, 80-86.

Price, A. (2015). Shake in Oak: an evidence review. Edinburgh: Forestry Commission Research Report.

Prince, H. (1987). ‘The changing rural landscape’, in G.Mingay (ed.) The Cambridge Agrarian History of England and Wales, Vol.6. Cambridge: Cambridge University Press, 7-83.

Rackham, O. (1980). Ancient Woodland. London: Edward Arnold).

Rackham, O. (1986). The History of the Countryside. London: Dent.

Rackham, O. (1986b). ‘The Ancient Woods of Norfolk’, Transactions of the Norfolk and Norwich Naturalists Society 27, 161-7.

Rackham, O. (2006). Woodlands. London: Collins.

Read, H.J. (2008). ‘Pollards and pollarding in Europe’, British Wildlife 19, 250-259.

Reed, M. (1981). ‘Pre-Parliamentary Enclosure in the , 1550 to 1750, and its Impact on the Landscape’, Landscape History 3, 60-68.

Repton, H. (1816). Fragments on the History and Practice of Landscape Gardening. London.

Rishbeth, J. (1963). ‘Stump protection against Fomes annosus: inoculation with Peniophora gigantea, Annals of Applied Biology 52, 63-77.

Roberts, B.K. and Wrathmell, S. (2002). Region and Place; a Study of English Rural Settlement. London: English Heritage.

Roden, D. (1973). ‘Field Systems of the Chiltern Hills and their Environs’. In A.H.R.Baker and R.A.Butlin (eds) Studies of Field Systems in the British Isles. Cambridge: Cambridge University Press, 325-374.

Rogerson, A. (1995). ‘Fransham: an Archaeological and Historical Study of a Parish on the Norfolk Boulder Clay’. Norwich: unpublished PhD thesis, University of East Anglia.

Rogerson, A. (2005). ‘Middle Saxon Norfolk’, in T.Ashwin and A.Davison (eds) An Historical Atlas of Norfolk, 3rd edn. Chichester: Philllimore,32-33.

226

Ross, J.M (1935). ‘Study of Pine Shoot Moth Damage’. Journal of the Forestry Commission 14, 56-64.

Rotherham, I.D. (2011). Peat and Peat Cutting. Princes Risborough: Shire.

Rotherham, I.D. and Lambert, R.A. (eds) (2011). Invasive and Introduced Plants and Animals: human perceptions, attitudes and approaches to management . London: Earthscan.

Rowe, A. (2009) Medieval Parks of Hertfordshire. Hatfield: University of Hertfordshire Press.

Rowe, A. (2015). ‘Pollards: living archaeology’, in K.Lockyear (ed.) Archaeology in Hertfordshire: recent research. Hatfield: University of Hertfordshire Press, 302-324.

Rowe, A. and Williamson, T. (2013). Hertfordshire: a landscape history. Hatfield: University of Hertfordshire Press.

Rowley, T. (2006). The English Landscape in the Twentieth Century. London: Hambledon Continuum.

Ruggles, T. (1786). ‘Picturesque farming’, Annals of Agriculture 6, 175-84.

Ryden, L., Pawel, M, and Anderson, M. (2003). Environmental Science: understanding, protecting and managing the environment of the Baltic Sea region. Uppsala: Baltic University Press.Rackham, O. (2006). Woodlands. London: Collins.

Ryle, G. (1969). Forest Service: the first forty-five years of the Forestry Commission in Great Britain. Newton Abbot: David and Charles.

Salisbury, E.J. (1918). ‘The oak-hornbeam woods of Hertfordshire’, Journal of Ecology 6, 11-52.

Saunders, C. (1977). ‘A sixteenth century tannery in St Albans’,Hertfordshire’s Past 3, 9–12.

Seebohm, F. (1883). The English Village Community. London.

Selby, J. (1842). A History of British Forest Trees. London.

Sheial, J. (1971. Rabbits and their History. Newton Abbot: David and Charles.

Sheail, J. (2002). An Environmental History of Twentieth-Century Britain. London: Palgrave.

Simpson, J. (1900). The New Forestry: or, the Continental system adapted to British woodlands and game preservation. Sheffield: Pawson & Brailsford.

Skipper, K. and Williamson, T. (1997). Thetford Forest: making a landscape, 1922-1997. Norwich: Centre of East Anglian Studies.

Slater, G. (1907). The English Peasantry and the Enclosure of Common Fields. London: Constable.

227

Smith, A. H., Baker, G. M. and Kenny, R. W. (1982). The Papers of Nathaniel Bacon of , Vol. 2. Norwich: Centre of East Anglian Studies.

Smith, J. (1670). England’s Improvement Reviv’d. London.

Spencer, J. and Kirby, K. (1992). ‘An inventory of ancient woodland for England and Wales’, Biological Conservation 62, 77-93.

Spray, M. (1981). 'Holly as fodder in England', Agricultural History Review 29 (2), 97-110. Steuart, H. (1828). The Planters Guide. Edinburgh.

Stoertz, C. (1997) Ancient Landscapes of the Yorkshire Wolds: Aerial Photographic Transciption and Analysis. Swindon: English Heritage.

Stone, A. and Williamson, T. (2013). ‘Pseudo-Ancient Woodland’ and the Ancient Woodland Inventory, Landscapes 14, 141-54.

Switzer, S. (1718) Ichnographica Rustica. London.

Straw, N.A. and Williams, D.T. (2013). ‘Impact of the leaf miner Cameraria ohridella(Lepidoptera: Gracillariidae) and bleeding canker disease on horse-chestnut: direct effects and interaction’, Agricultural and Forest Entomology 15 (3), 321-333.

Tabbush, P. and Beaton, A. (1998). ‘Hybrid poplars: present status and potential in Britain’, Forestry 71 (4), 355-64.

Takamatsu S., Braun, U., Limkaisang, S., Kom-Un, S., Sato, Y. and Cunnington, J.H. (2007). ’Phylogeny and taxonomy of the oak powdery mildew Erysiphe alphitoides sensu lato’, Mycological Research 111, 809-826.

Taylor, C. (1975) Fields in the English Landscape. London: Dent.

Tennyson, J. (1939). Suffolk Scene: a book of description and adventure. London: Blackie.

Thirsk , J. (1987). England's Agricultural Regions and Agrarian History 1500-1750. London: MacMillan

Thomas, F.M., Blank, R., and Hartmann G. (2002) ‘Abiotic and biotic factors and their interactions as causes of oak decline in Central Europe’, Forest Pathology 32, 277-307.

Thomas, K. (1983). Man and the Natural World: changing attitudes in England 1500-1800. London: Allen Lane.

Thompson, F.M.L.(1963). English Landed Society in the Nineteenth Century. London: Routledge and Kegan Paul.

Thompson, F.M.L. (1968). ‘The second Agricultural Revolution’, Economic History Review 21, 62-77.

228

Tilbury, C. and Evans, H. (2003). ‘Exotic pest alert. Horse Chestnut leaf miner’. Edinburgh: Forestry Commission.

Toseland, A. (2013). Estate Letters from the Time of the 2nd Duke of Montagu, 1709-32. Northampton: Northamtonshire Record Society Vo. 46.

Toulmin-Smith, L. (ed.) (1907) The itinerary of John Leland in or about the years 1535-1543. London: George Bell.

Tuke, J. (1800). General View of the Agriculture of the North Riding of Yorkshire. London.

Turner, M.E. (1982). Volume 190: Home Office Acreage Returns HO67. List and Analysis, three parts. London: List and Index Society.

Tusser, T. (1573). Five Hundred Good Points of Husbandrie. London.

Vera, F. (2002). Grazing Ecology and Forest History. Wallingford: Cabi.

Wade, E. (1755). A Proposal for Improving and Adorning the Island of Great Britain; for the maintenance of our navy and shipping. London.

Wade Martins, S. and Williamson, T. (eds) (1995). The Farming Journal of Randall Burroughes of Wyomondham, 1794-99. Norwich: Norfolk Records Society.

Wade Martins, s. and Williamson, T. (1999). Roots of Change: farming and the landscape in East Anglia, c.1700-187. Exeter: Agricultural History Society.

Wade Martins, S. and Williamson, T. (2008). The Countryside of East Anglia: changing landscapes, 1870-1950. Woodbridge: Boydell.

Walsingham, Lord and Payne-Gallwey , R. (1983). Shooting: field and covert. London.

Wake, J. and Webster, D.C. (1971). The Letters of Daniel Eaton to the Third Earl of Cardigan, 1725- 1732. Northampton: Northamptonshire Record Society.

Ward, H. Marshall (1892). The Oak: a popular introduction to forest-botany. London.

Ward, J.R. (1967). East Yorkshire Landed Estates in the Nineteenth Century. York: East Yorkshire Local History Society.

Warde, P. and Williamson, T. (2014). Fuel supply and agriculture in post-medieval England. Agricultural History Review, 62, 61-82.

Wargo, P.M. (1996). ‘Consequences of environmental stress on oak: predisposition to pathogens’, Annals of Forest Science 53, 339-368.

Warner, P. (1987). Greens, Commons and Clayland Colonization. Leicester: Leicester University Press.

229

Wass, J.G. (1956). ‘Fomes Annosus in East Anglian Pine Sample Plots’. Journal of the Forestry Commission 9, 75.

Waterson, M. (1994). The National Trust: The First Hundred Years. London: BBC Publications.

Westmacott, R. and Worthington, T. (1974). New Agricultural Landscapes. London: Countryside Commission.

Woodward, D. (ed.) (1985). Descriptions of East Yorkshire: From Leland to Defoe. York: East Yorkshire Local History Society

Wheeler, J. (1747). The Modern Druid. London.

Whynbrow, G.H. (1934). The History of Berkhamsted Common. London: Commons, Open Spaces and Footpaths Preservation Society.

Whyte, N. (2009). Inhabiting the Landscape: Place, Custom and Memory, 1500 – 1800. Oxford: Windgather.

Williams-Ellis, C. (1928). England and the Octopus. London:G. Bles.

Williamson, T.M. (1998). The Archaeology of the Landscape Park: garden design in Norfolk, England, c.1680-1840. Oxford: British Archaeological Reports, British Series, 268.

Williamson, T. (2002). The Transformation of Rural England: farming and the landscape 1700-1870. Exeter: Exeter University Press.

Williamson, T. (2007). ‘Archaeological perspectives on landed estates: research agendas’, in J.Finch and K.Giles (eds) Estate landscapes: design, improvement and power in the post-medieval landscape. Woodbridge: Boydell, 1-18.

Williamson, T. (2013). Environment, Society and andscappe in Early Medieval England: time and topography. Woodbridge: Boydell.

Williamson, T. (2013b). An Environmental History of Wildlife in England, 1550-19050. London: Bloomsbury.

Williamson, T., Liddiard, R., and Partida, T. (3013) Champion. The making and Unmaking of the English Midland Landscape. Liverpool: Liverpool University Press.

Willmott, A. (1980). ‘The Woody Species of Hedge with Special Reference to Age in Church Broughton Parish, Derbyshire’, Journal of Ecology 68, 269 – 286.

Wiltshire, P. and Murphy, P. (1999). ‘Current knowledge of the Iron Age environment and agrarian economy of Norfolk and adjacent areas’, in J.Davies and T.Williamson (eds), Land of the Iceni: the Iron Age in northern East Anglia. Norwich: Centre of East Anglian Studies, 132-161.

230

Witney, K.P. (1998). ‘The Woodland Economy of Kent, 1066-1348’, Agricultural History Review 38, 20-39.

Wooldridge, S.N. (1945). The Land of Britain, Part 51: Yorkshire, North Riding. London: Geographical Publications.

Worlidge, J. (1660). A Compleat System of Husbandry and Gardening. London.

Wrathmell, S. and Roberst, B. (2000). An Atlas of Rural Settlement in England. London: English Heritage.

Yelling, J.A. 1977. Common Field and Enclosure in England 1450-1850. London: Archon.

Young, A. (1797). General View of the Agriculture of the County of Suffolk. London.

Young, A. (1804). General View of the Agriculture of Hertfordshire. London.

Young, A. (1813). General View of the Agriculture of Norfolk. London.

231

Principal Primary Sources.

Hertfordshire Archives and Local History

HALS DE/Bb/E27. Valuation of standing timber and notes of the amounts cut. 1784-1806 HALS

A/2832. Map of the Manor of Great Barwick 1774.

HALS A/2833. Map of the Manor of Standon 1779.

HALS 43754. Land at Puckeridge, Standon, 1778.

HALS DE/Ha/B2112. Surveys of timber on the lands of Ralph Skynner’s charity at Kelshall. 1724-1774.

HALS B479. Map of lands on the Broxbourne Mills estate, late eighteenth century.

HALS DE/B944/E1-2. Survey of Sheeps Farm, London Colney. 1852.

HALS DP/117 25/7. Three fields called Church Lands, Watford. 1754.

HALS DE/Cr/55 Valuations of timber, Hemel Hempstead, Redbourne, Albury. 1838-1839.

HALS DE/X905/P1. Map of Flaunden and Sarratt.

HALS DE/Ha/P1. Survey of manor and parish of Ickleford and adjoining parishes. 1771.

HALS DE/H/P16 Plan of Stockers and Pearces Farms. 1807.

HALS DE/P/P4. Map of the estate of William Cowper, 1703-4.

HALS 51124. Papers concerning the estate of Anne Catharine Shapherd. 1769-1782.

HALS DE/Am/E3 Leases and related papers. Melbourn Farm. 1694-1838.

HALS DE/B983/E1. Survey of the estate of Edward Beeton Long, esq. 1787.

HALS DE/Db/E26 accounts for cutting and sale of timber. 1731-1795.

HALS 15594. Plan of Herons Manor Farm. Wheathampstead. 1768.

HALS DE/P/T262. Sale of Wood. Hertingfordbury. 1637.

HALS DE/LW/E33. Account of timber sold from Marshall and Butlers. 1539.

HALS DE/12/E114/1-10. Accounts of John Merritt, Woodward. 1739-1767.

HALS 27252. Tarrants Farm. Lower estate, 1734.

HALS 1985. Map of Berkhampstead Heath. 1638.

232

HALS 71014. Parish map of Broxbourne. 1829.

HALS DSA4/25/2. Tithe map of Broxbourne. 1829.

HALS DE/H/P11. Plan of Much Hadham. 1851.

HALS DE/B11768/P2 Map of Colliers End. 1794.

HALS DE/X216/B2. Valuations of timber on lands allotted and exchanged. 1802-1803.

HALS DE/Bn(Add) B13. Timber sales, Tring. 1880.

HALS DE/BN(Add) B19. Timber valuation, Rickmansworth. 1859-1861.

HALS 80226. Court Rolls. Broxbourne and Hoddesdon.

HALS DE/X713/P1. Map of the whole parish of Much Hadham. 1833.

HALS 81750. Map of Whempstead. 1808.

Norfolk

NRO MC 1777/1 Map of the manor of Chanonz (in Tibenham, parts marked as Aslacton and Bunwell), part of the possessions of John Buxton, esq., by John Harrison, surveyor. 1640.

NRO WIS 138,166X3, Map and Terrier of estate in Beeston, property of William Clarke Woodbine of Swaffham, gent. 1761.

NRO DCN 59/21/1 Notes re timber on Hindolveston estate, 1731.

NRO BR 90/35/22 Valuation of timber, fences, etc, Skeyton and Burgh. 1816.

NRO 21128, 179X4. Survey of the Manor of West Lexham. 1575.

NRO HIL 3/34/1-27,879X3. Map of Whissonsett, 1724.

NRO EVL 348, 459X1. Particular and valuation of timber, Wendling. 1777.

NRO HNR 149/3. Timber account book, 1813-1824.

NRO MC 257/23/5, 638X8 Timber survey, Palgrave, 1792.

NRO BUL 11/283, 617X2. Survey of timber on Erasmus Earle’s estates, 1722.

NRO BL/CS 1/20/3/1-5 Particulars of several estates of Miss Williams, with separate valuation of timber, 1775.

NRO WLS XXVI/4, 414X6. Parliamentary survey of the manor of Ickburgh, 1652.

NRO BR 90/34/16 Bundle of timber accounts.

233

NRO MS 13751/ 40E3 letters and timber at Munford. 1805-1806.

NRO WLS LXX/22/1-41, 481X7. Papers re exchange of Merton glebes for part of Woodfield Farm, Thompson. 1839-1976.

NRO HARE 5500,223X1. Valuation of timber. Stow Hall Estate. 1815.

NRO 16112, 32A6. Valuation of timber at Hunworth, Stody and Brinningham. 1739.

NRO 20757A Valuation by Nathaniel Kent of estates belonging to Thomas William Coke. 1789.

NRO MC 3/252, 468X4. Valuation by Thomas Browne of the estate of the Countess of Buckingham. 1756.

NRO MS 13167, 40A7. Copy of Court Roll for the Manor of Little Fransham. 1494-1597.

NRO 13279, 40C2. Article of agreement for a lease. 1603.

NRO 13159, 40A6. Survey of the Manor of Little Fransham, Kirkhams and Wylcockes. 1605.

NRO MS 13326, 40C2. Indenture-lease for 14 years. 1636.

NRO FEL 1076. Map of easterly part of Shotesham St Mary, St Martins and All Saints. 1650.

NRO BR 139/1 Map of the parish of Saxlingham and lands in Shropham. 1806.

NRO PD 3/108 Map of Morley St Botolph with St Peter.1629.

NRO PD 3/109 map of Morley St Botolph and St Peter, with Inclosure award and map. 1815.

NRO 4087 Plan of the tree felling at Foxley park. 1815.

NRO BER 336. Survey of Hockering, Morton, Ringland, Easton, Weston and Costessey estates. 1828.

NRO 336 291/7. Hockering Wood Indenture.

NRO 20888. Estate papers re Sporle and Palgrave.

NRO KNY 571, 372X3 Copy of Bailiff Edward Harrison’s account of timber, underwood, etc. Ashwellthorpe Estate. 1724.

NRO MEA 7/4. Articles of agreement to let Attleborough Wood. 1801.

NRO EVL 650,464X2. Honeypot Wood, lease for 10 years. 1835.

NRO MR 211, 241X6. Draft indenture of bargain and sale 1612.

NRO BRA 301/1 Manor of Fersfield. Lease agreements 1682-1712.

NRO 13253, 40 B4. Indenture, lease for 21 years. 1530.

234

NRO WLS LXIX/15, 481X4. Memorandum book of William de Grey. 1675.

NRO WLS IV/6, 407X6. Interogations on behalf of the Queen versus Francis Woodhouse relating to Wayland Wood and Barnyards Wood, Bunwell. 1594.

NRO MC 600/4-6, 780X9. Bargain and sale of 75 timber trees at Redenhall. 1736/7.

NRO 59/30/12. Request by Edmund Rolfe lessee of Chapter estate, to plant wood and underwood. Nineteenth century.

NRO HOW 757/9. Estate matters, Castle Rising. 1785-1799.

NRO MR 211, 241X6. Indenture of Bargain and Sale. Sir Hammon Le Strange, Lord of the Manor of Gressenhall, 1612.

NRO HNR 135/3/18. Woodmans Cash Account.

NRO MC 3/252, 468X4. Valuation of the estate of the Countess of Buckingham. 1756.

NRO MC 22/11. Map of New Buckenham following dispute over commons. 1597.

NRO MS4521. Maps of the parishes of Cawston and Haveringland, 1600.

NRO BL71. Copy of 1588 map of Weston, Norfolk from King’s Lynn to Flitcham.

NRO BRA 2524/6. Map of the Manor of Appleton, 1595.

NRO Hayes and Storr72. Map of Gressenhall. 1624.

NRO 149/3 Timber Account Book. 1813-1824.

NRO MR 61, 241X1. Gressenhall Surveys. Late sixteenth century.

NRO BR 90/14/2. Statement of Claims for Inclosure. 1807.

NRO WIS 138, 166X3 map and terrier of estate in Beeston, 1761.

NRO MS 9316, 7B9. Timber valuation and extents of lands, Whissonsett, 1762.

NRO BL/BG 5/3/4 Three timber notebooks 1813-1821.

NRO MC 687/29-31, 813X3. Catalogue of agreement for sale by John Lock to Humfrey. 1813-1821.

NRO MC3/47 Blickling Hall building accounts. 1619-1622.

NRO AYL 900. Salle, case and opinion re timber felling by copyhold tenant. 1805.

NRO MC 688/11, 813X5. Particulars of timber trees growing on estates of William Harbord in Hempstead, Plumstead and Erpingham to be sold in 1799.

235

NRO NRS 11126, 25E5. Papers re timber sales from estates at Blickling, Langley, Chedgrave and Intwood. Late seventeenth century.

NRO DN/EST 52/22. Account book for timber in Curples Wood. 1771.

NRO HARE 5499, 223X1. Valuation of timber. Stow Hall. 1811-1815.

NRO BUL 2/3, 604X7. Lease for Southhaw Wood. 1612.

NRO MC 77/1, 521X7. Sale particular.

NRO BUL 118/11, 615X9 Timber catalogue, Heydon.

NRO BL/BG 5/3/18. Timber notes. 1916.

NRO HNR 736/9. Hickling Wood timber account. 1872-1894.

NRO WLS XVIII/7/1 Correspondence and papers of Thomas the fifth Earl. 1831-1870.

NRO HNR 112/19. Timber Book. 1847-1850.

NRO BUL 11/436 Paper in Chancery Suit. 1813-1840.

NRO HNR 112/19. Timber Book.1847-1850.

NRO 111/48, 581X5. Sale catalogues of trees on Wallington Hall Estate.

NRO RQG 178, 489X1. Sale Particulars re timber sale 1903.

NRO BL/BG 5/3/22. Sale catalogue of timber in West Norfolk. 1926.

NRO RQG 226, 489X4. Timber sale catalogue, 1906.

NRO BUL 11/118/1. Timber catalogues. Heydon. 1911-1928.

NRO WIS 138. Map and terrier of estate in Beeston.

NRO HNR 149/2 Timber account book. 1824-1837.

NRO MEA 3/549, 658X8. Timber Account Books 1811-1823.

NRO MC 547/81-104, 768X8. Agreements, survey etc, re timber.

NRO HNR 64/1. Foresty and timber papers. Hevingham woods. 1939-41.

NRO MC 166/238. Notes re Ditchingham Hall Estates. 1873-1900.

NRO HNR 465/3/1 Bradenham Hall Estate timber book. 1892-1917.

236

NRO MC 547/81-104. Agreements, survey etc re timber.

NRO Y/C 36/15/18. Miscellaneous Correspondence of Richard Ferrier.

Northamptonshire

NHRO LT 64. Ravensthorpe 1377

NHRO A/089 Lease of land and property in Crick. 1582.

NHRO 2920/001 Memorandum for planting ash trees. 1740.

NHRO 3133 Map of Upper Boddington 1758.

NHRO E(S) Box X1071. Enclosure award and other documents. Ecton.

NHRO 1349. Map of Weekley.

NHRO 2895 Map of Quinton.

NHRO 6-p/504 Map of Easton-on-the-Hill.

NHRO 4447. Map of Wallaston.

NHRO Inclosure Vol Ap9. Enclosure map of Ecton. 1759.

NHRO ZA 0906 Valuation of hedges, trees at Irthlingborough at the time of inclosure. 1808.

NHRO ZA 0898 Valuation of hedge, trees at Finedon at enclosure, 1806.

NHRO X1657. Wilby 1764.

NHRO 350P/90. Wellingborough 1735.

NHRO C(A) Box 104 4 Papers of the Cartwright family of Aynho. 1656.

NHRO FH 272Map of Lower Radstone. 1586.

NHRO ZB 1837. Sale catalogue of Freehold estates of late William Duke of Powis. 1758.

NHRO YZ 2183. Survey of John Darkers estates in Gayton, Tiffield, Kislingbury, Milton, Litchborough and Upper Heyford. 1791.

NHRO HOLT 655. Helidon 1630.

NHRO G4167.Whittlewood. 1853.

237

NHRO G3909. Geddington.

NHRO Survey and paparer, Brooke of Oakley 318/1 NHRO

Surveys and papers, Brooke of Oakley 313/21 NHRO Bru

OV 229. Geddington. 1659.

NHRO B(G) 1. Wilby survey 1764.

NHRO ZB 887 Survey of estate at Staverton 1835.

NHRO ASL 392. Nineteenth century deeds, Welton.

NHRO B(HH) 461 valuation of timber at Teaton. 1812.

NHRO X1693. Bundle 21. G3903.

NHRO Map 5704. Map of Brockhall 1672.

Boughton House archives: map of Broughton 1728.

Boughton House archives: map of Newton le Willows 1717.

NHRO Map 1388. Map of Cranford St Andrew 1748.

NHRO FH 272. Map of Deenethorpe 1585.

NHRO Bru Map 8. Map of Deene and Deenethorpe 1739.

British Library, London: AddMss 78136A. Map of Elkington 1775.

British Library, London: AddMss 78143. Map of Nobottle 1715.

NHRO Map 853. Map of Fawsley 1741.

NHRO Inclosure Plan 19. Map of Hannington 1802.

NHRO T154. Map of Loddington 1842.

NHRO Map 2221. Map of Papley 1632.

NHRO 1349. Map of Quinton 1723.

Yorkshire

In contrast to the other counties studied in this research project, Yorkshire’s archives are dispersed across a number of separate public repositories.

238

Beverley (Bev)

Bev DDX3/15 Map of Howsham 1705.

Bev DDX 73/51: survey of timber at Ellerton, 1768.

Bev DD CROM 6/14 timber valuation, Sprotborough, 1719 and 1720.

Bev DDBD 19/19 timber valuation, Ellerker and Brough, 1860.

Bev DDCC 135/71 timber valuation, Burton Constable, 1799.

Bev DDCS 4/5 timber valuation, Melbourne Hall, 1880.

Bev DDHB 57/5, timber valuation, Hessle and West Ella, 1775.

Bev DDSE 1/11, sale, Darrington, 1887.

Bev DDX 73/51, timber valuation, Ellerton, 1768.

Barnsley (Barn)

Barn EM 291: Timber Valuations, Thorne, 1793-1813.

Barn EM 791 timber valuation, Thorne, 1818.

Barn EM 812 timber valuation, Dodworth etc, 1829.

Bradford (Brad)

Brad 23D98/3/5: Valuation of wood, Blacketts Estates, 1750.

Brad 15D74/17/2/35, timber valuation, various parishes, 1810.

Brad 15D74/17/2/55, timber valuation, South Duffield.

Brad 23D98/3/5, valuation, Esholt etc, 1750.

Brad BAR/3/3/3, valuation, Faweather, 1812.

Brad SpSt 5/4/1/20, sale, Horsforth, 1875.

Brad Tong MSS 4f/5, valuation, Tong, 1829.

Brad Tong MSS 4f/11, valuation, Tong, 1831.

Brad Tong MSS 4f/13, valuation, Tong, 1832 .

Brad Tong MSS 4f/14, valuation, Tong, 1832.

Brad Tong MSS 4f/4, valuation, Tong, 1832.

239

Brad Tong MSS 4f/15, valuation, Tong, 1834.

Brad Tong MSS 4f16, valuation, Tong, 1837.

Calderdale (Cald)

Cald SH1/YT/1772/Jan 15: Valuation, Nr. Halifax, 1771.

Cald FW:50/27-28, sale, Shelf, 1876.

Cald FW:50/51, valuation, Warley Town, 1882.

Cald FW:50/51, valuation, Ovenden, 1882.

Cald FW:50/51, valuation, , 1882.

Cald KMA:1208, valuation, Clifton, Hartshead, 1790.

Cald KMA:1208, valuation, Briestwistle, Middlestown, Denby, 1763.

Cald KMA:1208, valuation, Middlestown, 1797.

Cald KMA:1208, valuation, Clifton, Hartshead, 1766.

Cald KMA:1208, valuation, , 1766.

Cald KMA:1208, valuation, Mirfield, 1765.

Cald SH1/NH/1835, valuation, Halifax, 1835.

Cald STN:159, sale, , 1811.

Cald SU:52, valuation, Walshaw, 1852.

Cald KMA 1208: Timber Surveys and Valuations, Armytage Estate, eighteenth and nineteenth century.

Doncaster (Don)

Don DD/BW/E11/7 A map of Warmsworth Lordship 1726.

Hull

Hull U DDWA/x1/4/3: valuation of timber, Nunburnholme, 1774.

Hull C DBHT/5/75, valuation, Welton, 1898.

Hull C DBHT/5/75, valuation, Brandsby, 1907.

Hull U DDBH/24/16, sale, Naburn Wood, 1867.

Hull U DDBH/24/16, valuation, Naburn Wood, 1868.

240

Hull U DDBH/26/7, valuation, Bell Hall Estate, 1857. Hull

U DDBH/76/13/65, valuation, Deighton, 1874.

Hull U DDCA/36/8, valuation, Carlton and Camblesforth, 1753.

Hull U DDFA/13/78, valuation, Riccall, 1814.

Hull U DDFA/13/78, valuation,, Osgodby, 1816.

Hull U DDFA/13/78, valuation, Skipwith, 1820.

Hull U DDFA/X6/1/4, valuation, Acaster, 1898.

Hull U DDHB/3/56, valuation, Deighton, 1804.

Hull U DDSQ3/14/5, valuation, Scamston, 1825.

Hull U DDSQ3/17/3, valuation, Lowthorpe and Thornholme, nineteenth century.

Hull U DDSQ3/17/3, valuation, Lowthorpe, 1841.

Hull U DDSQ3/17/3, valuation, Scampston, 1844.

Hull U DDSQ3/17/3, valuation, Scampston, 1847.

Hull U DDWA/x1/4/3, valuation, Nunburnholme, 1774.

Kirklees , Huddersfield (Kirk)

Kirk DD/RE/C/33/14A, valuation, , 1847.

Kirk DD/S/II/192, valuation, Wadsworth, 1900.

Kirk KC174/5/3, valuation, , 1763.

Kirk KC694, sale, , 1855.

Leeds (Lds)

Lds WYL68/63 Map of the Manor of Healaugh and Catterton 1718.

Lds WYL50 13/2234: Timber sold, Newby Hall, 1829-32.

Lds WYL160/220/157: Agreement regarding woodland, Arthington Hall, 1708.

Lds WYL230/2828: Purchase of trees, Ripley Castle, 1738.

Lds WYL500/292: Agreement regarding brushwood, Stansfield of Esholt Estate, 1724.

Lds WYL500/939: Valuation of Milford Estate, eighteenth century.

241

Lds WYL639/339: Timber Sales (various locations), 1835-6.

Lds WYL189/B/61: Valuation of Timber, Temple Newsam Manor, 1820-34.

Lds WYL68/4: Healaugh Survey, 1757. Lds WYL68/33: Healaugh Estate Papers, eighteenth century.

Lds WYL100/EA/13/38, sale, Temple Newsam, 1632.

Lds WYL100/EA/13/72, sale, Temple Newsam, 1795.

Lds WYL115/XB/5/3, sale, Parlington, 1870.

Lds WYL115/XB/5/4, sale, Parlington, 1840.

Lds WYL115/XC/Box71a/25, valuation, Parlington, nineteenth century.

Lds WYL160/220/179, valuation, Arthington Hall, 1835.

Lds WYL189/B/61, valuation, Temple Newsam, 1823.

Lds WYL230/2826, sale, Ripley Castle (Ingilby), 1705.

Lds WYL230/2840, sale, North Deighton, 1801.

Lds WYL500/939, valuation, Milford Estate, eighteenth century.

Lds WYL5013/2234, sale, Newby Hall, 1832.

Lds WYL639/339, sale, Shipley, 1835.

Lds WYL639/339, sale, , 1836.

Lds WYL68/33, valuation, Healaugh, 1751.

Northallerton (Nrth):

Nrth ZQG IV/16 Map of Brandsby 1746.

Nrth ZDV Map of Newburgh 1605.

Nrth ZAZ (M) 1 Map of Walburn c.1700-1743.

Nrth ZBO (M) 1/2 A map of lands belonging to His Grace the Duke of Bolton at Leyburn, Harmby & Bellerby [1730].

Nrth ZBO (M) 1/5 A map of the Manor of Downholme 1738.

Nrth ZMI Map of Skeeby 1779.

Nrth ZEW M 13 Pockley and Beadlam Map 1 1785.

Nrth ZEW M 14a The Township of Pockley Map 2 1785.

Nrth ZIQ Map of Sinderby and Pickhill 1778. 242

Nrth ZNS A plan of Harlsey 1762.

Nrth ZDS M 2/12 A map of the Lordship of Birdforth 1770. Nrth ZDS M 2/1 A survey of the Manor of Thormanby 1785.

Nrth ZEW IV 1/6: Copy (1647) Survey of the Manor of Helmsley, 1642.

Nrth ZEW IV 6/4/6: Wood accounts, Duncombe Estate, 1786-91.

Nrth ZEW III 7/2: Manorial Court papers, Pockley, Beadlam, Eastmoor, eighteenth century.

Nrth ZEW IV 7/25: Timber Sales, Duncombe Estate, 1724, 1917, 1944.

Nrth ZDV V 30: Coxwold Manor Court Summaries, 1635-55.

Nrth ZDV: Newburgh timber account, 1776.

Nrth ZDV: Newburgh Timber, 1781.

Nrth ZQG IV 1/1: Survey of Brandsby, 1746.

Nrth ZQG IV 2/8: Timber Valuation, Brandsby, 1733.

Nrth ZBO [4404]: Letter, Thomas Sadler to Lord Bolton, 1807.

Nrth ZDS IV 14/6: Wood valuation and letter, Danby 1789.

Nrth ZCG IV 5/2/10: Howsham Woods Management, 1803.

Nrth ZBO IX 1/12/5, valuation, Wensley, 1784.

Nrth ZBO IX 1/24, valuation, Downholme Woods, 1808.

Nrth ZBO IX 1/24, valuation, Well Gill, 1808.

Nrth ZBO IX 14/8, valuation, Bolton Hall, 1793.

Nrth ZDS IV 14/6, valuation, Danby Park Wood, 1789.

Nrth ZDS IV 14/6, valuation, Liverton, 1759.

Nrth ZDS IV 14/6, valuation, Danby Park, 1762.

Nrth ZDS IV 14/6, valuation, Liverton, 1762.

Nrth ZDS IV 14/6, valuation, Danby Park, 1789.

Nrth ZDV, valuation, 1774.

Nrth ZDV, valuation, Newburgh, 1776.

Nrth ZDV, valuation, Newburgh, 1781.

Nrth ZDV, sale, Newburgh, 1793.

Nrth ZDV, sale, Newburgh, 1789.

243

Nrth ZDV, sale, Newburgh, 1786.

Nrth ZDV, sale, Lund, 1794 . Nrth ZEW IV 1/6, valuation, Helmsley, 1642.

Nrth ZQG IV 2/8, valuation, Brandsby, 1733.

Sheffield (Shf)

Shf ACM 2/279: Account of cordwood cut, Arundel Castle Estate, 1683-4.

Shf ACM 2/280: Accounts for herbage, Arundel Castle Estate, 1692-3.

Shf ACM/MAPS/Shef/169: Maps of Sheffield Woods, eighteenth and nineteenth century.

Shf WWM A1273: Marquis of Rockingham 'A Scheme for a Regular Fall of Wood for 21 years…', 1749.

Shf FB/CP/25/11: Enquiry regarding the legal definition of coppice and timber, Mid nineteenth century.

Shf FB/CP/25/1-2: Remarks on woods and plantations, early nineteenth century.

Shf MD3520: Notes on tree cultivation, early nineteenth century.

Parliamentary Papers (online at: http://parlipapers.proquest.com)

The Eleventh Report Of The Commissioners Appointed To Enquire Into The State and Condition Of The Woods, Forests, and Land Revenues Of The Crown, And To Sell Or Alienate Fee Farm and other Unimproveable Rents (6th February 1792).

Board of Agriculture Returns, 1896.

Board of Agriculture Statistics, 1910.

Board of Agriculture Statistics, 1914.

Forestry Commission 4th Annual Report, 1924.

Forestry Commission 15th Annual Report, 1935.

Forestry Commission Annual Report, 1945.

244