Forestry Commission

Bulletin 115

in Britain: A Review

Cyril Hart

Forestry Commission ARCHIVE

FORESTRY COMMISSION BULLETIN 115

Alternative Silvicultural Systems to Clear Cutting in Britain: A Review

Cyril Hart

LONDON : HMSO © Crown copyright 1995 Applications for reproduction should be made to HMSO Copyright Unit, Duke Street, Norwich NR3 1PD ISBN 0 11 710334 9 Hart, C.; 1995. Alternative Silvicultural Systems to Clear Cutting in Britain: A Review. Bulletin 115, HMSO, London, xiv + 101pp. FDC 22: 221.1: (410): (4) KEYWORDS: Amenity, Broadleaves, Conifers, Conservation, Establishment, Forestry, Management

Cyril Hart Dr Cyril Hart OBE is a Chartered Surveyor and Chartered Forester and is the Queen’s Senior Verderer for the Forest of Dean. He is author of several books on Britain’s forests and trees of which the best known is his authoritative Practical forestry for the agent and surveyor.

Professor Julian Evans and Mr Gary Kerr of the Forestry Commission’s Research Division contributed to this important review through editorial involvement and text revision.

Front cover: Weasenham New Wood, Please address enquiries Norfolk: mixed conifer selection system; about this publication to: in the picture, Major Richard Coke Research Publications Officer DSO MC, the present owner. The Forestry Commission (P e t e r G o o d w i n ) Research Division Alice Holt Lodge, Wrecclesham Back cover: Dr Cyril Hart OBE Farnham, Surrey GU10 4LH (D a v id G r e e n ) Acknowledgements

This Bulletin was written with the benefit of Professor J.D. Matthews’s updat­ ing (1989) of Professor R.S. Troup’s two editions of Silvicultural systems (1928, 1952), the latter edition by Dr E.W. Jones. Eminently rewarding too has been co-operation received from David Paterson, recently retired from the Forestry Commission, whose publications and erudite comments, based on a wide expe­ rience of practice and research in silviculture, have been invaluable. Likewise acknowledged are texts of Professor M.L. Anderson, and of Forestry Commission staff, in particular: Julian Evans, John Everard OBE, Gary Kerr, Mark Yorke and Adrian Whiteman. Additional assistance relating to silvicul­ tural systems was received directly or indirectly from Simon Bell, John Blyth, Paul Bond, Graham Booth, Richard Broadhurst, Robert Budden, Vernon Daykin, Christine Cahalan, Murray Carter, Lord Cawdor, Major Richard Coke, Wilfred Fox, Bob Furness, Michael Gane, Ted Garfitt, Peter Garthwaite OBE, David Goss, Sandy Greig, Michael Harley OBE, Esmond Harris, Anthony Hart, Rodney Helliwell, Paul Hill-Tout, Peter Hughes, Phillip Hutt MBE, Robert Jordan, Talis Kalnars, Felix Karthaus, Steve Lee, Arthur Lloyd, Roy Lorrain-Smith, David Lowe, John MacKenzie, Douglas Malcolm OBE, John McHardy, Robert McIntosh, Philippe Morgan, Andrew Neustein, John Newcomb, Trefor Owen, Colin Price, Chris Quine, Michael Reade DSC, John Robinson, Martin Rogers, Irvine Ross, Roddy Ross, Robin Satow, Derek Stickler, Ken Stott, Charles Taylor, David Taylor, Keith Wallis, Guy Watt, Michael Wemyss, Helen Whitney Mclver, John Workman OBE, and Richard Worrell. References to many of them are made in the text. For useful com­ ments and editorial assistance throughout I thank Jenny Claridge. Literature of earlier years, particularly for the purpose of describing the few British examples of alternative silvicultural systems, indicated some con­ fusion in their classification. Professor Matthews’s admirable text (1989) effected clarification, and with his permission and that of Clarendon Press, Oxford his classification and descriptions of systems have been followed. The courtesy and assistance thus extended has enabled this Bulletin to benefit from a sound base on which to record past and present and suggest potential relevant practices in Britain. Nonetheless, readers will benefit by study of Professor Matthews’s text as well as those of Evans (1988) and Kerr and Evans (1993). Information gained by membership of the Continuous Cover Forestry Group, inaugurated in 1991, and from the Edinburgh University database on examples of irregular silvicultural systems in Britain is warmly acknowl­ edged. Profound thanks are recorded to the owners and/or their foresters of many private sector woodland stands visited; also to managers of Forest

iii Enterprise’s relevant forest stands, where my visits have been welcomed and proved rewarding to me. I not only ended up with many friends but with a very large number of obligations. This Bulletin is written for landowners and foresters who wish to convert all or part of a wood or forest from pure, regular, uniform, even-aged stands to mixed, irregular, uneven-aged stands - in particular for those who desire to use silvicultural systems alternative to that of extensive clear cutting, with a view to achieving diverse structure leading to biological diversity and semi-perma- nent or continuous forest cover. The information provided can be adopted to suit individual owners’ or foresters’ own constraints and objectives. Other aims are to draw attention to what has been achieved in Britain, and the opportunities for further attempts, and to highlight the need for further research, trials and demonstration areas. A few of the plates were taken by myself; others were taken by my son, Anthony, and by friends in the forestry world; the remainder were selected from the Forestry Commission’s photographic library. Permission of John Matthews CBE (Emeritus Professor, University of Aberdeen), David Paterson and Chris Quine to include certain tables (amended) is gratefully acknowl­ edged: the sources are recorded in appropriate places in the text.

Cyril Hart 1995 Foreword

Compared with much of Europe, British forestry has largely been dominated by just one silvicultural system in recent decades, viz. clearfelling and replant­ ing. There are sound silvicultural reasons for this, in particular the windiness of our upland climate. Nevertheless, today, there is increasing interest in extending rotation lengths and in diversifying forest structure which, along with greater attention to lowland woods, create more opportunities for alter­ native silvicultural systems. This Bulletin draws together experience in Britain with many different types of silviculture which have been tried out or practised during the present century. I hope foresters, woodland owners, and all those interested in the management of Britain’s woodlands and forests will find the assemblage of experience of considerable use. The Forestry Commission is deeply indebted to Dr Cyril Hart OBE for bringing together, analysing and synthesising experience with these alterna­ tive silvicultural systems from all parts of the country. This task has not been easy and the Forestry Commission is most grateful that the job has been undertaken by one so eminent. I hope that readers will find the text itself informative and interesting and the plates instructive of the different kinds of silviculture discussed. Each detailed account of individual woods and the silvicultural system practised is headlined in colour. It is these descriptions, brought together for the first time, which are at the heart of this Bulletin and, I believe, will make it such an invaluable reference work in the future.

T.R. Cutler CBE Director General Forestry Commission June 1995 Contents

Page Acknowledgements iii Foreword v List of plates viii Summary x Resume xi Zusammenfassung xiii Resumen xiv

1 Diverse woods and forests: the roleof silvicultural systems 1 Introduction 1 Alternative silvicultural systems to clear cutting 2 Preparatory work in conversion 5

2 Shelterwood systems 6 Introduction 6 Use in Britain: broadleaves 8 Uniform shelterwood 8 Group shelterwood 10 Use in Britain: conifers 12 Uniform and group shelterwood 12 Conclusions on shelterwood systems 14

3 The group selection system 17 Introduction 17 Use in Britain: broadleaves 17 Use in Britain: conifers and mixed stands 25 Conclusions on the group selection system 35

4 The single tree selection system 39 Introduction 39 Application on the Continent 39 Use in Britain 41 Conclusions on the single tree selection system 44

5 Other silvicultural systems 46 Two-storied high forest 46 High forest with reserves 48 Coppice and coppice with standards 49 Restructuring extensive conifer forest by patch clear cutting 50

VI Restructuring beechwoods by (mainly) group regeneration 52 Conclusions on other silvicultural systems 53 6 The economics of irregular forestry 54 Introduction 54 The single tree selection system 55 The shelterwood uniform and group systems 56 The economic justification of uneven-aged and continuous forest cover 56 Conclusions on the economics of irregular forestry 57 7 Influence of environmental factors on alternative silvicultural systems and conversion 58 Effect of wind 58 Effect of damaging mammals 60 Deer 60 Grey squirrels 61 Effects on landscape 61 Impact on nature conservation 62 Effect on recreational potential 63 Conclusions on the influence of environmental factors 63

8 Conclusions: resources and benefits, opportunities and recommendations 64 General 64 Experience in Britain 65 Opportunities in Britain 65 Conversion 68 Summary of benefits of alternative systems 69 Environmental benefits 69 Silvicultural benefits 69 Management economics of alternative systems 70 Postscript 71 References 72 Glossary 1 Definition of terms 78 Glossary 2 Scientific names and authorities of the main English names used in the text 80 Appendix 1 Classification of silvicultural systems 81 Appendix 2 Summary of factors to consider in choosing silvicultural systems 82 Appendix 3 Need for research, trials and demonstration areas 84 Appendix 4 Suggestions for trials and demonstration areas 88 Appendix 5 History and personalities 89

vii List o f plates

1 Natural regeneration of oak and beech at Churchill East, Forest of Dean 2 Beech natural regeneration at Savernake Forest, Wiltshire, showing older groups becoming diverse in canopy structure and age 3 Pedunculate oak natural regeneration at High Standing Hill Wood, Windsor Forest 4 Natural regeneration of oak, improved by fencing to exclude mammals, at Wyre Forest 5 Douglas fir and Norway spruce group shelterwood system at The Hermitage Wood, near Dunkeld 6 Natural regeneration of grand fir in a uniform shelterwood system at High Standing Hill Wood, Windsor Forest 7 Scots pine natural regeneration under a uniform shelterwood system at the Crown Estate, Swinley 8 Beech natural regeneration under 110-year-old trees at Workmans Wood, Ebworth, Gloucestershire 9 European larch at Cawdor Big Wood, near Nairn 10 Oak trees in fog at Cawdor Big Wood, near Nairn 11 The Glentress Trial, Glentress Forest, near Peebles: conversion of conifer stands from uniform even-aged clear cutting system to uneven-aged group selection 12 Conversion of conifers to group selection system at Corrour Forest, Inverness-shire 13 Old Scots pine at Faskally Wood, near Pitlochry, showing conversion towards group selection in conifers 14 Group or random regeneration of indigenous Scots pine at Benevean Wood, Glen Affric 15 Douglas fir at Diosgydd, Gwynedd, illustrating conversion of conifer plan­ tations to group selection 16 Sitka spruce and beech natural regeneration under mature larch at Nant BH, Gwynedd 17 Shelterwood-cum-group selection system at Longleat, Wiltshire, showing natural regeneration under Douglas fir 18 Prolific natural regeneration of Western hemlock in a single tree selec­ tion system at Kyloe Wood, Northumberland 19 Kielder Forest: restructuring for increasingly multiple purpose objectives 20 Hamsterley Forest: smaller scale restructuring 21 Group regeneration of beech in the Chilterns 22 A mixed storey formed by regenerating beech in the Chilterns 23 Restructuring of beechwoods in the Chilterns by group regeneration under a stand of c. 150-year-old beech 24 Professor Mark Loudon Anderson

IX Alternative Silvicultural Systems to Clear Cutting in Britain: A Review

Summary

The objective of this Bulletin is to summarise British experience of alternative silvicultural systems to clear cutting. Management of woodlands for diverse objectives such as recreation, landscape, timber production and conservation is becoming increasingly important. Irregular forestry systems have been pro­ moted as one way of maximising the potential benefits woodlands have to offer. This Bulletin explores this assumption and highlights the lessons which have been learnt in managing irregular forestry, and in converting high forest managed by clear cutting to other systems. The examples described have been classified into: shelterwood, group selec­ tion, single tree selection and other silvicultural systems. In total 44 examples of managing forest areas using alternative silvicultural systems, in upland and lowland Britain, are described; these include the well-known examples of Glentress, Ebworth, Weasenham and Dartington. Brief consideration is also given to the economics of irregular forestry and the effects of environmental factors on transformation to alternative silvicultural systems. The Bulletin emphasises four key considerations influencing the use of irregular forestry: 1. The use of irregular silvicultural systems is possible on a wide range of sites in upland and lowland Britain. 2. The key to success for irregular forestry in Britain is continuity of man­ agement. The forester must possess a sound knowledge of silviculture, have an intimate acquaintance with the stands in question, and be prepared to be innovative, patient and opportunistic in the application of, or conversion to, the chosen silvicultural system. 3. The costs of management of irregular forestry can be high compared with the clear cutting system, hence it is not favoured by conventional econom­ ic analysis based on financial yield. Such analyses invariably favour sys­ tems of low management cost. 4. The potential of irregular forestry to produce non-market benefits can be high compared with the clear cutting system. More research into the eval­ uation of such benefits is required so that irregular forestry can be com­ pared objectively with other silvicultural systems. The Bulletin concludes that there are many opportunities to pursue irregular forestry practices more widely throughout Britain. Regimes de sylviculture utilises en alternative a la coupe rase en Grande-Bretagne: une revue

Resume

L’objectif de ce bulletin est de resumer l’experience effectuee en Grande- Bretagne sur les regimes de sylviculture offrant une alternative a la coupe rase. La gestion des bois a des fins diverses telles que la detente, le paysage, la production de bois et la conservation devient de plus en plus importante. Les regimes de foresterie irreguliere ont ete encourages en tant que moyen de maximiser les avantages que les bois ont le potentiel d’offrir. Ce bulletin etudie cette hypothese et met en relief les enseignements tires de la gestion de la sylviculture irreguliere et de la conversion a d’autres regimes des hautes futaies precedemment gerees a l’aide de la coupe rase. Les examples decrits ont ete classes de la maniere suivante: abri (coupes progressives), jardinage par bouquets, jardinage et autres regimes de sylvi­ culture. Au total, 44 exemples de gestion de la foret a l’aide de regimes de sylviculture utilises comme alternative dans les hautes et basses terres de Grande-Bretagne se trouvent decrits. Parmi ceux-ci figurent les exemples bien connus de Glentress, Ebworth, Weasenham et Dartington. La rentabilite de la foresterie irreguliere est aussi brievement abordee de meme que les effets qu’ont les facteurs ecologiques sur la transformation en faveur des regimes de sylviculture utilises comme alternative. Le bulletin met l’accent sur quatre considerations cles influenfant le recours a la foresterie irreguliere: 1. L’utilisation de regimes de sylviculture irreguliere est possible dans une grande variete de sites se trouvant dans les hautes et basses terres de Grande-Bretagne. 2. La cle du succes en matiere de foresterie irreguliere en Grande-Bretagne se trouve dans la continuity de la gestion. Quand il s’agit d’operer la conver­ sion au regime de sylviculture choisi ou d’appliquer ce dernier, le forestier doit posseder une bonne connaissance de la sylviculture, etre tres au courant des positions en question et etre pret a innover, a se montrer patient et a saisir les occasions se presentant. 3. Les couts de gestion de la foresterie irreguliere peuvent etre eleves par comparaison a ceux du regime utilisant la coupe rase. De ce fait, l’analyse economique conventionnelle basee sur le rendement financier ne lui est pas favorable. De telles analyses favorisent invariablement les systemes ayant des couts de gestion peu eleves. 4. Le potentiel montre par la foresterie irreguliere lorsqu’il s’agit de produire des benefices n’ayant rien a voir avec le marche peut etre eleve par com-

XI paraison avec le regime utilisant la coupe rase. II s’avere necessaire de proceder a une evaluation plus approfondie de tels benefices afin de com­ parer objectivement la foresterie irreguliere aux autres regimes de sylvi­ culture. Le Bulletin conclut qu’il existe de nombreuses occasions d’accroitre la mise en pratique de la foresterie irreguliere dans toute la Grande-Bretagne. Forstwirtschaftliche Alternativen zum Kahlschlag in Britannien: Eine Ubersicht

Zusammenfassung

Der Zweck dieses Bulletins ist die Darlegung de britischen Erfahrungen mit forstwirtschaftlichen Altemativen zum Kahlschlag. Es wird immer wichtiger, Forstwirtschaft so zu betreiben, dap diverse Ziele wie etwa Erhohlung, Landschaftsgestaltung, Holzproduktion und Naturschutz erreicht werden konnen. Unregelmapige Forstwirtschaftssysteme werden als eine Art zur Maximierung der moglichen Nutzen, die Walder bitten, vorgeschlagen. Diese Bulletin untersucht diese Ansicht und hebt die Erfahrungen hervor, die durch unregelmapige Forstwirtschaft und die Umstellung von Hochwaldbewirtschaftung dureh Kahlschlag auf andere Systeme, gewonnen wurden. Die Beispiele die hier beschrieben werden, sind wie folgt unterteilt: Schutzwald, Gruppenauswahl, Einzelbaumauswahl oder andere Systeme. Insgesamt 44 Hoch- und Niederland - Forstgebiete, beschrieben die nach alternativen Systemen bewirtschaftet werden. Darunter sind unter anderem die bekannten Beispiele von Glentress, Ebworth, Weasenham und Dartington. Die Wirtschaftlichkeit unregelmapiger Forstwirtschaft wird kurz behandelt, sowie die Wirkung von Umweltfaktoren auf die Umstellung auf alternative Systeme. Dieses Bulletin unterstreicht die vier Hauptfaktoren, die den Gebrauch von unregelmapiger Forstwirtschaft beeinflussen: 1. Der Gebrauch von unregelmaPigen Forstwirtschaftssystemen ist bei einer groPen Anzahl von Standorten im Hoch- und Niederland Britanniens moglich. 2. Der Schliissel zum Erfolg von unregelmapiger Forstwirtschaft liegt in der Stetigkeit der Pflege. Der Forster muP ein gutes Kenntnis der Forstkultur haben, die betroffenen Stande genau kennen und bereit sein, innovativ, geduldig und opportunistisch in der Anwendung, oder Umstellung, des gewahlten forstwirtschaftlichen Systems, vorzugehen. 3. Die Kosten der unregelmaPigen Forstwirtschaft konnen, im Vergleich zu Kahlschlag, hoch sein, dadurch wird es bei normaler finanzieller Analyse der Ertrage nicht vorgezogen. Solche Analysen ziehen unweigerlich Systeme mit niedrigen Pflegekosten vor. 4. Unregelmapige Forstwirtschaft kann, im Vergleich zum Kahlschlagsystem, ein groPes Potential an nichtwirtschaftlichen Nutzen bieten. Es sind mehr Untersuchungen in die Bewertung dieser anderen Nutzen notig, so dap unregelmapige Forstwirtschaft objektiv mit anderen Systemen verglichen werden kann. Dieses Bulletin zieht den Schlup, daP es viele Gelegenheiten gibt, unregelmapige Forstwirtschaft in groPerem Umfang zu benutzen liberall Gropbrittanien.

xiii Sistemas Silviculturales Como Alternatives a la Corta a Hecho en Gran Bret ana: Una Re vista

Resumen

Este Boletln tiene por objeto resumir la experiencia britanica con sistemas sil­ viculturales como altemativas a la corta a hecho. La ordenacion de bosques para objetos diversos como recreo, paisaje, production maderable, y conservation llega a ser cada vez mas importante. Se fomentan sistemas forestales irregu- lares como un metodo de maximar los beneficios potenciales que ofrecen los bosques. Este Boletln trata de esta suposicion, y destaca las conclusiones, apren- didas en la ordenacion de dasonoima irregular, y en la conversion de un regimen en monte alto con corta a hecho a otros sistemas. Se clasifican los ejemplos descritos como sigue: cortas por aclareos sucesivos; entresaca por bosquetes; selection (entresaca) individual; y otros sistemas silvi­ culturales. En suma, se describe 44 ejemplos del tratamiento de bosques por sis­ temas silviculturales altemativos, en las tierras altas y en las tierras bajas de Gran Bretana; estos ejemplos incluyen los bosques conocidos de Glentress, Ebworth, Weasenham y Dartington. Se tratan brevemente tambien los aspectos economicos de dasonomia irregular, y los efectos de factores environmentales sobre la transformation en sistemas silviculturales altemativos. Este Boletln hace resaltar cuatro consideraciones principales que afectan el uso de dasonomia irregular: 1. El uso de sistemas silviculturales irregulares es posible en una gama ancha de sitios en las tierras altas y bajas en Gran Bretana. 2. La llave al exito para la dasonomia irregular en Gran Bretana es la con- tinuidad del tratamiento. El ingeniero forestal debe ser experto en la silvi- cultura, debe conocer Intimamente las masas particulares, y debe ser innovador, paciente y oportunista en la aplicacion del sistema silvicultural elegido, o en la conversion del sistema. 3. Los costes de la ordenacion de dasonomia irregular pueden ser altos en comparacion con la corta a hecho, y por esto no se favorece por un analisis economico convencionial a base del rendimiento en dinero. Tales analises siempre favorecen los sistemas con costes bajos de ordenacion. 4. El potencial de dasonomia irregular para producir beneficios no-mercados puede ser alto en comparacion con el sistema corta a hecho. Se necesita mas investigaciones en la valuation de tales beneficios, para hacer una compara­ cion objetiva de la dasonomia irregular con otros sistemas silviculturales. Este Boletln concluye que hay muchas oportunidades para practicar tratamientos silviculturales irregulares en escala mas grande a traves de todo la Gran Bretana.

xiv Chapter 1 Diverse woods and forests: the role of silvicultural systems

Introduction tems are ecological means of achieving manage­ ment objectives. They specify the method of In Europe many silvicultural systems have been regenerating new stands, the form of stand pro­ used for centuries; and in recent years the organ­ duced, and their spatial arrangement over the isation Pro Silva has given increased attention forest. This planned arrangement takes care of to extension of uneven-aged, irregular systems an essential group of requirements, namely the including ‘continuous cover forest’ and so-called efficient harvesting of the produce, the protec­ ‘near to nature forestry. In Britain, during this tion of the wood or forest against damage by century, most silviculture has been under the wind, fire and animals, and reaction to attacks of extensive clear cutting system, supplemented by insect pests and diseases. Where a forest is part the coppice and coppice with standards systems. of a water catchment, the arrangement of stands However, in recent decades, the clear cutting can sometimes be used to assist in regulating system, despite its many valuable attributes, water supplies. has received some disapproval because of sud­ In the great majority of older woods and den changes on landscape, amenity, wildlife con­ forests in Britain timber production has been servation and recreation. Continuous forest and remains of prime importance but experience cover and ‘naturalness’ are increasingly thought has shown that the sustained yield of timber can of as desirable for multi-purpose objectives be integrated with aesthetic, recreational and where locations, sites and conditions are suit­ environmental benefits when a range of age able. classes and size classes exist. This is because, in These changes are reflected in the present the traditional clear cutting and restocking silvi­ policy of Government on woods and forests in cultural system, each stage of development of a Britain which emphasises that foresters should stand has certain distinct characters. When the integrate the production of timber with the trees are young, there is plenty of open space desired aesthetic, recreational and environmen­ and abundant habitat and food for birds and ani­ tal benefits; also that they should recognise their mals. But when the trees form canopy and enter protective functions in relation to soil, water and the thicket stage they dominate the site and sup­ air quality, and manage to protect and enhance press or exclude ground flora and restrict the them. Part of the process of meeting these diversity of wildlife. Then, as the trees gain in requirements is to accelerate the diversity of stature, the structure of the stand becomes more young woods and forests and develop their struc­ open; as the five crowns move up above the ture to achieve harmony with the landscape and ground and the processes of thinning and har­ gain environmental benefits. vesting trees marked for removal admit light, All these topics fall within the province of sil­ the habitat is improved for a greater variety of vicultural systems which result in the produc­ wildlife on the ground, in the crowns and around tion of stands of distinctive form; and the indi­ the edges. The ideal for recreational and envi­ vidual stands together make up whole woods ronmental benefits is to create open space in and and forests of distinctive form. Silvicultural sys­ around the stands of trees, and to arrange newly

1 regenerated stands close to those approaching 2. A stand with only one canopy of foliage is uni­ maturity. These benefits follow from forests that form. If there are two distinct canopies, the are close to the stage of sustained yield of timber; term used is two-storied high forest, but this in other words, the forests have a balanced grow­ structure is not permanent, although it often ing stock that can produce timber as well as the persists for an appreciable time. Coppice with other benefits already mentioned, in perpetuity. standards has an overwood of mixed ages and A medium sized wood or small forest of (say) an even-aged underwood of coppice; this is 150 to 360 hectares (ha) in extent, managed on permanent provided there is periodic recruit­ the more recently introduced patch clear cutting ment to the standards and the coppice is not system (see Chapter 5) with a rotation of 50 to 60 stored. If there are more than two canopies of years, may comprise 50 to 60 stands, each of 3 to foliage, the term used is selection forest or 6 ha. If the boundaries of individual stands or of sometimes continuous canopy forest (although groups of stands have been designed to fit the the latter is a general description and not a landscape, the edges of extraction routes and definition). other open spaces are treated to favour 3. When the ages of the trees in a stand are broadleaved trees, shrubs and interesting within a few years of each other, the term ground flora, and young stands adjoin older used is regular or even-aged; when the ages ones, there will be considerable diversity in differ by more than 15 to 20 years, the stand appearance, structure and habitat. If two or is irregular or uneven-aged. more species are used, the diversity will increase A classification of silvicultural systems (follow­ further. In larger forests, use of the clear cutting ing Matthews, 1989) is given in Appendix 1. A system will give the impression of less variety, brief history of attempts at alternative systems to but this does not mean that diversity is absent: clear cutting in Britain, and an introduction to it is just more dispersed and on a larger scale. some of the earlier practitioners, are given in This is a problem particularly of large forests in Appendix 5. the uplands, and examples of restructuring and alternative silvicultural systems for the uplands are described in Chapters 2-5. Alternative silvicultural systems Lucas (1991) states that diversity is ‘the to clear cutting degree and number of differences in a composi­ tion’; and that ‘an important principle of design Several alternative silvicultural systems can be is to achieve sufficient diversity to create inter­ used in suitable locations in place of clear cut­ est without causing loss of unity’. The compo­ ting, and these are distinguished by the length of nents of diversity in forestry include species the period of regeneration, the proportion of the composition, canopy structure, size and shape of previous stand that is retained as seedbearers or stands, the use of open space (roads, streams, upper canopy, and the preferred scale of opera­ pools and glades), the spatial arrangement of tions. Because of the familiarity of clear cutting, individual stands, and the provision of habitat it will be convenient to make comparisons with for wildlife. However, a diverse structure is not it to provide a reference point. permanent: it is in fact very dynamic with shift­ In the clear cutting system all the old stand is ing time components of stands as they age, felled in one operation. The new stand is regen­ develop, mature and are regenerated on a small erated, usually by planting but sometimes natu­ scale. Forest cover is the permanence. In prac­ rally, and generally establishes itself without tice, some of these components are classified as any shelter from the old canopy. This simple pro­ follows: cedure is acceptable for species that are hardy and demand full light to grow, e.g. larches, pines, 1. Stands are classed pure when they consist of birches and oaks, and it is a reliable way of one principal species, and mixed when there restocking with pure or mixed species of light- are more than one. demanders. It provides an opportunity to intro­

2 duce new cultivars with improved performance. same site for several years. In the pines, the The range in scale of operations is varied but, regeneration period is normally up to 10 to 15 when the individual coupes decline in area to years; in oak and beech and mixtures of these about 2 ha, the benefits of scale tend to be lost two species, the time taken to remove the old and the system gives way to others. trees is usually 20 to 30 years but can be short­ When a forester wishes to use the old stand to er if good regeneration is obtained early. During provide seed and shelter for the regeneration the regeneration period the stand is two-storied and early development of a new crop, a group of and may consist of two or more species. The silvicultural systems called the shelterwood sys­ diversity in appearance and habitat can be rich; tems (see Chapter 2) can be used. The old stand moreover, in some cases some of the best seed­ is felled in two or three stages during 10 to 30 bearers may be retained as standards and the years depending on circumstances. The new crop dense lower canopy gradually ascends to meet regenerates and becomes established under the the sparse upper one. changing conditions produced by manipulating When one of the shelterwood systems is the retained seedbearers in the old stand. The applied to coupes of (say) 10 to 25 ha, seed pro­ amount of shelter produced by overhead shading duction is regular, and good prices are obtained and/or side shading can be used to stimulate the for large, high quality logs, the time and atten­ formation of mixed stands. The systems com­ tion to detail demanded from the forester is monly include the use of planting to reinforce the repaid. Great emphasis is placed on the ground stocking from natural regeneration, or to adjust conditions so that weed growth does not inhibit the species mixture. Species which are initially the germination and growth of natural regener­ frost tender and prefer shade during regenera­ ation; and the stands are carefully thinned so tion are favoured in the longer regeneration that the seedbearers can be selected from among periods, e.g. beech, Douglas fir, grand fir, west­ windfirm trees with stems of high quality. A high ern hemlock. Shorter regeneration periods and managerial input is ultimately rewarded by good more uniform regeneration occur with oak, ash financial returns. But in Britain there are few and sycamore. places where large stands of quality broadleaves The phased removal of the old stand may be exist that will repay adoption of the shelterwood effected over quite large areas, the uniform shel­ systems; a notable exception is part of the Forest terwood system, or may start from several small of Dean. The systems can be widely applied to nuclei which are gradually extended and even­ light-demanding conifers provided that the tually merge, the group shelterwoood system. windthrow hazard classification of the site does Alternatively, when it is done in strips, or in not exceed 3 (or perhaps 4 with stand prepara­ groups organised into strips the shelterwood tion) and the soil is freely drained. Examples strip system and strip and group system are given in Chapter 2 will allow readers to judge for being used. If adequate seed production occurs themselves. at short intervals or prompt recourse is made to Some silvicultural systems can create similar enrichment planting, the new crop is quickly diversity in species, structure, appearance and established, and is even-aged and regular in habitat in all parts of the forest. Instead of large appearance like that in the clear cutting system. open spaces at regeneration time, as are But seed production can be spasmodic and less required by the extensive clear cutting and shel­ abundant, so several seed years may be needed terwood systems, the forester manipulates a con­ and eventually planting done to fill the gappy tinuous canopy, which is only partially broken regeneration; in the latter case the new stand when regeneration is to be favoured, and then tends to be irregular in age and size; but after over quite small areas. The forester seeks to cre­ cleanings and early thinnings the stand becomes ate a ‘natural’ appearance which is sustained for more regular in appearance. as long as possible, and ideally in perpetuity. In all shelterwood systems the two stands, When the regeneration period is lengthened, one mature and the other young, grow on the the stand becomes more varied in age and size,

3 and more irregular in structure. If it takes half and very deep, extending right down to ground the rotation to restock part of a forest, that part level; there are usually three canopy layers will contain trees ranging in age from (say) 1 to which approximate to small-sized, middle-sized 50 years. If the regeneration period is equal to and large trees (classes often used in enumera­ the time taken to reach an exploitable diameter, tions). The most obvious size classes in the mix­ the whole forest is being regenerated continual­ ture are the middle and large trees and it is ly and becomes completely irregular in struc­ among these that thinnings are undertaken to ture. In practice, the whole management plan adjust the distribution of size classes, which in area is usually divided into parts or blocks and theory must correspond in numbers to a stem- the work of felling mature trees, making thin­ number curve for the stand type. The stems of nings and cleanings and favouring new regener­ large trees in the upper canopy are usually ation is done in different parts, usually on a 6- strongly tapered, but of a good size and well suit­ year cycle. The first situation described above ed to the market. The species will vary, but in leads to the irregular shelterwood system, and Switzerland and other parts of central Europe a the second to the two selection systems (see mixture of Norway spruce, European silver fir Chapters 3 and 4). If the sites being regenerated and beech is commonly found. In Britain, are each roughly equal to the crown spread of European silver fir is usually attacked by one or two mature trees, the system is single tree Adelges nusslini and is not widely planted; how­ selection (see Chapter 4); and if each area ever, it can be replaced by some useful exotics cleared is made larger so as to allow for light- such as noble, grand, Caucasian and Pacific firs, demanding and moderate shade-bearing species and western red cedar and western hemlock. the system is group selection (see Chapter 3). Sycamore and Norway maple may supplement The ideal condition of a forest produced by the beech. irregular shelterwood system is a varied canopy The group selection system (see Chapter 3) of groups of the favoured species, retention of seems to be well suited to the small scale of mature and maturing stands so long as they ful­ much forestry in the lowlands of Britain. Here a fil their function, whether of sustained produc­ great deal of the broadleaved silviculture tion of timber, or maintaining site conditions, depends on light-demanding species including filling gaps in a range of age classes and size ash and wild cherry (gean) but the system is also classes or providing diversity. There is no formal relevant, using smaller groups, to beech, statement of rotation. The system is often used sycamore and Norway maple. A wood or forest in Europe to introduce a shade-bearing mixture managed by the group selection system reveals a of beech and European silver fir into stands of series of groups originating at different times, up pure, even-aged Norway spruce. The result is to 0.6 and even 1.0 ha in size, growing side by often a forest of fine appearance, which because side and in various stages of treatment, e.g. of the emphasis placed on improving the site cleaning, thinning, and others restocked after conditions and the quality of the growing stock is felling. This diversity over small distances is also productive in both yield and income. made possible by the small size of the groups However, the conventional irregular shelterwood which are encouraged to blend together to system (Femelschlag in German) has so far been exploit the mutual synergy between species. untested in Britain, and only small scale vari­ Not all the benefits claimed for mixed, ants of it are mentioned in this Bulletin. uneven-aged forests produced by the two selec­ The conditions of single tree selection forest tion systems can be completely substantiated. (see Chapter 4) can be appreciated by imagining Such forests are said to be resistant to fire, but being in a place surrounded by trees ranging most are used in areas where fires are infre­ from the smallest to the largest in the forest. The quent. They are also claimed to be more resis­ young trees are rather slowly grown but are of tant to wind and snow and to be more flexible to shade-tolerant species able to recover when they meet storm or pest events than forests produced receive adequate light. The canopy is complete by clear cutting and shelterwood systems, a

4 claim that seems to be home out in practice. The lengthening the regeneration period, breaking argument that uneven-aged systems produce up the existing structure, and using systems similar or even greater yields than regular other than extensive clear cutting. The process forests can be resolved rather simply as follows: must begin early and at not later than 40 years if the ground in the cases being compared is cov­ in conifer stands and 60 to 80 years in ered by a complete canopy of foliage of the same broadleaved stands. species, why should there not be roughly equal Remaining after the selection felling of production of dry matter or timber? If both broadleaves during World War II (1939-1945) stands are fully stocked, there should be similar was a jumble of ‘derelict woodland’ (see Chapter yields; in practice, however, it is easier to control 3). Some of this area was converted by clearing stocking in an even-aged stand than in an and replanting, enrichment, or underplanting, uneven-aged one, but the structure in the former together with cleaning and thinning of well- can be more vulnerable to storm events. stocked and promising groups of advanced nat­ One great advantage of selection forest is that ural growth and older stands. The success of good quality trees can be retained for as long as these operations during the post-war years pro­ they are producing an acceptable increment, and vided good examples for foresters in how to deal this leads to a potent argument that is easy to with degraded broadleaved stands by adapting accept, namely that the financial yield of irregu­ these silvicultural techniques and using alterna­ lar forest is at least equal to that of comparable tive systems of silviculture to create new forms regular forest. It is also possible to resolve in of stands, woods and forests. Thus, if there are similar unsophisticated terms the argument existing woods which contain small groups of that the costs of management of irregular forest regeneration with some straight, vigorous and are high. The counter-argument runs as follows: well-formed trees, or larger stands either pure or given that the silvicultural costs and those of with a mixture of species, or old trees that will management are higher than those of compara­ survive for some years, these can form a basis for ble regular forest, the costs could match the ben­ conversion to a new system. Old poor quality efits if a value was placed on permanent ameni­ trees or groups can be felled to form new regen­ ty, recreation and the provision of habitat for eration groups. Glades can either be planted or wildlife. Irregular forest and the selection sys­ left unstocked as part of the conservation man­ tems are clearly the better providers of these agement plan. The details of operations done to benefits. make conversion from one form of such wood­ land to another are described by Garfitt (1980, Preparatory work in conversion 1984), Evans (1984), Matthews (1989) and Hart The particular features of diversity being sought (1991), and are further discussed in this will normally be obtained by taking an area of Bulletin. woodland or forest and converting (or transform­ The attempts at uneven-aged forestry in ing) it to an alternative silvicultural system. Britain have been made by foresters who have British foresters are normally not familiar with been enthusiastic and patient, also opportunistic converting plantations into forests because so and flexible in management, not always adher­ many of our stands are pure, even-aged and reg­ ing to any formal system, so that it is difficult to ular, and clear cutting has been the system classify some of the examples in this Bulletin adopted previously. The object is to make some of because distinctions between variants and adap­ the area mixed, uneven-aged and irregular by tations have become blurred. Chapter 2 Shelterwood systems

Introduction ment unit shall be felled and regenerated during the course of a rotation, the tradi­ Shelterwood systems have many variants but tional approach is to divide a forest into a essentially involve partially felling a stand to number of ‘periodic blocks’, each to be felled leave a scattered overstorey of mature trees (as and regenerated in turn during the course of seedbearers); regeneration then takes place in successive ‘periods’ (Matthews, 1989; the overhead or side shelter of the seedbearing Benezit, 1991). The system practised on the trees. Planting is used only to supplement Continent with oak and beech has been dis­ areas where natural regeneration is inade­ cussed by Everard (1985), and with Scots quate or has failed. In Britain the systems pine, larch, European silver fir or Norway have been used for conifers, particularly Scots spruce by Yorke (1991a, b). pine, but less used for broadleaves; however, • Group shelterwood system Group shelter­ under the right conditions they have succeed­ wood is similar to uniform shelterwood but ed, and are particularly suitable for heavy regeneration is in groups, usually under 0.6 seeded species such as oak and beech (Table ha. It involves the opening up of the canopy 2.1), although the shade regime has to be han­ by scattered groups which may need prepa­ dled carefully for the latter (Evans, 1984, 1988; ration by thinning of the margins. Kerr and Evans, 1993). The traditional method • Shelterwood strip systems The shelterwood comprises seeding, secondary and final strip systems include strip and group, and fellings, which are described in detail by wedge. They are devised for regions with Matthews (1989). Its use for conifers is most more windy climates, on sites with brown common in regions with warm summers earths, podzols, and occasionally gleyed favouring seed production, on sites with low soils, and on flat, moderate or occasionally windthrow hazard (WHC 1-3), deep soils, light steep terrain. Foresters can use the systems weed growth and flat to moderate terrain. The in mixtures containing light-demanders systems’ main advantages are that they pro­ duce uniform pure or mixed crops over large (pine or larch), moderate shade-bearers (spruces or sycamore) and shade-bearers areas; and the seedbearers add valuable incre­ (beech or silver firs). The main advantages ment to compensate for any longer regenera­ are that they provide sheltered regeneration tion periods. Their main disadvantage is that a long period of preparation is required to make sites and control of the micro-climate dominants into windfirm seedbearers. through use of overhead and side shade. Skilled directional felling is involved but • Uniform shelterwood system In the uniform harvesting is easy to organise. When used in shelterwood system the whole stand is progressive fashion, the systems can be opened up uniformly for regeneration pur­ visually attractive. The main disadvantage poses, and promotes a largely uniform and is the requirement for careful thinning to even-aged young crop. To systematise opera­ create stable edges on the inner and outer tions and ensure that the whole manage­ strips. The wedge system can be used on

6 Table 2.1. The location of stands treated by the uniform and group shelterwood systems OwnershipWind zone 3 Main Main shelterwood species system

England

Forest of Dean Forest Enterprise 3 Uniform-Group Pedunculate oak

Savernake Forest Forest Enterprise 1 Group Beech (lease)

Slindon Park, National Trust 2-3 Uniform Beech West Sussex - terminated Windsor Forest

Cranbourne Wood Crown Estates 2 Uniform ' Pedunculate oak

High Standing Hill Wood Crown Estates 2 Uniform Pedunculate oak

Wakefield Estate (Northants) Private 3 Uniform Ash

Wyre Forest English Nature 3 Group Sessile oak

Wyre Forest Forest Enterprise 3 Uniform Douglas fir

Watlington Forest Forest Enterprise 1 Group Beech Queen Wood Windsor Forest

High Standing Hill Wood Crown Estates 2 Uniform Grand fir

Swinley Crown Estates 2 Uniform Scots pine

Scotland

Glen Tanar Estate Michael Bruce 5 Uniform Scots pine Aboyne

Craigvinean Forest, Forest Enterprise 3 Group Douglas fir Dunkeld

The Hermitage Wood, National Trust for 3 Group Douglas fir Dunkeld Scotland

Grampian and Tayside Forest Enterprise 3-5 Underplant Firs etc. a Wind zone describes the regional windiness for the site; values obtained from Quine and White (1993) and Quine (person­ al communications for areas in SE England). See Glossary 1 for further definition.

7 sites with moderate windthrow hazard resulted from an abundance of mast in 1899, (WHC 3-4) provided the soils are quite deep; protected with fencing against sheep (Troup, in Britain it is worth testing on gleyed soils 1928, 1952; Hart, 1966, 1991). Some 160 ha with Sitka and Norway spruces in moist were established in this way. At Sutton (not wet) regions. Bottom, 11 ha were regenerated in 1909. A gradual removal of the overstorey oak was made. Gaps were planted with Norway Use in Britain : broadleaves spruce, sycamore and European larch. Fine stands of oak resulted and these were under- Uniform shelterwood planted with beech by 1952. The stands are The uniform shelterwood system has been little referred to as ‘Hanson’s oak’. used on broadleaves in Britain. Few British At Churchill East (Plate 1), the uniform foresters have gained the experience and skill shelterwood system is being used to regener­ needed to promote and manage natural regen­ ate a 1857 stand of pedunculate oak and beech eration. Other factors which make the system on shallow brown earth over broken rock. less easy to apply to oak and beech in Britain Following a heavy oak mast in 1989, the are high densities of deer and rabbits, and pre­ broadleaved understorey on 3.5 ha was dation from pigeons and mice is also very high. removed in 1990, and a seeding regeneration In addition, an important difference between felling of the overwood left a 60% canopy. A Britain and the Continent is the lower frequen­ secondary felling followed in 1991. In each cy of good mast years; also absence of ground felling the hardwood lop and top was placed ‘cultivation’ by wild boar. Troup (1952) and around the base of the remaining parent trees Evans (1988) listed good mast years for beech in so as not to hinder weeding of the young crop. Britain in the first half of this century, and The natural regeneration of oak at present is emphasised that they are not only less frequent 10-40 cm tall and equalled by a similar but also more erratic in occurrence than on the amount of beech, though slight damage has Continent. Foresters can do little to increase been done by caterpillars and oak mildew. The this supply of seed apart from increasing the young plants are lightly top-weeded, mainly size and vigour of crowns of seedbearers by in bracken. A further secondary felling was crown thinning (Matthews, 1963; Evans, 1988). planned for 1993. The final felling, with the But, apart from this difference in seed supply, exception of 5-10 large oak reserves per there are no obvious ecological differences hectare, will be done in 3-5 years time. The between Continental and British conditions, eventual aim for the new stand is a 50/50 mix although shade is better used to keep a clear for­ of oak and beech, most of the latter to be man­ est floor in Europe. Selective felling and lack of aged as an understorey for the benefit of the understorey have often allowed the forest floor oak (J. Everard, personal communication). to receive too much light and to become invaded by grasses and bramble. Several foresters have attempted to use the uniform shelterwood sys­ tem for broadleaves in Britain. Five notable Savernake Forest, Wiltshire examples are described.

Forest of Dean, This former ancient royal forest is now pri­ vately owned but was leased to the Forestry Gloucestershire Commission from 1939. For several centuries it appears to have been used mainly as wood­ land pasture with small copses and woods At Blakeney Hill West, after seeding fellings, within. Old beech from the 1740s indicate excellent regeneration of pedunculate oak that it can grow fine trees, although the

8 provenance appears not to be good. Windsor Forest, Berkshire Management this century has included suc­ cessful regeneration of many of the over­ mature 18th century beech stands, notably along the 1750 Grand Avenue which had The Crown Estate has one of the most deteriorated through drought, beech-snap impressive examples of pedunculate oak nat­ disease and age; and, near the Eight Walks ural regeneration in Britain. In both centrepiece, by using a group shelterwood Cranbourne Wood (Cpt. 107) and High system (R. Budden, personal communica­ Standing Hill Wood (Cpt. 21d) the fine regen­ tion). eration is now (1995) 6-8 m tall, and very Compartments 74b and 75b received a dense (Plate 3). The old overwood was felled more uniform regeneration felling of beech, during World Wars I and II. taking advantage of advance natural regener­ Natural regeneration appeared in 1962 at ation during the past 25 years and associated a density of 20000-25000 ha1. Seeding losses in part with clearing up after the gale in due to natural causes, and mortality due to January 1990. Some older groups of regener­ suppression by dominants, followed by a ation are becoming diverse in canopy struc­ recent cleaning, reduced the density to an ture and age (Plate 2). acceptable 6000 live quality stems ha1 in The old stands are generally too even in 1992. Re-spacings will be done as necessary, age and the plan now is to break up the older while maintaining stem and crown density to age classes and achieve diversity by a mix of restrict direct light and the formation of epi- natural seeding and planting. The best cormic growth. stands are good examples of converting old Details of the early treatments applied beechwoods by a mixture of uniform and were not recorded, but Peter Garthwaite group shelterwood to diverse age and canopy (1972) describes some of the techniques of structure by natural regeneration, but much establishment used by Robert Lindsay who patience is needed to achieve adequate natur­ was Director of Forestry when these compart­ al regeneration, followed by some respacing. ments were established. Acorns were often The high value of Savernake Forest for recre­ broadcast to boost natural seed supplies, and ation, wildlife and timber production will be a large rotovator was usually used to ensure greatly enhanced in the future. this seed was buried. Oak seedbearers were usually retained, but the number, and the length of time they were retained, is not known. It may be that these oak crops were Slindon Park, West Sussex established with very little shelter from an overstorey (F.G. Stickler, personal communi­ cation). In 1978, at Slindon, owned by The National Trust, Aldhous found ‘plentiful’ natural regeneration under fine mature beech which Wakefield, Northamptonshire had been thinned moderately from time to time since 1950 (Aldhous, 1981). The forest floor had been scarified before the fall of In 1982 Peter Garthwaite made an interest­ heavy beech masts in 1970 and 1976. In 1978 ing attempt to manage ash on a shelterwood some 20% of the older beech were severely system. Most of the wood was mixed ash, oak, affected or killed by drought and beech bark lime and sycamore coppice 30-50 years old, disease, hence were felled. Sadly the over­ with some oak and ash standards. A heavy storey and the regeneration were ruined by thinning was done in compartment 1 of 14 ha the severe storm of 1987. reducing the basal area to 19 m2 h a 1. The sys-

9 tern was continued until 1987 when the good stands have been established by this October storm in southern England caused a low-cost regime despite some sheep browsing. glut in the market for ash so that felling was However, in some areas the density of oak suspended. During this period the majority of and beech regeneration can still be rather compartments 1-6 (79 ha) had been selectively sparse, with birch and coppice regrowth pre­ thinned. The response has been quite dramat­ dominating. Hopefully, within this low-grade ic both on the trees remaining and on the matrix there are sufficient oak seedlings to extent of ash natural regeneration. The latter form a final stand. The use of the coppice is entirely dependent upon light availability, regrowth as a nurse to such seedlings there being no shortage of viable seed from the requires careful timing of cleanings and standing ash. There has been little or no oak respacings, if the oak seedlings are not to be regeneration but some sycamore where seed overtopped. More recent management of this trees exist nearby. John Newcomb took over part of the Forest is probably best classified the management in 1988 and the thinning as ‘uniform shelterwood’ (J. Everard, person­ recommenced in 1989-90 when a further 45 ha al communication). In summary, the system was worked. About 25 ha remain to be thinned is used to regenerate broadleaves, but is as soon as the market for ash improves. described as ‘shelterwood with retentions’. The size of regeneration unit is 1 ha in the Conservation Working Circle, and up to 5 ha Group shelterwood in the Production Working Circle. Seeding fellings now aim to leave 60% canopy. In the lowlands of England, the examples of group shelterwood systems have tended to merge with uniform shelterwood systems. Management has been more informal than in Wyre Forest, Europe, and use of seeding, secondary and final fellings and precise regeneration periods north Worcestershire and are unusual. The following three adaptations south-east Shropshire are noteworthy.

Wyre contains the remnants of a former royal Forest of Dean, forest which later came into private owner­ ship. Now, much reduced in size by clearance Gloucestershire and modified by reforestation, it still repre­ sents one of the largest surviving areas of ancient broadleaved woodland in lowland Here the regeneration and management Britain. Part is owned privately, many of the regime followed by the Forestry Commission areas being designated as Sites of Special for 1200 ha of broadleaves (Joslin, 1982) was Scientific Interest or National Nature based on ‘group shelterwood’. Seeding Reserves, and largely managed by English fellings, removing 60% of the volume were Nature. Other parts are owned and managed made to create and enlarge groups, followed 5 by the Forestry Commission. to 7 years later by one or more secondary The area managed by English Nature, fellings. The timing of the final felling some 570 ha, consists largely of stored cop­ depended on the progress of regeneration, pice sessile oak, mainly 40-50 years in age amenity and other considerations. After 11 with some up to 100 years old, mostly with years of applying this system, 12% of the area little productive understorey. There is a range was successfully regenerated, and 48% of soil conditions, mainly acidic and often stocked with advance regeneration. Many poor in nutrients, and the field layer is var­

10 ied. Throughout this century the original and excessive bramble) and the invasion of coppices have grown into high forest of oak, birch. Throughout much of the woodland plus a limited number of other species which there is good regeneration of oak, but unless have become adapted to this type of wood­ fenced it is continually damaged by fallow land condition. Some of the less common deer; and even after 6 years is rarely taller broadleaved species are responding to man­ than a few centimetres. agement and their numbers are increasing. The main aim of management is to create high forest in 90 to 100 years with a diverse Queen Wood, Watlington structure instead of the present random range of age classes, about half by the group Forest, The Chilterns, shelterwood method; but about one-third is Oxfordshire reserved as pure coppice on a 30-year cycle (coupes about 1 ha in area, 3.5 ha cut each year); the remainder will include coppice with standards, and some areas of limited The wood demonstrates several successful silvicultural treatment. Much attention is examples o f‘group regeneration’ of beech. In given to sensitive edge treatment of rides, 1945, 27-36 m diameter groups were created roads and pathways. In any case, owing to by felling, resulting in adequate beech nat­ changes in soil and site conditions, part will ural regeneration around the perimeters; become diverse in structure during the next the groups were too large and prevented century (J. Robinson, personal communica­ beech mast reaching the centres (Brown, tion). 1953). Queen Wood was purchased by the Forestry Commission in 1947 as a demon­ In Shelf Held Coppice, much of the con­ stration area for beech regeneration. By version from unbalanced age classes to more then, the clearings were occupied by a vari­ diverse conditions of age classes and size ety of low shrubs and tall herbs, and few classes is planned by group selection felling beech seedlings. In 1952 and 1961 parts of (see Chapter 3) and regenerating by natural the mature and over-mature broadleaves, seedlings or coppice regrowth. The system is predominantly beech, were felled in strips being used with the expectation that it will and replanted mainly with beech but also be worth repeating throughout one half of with some oak and other broadleaves. More the Forest. In the 1980s, a compartment of recently, coupes of varying shape and size 50 to 60-year-old sessile oak derived from but typically 0.5-1.0 ha were created by stored coppice was heavily thinned to group selection fellings in some of the poorer encourage crown development and seed pro­ stands and successfully restocked using duction. Seventeen groups, each of about 30 treeshelters mainly containing beech but m diameter (0.07 ha), were felled in 1985; in also oak and wild cherry. Despite lack of sup­ 1986 there was an abundance of oak port from natural seeding, and some damage seedlings, approximately 10 per m2 (Pryor, by deer and adjacent windthrow in January 1985). Subsequently, six larger groups of 1990, the foregoing are commendable exam­ about 0.5 ha have been felled and fenced ples of conversion of mature and over­ each year in the 40 to 90-year-old stands. mature beechwoods. Limited enclosure of The success of the restocking has depended small coupes or groups in shelterwood largely on site condition, not only with oak stands is intended in order to benefit from but also with other broadleaved species, future beech masts (K. Wallis, personal com­ notably birch. Within the fenced groups munication.). This example could be termed (Plate 4), the natural regeneration depends either group shelterwood or group selection. largely on the soil fertility and on the vary­ ing field layer (bracken, heather, bilberry Use in Britain : conifers provide a regeneration surface is necessary to overcome the effects of Calluna and Uniform and group shelterwood Sphagnum mosses (Low, 1988). The appropri­ ate structure may evolve after continued shel­ In the uplands and heathlands of eastern terwood experience and after research on light Scotland, in wind zones 3-5, the mainly uniform requirements. The forest manager since 1977 shelterwood system has been successful on has been Irvine Ross. Scots pine natural regeneration on podzols, where Vaccinium spp. are dominant over the forest floor. Besides Seafield Estate (notably Craigvinean, Perthshire Curr and Grantown Woods), examples are Glen Tanar Estate, Craigvinean and The Hermitage.

Craigvinean is an extensive Forest Enterprise forest, about 1 mile (IV 2 km) west Glen Tanar, Aberdeenshire of Dunkeld. It adjoins the upper side of The Hermitage Wood, described below. Compartment 3040e (2.9 ha) comprises Since 1970, Michael Bruce’s native pinewood mature Douglas fir at the eastern end with near Aboyne has applied both the uniform younger planting of the same species in P58 shelterwood system using scarification and and P80. Much of the old stand has produced small group regeneration fellings of 0.3 ha — natural regeneration of mixed ages and as irregular shelterwood — in stands aged heights, in effect being an unplanned varia­ 150-170 years. The success of natural regener­ tion of group and uniform shelterwood sys­ ation in extensive clearings had been well tems. The 1980 stands were formed by plant­ demonstrated several times in the recorded ing in gaps following windthrow. The appear­ history: following heavy selective fellings from ance now is of irregular shelterwood. The 1810 to 1840, protected from deer until 1865, management plan is to thin the 1958 stands, and following an extensive severe fire in 1920 and to respace the understorey of natural and after the 1953 storm damage (Edwards, regeneration. The system will be extended to 1981; Miller and Ross, 1990; I. Ross, personal a further 113 ha in Craigvinean to achieve communication). Stands were prepared for uneven-aged stands in Sitka spruce, Norway regeneration by thinning in the late pole stage spruce, Douglas fir and larch. In part of the to make seed trees windfirm. The group regen­ adjacent compartment 3040g with old Scots eration has suffered from severe deer browsing pine, beech and oak and younger Douglas fir, and the second stage of treating this as irregu­ the management plan is to maintain diversi­ lar shelterwood has had to be postponed. The ty of species and structure. two stands treated by uniform seeding fellings In 1993, Craigvinean Forest (Forest and mechanical scarification show promising Enterprise District Manager, Charles Taylor) regeneration. It has been found that a stocking won the new ‘Landscape Institute Award’ for of 100 seed trees ha-1 can cast excessive shade the design plan. if crowns are well developed. The most successful regeneration at Glen Tanar, extending to over 100 ha, has taken place alongside the pinewood on heather moor The Hermitage Wood, Dunkeld treated by muirbum progressively over 20-30 years. Indications are that regeneration groups for Scots pine may need to be extensive Alongside the River Braan, a tributary of the in size or of few seed trees per hectare to pro­ River Tay, about 1 mile (IV 2 km) west of vide adequate light. A degree of scarifying to Dunkeld, lies The Hermitage Wood, owned by

12 The National Trust for Scotland. It comprises 1. too short a period for preparation and iso­ about 15 ha of mixed conifers and lation of seed trees and thinning on bound­ broadleaves of great beauty and wildlife aries; interest. The south-west upper part (com­ 2. poor provenance of original stand; partment 4, 4.3 ha) contains stands of large 3. climatic conditions causing frequent failure Douglas fir, Norway spruce and Scots pine 70- of cone crops; 80 years old, some standing above sporadical­ 4. competition from Deschampsia flexuosa ly formed groups and single trees of natural­ sward; ly regenerated Douglas fir (0.2-4.0 m in 5. inadequate protection against deer browse. height) with smaller and less thriving Norway spruce. These are examples of Treatment resulted in sparse inadequate unplanned variants of both uneven-aged and stocking of regeneration. even-aged shelterwood. Slightly lower down In the lowlands of England, three examples the slope are small groups of mid-rotation of mainly uniform shelterwood, using conifer Douglas fir and Norway spruce, with a few natural regeneration are at Windsor, Swinley grand fir, of varying ages believed to have and Wyre. been naturally regenerated (Plate 5). Windsor Forest, Berkshire Grampian and Tayside Regions

On the Crown Estate in High Standing Hill During 1960-1970, in the Grampian and Wood is an impressive example of natural Tayside regions of Scotland, 40-year-old larch regeneration of grand fir managed under a more stands were successfully underplanted on a 5 or less uniform shelterwood system (Plate 6). to 10-year regeneration period by the Forestry Following heavy windthrow in January 1990, a Commission under the influence of Frank P52 stand of grand fir was thinned and the Oliver and Morley Penistan (see Appendix 5), branchwood cleared in spring 1990. This coin­ using Douglas fir, grand fir, western hemlock cided with a good mast, which dining 1990-1991 and Norway spruce. Removal of the upper led to dense natural regeneration, now (1995) 5 storey was often delayed owing to low demand years old (F.G. Stickler, personal communica­ for larch sawlogs and this occasionally result­ tion). The overwood, which now stands at 575 ed in two-storied high forest. The system fell stems ha1, will be progressively removed. A into disuse due to: thriving and inexpensive new stand will be gained by skilled management of this timely use • reduced intensity of supervision; of natural regeneration. • changes in extraction methods; • loss of skilled fellers; • improved weed control in clear cutting by Crown Estate, Swinley use of herbicides; • a preference for Sitka spruce which did not require protection of an overwood. There are fine examples of Scots pine growing on Bagshot heath, established over a period However, in England the shelterwood sys­ from 1950 to 1970 under the uniform shelter­ tem has been unsuccessful with Japanese wood system (Plate 7), currently in use to larch, aged about 40-50 years on brown earths, restock areas devastated by the storms of when aimed at using natural regeneration, e.g. 1987 and 1990. at Grizedale, for reasons advanced by Paterson Scarification is undertaken to encourage (1990): germination, which typically amounts to

13 about 25 000 seedlings h a1 following removal large areas of conifer forest or broadleaved of the parent trees. Costly early respacing is woodland, have turned to the uniform shel­ unnecessary due to natural suppression and terwood system which was designed to emergence of dominants; so stands are left restock large areas with regular even-aged untouched for 25 years. Following an early stands like those produced through the clear thinning to remove dead and suppressed cutting system. This system can work in stems, stocking is likely to be about 5000 Britain and examples are provided by the stems ha-1. Periodic thinning is practised and Scots pine and pedunculate oak on the regeneration fellings are commenced at age Crown Estate at Windsor, pedunculate oak 55. Final felling is carried out when stands in the Forest of Dean and beech at Slindon are fully stocked and the new crop is about 1 (although the last named has been lost by m tall. These fine examples of uniform shel­ windthrow). The small areas of grand fir at terwood system have proved to be more prof­ Windsor and Douglas fir in Wyre Forest were itable, more windfirm, and of better quality responses to windthrow in the old stand. (straighter and cleaner poles, naturally 3. Regeneration: influence of combination of pruned) than planted stands. The marketing soil and species. The relatively strict of produce is facilitated by a good local sequence of regeneration fellings, i.e. seed­ demand for transmission poles (F.G. Stickler, ing, secondary and final, are used for Scots personal communication). pine at Windsor, where the favourable com­ bination of soil and species has prompted use of the uniform shelterwood system to Wyre Forest produce abundant regeneration which can be selectively grown on as transmission poles for a demanding but profitable mar­ Within the Forest Enterprise management of ket. (This was the practice followed histori­ part of this ancient forest, a stand of 17 ha of cally on the Seafield Estate, Strathspey.) P26 Douglas fir which suffered severe wind­ Elsewhere, regeneration fellings have fol­ throw in 1974 was followed that year by nat­ lowed a less formal pattern, except in the ural regeneration of Douglas fir (with some case of recent treatment of oak and beech in Corsican pine and European larch) and mixed the Forest of Dean. broadleaves, mainly birch. Further seeding 4. Regeneration fellings. In the group shelter­ continued. In 1988 the overstorey was lightly wood system the regeneration fellings are thinned, removing about 50 stems ha1. The used to enlarge groups of regeneration so result (1995) is a heavily stocked thriving that eventually they coalesce, but so far example of shelterwood. Respacing is soon to British foresters have used them in an infor­ be undertaken and final felling of the over­ mal and even short-term manner. This sys­ wood will follow in due course. tem, as adapted to British conditions, appears suited to small mixed woods provid­ Conclusions on shelterwood ed they are sheltered from wind, and the systems soils are light to medium in texture and quite deep. Sometimes there is a distinct 1. Application to light-demanding conifers. feeling that the owners and foresters are on Shelterwood systems can be widely applied to the verge of using variants of the irregular light-demanding conifers provided that the shelterwood system and this is the case in windthrow hazard classification of the site The Hermitage Wood near Dunkeld, and does not exceed 3 (or perhaps 4 with stand parts of the Glen Tanar native pinewoods. preparation) and the soil is freely drained. When using the irregular shelterwood sys­ 2. Regeneration in large areas. Foresters on tem it is not necessary to specify the regen­ the Continent, when having to regenerate eration period closely, nor is the rotation

14 fixed; the seedbearers can be left until the encourage appearance of the shade-bearers; current annual increment falls below an and further reducing the stocking of the old acceptable level. But irregular shelterwood trees to 100 h a 1 should induce light-deman­ produces a wood or forest that is close to ders to appear. It is important that these gen­ group selection, and the owner or forester eral prescriptions should be tested on repre­ might consider using continuous inventory sentative trial areas before embarking on of three size classes to regulate yield. operations using the shelterwood systems. It will also be seen that controlling the stocking 5. Small stands. In parts of Britain there are of the seedbearers or overstorey provides the admirable small stands of large old forester with control of the composition of broadleaves and conifer species on sheltered mixtures. Co-ordination of fellings with mast sites (the old beeches at Savernake provide a good example). The group shelterwood sys­ years should be attempted. tem has already been proved useful in sev­ 8. Size o f the regeneration block. The areas eral of these and could be used more widely treated by the group shelterwood system to keep representatives of the old stands have varied from 10 to 50 ha, which is a rea­ and enable the next generation to conserve son for suggesting that the system could be their pool of genes. adapted to the conditions in small woods. Scope is considerable in Britain’s eastern 6. Silvicultural prescriptions. The number of pinewoods and birchwoods generally. It is seedbearers to be retained in a seeding unlikely to be of value in upland spruce felling for the uniform shelterwood and forests because of instability in seedbearers group shelterwood systems can only be on wet soils. finally determined by local experience. As a guide, if maturing stands of the major 9. Protection. Because large (and perhaps small) mammals prefer feeding in small species used in Britain are stocked at 300- 500 stems ha1, the canopy is too dense for clearings surrounded by taller stands which provide cover, regeneration tends to suffer young plants of light-demanders to survive relatively greater damage on coupes of up to for long; the seeding felling should remove 2 ha unless the numbers of deer and other more than half the old stand, leaving 75 to animals are controlled or fencing is erected, 120 seedbearers of oak, ash, birch, Scots or both measures are taken. In Wyre Forest, pine or Sitka spruce. As the regeneration fallow deer browsed unprotected plants of (whether natural, direct seeded or planted) oak to a stunted condition and at Glen becomes established, secondary and final Tanar Forest unprotected native Scots pine fellings are made. For the shade-enduring regeneration is browsed by red or roe deer and moderately shade-enduring species — down to the top level of the heather sward. beech, sycamore, grand fir, Norway spruce, Large groups tend to have greater weed Douglas fir, western hemlock and western growth in the centre of the clearing and red cedar — a seeding felling should retain fewer seedlings, while smaller groups have seedbearers of 150 to 200 trees ha-1. a long perimeter of edge trees for their area 7. Thinning prescriptions. In uniform shelter­ and may suffer from more endemic wind wood, stands should be prepared by heavy damage. The success of the group shelter­ selection crown thinnings to make seedbear­ wood system in restocking the whole area ers more stable. In group shelterwood, thin­ depends on being able to enlarge the initial ning around groups must be undertaken clearings by further regeneration fellings when required. If ‘advance growth’ of shade- around their edges so that the individual enduring natural seedlings appears in groups groups coalesce. If this process is slow, the on the forest floor beneath a canopy of matur­ new stand becomes somewhat irregular but ing trees stocked at 300-500 stems ha1, this is corrected when cleanings and thin­ reducing it to around 200 trees should nings are done.

15 10. Rotation. In the shelterwood systems the intensity of supervision. These drawbacks regeneration period is usually 5 to 10 years also apply to the creation of two-storied for Scots pine and 20 to 30 years for beech high forest (see Chapter 5). The length of and oak. If the rotation includes the period rotation can exceed the rotation of maxi­ for regeneration, the rotation for average mum mean annual increment (MAI) by 15 yield classes will total 60 to 65 or more or more years, entailing some loss of finan­ years for Scots pine and 80 to 100 years or cial performance if these criteria and tim­ more for beech and oak. In the group shel­ ber production revenues are the only ones terwood system the rotation may be longer measured and valued. unless the groups can be enlarged quite 12. Factors leading to success with the shelter­ rapidly. The drawback of a lengthened wood systems. If large logs of the principal rotation can usually be offset by a greater species are in good demand in the long timber value for the seedbearers and an term, seed production of the desired species extended period of having the benefits of a regular and plentiful, weed growth sparce, diverse structure. and regeneration appears readily, the shel­ 11. Factors causing difficulty in applying the terwood systems can provide an efficient shelterwood systems. Experience in the means of forming new stands. Moreover, north-east of Scotland during 1960 to 1970 the appearance of an area of woodland or suggested that the conversion of pure forest is attractive because the seedbearers stands of larch by underplanting them with give enhanced height and stature to the several conifer species could run into diffi­ temporary two-storied stands. But the culty through delayed removal of the over­ effect does not last and a return is made to storey, changing methods of harvesting, a the aspect of regular more or less even- preference for species that did not thrive aged stands, sometimes with a few old under a larch overstorey, and reduced standards reserved for special markets. Chapter 3 The group selection system

Introduction well-suited to small stands where intensive In the group selection system the growing working and close supervision are possible; stock of trees is arranged in groups of varying accommodating a wide variety of species rang­ dimensions appropriate to the requirements of ing from light-demanders, moderate shade- each species: small for shade-tolerant species, bearers to the shade-tolerant species. and larger for light-demanders. These groups Aesthetically the open glades, thickets and tall are dispersed through the forest to form a con­ trees are pleasing. The disadvantages are: tinuous mosaic, so that a recently regenerated skill is needed in selecting groups of trees in group of (say) Norway spruce could adjoin a the existing stand for retention; felling, maturing group of Douglas fir, a pole-stage extracting, regenerating and tending are scat­ group of beech, and a thicket-stage group of tered and on a small-scale. In short, experi­ ash. There will be interaction between these enced and knowledgeable supervision is various groups; and to ensure that they keep required. The costs are likely to be higher than their identities for tending, protection, yield those in a comparable area of even-aged forest; control and harvesting, the dimensions corre­ control of browsing and bark-stripping mam­ sponding to these areas are suggested in mals is more difficult and damage done can be Tables 3.1 and 3.2. Access to the groups is pro­ more severe; and timber quality problems can vided by a network of rides and extraction arise from heavy branching on edge trees. paths, and the groups themselves may be established according to a formal pattern or, Use in Britain: broadleaves more commonly, on a less formal layout. The Census of English Woodlands taken between The formation of such a forest may start 1947 and 1949 revealed the dire condition of sub­ either by converting existing stands of pure, stantial areas of broadleaved woodland, largely even-aged, uniform canopies, or by restoring as a result of selective felling during World Wars exploited ‘derelict woodland’ which contains I and II and of intervening periods of neglect, or scrub, untended groups of regeneration, a few of management mainly for field sports. In 1950 productive stands, single older trees that were the Forestry Commission tested three methods too small or too low quality for exploitation, for converting unproductive, chiefly broadleaved and bare unstocked ground. In both cases the stands to high forest (Wood, 1950): objective of conversion to the group selection 1. clear cutting and restocking with broadleaved system is to create productive forest which is species, either pure or in mixtures; diverse in structure and appearance, uneven- aged, provides diverse habitat for wildlife, is 2. new planting under the shelter of existing suitable for field sports and recreation, and trees or in strips and groups cleared through produces a sustained yield of timber and other degraded woodland; benefits in perpetuity. 3. enrichment with single trees and small The advantages of the system according to groups of commercial species, favouring the Matthews (1989) are: flexible and particularly best trees of the existing stand.

17 Table 3.1Dimensions of groups for different species Species Rotation aCrown diameter Crown projection Size of square (years) (m) area (m2) plot (m2) Douglas fir 50-60 6 20 36 Ash and sycamore 60 9-10 64-78 01 Oak 100 14 164 196 a Strictly speaking, group selection does not have a ‘rotation’; the criterion is the exploitable final tree diameter aimed for. To enlarge the table and adapt it for local use, make measurements of the crown diameters and crown projection areas for mature trees of the desired species in a given wood or forest, and then decide on the desired number of final crop trees in each group.

Table 3.2 Dimensions of clearings to provide growing space for trees of final size. (Sources: Matthews, 1986, 1989) Diameter of Area (ha) Number of______Number of trees of final______size group (m) plants (2 x 2 m Douglas fir Ash, Oak spacing) sycamore 11-13 0.01 25-35 3 1 1 15-17 0.02 45-55 5 2 1 19-21 0.03 70-85 8 4 2 22-24 0.04 95-115 11 5 2 25-26 0.05 120-130 14 6 3 27-20 0.06 145-155 16 7 3 29-30 0.07 165-175 10 8 4 31-32 0.00 100-200 21 9 4 33-34 0.09 215-230 24 10 5 35-36 0.10 240-255 27 12 5 37-38 0.11 270-285 30 13 6 39-40 0.12 300-315 33 15 6 41 0.13 330 36 16 7 42-43 0.14 345-365 30 17 7 44 0.15 300 42 19 8 45-46 0.16 400 44 20 9 47 0.17 435 40 21 9 40-49 0.10 450-470 50 22 10 50-51 0.20 490-510 54 24 15 62 0.30 755 03 37 20 72 0.40 1020 113 50 25 00 0.50 1255 139 62 30 88 0.60 1520 169 75 32 90 0.63 1590 177 78 -

The results of this work became the basis of overwood to establish beech on thin chalk recommendations for improving unproduc­ downland soils in southern England; also of tive State and private stands in the lowlands 2 row/2 row or 3 row/3 row mixtures of (Wood et al., 1967; Evans, 1984). At this larch/oak or Scots pine/beech in cleared time, much use was also made of birch as an areas.

18 Plate 1. Natural regeneration of pedunculate oak and beech under the uniform shelterwood system at Churchill East, Forest o f Dean. (41312)

Plate 2. Beech natural regeneration at Savernake Forest, Wiltshire. Older groups of regeneration are becoming diverse in canopy structure and age. ( t o m B u c h a n a n )

Plate 3. Pedunculate oak natural regeneration at High Standing Hill Wood, Windsor Forest. The fine dense regeneration is now 5-6 m tall. (41314) Plate 4. Natural regeneration of oak at Wyre Forest. Fencing has improved the success of natural regeneration by excluding mammals but the birch is intrusive. (CYRIL HART)

Plate 6. Impressive natural regeneration of grand fir in a uniform shelterwood system at High Standing Hill Wood, Windsor Forest. (41313)

Plate 5. Douglas fir and Norway spruce in a group shelterwood system at The Hermitage Wood, near Dunkeld, of great beauty and wildlife interest. (ANTHONY HART) Plate 7. Fine Scots pine natural regeneration under a uniform shelterwood system at the Crown Estate, Swinley. (41315)

Plate 8. Beech natural regeneration under 110-year- old trees at Workmans Wood, Ebworth, Gloucestershire: an impressive example of continuing group selection of broadleaves. (41270)

Plate 9. European larch in a superb stand at Cawdor Big Wood, near Nairn. (CYRIL HART) Plate 10. Oak trees in fog at Cawdor Big Wood, near Nairn. This famous Scottish wood of great scenic beauty is predominantly pedunculate and sessile oak with some mixed stands of conifers. (ALAN WATSON)

Plate 11. The Glentress Trial, Glentress Forest, near Peebles: a conversion of conifer stands from uniform even-aged clear cutting system to uneven-aged group selection, producing a wide range of benefits including high standard of amenity, nature conservation and forest recreation. (CYRIL HART) Plate 12.Conversion of conifers to group selection system at Corrour Forest, Inverness-shire, showing diverse stand structure initiated in compartment 3. (FELIX KARTHAUS)

Plate 13. Old Scots pine at Faskally Wood, near Pitlochry, Perthshire. An example of conversion by M. L. Anderson and others towards group selection in conifers. (ANTHONY HART)

Plate 14. Benevean Wood in Glen Affric, Inverness-shire: an example of group or random regeneration of indigenous Scots pine. (39518) Plate 15.Douglas fir naturally regenerating under Plate 16.Sitka spruce and beech regeneration mature crop at Diosgydd, Gwynedd: an example of under mature larch at Nant BH, Gwynedd. Note conversion of conifer plantations to group selection. carpet of bilberry. (41267) (41265)

Plate 17. Application of a shelterwood-cum-group selection system to conifers at Longleat, Wiltshire, showing natural regeneration under Douglas fir. (41275) Plate 18. Prolific natural regeneration of western hemlock in a single tree selection system at Kyloe Wood, south o f Berwick-on- Tweed, Northumberland. (H. WHITNEY MCIVEH)

Plate 20.Hamsterley Forest: smaller scale restructuring. (38812)

Plate 19.Kielder Forest: restructuring for increasingly multiple purpose objectives. (39879) Plate 21.Group regeneration o f beech in the Chilterns showing a well- established group of young trees in an opening in the canopy of mature beech. (E7785)

Plate 22. Amixed storey formed by regenerating beech in the Chilterns. (E10357)

Plate 23.Natural regeneration under a stand of c. 150-year-old beech in the Chilterns: an example of restructuring of beechwoods by group regeneration. (10244) In the regeneration of such woodlands, England, concluded that the only successful adaptations of the group shelterwood system method to obtain natural regeneration was to and the group selection system have been used secure advance regeneration and fell only more frequently than other alternative sys­ where it is well established, an assertion con­ tems. The motives have been to achieve forest firmed by some of the following fine woodlands. cover for reasons of silviculture, protection or aesthetic and environmental benefit. The foresters concerned have been enthusiastic Ebworth, Gloucestershire and patient, also opportunistic and flexible in management, not always adhering to any for­ mal system, so that it is difficult to classify Many fine stands of beech are growing on sev­ some of the examples because distinctions eral private estates in the Cotswolds. Most between the variants have become blurred. notably, good quality beech is grown in an Group selection is a term that has been environmentally sensitive manner at widely used in Britain and loosely applied to Ebworth, owned since 1989 by the National almost any regeneration obtained by clearing Trust and managed by English Nature with areas in groups, but in this Bulletin the term the continued assistance of the donor, John refers to systems in which the management Workman OBE. Workmans Wood, occupying area is subdivided into groups or components 120 ha on steep slopes, is an ancient semi­ that are usually even-aged and have one natural beechwood, generally high forest of storey. A group selection stand differs from a diverse age and size, with trees of ages from single tree selection stand (see Chapter 4) in newly seeded up to about 200 years old. The that a time comes when all the trees in a group only current lack is that of 50 to 150-year-old or component must be removed because they trees. It is the most impressive example of are judged mature, whereas in single tree continuing ‘group selection’ of broadleaves in selection working, trees judged to be mature Britain, comprising high quality timber pro­ are felled singly through the forest or section of duction while achieving continuous forest a forest. In group selection there is still a shel­ cover and maintaining rich and diverse terwood notion: an older stand provides protec­ wildlife interest (Plate 8). tion for young crops that are developing along­ The basis of the silviculture is securing side and the period of side shelter may be as advance natural regeneration and, when it is long as 30 years or more. It differs from a sin­ well established, felling the mature trees over gle tree selection stand in that regeneration is the regeneration. Where there has been a gap tended in groups, which take the place of indi­ in the beech canopy, sycamore and ash regen­ vidual trees. The individual groups will not dif­ eration may have come in initially, but usual­ fer much in structure and average age but the ly sufficient beech is awaited before under­ group selection structure will be clearly visible taking small group fellings. A few potentially overall as small stands varying in size, age and fine ash which benefit from mixture with species. In contrast to the shelterwood group other broadleaves are welcomed. Occasionally system, the size of each group or component is beech regeneration will appear when the fixed at the time of regeneration, instead of overstorey is not mature (100 years), in which being steadily enlarged so that the groups coa­ case thinnings are made to favour this, and lesce to form one large stand. eventually the overstorey may be felled The group selection system has been used before it is fully mature to ensure that the more frequently (Table 3.3) than the group advance regeneration does not suffer. shelterwood system because long or indefinite Typically, it takes around 30 years to natural­ regeneration periods have been adopted due to ly regenerate a stand, and towards the end of poor initial regeneration. Aldhous (1981), hav­ this period some infilling with beech trans­ ing visited many beechwoods in southern plants may be necessary. No weeding is done

19 Table 3.3The location of stands treated by the group selection system Wind zone a Site Species Conversion period

England Ebworth Estate, Gloucestershire 3-4 Jurassic limestone Broadleaved 40 years Workmans Wood Hockeridge Estate, Hertfordshire 2 Mixed Cirencester Park, Gloucestershire 1 Jurassic limestone Broadleaved 50 years Dartington Estate, Devon 3 South-west peninsula Conifers Longleat Estate, Wiltshire 2 Lower Greensand Conifers 45 years Tavistock, Devon 3-4 Medium to high fertility Conifers 54 years

Wales Cefn Llwyd Forest, Bala 5 Brown earths Conifers Coed y Brenin, Dolgellau 5 Brown earths Conifers Ffrwdgrech Woodlands, Brecon 3 Brown earths Conifers Clocaenog Forest, north-east Wales 5 Brown earths Conifers

Scotland Bowhill and Eildon Woods, Borders 5-6 Eastern, brown earths Mixed and surface-water gleys Cawdor Big Wood, Nairn 4 Eastern, brown earths Mixed Glentress Forest, Peebles 5-6 Eastern, brown earths Conifers 60 years and surface-water gleys Corrour Forest, Inverness-shire 6-7 High moorland Conifers Drumlanrig Estate, Thornhill 6 Western, brown earths Mixed and surface-water gleys Faskally Wood, Pitlochry 3 Eastern, brown earths Conifers and surface-water gleys Benevean Wood, Affric Forest 5 Northern, wet heath Conifers 70 years a Wind zone describes the regional windiness for the site; values obtained from Quine and White (1993) and Quine (person­ al communications for areas in SE England). See Glossary 1 for further definition.

20 as ground vegetation is controlled by manip­ In 1993, John Workman won one of the two ulation of the overstorey but the regeneration first national Forestry Authority Awards is rigorously cleaned once the overstorey has identifying England’s best forestry. been removed, and secondary species are retained to nurse the beech. Respacing is gen­ erally completed by the 15th year. The groups Since the end of the World War II, J. E. Garfitt are usually less than 0.3 ha in size. (see Appendix 5), following Ray Bourne (see The natural regeneration is mainly of Appendix 5) has been a vigorous and success­ beech of good genetic stock with some planti­ ful designer of variants of the group selection ng from local sources, thus operating an system, particularly for broadleaves, in mid uneven-aged system of groups of trees of dif­ and southern lowland England. The essence of ferent ages. The whole concept is based on Garfitt’s system is that the trees should be fostering regeneration, i.e. observing where allowed to grow in full light, the groups being beech seedlings are appearing and gradually a certain minimum size, defined as having a felling back from these initial patches, allow­ diameter at least one-and-a-half times the ing side-light to penetrate below the canopy height of the surrounding stand. In most con­ and so provide the critical conditions essen­ ditions and on most soils in lowland England tial for successful establishment and early and Wales the optimum size is usually between growth of beech seedlings. The long-term 0.1 and 0.5 ha. Except where total reliance is objective is sustained yield from continuous placed on natural regeneration, the distribu­ forest cover. The notable example at Ebworth tion of the species within such groups is con­ has been achieved by John Workman over trolled by the forester. Since the groups consist some 40 years by skilled, patient and oppor­ of more or less even-aged trees, the distribu­ tunist management in the face of many prob­ tion of the age and size classes throughout the lems: damage by wood-pigeon, grey squirrel, forest is determined by the groups which take wood mice, fallow and roe deer, excessive the place of individual stems in the single tree bramble and wild clematis, and extremes of selection system (see Chapter 4). In group temperature (Workman, 1986). Among rele­ selection it is convenient to avoid small annual vant lessons learned are: fellings and plantings, and to work on a 5- or • voles respace regeneration; 10-year cycle. The choice of this cycle will determine the extent of the felling which is to • badgers aid regeneration by disturbing the be carried out in one year. This will be made up forest floor; of a number of groups in the forest. Co-ordina­ • if the canopy is too dense, the forest floor is tion of fellings with mast years should be clean and inhospitable, but if opened too attempted. much dense bramble and grass smother Garfitt (1984) discusses application of his the seedlings; system to three cases of conversion: • dog’s mercury and nettle hinder regenera­ • from uniform even-aged woodland tion. • from unmanaged woodland The balance is precarious and depends on inti­ • from scrub. mate knowledge of the overstorey and the site. The result of management is a woodland her­ The result of the application will be the cre­ itage of national renown: evidence of one ation of a forest which, once established, will man’s vision and commitment now partnered not change overall. Individual parts will go by the resources of an enlightened national through the familiar stages of young crop, organisation, growing fine quality broadleaved thicket, young pole stage, pole stage, and large timber, and making special provision for land­ trees; but the forest as a whole will remain scape, public access, education and wildlife. essentially the same, all the constituents

21 being constant in the proportion of ground considerable areas in which existing regener­ that they occupy. With the limited size of ation of ash and sycamore occurred in patches. group envisaged, at any one point the full European larch, Scots pine and beech were range of size classes will be visible, producing planted between the patches, using pure effects desired both from nature conservation groups similar in size to those of the regener­ and amenity points of view, that is, open ation groups. This involved the removal of glades, among stately trees of many sizes. At existing inferior growth, including hazel the same time, it will be possible at any stage underwood. Some fine groups of pole-stage to initiate changes in the species used and in stands have resulted under Arthur Lloyd’s any mixtures employed; and by judicious management. manipulation to retain fine specimen trees to In the second application at Cirencester a considerable age, to create vistas, to pre­ Park, Garfitt (1953) applied a form of group serve shelter effects and to preserve important selection in areas where the beech overstorey ecosystems. Three applications by Garfitt are had been, removed in wartime fellings. noted, one at Hockeridge and two at Clearance of the hazel underwood was essen­ Cirencester Park. tial before planting of beech, European larch and Scots pine. No planting was done under the shade of old hazel or other growth. Hockeridge, near Management has changed under subsequent regimes but many of the stages in the original Besrkkamsted, Hertfordshire system can still be traced, although much of the planted conifer was removed in early thinnings, resulting in a greater proportion of This wood was gifted by Miss Mary Wellesley sycamore and ash. Current management is in 1986 to the Royal Forestry Society of tending towards a form of group shelterwood England, Wales and Northern Ireland. Here system leading to final stands of a ‘uniform’ Garfitt (1953, 1984) developed a variant of character. Groups of regeneration are typical­ the group selection system. Management, ly 0.05 to 0.1 ha in size, and the range of ages begun in 1954, consisted of cleaning and within each group is less than 10 years. The enriching any acceptable beech regeneration, aim of the present management is to spread but most was scrub or birch woodland. the felling of the remaining areas of beech Garfitt aimed to retain the benefit of shelter overstorey over 20 to 30 years, by which stage and to create an uneven-aged wood of diverse all the groups will have merged. The most structure through small blocks of contrasting abundant seedlings are ash and sycamore, ages; therefore he clear felled and regenerat­ but through the selection applied in three ed rectangular areas of 0.4 to 0.6 ha. Of par­ cleanings in the first 15 years, the proportion ticular benefit has been the extensive inter­ of beech and wild cherry is considerably nal edges throughout the wood resulting increased. Ash and sycamore are, in effect, from the small groups, making it ideal for used as a matrix or nurse for the final stand wildlife conservation. species which are less abundant. Thus Garfitt’s original group selection system has been converted into a mixture of shelterwood Cirencester Park, systems.

Gloucestershire In the lowlands of Scotland, two notable examples of attempted and projected conver­ sion of mainly broadleaved woodland towards In the first of the two applications at group selection are: Bowhill and Eildon Cirencester Park, Garfitt (1953) managed Woods, and Cawdor Big Wood.

22 Bowhill and Eildon Woods, Currently the object of timber production is primary in 83% of the area, and the remain­ Borders der is managed in the interests of shooting, forest recreation, amenity and shelter. For each of the objectives, the timing and location In the neighbourhood of Selkirk and Melrose of forestry operations are variable and impor­ there are two notable examples of conversion tant. A long-term working plan (using a 4- from even-aged to uneven-aged (or more cor­ year thinning cycle) is being developed, par­ rectly uneven tree size) forestry using the ticularly in those areas to be run on mixed group selection system. Conversion of mainly uneven-aged principles. The silvicultural sys­ broadleaved stands (oak, beech, sycamore, tems adopted must be sufficiently flexible to ash) surrounding the mansionhouses of His encompass storm damage and allow a mini­ Grace the Duke of Buccleuch at Bowhill and mum of management input. The systems to Eildon began in 1926 under the Duke of the be used will comprise group selection (80%), time, and continued under the 8th Duke, the single tree selection (10%) and in beech present Duke’s father, who died in 1973. stands irregular shelterwood (10%). Since 1956, under the head forester, Jack Considerations as to choice of system will Watson, there has been continual establish­ include rotation, profitability, amenity, land­ ment of groups of conifers and broadleaves in scape, wildlife, sporting, recreation, and the various sizes of clearings, usually on a 6-year need for sensitivity of management, all cycle. Groups of the upper storey trees were dependent on site conditions, the state of retained to accentuate the group selection existing stands and the need for sustained structure which now extends over more than yield and continuous forest cover. 300 ha and is the largest in Britain The Estate’s aim is for this diverse struc­ (Matthews, 1989). During the 1980s, the ture to amount to about 17% of the total for­ number and size of the groups were reduced; est area. The total aim for the extent of and the area of small group plantings spread diverse structure is about 300 ha (Bowhill into adjoining commercial conifer stands, with Eildon). Group size will be less than 1.0 usually following the windthrow of small ha. Planting will be done, but natural regen­ pockets of trees. Thinning (always light) eration of grand fir, Sitka spruce and declined in the 1970s, and did not increase in sycamore will be accepted. the 1980s. Along with this, there was a prac­ At both Bowhill and Eildon, stands of tice of retaining smaller trees without open­ beech will be managed on a variant of the ing out the higher canopy to release them, irregular shelterwood system, over a period of resulting in many drawn stems and mori­ 50 years, with the aim to continue the ameni­ bund trees. The woods became overstocked ty benefit to each mansionhouse. At Eildon, a partly through underthinning and there was stand of mixed uneven-aged conifers and little planting during the 1970s and 1980s; broadleaves is to be converted to the single diverse structure was developing but not as tree selection system, to include some plant­ planned. ing. Elsewhere the group selection system Following Jack Watson’s retirement in will be used. 1988 after 32 years’ distinguished service, Graham Booth was appointed as head forester and began a 3-year thinning cycle, Cavtfdor Big Wood, near Nairn which included the mixed stands tending to irregularity. Attempts were made to open out the woods, leaving trees of diverse sizes and Cawdor Big Wood, also called Cawdor Old species wherever possible. To date (1995) the Wood, lies immediately south of the Castle, diverse structure has been emphasised. owned by Lord Cawdor (the 7th Earl).

23 Probably the most famous oakwood in small areas were planted in 1920 and 1938 Scotland (Goss, 1992), it extends to some 303 (Cawdor, 1985). ha. It is predominantly pedunculate and ses­ The Big Wood has not been fully stocked sile oak (a few having been coppiced) but for many decades, so its potential for forestry there are some mixed stands of conifers. has not been fully used. The oak is ageing and These include superb stands of European of relatively poor form, influenced by the use larch, arguably of the best quality in Britain of pedunculate oak introduced by planting. (Plate 9). The Big Wood (Plate 10) lies The Big Wood is capable of growing good tim­ between, and straddles, the arms of an ber as shown particularly by its sessile oak, inverted ‘Y’ formed by the confluence of two Scots pine and European larch, hence its con­ burns flowing through rocky gorges of great tinuity is fully justified as well as for ameni­ scenic beauty and interest. ty and wildlife conservation. For many The main part of the Big Wood is sited on decades thought has been given to conversion Middle Old Red sandstone conglomerate of it under the group selection system. overlain with fluvio-glacial material which In 1967, Professor M. L. Anderson, refer­ has produced a deep freely drained podzolised ring to experiments within the private sector brown earth of moderate fertility. The eleva­ ‘concerned with the establishment of an irreg­ tion is 60-180 m with a northern aspect. The ular group structure’, noted that the Forestry region has the second coldest climate in Department in the University of Edinburgh Scotland and comes within the lowest rainfall in 1954-55 began such experiments in ‘the zone in Scotland with a mean annual figure of Old Wood at Cawdor belonging to Lord 768 mm. Frosts are recorded on more than 50 Cawdor [the 5th Earl] in old oak and birch days per year, with early June frosts being mixed with conifers’. The objectives of using not uncommon. The Big Wood experienced the group selection system were: gales of storm intensity in 1953, 1961 and • To bring all parts of the Wood into maxi­ February 1989, but extensive damage has not mum value production. occurred in the oak stands. The vegetation type is oak-birch-juniper association. There • To maintain the existing aesthetic and pro­ are several introduced exotic species. tective amenities so far as they are com­ patible with good management. Maps show the area as a wood as early as 1660 and parts may have been wooded since • To develop the Wood in such a way as to the period of general cover of the lowlands provide the maximum experimental and with natural sessile oakwood sometime after instructional value. glaciation (Goss, 1992). Parts were once used The Big Wood was divided into 48 compart­ for arable and pastoral agriculture, with ments, grouped into 6 blocks of about 50 ha patch farming in places. The planting was each, corresponding to a treatment cycle of 6 probably done by John (or ‘Jack’) Campbell years. The intention was to convert it during (1st Baron Cawdor, from 1768, died 1821). a period of 120 years by regenerating about Much of the original remnant was natural 2.5 ha each year in the current block using 25 sessile oak which had been depleted by selec­ irregularly spaced small groups. The staff of tive felling and renewed by planting pedun­ the Department of Forestry, University of culate oak, Scots pine and European larch, Edinburgh, assisted by students made the some 200 years ago. The present stands with­ annual 100% enumeration, marked the in the Big Wood are mainly from 1830-1850 fellings, selected planting groups and made (oak), with small areas planted between 1859 the choice of species. and 1865 (Norway spruce, Douglas fir, Scots Anderson’s planned 120-year conversion pine and European larch). In 1911, 4 ha of began in 1954-1955 in the old oak, with a Japanese larch, European larch, Norway group size of about 0.1 ha (Taylor, 1967). He spruce and Douglas fir were planted. Other used grand fir or western hemlock and beech

24 in the south part of groups; and Norway to reach the soil and some seedlings may be spruce, oak, sycamore and European larch in suppressed by the vegetation. The serious the north, the trees being closely planted in inhibiting factors are the wood-rush sward, ‘nests’ of species, about nine in each. Roe deer infrequency of acorn production and preda­ damage was severe as the groups were not tion by wood-pigeons, etc., insufficient light fenced. In 1962 W. A. Fairbairn and W. E. S. reaching the forest floor and browsing by roe Mutch of the staff of the University produced deer. a management plan for 1962-74. Anderson’s In 1984, when further experimenting to plan was changed because the ageing Big achieve a method of converting and perpetu­ Wood was suffering from windthrow, the ating the oakwood, a 0.4 ha coupe was felled regeneration groups were too small, the to admit light and fenced against roe deer. A planting spacings too close and the species in section was rotovated, giving rise to a dense a single group were too diversified. mass of birch and broom seedlings with only Accordingly, the conversion period was few oak interspersed; the remainder, not roto­ reduced to 60 years, the annual regeneration vated, contains a scattering of birch and some allotment doubled, the size of a group good established sapling oak. increased to 0.4 ha or even more if necessary, The Big Wood of Cawdor is now designated and the planting spacing set at 1.4 x 1.4 m as a Site of Special Scientific Interest. The but 1.8 x 1.8 m for Douglas fir. Windthrow Nature Conservancy Council’s preference was caused areas to be treated out of phase. to plan regeneration over a very long period In 1967 under a new management plan the allowing gaps to occur when over-mature Big Wood was divided into two parts for, specimens blew down. Lord Cawdor, however, mainly, timber production, but with an eighth felt that such regeneration would be difficult of the area managed for amenity. From 1970 to protect, and that parts of the ageing Big the ‘Anderson trial’ was discontinued Wood might well succumb before it was ade­ (Newton, 1986). The groups now require thin­ quately regenerated. Subsequently, current ning to relieve the young stands. negotiations with Scottish National Heritage Today, the Big Wood has a major shortage may favour the group selection system with of recently established and middle-aged trees. fencing to help achieve conversion and per­ However, under the old oak there are small petuation of this ancient oakwood. Although patches of naturally regenerated Norway wood-rush hinders natural regeneration, spruce, Douglas fir, western hemlock and some oak, as well as shade-tolerant conifers, Noble fir along with natural common juniper, survive; swiping or scarification may increase beech, oak, holly and a few aspen. The Big this. The head forester is Roddy Ross. Wood lacks an understorey because of the absence of fencing; intense roe deer browsing Use in Britain: conifers and is very evident. The great wood-rush (Luzula sylvatica mixed stands Gaud.) covers quite large areas. Its dense mat Research in Scotland on the application of the intercepts rain water, and uses a great deal of group selection system to conifers was initiat­ soil water, so is inimical to natural regenera­ ed during the 1950s by Professor M. L. tion of tree species. Its presence creates a Anderson (see Appendix 5) notably in his work need for screefing or scarification when plant­ on (a) ‘Anderson groups’ (Anderson, 1951) and ing is undertaken. Being a light-demander, (b) conversions (transformations) from 1953 the wood-rush disappears following the effec­ (Anderson, 1960). He believed that forestry of tive closing of the new crop tree canopy. Also mixed species and ages when regenerated in it has been found that wood-rush does not small groups would provide and maintain eco­ wholly inhibit regeneration of oak if there are logical conditions necessary for sustained plenty of acorns; however, acorns usually fail yield, with some protection from wind and

25 other disturbance factors. This approach Glentress Forest, near Peebles seemed especially suitable on higher eleva­ tions where diversity would inevitably devel­ op in even-aged plantations through endemic windthrow, disease and site variability. He Glentress Forest lies in the uplands of south- wished to create some diverse structured for­ central Scotland. A trial of conversion of 30- est in Britain and thereby gain experience in year-old conifer stands from uniform even- its management. aged clear cutting system to uneven-aged Also in Scotland, Fairbairn (1963), follow­ group selection was begun in 1952. The ing up Anderson’s practical work on groups, experimental area was established by an clearings or fellings, made a study of wind, informal agreement between Professor M. L. aspect, insolation, incident rain and temper­ Anderson, on behalf of the Forestry ature at different points within coniferous Department of the University of Edinburgh, groups; and made recommendations for the and Sir H. Beresford Peirse, Director of the distribution of planted species within the Forestry Commission in Scotland. The aim group. In 1966, McNeill and Thompson was to create a mixed forest of irregular (1982) began to monitor the occurrence of structure similar to the beech-spruce-fir pro­ Sitka spruce seedlings in the smallest size of tection forests of the Swiss Jura, converting clearing (0.04 ha) at Forest of Ae and their the existing even-aged stands over a period of survival in relation to their position in the 60 years. clearings and the vegetation competing with The total area of the Glentress Trial is 117 them. Low (1985) reported the occurrence of ha, the altitudinal range 240-560 m, the natural regeneration of Sitka spruce in west annual precipitation 1000-1500 mm, the gen­ and south-west Scotland, but considered that eral aspect south-west, and the windthrow it did not provide a practical alternative to hazard classes 3-4 (exposure to wind is accen­ planting because seed years are not regular tuated by funnelling up valleys). At the start and the seedbearers are prone to windthrow. of the trial the species were Douglas fir on the During the 1930s and 1950s on frosty sites in lower ground, Japanese and European larch Scotland and Wales, use was made of oak and on the middle slopes and Scots pine on the birch as an overwood to establish Douglas fir, upper slopes; some stands had been partially western red cedar, Norway spruce and grand beaten-up or replaced with Sitka and Norway fir. Through a period from 1964 to 1970 spruce in the 1940s. There were remnants of (Neustein, 1970) studies were made on old high altitude shelterbelts of European underplanting for conversion of stands of larch and Scots pine; and some 1920 Sitka Scots pine and larches to Douglas fir, western spruce and Douglas fir below 300 m. hemlock, grand fir and spruces, using various Anderson’s main objectives were: densities of overwood and strip felling com­ • To create mixed stands of diverse structure pared with clear cut plots. There exists a con­ Taest adapted to the natural conditions’ in siderable amount of unpublished material order to maintain and increase soil fertility. from Neustein’s series of experiments on con­ • To maximise value production by achieving version by underplanting Scots pine and the highest possible volume increment and larch in sandy sites and other more fertile timber quality. sites bearing larch, using Douglas fir, Norway spruce and species of silver fir • To determine the optimum growing stock (Paterson, 1990). composition, avoiding the use of extensive clear cutting, for the provision of sustained In upland Scotland, five examples of con- yield. vesion towards group selection in conifers — Glentress Forest, Corrour, Drumlanrig, Faskally • To obtain experience in the management of and Benevean — are particularly notable. diverse structure stands.

26 • To improve internal access to ensure satis­ • To provide information for education and factory harvesting. research relating to the management of Anderson was particularly seeking diverse diverse structure forest. structure and mixed species, stability to From 1964, changes have occurred in silvicul­ wind, and sustained yield; no apparent tural and management practices. Annual enu­ emphasis was placed on amenity, landscape, merations were discontinued (because the wildlife conservation and sporting/game stands were far from achieving a diverse struc­ management. He appears to have underesti­ ture) and replaced by periodic sampling (the mated the cost of controlling browse damage latest sample enumeration was done in 1991). (after more than 40 years, some Norway Group size was increased to reduce shading spruce and beech are still below 1 m in from the surrounding trees, with larger groups height). at lower elevations. Groups are made roughly The silvicultural system used was group circular with an effective diameter at least selection: groups of 20 x 10 m or smaller were twice the surrounding stand height, giving a located on a regular basis with the long axis minimum group size of about 0.1 ha. Choice of oriented north-south. Groups in the existing species has changed: beech and European sil­ stands were felled and replanted using ver fir were excluded because of deer browsing; shade-bearing species at the southern end of Sitka spruce, Noble fir, western red cedar and each group. Plant spacing was 0.9 m to min­ hybrid larch were introduced. There will be lit­ imise beating-up and weeding. A total of tle further planting of broadleaves because of approximately 2 ha annually were regenerat­ adequate natural regeneration of sycamore, ed in the treatment block (total area/conver­ birch and rowan. The present aim is to estab­ sion period). Species planted initially were lish, over the whole trial, approximately 50% Norway spruce, European silver fir, beech, Sitka spruce, 40% other conifers and 10% grand fir, western hemlock and broadleaved broadleaved species by area. Natural regener­ species including sessile oak, ash, sycamore, ation of conifer and broadleaved species is alder, hazel, rowan and whitebeam. occurring on the upper slopes. Several areas of The ‘check’ method of yield control was young growth have been respaced to assist in used: the area was divided into six blocks achieving the desired species diversity. Losses averaging 19.5 ha each and worked on a 6- from windthrow have been very light. Yield is year cycle with all operations, including enu­ obtained primarily from the 2 ha annual cut, meration, done in the year of treatment. together with periodic thinning of those stands Edinburgh University were and are respon­ in mid-rotation. As to economics, it is estimat­ sible for complete enumerations, for mark­ ed that establishment costs may be up to 50% ing thinnings, selecting the regeneration higher and harvesting costs 10-20% higher sites, and species. Forest Enterprise under­ than Forestry Commission standard rates take harvesting, site preparation, planting, (Shrimpton, 1986, 1988). During conversion tending and protection. Under recent man­ there is a cost due to production and revenue agement (1964 to present) the current objec­ foregone; and there is some loss of quality in tives are (Whitney Mclver et al., 1992): the edge trees of groups which develop large side branches. • To achieve mixed species stands of diverse The Glentress Trial is fortunate in being structure in order to maintain permanent located mainly on freely draining brown earth forest conditions throughout the area. and ironpan podzol, where windthrow is not a • To achieve sustained production of high great problem, and in starting with mid-rota­ quality sawlogs of 40-65 cm dbh, depend­ tion conifers which provide scope for develop­ ing on site type. ing diverse structure. Natural regeneration • To enhance recreation, amenity and con­ has been helpful, and is now tending to servation values. replace planting on all sites except those with

27 dense soft grasses ( Holcus) on moist gley soils. structure was associated with change of The early experience of sheep grazing and species, checked growth and site quality. deer browsing emphasises that strict control Each year one block was completely enumer­ of these animals is fundamental to such group ated and trees were marked for felling. The selection working. Only after about half the equivalent of heavy crown thinnings was conversion period does the diverse structure made to break up the stands into groups of become apparent (Fairbairn, 1963; Blyth different diameter classes and all these 1986; Blyth and Malcolm, 1988; Shrimpton, fellings were completed by the end of the sec­ 1986, 1988; Whitney Mclver, 1991; Whitney ond cycle of inventory in 1964. Mclver et al., 1992; Malcolm, 1992). In 1965 the Forestry Commission acquired The Glentress Trial is an excellent educa­ the woodlands, the enumerations continued tion resource; it has provided a wide range of and the third cycle was completed, so that research topics. Other benefits are the high data from 19 years were available. Malcolm standard of amenity, nature conservation and (1971) reported that the current annual incre­ forest recreation values (Plate 11). In manag­ ment of the trial was 10.7 m3 ha'1, so that a ing this important trial, the Forestry mean annual increment of 7.1 m3 ha'1 could be Department of the University of Edinburgh assumed; this was reasonable for a high ele­ — particularly in recent years through vation site. When fully converted, the trial for­ Douglas Malcolm, John Blyth and Helen est was expected to yield about 45 m3 ha'1 at Whitney Mclver — has played a leading role each felling on a 6-year cycle. Trees of about in demonstrating conversion to diverse struc­ 50 cm diameter at breast height were expect­ ture forestry in Scotland. ed in 80 years. Regeneration of Sitka spruce was appearing and the size of clearings was Corrour Forest, Inverness-shire 0.04 to 0.2 ha, with the larger sizes preferred. This example of conversion to group selec­ tion system would have had particular rele­ Corrour Forest lies in wet uplands in the vance to the type of stand structure which may West Highlands of Scotland. Sir John Stirling be necessary for the protection of forest fringes Maxwell began the afforestation in 1892 at higher elevation. Unfortunately, work on the using Scots pine; but planting of mixed west groups ceased owing to changes of ownership North American conifers followed (Anderson, and distance from research centres (Paterson, 1967). The forest lies at 380 to 460 m; the 1990). The groups suffered from inadequate moorland soils over granite are of low to mod­ fencing against deer. A diverse stand structure erate fertility. M. L. Anderson in 1952 began had been initiated in compartment 3 (Plate 12). a trial of conversion to group selection forest, In 1983, following the State’s disposal, much of using the Methode du Controle or ‘Check’ the mature timber was clear felled. Some Method to determine increment and control replanting took place but the trial terminated yield. The trial was located on the better soils as far as inventory control is concerned. The south of Loch Ossian. present owner is Donald Maxwell Macdonald The main object was to create a balanced and the forest manager is Felix Karthaus. growing stock of exotic species capable of con­ tinuous production. Recurrent inventories Drumlanrig, near Thornhill, began in 1952 and the area to be included in the trial was finalised in 1958 as 54 ha in six Dumfriesshire blocks. In 1958 the species composing the stands were Norway and Sitka spruce (73%), grand and noble firs (7%) and the remainder Drumlanrig, the estate of His Grace the Duke Scots pine, European larch, birch and other of Buccleuch, has examples of group selection broadleaves. Much of the initial variance in in Bogrie Wood and Mallyford extending to

28 some 173 ha on rising ground and important there are even-aged areas which were origi­ to the landscape setting of the Castle. The nally planted for shelter, amenity and sport­ woodland manager is Bob Jordan. ing, and therefore will require to be regener­ Between the years 1900 and 1949, sporadic ated on a group selection system aiming for planting into old broadleaved and mixed continuous forest cover. stands extended to some 11 ha, mainly of European larch, hybrid larch, Norway spruce and Douglas fir. Between 1953 and 1973 Faskally Wood, Perthshire group felling and replanting took place annu­ ally, totalling 108 ha, the main species being Japanese larch, Norway spruce, Sitka spruce and Douglas fir, "with small quantities of North-west of Pitlochry, Faskally Wood, con­ Scots pine, European larch, western hemlock, sisting of mixed conifers and broadleaves, oak and beech. In the early part of this peri­ was originally planted as policy woods for a od the coupes were small, following M. L. Victorian mansion, now Faskally House. The Anderson’s recommendations but later soils are mainly freely drained fertile coupes were larger, up to 0.5 ha each. Where morainic. In 1947 the estate was purchased large broadleaves, Scots pine or Wellingtonia by the North of Scotland Hydro-Electric stood in the coupe area they were retained. Board; and for several years the House was These plantings are now overdue for thinning used as a staff college, and Faskally Wood but do present marketing problems due to the formed part of a public relations campaign to diversity of species and size class, giving only convince the public that the Board was sensi­ small quantities of any one product. tive to environmental concerns (Newton, Planting during the period 1974 to 1992 1986). amounted to some 17 ha in the three years In 1953 the Forestry Commission acquired 1977, 1979 and 1988, using only Scots pine ownership, under the proviso that no clear and Sitka spruce. Current policy is to replant cutting would occur. From 1955 until 1969- following windthrow rather than create new 70, Faskally House was the site of one of the coupes on a planned basis. Browsing by deer Foresters’ Training Schools. The manage­ is the main disincentive to fully pursuing the ment of Faskally Wood from 1953 was within group selection system. a plan designed by M. L. Anderson of the Drumlanrig illustrates the value of group Department of Forestry, University of selection in a sensitive landscape. It also Edinburgh. With the co-operation of Frank highlights continuity of management, felling Oliver of the Forestry Commission’s Aberdeen and relief thinning around groups, replanting office, he planned a group selection system in and appropriate sized coupes as vital for suc­ the long term and suggested detailed pre­ cess; but overriding these factors is the need scriptions for the conversion. The group sizes for practical and economic control of browsing ranged from 0.02-0.05 ha, later increased to and bark-stripping mammals. 0.05-0.10 ha, the whole to be converted over a These woods are commercial and are 120-year period. worked within the whole management plan of Anderson recommended in 1953 that about the estate. Elsewhere, the planned rotation is 75 ha for ‘irregular working’ should be sub­ 60 years for the even-aged clear fell and divided into blocks on a 6-year cycle, one replant system but in the Drumlanrig policy block to be dealt with each year. Groups in woods a diverse scene was quickly created general were to be ‘dispersed over each com­ through the use of the group selection system, partment’. Where natural regeneration was to be improved in age and species over a hoped for but did not appear in 3 years, plant­ longer period. Throughout the estate, this ing was to be undertaken; birch regeneration type of treatment is to be continued where was to be made use of for nursing. Two years

29 later, in 1955, Anderson confirmed to H. 2. To provide adequate essential visitor facil­ Cruickshank in charge of the School that the ities to meet present and future recre­ intention was that the groups of small ational demand. stands among large blocks of even-aged 3. To conserve the environment in Faskally Scots pine, larch, Douglas fir and Norway Wood primarily as a means of enhancing spruce should consist of shade-bearers and its amenity value and visitor enjoyment. semi-shade-bearers; that ‘it would be a mis­ take to make the openings and groups too 4. To provide high quality timber as effi­ large’; and that ‘a small proportion of larch ciently as the above-mentioned objectives and Scots pine may be desirable and a good permit. deal of oak [in the oak stands] which may Disposals have reduced the net management require larger openings’. No group should be area of Faskally Wood to about 51 ha. The less than 0.02-0.05 ha. Anything that would plan was only partially followed and little make Faskally Wood more uniform should new planting undertaken. In 1987 a new be avoided. The system of group selection in Forestry Commission plan was formulated mixed conifers and broadleaves, primarily and adopted. Recommended group size was for amenity and recreation as well as for 0.05-0.50 ha; abundant regeneration of continuous forest cover, was applied sensi­ Douglas fir, sycamore and beech was avail­ tively to the environment. able; the dominant species included Scots Today after 40 years the original pine, European larch, Douglas fir, oak, birch ‘Anderson groups’ are sometimes difficult to and beech. A proportion of the older trees recognise. They have become tall, cramped were felled as mature and a proportion of and require relief by thinning; and in gener­ younger trees are removed periodically. al have proved to be too small and too Replanting has been of small areas where detailed in species layout. However, they are there is insufficient natural regeneration. growing well and will be improved by thin­ Today, there are about 23 different species of ning and freeing of their edges. They still trees, a pleasing mosaic of conifers and enhance the interest and productivity of broadleaves with a great variety of ages and Faskally Wood, and diversify the structure sizes including many fine examples of (Plate 13). A detailed stand inventory was Douglas fir and silver firs, some over 100 made in each working block in turn by the years old, and wild cherry, oak, birch, rowan, staff and students of the Forestry School, lime, elm, ash, sycamore, beech, alder and but ceased when it closed. poplar. The whole is a splendid area for In 1971 Faskally House was bought by amenity and recreation, sensitively man­ Glasgow Corporation, and in 1973 opened as aged to complement timber production. a Schools’ Outdoor Centre. In the 1970s Current Forest Enterprise intentions for the Faskally Wood was incorporated in the Wood are: Forestry Commission’s Tummel Forest • To continue to sensitively manage for administration. It was the first wood in East amenity and recreation purposes, includ­ Scotland Conservancy to have timber pro­ ing semi-permanent or continuous forest duction ranked lower in objectives than cover. recreation and diversity for internal land­ • To thin the ‘Anderson groups’ (horse scaping of high quality. In 1977 a new man­ extraction may be used). agement plan was prepared for the period 1977-86, emphasising four objectives: • To create some additional regeneration groups. 1. To provide and enhance the structure and species diversity of the area, creating a • To sensitively remove by selection some of recreational wood of high amenity. the fine mature Scots pine and larch.

30 Benevean Wood, Inverness-shire Cefn Llwyd Forest, Dyfed

Owned by the Forestry Commission, Situated south-east of Bala, the 439 ha Cefn Benevean Wood in Glen Affric is an example Llywd Forest has been owned since 1986 by of group or random regeneration of indige­ Shotton Paper Company pic and managed by nous Scots pine (Plate, 14). Stands had arisen Talis Kalnars. (The Forest Manager since from extensive natural regeneration follow­ January 1993 is Philippe Morgan.) It is ing exploitation of all mid-sized and large mainly of uniform even-aged conifer planted trees of good quality around 1770. An area of in 1952-63 with 100 ha in 1988-91. The site about 850 ha was brought into a ring deer types are 247 ha upland with peaty gleys and fence in 1959-60 and later subdivided into ironpan soils, 147 ha lowland (brown earth). three sections to improve deer control. An The basic geology is Silurian and the annual annual programme of about 12 ha in six sep­ rainfall 1500 mm. The main species is Sitka arate felled 2 ha groups was initiated from spruce (63%) with some Douglas fir, larch and 1960 on the premise that mean age was Scots pine, and about 1% broadleaves. The already 140-180 years and few trees could be management aims are: expected to remain healthy beyond another 1. To maximise the return on the financial 70 years. Therefore, the regeneration period investment by the production of the high­ envisaged was 70 years. Many trees were est volume of quality timber. - already thin in the crown and being killed by Trametes pini crown and stem rot and 2. To convert from uniform even-aged planta­ Stereum butt rot. The 2 ha groups were cen­ tion to uneven-aged mixed species forest. tred on the more open sickly parts of these 3. To normalise production, cash flow, and old stands. Artificial ground preparation was aesthetic and environmental benefits by used: normally turfing in deep Sphagnum; the practice of sustainable forestry and ploughing was attempted but Calluna and achievement of continuous forest cover. mosses fouled the ploughs. Later, Affric Scots pine seed was collected and plants from this Since 1986 all compartments have been seed were used. In fact, a lot of regeneration thinned on a graduated density pattern that emerged generally from the Calluna after aims to increase production of high quality improved protection from red deer. This timber and to introduce irregularity into group felling was continued for only 5 to 10 stands. The larger part of the Forest is man­ years. Thereafter, random regeneration was aged under a progressive small coupe felling as accepted although it appears to be less part of the 20-year restructure plan begun in dependable where there is no extraction 1993. The remainder is managed under other skidding in harvesting to expose a seedbed. more specialised systems, each with a degree Occasional windthrow is presenting new of irregularity. The progressive small clear openings and soil disturbance for regenera­ cuts will be limited to 0.5-1.0 ha, opened on tion. windfirm edges, and advancing progressively towards the prevailing wind. Topographical variation and physical division by tracks and In upland Wales, the two examples chosen roads allow felling of coupes to start from of conversion of conifer plantations to group numerous points to accommodate the annual selection are: Cefn Llwyd Forest and Coed y cut determined by the length of the conversion Brenin. Two further examples (not described period which itself is constrained by top height here) are Diosgydd (Plate 15) and Nant BH and windthrow risk. Small coupes maintain (Plate 16), both in Gwynedd. the forest environment and assist in the man­ agement of natural regeneration, which is the

31 preferred means of restocking. Planting or exploitable diameter of 42 cm is set so that direct seeding after cultivation are options only trees that have reached their where regeneration fails or is constrained by exploitable diameter are selectively felled the terms of the field sporting agreement. creating gaps in which regeneration is Site type and species productivity deter­ accepted. Where two or more adjacent trees mine the choice of species. A particular aim is are felled together, the system tends towards to increase the area under Douglas fir in the group selection and allows retention of light- lower sites. With natural regeneration, species demanding species. choice and composition is made during The aim of conversion is to achieve normal­ respacing to about 2 m x 2 m, when seedlings ity. Several restructuring phases will be are about 1 m tall. Choice and proportion are required to achieve this. The first plan spans varied to suit site requirements, silvicultural 20 years. A second plan (again first rotation) system employed, amenity, and water guide­ will restructure the recently planted areas. lines. An element of broadleaves will be includ­ Further restructuring will be required beyond ed to satisfy conditions of the Woodland Grant this point before the Forest produces sustain­ Scheme. Current species composition is typi­ able supplies of timber and non-wood benefits. cally Sitka spruce (90%), Japanese larch, Scots Continuous production will help to ensure pine and broadleaves (10%). sustained supplies of logs for sawmilling, and Some compartments are managed under a smaller produce for pulpwood, chipboard and variety of irregular systems that have been fencing. The Forest as a whole will be mixed chosen to suit site requirements and amenity. in species. The stands managed under the They are single tree, group selection, strip small coupe felling system will be generally and irregular shelterwood systems. The sys­ even-aged. Stands managed under the more tem combines features of different systems irregular systems will have a greater range of and is best described as ‘a strip and group age classes with a tendency to normalise system’. Felling is done on a strip of width within compartments. The effect will be vir­ twice top height. Use is made of advance tually continuous forest cover. regeneration (typically Sitka spruce, Japanese larch, Scots pine, birch and rowan). The aim in these high production areas is a Coed y Brenin, North Wales species mix of Sitka spruce, Japanese larch, Scots pine and broadleaves. Strips are felled progressively against the prevailing wind, Coed y Brenin lies within Forest Enterprise’s either on a 6-year cycle or when natural Dolgellau Forest District (c. 17 000 ha) in regeneration enters the thicket stage. North Wales. Site conditions are generally Thinning will take place in adjacent strips; very favourable: highly productive brown 80% of the new strip will be retained as earths on sheltered sites dominate the mid groups and shelterwood and the retained por­ and lower slopes. Gradual introduction of tion of the previously felled strip removed, alternative silvicultural systems is being providing access through the regeneration. A made to some sites throughout the District. portion of the original retained areas may be Something approaching 1% might be suitable reserved at each stage depending on the con­ for conversion during the next decade. Much straints of windthrow or the requirements of consideration is being given to achieving yield control and markets. some biological diversity. The aim is to One compartment is managed under a sin­ achieve both structural and species diversity, gle tree or group selection system. Two mainly by the group selection system, paying preparatory thinning operations were done special regard to aesthetic and environmental in which selection was across the dbh range benefits, and matching species to site condi­ and not to favour large trees. A minimum tions. At the same time, timber production is

32 to be sustained. An important principle to be extending over the last 30 years. Ninety followed is what is considered best for the hectares of well-stocked woodlands, managed future of the Forest and the set objectives, not by Talis Kalnars, contain a wide variety of least achieving environmental objectives, as both conifer (80%) and broadleaved species well as meeting timber production require­ (20%), the preferred species being Douglas fir ments. Appropriate thought is given as to the and ash. They are run on the group selection opportunity cost of conversion (T. Owen, per­ system with continuous cover being main­ sonal communication). tained, as far as possible, by the annual use of Particular experiments and actions are group fellings and selection fellings followed aimed at the achievement of virtually continu­ by planting. The harvesting is largely orient­ ous forest cover by (a) converting some uni­ ed to exploitable diameter and is not related form, even-aged mid-rotation or maturing to age. There are regular checks, using cal­ conifers into uneven-aged diverse structure by lipers and counters, every 5 years, to assess the group selection system, and (b) recruiting the growing stock, thereby monitoring the natural regeneration into maturing or mature production levels, the profitability, and the high quality conifer stands (especially Douglas gradual increase in standing volume and fir and Japanese larch which will have extend­ value. The woodlands are of high amenity ed rotation age) in areas requiring sensitive value. The group selection areas as well as management for reasons of landscape, ameni­ natural regeneration are particularly impres­ ty, wildlife habitat and recreational potential. sive. Wildlife habitats are being improved as At the same time, timber production will meet well as soil structure and fertility.- ~ set objectives. Achievement of the foregoing will be sought mainly by manipulation of conifer stands by Clocaenog, North-east Wales careful selective removal of individual large stems (significantly Douglas fir) or by felling small coupes (0.1 to 0.2 ha), restocking if pos­ The main block lies within Forest sible by natural regeneration or by planting, Enterprise’s Clwyd Forest District (c. 9250 or, experimentally, by new methods of direct ha). Soil conditions on the wet and exposed seeding. Another aim is to perpetuate part of Hiraethog plateau tend to favour Sitka the forest cover by changing the thinning spruce, although larch and Norway spruce regime to retain mixed size classes, including grow on the lower southern levels. In recent small stems. Wide choices are intended, all years as clear cutting of the older spruce has expected to be of help in future management. evolved, opportunities have arisen to plan the second generation woodlands. This planning Additionally in Wales, there are two notable takes account of environmental principles, examples of diverse structure forestry, in part embracing landscape, conservation, water, using variants of group selection and uniform public access and recreation, and reinforces shelterwood. the concept of multi-purpose forestry (P. Hughes, personal communication). Natural regeneration of Sitka spruce is Ffrwdgrech Woodlands, South prolific on most clear fell sites in Clocaenog, and it will be the main species in the Wales restocking process. Larch and pine also regenerate on some areas, and by careful management, an acceptable mixture of Owned by Major W. D. D. Evans, the conifer species can be achieved. Further Ffrwdgrech Woodlands near Brecon comprise diversification is realised by the introduc­ a fine example of continuous forest cover tion of broadleaved species on all sites. Even

33 on those areas where Sitka spruce regenera­ of rampant weed growth (particularly bram­ tion dominates, it is possible to find space ble) which smothered the seedlings, and that and the opportunity to plant larch, Norway the small size of the groups would make their spruce and pine. Some long-term retention, tending and thinning too expensive. the combination of overstorey with an Hiley had used the Methode du Controle understorey crop, also contribute to the (Hiley, 1967), considering this to be the most diverse structure of the forest block. practicable, if rather laborious, means of reg­ Producing and maintaining a programmed ulating yield. It required intensive measure­ supply of timber is one of the main objectives ment and management and took much time to and the main source of revenue. Clear fell achieve reasonable accuracy since it involved coupes are planned to achieve this pro­ the periodic measurement (say every 5 or 10 gramme but the process is arranged so that years) of all the trees above a certain mini­ restocked crops reach heights of 2 m or more mum size. By a comparison of the results of a before an adjacent coupe is felled. This in progression of periodic measurements called itself ensures sound diversity, and in conjunc­ recurrent inventory, it is possible to achieve tion with the species mixture results in the reliable estimates of the growing stock vol­ gradual change from even-aged stands to for­ ume, the distribution of size classes, the incre­ est of mixed age and species. ment and hence the periodic cut of timber and its distribution among the size classes and In the lowlands of England, adaptations or species (Paterson, 1958; Howell et al., 1983). variants of group selection in conifers are In spite of Hilej^s rejection of his system, a demonstrated in the following three examples, part of compartments 1 and 2 at North Wood, two in Devon, one in Wiltshire. which includes his original sample plot, has continued to be treated in a way consistent with group selection principles. As a result, Dartington, Devon this is a particularly attractive part of the woodlands, much frequented by the public. Subsequently, it was decided to maintain the character of this area with an attempt to W. E. Hiley in 1943 tried a version of the operate the system (Howell et al., 1983). The group selection system (Hiley, 1953, 1964, original experimental plot withstood the 1967). In a stand of Douglas fir, Japanese gales of 1989-90 whereas an even-aged stand larch, Norway spruce and Sitka spruce of var­ of coast redwood nearby was flattened. The ied ages from 22 to 46 years, gaps were creat­ main problems Hiley foresaw in applying the ed by heavy thinning to 370 stems ha'1, and system concerned economics, extraction costs, planted with groups of Douglas fir, Japanese and the possible difficulties in using it to larch and sweet chestnut. During the 1950s, grow his most profitable species, Douglas fir further group planting took place, mainly of and Japanese larch, because of their limited Douglas fir and western red cedar. Hiley was shade tolerance after early youth. particularly pleased with the good growth shown by the isolated overstorey trees, and there was very little wind damage. However, by the early 1960s he appeared to have reject­ Longleat, Wiltshire ed any further application of the system on sil­ vicultural and management grounds (Howell et al., 1983). Although he considered that This example is on the estate of Lord Bath, on enough had been done to show that Douglas freely drained podzolised brown earth fir and Japanese larch could be grown under a derived from Greensand with light weed form of group selection, he felt that natural growth on low ridges. John McHardy, the for­ regeneration could not be relied upon because est manager, has applied to conifers a shel-

34 terwood-cum-group selection system with divided into nine square ‘plots’ measuring 6 m evolving stand structure on some 300 ha x 6 m. Plots are intended to contain one final (Hart, 1991). It is, in fact, conversion to stand tree, and are initially planted with nine groups from even-aged plantations of Scots shade-bearing transplants consisting of pine, Douglas fir and larch. A 10-year treat­ conifers apart from a few Nothofagus species. ment cycle is used, and stands are kept very The existing stand within a unit is felled and open to accommodate such light-demanding replanted at 6-year intervals, so each unit species as larch and pines (Plate 17). The sys­ eventually will contain plots of all ages. The tem has partly arisen as a side-product of the uneven-aged units taken together as a stand very heavy crown thinnings applied to all the are considered as the basic group, and the Plan 20 to 30-year-old larch, Scots pine and qualifies as a group selection system. The con­ Douglas fir plantations. These thinnings have tinuous canopy is particularly important. Part encouraged development of fairly plentiful of the Tavistock Woodland Estate has recently natural regeneration of those species which been sold and management is now fragmented are focused into groups by group fellings. The but continuation of the Plan, which is half regeneration is respaced with a clearing saw completed, is likely. when 1.5-2 m in height. Apart from applica­ tions of ‘Asulox’ to control bracken every few years in the regeneration groups, this is the only tending undertaken. The aim is to estab­ Conclusions on the group lish and respace a new crop of regeneration selection system every decade, although fellings or thinnings of the overstorey are generally every 5 years. The conditions of climate, site and vegetation Stocking in the older stands at present is being remarkably varied in Britain, it is not light and there is some sacrifice of volume surprising that experience gained from the production; however, as more age or size use of the group selection system has also classes become established, canopy should been varied. Sometimes it has been difficult become more complete, and weed growth to classify the examples because distinctions reduced. The low stocking compared with between variants and adaptations of systems Continental selection stands means that high have become blurred. But it is important to pruning is necessary on final stand trees but use recognisable labels to establish norms. diameter increment is outstandingly high, Nevertheless, individual stands studied for and the expected age of exploitation is 40-50 this Bulletin, together with many relevant years. discussions with their managers, reveal much common ground and this is now identi­ fied. Tavistock Woodland Estate, 1. The objectives o f conversion. As might be South Devon expected, almost every owner or forester gave the production and sustained yield of timber as the prime aim but several empha­ sised quality of timber, especially in The Bradford-Hutt Plan pursued on the broadleaves, in addition to quantity of tim­ Tavistock Woodland Estate has been the ber. The statements of other objectives usu­ most intensive organised conversion to the ally gave as important the creation of con­ group selection system recently evolved and tinuous forest cover; diversity in species, age practised in Britain (Hutt and Watkins, and canopy structure; amenity; improved 1971; Hart, 1991). In brief, each stand is habitat for wildlife; facilities for field sports; divided geometrically into square ‘units’ and access for recreation, the emphasis (18 m x 18 m) and each of these units is further given varying with circumstances.

35 2. Choice of sites. It is notable (see Table 3.3) lowing heavy thinnings. Areas of ‘advance that most owners and their foresters have growth’ are encouraged to expand by further played safe and selected relatively sheltered removals of the overstorey, and this is sys- sites in favourable wind zones (Quine and tematised into a plan for group selection. White, 1993) which should allow the stands to have a low risk of strong gales, but does not 5 .Silvicultural prescriptions. When the starting rule out endemic wind damage; the sites usu­ point has been selectively exploited woodland ally have soils that can be penetrated by tree or ‘derelict’ woodland, the bare land not roots to a good depth. Soils that are wet and already stocked with trees has been regenerat­ cannot be deeply rooted have been avoided ed by planting groups of one or sometimes two and the deeper fertile loams — brown earths, species. When the starting point has been soils with sandy texture and podzols — have maturing, mature or over-mature woodland, been preferred. Natural regeneration is often ‘advance growth’ has been augmented by easier to obtain on such soils. The techniques planting with the same or different species. commonly used in the group selection system Experience suggests that it is a mistake to use have not been tested on more exposed sites, several species in underplanted groups or to with the possible exceptions of Dartington in plant them too densely (normal planting dis­ Devon and Corrour Forest in Inverness-shire. tances are adequate). However, ‘advance growth’ cannot always be expected to recover 3. Species. Existing broadleaved species to be when light is admitted by thinning and felling continued in the converted wood or forest the overstorey which is sometimes the case include oak, ash, wild cherry, sycamore, lime with beech. Much depends on species and the and beech; and Scots pine, larches, Douglas condition of the regeneration at time of fir, Norway spruce and Sitka spruce among release. conifers. Species that have been introduced in the course of conversion are often shade- 6. Size of group. Almost all owners and foresters tolerant conifers including grand fir, Noble contacted for this Bulletin began with small fir, western hemlock and western red cedar. groups of less than 0.2 ha but as they gained Beech has proved difficult to introduce in experience the size of group used has enlarged. conifer forest owing to deer browse. Some consider that Britain’s summers are 4. Age of stand at the start of conversion. more cloudy than those common on the Conversion has followed two main routes. Continent so there is less light at ground level; Either ‘derelict woodland’ of varying age and others wish to retain light-demanding species condition has been cleared of poor quality in the new crop; others again wish to ensure stems, while stands and trees of good that the groups created will survive without growth and quality have been favoured, excessive treatment of the overstorey and sur­ tended and augmented; or clearings have round. Typical sizes of groups are now 0.3 to been made in existing productive stands and 1.0 ha, which can be treated as units for tend­ regenerated naturally or by planting. The ing, harvesting and yield control (see Tables former route is common in broadleaved 3.1 and 3.2). stands, while the latter route has been fol­ In all group adaptations, the size of lowed in both broadleaved and conifer groups, their shape, orientation, distribu­ stands. When existing productive stands are tion and protection have been critical ele­ to be converted, the process has begun at all ments in prescriptions; size is particularly ages from 30 to 100 years or even more. important. Natural regeneration of large- However, most foresters agree that conver­ seeded species (notably oak and beech) may sion should begin earlier (at 30 to 50 years) not spread to the centre of large groups. On rather than later (at 80 years or more). A the other hand, Brown (1953) commented on common stimulus to conversion has been the ‘pudding-shaped profile’ of beech fre­ the appearance of natural regeneration fol­ quently found in groups, with taller regen­

36 eration in the centre of the group; this is typ­ Because the group selection forest develops ical of groups that are too small or have not as a mosaic of groups of different species, been enlarged when regeneration has ages and canopies, some planned skeleton appeared. Larger groups are needed in taller layout of extraction routes is desirable. and more uneven-aged stands as well as for Much depends on the existing road system full and partial light-demanders. The limita­ and how it is to be used. tions imposed by natural regeneration do not 8. The treatment cycle. Continental foresters apply in the same degree where planting is consider that a cycle of 6 years is about used to augment the stocking, and this will right, so they have divided the whole area to normally be the case in most situations. be treated by the group selection system Scrutiny of European literature and dis­ into six roughly equal ‘blocks’ or areas. Each cussions with foresters in Britain on group block is treated at intervals of 6 years, cuttings reveals wide divergence of opinion cleaning and respacing stands that have about suitable dimensions. The size of clear­ reached the thicket stage, thinning, occa­ ings depends largely on features of the cli­ sionally high pruning, and clear felling fol­ mate, soil, final tree height at harvesting, lowed by regeneration. One forester would and species. Suitable dimensions must be emphasise the importance of thorough worked out for each location using experi­ cleaning of young stands, another would rec­ ence gained elsewhere as a starting point. ommend delaying cleaning until first thin­ One working rule emerging from the range ning, and yet another would prefer ^election of opinion is that the first sizes of clearings and favouring of the best saplings only, a chosen are commonly too small and must often be doubled usually by thinning the form of respacing. There seems to be ample overstorey (Matthews, 1986). Table 3.1 is room for inventiveness on the part of foresters. The areas being treated by the intended as a guide to assessing possible group selection system range from 6 to 300 sizes of clearings sufficient to provide grow­ ing space for final crop trees of Douglas fir ha, with many less than 100 ha. In Britain a (Hutt and Watkins, 1971; Bradford, 1981), large proportion of private woodlands fall ash and sycamore (Garfitt, 1984) and oak within the range of 50 to 100 ha and it (Jobling and Pearce, 1977). seems probable that the group selection sys­ Matthews (1986, 1989) gives a table of tem will suit this extent, where sites are ‘dimensions of clearings to provide growing suitable. space for trees of final size’ (see Table 3.2). 9. Protection. Foresters are adamant that deer, Groups in broadleaves need to be larger for rabbits and grey squirrels must be con­ ash, oak and sycamore than for beech owing trolled in numbers. Some have used tempo­ to their different light requirements. In view rary internal fencing to protect regenera­ of the foregoing information, after consider­ tion, but rarely are treeshelters used for this ing wind effect and deer browsing probabili­ purpose. If group selection is to become ties, groups should be used that are as large more widespread, rangers and gamekeepers as the site conditions and species require­ are likely to be in good demand and field ments will allow but generally will be less sporting may need to be encouraged. than 1 ha to maximise the benefits of natur­ 10. Vegetation control. An interesting feature al regeneration and side shelter and protec­ is provided by the reference to the barriers tion from excessive insolation. to regeneration caused by particular plant 7. Access and extraction. There is sparse infor­ species, e.g. the great wood-rush at Cawdor mation about access for tending, but some and Holcus and grass swards generally, owners and foresters emphasise that tend­ and predation by small mammals such as ing and extraction paths must be integrated mice and voles. Generally it is essential to with the spatial arrangement of groups. control vegetation by herbicides or mulches

37 or manipulation of the overstorey, or by use that are too small in area is a common of an understorey. problem. No forester wishes to open the 11. Management. A rotation for group selec­ canopy unduly and admit wind, so the tion is rarely specified, because a fine tree group selection system must normally be is allowed to grow on until the value incre­ restricted to windthrow hazard class 1-3, ment ceases. If this practice became gener­ and to deep soils that permit good root al, it could be a good advertisement for development. Typically, one or two tree forestry. Precise costs and benefits are also species will dominate the stand and site, lacking, partly because mixed, uneven- making it difficult to keep others in a aged and irregular stands have so far been mixed group. To adequately satisfy the a small part of British forestry and no val­ needs of two species is sometimes difficult; ues have been placed on high amenity in to meet the needs of more than three is full economic appraisal. There is need for of problems. inventory and periodic measurement of 13. Factors leading to success. Besides conti­ development of the desired structure (see nuity of management, sustained owner­ Appendix 3). ship by enthusiastic and patient practi­ 12. Factors causing difficulty. Continuity of tioners is essential; so too is staff training management is essential and, if it falters, to impart the skills of conversion and several unfortunate things happen. The management. The period of conversion for most obvious is that groups of young regen­ conifers must not exceed 30 to 40 years, eration begin to run short of light and need which is roughly equal to the working relief by thinning of the upper storey or lives of two foresters. It is likely that three cleaning within the groups. Again, if thin­ principal species which are compatible nings are delayed, the older trees dominate with each other in their light require­ and trees of the lower size classes soon suf­ ments are sufficient; any others should be fer. A possible useful guide is that there secondary and occupy the lower canopies. must not be weakly co-dominants and sub­ Any suggestion that a group selection dominants in the stand; the trees must stand constructed by patient care and thrive and make their contribution to the attention is ‘self-perpetuating’ must be whole canopy. A forester used to thinning a forgotten, although readiness to intervene uniform stand from below must instead and experiment to gain experience will do thin from above and concentrate on the much to ease what is obviously a satisfy­ crowns and the canopies. Use of groups ing task.

38 Chapter 4 The single tree selection system

Introduction tinuous forest cover. It involves regular removal of single trees of marketable size with­ In the single tree selection system, felling and in the dominants, distributed throughout the regeneration are distributed over the whole stand and adjustment of stem-size class num­ management area. The fellings are called bers to conform with a stem-number curve. A selection fellings; they remove single trees and system of extraction paths aids harvesting and are combined with thinning among the various reduces damage to regeneration. However, size classes, and aim to perpetuate a structure when experienced staff and directional felling in which all age and size classes are mixed are employed, damage is much less than might together over every part of the stand. Unlike be expected. Where conditions are favourable, the clear cutting or shelterwood systems, the natural regeneration springs up in the small single tree selection system maintains a con­ gaps created by the felled trees. Measures to tinuous forest cover: viewed from both without assist regeneration may include removal of and within, the scene and the environment excess raw humus, soil cultivation and even appear stable and unchanging. The fellings planting in gaps caused by felling or accident. and thinnings create quite small gaps to be Under ideal conditions, this process goes on regenerated by natural seeding, and on the year after year throughout the whole stand, Continent the species are virtually all moder­ resulting in the constant maintenance over the ately or very shade-tolerant, e.g. European sil­ whole management area of the uneven-aged ver fir, Norway spruce and beech. structure in which trees of all sizes are mixed together in numbers conforming to the ideal Application on the Continent stem-number curve (age is not recognised, just as rotation is not). Such perfect distribution is The system is commonly used in Europe seldom actually found, the small size classes where important management objectives are occurring more usually in patches resulting site protection and enhancement of the exter­ from regeneration springing up in small gaps. nal landscape. It is often advocated as ‘natur­ Thinnings among the various size classes are al’ and ‘perpetual’ forest. Stands managed necessary to provide for sustained yield and under single tree selection are protective in quality, ensuring that: several senses: continuous forest cover pro­ tects the soil; frost-sensitive species such as • the various classes are maintained in their beech are sheltered; and the lack of sharp correct proportions; edges to the canopy plus the strong taper of • a suitable mixture of species (where the tree stems, generally give increased firm­ required) is maintained; ness against wind and snow. Another impor­ • young growth is freed from suppression. tant attribute is its help in regulating water Cleanings are also undertaken among the flow into catchments. younger age classes. The single tree selection system as applied The selection and felling of single trees scat­ in central Europe is the classic example of con­ tered over the whole management area would

39 be impracticable except in stands of small size, number curves are a simple and easy to use and hence, on the Continent, this procedure guideline (Philip, 1994). (sometimes called the ‘ideal’ selection system) Advantages of the single tree selection sys­ is usually confined to woods or small forests tem are: by maintaining a continuous forest belonging to private individuals or communi­ cover, exposure of the soil is reduced and pro­ ties. In forests of larger size, it is customary to tection provided against erosion, landslips and divide the area into several more or less equal avalanches; and damage by wind and snow is ‘blocks’, in one of which selection fellings are minimised. All seed years can yield regenera­ made each year, so that the whole forest is tion, however infrequent or partial, and the worked over during a period of years (‘the seedlings are protected. It is possible to pro­ felling cycle’ is equal to the number of blocks). mote the development of individual trees with Under this ‘periodic’ selection system, fellings good form and branching habit and to retain are more concentrated, there being an accumu­ them as long as they are making valuable lation of mature trees during the interval increment because the crowns of the dominant between two successive interventions in the trees are long and well-developed and stand same “block’, so that a larger volume per more or less isolated. The lower parts of the hectare is available than in the case of annual stems of trees of large size yield timber of high fellings extending over the whole forest. The quality but the upper exposed parts do produce felling cycle is fixed according to the circum­ large knots, and taper markedly. stances of each case. It is generally not more Selection forest, aesthetically and environ­ than 10 years and is often less; in France, usu­ mentally, is usually very attractive, bringing ally between 5 and 8 years, rarely more than 8 benefits to the owner and the general public. A years; and in Switzerland 6 to 10 years. In further advantage sometimes argued is that most instances the periodic yield is relatively the system is ‘natural’, resembling natural diverse in size classes and may be more diffi­ indigenous forest. This is not widely accepted. cult to market. Peterken (1981) and Peterken et al. (1992) In Switzerland, the regulation of the yield of assert that such stands are unnatural, being timber from selection forests is by the Methode the product of sustained, skilled and highly du Controle or ‘Check’ Method devised by the regulated forestry working with a small group French forester Biolley (1920) and described in of shade-tolerant species. The system is partic­ detail by Kniichel (1953) and in outline by ularly well suited to small woods or forests Anderson (1953), Paterson (1958) and where intensive working and close supervision Osmaston (1968). Biolley adopted the principle are possible. of gaining maximum volume increment from Disadvantages of the single tree selection the smallest possible growing stock, and set system are that it is one of the most intensive, out to maintain a definite distribution of size needing considerable silvicultural skill, conti­ classes so as to establish a normal forest and nuity in management and ownership, and care ensure sustained yields. He divided the forest in marking, felling and extraction. The system into compartments and made recurrent inven­ is applicable chiefly to shade-tolerant species tories of the growing stock in three size classes on well-drained soils, and is unsuitable for (large, medium and small) at intervals of 6 to areas where windthrow risk is high. The work 10 years, to determine the relation between of tending, felling and extraction is scattered increment and growing stock, fix the yield for and on a small scale, needing experienced and the next period and plan the fellings and thin­ knowledgeable supervision, and the costs are nings to work towards the normal distribution likely to be higher than those in a comparable of size classes. It is difficult to appreciate the area of even-aged forest. Control of browsing significance of the full control system in and bark-stripping mammals is probably more Britain, where most woods are in a far from difficult than in even-aged forest, hence dam­ normal state (Hart, 1991). However, stem- age done can be more severe.

40 The single tree selection system is impor­ diameter at breast height can be very high. tant in the mountainous regions of Europe, Thus, the diverse structure forest is produc­ especially in Switzerland where it must be tive, stable, and continuous and yet is capa­ practised by legislation for protection purpos­ ble of meeting markets where large size and es; examples are the mixed conifer and high quality are greatly valued (Matthews, broadleaved selection forests of the Vosges, 1989). Jura, Alps and Carpathians. Often the main purpose is to divert snow avalanches and rock slides away from villages. The main species are Use in Britain Norway spruce, European silver fir and beech, Britain has no examples of the ‘true’ selection with some sycamore, rowan, lime and occa­ stands found in central Europe, because it has sional other species. The beech is used to main­ no comparable mixed natural indigenous tain fertility of the soil and depress weed forests. This has not prevented a few British growth. There are several variants of the main foresters from attempting to create single tree forest type and all are typically montane, being selection forest with the species and sites confined to hills and the middle slopes of the available to them (see Table 4.1). They have mountains and rising to elevations of 1500 m used light-demanding species like Scots pine or where the geology is varied. A major feature of ash, moderate shade-bearers like sycamore or the climate is the long and often very severe Douglas fir, and shade-bearers like grand fir, winters usually with more than four months of western hemlock or beech to provide compo­ snow. Rain is plentiful and evenly distributed. nents of the different canopy levels -described The proportion of species varies from site to earlier. site; in general, that of beech is greater on Another difficulty in Britain is that sunny or south-facing slopes than on the shad­ although foresters are familiar with the silvi­ ed, north-facing slopes. The storied structure cultural characters of many species when usually has the conifers projecting above the grown pure, they are less informed about broadleaved trees of the lower storey. Gaps in their behaviour in mixtures or in two-storied the stands are filled with young mixed regen­ or multi-storied forest. So, attempts to use the eration. selection systems have gone in two main On a favourable site the mean annual directions: either to create mosaics of small increment of a selection forest composed groups or to proceed by underplanting to mainly of Norway spruce and European silver achieve several storeys. The four examples fir with some beech is 10-11 m3 h a 1 year1, and which follow suggest that both these routes the proportion of trees exceeding 50 cm in are viable.

Table 4.1 The location of stands treated by the single tree and/or group selection systems. Wind Main soil Species Age and zone 3 type conversion Weasenham New Wood, Swaffham 3 Podzol Mixed conifers Various Kyloe Wood, Berwick-on-Tweed 6 Brown earths and surface-water gleys Mixed conifers Various Ipsden Estate, Oxfordshire 1 Clay-with-flint Mixed conifers 40 years and broadleaves Rossie Priory, Dundee 3-4 Brown earths Mixed broadleaves Various a Wind zone describes the regional windiness for the site; values obtained from Quine and White (1993) and Quine (person­ al communications for areas in SE England). See Glossary 1 for further definition.

41 Weasenham New Wood, The result of almost a century of manage­ ment is a single tree-cum-group selection N orfolk stand capable of sustained production of high-quality timber, with high amenity value, and encouragement of wildlife (front Weasenham New Wood is situated near cover plate). Some of the best trees have Swaffham in west Norfolk. A stand of 20 ha been high-pruned, but the growth of the is a well-established example of selection for­ lower canopy and shading of stems has est composed of mixed conifers (Douglas fir, made this unnecessary for many other Scots pine, larches, western red cedar, west­ species. The intention is to continue to grow ern hemlock and species of Silver fir) togeth­ mainly conifers with a small number of er with a few broadleaved trees. It is part broadleaves such as oak, sweet chestnut group selection and part single tree selection. and birch in a diverse structure with no Access to the stand is well served by an inten­ clear cutting. The New Wood is ‘self-perpet­ sive system of tracks. uating’ provided that the skilled manage­ This stand and the woods around it are the ment continues. It has great beauty, aug­ result of original thinking and skilled man­ mented by azaleas and hybrid rhododen­ agement by three generations of the Coke drons at ride intersections. Major Coke family. In the 1880s the second Earl of affirms that ‘the timber in the uneven-aged Leicester (grandfather of the present owner, woods is far superior to that in his even- Major R. L. Coke) planted west North aged woods, and the former has suffered American conifers on heathland with ironpan less from windthrow. The uneven-aged podzol in gaps among Scots pine self-seeded woods have also been considerably more from old clumps. Many of the new plants profitable’. Major Coke’s experience at were killed by frost, but after a gale in 1885 Weasenham New Wood enabled him in 1980 had removed most of the old trees, Scots pine to urge ‘more irregular forestry in Britain’ and birch invaded the site. The New Wood to a Government Select Committee on was fenced against rabbits and in 1905 varied ‘Scientific Aspects of Forestry’. species of conifers were planted. In 1907 many more trees were planted, and all grew well to provide shelter and the backbone of Kyloe Wood, Northumberland today’s New Wood. After 1920 more trees were added and natural regeneration was encouraged by making thinnings in the devel­ Nine miles south of Berwick-on-Tweed, Kyloe oping stand. Wood (391 ha) is owned by the Fleming family. Weasenham New Wood was first substan­ It includes many fine conifer stands of diverse tially recorded by Paterson (1958) by which structure, consisting largely of maturing and time it had evolved from a two-storied stand mature overwood with an understorey of nat­ into an irregular stand with many species ural regeneration. The top canopy of conifers and several canopy levels. There was also a (chiefly Douglas fir, grand fir, Scots pine, west­ notable increase in the proportion of shade- ern red cedar and western hemlock) were bearing and moderate shade-bearing species. mainly planted in 1915 but some trees are Establishment, initially by underplanting, older. Prolific western hemlock natural regen­ was increasingly by natural regeneration, eration appeared following thinnings during with some large gaps being provided to the early 1970s and even more abundantly encourage regeneration of larches, which later in that decade (Plate 18). These thin­ locally were the most valuable species. nings and others made during the 1980s did Selective fellings were made on a 3 to 5-year some damage to the understorey but it was cycle. quickly made good by new regeneration.

42 There is a lack of 20 to 40-year-old trees. predominantly of beech or oak which are Some of the stands are now two-storied and developing the multi-storied condition of a irregular in appearance. Earlier attempts at selection forest, largely as a result of thin­ group selection and the single tree selection ning, underplanting and natural regenera­ systems were not successful. Almost all the tion. Continental recommendations are being mixed, two-storied or multi-canopied stands followed in keeping natural regeneration in a are dominated by western hemlock which is suppressed state for periods of 20 years or sometimes so aggressive as to prevent the sur­ more. The trees then tend to develop a char­ vival of other species. For the past 20 years it acteristic flat-topped form, but are expected has covered much of the forest floor in a dense to respond to light when the overstorey is mat often of about 2000 seedlings per square thinned. The treatment cycle is variable, usu­ metre. Kyloe Wood could consist of western ally between 6 and 12 years. In other areas a hemlock and Noble fir since the other species group selection structure is more apparent regenerating— notably Douglas fir, Scots pine, (as in compartment 10). The old trees in larch and western red cedar — provide only groups of 0.2 to 0.6 ha are heavily thinned sporadic cover. The regeneration overall has and underplanted, usually with beech, oak, been successful on the soils derived from sand­ pine and grand fir. Oak grows strongly on the stone and less frequent on the heavier clay deeper clay-with-flint soils and outperforms soils. beech. The objectives of management are to main­ Since 1953 Reade has monitored his woods tain uneven-aged stands of mixed conifer by the ‘Check’ Method or Methode du Gontrole, species as well as notable specimen trees and in this instance recording the number, species to achieve sustained yield. The dominance of and stem diameter of all trees larger than 16 western hemlock poses problems in species cm dbh at intervals of 6 to 10 years. About diversity which will require intervention. The one-tenth of the 142 ha is enumerated, requir­ diverse, mixed, uneven-aged stands supple­ ing about 3 man-hours per hectare. Local tar­ ment a wide range of fine large conifers, iff tables are used to convert diameters to esti­ many high-pruned and up to 90 years old. mated volumes. The growing stock has been The whole forms a woodland of beauty and rising steadily in volume, and is expected to interest. Much credit is due to past foresters, increase further for the next 30 to 40 years. notably Wilfred Fox. Stabilising the growing-stock at a level com­ patible with the planned distribution of size- classes is the prime aim of management; but Ipsden Estate, Oxfordshire the interests of all other aspects and attribut­ es of the forest are also considered.

Ipsden Estate is part of the Chiltern beech- . woods. The owner, M. G. Reade DSC (see Rossie Priory Estate, Angus Appendix 5), has consistently established what he describes as ‘irregular selection for- ^ est intermediate between single tree selection The Rossie Priory Estate, owned by the and group selection’ (Reade, 1957, 1960, 1965, Honourable Mrs C. Best, daughter of Lord 1990). The emphasis varies between one site Kinnaird, lies about 9 miles west of Dundee, and the next. At present, some areas (e.g. An uneven-aged and irregular stand in com- compartments 5 and 7) are conventional partment 39 extends to 6 ha of mixed stands of mixed broadleaved species, broadleaves growing on a deep moist brown although an attempt is made to diversify the earth. The main species was sycamore, and age distribution by retaining some overwood. the older trees included beech, oak, ash and There are also areas (compartments 2 and 3) wych elm, about 170 years old, the remains of

43 a former plantation. The stand was first sub­ bined features of group selection and single stantially recorded by Paterson (1958) when tree selection; and Weasenham New Wood, ash and sycamore regeneration had occurred given continued care, will remain as multi- throughout, being abundant in clearings cre­ storied mixtures of light-demanders, moder­ ated by removal of old, unhealthy or defective ate shade-bearers and shade-bearers. It is trees. Group size was 0.1 to 0.2 ha and the significant that both stands traversed a edges were rapidly extended after successful period when they were two-storied, and it regeneration. Temporary fencing against roe may be that creating two-storied high forest deer and rabbits was used frequently to pro­ is a possible or even a desirable route to sin­ tect the regeneration. The system which gle tree selection forest. appeared to be best suited to the light- demanding behaviour of ash and sycamore 2. Creating mosaics. Ipsden Estate and Kyloe created a mosaic of single-storied groups. Wood provide two contrasting examples. In the former, the owner has a wide variety of Although in 1958 the growing stock was soils to contend with and conversion may well distributed over the size classes, there was a lack of information which would pro­ have been from scrub and poor quality vide a reliable guide to treatments and woodlands to something better. Parts of fellings: periodic inventories and increment Kyloe Wood have to contend with an aggres­ calculations would have been of value. Sadly, sive shade-enduring climax species, western the stand was heavily damaged by a storm in hemlock, and the few examples of attempted 1968 but proved resilient. The storm created selection forest are reverting to even-aged new clearings which were filled by dense stands chiefly of that species and some of regeneration of ash and sycamore. Attention grand fir. to cleaning and thinning ensured that high 3. Soil preferences. Selection forest can be cre­ quality poles of both species were recruited ated on soils ranging from podzols with iron- into the middle and upper canopy. The system pan to deep moist brown earths. So far, it can thus be applied to vigorous, light- has been thought that selection forest demanding broadleaves provided that consis­ should be attempted only on fertile shel­ tent, regular and positive silvicultural treat­ tered sites but this may not always be so. ment is done. This example shows the value However, sites with high windthrow possi­ of retaining in mixed broadleaved stands bilities, and wet soils, should be avoided. some of the older trees to form a shelterwood for group regeneration, in this case of ash and 4. Influence of soil texture on natural regener­ sycamore and replanting of oak in patches. ation. Natural regeneration is more plenti­ ful on soils of light texture than on heavier Conclusions on the single tree soils, perhaps due to lesser weed growth, and often appears after disturbance of the selection system canopy by thinnings and skidder extrac­ tion. 1. Light-demanders and shade-hearers. The examples of Weasenham New Wood and 5. Achieving the balance in broadleaved selec­ Rossie Priory Estate have been managed for tion forest. For broadleaved selection forest many years to encourage diversity through there must be at least one, and preferably use of selection forest principles. Although two, light-demanding species that produce the Weasenham New Wood is largely conif­ valuable timber. If there are two shade- erous and Rossie Priory was composed of enduring species in the mixture, they must broadleaves, the former still contains, and be controlled. This is done by denying light the latter did contain, an important element until the one that needs more light is well of light-demanding species providing good established, or by removing some of them timber readily marketed. Thus, they com­ before they get too numerous.

44 6. Trial and error period. There has to be a peri­ sure the development of the stands and to od of trial and error before a suitable form of guide the application of thin/fell to lead selection forest is found for each set of site con­ towards the ideal stem-number curve (see ditions. However, it seems that multi-storied Appendix 3). British examples have been con­ structure, suitably stocked with compatible ducted without this; an important exception is species, will be useful although likely to fall atIpsden, by Reade. half-way between single tree selection and 8. Continuity of management. The benefit of con­ group selection. tinuity of management is patently demon­ 7. Importance of inventory. It is important to strated at Weasenham. have initial and subsequent inventory to mea­

45 Chapter 5 Other silvicultural systems

The four silvicultural systems considered in Two-storied high forest this chapter are: two-storied high forest, high forest with reserves, coppice and coppice with This system is composed of an upper and lower standards; also included is restructuring storey of trees growing on the same site. extensive conifer forest by patch clear cutting. Generally, two species are involved, the upper Particular reference is given to the Chilterns storey usually comprising a light-demander as being an important beechwood region under which a shade-tolerant species can grow requiring special treatment for structure, without being suppressed when introduced at access, recreation, wildlife conservation and a later date. Trees of the upper storey are landscape. Table 5.1 gives examples of stands treated as an even-aged stand in terms of treated by these systems. stocking and thinning until they approach

Table 5.1 The location of stands treated by other silvicultural systems. Wind Main species zone?

Two-storied high forest Dartington, Devon 2 Japanese larch over western hemlock, western red cedar and Douglas fir

High forest with reserves Longleat, Wiltshire 2 Douglas fir and larch Bolderwood, New Forest, Hampshire 2-3 Douglas fir Reelig Glen, near Inverness, Inverness-shire 4 Douglas fir and mixed conifers Plodda Falls, Guinsachan, Inverness-shire 4 Douglas fir

Coppice and coppice with standards Numerous examples available in south-east and south-west England Broadleaves

Restructuring by patch clear cutting Kielder, Northumberland 5-6 Sitka spruce North York Moors District 3-4 Sitka spruce

Restructuring by (mainly) group regeneration The Chilterns, Buckinghamshire, Hertfordshire, Oxfordshire, Bedfordshire 1 Beech a Wind zone describes the regional windiness for the site; values obtained from Quine and White (1993) and Quine (person­ al communications for areas in SE England). See Glossary 1 for further definition.

46 mid-rotation, when a final heavy thinning and damage can occur to the lower storey unless possibly some high pruning takes place. At care is taken (appropriate thinning of the this stage, the stand is underplanted with a lower storey will allow felling into prepared shade-tolerant species; occasionally the lower gaps); and the system is more difficult to man­ storey is established through natural regener­ age than a uniform even-aged one, both in ation before or soon after the heavy thinning. terms of balancing the growth of each storey Both storeys are allowed to grow up and sub­ and in carrying out regeneration and tending sequent thinnings are undertaken in the lower operations. storey. They may be felled together, or the On the Continent, the most common reason upper storey may be removed to leave the for using the system is to encourage diameter lower as an even-aged maturing stand. Two- increment in selected upper storey trees while storied high forest is frequently limited to only protecting the soil with the lower storey. The one rotation because of the difficulty of forming objective is most frequently met where Scots a second understorey under the previous pine is underplanted with European silver fir, shade-tolerant one. The trees to be retained Douglas fir or Norway spruce; or where Scots must be prepared for this several years in pine or oak is underplanted with beech. advance so that they develop strongly tapered Miegroet (1962), after studying two-storied stems and good root systems, and this implies high forest stands for small estates in Belgium, early commitment to the establishment of a calculated minimum areas for regeneration, lower storey rather than reliance on the sys­ tending and economic management. For areas tem to resolve a silvicultural or management of 5 ha or less, he considered the system was difficulty at some later stage. promising, using as the upper storey light- The advantages of two-storied high forest demanding species with rapid growth and according to Matthews (1989) are: short rotations, such as ash, sycamore and • The lower storey protects the soil, controls wild cherry, and as the lower storey Douglas weeds and gives freedom to thin the upper fir, western hemlock and grand fir. storey heavily to encourage the best stems to add increment. • The upper storey provides protective cover Dartington, Devon for establishing those shade-tolerant species which are subject to frost damage. • Where a mixed stand is required but the In the 1950s W. E. Hiley (1956, 1959, 1964, growth rates of the species differ, the system 1967) experimented with two-storied high can be used to give the slower growing forest. He wished to grow Japanese larch to species an early start. large diameters but noted that in well- stocked stands current annual diameter • There may be early income from the final increment culminated quite soon; and in heavy thinning at a time when the trees are order to maintain rapid growth in diameter still of moderate size. the trees needed large crowns, requiring vir­ • The system can be used to effect a gradual tual isolation at an early age. Hiley decided change of species and products through the that underplanting with a shade-tolerant lower storey. species would make fuller use of the site • At least in its early stages the visual while allowing rapid growth of the Japanese appearance can be pleasing. larch. He had hoped to exploit a strong local The disadvantages of the system include: in market for mining timber with the produce windy climates the sudden exposure of the from the heavy thinning of the upper storey, upper storey by the heavy thinning can result avoid less profitable later thinnings and com­ in windthrow even on brown earths; when the bine the advantages of two-storey forest with upper storey trees reach the age of final felling, the economies of the clear cutting system.

47 Site protection and weed suppression were to felling more of the overstorey, and cleaning and be achieved by means of the lower storey of thinning the groups as required. Then, make shade-tolerant species which would be tended space for further group planting by felling more as an even-aged stand. upper storey larch, encouraging natural regen­ In pursuance of those objectives, in 1955 eration of larch and introducing more species Hiley heavily thinned a compartment of 25- such as Douglas fir and Norway spruce. The year-old Japanese larch to 120 stems ha-1 and size of the groups should not be less than 0.5 ha underplanted with western hemlock, western and preferably larger for the light-demanders. red cedar and Douglas fir (Howell et al., 1983). The total time for conversion would be approxi­ The first two species succeeded in the under­ mately 30 years, at which time the forester will storey but the Douglas fir did not. Although have to decide on the principal species. Hiley achieved his silvicultural and marketing objectives in the two-storied stand, his finan­ High forest with reserves cial predictions proved optimistic. The system was discontinued after 1965, partly due to This is produced by retaining selected trees of a changes in the dimensions of timber required parent stand over a young crop established by the local mining market. Hiley, besides prov­ below it through natural regeneration. These ing that the overstorey must be further reduced large trees, known as ‘reserves’, may be if Douglas fir is to grow healthily as an under­ retained, scattered or in small groups, for the storey, saw several benefits in the two-storied whole or part of a second rotation. The objective system. The ground is never completely uncov­ is for them to add increment and produce large­ ered and a small selection of the best trees is sized timber for use in boat-building and con­ given the necessary room for rapid growth in the struction and some to produce veneer quality. overstorey. The system enables adjustment of Thus, reserves are normally of species in the distribution of age classes within a stand. demand for those purposes, such as Scots pine, Where there is an excess of young age classes, European larch, Douglas fir and oak. Reserves early maturation of the open upper storey pro­ may also act as seedbearers to fill blanks in vides a final crop sooner than is possible in a young crops. Other important benefits are dense uniform stand. Moreover, the lower amenity by producing stands with a degree of storey helps to fill the gap in the distribution of natural appearance and increasing diversity. size classes which may otherwise be caused by The selection of reserves requires care. Large not having any land to plant. Historically, it trees suddenly isolated are liable to be thrown was thought that some species were ‘incapable by wind or in certain cases become unhealthy; of maintaining soil fertility’ and that growing a and hence reserves should be of windfirm lower storey of a different, usually broadleaved, species and have well-tapered stems, in addi­ species would help to protect the soil from tion to strong crowns and firm root systems. degrade. However, at Dartington, the poorer Trees in some instances are prepared for isola­ sites gave little gain and did not justify estab­ tion by selecting prospective reserves many lishing a lower storey (Howell et al., 1983; years ahead so that they may receive special Matthews, 1989). attention during thinnings. Oak is often retained as a reserve, but is liable to become Two-storied high forest can provide a quick stag-headed and produce epicormic branches route to selection forest. As an example, take a when isolated. Beech is generally unsuitable, 30 to 40-year-old stand of larch. Thin to vari­ as it has a heavy crown and is liable to sun- able intensities to suit the species to be under- scorch when suddenly isolated. Spruces are gen­ planted in groups, i.e. to 300 stems h a1 for beech erally unsuitable. Douglas fir can be accepted in and grand fir; and 100 stems h a 1 for Scots pine, reasonably sheltered areas. ash and sycamore. Allow the underplant groups Advantages and disadvantages of high forest to grow on for 15 to 20 years, progressively with reserves are somewhat similar to those of

48 two-storied high forest although harvesting Coppice and coppice with damage to natural regeneration and under- planting is less. The greatest danger being the standards possibility of windthrow has resulted in a The systems of broadleaved coppice and cop­ decline in the use of the system following some pice with standards (nowadays used mainly in major storms on the Continent during the past England, but previously also in Wales and 20 years. In France, this system is infrequent­ Scotland) are well known. For general ly used, whereas in Germany it is more com­ accounts of them and associated techniques of mon. Four notable examples in Britain are out­ silviculture, the reader is referred to Rackham lined below. (1980), Peterken (1981), Evans (1984), Pryor and Savill (1986), Crowther and Evans (1986), Longleat, Wiltshire Matthews (1989), Hart (1991) and Buckley (1992). Both systems are always the result of human management and are applicable main­ In this adaption of the system small areas of ly to sweet chestnut, oak, ash, lime and hazel. large Douglas firs and larches are retained That of pure coppice, when used for sustained as reserves, for amenity and landscape pur­ yield, led to the idea of rotations to suit differ­ poses, to serve as seedbearers, and to add ent species and local markets, and to the prac­ value increment; some trees are intended to tice of felling in roughly equal and contiguous become ‘monarchs of the forest’. Windthrow annual ‘cants’ which together made up a felling is usually a risk. series. Coppice with standards met so many needs of the medieval rural economy that it was soon fteelig Glen, Inverness-shire systematised into a carefully managed under­ wood of selected species such as hazel and lime (later expanded to sweet chestnut coppice by At Reelig Glen near Inverness, Douglas fir is planting that species), and of an overwood with a major constituent of a mixed and somewhat a balanced growing stock of standards com­ diverse structure stand growing on a fertile, prising chiefly oak, sweet chestnut and ash. moist and sheltered site. It is also popular The conversion of coppice to high forest, by with the public because of the varied habitat ‘storing’, is an alternative management system and range of bird life (M. Rodgers, personal for most species of coppice, but in Britain such communication). conversion has been effected predominantly through neglect or from decline in markets for Plodda Falls, Inverness-shire the coppice products. The two coppice systems are now regarded as being among the most desirable silvicultur­ al treatments of broadleaves from the view­ The third example with Douglas fir is at point of nature conservation on lowland sites. Plodda Falls near Guisachan. When standards are present, particularly if they are diverse in age, they not only provide a Bolderwood, New Forest, deep canopy but also continuity of woodland Hampshire conditions. In addition they have attributes for landscape, whereas pure coppice— although it supplies benefits of plantlife conservation— can over time be temporarily undesirable scenically, At Bolderwood the reserved Douglas fir have if too dominant in the local landscape; the bare commanded very high prices as ships’ masts ground condition is repeated much more fre­ and for large flag poles. quently than with high forest. Coppice occasion­

49 ally can be thinned, especially if the introduc­ Thetford Chase, Norfolk), or by major recre­ tion of light would have beneficial effects on ational pressure (e.g. Grizedale, Cumbria). yield/size of produce or nature conservation. Restructuring uniform even-aged conifer for­ Coppice woods can thus be managed on a kind est in the uplands using patch clear cutting (as of selection basis (the coppice selection system: well as some other methods of regeneration), Matthews, 1989), with a range of ages of patch size being varied according to altitude and regrowth on the same area, giving greater con­ aspect, is now progressing, adhering to felling tinuity of woodland cover and more internal and landscape design plans. These are relative­ structural diversity than conventional coppice. ly recent attempts, and it is uncertain that the This procedure, though requiring extra man­ planned structure will be fully achieved owing agement, may have advantages for landscape to the unpredictable effects of windthrow and and wildlife. Conservationists present a strong lack of preparation of coupe edge. In north-east case for the retention of the coppice systems as England, two good examples of such restructur­ they are the traditional management for many ing by Forest Enterprise are Kielder Forest and of Britain’s broadleaved woodlands and that North York Moors District. which gives high biological diversity. Many older broadleaved stands have the appearance of high forest but actually are of Kielder, Northumberland coppice or pollard origin. When the main demand for coppice products almost ceased around the 1920s, the stems/poles were left to This extensive State forest, managed by Forest grow on to full size trees that had some sale Enterprise, consists mainly of 40 to 50-year-old value. The timber yield from these ‘stored cop­ Sitka and Norway spruce (Plate 19). It is man­ pice’ stands is similar to comparable high forest aged under a design plan to restructure uniform stands, though quality, particularly in the base even-aged plantations of the first rotation into of the tree, may be inferior. Coppicing may be diverse structure to serve increasingly multiple possible in these stands, even after 100 years. purpose objectives. The design concepts are a reflection both of the physical limitations Restructuring extensive conifer imposed by windthrow and the site (gley soils) forest by patch clear cutting and the aim to achieve a reasonable balance between timber production and environmental The concept of introducing diversity into exten­ enhancement (McIntosh, 1991). The design plan sive uniform even-aged State conifer forest by seeks to integrate the requirements of efficient restructuring using medium-sized patch clear operational activities with a planned increase in cut coupes was initiated by a Forestry wildlife and landscape diversity. The current Commission Conservator, J. A. Spencer, in extents of felling coupes by altitudinal zone are north-east England in the early 1980s. The generally 5-10 ha in the valley bottom, 15-45 ha stimulus to restructuring can be associated on the mid slopes and 50-100 ha on the upper with the realisation that additional land will slopes. not be released to forest use around a given The main elements in the process of holding/district, and plans must be laid to restructuring are the identification and obtain a sustained yield and forecasted sup­ design of future felling coupes, manipulation plies of small diameter roundwood and of the felling dates to achieve a ‘patchwork sawlogs. Concentrating the felling into 10-15 quilt’ pattern of stands of different ages, years would not serve these objectives. indefinite retention of a small proportion of Elsewhere in State forests, restructuring giv­ stands, subject to possible windthrow, and ing rise to diverse structure tended to be initi­ the introduction of both broadleaved trees ated either by landscape plans whereby stands and open space. Considerable commitment were reserved to improve the design (e.g. and planning are required to achieve the

50 objectives set, and costs are incurred (mainly Firstly, the more long standing practice of loss of some revenue from timber production), restructuring is used, similar to the system but these relate largely to the first rotation. recently being undertaken at Kielder Forest, If some of the age-class manipulation can be but in the North York Moors and Hamsterley achieved in this period, then subsequent Forest in south Durham (Plate 20) the scale felling can take place at nearer the optimum is smaller, resulting in the development of a economic time without adversely affecting more intricate mosaic forest structure. Clear the desired pattern. It is not known what the fell coupes vary from 3 ha to, occasionally, as effect of this design will be on windflow pat­ much as 25 ha, and have been replanted. terns over the Forest, but the increased Natural regeneration has appeared on many roughness may cause more turbulence and occasions and supplemented planting. Where the risk of windthrow may be increased. practicable, the size and shape take into The whole is a conventional clear cutting account the principles of forest design per­ system but with constraints on the size of the taining to land form, scale and prominence in coupe and the timing of its felling in relation to the local landscape. This system suits the surrounding coupes. It will be interesting to management of the light-demanding species see whether the pressure over time is to sim­ growing in these forest areas, namely larches, ply reduce the size of the coupes without pines and Sitka spruce. changing the basic silvicultural system of clear Secondly, consideration is being given to cutting; already there are areas where the sys­ the introduction of more intensive selection tem is not sensitive enough and where more silvicultural systems to provide a degree of traditional continuous forest cover systems are naturalness in a few areas, particularly planned. There appears to be scope for more where adjacent to recreation facilities. This use of natural regeneration in smaller groups could not be introduced on a large scale on sites where weed re-colonisation is slow. because of the need for operating cost-effec­ tive management systems; but there is scope for small scale development of such stand North York Moors District structures along selected edge zones on forest walks or public rights of way, as well as applying it to larger areas of stands which This eastern heathland area of some 19 000 have a high usage by the public. ha, under Forest Enterprise management, In the North York Moors the problem lies presents examples of two forms of uneven- in the effective application of these more visu­ aged forest structure (J. MacKenzie, personal ally attractive and more intensively managed communication). The site conditions con­ systems within stands of light-demanding tribute to relatively good stability; there are species, such as pines and larches. This brown earths and podzols on the scarp slopes would normally be done by uniform seeding and extensive ironpan soils over the plateaux, felling or group shelterwood but there have with pine and larch stands dominant at mid­ been considerable windthrow losses on dle to late pole stages. This gives the oppor­ ploughed ironpan soils. A trial has been tunity for implementing systems alternative established to introduce a group selection to large scale clear cutting through the use of system into a heavily thinned Scots pine natural regeneration or by underplanting fol­ stand at Dalby Forest. Also at Dalby and in lowing preparatory heavy late pole-stage Cropton Forest, two-tiered larch stands are thinning. More gradual replacement also to be managed to develop further diverse favours long-term log supply to sawmills as structure on a group basis. the total forest area in the District is not Where Douglas fir already occurs in expanding significantly owing to the con­ appropriate areas there is more scope for the straints on land use in the National Park. development of selection systems as initially

51 Douglas fir is more tolerant of some shade often wrongly cited as an example of ‘selection when young than pine or larch. At Hamsterley forest’. Several centuries ago, the French prac­ and Dalby Forests, management for diversity tice of furetage was followed. This is really an of structure on a small scale has been intro­ ‘exploitation felling’, single good quality trees duced into selected stands where natural being removed here and there in an unregu­ regeneration has occurred. The same principle lated way, resulting in impoverishment of the is being considered for suitable stands of grand stand. Jones (1952) gives historical and silvi­ fir and western hemlock at Hamsterley and cultural evidence which shows that ‘exploita­ Wykeham Forests. tion felling’, rather than true selection man­ Another possible variant is to achieve a agement, has dominated these woods for the diverse structure in pine stands within a last few centuries. The woodlands were heathland ecosystem, the aim being to allow a steadily creamed of the best timber-sized more open stand to develop to encourage beech (along with smaller poles for wood turn­ Calluna and Vaccinium to be sustained by ers or TDodgers’) as well as almost all the oak. manipulating natural regeneration of pine. Some of the stands are of coppice origin. This could suit the North York Moors The Chiltems include numerous stands of National Park’s policies for woodland by beechwoods (Plates 21, 22 and 23), some State developing a more natural transition zone and many others in private hands. between forest and moorland, as well as with­ Predominantly of beech, there are also good oak, in appropriate open ‘space zones’ in the forest. ash and wild cherry. Sycamore, birch and horn­ Other examples of restructuring by patch beam are present in some areas. Less frequent­ clear cutting are with Corsican pine at ly, introduced conifers are found — European Thetford Chase (Norfolk) and with larch and larch, western hemlock and western red cedar, mixed conifers at Grizedale (Cumbria). and occasionally Norway spruce, grand fir, Douglas fir and coast redwood — all helpful for diversification, amenity and conservation rea­ Restructuring beechwoods by sons (K. Wallis, personal communication). Beech regenerates naturally, but rarely pre­ (mainly) group regeneration dictably, and has to compete with several dam­ aging agents including frost, deer, rabbits, grey The example chosen to illustrate this form of squirrel, voles, wood-pigeons and insects. The restructuring is the Chilterns. This is an provenance of the beech is not good, probably important woodland region requiring special due either to poor genotype from selective silvicultural treatment for timber production, exploitation or to unsuitable past management, structure, access, recreation, wildlife conserva­ so phenotypes are poor. However, some users of tion and landscape. It can be regarded as the timber assert that it is whiter and more flex­ attempting group shelterwood restocking on ible than that grown elsewhere. an extensive scale in multiple ownerships. Many of the beechwoods have been allowed to grow to financial maturity and beyond without any form of controlled replacement or The Chilterns, conversion. From the age of about 130 years, Hertfordshire/Buckinghamshire / depending on site conditions, beech on the heavier plateau soils start to deteriorate; this Bedfordshire/Oxfordshire process begins earlier and, especially following droughts, can be more rapid on the poorer soils of the scarp and combe sides. The result can The largest area of beechwoods in Britain, the be serious deterioration of the amenity, conser­ Chiltems, provides examples of the need for vation, landscape and recreational values of and use of regeneration. Its woodlands are the woods, as well as of timber production.

52 Almost half of the Chiltem woodlands are its uncertainty (beech mast frequency being mature or over-mature. They are remarkably some 5 to 7 years). Where natural regeneration even-aged with very few stands in the age arises at the time of felling, it can be used to range 30-80 years, being more in the range of advantage but, in general, planting— typically 80-150 years, with a few trees reaching 180 with mixtures of beech, oak, ash, wild cherry and years. Some good quality stands exist, but the European larch— is the main means of re-estab- general impression is one of even-aged stands lishment. Other relevant Forest Enterprise of low quality and poor stocking, lacking diverse structure stands on the Chiltems are at diverse age and size structure, and now in Queen Wood (see Chapter 2). (Private sector deteriorating health in parts. examples on the Chiltems by Bourne, Garfitt, The increasingly strong public pressure on and Reade have been noted earlier in Chapters 2 the Chilterns for access, recreation and land­ and 3.) scape retention has raised concern about how best to perpetuate this important woodland region (Penistan, 1974; Evans, 1984; Tilney- Conclusions on other Bassett, 1988). The Chiltem Plan (Chiltems Standing Conference, 1971, 1988, 1993) advo­ silvicultural systems cated that the beechwoods should be managed 1. Two-storied high forest can provide a rela­ with the objective of perpetuating a tively quick route to selection forest on suit­ broadleaved forest. The emphasis would be on able sites. - ' beech where appropriate, and on limiting the 2. High forest with reserves can provide stands area of each coupe to about 1 ha and to spread with a degree of natural appearance and the regeneration of individual larger woods (8 interesting diversity, the greatest danger ha or more) over a period of 30-40 years. being the possibility of windthrow. Conifers would be allowed as nurses provided 3. Coppice and coppice with standards are con­ they did not prejudice the main principle. In sidered to be among the most desirable silvi­ some areas of high conservation value, the cultural systems from the point of view of coupe size would be limited to 0.5 ha. nature conservation on lowland sites, and Throughout this century, continual attempts continuity of woodland conditions. Good spa­ to manage the numerous component stands of tial planning of the coupes can improve struc­ the Chiltem woodlands on a so-called ‘selection tural diversity. system’ mostly failed, resulting in the urgency during recent decades to regenerate effectively 4. Restructuring extensive conifer forest by patch large areas of such stands. For the bulk of the clear cutting gives rise to diverse structure; it over-mature stands the only management should adhere to felling and landscape option is to fell and restock them, either by designs. The effects of windthrow and coupe planting or natural seeding (despite its prob­ edge preparation are pertinent. Mosaics of lems) or a mixture of the two. In stands where small patch clear cutting which are naturally only some trees appear unhealthy and lacking regenerated would greatly soften the impact in vigour, group shelterwood fellings of the poor­ of the restocking stage on conservation and est parts can be carried out to diversify struc­ landscape. Such restructuring is now very ture, the size of opening generally being 0.2-0.4 widely applied, and is certain to increase. ha, though it may extend to over 1 ha. In stands 5. The Chiltern beechwoods contain examples of which appear completely healthy and still grow­ the treatments necessary for regenerating ing vigorously, very careful thinning may be car­ large areas of environmentally sensitive, ried out with a view to preparing for their regen­ aged, low quality broadleaved woodlands in eration in 20 to 30 years’ time. The use of nat­ the south of England. Forms of group shel­ ural regeneration has been successful on sever­ terwood and group selection seem most al sites but it is not generally followed owing to appropriate, given adequate protection.

53 Chapter 6 The economics of irregular forestry

Introduction native silvicultural systems estimated against them (Whiteman, 1991). Economic analyses of forestry until recent Economic analysis, and discounting in partic­ years have been chiefly an assessment of ular, has never featured as much in Continental financial yield, reflecting that forestry has, in forestry as it has in Britain, although it origi­ the past, predominantly been directed at prof­ nated around 1849 in Germany. Continental itable timber production. With an environmen­ foresters have not needed discounting to man­ tally aware society there is increasing interest age existing, often natural forests: the felling of in non-wood benefits, which often complement mature stands provides profit and pays for timber production. There is increasing need for restocking. In Britain, renewal of the forest these benefits to be taken into account by eco­ resource has required the injection of capital nomic analysis which could lead to changes in and this has been charged by discounting. The methods of silviculture, stand structure, economics of forestry are highly dependent on species choice and rotation. the discount rate employed, and the values A wide range of silvicultural options should which are accorded to non-wood benefits (Price, be tested; and analyses are required to deter­ 1989; Worrell, 1991). These are not unique to mine optimal, desirable, or at least acceptable uneven-aged forest, but are likely to be espe­ silvicultural systems on specific sites, and for cially important due to: desired objectives. The implicit objective, at least of State forestry, is to make forests useful 1. the assumed higher value of non-wood ben­ to society. Social returns are affected by a com­ efits which frequently may be expected; plex range of factors including: land capability, 2. the financial loss which may be sustained nature conservation attributes, recreation during any conversion period. potential, landscape, and watershed character­ The most commonly assumed objective of most istics. The overall social objective is still to State forestry until recent years was that of max­ operate so that the value of benefits derived imising net discounted revenue (NDR); and prof­ from a forest exceeds the costs expended, but itability remains of central importance, even now social and environmental factors should when some of the benefits are non-financial. be brought within the decision-making Therefore, careful economic comparison is needed process. of costs and revenues between uniform and One of the contenders for fresh appraisal is uneven-aged systems (Lorrain-Smith, 1986). The uneven-aged diverse structure forest. In its following questions need to be addressed: widest sense this comprises a group of silvicul­ tural systems that have not been practised to • In irregular forestry does management cost any significant extent in Britain, although more per unit area, e.g. because of relative­ they have long been practised successfully in ly small scale, complexity of working and western and central Europe. In making these other management decisions, and the need new, wider assessments, objectives of forestry for additional operations or higher over­ must be clearly stated and the value of alter­ heads (i.e. management time)?

54 • Are volume and value production and quali­ stems and reduce the number and size of lat­ ty of timber higher or lower and is there a eral branches. However, in later years when a different assortment of sizes? dominant tree becomes more exposed, there is • Are timber prices affected (much will likely to be an increase in the amount of taper, depend on the rotation), for example by as well as ring width, and, depending on the small scale and mixed assortments? species, stocking and any artificial pruning, • Are the timing of costs and receipts different there may be more live branches and hence between the two systems? larger knots in the crown wood and on trees on the edge of any groups (Helliwell, 1982); this • Do better non-wood benefits accrue to the would be apparent in the crown rather than in State and society or to the private owner? the more valuable butt lengths but these • What is the influence on cash flow (stored effects on wood quality may affect revenue. capital)? Debate has also occurred about the total vol­ Answers to these questions will only be ume production of even-aged forest compared acceptable when they are based on adequate with uneven-aged forest (Savill and Evans, data. Evidence from Glentress Forest suggests 1986). Productivity depends upon the species that costs of establishment may be up to 50% that are grown because there are differences in higher when planting is the method of regen­ efficiency of assimilation between species and eration, and of harvesting 10-20% higher, than even between genotypes of the same species on Forestry Commission standard rates the same site. In the selection system there (Shrimpton, 1986, 1988). Compensatory are may be greater productivity due to more effi­ the increased values of amenity, landscape, cient utilisation of incident light, and other wildlife conservation and forest recreation. resources, by the multi-storied canopy. The proportion of small-sized timber production is The single tree selection system probably lower, and the maintenance of contin­ uous forest cover means that the periods of low Estimates of costs or revenues have rarely been current annual increment at the beginning of published from uneven-aged stands in Britain, the rotation of an even-aged stand are largely but net timber revenues per hectare are likely to be lower than in the clear cutting system due to avoided. the need for directional felling, the increased care The conflicting evidence was reviewed by needed in extraction, and the diversity of size Johnston et al. (1967), Johnston (1978) and assortment and species of marketable trees. Helliwell (1982) but no clear conclusions are evi­ Where a selection forest is already fully estab­ dent. Differences in timber value are likely to be lished, annual management and tending costs more important. There has also been consider­ known, as well as the volume and value of the able discussion on size assortments and it is sustainable yield, it is possible to compare its usual that a much higher proportion of the total financial yield with that of a forest under the clear volume production of selection forest is in large cutting system. Experience is needed in order to sizes, and hence high value, compared with uni­ be sure of these factors, but modelling could help form stands. However, further research in this this work in advance of empirical knowledge. area is required (see Appendix 3). There has been debate as to the relative Price (1989) lists claimed financial advan­ quality and value of timber produced under the tages of uneven-aged stands (significantly the selection system. The trees grow under rela­ selection system), relating particularly to tively sheltered conditions, and receive a assumed increased yield, better quality of tim­ degree of shade in the early years, the effect ber, reduced susceptibility to storms and being to reduce (a) early growth rates/ring improvement of site by lesser soil insolation. width, and thereby the size of the juvenile core, He claims the system can save much cost of and (b) the amount of reaction wood. The sys­ insecticide treatment, sometimes necessary in tem is likely to increase the straightness of conifers when regenerating under the clear

55 cutting system, there being a low density of The economic justification of felled tree stumps, and the population ‘explo­ uneven-aged and continuous sion’ of pine weevil and black pine beetle does not occur. On the other hand, Price lists forest cover claimed financial disadvantages: the difficult, Texts touching on the economics of uneven- small-scale harvesting and more dispersed aged forest mostly agree that operations with­ working; extending rotations; planting low- in it are more expensive than in even-aged for­ yielding species to promote environmental est, but there is disagreement about the degree interest; or leaving areas unplanted for infor­ of such differences. mal recreation. However, he notes that even- Matthews (1986, 1989) found that the evi­ aged forest could suffer these penalties also in dence gathered by several Continental areas of high social pressure. foresters suggested that uneven-aged stands To create any form of uneven-aged structure produced about the same or slightly less tim­ takes significant time. Creation by afforesta­ ber than even-aged stands on similar sites, and tion could take several decades and entail that foresters were consistent in their claims of major financial penalty by delayed planting, higher monetary yields due to the superior extended rotations or felling of immature qualities of the timber. Specific tentative stands. Conversion of a uniform even-aged claims about the costs of restructuring State pole-stage stand entails felling some trees forests in Britain include: before optimal rotation and prolonged reten­ tion of other trees that may become financially 1. At Afon in Wales, where Forest Enterprise over-mature. The opportunity costs can be are restructuring upland conifer forest, abated by spreading conversion over several Farmer (1991) expected the discounted rev­ rotations, so that no tree is ever felled signifi­ enue of timber foregone to be about 15% cantly early or late, but the advantages of compared with clear cutting. uneven-aged forestry are then very long 2. Greig (1990) suggests that Forest Enterprise delayed (Price, 1989). conversion or restructuring of 15 separate lowland pine forests might be achieved for The shelterwood uniform and less than £5 ha1 year1 in profit foregone from timber production. group systems Whiteman (1991) sets out an economic These systems differ in several ways from the framework for evaluating different silvicul­ clear cutting and selection systems. The tural systems in State conifer forests includ­ absence of uniformity will increase the manage­ ing non-wood benefits. He suggests the like­ ment costs. However, the division of a forest ly levels of costs under each possible system, management unit into a number of different estimates the unpriced benefits to society, age- or size-class stands should give a more uni­ and concludes that diverse structure or con­ form distribution of costs and revenues through tinuous forest cover may in some cases be time. The shelterwood group system would have justified but that this is very sensitive to the a similar cash flow to that of a shelterwood uni­ method of assessing non-wood benefits; form system, although the regeneration period hence further research on methods to value would be longer and the management expenses these benefits is required. This deficiency is during the period greater. A group system would gradually being rectified by research, e.g. probably be little different from a selection sys­ Shrimpton (1986, 1988), Pryor and Savill tem in costs and revenues, provided the whole (1986), Benson and Willis (1992), Willis and forest is divided up into several different age or Garrod (1992). A better economic case for size classes, so that their regeneration periods alternative silvicultural systems is made overlap. Over the forest as a whole this would when appropriate data and analyses have result in similar annual costs and revenues. been developed and are accepted.

56 Conclusions on the economics of cases where forest is owned and managed on irregular forestry behalf of the State, because its owners will be acting wholly in society’s interests. Until conflicting objectives, and the ranking of 1. Overriding factors. Initial conclusions would them, are clarified, economics can only par­ seem to suggest that silvicultural, aesthetic and environmental reasons rather than eco­ tially solve the problem of deciding which nomic viability may be the overriding factors different uneven-aged systems could be used in determining the extent of continuous for­ (Whiteman, 1991) and a fully convincing eco­ nomic case can be made for them in Britain est cover in Britain. Where an uneven-aged (see Appendix 3). system will produce high-quality timber at a regular and sustained level without a large 2. Importance of the starting point. The prac­ increase in costs, it may appeal to the State tice of conversion much depends on the and other owners. This will also be the case starting point. Some foresters have started where the benefits are high, if it is possible from (a) uniform even-aged woodland (likely for the owner to capture some revenue from to pay for itself); others from (b) unmanaged them as, for example, from commercial woodland or (c) scrub (both bound to be recreation. It should be easier to justify the expensive).

57 Chapter 7 Influence of environmental factors on alternative silvicultural systems and conversion

The term ‘conversion’ is taken to mean a England, approximately 5 years for mid slope change from the even-aged clear cutting sys­ sites in the Scottish Borders, and less than 5 tem to an alternative system more diverse in years for similar sites in north Scotland. stand structure and species. In most conver­ The wind input into a silvicultural system is sions (and in restructuring) the main con­ therefore dependent upon the region in which it is straints are protecting the forest from the rav­ located i.e wind zone. Wind input also depends ages of wind and damaging mammals. At the upon elevation, topographic position, and aspect same time conversions have an impact on land­ relative to the prevailing winds (Quine and scape, nature conservation and recreational White, 1993). The character of the wind, e.g. potential. A desirable element, where possible, gustiness, is conditioned by the local terrain is natural regeneration. upwind, e.g. open water, open country, forest. The transition from one silvicultural system Effect of wind to another is likely to involve significant risk. Where the return periods or average interval The severity of Britain’s wind climate is impor­ between damaging storms are less than or of tant when discussing appropriate silvicultural similar magnitude to the duration of the transi­ systems, many of which originate from conti­ tion period then windthrow must be a serious nental Europe. Britain’s wind climate is domi­ consideration in decision-making. The answer nated by the passage of Atlantic depressions, is to begin early before the trees in the existing whose accumulated influence bring the crops have reached a height that brings with it strongest winds to the north and west of susceptibility to windthrow (see Chapter 3). Britain; less common tracks bring strong winds The windthrow hazard classification (Miller, to southern Britain. The general windiness is 1985) was developed for even-aged stands sub­ higher and the occurrence of damaging winds ject to the clear cutting system. It gives a prac­ more common than on mainland Europe (see tical method of assessing windthrow vulnerabil­ Figure 7.1). ity, and gives reasonable predictions of damage Recent experience in Britain suggests that occurrence for use in management planning; areas affected by gusts of 30-50 m s_1 will expe­ refinements have been made (Quine and White, rience sporadic damage to trees but that this 1993; Quine and Wright, 1993). However, the will rarely total more than 1% of the forest area; application of the current classification to more areas affected by 35-40 m s_1 knot gusts will sophisticated silvicultural systems is likely to be experience significant damage with 1-10% of the limited. While it will still provide valid compar­ forest over 12 m in height affected; areas affect­ isons of site vulnerability, the defined critical ed by 40-45 m s_1 gusts will experience cata­ and terminal heights will not be appropriate. strophic damage with 10-30% or more of the for­ Quine and Miller (1990) derived a set of rec­ est over 12 m in height affected (Quine, 1991). ommendations for silvicultural systems and The return period for gusts of around 35 m s_1 is site type with reference to windthrow vulnera­ 50 years on low lying sites in south-east bility (Table 7.1). Only the broad system

58 -10 -5 0 5 10

Figure 7.1 Wind zonation of Europe in the manner of the zonation of Britain included in the windthrow hazard classification. This is on scale of A (most windy) to K (least windy).

59 groups are covered, as it is impractical to allow must be free of excessive browsing and bark- for the many possible refinements to the sys­ stripping (Gill, 1992). Conversion of an even- tems. Windthrow hazard class (WHC) in this aged stand to the establishment phase is the context is taken to reflect the frequency and most difficult time when coping with deer, strength of damaging winds, although in prac­ because of the sudden change in forest struc­ tice it also reflects soil-induced vulnerability. ture and consequent effects upon deer behav­ The scope for choice of alternative systems on iour. Inadequate recognition of the latter and the high windthrow hazard classes 5-6 is limited, of the requirements of deer can pose severe and is also constrained on the moderate wind­ problems for the forester. A relationship exists throw hazard classes 3-4 because of the serious between forest structure and the population risk of windthrow. On the low windthrow hazard dynamics of all deer species. Sequential classes 1-2, choice of silvicultural systems will be- changes in forest structure are associated with largely determined by other factors, although stand succession from establishment to pre­ perception of risk and recent experiences will thicket, thicket, and maturing and mature undoubtedly influence decisions. stages (Ratcliffe, 1988). On a local scale, per­ tinent issues are the relative palatability of Effect of damaging mammals species, stand vigour (and hence tolerance of browsing), site productivity, availability of Deer alternative vegetation browse and the inti­ Any silvicultural system dependent on regen­ mate distribution of shelter/cover and food eration either by planting or natural seeding supply.

Table 7.1 Appropriateness of silvicultural systems with reference to windthrow vulnerability of sites. (Source: Quine and Miller, 1990) Low hazard Moderate hazard High hazard class sites class sites class sites e.g. WHC 1 and 2 e.g. WHC 3 and 4 e.g. WHC 5 and 6

Clear fell S1 Si S 2&3 without thinning Clear fell S2 S2 U with thinning Uniform S U u shelterwood Group SU u shelterwood Strip S S3 u Wedge S S2 u Selection S SS’ Coppice s SS’ Coppice s U u with standards S: Suitable. 1 Unlikely to be widely applicable; 2consider diversification of age structure towards normality; Requires careful selection and alignment of edges and/or severance cuts. U: Unsuitable.

60 Hill-Tout (1990) explains the factors relating Woods close to, or visible from, main roads or to deer influencing the choice of silvicultural towns or scenic viewpoints have special impor­ systems, especially concerning conifers in the tance as valued landscape features. One of the uplands. Problems addressed and suggestions main reasons advanced for considering the use given for solving them include: fencing or indi­ of diverse structure systems is the effect on the vidual protection is often inevitable for suscep­ landscape. Sudden and drastic change is gen­ tible tree species in vulnerable situations; erally deplored by the public, whether it shooting is often effective in a comprehensive involves afforestation, line thinning, or exten­ deer management strategy; and chemical repel­ sive clear cutting. Gradual change is much lents are generally ineffective. The whole of more likely to be acceptable. Uniform even- deer management should be seen as part of an aged high forest under the extensive clear cut­ integrated forest design strategy incorporating ting system is widely regarded as the least also timber production, nature conservation, desirable for landscape value. The sudden lack landscape and recreational interests. Coupe of continuity of tree cover, and the change to size is pertinent; during their regeneration it is bare ground with branchwood and stumps, are difficult and costly to reduce deer browse and major disadvantages. bark-stripping in small coupes. Shelterwood systems, and certain variants of group systems, achieve a degree of semi-perma­ Grey squirrels nency. They are most acceptable visually where A deterrent to the successful economic growth trees have well-developed crowns and give a of broadleaved stands of some species is the degree of coalescence. Strip systems can cause stripping of bark by grey squirrels. Control is a significant adverse effect: coupes may appear necessary; this is not peculiar to uneven-aged ‘moth-eaten’ or fragmented and the geometric systems though they prolong the period of risk. patterns are difficult to accommodate in the Methods of controlling populations are avail­ landscape, especially parallel linear openings able and have been shown to be effective in perpendicular to the contours. In the single preventing damage if properly applied and car­ tree selection system full permanence of cover is ried out at the right time although it has to be achieved: the appearance as a whole remains recognised that many woodland owners and unchanged; its visual advantages externally farmers neglect to take the necessary action. and internally are retained; and to the casual High costs and many difficulties are experi­ observer the woodland appearance remains enced in growing some species of broadleaves constant. However, some people deplore the in areas when populations of grey squirrel are absence of a distant or even mid-length view. high. Sycamore, beech, ash, birch and Norway The selection system produces the greatest maple are the most frequently and severely small scale diversity which itself may some­ damaged. Oak and sweet chestnut are less fre­ times be monotonous if too extensive from the quently attacked and wild cherry is not often point of view of people walking in the forest, damaged. Of conifers, Scots, Corsican and and the lack of views out; the views internally lodgepole pine are the most frequently and to the outside can be inadequate. Group attacked. Thicket- and pole-stage trees aged systems have a pleasing irregular appearance 10 to 40 years are the most vulnerable. There both during and after the regeneration period; is no evidence that damage is significantly the resulting young stands of trees often have greater or less between silvicultural systems. an age range of 30 years or more. Under the clear cutting system, in order to Effects on landscape produce natural looking woodland landscapes, Woods and forests are among the most attrac­ certain basic design criteria need to be satis­ tive features of a landscape, their significance fied even when using a patch cutting system. being related to location, topography, extent, The key principles are shape, scale, and diver­ outline, species, age, texture and colour. sity, aiming to produce a pattern which reflects

61 the characteristics of the landform (Crowe, woods to be of lower conservation value than 1978; McIntosh, E, 1990; Lucas, 1991; Forestry those with an uneven-aged diverse structure Commission, 1991, 1992; Bell, 1992). Time (Peterken, 1981). While the latter offer the and space separation have to be used to create desired permanence and diverse habitat, uni­ diversity by limiting the area to be clear cut in form even-aged forest under the extensive clear any one place and setting a period before an cutting system can be less conducive to certain adjoining stand is felled and regenerated. The types of wildlife. Stands managed under shel­ foregoing principles also apply to shelterwood terwood systems likewise can be inadequate. systems based on group felling (in effect, felling Coppice and coppice with standards can of small coupes), because of the textural and encourage a rich woodland fauna and flora, tree height variations between management depending on the age of the trees. Being decid­ areas. Both the external and internal appear­ uous, they can be dull and gloomy in winter, ance of the forest should be considered in the but in spring the field layer of flowering plants design process. McIntosh (1991), the Forestry can be very rich particularly with bluebells, Commission (1991, 1992a, 1992b) and Bell wood anemones, oxlips and celandines. The (1992) give landscape recommendations for practice of coppicing develops a distinctive each silvicultural system. The course to reach flora and fauna; with the repeated cycle of sun­ an acceptable solution will be by the recogni­ light and dense shade providing ideal condi­ tion and application of good design standards. tions for both sun-loving and shade-tolerant Critical areas in the landscape, such as along species. There is a regular cycling of stored public access routes and on hilltops and sky­ seed banks of herbs. lines may need special attention to avoid nega­ The single tree selection system is recom­ tive and intrusive effects. mended for its continuity of habitat conditions. Landscape will not be the only consideration A possible disadvantage for nature conserva­ affecting the choice of silvicultural system. The tion is the lack of diversity. However, the multi- use of alternatives to clear cutting will not neces­ storied diverse structure is highly desirable for sarily produce an acceptable solution in the land­ some species of breeding birds, and within any scape, and basic principles should be applied at stand there will be small-scale spatial varia­ the outset of the design process. Woods and tions in shading and hence of flora and shrub forests should be viewed as an element in the layer communities. The intimate mixing of old wider landscape and should be designed as an and young trees, and the continuous woodland integral part of it, to respect landscape character. conditions, are ideal for sustaining a diverse Improvements to poorly designed woods and epiphytic flora. In group systems, wildlife con­ forests will not be brought about by choice of sil­ servation is favoured by two attributes of a vicultural system alone. In their Lowland land­ degree of continuity of forest cover and diversi­ scape guidelines the Forestry Commission ty of structure and species. In all systems, the (1992b) give detailed advice on the visual aspects understorey, as well as shrubs, herbs and field of Various options for felling and regeneration. layer— particularly brambles — can lessen dis­ turbance to wildlife when it provides safe cover for mammals and helps to deter people and Impact on nature conservation dogs straying too far from paths. Presence of wildlife in its many forms is a The use of natural regeneration brings major attribute to woods and forests. many wildlife conservation benefits, favouring Silvicultural systems have a great impact on species already on the site, tending to generate mammals, reptiles, butterflies, insects, birds mixed stands, and creating a more diverse and plants. Protecting sensitive habitat areas structure than plantations. The selection, is an important part of maintaining diversity. group shelterwood, and coppice systems are Some ecologists, as well as some discerning lay increasingly being appreciated by conserva­ people, have considered uniform even-aged tionists. Forest nature conservation guidelines

62 (Forestry Commission, 1990) is a highly bene­ ation guidelines (Forestry Commission, 1992a) ficial publication. The relevant treatment of provides some essential reading. In general ancient and semi-natural woodlands, which any silvicultural system which gives due atten­ contribute in a significant way to the welfare of tion to landscape and nature conservation is wildlife, is discussed in the Guidelines for the also likely to provide acceptable conditions for management of broadleaved woodlands and recreation potential. The management of semi-natural woodlands (Forestry Commission, 1985, 1994). Conclusions on the influence of environmental factors Effect on recreational potential 1. Effect o f wind. The wind input into a silvi­ Woods and forests are a major asset for outdoor cultural system is dependent upon the recreation, their importance depending on region in which it is located and local topog­ extent, age, structure, species mix, internal raphy. As the transition from one silvicul­ visual quality, accessibility, nearness to urban tural system to another involves significant areas, and tourism in the locality (Broadhurst, risk, windthrow should be given serious con­ 1993). Public interest in forest recreation has sideration in selecting appropriate sites and built up over the last few decades and will con­ systems. tinue. Access must ensure that it is compatible with environmental and wildlife protection 2. Effect of damaging mammals. Adequate recognition of deer behaviour and require­ and the interests of all users of the forest. The ments are vital during conversion, with deer most popular and compatible forms are those described as informal, that is, they depend for management being seen as part of an inte­ grated forest design strategy. To avoid dam­ their appeal on natural factors. Extensive age by grey squirrels methods of controlling areas of clear cutting are significantly disliked. Silvicultural systems based on at least a populations are available and effective if properly applied. degree of permanence of parts of the upper canopy and with varied height and structure of 3. Impact on landscape, nature conservation stand appear to find most favour. Despite all and recreation potential. A detailed assess­ the foregoing, some individuals prefer open ment is necessary of the impact of silvicul­ access to forest using a clear cutting system tural systems and operations. The nearer rather than being restricted to forest using the approach to achievement of continuous alternative systems; their highest priority forest cover, the more is the likelihood of sat­ being open access. Evaluation of recreation isfying aesthetic and environmental objec­ has been undertaken by Benson and Willis tives. Although restructuring (see Chapter (1992) and useful consideration given to the 5) provides an improved landscape and subject by Ruffell and Hanley (1993). slightly greater diversity than the original Management for recreation is discussed by planting, it is still generally based on the Hart (1991) and Hummel (1992). Forest recre­ clear cutting system.

63 Chapter 8 Conclusions: resources and benefits, opportunities and recommendations

General Weasenham, has been continuity of owner­ ship and management. Large areas of forest • Britain’s woods and forests are a valuable are managed by the State and private com­ renewable natural resource, producing tim­ panies who frequently transfer personnel to ber and at the same time providing impor­ provide breadth of experience and flexibility. tant non-wood benefits. The public are Such movement of staff is not conducive to demanding that greater emphasis be placed developing the skills for conversion to irreg­ on the latter in forest management. There is ular forestry, or maintaining such systems. growing concern to create ‘near to nature’ Costs of management of irregular forestry forest, and advocation of an extension of sil­ are thought to be high; but this is not so if a vicultural systems giving continuous forest value is placed on permanent amenity, recre­ cover. This has led to renewed interest in ation and the provision of habitat for irregular uneven-aged forest, as opposed to wildlife. (Irregular forestry and the selection the extensive clear cutting system normally systems are clearly the better providers of adopted in larger woods and forests in these benefits.) Britain. The former is claimed to achieve • There is a need for thought to accommo­ ‘non-wood’ benefits better than uniform date changes to irregular systems in the even-aged forest. Debate continues as to the economics of forest management. There scale and kind of adjustment desired to will be financial gains and losses between meet the demands for more beautiful, more even-aged and irregular systems especial­ varied, more diversified and more wildlife- ly during the conversion phase; the values rich woods and forests. of the relevant costs and benefits are vari­ • A forester skilled in operating the clear cut­ able and uncertain. More research is ting system may not be the most suited to required to decide if a fully convincing eco­ the slower pace of regeneration and smaller nomic case can be made for irregular sys­ scale of the alternative systems, or be best tems to be more widely used. Such systems prepared for some of the setbacks which do appear to be worth while and affordable may occur when attempting conversion from on smaller areas for certain objectives of even-aged to irregular systems of manage­ high multi-purpose use, and sometimes in ment. The forester must be innovative, sheltered upland sites with a moderate patient and opportunistic in gaining silvi­ extent of well-drained soils. There is a cultural and management advantages. If strong case for them in areas of high recre­ such attributes are not provided continually ational use. the new stands may soon revert to a uniform • As forest conditions on the Continent are canopy. often different to those in Britain, the alter­ • A key factor in the success of examples of native systems should not be an exact model irregular forestry in Britain, notably at for Britain; they should be used intelligently

64 as useful guides, as in many of the examples 1. Failures have occurred through (a) group cited in this Bulletin. However, more use of clearings being too small, (b) mixtures over­ natural regeneration is justified, especially complex, (c) inadequate control of deer with broadleaves in lowland Britain and (sometimes in the interests of sporting) and similarly with birch and Scots pine on north­ grey squirrels, and (d) lack of appreciation of ern and eastern heaths, Douglas fir in shel­ need for lateral light and for sustained tered locations, and Sitka spruce in uplands. action to develop the groups (e.g. ‘relief’ Suitable systems (shelterwood uniform and thinning on their edges). shelterwood group) and mosaics of small 2. Foresters can be unsuccessful if they (a) patch clear cutting which are naturally begin conversion too late, (b) wait too long regenerated would greatly soften the impact for natural regeneration and unduly delay of the restocking stage on conservation and planting, and (c) allow expanding group landscape. These systems can be applied regeneration to seed the whole stand, and, with suitable variation to broadleaves, by not reserving groups of large sized trees, conifers and mixed woods, and adapted to risk reversion to even-aged structure; con­ shade-bearers and light-demanders. tinual attention is essential. 3. Adequate access is required for the frequent Experience in Britain small scale cutting and tending operations needed to operate the systems. • Several examples of conversion to alterna­ 4. Deer, grey squirrels, hares and rabbits can tive systems or adaptations of them are seriously damage all systems. A combina­ available for study. Some have reached a tion of effective fencing and control of ani­ halfway stage, some are showing promise, mal numbers is essential. but it may be too soon to be certain that the desired objectives will always be fully 5. Conversion to, and management of, uneven- attained. Regrettably, adequate records are aged systems can have some problems and lacking (exceptions are the Ipsden Estate can generate additional costs. and Glentress). Foresters should make 6. There has to be investment in staff training themselves more familiar with these demon­ and continuity of staff is highly desirable. strations and other examples should be Frequent changes of ownership and objec­ introduced and research undertaken. tives are adverse. • Notable examples of the group shelterwood system exist in the lowlands involving nat­ Opportunities in Britain ural regeneration, both conifer and • It is possible to achieve an uneven-aged broadleaved. In the uplands, use of natural structure on a range of sites in Britain regeneration has been less significant, and (Paterson, 1990) using the shade-tolerant then mainly of Scots pine or Sitka spruce but species that can be grown here and which there are relatively small scale examples of regenerate satisfactorily on certain site western hemlock (Kyloe), Douglas fir (Coed y types. Britain’s woods and forests, especial­ Brenin, The Hermitage, Craigvinean and ly in the uplands, are very varied in land Faskally), and Sitka spruce, Norway spruce form, climate, soil type, vegetation type and and larch (Glentress). The longer each sys­ tree species mix: a variety encouraging dif­ tem has been employed, the more depend­ ferent approaches to stand structure and able natural regeneration has become as the diversification. As a result of its multi-pur­ means of restocking. pose objectives, Forest Enterprise is identi­ • The examples in Britain have taught sever­ fying parts of its forest to be managed as al lessons, significantly: permanent structures.

65 • There will be differences between the type of patch clear cutting and restocking, or irregular systems and management appro­ restructuring, are procedures which can help priate for (a) extensive forests on upland to diversify even-aged structure to meet mul­ and lowland heaths chiefly comprising tiple objectives. There may be scope for using monocultures of uniform even-aged conifers, group or wedge shelterwood systems and nat­ and (b) traditional small broadleaved or ural regeneration (e.g. in areas of WHC 4) in mixed species stands in the lowlands. In the the conversion of these plantations. first case, where most forests have been • In the uplands, there are areas with well- developed as even-aged plantations on a drained soils suitable for some form of alter­ large scale, foresters will continue to mainly native to the clear cutting system, for exam­ use simple but less extensive methods, e.g. ple shelterwood systems on low windthrow shelterwood systems as at Windsor Forest. hazard class sites (WHC 1 and 2), or strip or By contrast, in the more populated low­ wedge systems on higher wind hazard sites lands, foresters must seek an intimate bal­ (WHC 3 and 4). It is essential to match ance between many objectives, i.e. commer­ appropriate silvicultural systems with sites cial, amenity, conservation and field sport. of different windthrow hazard class (see This may require more intensive silvicul­ Table 7.1). Detailed advice on this point is ture and management, and may be more given by Paterson (1990): see Table 8.1. appropriate for systems such as group selec­ • In the lowlands, conversion to group selec­ tion or single tree selection. tion or shelterwood systems could be appro­ priate for broadleaved and mixed stands, • Helpfully, there is already some unevenness especially in sensitive locations, mainly by of structure/development stage among the post-1950 growing-stock in the uplands traditional private sector woodland owners who appreciate and desire the relevant ben­ because of planting dates, species differ­ efits. The development of many variants can ences, provenance differences, variations in be expected. There is also ample opportuni­ site quality, yield class, ‘check’ and sporadic ty for use of the uniform shelterwood system windthrow. in plantation pinewoods, particularly on the • Furthermore, advantage can be taken of the heaths and breckland of lowland England. ameliorated microclimate and improved site Although the system results in even-aged conditions created by the extensive produc­ stands, it is a much more environmentally tive, mainly coniferous, plantations; these acceptable method than clear cutting during conditions now permit the regeneration of a the regeneration stage. wider choice of species, provided that high • Only where significantly advanced uneven- standards of protection from browsing ani­ aged diverse structure is available will mals are rigorously pursued. Moreover, the attempts be undertaken to create the ‘true’ mid-rotation stands provide opportunities to single tree selection system, which needs convert some of them from a uniform to a fertile soils, sheltered sites, a higher inten­ diverse structure, which may lead to ‘semi­ sity of access for extraction and a higher permanent’ or continuous forest cover. proportion of shade-bearers. • On windswept upland sites, the clear cutting • In conifers, Paterson (1990) considers that silvicultural system, despite its drawbacks to about 25% of existing upland conifer planta­ aesthetic and conservation interests, remains tions, located mainly on brown earths, pod- the most effective system suitable for the zols and ironpan soils, are likely to be suit­ light-demanding species, mainly Sitka able for conversion to alternative systems. spruce, extensively planted to produce timber (There will be additional potential following at competitive prices. Opportunities for irreg­ an extension of conifer forests ‘down-the-hill’ ular forestry are constrained in these situa­ to better soils and site conditions.) Quine tions by high windthrow hazard class, but (1991) appears to support Paterson’s view as

66 Table 8.1 Possible silvicultural systems for various site groups in conifer upland forests. (Source: Paterson, 1990, amended) co (S' £ .c "O . 5 j§ *•§ £ Q. :0 52 1 .O T3 S' t i f 52 E Sis 0 o> 2 & 0 0 o 0 CO a? g> o co co

o O)

CD o Q "O Q o in O CM CM O o 3 CO 0 0 o o O O) V 3 CO CO z < o o m co m E X v + E Q B TJ CL £ - o o m CM in c ° D)0 io ^ o I o S O 3 c o o. 0 _r r i_ CO 0 0 0 3 £ v E s: -9 ^ 2 to 5 0 0 o O) o

Z 0. CM < ■a o o o in o ib 0 co w CO o 5 CO CO V 1 CO 1 >.jg + < B CL N CO E Q J o ^ cm C Q. CO ■C "0 ■D O CM CO Tf up Q. £ £ o o 0 0 O) io 1 3 01 o E W 0 1 £ 0 o-g .c 3 - ^ _- 3 C0 c E

^ CL -9 2 to 0 5 0 0 2

< .c £L CO CO E Q J = O y= CM o CM d CM CM -c ~ 52 o O Q. c 0 .c O 3 0 co co i_- 3 0 co 0 0 52 5 ^ o | o 3 E *0 -9 to 5 0 0 c-

z < ro — 5 0 ■g CO 5 (D CM w 52 CM h- CO Q. 8- (o - d 3 g 0 « + o) i o) o 0 o 0 o .c 0 5 . a CO 3

< CO *o CO CO CM r^- + 0 0 5 CO < d in O CO CO CL o ±s o in + c CO < co in j CO XZ 0. *5 o in in CT> c co O W . Q 3 £ o 0 + CO . * Q. CO 0 .9 0 * C 0 o l

< CO CM in CL CO O jn CO ts 0. Q_ d o> in o 'M- CO 0 ‘ Q. + CO c 0 9 E 2 c o 0 0 0 0 * . J Q. 0 . Q m o

73 .o o 0 0 CL 0 c 0 2

< in co CL .c 52 CO £L H d co in o + o CO O £ CO o * CO Q. ® 2 -o 0 o 0 0 0 5 0 <■>

* 0 >»

67 BE: brown earth, PZ: podzol, IPS: ironpan soil, SWG: surface-water gley non-peaty, PG: peaty gley, PIPS: peaty ironpan soil, PT: peat. A: planting, N: natural regeneration, S: length of strip variable. there are many sites in the lower windthrow able for specific objectives by an earlier con­ hazard classes and on better soils where version of existing pole-stage/mid-rotation endemic damage from normal winter storms stands (especially of seedbearing age) on will only rarely be experienced. suitable sites in appropriate locations, tak­ • Broadleaved species in the lowlands offer ing a longer time but achieving a wider much scope for application of diverse struc­ range of size classes. It entails some sacri­ ture systems, especially with the aid of nat­ ficed revenue from premature felling. ural regeneration, and can provide or • The period of conversion for conifers should enhance non-wood benefits. Notable exam­ not exceed 30 years because of cost and the ples have been recorded in Chapters 3 and 4. need for continuity of management; some Variants of group selection can be applied on foresters suggest this is too short, and rec­ sites with soils capable of supporting good ommend 40-50 years. Others point out that growth of oak, beech, ash, wild cherry, pine and larch can easily stand for 90-120 sycamore and birch. The windthrow hazard years, and broadleaves 120-140 years. If for needs generally to be low. Most species can be some reason conversion is delayed, it is gen­ accommodated by adjusting the size of the erally because some difficulty has appeared groups in the range of 0.3-1.0 ha. Small groups that was not foreseen at the outset. will accommodate beech, medium size • It is probably unrealistic to attempt complex sycamore and ash, and large size oak and conversions unless conditions of ownership, birch. Rabbits, hares, grey squirrels and deer forest management, existing growing stock, must be controlled. Such potential will appeal and site are very favourable. Although care­ to owners whose commitment to amenity, fully constructed conversion schemes (e.g. sporting and public recreation is strong. the Bradford-Hutt Plan at Tavistock) are Shelterwood systems using broadleaves need commendable, more simple procedures have to be on deep soils and on wind-stable sites of also proved successful. It is to be expected low risk windthrow hazard (WHC 1 and 2). that the majority of conversions will rarely They can be successful with oak, beech, ash, exceed 100 ha, and that rotations, whether sycamore and birch. The single tree selection nominal or formal, will be in the range of 80 system, when using broadleaves, is difficult to to 120 years for broadleaves, 50 to 90 years achieve, though at one time successful at for pine and larch and 50 to 70 years for Rossie Priory. In all such systems using other conifers. However, rotations over 80 broadleaves, the benefits of an understorey years may not be financially viable. (distinct from a young crop) are important • Associated interests or activities are needed (Everard, 1985; Reade, 1991), suppressing to provide income which accrues directly or branches and epicormic shoots on the stems of indirectly to the irregular forest. Particularly favoured trees, and protecting the soil. in the private sector, fine landscape, beautiful dwellings, an historic monument, or out­ Conversion standing recreational or wildlife features • To create, by planting groups, a stand to be should generally be present to enhance managed under the selection system or potential income or justify in some other under any other irregular system, would way the increased efforts devoted to irregu­ take many decades and generate prohibitive lar forest. costs. The creation of group selection and • During the period when an area with its shelterwood systems from middle-aged reg­ existing stands is being converted to a new ular stands on well-drained soils is a better system, particularly for continuous cover, approach. It is necessary to use heavy crown the forester must concentrate on restraining thinning in the run up to regeneration. costs because benefits are likely to be few in However, irregular structure is also achiev­ the early years. The best that can be hoped

68 for is that income and expenditure are kept Summary of benefits of in balance. When the structure of the whole forest takes shape the benefits will increase, alternative systems and at this stage it will be possible to • Uneven-aged systems are claimed to have appraise the financial conditions needed to benefits which could make them more make the new system permanent. Until that important for Britain in future. The follow­ stage is reached, estimates of the long-term ing environmental and silvicultural benefits costs and benefits are likely to be unrealis­ are claimed as more readily obtained, com­ tic. The work done during conversion will pared with the clear cutting system. (They not be wasted: this is a suitable point in chiefly refer to the group selection and sin­ time at which to consider whether the case gle tree selection systems and to conifers in remains strong for continuation with the the uplands.) proposed system. Environmental benefits • When the conversion is approximately 40% complete, income from the forest should be 1. Less change in internal and external land­ significant: based on the fact that the main scape and a more natural and diverse sources of income from the forest itself will appearance which is better habitat for arise from thinnings and large diameter plants, animals, birds and other wildlife, and fellings, and from other charged-for goods often a greater variety of them. Pleasant sur­ and services including recreation. roundings for field sport and recreation. 2. Site productivity improved as a result of soil • The timber and the non-wood values of the conservation and maintenance of stability of uneven-aged shelterwood, single tree selec­ the ecosystem. Soil temperatures increased, tion and group selection systems appear to prolonging growth and biological activity. be comparable. The experience gained so far Soil protection against erosion, particularly is greatest with the group selection and on steep slopes. Efficient use of light group shelterwood systems, which have through absorption by a number of photo- worked well both in the less exposed synthetically active canopy layers, increas­ uplands and the lowlands, and can accom­ ing the productivity. A stable ground water modate the requirements of a wide range of regime, with no sudden movement of sedi­ species, from light-demanding to shade- ments, and no significant change in flow enduring. The appearance of woods and rate and water quality. forests run on these systems in Britain can­ not be claimed as equal to Continental Silvicultural benefits stands because they have been developed under different conditions. 1. When regenerating by planting, the lower concentration of fresh felled stumps may limit the population of pine weevil and black • It appears probable that a good route to pine beetle, thus reducing damage to the uneven-aged forest is through periodic young plants. Damage by insects and dis­ underplanting (plus natural seeding) of parts of heavily thinned mid pole-stage ease is theoretically reduced, but manageri­ stands (e.g. larch, Scots pine) to produce al practices are most important in determin­ ing the level of damage. multi-storied high forest. A good structure for the group selection system consists of a 2. Where natural regeneration is adequate, mosaic of even-aged stands of varying there is no need for conventional ground species, ages and extents (this is the preparation, planting and beating up; and essence of Garfitt’s approach; likewise at the cost of weeding can be reduced (although Longleat). respacing and cleaning are necessary).

69 3. Tree stability is improved and there is less higher value than many of those in the clear risk of snowbreak. The relationship between cutting system. However, this may be offset stand structure and wind stability depends by the proportion of timber produced from largely on the soil type, ground preparation, the upper part of large dominants and edge the thinning regime and the manner in of group trees which are growing free of which ‘forest edge’ is exposed and the stand competition and therefore producing a high­ prepared for it. Anchorage characteristics er degree of knotty timber. are probably improved by the rougher 3. Harvesting: this may be scattered leading to canopy of diverse structure stressing trees higher cost operations, and constraints on and reducing height/diameter ratio com­ the use of machinery; and supervision needs pared with even-aged. to be more intensive. Costs per cubic metre 4. The claim that irregular forestry leads to a tend to fall considerably as trees increase in reduced fire risk is difficult to verify as most size; the mean volume of trees removed from Continental systems are used in areas of low some forms of uneven-aged forest (in partic­ fire risk. ular, selection forest) is likely to be greater 5. Deer population may be on a more stable than that of trees grown under an even-aged and predictable level, due to an absence of system. Sorting of mixed size assortments the otherwise typical peaks and troughs and species is more costly. arising from violent changes in browse and 4. Profitability: much depends on species and cover. Control may be easier as a result of site conditions and the value attributed to less disturbance of the habitat and range. non-wood benefits. Some factors may make However, appropriate control is essential significant contributions to profitability. over the whole area. Management of uneven-aged systems (espe­ cially where flora and fauna are important Management economics of along with timber production) generally requires a high level of skills, and adequate alternative systems knowledge of structure of stands to ascer­ • Conversion of even-aged to uneven-aged tain and undertake timely silvicultural systems may involve loss in timber revenue treatment. There may be a regular cash flow through premature felling or deferred in uneven-aged forests from sustained yield income from prolonged retention, or both. of marketable timber. However, it is wrong Relevant factors include: to suggest that uneven-aged structure on a 1. Timber yield: whereas in the clear cutting large scale is the sole route to sustained system, conventionally there is a periodic income; this could be attained by a collection series of thinnings from approximately age of even-aged stands giving a normal forest. 20-30 in conifers, in selection forest there is Some present day management, financial usually a more consistent and higher aver­ and industrial factors obstruct the use of age tree size harvested, and hence of higher alternative silvicultural systems, for exam­ value, but there is a wide assortment range ple instability of ownership/supervision, and difficulties with many species in a batch emphasis on short-term income, and con­ of produce. It is difficult to calculate the tracted commitments to supply to the small potential yield of timber for uneven-aged roundwood market, and even the grant-aid stands, without relevant management mod­ structure. els but little difference in gross yield • The clear cutting and replant system will appears likely. remain appropriate in many places, but 2. Timber quality: the canopy, being more open there will be an increasing use of alternative contains a greater number of trees of domi­ silvicultural systems, especially in areas nant, large diameter category, usually of which are visually and biologically sensitive

70 and where the risk of windthrow is not too • The growing-stock contains trees of good great. growth and quality. • The different age or size classes essential to Postscript sustained yield are developing. • The external and internal appearance of the The conversion of unmanaged or derelect stands is pleasing. woodland, and of even-aged plantations, to a more uneven-aged forest, often of several In short, the examples show the expected ben­ species, has been described in this Bulletin efits of careful silviculture with a clear environ­ from 44 different sites in England, Wales and mental objective. They are modest attempts to Scotland. The silvicultural systems used in 15 create ‘forests with true structure and diversity cases are uniform or group shelterwood, in 17 that more accurately mimic natural and sustain­ cases group selection, in 4 cases single tree able [forest] ecosystems’ (Whitfield, 1993). selection, and in 8 cases various other systems There are uncertainties that must still be including the technique called ‘restructuring’. resolved. The lack of information concerning vol­ The pioneer conversion of one stand (New ume increment and value of the benefits obtained Wood, Weasenham Estate) is virtually com­ from uneven-aged stands in Britain, and of the plete; that of several others is at a halfway costs of conversion to gain the desired benefits, stage or slightly more advanced; while the con­ makes the collection and analysis of such infor­ version of the remainder began more recently. mation essential. Recommendations are made for this in Appendix 3. The sustained effort of owners and their Finally, appreciation is recorded to the owners foresters during several decades has produced and foresters who have explored for everyone’s many of the desired ends: benefit the idea of continuous cover forestry in • Natural regeneration is often present. Britain during the past 40 years.

71 REFERENCES Broadhurst, R. (1993). Forests, woodlands and Aldhous, J.R. (1981). Beech in Wessex - a per­ people’s preferences. Countryside Recreation spective on present health and silviculture. Network News No. 1, 6-7. Forestry 54, 197-210. Brown, J.M.B. (1953). Studies on British Anderson, M.L. (1930). A new system of plant­ beechwoods. Forestry Commission Bulletin ing. Scottish Forestry 44, 78-87. 20. HMSO, London. Anderson, M.L. (1931). Planting in dense Buckley, G.P. (ed.) (1992). Ecology and man­ groups spaced at intervals. Quarterly agement of coppice woodlands. Chapman & Journal of Forestry 25, 312-316. Hall, London. Anderson, M.L. (1951). Spaced group planting Cawdor, Lord (1985). Historical notes on man­ and irregularity of stand structure. Empire agement (of Cawdor Wood). Cawdor Archives. Forestry Review 30, 328-341. Chilterns Standing Conference (1971). A plan Anderson, M.L. (1953). Plea for the adoption of for the Chilterns. Buckinghamshire County the standing control or check in woodland Council, Aylesbury, Bucks. management. Scottish Forestry 7, 38-47. Chilterns Standing Conference (1988). A Anderson, M.L. (1960). Norway spruce-silver plan for the Chilterns woodlands policy. fir-beech mixed selection forest. Scottish Buckinghamshire County Council, Ayles­ Forestry 14, 328-341. bury, Bucks. Anderson, M.L. (1967). A history of Scottish Chilterns Standing Conference (1993). forestry. Nelson, London. Introduction to the management of small Bell, S. (1992). Landscape design for irregular woods in the Chilterns. Buckinghamshire forestry. Continuous Cover Forestry County Council, Aylesbury, Bucks. Newsletter 2, 6-7. Crowe, S. (1978). The landscape of forests and Benezit, J.J. (1991). Soils and natural regener­ woods. Forestry Commission Booklet 44. ation in Normandy. Quarterly Journal of HMSO, London. Forestry 85, 30-36. Crowther, R.E. and Evans, J. (1986). Coppice. Benson, J.F. and Willis, K.G. (1992). Valuing Forestry Commission Leaflet 83 (second edi­ informal recreation on the Forestry tion). HMSO, London. Commission estate. Forestry Commission Edwards, I.D. (1981). The conservation of the Bulletin 104. HMSO, London. Glen Tanar native pinewoods, near Aboyne, Biolley, H. (1920). The planning of managed forests Aberdeenshire. Scottish Forestry 35, 173-178. by the experimental method and especially the Evans, J. (1984). Silviculture of broadleaved check method. Translated by M. L. Anderson woodland. Forestry Commission Bulletin (1954). The Schrivener Press, Oxford. 62. HMSO, London. Blyth, J. (1986). Edinburgh University experi­ Evans, J. (1988). Natural regeneration of mental area, Glentress Forest. Internal broadleaves. Forestry Commission Bulletin Report, Department of Forestry and Natural 78. HMSO, London. Resources, University of Edinburgh. Everard, J. (1985). Management of Blyth, J. and Malcolm, D.C. (1988). The devel­ broadleaved forests in Western Europe. opment of a transformation to irregular for­ Forestry Commission, Edinburgh. est: 35 years’ experience at the Glentress Everard, J. (1992). Natural regeneration of oak Trial Area. O.F.I. Occasional Paper No. 37. in Normandy. Quarterly Journal of Forestry Oxford Forestry Institute, Oxford, 33-41. 86, 173-180. Bourne, R. (1945). The neglect of natural Fairbairn, W.A. (1963). Some observations on regeneration. Forestry 19 33-40. group regeneration. Forestry 36, 113-123. Bourne, R. (1951). A fallacy in the theory of Farmer, R. (1991). Achieving multi-benefits in growing stock. Forestry 24, 6-18. upland forestry. ICF Wales and Marches Bradford, Lord (1981). An experiment in irreg­ Regional Group Discussion, Llandrindod ular forestry. Y Coedwigr 33, 1, 6-18. Wells, 29 November 1991.

72 Forestry Commission (1985). Guidelines for the Gibson, W.N. (1953). Exotic conifers and their management of broadleaved woodland. importance to forestry in Northern Ireland Forestry Commission, Edinburgh. and the West of Scotland. Scottish Forestry Forestry Commission (1990). Forest nature 7, 3-10. conservation guidelines. HMSO, London. Gill, R.M.A. (1992). A review of damage by Forestry Commission (1991). Community mammals in north temperate forests. 3: woodland design guidelines. HMSO, Impact on trees and forests. Forestry 65, London. 363-88. Forestry Commission (1992a). Forest recre­ Goss, D.W. (1992). The silviculture of oak in ation guidelines. HMSO, London. Scotland. Scottish Forestry 46, 212-222. Forestry Commission (1992b). Lowland land­ Greig, D.A. (1990). The lowland pine forests of scape guidelines. HMSO, London. England and Wales. In Silvicultural sys­ Forestry Commission (1994a). Forest land­ tems, ed. P. Gordon. Proceedings of a scape design guidelines, 2nd edition. HMSO, Discussion Meeting, University of York 6-8 London. April. Institute of Chartered Foresters, Forestry Commission (1994b). The manage­ Edinburgh, 139-146. ment of semi-natural woodlands. Forestry Hart, C.E. (1966). Royal forest: A history of Practice Guides 1-8. Forestry Commission, Dean’s Woods as producers of timber. Edinburgh. Clarendon Press, Oxford. Garfitt, J.E. (1953). The rehabilitation of dev­ Hart, C.E. (1991). Practical forestry for the astated woodlands. Forestry 26, 28-32. agent and surveyor, 3rd edition. Alan Sutton Garfitt, J.E. (1963). Treatment of natural Publishing, Stroud. regeneration. Forestry 36, 109-112. Helliwell, D.R. (1982). Options in forestry. Garfitt, J.E. (1966). Edge-effect. Forestry 39, Packard, Chichester. 189-190. Helliwell, D.R. (1985). The need for an experi­ Garfitt, J.E. (1977). Irregular forestry in the mental study of different silvicultural sys­ service of amenity. Quarterly Journal of tems. Scottish Forestry, 39, 9-12. Forestry 71, 82-85. Hibberd, B.G. (1985). Restructuring of planta­ Garfitt, J.E. (1979). The importance of brashing. tions in Kielder forest district. Forestry 58, Quarterly Journal of Forestry 73, 153-154. 119-129. Garfitt, J.E. (1980). Treatment of natural regeneration and young broadleaved crops. Hiley, W.E. (1953). Irregular forestry. Quarterly Journal of Forestry 74, 236-239. Quarterly Journal of Forestry 47, 231-237. Garfitt, J.E. (1984). The group selection system. Hiley, W.E. (1956). Underplanting of Japanese Quarterly Journal of Forestry 78, 155-158. larch. Quarterly Journal of Forestry 50, 189- Garfitt, J.E. (1987). Yield control of irregular 196. woodlands. Quarterly Journal of Forestry Hiley, W.E. (1959). Two-storied high forest. 81, 181-184. Forestry 32, 113-116. Garfitt, J.E. (1988). Irregular systems of silvi­ Hiley, W.E. (1964). A forestry venture. Faber culture for broadleaves and mixed crops: do and Faber, London. they work in an English context? O.F.I. Hiley, W.E. (1967). Woodland management, Occasional Paper No. 37. Oxford Forestry 2nd edition. Faber and Faber, London. Institute, Oxford, 47-51. Hill-Tout, P. (1990). Deer and restocking - fac­ Garfitt, J.E. (1995). Natural management of tors influencing the choice of silvicultural woods: continuous cover forestry. Research system. In Silvicultural systems, ed. P. Studies Press Ltd, Taunton. Gordon. Proceedings of a Discussion Garthwaite, P.F. (1972). Windsor Forest. Meeting, University of York 6-8 April. Internal Report to the Nature Conservancy Institute of Chartered Foresters, Council (unpublished). Edinburgh, 82-84.

73 Howell, B.N., Harley, R.M., White, R.D.F. and Lucas, O.W.R. (1991). The design of forest land­ Lamb, R.G.M. (1983). The Dartington story, scapes. Oxford University Press, Oxford. II. Quarterly Journal of Forestry 77, 5-16. McIntosh, E. (1990). Landscape considerations Hummel, F.C. (1992). Aspects of forest recreation for various silvicultural systems in the U.K. in western Europe. Forestry 65, 237-251. In Silvicultural systems, ed. P. Gordon. Hutt, P. and Watkins, K. (1971). The Bradford Proceedings of a Discussion Meeting, plan for continuous forest cover. Journal o f the University of York 6-8 April. Institute of Devon Trust for Nature Conservation 3, 69-74. Chartered Foresters, Edinburgh, 88-110. James, N. D. G. (1981). A history of English McIntosh, R. (1991). Planning for the second forestry. Basil Blackwell, Oxford. rotation. In Forestry practice, ed. B.G. Jobling, J. and Pearce, M.L. (1977). Free Hibberd. Forestry Commission Handbook 6. growth of oak. Forestry Commission Forest HMSO, London, 211-222. Record 113. HMSO, London. McNeill, J.D. and Thompson, D.A. (1982). Johnston, D.R. (1978). Irregularity in British Natural regeneration of Sitka spruce in the forestry. Forestry 51, 163-169. forest of Ae. Scottish Forestry 36, 269-282. Johnston, D.R., Grayson, A.J. and Bradley, Malcolm, D.C. (1971). Corrour management R.T. (1967). Forest planning. Faber and trial. Scottish Forestry 25, 262-269. Faber, London. Malcolm, D.C. (1992). The development of a Jones, E.W. (1952). Silvicultural systems, 2nd transformation from even-aged plantations edition. Clarendon Press, Oxford. to an irregularly-structured forest. Report to Joslin, A. (1982). Management of broadleaves the Scottish Forestry Trust. School of in the Forest of Dean, with special reference Forestry, Institute of Ecology and Resource to regeneration. In Broadleaves in Britain: Management, University of Edinburgh. future management and research, eds D.C. Matthews, J.D. (1990). The evolution of silvi­ Malcolm, J. Evans, and P.N. Edwards. cultural systems in western and central Institute of Chartered Foresters, Europe. In Silvicultural systems, ed. P. Edinburgh, 53-60. Gordon. Proceedings of a Discussion Kerr, G. and Evans, J. (1993). Growing Meeting, University of York 6-8 April. broadleaves for timber. Forestry Institute of Chartered Foresters, Commission Handbook 9. HMSO, London. Edinburgh, 1-19. Kniichel, H. (1953). Planning and control in Matthews, J.D. (1963). Factors affecting the the managed forest. Translated by M.L. production of seed of forest trees. Forestry Anderson. Oliver and Boyd, Edinburgh. Abstracts 24, i-xiii. Kostler, J. (1956). Silviculture. Translated by Matthews, J.D. (1986). The history and status of M.L. Anderson. Oliver and Boyd, uneven-aged forestry in Europe and Britain. Edinburgh. Proceedings of the discussion group on Lorrain-Smith, R. (1986). Economic aspects of uneven-aged silviculture. Report of a Meeting, uneven-aged forestry. Proceedings of the dis­ Pershore, 29 October 1986, 3-16. (Obtainable cussion group on uneven-aged silviculture. from D.R. Helliwell, Yokecliffe House, West Report of a Meeting, Pershore, 29 October End, Wirksworth, Derbyshire DE4 4EG.) 1986. (Obtainable from D.R. Helliwell, Matthews, J.D. (1989, 1992). Silvicultural sys­ Yokecliffe House, West End, Wirksworth, tems, 2nd edition. Clarendon Press, Oxford. Derbyshire DE4 4EG.) Matthews, J.D. (1991). Silvicultural systems Low, A.J. (ed.) (1985). Guide to upland restock­ for the 1990s. ICF Wales and Marches ing practice. Forestry Commission Leaflet Regional Group Discussion, Llandrindod 84. HMSO, London. Wells, 29 November 1991. Low, A.J. (1988). Scarification as an aid to nat­ Miegroet, M. van (1962). The silvicultural treat­ ural regeneration in the Glen Tanar native ment of small woodlands. Bulletin de la Societe pinewood. Scottish Forestry 42, 15-20. Royale Forestiere de Belgique 69, 437-456.

74 Miller, H.G. and Ross, I. (1990). Management Peterken, G.F. (1981). Woodland conservation and silviculture of the forests of Deeside. In and management. Chapman and Hall, Silvicultural systems, ed. P. Gordon. London. Proceedings of a Discussion Meeting, Peterken, G.F., Ausherman, D., Buchenau, M. University of York 6-8 April. Institute of and Forman, R.T.T. (1992). Old-growth con­ Chartered Foresters, Edinburgh. servation within British upland conifer Miller, K.F. (1985). Windthrow hazard classifi­ plantations. Journal of Forestry 65, 127- cation. Forestry Commission Leaflet 85. 144. HMSO, London. Philip, M.S. (1994). Measuring trees and Neustein, S.A. (1964). Windthrow on the mar­ forests, 2nd edition. CAB International, gins of various sizes of felling area. Forestry Wallingford. Commission Report on Forest Research Price, C. (1989). The theory and application of 1964. HMSO, London, 166-171. forest economics. Basil Blackwell, Oxford. Neustein, S.A. (1970). In Forestry Commission Pryor, S.N. (1985). An evaluation of silvicultur­ Report on Forest Research 1970. HMSO, al options for broadleaved woodland. DPhil London. Thesis, University of Oxford. Newton, J.P. (1986). The suitability of irregu­ Pryor, S.N. (1990). Practical aspects of lar forestry for special areas in Britain - a irregular silviculture for British preliminary study. Unpublished thesis, broadleaves. In Silvicultural systems, ed. University of Edinburgh. P. Gordon. Proceedings of a Discussion Osmaston, F.C. (1968). The management of Meeting, University of York 6-8 April. forests. Allen and Unwin, London. Institute of Chartered Foresters, Paterson, D.B. (1953). A study in stand struc­ Edinburgh, 147-166. ture and management of irregular forests. Pryor, S.N. and Savill, P.S. (1986). Thesis. University of Edinburgh. Silvicultural systems for broadleaved wood­ Paterson, D.B. (1958). A study in stand struc­ land in Britain. O.F.I. Occasional Paper No. ture and management of irregular forests. 32. Oxford Forestry Institute, Oxford. Bulletin 5. Forestry Department, University Quine, C.P. (1991). Windthrow as a constraint of Edinburgh. on silviculture. ICF Wales and Marches Paterson, D.B. (1990). The potential to apply Regional Group Discussion, Llandrindod different silvicultural systems to upland Wells, 29 November 1991. British forests, predominantly of Sitka Quine, C.P. and Gardiner, B.A. (1992). spruce. In Silvicultural systems, ed. P. Incorporating the threat of windthrow into Gordon. Proceedings of a Discussion forest design plans. Forestry Commission Meeting, University of York 6-8 April. Research Information Note 220. Forestry Institute of Chartered Foresters, Commission, Edinburgh. Edinburgh, 120-138. Quine, C.P. and Miller, K.F. (1990). Windthrow Penistan, M.J. (1952). The alternative to - a factor influencing the choice of silvicul­ extensive regular clear felling. Scottish tural systems. In Silvicultural systems, ed. Forestry 6. P. Gordon. Proceedings of a Discussion Penistan, M.J. (1960). Forestry in the Belgian Meeting, University of York 6-8 April 1990. uplands. Forestry 33, 1-7. Institute of Chartered Foresters, Penistan, M.J. (1968). The 75th anniversary of Edinburgh, 71-81. the Societe Royale Forestiere de Belgique. Quine, C.P. and White, I.M.S. (1993). Revised Quarterly Journal of Forestry 62, 304-310. windiness scores for the windthrow hazard Penistan, M.J. (1974). The silviculture of beech classification: the revised scoring method. woodland. In The management of Forestry Commission Research Information broadleaved woodlands. Supplement to Note 230. Forestry Commission, Forestry 47, 71-78. Edinburgh.

75 Quine, C.P. and Wright, J.A. (1993). The effects Taylor, C.J. (1967). Management Plan for of revised windiness scores on the calcula­ Cawdor Wood 1967-1985. Unpublished. tion and distribution of windthrow hazard Department of Forestry and Natural class. Forestry Commission Research Resources, University of Edinburgh. Information Note 231. Forestry Tilney-Bassett, H.A.E. (1988). Forestry in the Commission, Edinburgh. region of the Chilterns. Forestry 61, 267-286. Quine, C. P., Coutts, M., Gardiner, B. and Troup, R.S. (1928). Silvicultural systems. Pyatt, G. (1995). Forests and wind: manage­ Oxford University Press, London. ment to minimise damage. Bulletin 114. Troup, R.S. (1952). Silvicultural systems, 2nd HMSO, London. edition, ed. E.W. Jones. Oxford University Rackham, O. (1980). Ancient woodland. Press, London. Edward Arnold, London. Turner, G. (1959). Note relative a la transformation Ratcliffe, PR. (1988). Protection against ani­ des pessieres en station. Bulletin de la Societe mals and management for game. In Farm Royale Forestiere de Belgique 66, 414-420. woodland practice, ed. B.G. Hibberd. Waters, T.L. and Savill, P.S. (1992). Ash and Forestry Commission Handbook 3. HMSO, sycamore regeneration and the phenomenon London, 46-57. of their alternation. Forestry 65, 417-433. Reade, M.G. (1957). Sustained yield from selec­ Whiteman, A. (1991). The economic justifica­ tion forest. Quarterly Journal of Forestry 51, tion of continuous cover forestry. Paper 51-62. delivered at a Discussion Meeting of the Reade, M.G. (1960). Lessons from Continuous Cover Forestry Group, Brecon Switzerland. Quarterly Journal of Forestry 22 October 1991. 54, 111-126. Whitfield, P. (1993). Forestry in British Reade, M.G. (1965). Natural regeneration of Columbia: a study of environmental issues. beech. Quarterly Journal of Forestry 59, Forestry 67, 103. 121-131. Whitney Mclver, H. (1991). An overview of Reade, M.G. (1990). Chiltern enumerations. uneven-aged forestry. Continuous Cover Quarterly Journal of Forestry 84, 9-22. Forestry Newsletter 1, 4-5. Ruffell, R.J. and Hanley, N. (1993). Whitney Mclver, H., Blyth, J.F. and Malcolm, Recreational use values of woodland fea­ D.C. (1992). The application of group selec­ tures. Forestry Commission Research tion working in an upland forest in south Information Note 229. Forestry Scotland [Glentress Forest], Scottish Commission, Edinburgh. Forestry 46, 202-211. Savill, P. S. and Evans, J. (1986). Plantation Willis, K.G. and Garrod, G.D. (1992). Amenity silviculture in temperate regions. Clarendon value of forests in Great Britain and its Press, Oxford. impact on the internal rate of return from Shrimpton, N.H. (1986). Studies on the eco­ Forestry. Forestry 65, 331-346. nomics of irregular stands. In Proceedings of Wood, R.F. (1950). Rehabilitation of devastated the discussion group on uneven-aged silvi­ and derelict woodlands. Quarterly Journal culture. Report of a Meeting, Pershore, 29 of Forestry 44, 5-10. October 1986. (Obtainable from D.R. Wood, R.F., Miller, A.D.S. and Nimmo, M. Helliwell, Yokecliffe House, West End, (1967). Experiments in the rehabilitation of Wirksworth, Derbyshire DE4 4EG.) uneconomic broadleaved woodlands. Shrimpton, N.H. (1988). Modelling the costs of Forestry Commission R & D Paper 51. uneven-aged forest management. O.F.I. HMSO, London. Occasional Paper No. 37. Oxford Forestry Workman, J. (1986). Experience in the man­ Institute, Oxford, 42-46. agement of beech woodlands. In Proceedings Steven, H.M. (1948). The late Mr Ray Bourne, of the discussion group on uneven-aged silvi­ M.A. Forestry 22, 244-245. culture. Report of a Meeting, Pershore, 29

76 October 1986, 17-20. (Obtainable from D.R. and management of uneven-aged coniferous Helliwell, Yokecliffe House, West End, forest in central Europe and its relevance to Wirksworth, Derbyshire DE4 4EG.) Great Britain. Forestry Commission inter­ Worrell, R. (1991). Trees and the Treasury - nal publication. valuing forests for society. World Wide Fund Yorke, D.M.B. (1992). Pros and cons of the ‘per­ for Nature, Godalming, Surrey. manent forest’. Forestry and British Timber, Yorke, D.M.B. (1991a). Scope for irregularity in October 1992, 20, 22, 24. Amended version upland coniferous forests in Britain. in Scottish Forestry 47, 6-14. Continuous Cover Forestry Group Yorke, D.M.B. (1992). The management of contin­ Newsletter 1, 5-6. uous cover conifer forests - an alternative to Yorke, D.M.B. (1991b). A study of regeneration clear felling. Continuous Cover Forestry Group.

77 Glossary 1

Definition of terms

Advance growth of natural regeneration Coupe The individual area of felling; an area Seedling growth which appears spontaneously of forest which has been felled for regenera­ before the planned regeneration stage in an tion. existing stand; sometimes achieved by reduc­ Crop A silvicultural or management unit of ing the canopy in advance of seeding fellings. young trees homogeneous in one or several Alternative system A silvicultural system respects. Later to develop to a stand. other than that of extensive clear cutting, Diverse structureSee Diversity. patch clear cutting and coppice systems. Diversity The use of mixtures or mosaics of Biodiversity The variability among living different tree species and/or of mixed organisms and the ecosystems of which they height/age structure to provide a variety of are part. (A multi-objective forest policy should habitat and continuously varying canopy. include the aims of maintaining and enhancing Extensive clear cuttingA silvicultural sys­ levels of biodiversity where appropriate in tem involving the complete removal of trees managed forests and in semi-natural wood­ before artificial or natural regeneration over lands. Spatial and structural diversity in extensive areas, ranging from about 30 to 50 forests may provide a surrogate measure of ha or more depending on the location and scale biodiversity.) of the forest. Felling is usually compressed into Canopy The mass of foliage and branches a period not exceeding 3 years. No seed trees formed collectively by the crowns of trees in a are retained. stand. Extraction rack A route created for tending Clear cutting See Extensive clear cutting and stands and extracting felled trees. Patch clear cutting. Femelschlag(German) The irregular shelter­ wood silvicultural system. (‘Femel’ means ‘to Continuous forest coverA general term cov­ select the best’.) ering several silvicultural systems which con­ serve the local forest canopy/environment dur­ Furetage Use of unregulated fellings in cop­ ing the regeneration phase. Coupe size is nor­ pice or mixed forest of selection type, thereby mally below 0.25 ha (50 x 50 m) in group sys­ improverishing the quality of the stand. tems; and in shelterwood - where used - is (Means ‘to rummage’.) retained for longer than 10 years. The general Group (a) An opening or gap made either nat­ aim of all systems within the concept is the urally or by felling; (b) a clump or patch of encouragement of diversity of structure and trees, planted or naturally regenerated. uneven age/size on an intimate scale. (The Irregular (= uneven-aged). Descriptive of a classical silvicultural system to create continu­ multi-storey structure derived from several ous forest cover is single tree selection.) regeneration periods on one area with only Conversion (= transformation) A change from partial removal of older/larger trees so that one silvicultural system to another, usually multiple canopies are present over a small from clear cutting or coppice systems to alter­ area. The area of coupe is usually restricted to native systems. 0.25 ha (50 x 50 m).

78 Naturalness The condition produced by the crop and soil at the time of regeneration, nature in balance at a primitive unexploitable and prepare the site for regeneration. stage. Refers to natural processes as well as Topex An index of topographic exposure native species. (Sensitively managed artificial obtained by measuring the angle of inclination ecosystems using native species can provide to the horizon at eight principal points of the valuable wildlife habitats, simulating natural compass; the measurements are used in the ones.) windthrow hazard classification to derive esti­ Normal forest One in which is present an mates of shelter, aspect and funnelling. See approximately equal area of each age-class up Quine and White (1993). to maturity. Transformation See Conversion. Patch clear cutting An application of the Silvicultural system The process by which clear cutting system in which coupe size is the crops making up a forest are tended, restricted to about 5-30 ha depending on the removed and replaced by new crops, resulting location and scale of the forest. No seed trees in the production of crops of distinctive form. are retained. Stand An area of forest comprising a more or Regular (= even-aged, uniform) Conven­ less homogeneous crop in age, species composi­ tionally, uniform extents of forest above 0.25 tion and condition. The essential management ha (50 x 50 m). unit of silviculture. Restructuring The arrangement and sched­ Windthrow hazard class A classification of uling of fellings and restocking in large even- hazard ranging from 1 (low) to 6 (high) derived aged forest so as to introduce more diversity by combining windiness scores (including and meet environmental objectives such as a topex) and soil score for a site. See Miller landscape design plan. The felling of some (1985) and Quine and White (1993). stands is advanced and of others postponed; Wind zone An estimate of regional windiness and coupes are dispersed over the area, usual­ derived from tatter flags. Site windiness for ly with a specified interval before adjacent the windthrow hazard classification is fellings are made. Also applied to an individual obtained by combining wind zone score with stand. scores representing topographic modifications Tending Cleaning, respacing, pruning and to the regional windiness, i.e. elevation, topex, thinning operations done in an immature crop aspect and funnelling. See Quine and White in so far as these operations affect the state of (1993).

79 Glossary 2

Scientific names and authorities of the main English names used in the text

Broadleaves Ash Fraxinus excelsior L. Beech Fagus sylvatica L. Birch, downy Betula pubescens Ehrh. Birch, silver Betula pendula Roth. Gean (wild cherry) Prunus avium L. Hazel Corylus avellana L. Hornbeam Carpinus betulus L. Lime, large-leaved Tilia platyphyllos Scop. Lime, small-leaved Tilia cordata Mill. Maple, Norway Acer platanoides L. Oak, pedunculate Quercus robur L. Oak, sessile Quercus petraea (Mattuschka) Liebl. Rowan Sorbus aucuparia L. Sweet chestnut Castanea sativa Mill. Sycamore Acer pseudoplatanus L.

Conifers Coast redwood Sequoia sempervirens (D. Don.) Endl. Corsican pine Pinus nigra var. maritima (Ait.) Melville. Caucasian fir Abies nordmanniana (Stev.) Spach. Douglas fir Pseudotsuga menziesii (Mirb.) Franco. European silver fir Abies alba Mill. Grand fir Abies grandis (Dougl.) Lindl. Larch, European Larix decidua L. Larch, Japanese Larix kaempferi (Lamb.) Carr. Larch, hybrid Larix Eurolepis Henry. Lodgepole pine, coastal Pinus contorta Dougl. var. contorta Lodgepole pine, inland Pinus contorta Dougl. var. latifolia Wats. Noble fir Abies procera Rehd. Norway spruce Picea abies (L.) Karst. Pacific fir Abies amabilis (Dougl.) Forbes Scots pine Pinus sylvestris L. Sitka spruce Picea sitchensis (Bong.) Carr. Wellingtonia Sequoiadendron giganteum (Lindl.) Buchholz. Western hemlock Tsuga heterophylla (Raf.) Sarg. Western red cedar Thuja plicata D. Don.

80 Appendix 1

Classification of silvicultural systems

Source: Matthews (1989) Fellings beginning in internal lines and advancing outwards in wedge formation - High forest systems. Crops normally of seedling Wedge system origin. Felling and regeneration distributed continu­ Felling and regeneration for the time being con­ ously over the whole area; crop wholly uneven- centrated on part of the forest area only: aged (irregular) - Selection systems Old crop cleared by a single felling; resulting Accessory systems arising out of other systems: crop even-aged - Clear cutting systems Form of forest produced by introducing a Systems of successive regeneration fellings. Old young crop beneath an existing immature crop cleared by two or more successive fellings; one - Two-storied high forest resulting crop more or less even-aged or some­ Form of forest produced by retaining certain what uneven-aged: trees of the old crop after regeneration is Regeneration fellings distributed over whole completed - High forest with reserves compartments or sub-compartments: Opening of canopy even; young crops more Coppice systems. Crops, in part at least, origi­ or less even-aged and uniform - Uniform nating from stool shoots (coppice) or by other systems vegetative means: Opening of canopy by scattered gaps; young Crop consisting entirely of vegetative shoots: crop more or less even-aged - Group system Crop removed by clear felling; even-aged - Opening of canopy irregular and gradual; Coppice system young crop somewhat uneven-aged -Irregular Only a portion of the shoots cut at each shelterwood system felling; crop uneven-aged - Coppice selection Regeneration fellings confined to certain por­ system tions of compartments or sub-compartments Crop consisting partly of vegetative shoots, at a time: partly of trees generally of seedling origin - Fellings in strips - Strip systems Coppice with standards system.

81 Appendix 2

Summary of factors to consider in choosing silvicultural systems

In Britain, when contemplating silvicultural 4. Soil types systems and their management, the follow­ (a) Soils in relation to stability (best in the ing significant factors or conditions are perti­ lowlands). nent. (b) Freely drained: brown earths and pod- 1. Scale State as compared with private forest zols. sector (the latter containing most of the (c) Imperfectly drained: ironpan soils and broadleaves). Large-scale as distinct from podzolic gleys (e.g. upland heaths). small-scale application: 90% of private sec­ (d) Inadequately drained: surface-water tor woodlands are very small (from 0.1 to and peaty gleys and peats (e.g. wet 100 ha), while the majority of Forest exposed uplands, Wales, the Scottish Enterprise and company forest manage­ Borders and West Highlands of ment units are large or very large (250 to Scotland). 10 000 ha); the larger the scale the more simplistic is the general silvicultural system Well-drained soils present the best prospect adopted. for uneven-aged forestry. 5. Animal damage The likely pressures of 2. Objectives In addition to timber production, deer, grey squirrels, hares and rabbits. The landscape, forest recreation potential and smaller the regeneration unit, the higher wildlife conservation. Where one of the are likely to be the costs of protection. objectives, whatever its ranking, is to create or enhance conditions for sporting and game 6. Species and systems management or shelter, an uneven-aged (a) Conifers vs broadleaves; pure stands vs system is likely to achieve that objective bet­ mixed stands; mixtures provide greater ter than the clear cutting system. opportunities for diverse structure and conversion. 3. Site conditions Location, elevation and (b) Light-demanding species as opposed to windthrow hazard class. shade-tolerant species; this affects (a) Exposed uplands (above 350 m), mainly mainly the size of regeneration unit and in Scotland, Wales and northern and tolerance of an overstorey. some other parts of England. (c) Planting as opposed to natural regener­ (b) Foothills (200-350 m). ation, coppice regrowth, suckering and direct seeding. (c) Lowlands (below 200 m), mainly in (d) High forest systems as distinct from cop­ England, and in parts of Scotland and pice systems. Wales. (e) Stands of even-age as opposed to stands of This particularly relates to species, terrain, uneven-age: in even-aged stands the exposure, aspect, climatic conditions and ground usually becomes bare at the end of probability of windthrow. Scope for uneven- the rotation, whereas in some uneven-age aged forestry is likely to be greatest in (b) stands there is prolonged or continuous and (c). forest cover which conserves a shrub layer.

82 (f) Treatment of overstoreys and under­ vicultural system. The practice of any system storeys of broadleaves (mainly in low­ presupposes at least a general knowledge of land southern England but including the foundations on which silviculture is based, lowlands generally) as distinct from along with some awareness of topics funda­ conifers (mainly in uplands). mental to many and often to all of them: species, forest ecology, genetics, the use and (g) Special treatment is required (and is manipulation of mixtures, the effect of wind, available) for ancient and semi-natural rainfall, disease, insects, mammals and air pol­ woodlands. lution. Always it is essential to match silvicul­ 7. Public pressure , e.g. Lake District in tural system with sites of different windthrow England as compared with Sutherland in hazard class. There is a need for adaptability Scotland. to accommodate prudent changes in forest The factors outlined relate to the choice of sil­ management.

83 Appendix 3

Need for research, trials and demonstration areas

In Britain, research into alternative silvicul­ • Causing minimal disturbance to the forest tural systems and stand structures for varied climate and ecological conditions. forest types has been infrequent, leaving a • Favouring long rotations. major gap in trials and demonstration areas, • Increasing the resistance of the forest to wind. and restricting experience and training in stand management. The costs and benefits of General requirements of research uneven-aged, diverse structure forestry cannot Further research, trials and demonstration be fully assessed, or the systems adequately areas are needed to discover whether a fully planned or managed without further knowl­ convincing economic case can be made for use edge and demonstration (Anderson (1953) and of alternative systems. Specific requirements Paterson (1958) were foremost among foresters of research, particularly in comparing uneven- drawing attention to this). In 1980, a aged diverse structure systems with the exten­ Government Select Committee on ‘Scientific sive clear cutting system, relate to: Aspects of Forestry’ received relevant deposi­ tions, and recommended diversification of • The respective costs and benefits. forest ‘both in age and composition’. • Variation in timber as to yield, size of assort­ Subsequently, the need for trials and demon­ ments, quality and value. stration areas has been advanced by Helliwell • Costs of regeneration and tending. (1982, 1985), Pryor and Savill (1986), Paterson • Costs of harvesting. (Harvesting practice (1990), Yorke (198la,b; 1982, 1992) and the needs to accommodate to silviculture.) Continuous Cover Forestry Group. Costs of sorting mixed species and sizes. Many aspects of silviculture are common to • Costs of management, inventory and super­ all types of systems; but there are many differ­ vision. (A plan is advisable to provide conti­ ences as to species choice appropriate to the nuity of objectives, codified to ensure conti­ site conditions and climate, the landscaping nuity of management.) and conservation objectives, the effects of • Damage by mammals, insects and disease. Cost severe gales, and the problems of damage by differences of protection, biotic and abiotic. insects, diseases and mammals. Systems have to be viable in the prevailing conditions of har­ • Physical and chemical condition of the soil, vesting and marketing. Nonetheless, basic rel­ through time. evant silvicultural principles which will gener­ • Windthrow incidence. Stability and resis­ ally apply to all diverse structure systems tance to storm winds on a stand basis; stem include: taper, crown and buttress differences. • Natural regeneration especially factors • Growing species which will suit the site and climate and predicted markets. relating to the effects of different light lev­ els, and of treatments to the soil surface and • Using natural regeneration where possible the forest layer. (provided the provenance is acceptable). • In groups, the effect of shading and compe­ • Aiming for structural diversity on as small a tition on the growth of trees and of weeds. scale as is compatible with the species, site Thereupon, to suggest optimum size of and the objectives. groups for all circumstances.

84 • Non-quantifiable benefits: amenity, inter­ type, topex and wind zone; and the develop­ nal and external landscape value, nature ment of different practices for each of these. conservation, ‘permanence’, soil and water In conifer forest (mainly uplands) stability and sporting value. • Mensurational and associated factors. There Paterson (personal communication) suggests is need for inventory and periodic measure­ several contexts in which the lack of trials and ment of development of structure (Philip, demonstration areas occur, in the uplands: 1994). Often involved will be changes in the 1. Brown earths on steep slopes: typically in way in which trees are selected, marked, England (the Lakes Forest District), Wales, felled, extracted and marketed. In the clear parts of Scotland (Great Glen, Argyll). cutting system there is a sequence of changes 2. Larch and pine stands on upland heaths which take place over the length of the rota­ suitable for: tion, for which adequate Forestry (a) natural regeneration by uniform, seed Commission management tables and models tree or strip systems (this has parallels are available, but for diverse structure sys­ for lowland pine forests); tems there are no such readily available aids (b) conversion by underplanting to silver firs. (neither for pure stands nor mixtures) and certainly no basis for comparison of silvicul­ 3. Surface-water gleys: peaty and non-peaty: tural systems and site conditions. Stem-num­ (a) large group regeneration (spruce); ber curves for various stand types are a pos­ restricting coupe size and defining Size sible aid. The Methode du Controle or ‘Check’ of group for maximum natural regener­ method may need to be superseded by a pro­ ation contribution in WHC 4 and 5; cedure more suited to conditions in Britain, (b) wedge systems or strip systems (spruce) especially for broadleaves: see Philip (1994). in WHC 3 and 4; stand preparation; dimensions. Research on the above would have a strong 4. Moraine-peat complex (West and North bearing on improving understanding of the Highlands and Galloway): restructuring; management and economics of uneven-aged diverse structure forest; yet some expected restricting coupe size; and exploring use of moraine sites for stabilising coupe margins advantages may not be fully proven for many and for species component mix. decades. (Current trials are only part way through their planned conversion.) State own­ 5. Moorland sites: conversion from lodgepole ership, offering the best possibility of continu­ pine. ity of management, is ideal for research as well 6. Lower slopes (i.e. sites ‘lower down the hill’), as for trials and demonstration areas. Already, ex-agricultural. as a result of its multi-purpose objectives, Paterson (personal communication) further Forest Enterprise is identifying parts of its for­ emphasises the need to marry landscape objec­ est to be managed as permanent structures. tives to the silvicultural systems necessary to achieve them. Of significant interest is to know Planning trials and demonstration (particularly for landscape reasons) the stabil­ areas of alternative systems ity of margins and the matrix left between ini­ tial coupes in WHC 5 and 6, gley/peat/peaty The need for trials and demonstration areas of ironpan soil zones (and even in WHC 4, alternative systems concerns conifers, gley/brown earth zones). Landscape design broadleaves, and mixed conifers/broadleaves, plans appear to envisage margins lasting for at as suggested hereafter. Starting points could least 10 years; if they do not - as is expected - be identification of sites which have the poten­ other stand treatments will have to be devised. tial for alternative systems, depending upon Work on severance thinning below critical broad site classes derived from land form, soil height may be involved.

85 Examples of strip and wedge systems for stand structure would be that several scientific regeneration of Sitka spruce and Norway teams could co-operate around them to study spruce are rare on locally more sheltered gleys effects of lateral shelter on growth, deer control (an exception is the attempt at Fernworthy on problems, internal and external landscape qual­ Dartmoor in England). Further experience is ity, intensity of pine weevil and black pine beetle needed to define widths and techniques in damage, and the benefits for wildlife and vege­ more windy climates. tation compared with similar areas treated Empirical practice in alternative systems under the clear cutting system. Information on appears to arise to meet defined recreational stability and wind behaviour in various stand and landscape objectives of Forest Enterprise structures would fill an important gap, as in or beside Forest Parks, for example: Douglas Britain’s wind climate is so different from that fir group regeneration and larch at Whinlatter, on the Continent of Europe. Operational eco­ in the Lakes Forest District; Sitka spruce nat­ nomics could also be studied by the Technical ural regeneration in coupes of 10-12 ha strip Branch of the Forestry Commission. uniform shelterwood in Hamsterley beside Publication would be welcomed of the Northumberland Park; and small coupes for results from Neustein’s series of Forestry natural regeneration in Douglas fir at Coed y Commission experiments (see Chapter 3) on Brenin in Wales. Research into public prefer­ strip shelterwood and conversion by under- ences could provide guidance to the forester for planting Scots pine and larch stands in both the management of areas much used for recre­ the littoral sand sites and some other better ation including urban conifer woodland. status larch sites, using Douglas fir, Norway Paterson (personal communication) further spruce and Abies spp. This does not detract suggests that it would be venturesome for from the need for setting up contemporary Forest Enterprise, in association with the working examples distributed around Britain Forestry Commission, to identify moderately and including some site types not covered sized blocks larger than Glentress (117 ha), of adequately by the experiments of that period, about P45-P65 in some of the site contexts sug­ e.g. moraine/peat complex, lodgepole pine gested above and to begin to manage them in conversion, and upland spruce on WHC 5 and co-operative plans with the Research Division, 6. exploring group size, group conversion, stand preparation for regeneration, underplanting In broadleaved forest (mainly lowlands) and natural regeneration, as appropriate. The Northern Research Station has unique oppor­ Helliwell (1982, 1985) and Pryor (1985, 1990) tunities to monitor large-scale demonstrations have made relevant suggestions regarding of alternative systems on various forest/site management of alternative systems for types. The range is expressed by Paterson broadleaves. The use of quite small groups (1990): see Table 8.1. The wide territorial dis­ may be adequate for beech but not for light- tribution of the Forestry Commission research demanding broadleaves. Useful work could be outstations forms a useful network to cover the undertaken as to the edge shade tolerance of necessary monitoring. Possible options for oak and other light-demanding species on dif­ large scale research demonstrations of alterna­ ferent aspects in treeshelter experiments tive silvicultural systems for conifers on vari­ which could lead to definition of appropriate ous forest/site types in upland state forests group sizes. Suitable methods for utilising the given by Paterson (1990, amended) are includ­ interplay of regeneration of ash and sycamore ed in Appendix 4 of this Bulletin. This is long­ also need to be explored and seem likely to be term work; a commitment by all parties con­ some form of group regeneration and group cerned needs to be envisaged over 15-30 years structure (Waters and Savill, 1992). Bowhill at least. and Eildon (in south central Scotland), An important benefit of engaging in studies of Ebworth (on the Cotswolds) and other stands

86 (in southern England) provide examples of Commission and Coed Cymru, addressing group selection experience in broadleaved appropriate silvicultural systems for western woodland. semi-natural oakwoods. It would have to be The need for trials and demonstration areas demonstrated that small group regeneration on lowland broadleaved forest (Evans, 1984) for old degraded oakwoods can be successful. are only different for the uplands by the pres­ Birch in Scotland seems flexible to several ence of extensive over-grazed semi-natural approaches— uniform seed tree, strip felling or birchwoods and less extensive upland oak- group regeneration — and may not merit woods. In Wales, initiatives for oakwoods, con­ research effort after protection is achieved cerning group size with or without seed trees, from sheep grazing and deer browsing and are being discussed by the Forestry bark-stripping.

87 Appendix 4

Suggestions for trials and demonstration areas

Possible options for large scale research demonstrations of alternative silvicultural systems for conifers on various forest/site types in upland State forests (Paterson, 1990, amended) Country Forest location Species System trial Research centre

Scotland Ae Sitka spruce (a) Group regeneration Mabie Solway, Ae Sitka spruce (b) Wedge/strip felling Mabie Wauchope Sitka spruce (a) Group regeneration WHC 5 Bush Sitka spruce/lodgepole pine Argyll Sitka spruce/Norway spruce (a) Wedge/strip felling WHC 3 Cairnbaan Glengarry or Shin Sitka spruce/lodgepole pine (a) Conversion of lodgepole pine, Newton (moorland) even-aged and snow-damaged plantations Ichnacardoch (Lon Mor) Mixed conifers (mainly (a) Group structure Newton lodgepole pine, Sitka spruce Anderson and Scots pine) Inshriach Larches, Scots pine (a) Uniform seed tree natural Newton regeneration Speymouth Scots pine (a) Strip felling Newton Tummel Larches, Scots pine (a) Conversion by underplanting Perth/Bush (Forest Park) with Douglas fir, beech, Pacific fir, Norway spruce Montreathmont Scots pine (a) Conversion by underplanting Perth/Bush (Angus lowlands) (b) Uniform seed tree natural regeneration Tummel Douglas fir, larches, Scots (a) Maintain conversion to small Perth/Bush (Faskally) pine, oak, beech group diverse structure (Craigvinean)

Wales Brechfa Larches (a) Conversion to Douglas fir, Talybont Pacific fir, beech Coed y Brenin Douglas fir, (a) Group regeneration Talybont Japanese larch Douglas fir, maple Twyi Sitka spruce (a) Group and strip regeneration Talybont Brecon WHC 3 and 4

England Hamsterley Sitka spruce/Scots pine (a) Group regeneration Kielder Kielder Sitka spruce (a) Group regeneration WHC 5 Kielder Newcastleton Norway spruce/Sitka spruce (a) Wedge/strip felling Kielder WHC 3 and 4

88 Appendix 5

History and Personalities

In Scotland, the first extensive planting by the 1923 the Forestry Commission accepted improving landowners of avenue and park responsibility (James, 1981). In the private trees, shelterbelts and policy woods for chiefly sector during the 19th century some private non-wood benefits took place mainly between woodland owners continued to give attention to 1810 and 1865, financed by profits from farm­ game management and shooting, but also ing, cotton, coal-mining, rents and town prop­ showed an increasing interest in growing exot­ erty development. These were predominantly ic conifers for timber. Much private planting even-aged broadleaved and mixed woods situ­ took place in many counties including ated in the lowlands and lower foothills, woods Cumberland, Devon, Durham and which often have the potential for continuous Lincolnshire. However, there are no indica­ forest cover. This period and continuing up to tions of use of alternative silvicultural systems 1910 was also when improving lairds (such as to clear cutting. Seafield and Atholl) used their estate and In Britain, the first sustained stage in the other revenues to plant extensively Scots pine, development of uneven-aged forest was moder­ European larch and Norway spruce for timber ately stimulated by foresters in the Forest of production. Some pine stands in the Spey and Dean from 1896 (Everard, 1992), and by Schlich Dee Valleys were naturally regenerated by the and Troup in the 1920s. From the early 1900s, uniform shelterwood system until emergency the Coke family of Norfolk, in their Weasenham clear cutting during World Wars I and II (1914- New Wood, had begun a selection system in 18 and 1939-45) terminated the use of such sil­ conifers (front cover plate). Subsequently, there vicultural systems. The Caithness policy wood­ have been several contributions— notably those lands, based on sycamore and beech, were per­ of Anderson, Bourne, Garfitt, Penistan, Hiley, haps the last of the policy or shelterwoods, dat­ Paterson and Reade — to advocate silvicultural ing from about 1890-1910. Shelter was very systems alternative to clear cutting. Among important there, but was also a common objec­ later practitioners have been Lord Bradford, tive elsewhere to protect farm crops and live­ Hutt, McHardy, Everard, Workman, Yorke and stock. There has been interest in finding ways Pryor. of renewing these woods and belts without M ark L. A nderson. From 1929, notable clear cutting in order to maintain amenity and encouragement of uneven-aged systems was shelter. In 1926 the Duke of Buccleuch at given by M.L. Anderson (born 1895, died 1961; Bowhill, near Selkirk, and Eildon, near Plate 24) who, after graduating from the Melrose, started a group selection system in University of Edinburgh in 1919 joined the broadleaves and conifers within his policy newly appointed Forestry Commission and woods. became a research officer. Later, he served as In England, meanwhile, management of the Director of the Forest Service in Eire, lecturer depleted Crown forests was transferred in in the University of Oxford and, from 1951 to 1810 from the Surveyor-General of Woods to 1961, Professor of Forestry in the University of the Commissioners of Woods and Forests (‘The Edinburgh. He was steeped in the literature of Office of Woods’). Sporadic planting, and some­ Continental forestry, and his varied experience what uninspired management, occurred in the of silviculture in western and central Europe Forest of Dean and the New Forest until in and Scandinavia gave authority to his teaching

89 and publications. In particular, his preparation of the English editions of silvicultural texts by Knuchel (1953) and Kostler (1956) made two detailed accounts of Continental silviculture available to English-speaking foresters. Anderson (1953) pointed out that, in conse­ quence of British foresters’ concentration on creating and managing forest on the clear cut­ ting even-aged system, ‘there has been little appreciation of the importance of the stand and the site, considered together as a single producing agent and not separately, and of the need for maintaining the production of that combination in perpetuity, or even of increas­ ing it . . . [and] instead of a perennial stand we have aimed at a succession of crop following crop with no clear idea of whether these crops are better or worse than their predecessors’. He asserted that Britain’s ‘silviculture can be improved only by experience. It cannot be stan­ dardised or made stereotyped, because in this country we have a very wide variety of condi­ rPHOTOGRAPH nOTJRTESY EDINBURGH UNIVERSITY) Plate 24.Professor Mark Louden Anderson MC tions and we are devoting our energies to the MA DSc FRSE (1895-1961) growing of a great variety of tree species, often influenced by the whims and vagaries of a number of rather inexperienced experts. All happening to their growing-stock or forest cap­ this leads to a great deal of confusion, largely ital. because we do not make use of some simple method of assessing results, which would at Anderson (1960) subsequently presented a least tell us what is happening within our thorough analysis of the climatic and site con­ woodlands as a result of the various treat­ ditions prevailing in the Vosges, Jura, Alps and ments that we apply to them. The time factor Carpathians where mixed forests of Norway does militate, of course, against the proper spruce, European silver fir and beech are man­ study of treatment effects, because any such aged under the single tree selection system. He study, to be of real value, must be carried on for speculated whether similar forest could be quite a long period of years.’ Anderson (1953) reproduced by planting in certain parts of further asserted that ‘progress in forest man­ Scotland, particularly at the higher elevations agement and silvicultural practice in this where exposure is severe. He concluded that country cannot be made until we adopt some parts of the uplands of east-central and south- method of readily determining the effects upon central Scotland with severe winters and mod­ production of the various techniques and treat­ erate though well-distributed rainfall would ments we apply’. He advocated the adoption of provide suitable sites, especially if the soils the Methode du Controle (which he termed the were derived from rocks of Upper Silurian age ‘standing control’ or ‘standing check’ method) in south Scotland and hornblende schists in especially for all woodland areas which it is the Highlands. Anderson further suggested proposed to run as uneven-aged forest, to that group selection would be the most suitable assess and control the growing-stock on a wide system. He proposed that dense groups of scale, and to enable forest owners and foresters birch, larch and alder, spaced quite widely, to obtain reliable knowledge of what is really should be established to suppress the ground

90 vegetation and to provide shelter for the later discussion went on for a year or two, in planting of spruce, European silver fir, the course of which Ponsonby, answering sycamore and other broadleaved species criticisms about going abroad for guid­ between the spaced groups in a progressive ance, asked had we not brought over, nat­ development towards an uneven-aged struc­ uralised and knighted Sir William ture. Between 1929 and 1932 he designed tri­ Schlich (1840-1925) in order to teach us als of a method of planting densely stocked, the lessons of Continental forestry? And well-separated, small groups of conifers on do we not go abroad to study the subject? upland sites being afforested in Scotland and In due course, on the basis of Continental northern England (Anderson, 1930, 1931, experience, [R.] Bourne in England 1951). Although the second stage was never spread the new gospel and his colleague, planted, the trials showed that a diverse forest [M.J.] Penistan, during his stay in the structure could be achieved by planting south of Scotland as a Forestry conifers in groups arranged in systematic fash­ Commission officer, found many disciples ion. The system did not become popular prac­ on private estates. In 1950 Penistan tice because of the slow build-up of production wrote in favour of methods of natural and the risk of browse damage to any sec­ regeneration and claimed that mixed ondary planting done between groups from woods were easier to manage. In 1952 he deer using the cover of the groups, and the fact was suggesting alternative methods [to] that the external trees in these so-called extensive clear-felling, and described" ‘Anderson Groups’ grew faster than the central improvement of woods in Dumfries and ones, thus suppressing them, and the sur­ Galloway. In 1949, Urquhart also came vivors were coarsely branched. The aim of out in favour of uneven structure. In these small groups was also to concentrate 1953, [W.N.] Gibson sounded a warning ground preparation on a fraction of the area. against even-aged structure, pointing out Anderson (1967) recorded the following addi­ that there were more pure stands than in tional relevant comments on alternative silvi­ the past. Gaps caused by windfalls would cultural systems in Britain: result in irregular stands. By 1960, Edinburgh University Forestry [In the late 1920s] ‘while the State Department had initiated six large-scale [Forest] service was still pondering over long-term experiments in the conversion [clear cutting], the new teachings of the of even-aged stands [into] a group struc­ Swiss school of Biolley and others were ture with the co-operation of the Forestry having their effects. These were against Commission and four private forest own­ clear-felling methods and in favour of ers. There was a growing appreciation selection or uneven-aged forest, especial­ that silvicultural treatment must be var­ ly for hilly regions. Ponsonby (1931), who ied for species other than light-deman­ had been an Irish representative on the ders.’ Forestry Commission, and who owned a wooded property in County Tipperary, Anderson’s further applications have been wrote an important article called ‘A noted in Chapter 3. He did not favour either System of Forestry for the British Isles’, underplanting or two-storied high forest; he advocating a departure from the pure was relieved that by 1960 they had ‘not invad­ even-aged stand to selection working. In ed Scotland’. [However, they had occurred. this he had the support of Major [the Morley Penistan and Frank Oliver during Honourable Richard] Coke who success­ 1945-60 encouraged the underplanting of larch fully applied such a method to his Norfolk stands in East Scotland, not with a view to estate [at Weasenham]. The challenge two-storied high forest but as a means of was taken up by A. C. Forbes and a lively reducing weeding costs and frost damage to

91 more productive successor species (such as tropical rain forest on a sustained yield basis Noble fir, grand fir, western hemlock, Douglas until this was interrupted by the Japanese fir and Norway spruce) which do not establish invasion. This type of management had a well in open extensive clear cutting in the strong group basis since natural regeneration semi-Continental climate over the regions of of the large species of dipterocarps tended in Tayside and Grampian owing to such factors as this direction. After 1945, Garfitt continued frost and May drought.] In the 1960s the the work initiated by Bourne on a number of Forestry Commission Research and lowland private estates and also developed his Development established large underplanting own ad hoc methods of group selection on the experiments in Drumtochty, Michaelston war devastated broadleaved woodlands at Drummond Hill, Allerston, Coed Morgannwg, Cirencester Park, Gloucestershire. These pro­ Culbin and Radnor. vided an ideal form of management for appli­ Ray Bourne. Bourne (born 1889, died cation to unproductive or unmanaged 1949) one of Sir William Schlich’s students, broadleaved woodlands in lowland England was the first forester to practise in Britain (Garfitt, 1953, 1963, 1966, 1977, 1979, 1980, variants of the selection system and to apply 1984, 1987, 1988, 1995). They provided the the Methode du Controle. He was appointed to opportunity to grow both shade-tolerant and the Indian Forest Service in 1911. From 1922 light-demanding species, with all the advan­ to 1937 he lectured in forest management at tages of the group selection system. Garfitt the University of Oxford. He then resigned to continued this work on estates, wherever the become a consultant to private woodlands. His conditions proved suitable and practicable, for extensive knowledge of Continental silvicul­ a period of 40 years until his retirement in ture and forest management was passed on to 1987. His group selection system and some of students in lectures and during study tours in his applications have been noted in Chapter 3. Germany, France and Switzerland as well as to M. J. Penistan. Morley Penistan (bom 1914, a wide audience of professional foresters died 1986), a one time Forestry Commission (Bourne, 1945, 1951; Hiley, 1964). Bourne ‘took Conservator, was an enthusiastic advocate of neglected and semi-neglected woodlands, uneven-aged forestry. He had studied under Ray extracted revenue from them, filled them with Bourne and was familiar with the work of M.L. natural regeneration, and progressively Anderson. His interest in this type of forestry in improved the growing stock in volume and still both England and Scotland increased following more in value. It was a highly personal form of an extensive tour of forests in France, Belgium management, and he appreciated the point of and Germany in 1956 (Penistan, 1960, 1968, criticism that it would be much more difficult 1974). During a visit by the then Society of for a successor to take over such work than if it Foresters of Great Britain to the Belgian had been organised on more orthodox lines’ Ardennes, which he helped to organise, atten­ (Steven, 1948). Bourne (1951) began from the tion was drawn to Turner’s (1959) method of con­ mid-1920s to convert several stands of beech­ verting uniform stands of Norway spruce to a woods in the Chilterns to a variant of selection diverse structure of mixed species in clearings. forest, which treatment resulted in an Penistan recognised its potential and advocated increased range of diameter classes and an its adoption in plantations of Norway spruce and increase in volume of the growing stock. Sitka spruce to diversify the structure of upland Regeneration of ash was adequate but that of forests in Britain. beech and oak were lacking. M. G. Reade DSC. Michael Reade, anoth­ J. E. Garfitt. Ted Garfitt succeeded to Ray er forester who recognised the potential of Bourne’s consultancy practice. He had studied variants of Continental diverse structure sys­ under Bourne at Oxford before joining the tems in British conditions, has consistently Malayan Forest Service in 1936, where he had practised since the 1950s what he terms ‘irreg­ five years’ experience in the management of ular selection forest’ or ‘irregular shelterwood’

92 in Chiltern beechwoods on his Ipsden Estate in Switzerland at that time (Reade, 1957, 1960, Oxfordshire. He wrote descriptions of the sys- 1965, 1990). His application has been noted in tems as they were being practised in Chapter 4.

Printed in the United Kingdom for HMSO Dd 0297433 C23 11/95 552 12521

93

Cyril Hart Dr Cyril Hart OBE is a Chartered Surveyor and Chartered Forester and is the Queen's Senior Verdererfor the Forest of Dean. He is a national authority on trees and forestry and the author of widely read books on Britain's forests and trees of which the best known is his authoritative Practical Forestry for the Agent and Surveyor, now in its third edition.

Published by HMSO and available from: HMSO Publications Centre (Mail, fax and telephone orders only) PO Box 276, London SW8 5DT Telephone orders 0171 873 9090 General enquiries 0171 873 0011 (queuing system in operation for both numbers) Fax orders 0171 873 8200 HMSO Bookshops 49 High Hoi born, London WC1V 6HB (counter service only) 0171 873 0011 Fax 0171 831 1326 68-69 Bull Street, Birmingham B4 6AD 0121 236 9696 Fax 0121 236 9699 33 Wine Street, Bristol BS1 2BQ 0117 9264306 Fax 0117 9294515 9-21 Princess Street, Manchester M60 8AS 0161 834 7201 Fax 0161 033 0634 16 Arthur Street, Belfast BT1 4GD 01232 238451 Fax 01232 235401 71 Lothian Road, Edinburgh EH3 9AZ 0131 2264181 Fax 0131 229 2734 The HMSO Oriel Bookshop ISBN 0-11- 710334-9 The Friary, Cardiff CF1 4AA 9780117103344 01222 395548 Fax 01222 384347 HMSO’s Accredited Agents (see Yellow Pages) and through good booksellers

9 780117 3344 £15.95 net