DIVERSIFYING NATIVE PINEWOODS USING ARTIFICIAL REGENERATION 311 Stages: Stand Initiation, Stem Exclusion and Can- Plots)

Diversifying native pinewoods using artiﬁ cial regeneration

TANYA K. OGILVY1 * , COLIN J. LEGG 1 and Downloaded from https://academic.oup.com/forestry/article/79/3/309/633189 by guest on 24 September 2021 JONATHAN W. HUMPHREY 2

1 Scottish Environment Protection Agency, Graesser House, Fodderty Way, Dingwall IV15 9XB, Scotland 2 Ecology Division, Forest Research, Northern Research Station, Roslin, Midlothian EH25 9SY, Scotland * Corresponding author. E-mail: [email protected]

Summary Little is known about broadleaf establishment within the native pinewoods of Scotland, yet such information is critical to current restoration initiatives. A 2-year fi eld trial was set up to assess the growth and performance of seedlings of four broadleaved species within different developmental stages of Pinus sylvestris (Scots pine) woodland in the Scottish Highlands. The developmental stages included stem initiation (establishment stage), stem exclusion (thicket/pole stage) and canopy break up (old-growth canopy), providing sites of varying shade and resource availability. It was predicted that differences in establishment between sites would occur due to species-specifi c differences in ‘ tolerance’ of shade, low-nutrient availability and poor soil drainage. Analysis of trial results showed high survival but growth rates (relative stem height and basal diameter growth) declined rapidly in year two. In year two, Ilex aquifolium achieved the highest mean growth rates in the stem exclusion stands (average 15 per cent light) where there was a lower percentage of Phytomyza ilicis (holly leaf miner) infection and frost damage. Sorbus aucuparia exhibited generalist behaviour with similar growth and performance across all stand treatments. Betula pubescens maintained similar growth rates across stand types but was more susceptible to insect damage in deeper shade. Overall, Alnus glutinosa grew more successfully than the other species and achieved highest mean growth rates in canopy break up (average 44 per cent light) and stand initiation (average 71 per cent light) stages. Broadleaf establishment within pinewood stands is challenging on account of acidic nutrient-poor soil regimes, and in places, poor drainage, aggressive understorey competition and canopies that restrict light availability and insect herbivory. Forest restoration techniques, including matching species to site, accelerating establishment and using framework species (individuals of high fi eld performance that are able to restore site productivity and nutrient cycles), are discussed.

Introduction 1991). Analysis of pollen cores has shown that pinewoods like Abernethy Forest (O’Sullivan, There is much evidence to suggest that broad- 1977) and East Glen Affric (Wolff and Tipping, leaved trees were formerly more important con- 1999; Shaw and Tipping, 2003) had a higher per- stituents of the oceanic boreal native pinewoods centage of broadleaved trees up until the nine- of Scotland (Steven and Carlisle, 1959; Rodwell, teenth century than they do today. Broadleaf

© Institute of Chartered Foresters, 2006. All rights reserved. Forestry, Vol. 79, No. 3, 2006. doi:10.1093/forestry/cpl018 For Permissions, please email: [email protected] Advance Access publication date 25 May 2006 310 FORESTRY species feature in the tree species descriptions of The importance of broadleaves in pinewoods all the pinewood remnants recorded by Steven is recognized in current forest policy. Guidelines and Carlisle (1959). Mosaics of birch and pine for native pinewoods state that the regeneration are more commonly described in western pine- of native species other than pine should be woods (e.g. Strath Glass, Wester Ross and Great encouraged to maintain or increase natural diver- Glen groups), but where mosaics are not found, sity ( Anonymous, 2003 ). These species include there is always reference to broadleaved species Betula pendula Roth, Betula pubescens Ehrh., occurring frequently or occasionally with the Sorbus aucuparia L., Alnus glutinosa (L.) Gaertn. pine. Pine and birch mixtures are seen today in and Ilex aquifolium L. as well as Salix L. spp, Downloaded from https://academic.oup.com/forestry/article/79/3/309/633189 by guest on 24 September 2021 the Strath Glass pinewoods of Glen Affric and Quercus petraea (Mattuschka) Liebl., Corylus Glen Strathfarrar. avellana L. and Populus tremula L. Furthermore, Scottish pinewood dynamics remain diffi cult at least 15 per cent of planted stock should be of to understand due to a combination of both cli- these species where natural regeneration is less matic change and anthropogenic disturbance. reliable. Thompson (2004) has also produced The successional status of mixed stands remains guidance for predicting successful natural re- unknown. Betula within the natural range of generation of P. sylvestris , Betula spp and S. Pinus sylvestris L. has been described as forming aucuparia adjacent to pinewood sites. Little a quasi-climax forest type but it also could be information is available, however, on the estab- successional to pine-dominated woodland or lishment of broadleaved species within pinewood mixed broadleaved woodland ( Worrell, 1996). sites where seedlings are subjected to limitations Furthermore, mixtures may occur as transitional in not only moisture and nutrients but also light. zones across altitude and edaphic gradients An understanding of their tolerance of shade, (Rodwell, 1991), as a product of human interfer- given the associated environmental conditions, is ence and/or stochastic disturbance events (Miles, especially important for restoration initiatives 1988). Many of the pinewoods today tend to be within woodland. dominated by P. sylvestris with pine and broad- The aim of this study was to develop an un- leaf woodland occurring in separate bioclimatic derstanding of the establishment of four broad- or edaphic zones. This segregation is believed to leaved species within pinewood sites of varying be a result of both direct and indirect human resource availability. Seedlings were planted infl uences, such as the selective removal of under different developmental stages of P. sylves- broadleaved trees (Summers et al. , 1999) and tris woodland and their growth and performance browsing by deer selectively restricting natural monitored over two growing seasons (April 2001– regeneration (Miller and Cummins, 1998; Scott September 2002). et al. , 2000). Broadleaves within pinewoods are believed to play an important role in promoting healthy eco- Materials and methods system functioning. There is evidence that broadleaves reverse podzolization (Malcolm, 1957; Study area Miles, 1985, 1986; Emmer et al. , 1998). Birch, for example, has a reputation as a soil improver The study was carried out in the Glen Affric Cal- known to reduce soil acidity, attract earthworms edonian Pinewood National Nature Reserve and increase A horizon depth due to organic which lies 56 km south-south-west of Inverness matter incorporation (Patterson, 1993). This in the Scottish Highlands. Glen Affric has a ‘ cool in turn provides more favourable conditions wet’ climate (Pyatt et al. , 2001) with a 4- to 5- for Scots pine seed germination, which prefers month growing season and an annual rainfall well-drained mineralized soils (McVean, 1963). ranging from 1600 mm in the east to 2800 mm in Broadleaves also enhance biodiversity through the west (Forest Enterprise, 1997 ). The study area infl uences on both structure and composition was restricted to seven woodland sites evenly such as providing an important habitat for spe- spaced around Loch Beinn a Mheadhoin at a cialist epiphytic lichens, fungi, invertebrates and mean elevation of 235 m a.s.l. Each site contained birds (Patterson, 1993). three P. sylvestris woodland developmental DIVERSIFYING NATIVE PINEWOODS USING ARTIFICIAL REGENERATION 311 stages: stand initiation, stem exclusion and can- plots). Each plot was divided into two randomized opy break up ( Oliver, 1981). Stands were selected Latin squares with every row and column of each on the basis of similarities in stand density, under- of the squares containing one of each of the four storey vegetation and slope. species (B. pubescens , S. aucuparia , A. glutinosa Stand initiation plots were within open wood- and I. aquifolium ) (32 individuals per plot). A land (distance of 8 m (average) from plot centre to hexagonal pattern was used for planting to ensure nearest P. sylvestris tree) with an understorey of that individuals were equidistance to each other degenerating Calluna vulgaris (L.) Hull, Vaccin- by 0.3 m. Each plot was also surrounded by a ium vitis idaea L. and Vaccinium myrtillus L. fence (2.72 × 1.95 m) to deter browsing from Downloaded from https://academic.oup.com/forestry/article/79/3/309/633189 by guest on 24 September 2021 growing on peat ( Table 1 ). Stem exclusion plots mammals such as deer, rabbits and hares. were characterized by a closed overstorey canopy One-year-old cell-grown seedlings were of P. sylvestris trees (average height: 13 m) with obtained from a local native tree nursery (Chris- thin ground vegetation dominated by Molinia cae- tie Elite Nurseries Ltd, Forres, Moray). All spe- rulea (L.) Moench on peat and peaty podzols. One cies were from seed provenance zone 201 to two old-growth P. sylvestris trees (average (Inverness) except for S. aucuparia , which was height: 17 m) characterized the canopy break up from provenance 102 (Caithness/Sutherland). plots with a fairly dense understorey of C. vulgaris , The seed zones are a non-statutory subdivision V. vitis idaea and V. myrtillus on peaty podzols of the statutory regions of provenance defi ned in and podzolic gleys. All sites were dominated by the Forest Reproductive Material Regulations ‘very poor’ soil nutrient regimes (Pyatt et al. , 2001) 1977 (Herbert et al. , 1999). with high soil acidity (mean pH 3.9). Although Seedlings were planted in the treatment plots in most sites were dominated by ‘wet ’ to ‘very moist’ April 2001 with a small amount of plain garden soil moisture regimes, an improvement in soil aer- peat and perlite to aid root establishment. Seed- ation was recorded with stand age (see Table 1 ). ling plugs were a mix of sphagnum moss and 10 per cent bark compost containing N:P:K (10:18:24) base nutrients and slow-release ferti- Experimental design lizer Sincrocell 12 (William Sinclair Horticulture A hierarchical factorial design was used. The Ltd, Lincoln). Any understorey vegetation that seven sites or blocks each contained the three was overtopping seedlings was trimmed back to stand developmental treatments (seven × three aid initial establishment.

Table 1 : Soil and tree data for stand initiation, stem exclusion and canopy break up stand treatments; all tree data are for stems >7 cm diameter at breast height (d.b.h.)

Variables Stand initiation Stem exclusion Canopy break up

Mean d.b.h. (cm)– 15.1 66.4 Mean height (m)– 12.8 16.7 Stem density (N ha − 1 )– 21.3 2.4 Total seasonal PPFD 2081 (±503) 424 (±136) 1254 (±325) Total seasonal PPFD % 71 (±12) 15 (±5) 44 (±14) pH 3.82 (±0.22) 3.82 (±0.19) 3.95 (±0.13) Aerobic soil depth (mm) 122 (±154) 211 (±136) 302 (±139) % of Calluna vulgaris 43 (±16) 3 (±4) 44 (±20) Soil nutrient regime VP P– VP VP Soil moisture regime VM– W MD – W F – W

Mean total seasonal photosynthetic photon ﬂ ux density (PPFD) and PPFD % from hemispherical photographs, mean soil pH, aerobic soil depth, mean abundance of C. vulgaris (calculated as a % per plot), soil nutrient regime (ranges from very poor (VP), poor (P), medium, rich to very rich) and soil moisture regime (ranges from wet (W), very moist (VM), fresh (F), slightly dry, moderately dry (MD) to very dry) (Ecological Site Classiﬁ cation: Ray, 2001). Means pooled across blocks (n = 7) (±SD) (see Ogilvy, 2004, for further details on methodology). 312 FORESTRY

Light ﬁ rmed by an absence of growth in the second year and was assigned a vigour score of 5. Type Seasonal estimates of total photosynthetic photon of above-ground damage (mechanical, mammal, ﬂ ux density were estimated from hemispherical fungal, insect, frost) was recorded with an esti- photographs of each treatment plot ( Table 1 ). mation of the amount of damage on a four-point Photographs were taken just above seedling level scale (I: <5 per cent, II: >5 per cent, III: >20 per and in the centre of each plot under uniformly cent and IV: >50 per cent). overcast sky conditions using an ISO-100 mono-

chrome T-grain fi lm (T-max 100, Kodak Ltd, Downloaded from https://academic.oup.com/forestry/article/79/3/309/633189 by guest on 24 September 2021 Statistical analysis Hemel Hempstead, Herts) and a Nikkor 8-mm fi sh-eye lens attached to a mechanical camera The effects of species, stand type and their interac- body (FM2 Nikon Corporation, Tokyo, Japan). tion on individual growth variables were analysed Negatives were scanned and the digital images using a general analysis of variance (ANOVA) processed using macros written for a commercial model in Genstat (Genstat/Windows edn 6, image analysis package (Optimas 5.2, Optimas, Release 6.1). Before ANOVAs were interpreted, Washington) to give estimates of diffuse and daily residual plots were drawn up for each growth direct transmission factors (see Ogilvy, 2004, for variable to check for normality and homogeneity further detail on methodology). of variance. Missing values in the orthogonal design were dealt with by substituting estimated Foliar nitrogen values generated by the missing value estimation function in Genstat ( Payne, 2002 ). Missing values Foliar nitrogen analysis was carried out on seed- included seedlings that died and seedlings with ling leaves to obtain a direct measure of nitrogen broken stems and damage from a low level of availability to the plants. One seedling of each herbivory (assumed to be by small mammals). species per treatment per block was selected in July 2001, taking enough leaves randomly from the apical top for 100 mg of oven-dried material. Samples were dried (24 h at 80°C) and crushed in Results and discussion preparation for a mixed acid digestion technique (sulphuric acid– hydrogen peroxide procedure) Growth and performance ( Rowell, 1994). Although survival remained high (<3 per cent mortality) ( Figure 1) growth rates declined sub- Growth and performance stantially in the second year of establishment for Before onset of growth in the April of each year, all species except A. glutinosa ( Table 3). It is pos- stem height (±0.5 cm) from soil level to the lead- sible that faster growth in the fi rst year is just an ing apex and basal diameter (±0.1 mm) were initial establishment effect caused by the fertilizer measured for each seedling. Seedlings were in the nursery plugs. In the second year, slow remeasured towards the end of the growing sea- growth can be explained by unfavourable site son in September to give estimates of relative conditions such as the acidic, nutrient-poor soils growth rates in stem height and basal diameter ( Table 1). Anaerobic soil conditions in the more

( RH , R D ) using R x = (ln x 2 − ln x 1 )/ t , where x 1 and open stand initiation sites most probably im - x2 were the initial and fi nal value of the measure- paired root function (Drew, 1997) and nutrient ments and the time, t , was 1 year. acquisition (Ohlson, 1995 ). In July of each year, the performance of the Of all seedlings, 43 per cent were scored as seedlings was recorded by assessing their vigour ‘ weakened’ from the vigour assessment in the and amount of damage. Each seedling was second year ( Figure 1). This score was assigned assigned a score (I– IV) for each of the three cate- primarily due to the yellow – green leaves, a gories: leaf colour, leaf loss and stem necrosis common indicator of nutrient defi ciency. A pre- ( Table 2). The highest score of any of the three liminary analysis of foliage nitrogen from a categories determined an overall vigour score sub-sample of seedlings revealed mean values to (1– 4) for each seedling. Seedling death was con- be lower than critical values, indicating defi ciency DIVERSIFYING NATIVE PINEWOODS USING ARTIFICIAL REGENERATION 313

Table 2 : Key to seedling vigour scores

Score Leaf colour Leaf loss Stem necrosis

1. Healthy I green I no loss I no necrosis 2. Weakened II green– yellow II <50% loss II <5% necrosis 3. Suffering III yellow– brown III >50% loss III >5– 20% necrosis 4. Dying IV brown IV total loss IV >20% necrosis 5. Dead Downloaded from https://academic.oup.com/forestry/article/79/3/309/633189 by guest on 24 September 2021

Ilex aquifolium Alnus glutinosa

Stand initiation (71% light) Healthy Canopy break-up (44% light) Weakened Stem exclusion (15% light) Suffering Dying Dead

Sorbus aucuparia Betula pubescens

Stand initiation (71% light) Canopy break-up (44% light) Stem exclusion (15% light)

020406080100 0 20 40 60 80 100 Vigour class % Vigour class %

Figure 1. Percentage of seedlings within each vigour class for four broadleaved species under stand initiation, stem exclusion and canopy break up Pinus sylvestris stands in July 2002. Seedlings browsed by mice and accidentally snapped have not been included. for three of the species (Evans, 1984 ) ( Table 4 ). is known to be very shade tolerant ( Savill, 1998 ), More than 50 per cent of seedlings showed signs the canopy provided added protection from late of damage in 2002, predominantly due to insects. (June) frosts and the understorey of C. vulgaris , Although herbivory was not substantial (mean which is known to compete vigorously for nitro- damage score: <20 per cent), any leaf damage gen ( Evans, 1984 ), was sparse ( Table 1 ). Sorbus would restrict photosynthesis to some degree. aucuparia exhibited generalist behaviour with Highly signiﬁ cant species × stand interactions similar performance ( Figure 1 ) and growth ( Fig- were found for mean relative growth rate in stem ure 2 ) across stand treatments in the second year. height (ANOVA: F = 8.13, df = 6, P < 0.001) and Betula pubescens also maintained similar growth basal diameter (ANOVA: F = 15.55, df = 6, P < rates across stands in the second year but had a 0.001) in 2002 ( Figure 2). Differences in both much lower percentage of healthy stems in the growth and performance of species between stand stem exclusion stage with a greater susceptibility treatments could be related to their ecologies. Ilex to leaf miner damage and fungal attack by rust. aquifolium achieved the highest mean growth Betula pubescens was expected to grow faster in rates in the stem exclusion plots (average 15 per stand initiation sites (71 per cent light) due to its cent light) ( Figure 2) where the plants were traditional classiﬁ cation as a shade intolerant healthier than in stands with higher irradiance ( Atkinson, 1992) combined with an ability to ( Figure 1). It was also less susceptible to frost and tolerate badly drained conditions (Savill, 1998). leaf damage caused by Phytomyza ilicis (holly However, it is obviously at the limit of its ability leaf miner) in the shade. Its success in these stands in these Glen Affric stands with very moist to wet can be explained by several factors: I. aquifolium soil moisture regimes. 314 FORESTRY

Table 3 : Mean stem height increment and basal diameter increment pooled across all stand treatments for four broadleaved species for 2001 and 2002

Height increment (cm/year) Diameter increment (mm/year)

2001 2002 2001 2002

Betula pubescens 13 4 1.4 0.3 Sorbus aucuparia 27 7 1.8 0.4

Alnus glutinosa 13 11 2.6 2.1 Downloaded from https://academic.oup.com/forestry/article/79/3/309/633189 by guest on 24 September 2021 Ilex aquifolium 9 3 1.7 0.8

Table 4 : Mean leaf nitrogen content of four broadleaved species growing in stand initiation, stem exclusion and canopy break up pinewood stages in 2001

Stand initiation Stem exclusion Canopy break up Deﬁ ciency level

Betula pubescens 5.8 (±1.3) 10.0 (±2.8) 6.2 (±0.7) <25 Sorbus aucuparia 4.8 (±1.3) 8.2 (±2.6) 4.9 (±2.4) <17 Alnus glutinosa 9.2 (±1.6) 9.9 (±3.6) 7.7 (±2.8) <25 Ilex aquifolium 5.4 (±1.8) 6.9 (±0.8) 4.3 (±0.7) –

1 Leaf nitrogen is expressed on a leaf mass basis (leaf Nmass (mg g − )). Means pooled across blocks (n = 7) (±SD). Critical values to indicate deﬁ ciency in foliar nitrogen for speciﬁ c broadleaved tree species are included where known ( Evans, 1984).

Alnus glutinosa achieved higher mean growth C. vulgaris is found to be aggressive. Hand- turfi ng rates in canopy break up (average 44 per cent around planting spots to improve aeration and light) and stand initiation (average 71 per cent planting with a less acidic, better-aerated soil light) stages. Overall, A. glutinosa was the most base would also assist establishment. In the trial, successful broadleaf across all pinewood sites. Its S. aucuparia plugs that were planted with a small success, particularly in stands of higher irradiance, amount of garden peat and perlite outgrew seed- can be related to its shade-intolerant physiology lings without these additions (not included in (Evans, 1988) but also its ability to tolerate anoxic the experimental design) by 40 per cent. waterlogged conditions due to a deep root system A natural alternative would be preferable to and specialized hypertrophied lenticels which are the use of fertilizers to address the problem of believed to increase the effi ciency of the plant’s poor soil nutrients. This introduces the ‘ frame- aeration system (McVean, 1956 ). Furthermore, work species method’ (Elliott et al. , 2003). This the ability of A. glutinosa to fi x atmospheric nitro- method involves planting tree species with high gen in association with bacteria in root nodules fi eld performance that are able to ‘ capture’ the (Savill, 1998) probably aided establishment. site by restoring productivity and nutrient cycles. In the trial, A. glutinosa was the one species that showed ‘ framework species potential’ with its Restoration approaches ability to fi x nitrogen, tolerate anaerobic soils The establishment of broadleaved trees under conditions, produce a shade canopy (high leaf pinewood sites in Glen Affric remains a challenge area ratio) and grow relatively well. Other studies due to the adverse site conditions. ‘ Acceleration’ have shown how nitrogen-fi xing trees can be used as a restoration concept is normally associated successfully to accelerate forest regeneration with natural regeneration ( Hardwick et al. , ( Rhoades et al. , 1998). 1997), but it can also be applied to planting. The Another important restoration approach growth and performance of planted seedlings involves accurate species– site matching for opti- could be accelerated by patch scarifi cation where mum rates of growth and survival to be obtained. DIVERSIFYING NATIVE PINEWOODS USING ARTIFICIAL REGENERATION 315 Downloaded from https://academic.oup.com/forestry/article/79/3/309/633189 by guest on 24 September 2021

Figure 2. ANOVA plots displaying species– stand interactions for relative growth rate in stem height and basal diameter for 2002. Least signiﬁ cant difference (LSD) error bars have been appended to each plot; the upper for the interaction between species and stand and the lower for the comparison of means within a stand treatment; stand initiation (open circle), old growth (open triangle pointing down) and stem exclusion (ﬁ lled circle).

Results showed how broadleaved species grew many different pinewood sites of varying irradi- and performed better in pinewood sites most ance, soil nutrient and soil moisture regimes are suited to their ecologies; thus, this study suggests required to produce more accurate guidance on that I. aquifolium would be best planted in the optimal sites for broadleaf establishment. shaded sites, while B. pubescens is more appro- priate in intermediate or open sites and A. gluti- Acknowledgements nosa and S. aucuparia have a wider tolerance. The work was supported by the National Environmen- However, it is acknowledged that results from tal Research Council and Forest Research. Forestry this trial are preliminary and restricted to low- Commission, Scotland (Fort Augustus Forest District), elevation Affric sites. Longer trials (>2 years) on gave permission for the ﬁ eldwork. 316 FORESTRY

References Miles, J. 1985 The pedogenic effects of different species and vegetation types and the implications of succes- Anonymous. 2003 Native pinewoods: 7. The manage- sion. J. Soil Sci. 36 , 571 – 584. ment of semi-natural woodlands. Forestry Practice Miles, J. 1986 What are the effects of trees on soils? Guide 7. The Forestry Commission, Edinburgh. In Trees and Wildlife in the Scottish Uplands. Atkinson, M.D. 1992 Betula pendula Roth (B. verru- D. Jenkins (ed.). Institute of Terrestrial Ecology, cosa Ehrh.) and Betula pubescens Ehrh. J. Ecol. 80 , Banchory, pp. 55 – 62. 837 – 870. Miles, J. 1988 Vegetation and soil change in the up-

Drew, M.C. 1997 Oxygen defi ciency and root me- Downloaded from https://academic.oup.com/forestry/article/79/3/309/633189 by guest on 24 September 2021 lands. In Ecological Change in the Uplands. M.B. tabolism: injury and acclimation under hypoxia and Usher and D.B.A. Thompson (eds). Blackwell anoxia. Annu. Rev. Plant Physiol. Plant Mol. Biol. Scientifi c Publications, London, pp. 57 – 70. 48 , 223 – 250. Miller, G.R. and Cummins, R.P. 1998 Browsing by Elliott, S., Navakitbumrung, P., Kuarak, C., Zang- red deer on naturally regenerated birch and juniper kum, S., Anusarnsunthorn, V. and Blakesley, D. saplings on wintering ground at Glen Feshie. Scott. 2003 Selecting framework tree species for restoring For. 52 , 138 – 145. seasonally dry tropical forests in northern Thailand Ogilvy, T.K. 2004 Regeneration ecology of broadleaved based on fi eld performance. For. Ecol. Manage. 184 , trees in Caledonian forest. Ph.D. thesis, University of 177 – 191. Edinburgh. Emmer, I.M., Fanta, J., Kobus, A.T., Kooijman, A. Ohlson, M. 1995 Growth and nutrient characteristics and Sevink, J. 1998 Reversing borealization as a in bog and fen populations of Scots pine ( Pinus syl- means to restore biodiversity in central-European vestris ). Plant Soil 172 , 235 – 245. mountain forests – an example from the Krkonose Oliver, C.D. 1981 Forest development in North Mountains, Czech Republic. Biodivers. Conserv. 7 , America following major disturbances. For. Ecol. 229 – 247. Manage. 3 , 153 – 168. Evans, J. 1984 Silviculture of broadleaved woodland. O’Sullivan, P.E. 1977 Vegetation history and na- Forestry Commission Bulletin No. 62. HMSO, tive pinewoods. In Native Pinewoods of Scotland. London. R.G.H. Bruce and J.N.R. Jeffers (eds). Institute of Evans, J. 1988 Natural regeneration of broadleaves. Terrestrial Ecology, Cambridge, pp. 60 – 69. Forestry Commission Bulletin No. 78. HMSO, Patterson, G. 1993 The value of birch in upland forests London. for wildlife conservation. Forestry Commission Forest Enterprise 1997 Glen Affric Caledonian For- Bulletin No. 109. Forestry Commission, Surrey. est Reserve Endangered Habitat Management Plan. Payne, R.W. 2002 The Guide to Genstat: 2. Statistics Unpublished internal report. Forest Enterprise, Fort (Genstat Release 6.1). VSN International, Oxford. Augustus. Pyatt, D.G., Ray, D. and Fletcher, J. 2001 An eco- Hardwick, K., Healey, J., Elliot, S., Garwood, N. and logical site classifi cation for forestry in Great Britain. Anusarnsunthorn, V. 1997 Understanding and as- Forestry Commission Bulletin No. 124. Forestry sisting natural regeneration processes in degraded Commission, Edinburgh. seasonal evergreen forests in northern Thailand. Ray, D. 2001 Ecological Site Classifi cation. A PC- For. Ecol. Manage. 99 , 203 – 214. based Decision Support System for British Forests. Herbert, R., Samuel, S. and Patterson, G. 1999 Using Forestry Commission, Edinburgh. local stock for planting native trees and shrubs. Rhoades, C.C., Eckert, G.E. and Coleman, D.C. 1998 Forestry Commission Practice Note 8. Forestry Effect of pasture trees on soil nitrogen and organic Commission, Edinburgh. matter: implications for tropical montane forest res- Malcolm, D.C. 1957 Site degradation in stands of toration. Restor. Ecol. 6 , 262 – 270. natural pine in Scotland. Forestry Department Rodwell, J.S. (ed.). 1991 British Plant Communities. I, Bulletin 4. University of Edinburgh. Woodlands and Scrub. Cambridge University Press, McVean, D.N. 1956 Ecology of Alnus glutinosa (L.) Cambridge. Gaertn.: IV. Root system. J. Ecol. 44 , 219 – 225. Rowell, D.L. 1994 Soil Science: Methods and Application. McVean, D.N. 1963 Growth and mineral nutrition of Longman Scientifi c and Technical, Harlow, Essex. Scots pine seedlings on some common peat types. Savill, P.S. 1998 The Silviculture of Trees Used in J. Ecol. 51 , 657 – 670. British Forestry. CAB International, Oxford. DIVERSIFYING NATIVE PINEWOODS USING ARTIFICIAL REGENERATION 317

Scott, D., Welch, D., Thurlow, M. and Elston, D.A. pinewoods and other woodlands in the Highlands of 2000 Regeneration of Pinus sylvestris in a natural Scotland. For. Ecol. Manage. 119 , 231 – 245. pinewood in NE Scotland following reduction in Thompson, R.N. 2004 Predicting site suitability for gazing by Cervus elaphus. For. Ecol. Manage. 130 , natural colonisation: upland birchwoods and native 199 – 211. pinewoods in northern Scotland. Forestry Commis- Shaw, H. and Tipping, R. 2003 Allt an Laghair, East sion Information Note 054. Forestry Commission, Glen Affric: recent woodland history from pollen Edinburgh. analysis. In The Quaternary of Glen Affric and Wolff, H. and Tipping, R. 1999 Recent vegetation Kintail. R. Tipping (ed.). Unpublished Field Guide, change within the pinewoods of East Glen Affric for Downloaded from https://academic.oup.com/forestry/article/79/3/309/633189 by guest on 24 September 2021 University of Stirling, pp. 105– 110. their conservation management. Internal Report. Steven, H.M. and Carlisle, A. 1959 The Native Pine- University of Stirling. woods of Scotland. Oliver and Boyd, Edinburgh. Worrell, R. 1996 The boreal forests of Scotland. Summers, R.W., Mavor, R.A., MacLennana, A.M. and Forestry Commission Technical Paper 14. Forestry Rebecca, G.W. 1999 The structure of ancient native Commission, Edinburgh.