Variation in Aspen in Scotland: genetics and silviculture

The Biodiversity and Management of Aspen woodlands: Proceedings of a one-day conference held in Kingussie, Scotland, on 25th May 2001

Bill Mason
Forest Research, Northern Research Station, Roslin, Midlothian, EH25 9SY. E-mail: bill.mason@forestry.gsi.gov.uk

Eric Easton and Richard Ennos
Institute of Ecology and Resource Management, University of Edinburgh, Darwin Building, The King's Buildings, Mayfield Road, Edinburgh, EH9 3JU. E-mail: r.ennos@ed.ac.uk

Introduction

European Aspen (Populus tremula L.) is a widely distributed tree species in the northern temperate zone with a range stretching from Scandinavia to north Africa and from Britain to Japan. While the species is found throughout Britain, it is commonest in northern Scotland (Worrell, 1995a). However, a tendency to grow in small groves on the edges of woodland, a lack of regular seed production (Ennos et al., 2000), plus a palatability to grazing, have meant that the total area of Aspen woodland in Britain is estimated at around 500ha (J. Gilbert FC, pers. comm.) and woods of more than 1ha are rare. As a consequence, there have been very few studies of the silviculture and genetics of Aspen in Britain (Worrell 1995 a, b) and knowledge of appropriate management to favour this species is limited. The lack of knowledge is the more unfortunate because the importance of the species for both nature conservation and landscape interests is considerable (see other papers in these Proceedings). The purpose of this paper is to summarise some recent studies which provide useful guidance to those interested in increasing the area of Aspen woodland, either through natural colonisation or through planting.

Propagation of Aspen

The recent upsurge of interest in Aspen partially stemmed from recommendations that it be accepted for planting in the Native Pinewood schemes of the late 1980s (Hollingsworth and Mason, 1991). An average of 9ha of Aspen has been planted per year in Scotland since 1996 under the Woodland Grant Scheme (D. Wright, FC pers. comm.). Concern was expressed that the only Aspen material that could be purchased from nurseries in Britain at the time was of non-native origin and probably not well adapted to Scottish conditions (Ennos et al., 2000). A small research project was carried out in the late 1980s to look at the possibilities of propagating Aspen by cuttings. This resulted in recommendations to excavate root sections in the field during the dormant season, place them in a heated greenhouse to promote suckering, and root the detached shoots as softwood cuttings in a mist house (Hollingsworth and Mason, 1991). Trials have generally shown high rooting percentages using this method (ie 90-100% rooting under mist) but there has been appreciable clonal variation in the number of shoots m^-1 of root, and in growth after rooting (Hollingsworth and Mason, 1993).

Although seed production by Aspen in Scotland is infrequent, there have been two recent reports of successful seed collections in Strathspey (Worrell, 1995b; Worrell et al., 1999). In addition, some of the native Aspen material offered by commercial nurseries in Scotland has been through at least one multiplication cycle in vitro before being grown on for forest planting (R.S.D. Ogilvy, Christie-Elite pers. comm.).

Distribution in Scotland

As a result of increasing confidence in the ability to propagate native Aspen, the next question was where could suitable locations of Aspen be found in Scotland. Discussions resulted in a small Forestry Commission/Scottish Natural Heritage project being established in 1992 with nine objectives (see Table 1). Some of these objectives are discussed in more detail below.

Worrell (1995a) recorded some 500 sites in Scotland where Aspen had been reported in the early 1990s. These locations¹ are plotted in Figure 1 together with a further 100 sites recorded since the original survey; the 40 sites identified by Wilson et al. (2000) are included. The results indicate particular concentrations of Aspen in Perthshire, Deeside, Badenoch, Strathspey, Easter Ross and eastern Sutherland, with lesser frequency elsewhere. The recorded distribution was divided into eight zones as shown in Figure 1. This used the boundary between Regions of Provenance 10 and 20 (see Herbert et al., 1999) as the first divider, with subsequent divisions along major watersheds.

1. These locations are held on a database at Forest Research - Contact the senior author for more details.

Figure 1. Map showing recorded locations for Aspen in Scotland plus the eight zones into which the distribution was divided.

Investigations of Scottish Aspen using genetic markers

Once the range of the distribution of Aspen in Scotland had been ascertained, it became possible to undertake genetic analysis to find out more about the population biology and genetic variability of Aspen in Scotland. As indicated earlier in this volume, the reproductive biology of Aspen in Scotland is very different from that of most other native tree species. Flowering of the species is rare, and significant seed set is found at distant and irregular intervals (Worrell 1995a, Worrell et al., 1999). Seed is short lived, and the disturbed and open conditions required for seedling establishment are rarely found. Sexual propagation of the species is presently very problematic. On the other hand, the species has a remarkable ability to persist and spread via root suckers, so that individual genotypes have the potential to be extremely large and long lived. The management of native Aspen woods for conservation requires an understanding of how these special features have affected the level and distribution of genetic variation within Scottish Aspen populations. Two studies using genetic markers were therefore conducted to investigate these topics (Easton 1997).

The objective of the first study was to look at the native Aspen resource in the whole of Scotland and assess its genetic variability relative to that of other tree species within their natural ranges. Three main questions were addressed:

• Has the Scottish population retained genetic variability despite its reduced powers of sexual reproduction?

• How is this genetic variability now distributed among the regions within Scotland?

• Is there any evidence that restriction of sexual reproduction has led to inbreeding within the resource?

The study was based on analysis of enzyme genetic markers. This represents genetic variation that is not affected by natural selection (selectively neutral variation [Ennos et al., 2000]), and is not involved in adaptation. This is important to remember when interpreting the results that were obtained.

Dormant Aspen buds were sampled from a total of 275 individuals across Scotland (Figure 2) and scored for their genotype at eight enzyme loci. Measures of genetic variability at these loci were very similar to those found for other long lived woody perennials, though rather lower than for other Populus species (Table 2). Analysis of data coming from the six regional populations in Scotland (Wester Ross, Sutherland, Strathspey, Deeside, Perthshire and Southern Scotland) showed that less than 2% of the genetic marker variation was accounted for by differences between these regions. An unexpected finding, however, was that significantly fewer heterozygous genotype were found than would be expected in an outcrossing dioecious species like Aspen (Inbreeding coefficient within populations FIS = 0.153, P< 0.001).

Figure 2. Location of six regional populations of Aspen sampled in the study of genetic marker variation within Scotland.
1 = Badenoch and Strathspey, 2 = Sutherland, 3 = Deeside, 4 = Perthshire, 5 = Wester Ross, 6 = Southern Scotland.

These results are consistent with postglacial invasion of Scottish Aspen under conditions where sexual reproduction was widespread and seed dispersal and establishment were unrestricted. It appears that this initial genetic structure and variability has been largely retained despite fragmentation and reduction of sexual reproduction as a consequence of vegetative persistence over long periods of time. During this period, limited sexual reproduction of the species has taken place among an increasingly related group of individuals, leading to a significant level of inbreeding in the population as a whole.

Having obtained a broad picture of genetic variation in Scottish Aspen, a second study was conducted to determine the extent and pattern of clonal diversity within native Aspen woodlands. The questions of interest were whether woodlands comprise single or multiple clones, and, if multiple clones are found, do they differ from one another in ecologically significant ways. The study was conducted in Tomnagowan wood, Strathspey, the largest Aspen dominated woodland in Scotland. Within a 7ha sample area, 198 stems of Aspen were mapped and scored for their genotype at seven variable enzyme genetic markers. Stems were also scored for the extent of leaf flushing in June (Figure 3), and for their sex in spring of 1996, an exceptionally prolific year for Aspen flowering (Easton 1997).

Figure 3. Leaf emergence score for Aspen used at Tomnagowhan wood.

Within the sample area, 21 different clones could be identified. These ranged in size from single stems to individuals spreading over 100m. The clones identified in this manner showed very different leaf flushing scores. Individuals within each of the putative clones were all of the same sex and the sex ratio was 3.33:1 male:female, significantly greater than the 1.5:1 ratio normally found within Scotland. These data indicate that high clonal diversity exists within native Aspen populations, and that clones differ widely in important ecological characteristics. In future conservation programmes aimed at restoring or recreating Aspen woods, this clonal diversity must be included within planting stock.

Suitable nursery production of Aspen © Bill Mason
There have been very few studies into the silviculture and genetics of Aspen in Great Britain. As a consequence, the practical problems associated with successful and appropriate commercial nursery production have not been widely investigated and opportunities to expand the area of Aspen in Northern Scotland have been missed due to a lack of suitable planting stock.

Studies of natural stands

Matthews (1993) carried out a small study comparing growth rates, ages, site characteristics, and incidence of fungal pathogens on 30 sites in northern Scotland, 15 in Wester Ross and 15 in Badenoch and Strathspey. A total of 226 trees was sampled. The oldest tree was 120 years old; only 2% of trees were more than 100 years old with 36% being 50-100 years, and the remainder less than 50 years. The largest tree in Wester Ross was 19m tall by 91cm dbh compared with 25m x 39cm in Badenoch and Strathspey. The best growth was found on clay loam and sandy loam soils and the poorest on sands. The incidence of bacterial canker (Xanthomonas populi) was relatively low overall (7%), but increased with age to nearly 20% in trees older than 70 years. Similarly, the incidence of butt and heart rots increased with age with the percentage of infected stems exceeding 50% for trees older than 50 years. Height-age relationships were interpreted to predict potential site indices (i.e. dominant height) at age 50 ranging from 6-18m; the former were more characteristic of exposed sites in Wester Ross and the latter of sheltered, fertile sites in both regions. Assuming that the appropriate yield models (Hamilton and Christie, 1971) are those for sycamore, Ash and Birch rather than those for hybrid Poplars (Worrell, 1995b), then potential yields from natural unmanaged stands in the two regions range from 2-6 m³ ha^-1 yr^-1.

Silvicultural performance of Aspen clones

Once satisfactory propagation techniques had been devised for Aspen, and a range of locations had been identified, a further stage was to examine the variation in growth rates and other parameters in comparative trials. Five field trials have been established with this aim in Scotland since 1993. These compared a range of Aspen clones - chosen systematically without selecting for superior phenotypes - with selected hybrid Poplar cultivars and, in some cases, selected clones for an Aspen breeding programme in Sweden (see Table 3). The trials were established using normal forestry methods in deer-fenced enclosures. The results after six years are briefly described below. Analyses are based on standard analysis of variance procedures with survival percentages first transformed to the arc-sine scale.

Bush 30 and Cowal 11

These were both small trials with a limited range of Aspen clones. At Bush, mean survival after six years was over 95% with no significant difference between clones. By contrast at Cowal, survival was only 75% overall, largely because of vole damage; there were significant differences (p<0.05) among the Aspen clones for survival with a range of 40-100%. Height growth after six years showed very highly significant (p<0.001) differences between the Aspen clones (see Table 4); the best clone at each site originated from Tummel in Perthshire. At the more fertile Bush site, the Aspen clones performed less well than the hybrid poplar standards, but this trend was not significant at the more oceanic site in Argyll.

Moray 43

There were very highly significant differences (p<0.001) between clones for height, survival and diameter. Survival here was also affected by vole damage with a mean value of c 70% and a range of 32-100%. However, perhaps of more interest is to examine the difference between zones (Table 5). The highest survivals were found in material from the southern Scottish zones (1 and 2), with best growth from clones of the east-central Scotland zone (3). However, the favourable performance of the latter will have been influenced by the fact that the best clone was from zone 3. This was the same Tummel clone which had grown well in the Bush and Cowal trials. The standards generally grew better at this site than the Aspen clones.

Kilmichael 37

Again, very highly significant differences occurred between clones (p<0.001) for survival percentage, height and diameter growth after six years. Survival was higher than at Moray, with a mean of 92%; the standards ranged from 23-57%, while the Swedish Aspen clones all had 100% survival. Comparisons between the different zones are shown in Table 6. The best survival is from zone 8 (northern Scotland) and the poorest is from zone 4 (Argyll). The best height and diameter growth is again in zone 3, but once again the mean figure for this zone is increased by the vigorous growth of the Tummel clone. The surviving standards were taller than the average of the Aspen clones. However, the most vigorous growth was shown by the selected Swedish Aspen clones.

Aberfoyle 9

Data is only available for the first three years after planting, when initial establishment could still be having an effect, so only preliminary results are noted here. Survival was generally 100%, with only five clones showing some losses (20% in each case). There were very highly significant (p<0.001) differences in height growth, with the Scottish Aspen ranging from 86-266cm. By contrast, the Swedish clones varied from 287-446cm. The fast growing Tummel clone was not included in this trial.

General

These five trials comprise the first systematic attempt to examine variation in establishment and growth rate in Scottish Aspen. Results are encouraging, since they suggest that good survival can be obtained, except where vole damage is a problem, and average height of 1.5-3.0m can be anticipated within six years, depending upon site quality. The possibility of improving growth rates by selection of superior clones, if timber production is an objective, is indicated by the vigorous growth of the Tummel clone at four of the sites and by the vigour of the selected Swedish material on the two sites where it has been planted. Examination of zonal performances at the Moray and Kilmichael sites does not provide conclusive evidence that local origins of Aspen are necessarily the best adapted.

Conclusions

The results reported in this paper only serve to re-emphasise Worrell's view (1995b) that "Fundamental questions concerning [Scottish Aspen's] ecology and silviculture remain largely unanswered." Over the last decade, progress has been made in two main areas. Firstly, we can be reasonably confident that native Aspen can be reliably propagated by cuttings and that this material will establish well and grow vigorously in the field without special attention. Given that we can also be more optimistic about obtaining seed from native Aspen (Worrell et al., 1999), there should therefore be no justification for using plants of non-native origin in native woodland restoration or creation schemes.

The other main area of progress is that we understand more about genetic variation in Scottish Aspen, both at a national level and in terms of the occurrence of clones within particular woodlands. As far as the former is concerned, we now know that there is appreciable genetic marker variability within the Scottish resource, little variation between the main regions of the Scottish population, and a low but significant level of inbreeding. This suggests that Scottish Aspen has been derived postglacially from a single source that was freely sexually reproducing at that time, but has since survived chiefly by vegetative propagation. The study of clonal variability has highlighted the need to use a mixture of clones when seeking to establish any sizeable areas of Aspen woodland. We would suggest a minimum of 10 clones in woods of 0.5-1.0ha, 15 clones in those of 1.0-5.0ha and >20 clones in those greater than 5.0ha. Within such schemes, the clones could be planted randomly (mimicking the situation expected with establishment from seed) or in pure blocks of perhaps 500-1000m² to mimic the patterns that are now found after the extensive vegetative spread of individuals in our few large Aspen woods.

Apart from these two areas, our understanding of the silviculture and dynamics of Aspen in Scottish woodlands is still rudimentary. The improved growth of the clone from Tummel indicates that there is improved vigour (i.e. 3.5-5.0m height at six years) in the Scottish population that could be exploited for timber production. This suggests that the growth rates of 6-10m³ ha^-1 yr^-1 quoted by Worrell (1995b) for Aspen in Norway on 40-60 year rotations could also be achieved in Scotland. The improved vigour observed with the small group of selected Swedish clones also suggests that there is a real possibility of increasing the productivity of Aspen stands where these are being used for timber. However, more systematic, comprehensive and longer-term trials would be necessary to confirm this potential for improved growth. Such trials could also consider other aspects of adaptive variation, such as salt tolerance of clones growing close to the sea (for use in reclamation schemes), and timing of bud burst and leaf senescence.

A further area of interest is the potential use of Aspen as a species for mixture in stands of non-native conifers. The species' tendency to grow in clumps, its vigorous early height growth and its tolerance to gleyed soils all suggest a potential role in enhancing biodiversity within spruce plantations. However, to date no experiments have been established to ascertain desirable patterns of mixture with non-native conifers.

A further limitation is our ignorance of stand dynamics in existing Aspen woodland. Worrell (1995a) cites MacGowan (1991) as indicating that woods of >4.5ha are desirable to provide a complete range of habitats for specialist Aspen invertebrates. Both Worrell (1995a) and Matthews (1993) suggest that trees over 100-120 years of age are rare and that older trees tend to be at risk from fungal pathogens. Based upon first principles and observations in the field, we propose that sustainable Aspen stands should contain a minimum of three age classes, each in discrete homogeneous cohorts. These would be an area of old trees perhaps 80-120 years of age characterised by abundant deadwood, a younger age range of perhaps 30-50 years which could be exploited selectively for a range of timber products, and a young cohort of perhaps 5-10 years to provide a successor stand. It is this last category which is so often absent in existing Aspen stands as a consequence of inappropriate grazing management.

In summary, we believe that these results give preliminary indications of the methods and potential for developing an expanded Aspen resource in Scotland, both as a component of native woodland and of restructured plantation forests. However, there is a considerable challenge for both researchers and forest managers if this potential is to be realised.

Acknowledgements

Dr Rick Worrell was responsible for the initial compilation of Aspen sites in Scotland with help from many individuals. Mike Hollingsworth propagated the material used in the field trials which have been managed by Forest Research at the Newton, Cairnbaan and Bush field stations. David Brown was responsible for collecting root sections from different parts of Scotland in 1993. The Swedish Aspen clones were kindly provided by Martin Werner and colleagues from Skogforsk, Ekebo. Statistical analysis and advice was provided by Ian White and Alvin Milner. Glenn Brearley provided the figures. The comments of Richard Thompson, Colin Edwards and Sam Samuel are gratefully acknowledged.

References

Easton, E. P. 1997. Genetic variation and conservation of the native Aspen (Populus tremula L.) resource in Scotland. Unpublished Ph.D. thesis, University of Edinburgh.

Ennos, R.A., Worrell, R., Arkle, P., and Malcolm, D.C. 2000. Genetic variation and conservation of British native trees and shrubs: current knowledge and policy implications. FC Technical Paper 31, FC Edinburgh.

Hamilton, G., and Christie, J.M. 1971. Forest Management Tables. FC Booklet 34, HMSO, London.

Herbert, R., Samuel, C.J.A. and Patterson, G.S. 1999. Using local stock for planting native trees and shrubs. FC Practice Note 8, FC Edinburgh.

Hollingsworth, M.K., and Mason, W.L. 1991. Vegetative propagation of Aspen. FC Research Information Note 200, FC Edinburgh.

Hollingsworth, M.K., and Mason, W.L. 1993. Vegetative propagation of native Aspen - update. Native Woodlands Discussion Group Newsletter 18.

Matthews, P. 1993. Growth rates, site indices and their environmental correlates for naturally occurring stands of European Aspen (Populus tremula) in 2 proposed Scottish provenance regions. Unpublished MSc thesis, University of Aberdeen.

Wilson, S. McG., Malcolm, D.C., Rook, D.A. 2000. Locations of populations of Scottish native trees. Scottish Forestry Trust, Edinburgh. 108pp.

Worrell, R. 1995a. European Aspen (Populus tremula L.): a review with particular reference to Scotland I. Distribution, ecology and genetic variation. Forestry 68: 93-106.

Worrell, R. 1995b. European Aspen (Populus tremula L.): a review with particular reference to Scotland II. Values, silviculture and utilisation. Forestry 68: 230-243.

Worrell, R., Gordon, A.G., Lee, R.S., and McInroy, A. 1999. Flowering and seed production of Aspen in Scotland during a heavy seed year. Forestry 72: 27-34.

Return to The Biodiversity and Management of Aspen Woodlands

For further information about aspen, please go to the Aspen Information Resource