Information and Advisory Note Number 102                                                Back to menu

A framework for forest development in the Cairngorms

1. Introduction

1.1 The response from the Secretary of State to the Cairngorms Working Party Report advised the Cairngorms Partnership to concentrate, in the first instance, on the montane and woodland zones in their Management Strategy. The stated priorities for woodlands are to "give further consideration to, and establish a framework for the protection, regeneration and re-establishment of the native woodlands, especially the Caledonian Pine Forest and the creation of the new Forests of Mar and Strathspey, together with the steps necessary to achieve this, including appropriate management of the deer population and of other grazing pressures."

1.2 The Cairngorms Partnership and the Forestry Authority are likely to counsel Scottish Natural Heritage (SNH) for guidance on forest expansion and management to secure natural heritage benefits. As a basis for providing this guidance in relation to enhancing woodland biodiversity, the concept of a Forest Habitat Network, as proposed by Peterken et al (1995), has been applied to the Cairngorms Area. The aim of a network approach is to effectively create the coherence of a large forest area with a relatively low percentage of tree cover while maintaining significant areas of open space. This requires patches of different woodland types and structure to be of sufficient size to sustain local populations of plants, animals and other organisms, and, that patches are sufficiently well connected to allow interaction between them.

1.3 This note summarises the findings of a recent study "A forest habitat network for the Cairngorms" by Ratcliffe et al (in press), which categorises the main woodland types and structures which occur in the Cairngorms area and selected a range of species which are characteristic of each. For each species habitat requirements, minimum patch size and dispersal distance were assessed. This information was then used to make specific recommendations on the size of individual woods and links between them that will be required to create an effective network of woodland habitats. Priorities for the expansion or restructuring of existing woodland and establishment of new areas of tree cover which contribute most to this network are proposed.

1.4 The network approach offers considerable flexibility to pursue woodland expansion and provide benefits for conservation, recreation, landscape and timber production, alongside other land-use and natural heritage interests. The recommendations from this study will form the starting point for dialogue with these wider interests to identify the opportunities and constraints for woodland expansion. Guidance on other natural heritage interests such as landscape character assessments and open ground habitats will also be considered in working up SNH's policy in relation to woodland development in the Cairngorms Partnership Area.


2. Designing forests in the landscape

2.1 There is a dearth of practical information on the potential effects of fragmentation, or of increased connectivity, in particular ecosystems. The benefit of developing habitat corridors, as a means of enhancing biodiversity is still open to debate. That said, an approach based on developing an interconnected forest habitat network appears to provide a means to satisfy the habitat requirements and dispersal characteristics of a range of species. It also offers a practical and targeted basis for providing advice on how best to meet the SoS's priorities for native woodland, with minimum disruption to other natural heritage and land-use interests.

2.2 The extent to which it is possible to create an effective network of woodland habitats is ultimately determined by the natural pattern of tree cover that could be expected in the area. Some sites may not be ecologically suitable for tree growth while some woodland types, such as wet woods, may be naturally scattered and isolated. An earlier phase of this study modelled the likely distribution of NVC woodland types for the Cairngorms Partnership area, using existing information on landcover, soils and tree cover, to indicate the potential area of woodland cover. The present and potential distribution of the woodland types in the Cairngorms is discussed later.

2.3 Biodiversity encompasses the range of variety in ecosystems, between species and within species and naturally varies quite markedly in response to adaphic and climatic conditions. It is virtually impossible to measure, and some form of sampling, or the use of indicator species is often used as an indicator. In practical applications spatial and structural variety can be used as a surrogate for biodiversity. Wildlife will usually benefit from a more diverse structure, especially when it mimics natural ecosystems. For this reason it is suggested that restoration of native woodlands in the Cairngorms should involve the creation and maintenance of high structural and spatial diversity to support the range of species characteristic of a large forest area. This highlights the need to consider not just networks of different woodland types but also the creation of different structural types within these networks.

2.4 The situation is further complicated by the fact that the character of individual woodland patches changes over time as the trees in them grow, senesce, die, decay and regenerate. If the variety of species is to be maintained, suitable habitat must be continually available for species to move into as the area they occupy changes around them, if they are to survive. Thus in designing new landscapes or deciding where to locate new woods, it is important to consider the character and distribution of woodland patches, and how they might change through time. Continuity in both time and space are equally important.
 

3. A species centred approach

3.1 While it is expected that species will benefit from the creation of more ecological niches, it is not generally known how particular species will be affected. For example, woodland patches maybe too remote to be colonised by some species or be of insufficient size to maintain viable populations. The extent to which this is so depends on dispersal distances and habitat requirements of the individual species concerned.

3.2 This presents a dilemma. While it is impossible to assess the requirements of all species in an ecosystem, it is difficult to proceed beyond the theoretical and general. In order to make recommendations on the creation woodland habitat networks, it is necessary to consider a representative range of species.

3.3 This study therefore selected a range of species on the basis of their rare and sensitive status, their potential to benefit from increased habitat connectivity and, their dependence on woodlands of particular types and structure. An attempt has been made to ensure that all components of the woodland ecosystem are represented, particularly with regard to decomposition.


4. Species selection

4.1 Although this was inevitably a subjective process, attempts were made to include a wide range of species. The Biodiversity Challenge exercise, which linked species listed under the UK Action Plan for Biodiversity, with habitats was used as a basis for establishing a list of rare or sensitive species. T his was modified by checking whether the particular species were recorded in the Cairngorms, and by asking a number of species specialists to point out omissions.


5. Autecological data

5.1 Having selected a list of species, information was collected for each on: current status, habitat requirements, causes of decline (relevant to habitat fragmentation), minimum patch size, mobility and dispersal capabilities. Frequently, the information was not available and literature reviews were not particularly helpful. The majority of the information was gleaned from informal interviews with species specialists. The data were analysed to seek common factors which might help plan future woodland restoration work and to determine whether species as a whole would benefit from forest habitat networks. Having reduced the approach from pursuing a general increase in structural diversity, it was now necessary to build back up to establishing a pattern of woodland which would benefit a wide range of the most sensitive and threatened species.


6. Statistical analysis of species requirements on a landscape scale

6.1 As well as the approach described above, a computer program (FRAGSTATS) was used with the Land Cover for Scotland 1988 (LCS 88) data in order to examine the existing woodland pattern in the Cairngorms through the eyes of some of the key species. This was done by applying the species requirements, in terms of habitat, minimum patch size and dispersal distance, to the dispersion of woodlands.


7. Findings

7.1 From the information available it appears as though 1 km is an achievable dispersal distance for many species. Once populations of many species are separated by more than 2km there is likely to be little interaction between them. Also the probability of finding suitable habitat decreases as the distance between patches increases. The degree to which species can disperse through the landscape depends not only on the mobility of individual species but also on the nature of the intervening land cover. Distances between patches of more than 1-2 km seem to seriously disadvantage a number of important species including, many saproxylic invertebrates, pinewood plants and crested tits.

7.2 Woodland patches of less than 5 Ha are less likely to support a representative range of key species. The benefits of patches less than 5 ha seem to diminish increasingly with decreasing size. Virtually all species will benefit from the creation of patches which exceed 5 ha particularly most insects, crested tit and pinewood plants.

7.3 Results from the FRAGSTATS analysis suggest that for red squirrel and capercaillie, current patch sizes are only marginally adequate, and that more patches greater than 500 ha would be beneficial.

7.4 Contiguity and continuity of habitats, such as old-growth and open, pine forest with a diverse shrub layer, within woodlands, are important for many species. Red deer grazing at a level of 4-6 deer/100 ha will maintain suitable habitats for a number of species, and so complete exclusion of grazing is not desirable, except where this may be necessary over short time scales to secure the desired tree regeneration.


8. Proposals - potential distribution of woodland types

8.1 The potential natural pattern of woodland as modelled for the Cairngorms indicates the degree of connectivity of different woodland
types which could theoretically be achieved.

• Pine birch woodland could be well connected around the Cairngorms massif.
• Both oak and birch woods and montane scrub could be well connected in certain localities.
• Riparian woods wet woods and the richer mixed broad-leaved woods are all naturally zonal and would remain isolated in the Cairngorms area.

8.2 Thus in the long term it seems feasible to aim towards creating a widespread network of pine/birch woods, localised networks of oak/birch woods in the straths and localised networks of montane scrub around the treeline. For riparian woods wet woods and richer mixed broad-leaved woods the long term aim could be to expand them where site conditions are suitable, in some areas this may allow then to link with neighbouring woods of a similar character, such as in gorges.

8.3 An impression of where different woodland types could be developed is given by the maps in appendix 2 of the main report. Specific recommendations for each of the main woodland types are discussed below.

8.4 Connectivity of woodland habitat can be created either by creating contiguous woodland areas or by increasing the density of woodland patches to the point where species utilise a number of patches as their home range, or it is possible for the local populations in each patch to interact. This latter approach offers considerable flexibility to pursue the long term aim of woodland expansion in sympathy with maintaining other interests.


9. Pine/birch woods

9.1 The short term (5-10 years) aim could be to expand the area of contiguous pine wood habitat in the Priority Forest Areas (PFAs) proposed by the Cairngorms Working Party, i.e. the Forests of Strathspey and Mar, and to establish new native pinewoods in the north eastern area between them.

9.2 In the PFAs priority could be given to encouraging the expansion of existing local origin pine/birch woods, particularly those recorded in the Caledonian Pinewood Inventory or the Ancient, Long-Established and Semi-Natural Woodland Inventories, through natural regeneration and restructuring existing plantations, to extend the areas of contiguous pinewood. Further expansion could be targeted to improve the connectivity between existing pinewood areas. Both these measures will increase the probability of survival, and range, of the larger species, such as red squirrel and capercaillie. Some individual patches would need to exceed 500ha and be well connected to achieve this. Pinewood specialists such as the crested tit, the pinewood plants and wood ants would also benefit from larger contiguous areas.

9.3 In the area to the north east of the PFAs, priority could be given to increasing the number of pine/birch wood patches to slowly create a mosaic of pinewood and open land. Restructuring existing plantations can be as effective as establishing new woods. Most of the existing pinewood plantations in the area are of unknown origin. In the absence of evidence which proves origin, planting with local origin stock should be the preferred method of establishment. Some new woods should be at least in the order of 100Ha in order to provide a substantial basis for the development of larger woods in the future. In general a minimum patch size of 25 Ha should be encouraged.

9.4 Within the pinewoods, connectivity of structure is necessary to support a number of important species, and in turn low levels of grazing are necessary to provide this. Low levels of grazing managed by manipulating local deer densities will allow the development of a woodland structure of high tree density in the early stages, with patchy, self-thinning later. This pattern mimics the natural development of forests. Tree densities exceeding 2000 stems/ha are common in the early regeneration stages in natural temperate forests, with less than 300 stems/ha in the mature and perhaps less than 100 in the old-growth stages. Overall deer densities in the range of 5-8 deer/km should be aimed for. Old Growth Core Areas (OGCAs) could be established with a minimal intervention regime, and these could be connected by Extended Rotation Areas (ERAs), where trees are removed at a range of ages beyond normally accepted rotations. This should provide the necessary connections to allow populations to disperse from the OGCAs into the areas where timber production is one of the main objectives. ERA connectivity should be planned to prevent gaps of more than 1km. The remaining matrix of woodlands derived from natural regeneration and plantations could be managed to provide timber. Markets for big logs could be expanded if possible.


10. Oak/birch woods

10.1 In the PFAs most of the potential oak/birch wood sites are concentrated along the floors of the Straths. Much of this land is presently in agricultural use or, particularly in the Dee PFA, carrying plantations of exotic conifers. Hence for the foreseeable future there is likely to be little opportunity to create extensive areas of contiguous broad-leaved woodland. In the short term priority could be given to consolidating and where possible expanding existing woods to a minimum of 5Ha especially on sites recorded in the Ancient, Long-Established and Semi-Natural
Woodland Inventories, and to increasing the density of broadleaved woodland patches. This will involve creating new woods in the agricultural landscape and diversifying the structure and species composition of existing plantations. Ideally patches should be within 1 km of each other. In some areas with a high density of patches it may be possible to create larger areas of contiguous broadleaved woodland.

10.2 Outside the PFAs most of the potential oak/birch sites are associated with Strath Avon, Strathdon, Glen Livet and the glens in the south of the Cairngorms area.

10.3 There are a number of existing broadleaved woods and considerable areas of regeneration along the Avon. This appears to be where, in the short term, there is greatest opportunity to create contiguous broad-leaved woodland cover, primarily through consolidating existing areas of regeneration and where appropriate, by encouraging further expansion of existing woods. Priority should be given to sites recorded in the Ancient, Long-Established and Semi-Natural Woodland Inventories.

10.4 In both Glen Livet and Strathdon the situation is similar to that in the PFAs. The short term priority should be the same.

10.5 Consideration could be given to the expansion of oak/birch and pine/birch woods in the southern apart of the Cairngorms area. Although there may only ever be weak links between the woodland in these areas and the rest of the Cairngorms, the fingers of wood up these glens represent outliers of larger woodland areas to the south.

10.6 Many of these, mainly broad-leaved, valley woods, are overgrazed presenting a savannah like habitat. Although it is recognised that some of these woods are important in providing shelter and food for stock, grazing needs to be reduced in at least some of them in order to provide the connectivity of habitat structure required by many of the invertebrates and plants.

10.7 Aspen woods in the Cairngorms are especially important in a British context, both in their own right, and with regard to the highly specialised and unique insect fauna associated with them. Priority should be given to increasing the size of existing remnants, particularly those along the Spey, and to managing for continuity of structural conditions within them. Many of the saproxylic invertebrates have very narrow niche requirements, requiring dead and decaying wood of particular dimensions. Minimum patch sizes of mature woods should be at least 2 ha, but 5 ha should be aimed for where possible.


11. Montane scrub

11.1 The potential distribution of these scrub types lies between the woodland and montane zone around the 600m contour. Pine and juniper scrub is likely to dominate with local occurrence of willow, particularly Salix aurita, and possibly scrub birch. There present distribution is very limited due mainly to high levels of grazing. In the longer term extensive areas of montane scrub could develop at the altitudinal limit of woodland areas to form Krummholtz woodland and gradually spread through some of the high passes such as between the Geldie and Feshie. In the short term priority should be given to consolidating expanding existing fragments through control of grazing. Montane scrub should be seen as an integral part of any woodland scheme which borders the montane zone. Further information can be found in "Subalpine Pine Scrub in the Cairngorms* French et al (1997).

11.2 The area between the forest and montane zones is particularly sensitive and there is considerable uncertainty over the type of vegetation which could be expected to occur, so tree and shrub cover should be allowed to develop in response to natural influences. Planting should only be considered as a means of establishing future sources of seed for natural regeneration.

 

12. Mixed broad-leaved woodland and riparian type woodland

It seems likely that many species are at risk in the low valleys and riparian areas. Gaps between existing fragments of the broad-leaved woodland types (mainly, the Riparian, Mesic broad-leaved and acid oak and birch woods) are in many cases too great to allow the effective dispersal and colonisation of many invertebrates. From the autecological data reviewed, distances between patches should be reduced to less than 1-2 km by the extension of existing woodlands and the establishment of new ones. Care should be taken to cater for the fishing interests through taking an integrated approach to woodland and riverine management. Generally there are likely to be benefits from developing links between these naturally discrete woodland types and other types of woods.

 

13. Additional points

13.1 Regeneration could be targeted to provide the pinewood mosaic described above. This mosaic should develop rather slowly to provide continuity of a variety of woodland structure through time. This can be achieved by encouraging targeted regeneration schemes to achieve the above pattern. Local deer densities should be reduced to 'kick start' regeneration.

13.2 In order to establish forests in as natural a process as possible, natural regeneration should be the preferred method, particularly where there is uncertainty over the likely composition and extent of tree cover, such as at the boundary between the woodland and montane zones. However, even in these sensitive areas, there are locations where tree cover has been eliminated for such long periods, and existing seed sources are so far away, that some planting may be necessary if tree cover is to be established. Small patches of woodland could be established, by planting a mixture of native species, suitable species would become established and act as seed sources in the future. Given appropriate grazing levels, scrubby tree cover could regenerate round these areas where site conditions were suitable.

13.3 The possible reintroduction of the beaver could create a potential conflict between the conservation requirements of the biodiversity associated with aspen woods and the beaver. It is considered important to treat the expansion of aspen woods on suitable sites as a priority so that the resource is sufficiently robust by the time possible reintroductions are contemplated. Ultimately, beavers might actually aid the regeneration of aspen by spreading propagules along the rivers and water courses, and they would almost certainly increase the diversity of riparian woodlands.

13.4 Care should be taken that the establishment of woodland connectivity does not at the same time, isolate moorland and montane species. Peterken et al (1995) draw attention to some of the mitigating aspects of forest networks on non-wooded areas. Essentially, it is important to develop an integrated landscape approach by ensuring that the forest network has minimum adverse impacts on other ecosystems or land-uses. The concept of minimum gaps, rather than completely contiguous woodland and 'stepping stones' of sub-optimal habitats allow considerable flexibility in the application of the woodland network, and it is not envisaged that adverse effects need be serious. The importance of fire breaks should be borne in mind when considering the location and distribution of open space, these can often be set out in a way that facilitates deer control and extraction.

13.5 In the long term, links between Strathspey, Strath Avon, Strathdon and Deeside will be important in creating a major Core Forest Area of Caledonian Forest in the Cairngorms area. Links could also be created in lower passes at such places as Ryvoan and by extending and linking high altitude scrub treelines, such as in the Lairig Ghru. Riparian ecosystems could be effectively linked with the fragmented birchwoods along the river straths, for example, from Spey dam to Glenbeg and Lettoch, downstream of Grantown on Spey. The broad areas for targeting are shown on the maps attached to the main report.

13.6 SNH views the proposals set out above as a useful starting point from which to develop policy on woodland development in the Cairngorms and to engage in a broader debate over the future of the forest resource in the Cairngorms Partnership area. In developing policy, the proposals set out above will need to be considered alongside other important aspects including non woodland habitats, landscape, recreation, community involvement, settlement patterns, archaeology and other land use interests.

 

14. References

French, D. Cummins, R., Miller, G., Bayfield, N. and McGowan, G. (1997). Subalpine pine scrub in the Cairngorms. In Gilbert, D. Horsfield, D. and Thompson, D.B. A. (Eds) 1997. The ecology and restoration of montane and subalpine scrub habitats in Scotland. Scottish Natural Heritage Review No 83.
Peterken, G.F., Baldock, D. and Hampson, A. (1995) A Forest Habitat Network For Scotland. Scottish Natural Heritage Research. Survey and Monitoring Report No. 44.
Ratcliffe, P.R., Peterken, G.F. and Hampson, A. (1998). A forest habitat network for the Cairngorms. Scottish Natural Heritage Research, Survey and Monitoring Report No.114

 

15. Author

Alan Hampson
Scottish Natural Heritage
2 Anderson Place
EDINBURGH
EH6 5NP
Tel: 0131-447 4784 Fax: 0131-446 2405


Back to menu