Appendix 3 Allocation of woodland types to different site conditions
W9 Upland mixed broadleaved with dog’s mercury, W11 Upland oak-birch with bluebell/wild hyacinth, W17 Upland oak-birch with bilberry/blaeberry, W18 Scots pine with heather
The site guidelines of Rodwell & Patterson (1994) for these woodlands, essentially the ‘dry’ woodland types, are quite precise and have been closely followed, albeit with the necessary interpretation to match the data. In many instances, i.e. where the site guidelines for individual woodland types overlap or where Rodwell (1991) indicates that one woodland type could succeed the other, these types are considered to be interchangeable with their ‘closest neighbour’ (i.e. the most similar in its site requirements). For example, W9 is considered potentially interchangeable with W11, but not with W17 or W18.
The ‘weighting’ used in the model favours geology, then soil, subsequently
modified by existing vegetation. All the above woodland types have potential
on base-rich rocks, although the podzolic soils suited for W17 and W18 are
very restricted, as podzolisation occurs infrequently on base-rich parent
material. Soils suitable for W9 are restricted to base-rich rocks (Table
4), so soils developed on acid parent materials are only considered
to have potential for W11, W17 and W18. In the second filter, the continuum
between the brown earths and peaty podzols (brown earths and brown rankers
brown
podzolic soils
podzols
peaty podzols) is used to refine the woodland prediction (Table 4).
 |
 |
|
Brown earth (Soil survey archive) |
Brown podzolic soil (Soil survey archive) |
 |
 |
|
Humus-iron podzol (Soil survey archive) |
Peaty podzol (Soil survey archive) |
The photographs above illustrate the principal soil changes along this continuum with greater horizon differentiation reflecting:
- Increasing leaching of nutrients
- Redistribution of organic matter
- Depth and type of surface organic matter.
More subtle refinements use existing vegetation as an indicator of inherent soil quality (nutrient status) to guide the final NVC allocation (Table 5), according to the ‘optimal precursor vegetation’ (Rodwell & Patterson (1994), supplemented by information in McVean & Ratcliffe (1962), Birse (1980) and Rodwell (1991). The final decisions on woodland/scrub potential are made with an understanding that vegetation cover can be related to grazing pressure, as well as inherent soil quality.
Brown podzolic soils, typical of steep valley sides in the Western Highlands and peaty podzols, common in the Eastern Grampians.
Table 4: Influence of geology and soil type on the native woodland suitability of soils of free drainage
Woodland Type: decreasing base status and nutrient requirement
 |
||||
| W9 | ||||
| Geology: decreasing base status |
Base-rich rocks |
|
||
| Intermediate rocks |  |
|||
| Acid rocks |  |
|||
| Soils: decreasing base-richness and nutrient status |
Basic brown earths |
|||
| W11 | ||||
| Geology: decreasing base status |
Base-rich rocks |  |
||
| Intermediate rocks | |
|||
| Acid rocks | |
|||
| Soils: decreasing base-richness and nutrient status |
Brown podzolic |
|||
| W17 | ||||
| Geology: decreasing base status |
Base-rich rocks | |
||
| Intermediate rocks | |
|||
| Acid rocks | |
|||
| Soils: decreasing base-richness and nutrient status |
Humus-iron podzols |
|||
| W18 | ||||
| Geology: decreasing base status |
Base-rich rocks | |
||
| Intermediate rocks | |
|||
| Acid rocks | |
|||
| Soils: decreasing base-richness and nutrient status |
Peaty podzols | |||
|
= suitable woodland types |
||||
= soils do not occur |
||||
Brown earths on steep grassy slopes - typical sites for oak and birchwoods (W.Towers)
Peaty podzols under Calluna-dominated moorland. (D.J. Henderson)
Table 5: Use of soil/wegetation cover relationships to guide woodland suitability
| Existing Vegetation: indicative of decreasing nutrient status* |
|||
| Soils: decreasing base-richness and nutrient status |
Grassland (Agrostis Festuca) and agricultural land) | ||
| Brown podzolics | W11 | ||
| Humus-iron podzols | W17/W11 | ||
| Peaty podzols | W17/W18 | ||
| Soils: decreasing base-richness and nutrient status |
Grassland-Moorland mosaics | ||
| Brown podzolics | W11/W17 | ||
| Humus-iron podzols | W17 | ||
| Peaty podzols | W18/W17 | ||
| Soils: decreasing base-richness and nutrient status |
Moorland-Grassland mosaics | ||
| Brown podzolics | W17/W11 | ||
| Humus-iron podzols | W17/W18 | ||
| Peaty podzols | W18 | ||
| Soils: decreasing base-richness and nutrient status |
Moorland (Calluna dominant) | ||
| Brown podzolics | W17 | ||
| Humus-iron podzols | W18/W17 | ||
| Peaty podzols | W18 | ||
| * unless clearly influenced by grazing pressure. | |||
W8 Lowland mixed broadleaved with dog’s mercury, W10 Lowland mixed broadleaved with bluebell/wild hyacinth
In general terms these are the lowland equivalents of W9 and W11, respectively, and the main difference in their site requirements is that W8 prefers more base-rich and warmer conditions than W10. Like their upland counterparts, both are favoured on brown earths with free or moderately free natural drainage.
Although one of the main differentiating criteria between these woodlands and W9 and W11 is climatic, most soil map units within the 1: 250 000 scale soil map are restricted to either the ‘upland’ or ‘lowland’ climatic zone of Rodwell and Patterson (1994). This is logical since climate is one of the primary soil-forming factors and therefore soil map units would not be expected to occur across a wide range of climate zones. Therefore soil map units function very well as a surrogate for climate in the NWM.
W6 Alder with stinging nettle
Rodwell & Patterson (1994) indicate that W6 is restricted to the climatically favoured southern and eastern lowland areas of the UK, including parts of Scotland, and to the larger valleys further north and west. We consider that its potential geographic distribution would be wider than that, as in our view it could also cover poorly and very poorly drained mineral and peaty alluvial soils in valley bottom sites. The land cover data enhanced the soils data in this example by indicating the inherent soil drainage characteristics in such areas; wetlands and rush pastures differentiate poorly drained alluvial soils from those with free or moderate drainage, which are often under arable cultivation, or improved, or semi-natural grassland.
W7 Alder-ash with yellow pimpernel
W7 is considered the most appropriate woodland type for poorly drained mineral gleys. Within the literature, there is some confusion over the base-richness of optimal sites, even within Rodwell & Patterson (1994) and between those authors and Pyatt (1995). Despite this confusion, this woodland type appears to be the most appropriate, and may be the only candidate category on these soils, perhaps with a bias towards alder under increasingly acidic conditions. Many of these gleys are now artificially drained and so some soils with efficient drainage systems may have an enhanced site capability for W11 compared to unimproved soils. Such detail cannot be readily identified from the data used for the model, although the presence of rushes indicates poorer drainage conditions. Gleys are identified by oxidation/reduction processes within the soil profile, causing mottling and grey/blue colours. With effective drainage, the gleying is to some extent relict (the process which caused it is no longer operating), but as they are the natural soil properties, it is important to recognise their role in determining the natural pattern of woodlands. If artificial drainage schemes are not maintained (as is likely if woodland is established), the original soil conditions are likely to return to some extent.
Juniper
W19 Juniper woodland with wood sorrel
This woodland type is predicted on brown earths developed on parent materials of intermediate base status. In most instances, W11 is the favoured woodland type under these site conditions.
W4 Birch with purple moor grass
W4 is considered to be the only potential NVC woodland candidate on peaty gleys, where the existing vegetation is wet heather moor or rough grassland (predominantly Nardus or Molinia dominated). Nevertheless, some reservations have been expressed about the widespread extrapolation of its existing distribution, which is concentrated in southern England (Rodwell 1991), to large areas of upland Scotland. McVean & Ratcliffe (1962) seemed to consider that re-establishment of woodland on this often Molinia-dominated, human-influenced, vegetation would be relatively difficult. In view of these reservations about W4, we have examined earlier woodland classifications which are more specific to Scotland. The most relevant type described by Birse (1982) seems to be the Sphagnum palustre-Betula pubescens type, within which tree growth is described as so poor that it is best described as ‘shrubs’. This does tie in with the descriptions of the quality and structure of W4 as ‘moribund’ and ‘rather decripit’ used by Rodwell (1991) and by Rodwell & Patterson (1994), respectively. McVean & Ratcliffe (1962) described a range of alder woodland types which Rodwell (1991) linked to their W4 description. We have thus used W4 as equivalent to these woodland types, but suggest that it would be restricted to flushed areas (this concurs to some extent with the guidance within Rodwell and Patterson 1994), normally with large areas of open ground remaining (see below: ‘W4 with open ground’). Furthermore, colonisation by a W4-type woodland could be relatively short-lived on this vegetation type; it is possible that woodland colonisation might rapidly change the vegetation and soil with succession moving towards drier woodland types (Rodwell 1991; R. Worrell pers. comm.).
W4 Birch with purple moor grass and open ground
W4 with open ground is most frequently predicted on poorly draining peaty gleys, the woodland being restricted to flushed sites within landscapes containing predominantly un-flushed soils. These sites are common throughout Central Sutherland and in the west Highlands. Flushing can be determined from the presence and dominance of species such as Molinia and Juncus spp. Peaty gleys are usually a component of very heterogeneous landscapes, therefore this woodland type is almost always predicted as part of a mosaic of different woodland types. This is described in more detail in the later section on ‘Soil Landscapes’.
Peaty gleys under wet moorland. (D.W. Futty)
Basin Bog woodland
This woodland type has been identified on areas mapped as basin peats on the soil map, irrespective of current vegetation cover. Woodland potential is considered to be much greater here than on blanket peat, primarily because of the greater degree of shelter. In addition, basin peats tend to be more flushed and hence richer in nutrients than blanket peat. A number of species could establish on these sites, including willow, birch and pine as at Abernethy (McVean and Ratcliffe, 1962) although the pine will be sparse and stunted on account of the inherently wet soil conditions.
Peatland with scattered trees/scrub
Peatland with scattered trees/scrub is used for areas where peat is dominant (from the soils data) and where the vegetation (from land cover data) is ‘blanket bog with other peatland vegetation’. Landscapes like this are common in Sutherland and Caithness. This category is not expected to carry dense woodland cover (but nevertheless contributes to woodland ecosystems), but the potential will vary considerably depending on altitude, degree of erosion, presence or absence of flushing, and wetness of the peat. For example, observation suggests that Betula nana might favour wet sites, whereas stunted Scots pine would establish more successfully where drainage is better. Large areas of upland peat with severe erosion, e.g. in the Monadhliaths, between the Great Glen and Strathspey, and the Knockfin Heights on the Caithness/Sutherland border, are considered to have very limited woodland or scrub potential. Although the soils and land cover data for this woodland type correspond in most areas, the land cover data has been given precedence over the soils data where:
- The soils indicate peat, but the vegetation suggests that there are other soils present as well at a finer scale, for example a land cover mosaic of heather moorland with rock and peatland vegetation. The land cover data, being at a finer resolution, is picking up heterogeneity which the coarser scale soils dataset cannot
- The soils have been mapped as a soil complex (in which peat is a component), but where the land cover data identifies the peat areas within it through the delineation of blanket bog and other peatland vegetation in the land cover dataset.
Eroded peatland with bare ground. (W. Towers)
Extensive peatland ecosystem –
suitable for scattered
woodland and scrub. (D.W. Futty)
Montane scrub
There is considerable uncertainty regarding both the former cover of, and the current potential for, montane scrub. McVean & Ratcliffe (1962), and more recently Hester (1995) stated that montane scrub communities have the potential, both climatically and edaphically, to be much more widespread than at present. Tipping (1997) and Horsfield & Thompson (1997) tended on the side of caution and the latter, in particular, argued that the widely held assumption that montane scrub in Scotland was once as extensive as it still is in Scandinavia is unfounded. There is little doubt that browsing by red deer and sheep has suppressed scrub growth in many parts of Scotland, and so it clearly has some potential for increased cover, but it is not known to what extent it could spread if not suppressed by herbivores.
We have predicted potential for montane scrub where current soils and land cover indicate that conditions are suitable, according to current knowledge and published literature (as described below and as listed in the References). However, given the uncertainty surrounding the potential extent of these montane scrub communities, caution is advised in the interpretation of scrub potential in the NWM maps. The decision rules created within the model to identify the different montane scrub types are as follows:
- Juniper scrub on sub-alpine or alpine soils with a ‘dry’ non-montane vegetation cover (for example, dry heather moor)
- Scattered juniper scrub on sub-alpine or alpine soils with a ‘dry’ non-montane vegetation cover in mosaic with montane vegetation
- Birch/willow scrub on sub-alpine or alpine soils with a ‘wet’ non-montane vegetation cover (for example, wet heather moor or Molinia grassland)
- Scattered birch/willow scrub on sub-alpine or alpine soils with a ‘wet’ non-montane vegetation cover in mosaic with montane vegetation
- Mixed montane scrub on sub-alpine or alpine soils with undifferentiated heather moor vegetation cover or dry heather moor dominant and peat or rushes subsidiary, i.e. potential for species preferring both freely drained and poorly drained soils (juniper, birch and willows)
- Scattered mixed montane scrub - sub-alpine or alpine soils with undifferentiated heather moor in mosaic with montane vegetation.
The growth form and height of the scrub is likely to vary considerably, from an upright form similar to that found in Krumholtz-type pine to very stunted, prostrate and sparse cover in the scattered scrub categories, especially where exposure is severe.
Krumholtz Scots Pine
Unsuitable for tree/scrub growth
Land considered unsuitable for tree/scrub growth is identified where sub-alpine or alpine soils are present and the land cover is montane vegetation. This includes some heath communities, in addition to the montane communities in the National Vegetation Classification. Vegetation cover is often very sparse. Bare rock and scree is also classified as unsuitable.
Model Predictions in heterogeneous landscapes
‘Soil Landscapes’ with associated woodland mosaics
A unique feature of the NWM is that the integrated soil and land cover data have been interpreted to explicitly take into consideration a number of ‘soil landscapes’ (Bibby et al 1982). Each of these has a different balance of soil components, and so can potentially support different mixtures of woodland types. Within each soil landscape, each soil component - peaty gleys, peaty rankers, peat and peaty podzols - occupies specific ‘niches’, and so the relative proportions of each soil type are determined by the topography and rockiness of the ground. Each soil landscape is considered to have a different native woodland potential, although the distinctions between them are often subtle. The judgements made (described below) are supported with other literature (McVean & Ratcliffe 1962; McVean 1963, 1966; Wormell 1968; Rodwell 1991) and personal experience.
Rock and scree-dominated land unsuitable for native woodland or scrub. (W. Towers)
The soil landscapes vary depending on rockiness, slope and the proportion of the ground occupied by peat. Some of the differences between them are relatively subtle; like soil profiles (four photographs on page 45), soil landscapes are a part of a continuum and changes between them can be gradual over relatively large distances. Descriptions of their main features and our assessment of their woodland potential are given below.
Soil Landscape 1. Very rocky ground with peaty soils, common in the West Highlands. (D.W. Futty)
Soil Landscape 1 (Figure 16). This has a moderately to very rocky terrain with wet and/or undifferentiated heather moor. The absence of peatland vegetation indicates a low proportion of blanket bog, and it is probable that drier rocky knolls will form a significant proportion of the landscape. In this soil landscape, there will be a higher proportion of peaty podzols and rankers than in soil landscapes 2 and 3 below, but nevertheless, peaty gleys under wet heath are still dominant. The predicted woodland mosaic on this terrain is ‘W4 with open ground’ as the major component (associated with the peaty gleys), but with potential for W17 or W18 on the drier, more heathery, rocky knolls and the steeper slopes associated with these outcrops. These landscapes are particularly prevalent on the western seaboard between Cape Wrath, in north-west Sutherland, the Island of Mull, and in the Western Isles.
Soil Landscape 2 (Figure 16). This has moderately to very rocky terrain with heather moor dominant and peatland vegetation subsidiary. This landscape tends to be associated with generally more gently sloping terrain than in Soil Landscape 1, and the rocky outcrops are generally separated by areas of peat. It has a similar distribution to Soil Landscape 1, but also occurs extensively in Argyll and north Sutherland. Peaty gleys remain dominant, but peat is a notable component of the landscape, and the proportion of drier soils such as peaty podzols and peaty rankers is reduced as compared to Soil Landscape 1. A complex mosaic of woodland types is predicted, reflecting the heterogeneity and complexity of the landscape, i.e. ‘W4 with open ground’ on the peaty gleys and the poorly drained peaty rankers, ‘peatland with scattered trees/scrub’ on the peat component, and ‘W17 and W18’ on the scattered drier rocky knolls and steeper slopes. Further fieldwork would be necessary to distinguish the exact arrangement of these components on the ground, but the area of individual components will generally be small, i.e. 5 hectares or less.
Soil Landscape 2. Moderately rocky and rocky ground with peat a subsidiary component. (D.W. Futty)
Figure 16 illustrates the how the differences in topography, vegetation and soil, interpreted from the combined dataset (see Figure 1) can be used to separate Soil Landscape 1 (to the left) and Soil Landscape 2 (to the right). The main differences in woodland potential are that in Soil Landscape 1 there is no potential for ‘peatland with scattered trees/scrub’ but there is a higher potential for acid birchwood/Scots Pine (the W17/W18 interchangeable category).
Figure 16: Native Woodland prediction on different soil landscapes
Soil Landscape 3. This third type is similar to Soil Landscape 2, but the peatland vegetation component is dominant and the moorland subsidiary. Peat is the dominant soil, with peaty gleys, peaty rankers and peaty podzols associated with the scattered rocky areas. The order of predicted woodland types follows the rationale described in Soil Landscape 2, with ‘peatland with scattered woodland/scrub’ dominant, followed by ‘W4 with open ground’. The ‘W17/W18’ component is considered to be very minor in this landscape.
Soil Landscape 3. Moderately rocky landscape with significant areas of peat. (D.W. Futty)
Soil Landscape 4. This is found on steep and very steep rocky slopes with peaty gleys, peaty rankers and peaty podzols. A mosaic of W17/W18 associated with the drier rocky knolls and slopes and ‘W4 with open ground’ on the peaty gleys and poorly drained peaty rankers is predicted. It is most commonly found on the steep lower slopes of the mountain ranges of central Ross-shire and Invernesshire.
Soil Landscape 4. Steep rocky ground, typical of steep Highland Glens. (G. Hudson)
Soil landscapes 5 and 6. These comprise non-rocky gentle and strong slopes with an intimate combination of peaty gleys and shallow peat. Both soil landscapes are most common in east and central Sutherland and the north part of Ross-shire. Compared to rocky ground, the different soils do not occupy distinct niches in the landscape, and they often subtly merge into each other, with no distinct topographic or vegetation change between them. ‘W4 with open ground’ and ‘peatland with scattered trees/scrub’ are considered to be the best potential woodland types, the former being dominant where wet moorland communities currently dominate, the latter where peatland communities prevail.
Soil Landscape 5. Non-rocky landscape with poorly drained peaty gleys predominant. (D. W. Futty)
Soil Landscape 6. Wet moorland landscape with many areas of peat (D. W. Futty)
Soil landscapes 7 and 8 (Figure 17). These are restricted to areas of moundy moraine where the woodland prediction depends on the proportions of peaty podzols on the mounds and the peat and occasional peaty gleys in the intervening channels and basins. These landscapes are best expressed on Rannoch Moor (on the Highland/Argyll and Bute border) although they are found in most valleys throughout the Highlands. The integration of the soils and land cover to make these distinctions has already been demonstrated (Figures 2 and 3). Figure 17 demonstrates how these subtle differences are carried forward in the woodland predictions.
Soil Landscape 7. Widely spaced moraines. (D. F. Futty)
Soil Landscape 8. Closely spaced moraines with extensive peatland (W. Towers)
Figure 17: Native Woodland predictions on moundy moraine landforms
