The end of the Late Glacial and the onset of the Holocene around 11,700 BP is characterised by an abrupt and profound rise in temperature, with temperatures like today achieved by around 11,000 BP (Ballantyne 2018). The rapid increase in temperature was first recognised in the ice-core studies on Greenland (Dansgaard et al 1989, 533), but has since been identified in a series of studies throughout the world. One of the earlier investigations at Whitrig Bog demonstrated that the rapid rise in temperature was not a localised event in Greenland but occurred also in the British Isles (Lowe et al 1997, 403).
Evidence of the incipient Holocene climate and environment is afforded in the aforementioned studies of Whitrig Bog. The earliest phases of the Holocene are represented by lake muds, that although undated, include increasing quantities of grass (Poaceae) pollen along with birch (Betula), and indications of a significant decrease in erosion as soils developed and stabilised (Mayle et al 1997, 293).
Following the end of the Loch Lomond Stadial around 11,700 BP, the open herbaceous vegetation cover was gradually colonised by closed-ground herb, shrub and shrubby tree/tree taxa, including grasses and dwarf shrubs such as crowberry (Empetrum), juniper (Juniperus) and birch. Juniper pollen is abundant at many sites in the region in the earliest postglacial. No temporal patterning in the first appearance or spread of Juniper is apparent, and dates of between 12,150 and 10,700 BP are common. It is generally accepted that the decline in pollen frequencies of shrub taxa like juniper and willow (Salix) in the early postglacial is the result of shading by a burgeoning tree population. These shrubs and associated shade-intolerant herbs are presumed to have retreated to cliffs inaccessible to grazing animals. The survival of arctic-alpine herbaceous elements in localities at the highest elevations is possible.
Postglacial Ecological Succession
There is very good agreement between sites in the region that birch was the first tree taxon to colonize after the Loch Lomond Stadial at c. 10,800–10,600 BP (Birks 1989; GUARD Archaeology Ltd 2014). Birch pollen is present during the Late Glacial, at sites such as Whitrig Bog (Mayle et al 1997), although most likely representing shrub (Betula nana) rather than tree birch.
The establishment of birch was followed by the rapid colonisation of hazel (Corylus avellana type) across the British Isles, with the first substantial appearance in the pollen record for South East Scotland dated at The Dod to c. 10,200 BP (Innes & Shennan 1991). This evidence is in agreement with several other sites from the South West Scottish mainland (Boyd and Dickson 1986). However, Tipping (2010) has highlighted that immigration of hazel coincides at some sites with hiatuses in sediment accumulation, including at Yetholm and Linton Loch. These hiatuses are considered to reflect phases of lowered lake levels, apparent more broadly in lakes across north-west Europe, and hypothesised to reflect a prolonged period of climate aridity. The impact of this phase of increased aridity on both peat growth and lake sediment accumulation, recorded at Yetholm between c. 10,100 and 9,700 BP, has been to produce anomalous radiocarbon ages for hazel expansion between sites.
Climate aridity is increasingly considered to be a contributory factor in facilitating hazel expansion (Tallantire 2002). Alternative hypotheses have explored the relationship with the increased incidence of fire, represented by microscopic charcoal particles, argued to reflect the manipulation of the environment by humans favouring the expansion of hazel as a food source (Huntley 1993). Although the remains of charred hazelnuts are frequently found on Mesolithic sites across Britain (Zvelebil 1994), there is no clear association between the incidence of fire and expansion of hazel in the palynological record (Tipping 1994). Moreover, where charcoal is recorded its interpretation requires a fuller consideration of the range of conditions under which fires may have naturally occurred, particularly during phases of increased aridity.
Birch and hazel are a consistent feature of woodlands across the region, with more limited competition from other tree species. Colonisation by temperate broadleaved trees such as oak (Quercus) and elm (Ulmus) generally follow the pattern apparent for the British Isles (Birks 1989; Brewer et al 2017). Both trees were more common in lowland environments, along with hazel. Oak first appears in the Din Moss sequence between c. 10480–9550 BP, increasing sometime after a date of 10,200–9430 (Hibbert & Switsur 1976). The radiocarbon dates from Din Moss (Hibbert & Switsur 1976) have large error terms (>±150 years) that place limits on the precision of the chronology, emphasising the need for robust chronologies and new dating programmes of key pollen sequences. Oak expansion is dated at Yetholm to c. 8800 BP. However, at Ravelrig Bog, an initial increase in oak has been dated from 10,000–9500 BP, with a more sustained increase from 8000 BP (GUARD Archaeology Ltd 2014). At both Din Moss and Yetholm, elm appears to have become established several centuries prior to oak (Hibbert & Switsur 1997, Tipping 2010).
Rare localised stands of pine (Pinus) are also likely to have formed a minor component of the region’s woodlands since the early Holocene, although their local growth is difficult to establish. Pine produces large quantities of widely dispersed pollen grains and is generally not considered to be growing locally unless pollen frequencies exceed 20% (Bennett 1984, 137). Innes and Shennan (1991) suggest that pine may only have colonised the surface of mires, and perhaps also deep within some valleys (Tipping 2010).
Alder (Alnus glutinosa) is the final tree to colonise South East Scotland during the early Holocene, although the date and rapidity of its establishment is highly variable between sites, occurring anywhere between c. 8100–7200 BP (Tipping 1994). For example, at Ravelrig Bog, alder increases locally around 7000 BP, before declining, with a more sustained increased not apparent until much later at c. 4000 BP (GUARD Archaeology Ltd 2014). The pattern of alder establishment is observable at the supra-national scale as well as within the region (Chambers and Elliot 1989; Bennett and Birks 1990), and has led to an appreciation that the alder rise is not a response to an increasingly wet climate as was long perceived (Godwin 1975). Alder at any one site is now seen to be able to colonise niches as small populations before it expands. Alder expansion can be highly site-specific, occurring in response to one or a combination of disturbance processes, including increased aridity promoting colonisation of drier peat surfaces and lake margins. Evidence for increased aridity, resulting in alterations to sediment accumulation, have been recognised at several sites across the region and northern England more generally (Mannion 1978; Tight 1987; Tisdall 2000; Barber et al 2003; Tipping 2010). Climate may therefore have been an important causal factor in the establishment of alder, in addition to other factors such as the manipulation of woodland landscapes by humans and animals (Smith 1984; Chambers & Price 1985; Tipping 1995).
Lime (Tilia), the most warmth-demanding of trees, was from c. 6000 BP relatively common as far north as the southern English Lake District and is argued to have reached a climatic (temperature-determined) threshold north of the border (Pigott and Huntley 1981). Some evidence from the Cheviots (Tipping 2010) supports a more northerly position for lime in the east of the country, probably determined by base-rich soils in the Lower Tweed basin. Traces of lime are noted throughout the Holocene sequence at Ravelrig Bog (<1%) at an altitude of 187m above sea level (GUARD Archaeology Ltd 2014).
In addition to noting the appearance for a time of pine in the west of the region, it is possible to identify at a basic level some spatial differentiation of the different tree types. At upland sites in the central Southern Uplands, at around 250–350 m above sea level, Tight (1987) considered the dominant woodland trees to have been hazel and also birch. Oak and elm seem to have played an increasingly important role at lower altitudes (Mannion 1978), although at The Dod (Innes & Shennan 1991) at c 200m, these broadleaf trees are still thought to have been initially restricted to better-drained soils. It is probable that the acid soils were never sufficiently base-rich to support stands of elm.
A similar pattern of decreasing arboreal diversity with altitude can be seen within one valley, the Bowmont Valley in the northern Cheviots (Tipping 2010) where oak-elm-lime woods near The Merse, around Yetholm Loch, give way to oak-hazel woods, without elm, on the acid andesites of the deep interior of the Cheviot valleys. Birch survived competition by being able to cope with wetter ground conditions, and was common on and close to the mires, mosses and infilling lake-basins with alder and willow. Birch also probably replaced broadleaf trees on dry ground with increasing altitude.
There is little evidence from published diagrams for altitudinal tree-line limits in the region, and no representation of sub-alpine scrub, heath or grassland communities that might have grown higher on the hillsides. A dense woodland probably surrounded the upland sites examined by Tight (1987). Turner (1984) has suggested that tree growth was possible on the summit of Cross Fell in the northern Pennines, at an altitude of 893m, and in the English Lake District, Pennington (1970) argued for a tree- line at around 760m. Birks’ (1988) estimate for the Galloway Hills suggests the natural tree-line to have lain above 457 m; this is a minimal estimate. At c 660m, on Rotten Bottom in the highest part of the Moffat Hills, there is thought to be a tree cover of sorts throughout most of the Holocene, possibly containing oak and elm. Macro-remains of elm have been identified at comparable altitudes to Rotten Bottom in the central Southern Uplands. Given these findings there would seem to be no reason to think that any part of southern Scotland lay above the natural tree-line. Unlike regions to the north, southern Scottish hills are not ubiquitously dominated by blanket peat.