4.8 Palaeoclimatology

North Atlantic climate – key processes

The atmospheric processes responsible for the existence of the North Atlantic climate place northern Europe within the North Atlantic storm track. During winter, this results in the passage of numerous frontal cyclones, the occurrence of high winds and associated periods of higher than average rainfall, usually associated with a positive North Atlantic Oscillation (NAO) Index. Regional patterns of air temperature change are also responsive to short-term changes in the NAO, itself based on monthly time series of air pressure differences between Iceland and the Azores. Over longer time periods, changes in the NAO are affected by fluctuations in the extent of sea ice across the Greenland Sea and northern North Atlantic. Similarly, climate changes during the Holocene appear to have been profoundly influenced by changes in oceanic thermohaline circulation that have alternately brought warmer salty water to higher latitudes and colder freshwater to lower latitudes. Greenland ice core research has shown that during the Holocene, there have been significant changes in both atmospheric and oceanic circulation across the North Atlantic region. These have been characterised by marked changes in regional storminess, rainfall, atmospheric and ocean temperature as well as in ocean circulation. Although past variation in air temperature has attracted most attention across the archaeological community, other changes in weather and climate may have had considerable significance to archaeology. In particular, ice core chemistry time series (e.g. sea-salt sodium (storminess) non sea-salt potassium (dust transport) as well as stable isotope data (e.g. deuterium excess (sea surface temperature)) have scarcely been used as contextual backgrounds in archaeological research.

The ice core data also demonstrates that past patterns of climate change have not been synchronous across the North Atlantic region. For example, the existence of a climate ‘seesaw’ between western Greenland and northern Europe has been known for over 200 years. The existence of this climate ‘seesaw’ is of fundamental importance to archaeology for two key reasons. First, it demonstrates that reconstructed changes in climate for Scotland for different time intervals during the Holocene need not be matched by similar changes elsewhere across northern Europe. Second, the oxygen isotope record of inferred palaeotemperatures derived from Greenland ice cores cannot be regarded as a template for past changes in temperature across northern Europe.

The patterns of climate change described above are particularly significant for the transition between Palaeolithic and Mesolithic. Thus the close of the last ice age across northern Europe was characterised by high-magnitude perturbations in climate the most significant of which was the Younger Dryas. This period of ca. 1,100 yr of extreme cold climate was accompanied by significant southward displacement across the North Atlantic of both the polar atmospheric and oceanic fronts, as well as the local development of ice caps and valley glaciers and the development of permafrost.

4.8.1 Dendroclimatology in Scotland

Dendrochronology is the scientific method of dating tree-rings by matching patterns of wide and thin rings between trees of the same species from the same region. There is a long history of using tree-ring dating in archaeology (Baillie 1982). The fact that trees of the same species show the same general pattern of wide and thin rings through a region and that this common variability can be used for dating implies a regional scale forcing upon growth. This regional forcing can only be climate related as all other influences on growth – whether ecological or management related – are often site specific. Therefore, by understanding how climate controls growth, it is possible to reconstruct past climate from measurements of various parameters (e.g. ring-width, density and stable isotopes) from tree-ring samples (Fritts 1976). Through the dating of preserved material from historical structures (e.g. beams) and sub-fossil material (e.g. preserved in lake sediments), living chronologies, and therefore climate reconstructions, can be extended back in time .

Schematic representation of crossdating in dendrochronology. Copyright R Wilson

In Scotland, Hughes et al. (1984) showed that a highly robust reconstruction of Edinburgh mean summer temperatures could be derived using ring-width and maximum latewood density data obtained from several pine sites throughout the Scottish Highlands. However, the Hughes et al. (1984) reconstruction only went back to AD 1721 and they concluded that:

“Further sampling of ancient living trees should allow the extension of the Edinburgh record back into the seventeenth century, the most severe phase of the Little Ice Age, and result in a valuable reconstruction. A proxy record for more distant times may be obtained using the large quantities of sub-fossil pine found in the British uplands.”

Despite this early study, further sampling for dendroclimatological purposes, has however not been made until recently. One of the current research foci of the Tree-Ring Laboratory at the University of St Andrews is to develop tree-ring chronologies for all semi-natural pine woodlands in the Scottish Highlands. The Scottish Pine Project (http://www.st-andrews.ac.uk/~rjsw/ScottishPine/) not only aims to update and improve the Hughes et al (1984) reconstruction, but to substantially push these quantified estimates of temperature much further back in time. The figure below presents preliminary analyses showing the potential for gleaning summer temperature information for the last 250 years using tree-ring data from the Cairngorm region. As with the Hughes et al. (1984) study, the good fit with the instrumental data is clear. However, the challenge for the future is to extend such reconstructions back prior to the period covered by the living trees.

A preliminary mean July-August temperature reconstruction for the Cairngorms. Copyright R Wilson

Finding appropriate preserved pine material in historical buildings is a challenge (Crone and Mills 2002) as most of the construction material used over the last ~500 years has been imported from the continent. The key to extending Scottish living pine chronologies back in time, therefore, lies in finding sources of complete stems with long tree-ring sequences. Woody macro-fossil material preserved in lake sediments provides such a source. The mountain lochs in regions which have been continuously forested for the last ~8000 years, such as the north-west Cairngorms and Glen Affric, are ideal target regions. ¹⁴C dates of pine macro-fossil material from lochs in the Rothiemurchus Estate have identified preserved pine material from the last ~8000 years with an abundant amount of material from the last ~1000 years. These early results indicate that a millennial length pine chronology from the north-west Cairngorm region is a feasible and realistic objective in the near future (Wilson et al. submitted).

St Andrew’s University has a tree-ring laboratory including two microscope based measuring systems as well as a scanner system and more advanced lab facilities for measuring stable isotopes and trace elements. There is also an X-ray densitometer system at Glasgow Chemistry Department. Active researchers in the field include Dr Rob Wilson (St Andrews University), Dr. Mike Jarvis (Glasgow University), Dr Coralie Mills (St Andrew’s University; specialising in cultural wooded landscapes), Dr Anne Crone (AOC Archaeology Group; specialising in earlier periods and wetland archaeology), Colin Edwards (Forestry Commission; interest in the effects of climate change on woodlands), and Dr Neil Loader (University of Swansea; Isotope dendro-chemist with a recent PhD student completing a doctoral thesis on an isotope based tree-ring reconstruction using material from Glen Loyne).

This is an exciting time for Scottish dendroclimatology as it will be soon possible to extend current temperature reconstructions back at least 1000 years with a long term aim for going back ~8000 years. Such climate proxy records will prove invaluable for identifying periods of warm/cool climate and due to their exact dating should prove useful for comparison to the archaeological record. It should also be emphasised that such long chronologies will also facilitate the dating of cultural artefacts and historical structures that are constructed from pine.

4.8.2 Speleothems

Speleothems are mineral deposits that form inside caves from dripping, flowing or standing water. Such caves are normaly hosted within carbonate rocks such as limestone or dolomite, that are karstified, i.e. partly dissolved with a range of scales of cavity. The familiar pendant stalactites and upstanding stalagmites form from water that drips from the cave ceiling, derived from pores or fractures in the overlying bedrock, which is normally composed of carbonates such as limestone or dolomite. Flowstones form from intermittent, larger flows of water on the floors of cave and are particularly likely to be intermittently associated with archaeological remains which they may underlie or encrust. The archaeological material can be introduced dry through overlying steep fissures or shafts, or be transported to the site by water flows, or humans or animals may have introduced the material directly themselves.

Speleothems provide two particular points of interest for archaeologists (Fairchild et al. 2006). Firstly they can often be precisely and accurately dated, normally using uranium-series isotopes and within the range of a few tens to a few hundred thousand years. Hence the age of archaeological remains can be constrained. Secondly measurements of growth rate, or one of several parameters of their chemical composition, can be proxies for past environments or climates at the time of deposition.

Applications and recent publications

There is a rapidly growing literature on linking speleothem and archaeological studies, ranging from pre-Quaternary deposits in South Africa right up to the present day (records of continued growth and recent or current human pollution can be obtained from speleothems). Representative examples of applications where speleothems are integral to the archaeological work are Bar-Matthews et al. (2010), Lundberg and McFarlane (2007), Marean et al. (2007) and Vacca and Delfino (2004), and an overview of relevant aspects of cave environments is to be found in Lewin and Woodward (2009).

The most useful Scottish caves for speleothem study have been in the Assynt district of Sutherland, although the most useful samples climatologically (Proctor et al. 2000, 2002; Trouet et al. 2009) are in different caves from those of most archaeological interest, although there are examples where they coincide (Hebdon et al. 1997).

The existing research capacity in Scotland and in the UK

There are no established speleothem groups in Scottish Universities, but there is plenty of appropriate infrastructure in Scottish Universities should an appropriate person be appointed there in due course. In the UK as a whole there are groups at the Universities of Birmingham, Bristol, Durham, London (Birkbeck College and Royal Holloway), Oxford, and UEA with some interest and expertise at several others.

A forward look

Speleothems have advanced rapidly in the last 10 years to become one of the pre-eminent avenues of investigation of past environments and climates and a new research-level text (Fairchild and Baker, 2012) marks this coming of age. The techniques of study have also advanced considerably in terms of sensitivity and limited destructivity. Whenever speleothem material occurs at archaeological sites, it is likely to provide useful information. Recent excavations on Skye have involved speleothem researchers; however the initial contact was by accident, suggesting that a wider awareness of the possibilities of speleothems would be useful in the archaeological community. In this respect, the Upland Caves Network has provided a useful forum for interchange of ideas in the last three years. The work of the British Cave Research Association and its linked caving clubs such as the Grampian Speleological Group are vital for the ongoing research.

4.8.3 Testate amoebae and chironomids

Techniques

Sedimentary deposits are excellent archives of environmental change, and can frequently be used to reconstruct past climate. Arguably, the best techniques from raised (ombrotrophic) peat bogs and lake sediments are testate amoebae and chironomid analyses respectively. Testate amoebae (Protozoa: Rhizopoda) are unicellular animals with a discrete shell enclosing the cytoplasm. These shells, or tests, are resistant to decay within peat and hence can be used as indicators of past hydrological change in raised bogs, specifically through the estimation of past water-table depth. Testate amoebae are abundant in wetlands, and have been used to reconstruct quantitatively the past bog water table depth through the transfer function approach (e.g. Woodland et al. 1998; Charman et al. 2007). Chironomids (Insecta: Diptera) are non-biting midges, and typically the dominant invertebrate within lake sediments after oligochaetes. The larvae possess a chitinous head capsule, which is replaced 4 times during the larval cycle through ecdysis. The head capsules of the 3^rd and 4^th larval instars are typically robust, and survive the processes of fossilization. They can be picked from sediment core samples and identified to genus level or sometimes higher taxonomic resolution (species group or species). Chironomids respond to a number of environmental drivers, but most datasets collected over large environmental gradients show summer temperature to be the key driver affecting distribution. This has allowed a number of regional transfer functions to be developed for reconstructing past air temperatures quantitatively (e.g. Brooks and Birks 2001; Larocque et al. 2001; Langdon et al. 2008).

Previous use in Scotland

There are many raised peat bogs in Scotland and a reasonable amount of research has been undertaken on them for reconstructing past climate change using a range of proxies, including testate amoebae. Langdon and Barber (2005) studied the palaeoclimate records from seven peat bogs across north-south and east-west geographical and climatological gradients in Scotland. Testate amoebae were analysed from five of the sites as they are typically abundant in all peats except the most decomposed and/or driest. It is usually difficult therefore to find enough fossil testate amoebae to count in blanket peats (e.g. Langdon and Barber, 2001). The sites that Langdon and Barber (2005) studied were also synthesised in a regional comparison of testate amoebae records from Northern Britain (Charman et al. 2006), which also included other Scottish sites such as Tore Hill Moss, Speyside (Blundell and Barber, 2005) and sites from the Scottish Borders (Mauquoy and Barber, 1999; 2002). The palaeoclimate records developed from these sequences have shown a clear Holocene event stratigraphy in terms of changes in bog surface wetness, with broad scale wetter/cooler climates occurring around 4500, 4200, 3800, 3300, 2700, 2200-2500, 1800, 1400, 800-1000, 500 and 150-250 cal. Years BP, but with some regional differences (Langdon and Barber, 2005). Furthermore, over the last 7500 years, a millennial scale cycle in wet/cool climates has been observed from Temple Hill Moss, SE Scotland (Langdon et al. 2003).

There has been relatively less work undertaken on chironomid analysis in Scotland. A Lateglacial sequence was analysed by Brooks and Birks (2000) from Whitrig Bog, Scottish Borders, that showed remarkable coherence between the chironomid inferred temperatures and Greenland ice core records, particularly noting the clear correlations of cool phases during the interstadial. In terms of Holocene records, the only sequence with quantitative summer temperature estimates is from Lochnagar, Grampians (Dalton et al. 2005). Although the authors argue that the major trends affecting the lake during the mid to late Holocene are due to pH changes, the temperature reconstruction does show similar general trends to a mid to late Holocene (last 6000 years) sequence from Talkin Tarn, in Cumbria (Langdon et al. 2004) suggesting that broad scale temperature patterns across northern Britain may be coherent for the mid to late-Holocene. There is one other notable chironomid record from Scotland; that from Lochan Uaine, Cairngorms (Brooks and Birks 2001) that has also been studied for a range of other palaeoclimate proxies (Battarbee et al. 2001). A climate deterioration around 2700 cal. BP was noted, although a chironomid inferred summer temperature quantitative reconstruction of this sequence has not been published.

Research capacity in Scotland

There is currently no one conducting research on chironomids in Scottish universities, and the only expert in testate amoebae analysis is Dmitri Mauquoy at the University of Aberdeen. However, the equipment needed for these analyses is not specialised or expensive, and will be present at all universities engaged in palaeoecological research. Testate amoebae analysis requires high power microscopes (the same as pollen) and minimal preparation (mainly washing in water and sieving) as strong chemicals can destroy the tests. Preparations for chironomid analysis are relatively easy as well, involving mild chemicals (KOH) and sometimes a sonic bath to disaggregate the lake sediments, and the analysis itself requires only low and high power microscopes. The research facilities are therefore present in Scotland, and it is only the expertise that is lacking.

Future research potential

There are some fundamental questions to be answered in the process of generating high quality, quantitative (e.g. summer temperatures), well resolved palaeoclimate records for Scotland. Currently very few exist, as highlighted by the recent review of British Isles palaeoclimatology in the mid to late-Holocene by Charman (2010). Filling this palaeoclimate ‘gap’ will be essential for understanding the natural and regional palaeoclimate variability on centennial to millennial timescales. More specifically, there are no current regional calibration sets, for chironomids or testate amoebae, in Scotland, and generating these datasets would further aid reconstructions through the understanding of regional variability within the modern and fossil data (cf Holmes et al. 2011).

See also the section on diatoms in the ScARF Palaeolithic and Mesolithic panel report.