UWLAS Archaeological Services

Soil/sediment analysis

Soils analysis is undertaken by Dr John Crowther, who has over 25 years’ experience of work on archaeological soils and sediments, and specialises in soil chemistry (especially phosphate) and magnetic susceptibility.

Photo: Bronze Age cairn, Carneddau, near Carno, Wales - excavated by Clwyd-Powys Archaeological Trust: phosphate and magnetic susceptibility analyses undertaken by UWLAS provided valuable insight into some enigmatic pit fills beneath the cairn, with evidence from certain pits suggesting that not only were cremation remains buried, but that soil was scraped from the surface on which cremations took place and placed on top of the cremation deposit (photo: John Crowther).

Service provided

• Site visits and advice on sampling and analytical programmes to ensure rigorous scientific investigation within the client’s budgetary constraints
• Wide range of laboratory analyses undertaken

  Key anthropogenic indicators:

  • Phosphate
  • Magnetic susceptibility and fractional conversion
  • Heavy metals (Cu, Pb and Zn)

  
  General soil/sediment characterisation:

  • Loss-on-ignition (organic matter)
  • pH and carbonate content
  • Particle size
  • Other soil/sediment properties (details on request)

• Integrated bulk sample analyses (undertaken by UWLAS) and soil micromorphology (Dr Richard Macphail, UCL)
• Assessments of representative subsets of samples
• No job too small or large – single property for an individual sample through to multi-property analyses of hundreds of samples
• Prompt reporting: publication-standard report, including statistical analysis (as appropriate), normally presented within 4 working weeks of sample receipt

See case studies (below) for illustrations of the range of work undertaken; and Background to analytical methods (pdf) for an outline of the archaeological significance of these soil properties and details of analytical techniques employed.

For further information/advice or quotation in relation to this service, please contact John Crowther (tel: +44(0)1570 424739; e-mail: j.crowther@trinitysaintdavid.ac.uk)

Soil/sediment analysis: Case studies

John Crowther has undertaken soil/sediment analysis on more than 100 sites in Britain (e.g. Middle Pleistocene hominid site at Boxgrove; the late-Mesolithic to Iron Age sediment and occupation sequence at Goldcliff on the Severn Estuary Levels; medieval depositional sequences in the Tower of London moat; William III’s Privy Garden, Hampton Court Palace; and the NERC-funded Experimental Earthworks projects at Overton Down and Wareham), including major development sites (e.g. Heathrow Terminal 5, Second Severn Crossing and Stanstead Airport); across Europe (e.g. early Neolithic settlement at Ecsegfalva, Hungary; Neolithic occupation/stabling deposits from Ramon IV rock shelter, Makhtesh Crater, Negev, Israel;  houses and yards from Regio I, Insula 9, Pompeii, Italy; early medieval stratigraphy, St Julien, Tours, France and Buraberg, Germany; medieval and later agricultural landscapes of the Plateau de Callerne, Alpes Maritimes, France - Fig. 1) and world-wide (e.g. pit fills, dated c. 2.36 million years, from the Maka’alitalu Basin, Hadar, Ethiopia; (Fig. 2) pre-historic pits, Nkang, Cameroon; excavations of enclosed hilltop site at Fiva, Central Madagascar Highlands; pit fills and old ground surfaces of Middle Neolithic Yangshao period occupation, Yilou, Huizui, China; (Fig. 3) and Indian burial site at Itapeva, southern Brazil).

The following British case studies illustrate the range of work undertaken (follow links for further details):

• Boncyn Ddol (Gwynedd, Wales): Lithics scatter site – spatial topsoil survey of phosphate and magnetic susceptibility
• Carneddau (Powys, Wales): Bronze Age cairn – investigation of old ground surface and cremation fills through phosphate and magnetic susceptibility analyses
• Guildhall (Roman amphitheatre) Excavation (London, England): Occupation surfaces/deposits from amphitheatre arena to medieval settlement – complementary use of multi-property analysis of bulk samples, soil micromorphology and palynology
• Oakham (Leicestershire, England): Field magnetic susceptibility survey of development site – use of fractional conversion in the interpretation of field survey data
• Plas Gogerddan (Ceredigion, Wales): Early Christian burial site – use of phosphate in investigation of possible graves
• Strata Florida (Ceredigion, Wales): Precinct of Cistercian Abbey – topsoil prospection survey of phosphate, heavy metals and magnetic susceptibility

Photos:

Fig. 1: Plateau de Callerne, Alpes Maritimes, France (survey and excavations supported by grant from French Ministry of Culture): investigation by UWLAS of variations in phosphate, magnetic susceptibility and organic matter concentrations through the deep colluvial sequences that have accumulated on the floors of dolines (small depressions - as pictured) and larger enclosed depressions provided insight into phases of clearance/agricultural activity on surrounding slopes (photo: John Crowther).

Fig. 2: General view of sites in Maka'alitalu Basin, Hadar, Ethiopia (project directed by Dr Erella Hovers, Institute of Archaeology, The Hebrew University of Jerusalem, Israel): Magnetic susceptibility analyses were undertaken by UWLAS on samples of reddened fills from a 'pit' feature (dated c. 2.63 m years) to complement mineralogical analyses (undertaken by Dr Francesco Berna, Universita delgi Studi di Bologna, Italy) in order to investigate whether the fills had been subject to heating/burning (photo: Erella Hovers).

Fig. 3: Yilou Excavation, Huizui, China (project directed by Professor Li Lui, La Trobe University, Melbourne, Australia and Professor Xingcan Chen, Chinese Academy of Social Sciences, Beijing, China): section showing some of lime plastered occupation surfaces investigated from Middle Neolithic Yangshao period in complementary investigations of soil chemistry/magnetic susceptibility (undertaken by UWLAS) and micromorphology (Dr Richard Macphail, UCL) (photo: Richard Macphail).

 

Boncyn Ddol (Gwynedd, Wales): Lithics scatter site

Often very little archaeological evidence remains on lithics scatter sites, other than the stone artefacts and associated debris (flakes etc.). In these circumstances spatial surveys of phosphate and magnetic susceptibility in topsoils (with samples usually sampled towards the base of the topsoil to minimise the effects of more recent activity) can potentially provide valuable insight into the way in which a site may have been used in the past and identify areas for more detailed investigation.

The lithics scatter site at Boncyn Ddol, near Roman Bridge, 5 km north of Blaenau Ffestiniog, is located on a low rocky knoll (a roche moutonnée) on the floor of the Lledr valley (Fig. 4). The lithics finds range from late Mesolithic to late Neolithic in age. In addition, trial excavation revealed two pits, with fills dating to the Early Bronze Age. A programme of investigations, which included geophysicaI surveys of the site and the analysis of pollen from peat sequences on the valley floor, were undertaken by Gwynedd Archaeological Trust (Project director: George Smith) in the hope of gaining additional insight into the site. As part of the study, samples of topsoil for phosphate and magnetic susceptibility analysis were taken at a 15 cm depth on a 10 m grid. This soil survey not only covered the ‘site’ as defined by the lithics finds, but extended beyond this, notably to the east (to provide ‘control’ samples).

Analysis undertaken on a small subset of the samples revealed a very strong relationship between (magnetic susceptibility) and (fractional conversion) (Fig. 5). In these circumstances variations in across the site (Fig. 6) can be assumed to closely reflect variations in susceptibility enhancement through burning etc., rather than natural underlying variability in the iron content of the soil. It should be noted, however, that the levels of enhancement were not especially high, with only one of the values (5.46%) exceeding 5% – which is often taken as being the lower threshold for clear signs of enhancement. The phosphate data are presented in Fig. 7. In view of the overall concentrations recorded across the site, values ≥2.50 mg/g are regarded as being indicative of enrichment.

The principal findings were that:

• Distinct areas of both magnetic susceptibility enhancement and phosphate enrichment can be identified, though these do not coincide – a finding that has been found on some other lithic scatter sites, where the areas of susceptibility enhancement have been tentatively attributed to areas of ‘domestic’ occupation and those of phosphate enrichment to nearby midden dumps.
• The highest phosphate-P concentrations were recorded in the area of control samples to the east of the ‘site’ as defined by the extent of the scatter – suggesting that this area merits further trial pitting/trenching. 

Clearly, caution must be exercised when attempting to relate spatial patterns in the properties of modern topsoils to such early phases of human activity, because of the likelihood of some degree of later phosphate enrichment and magnetic susceptibility enhancement. Here, for example, there is some evidence of EBA activity. Nonetheless, the results are valuable in that they provide independent anthropogenic indicators that need to be taken into account in the interpretation of the site.

 

Carneddau (Powys, Wales): Bronze Age cairn

The cairn is located at 419 m OD on a ridge crest c. 4 km north-east of Carno. Excavation (Fig. 8) undertaken by Clwyd-Powys Archaeological Trust (Excavation director: Dr Alex Gibson) revealed an extensive charcoal spread covering the old ground surface (OGS) beneath cairn in N quadrant. The origin of the charcoal spread was unclear from field evidence.

Magnetic susceptibility analyses on bulk samples taken from a 1-m spatial grid across the OGS (Fig. 9) provided clear evidence of localised burning close to and extending south east of a known hearth, but not of extensive burning. Thus, it would seem that the charcoal was not the product of extensive in situ burning, but resulted from the spread (deliberate?) of charcoal from specific locations (perhaps associated with cremation activity on the site).    

Several pits beneath the cairn revealed well-preserved cremation urns (as illustrated in Fig. 10). Other pits (e.g. pit 4) contained assemblages of cremation deposits and enigmatic fills. Bulk samples from these various fills were investigated using phosphate-P (total phosphate), (magnetic susceptibility), (maximum potential susceptibility) and (fractional conversion).

The results from pit 4 (presented in Table 1):

• Confirm the interpretation of the cremation deposit (note the high phosphate-P compared with the natural)
• Show the enigmatic lower fill to be highly enriched in phosphate (more than the cremation deposit) and to have an exceptionally strongly enhanced magnetic susceptibility ( and ) – clearly indicating the lower fill to be associated with cremation and likely derived from scrapings of soils from  the ground surface on which the cremation took place   

Further details Archaeological Journal, 150, 1-45 and Crowther (2003) Archaeometry, 45, 685-701. 

Guildhall (Roman amphitheatre) Excavation (London, England): Occupation surfaces/deposits from amphitheatre arena to medieval settlement

Excavations undertaken by Museum of London Archaeology Service (Excavation director: Nick Bateman) in the area of Guildhall Yard revealed a complex sequence of soils/deposits. These ranged from first century AD buried soils formed in deposits associated with the Roman amphitheatre, such as arena floor deposits and sediments related to its abandonment, through to post-Roman dark earth and 11th and 12th century moist, humic occupation deposits associated with more substantial settlement (Fig. 11), including timber structures. 

An integrated programme of soil bulk sample analysis (undertaken by UWLAS), soil micromorphology (Dr Richard Macphail; Fig. 12) and pollen analysis (Dr Gill Cruise) was undertaken in order to characterise many of the critical contexts present at the Guildhall site and to gain detailed insight into the various phases of occupation. Samples from both the Roman and medieval sequences were examined.

The 48 bulk samples analysed from the early medieval contexts revealed very marked variability in the various properties analysed (Table 2), with some exceptionally high values being recorded for the key anthropogenic indicators: phosphate-P (total phosphate), heavy metals (Pb, Zn and Cu) and magnetic susceptibility enhancement (). As is commonly the case with archaeological contexts, much of the phosphate is present in an inorganic (mineralised) form, as is reflected in the high phosphate-P:P ratios. Additionally, some notably high organic matter (LOI) and carbonate concentrations and pH values were recorded in certain contexts – the first being associated with dung and cess type deposits, and the carbonate/pH with materials derived from lime plaster etc. Very strong correlations were found between many of the properties analysed (Table 3). Most notably, all of the anthropogenic indicators (except ), including the heavy metals, show significant positive correlations with LOI, which reflects the way in which the strong anthropogenic signals at this site are closely associated with dung, cess and bone-rich deposits. This finding contrasts, for example, with previous work undertaken by UWLAS on late medieval deposits from the Tower of London moat, where heavy metal enrichment is strongly associated with industrial activity (e.g. Royal Ordnance and Mint) and craft work involving the use of metals and alloys (Macphail and Crowther in Tower of London Moat Excavation, G. Keevill, ed., 2004. Historic Royal Palaces, Monograph 1. Oxford: Oxford Archaeology).

Further information on the Guildhall site and the results of the soils analysis are presented in two monographs, both due to appear in 2007:
Bowsher, D., Dyson, T., Holder, N., and Howell, I. (in prep). The London Guildhall: an archaeological history of a neighbourhood from early medieval to modern times (MoLAS Monograph Series)
Bateman, N., Cowan, C., and Wroe-Brown, R. (in prep) London’s Roman amphitheatre: excavations at the Guildhall  (MoLAS Monograph Series)

The results of the work are also discussed in:
Goldberg, P. & Macphail, R.I. (2006) Practical and Theoretical Archaeology. Oxford: Blackwell Publishing.

 

Oakham (Leicestershire, England): Field magnetic susceptibility survey of development site

Field magnetic susceptibility surveys, such as that at Oakham (Fig. 13), are routinely undertaken as relatively quick and cost-effective means of assessing the likely archaeological potential of an area in advance of development/construction. However, one of the limitations of the technique is that magnetic susceptibility is not only affected by the degree of enhancement that may have occurred through heating/burning, but also by the background concentrations of Fe (iron) in the soil – which can exhibit wide variability within a survey area. Thus, ‘hot spots’ in magnetic susceptibility plots could potentially be the result of human activity and/or natural factors. In interpreting such plots it is therefore important to establish the likely magnitude of natural variability.

John Crowther, in collaboration with Peter Barker (Stratascan Geophysical & Specialist Survey Services) and following some earlier work by the late Tony Clark, has developed a novel method of resolving this issue. This involves laboratory determinations of (low frequency mass-specific magnetic susceptibility) and (the maximum potential susceptibility, which closely reflects the overall Fe content) for a representative subset of samples, and calculation of (percentage fractional conversion, viz:  x 100/ ).

A very strong positive linear relationship between and (e.g. at Clyst Honiton, Fig. 14c) indicates that natural background variability in Fe concentration is a minor factor, and therefore the ‘hot spots’ in the field survey are likely to associated with enhancement through burning. In contrast, and more problematic, are cases where the relationship is rather weaker (e.g. at Oakham, Fig. 14d). In these circumstances, some of the ‘hot spots’ may be attributable to higher Fe concentrations. Indeed, at Oakham, the results suggest that a band of more Fe-rich strata crosses the central part of the survey area. Even greater caution is needed when interpreting magnetic susceptibility data from different contexts on a site, where the Fe content may be highly variable from one context to another (e.g. Tower of London Moat, Fig. 14b).

Further details and illustrations of this approach are presented in Crowther & Barker (1995) Archaeological Prospection, 2, 207-15 and Crowther (2003) Archaeometry, 45, 685-701.

Plas Gogerddan (Ceredigion, Wales): Early Christian burial site

The site (Fig. 15), located close to Bow Street, 5 km north of Aberystwyth, revealed a series shallow, elongated pits (Fig. 16). Because the substrate comprises free-draining, slightly acid, fluvio-glacial deposits, virtually no traces of bone remained. During excavations undertaken by Cambria Archaeology (Excavation director: Ken Murphy) along the line of a new pipeline, various pits, ring ditches and other features were discovered. Certain pits could be clearly identified as graves from the presence of coffin stains. However, in other cases there was no visible evidence, and a programme of phosphate analysis was undertaken in the hope of gaining additional insight into these.

Unfortunately, the fluvio-glacial sands and gravels (Fig. 17) not only have a limited phosphate-retention capacity (phosphate tends to be fixed in the finer clay fraction of soils/sediments, and be readily leached through coarser matrices), but also exhibit very marked local variability in texture (with adjacent samples having widely differing proportions of finer sediments). In these seemingly unpromising circumstances, some initial exploratory work was undertaken, which established that better results were obtained by analysing the clay+silt (<63 m) fraction of samples, rather than the conventional fine earth (<2 mm) fraction; and that clear signs of phosphate enrichment was evident in a ‘control’ pit (Grave 136, Fig. 18) that was known to have contained a burial.

Having demonstrated that phosphate analysis could be used to identify burials at Gorgeddan, it was then applied to other, more enigmatic, pits. In several cases (e.g. Pit 248, Fig. 19), strong signs of phosphate enrichment (equivalent to those in the ‘control’ pit) were recorded in the central part of the feature – clearly suggesting that these were also graves.

Further details of the site are presented in K. Murphy (1992) Archaeological Journal, 149, 1-38.

Strata Florida: precinct of Cistercian Abbey

Human activities in the past often led to patterns of chemical enrichment and magnetic susceptibility enhancement (through burning) on former ground surfaces. The resulting spatial patterns, although inevitably ‘diluted’ by subsequent bioturbation, leaching, etc. and distorted to some extent by later activities, can often still be detected in modern topsoils. In view of the very limited topographic evidence within the Precinct of the Abbey, a trial topsoil prospection survey (partly funded by KEF grant: HE-08-FSP-1001) of phosphate, lead, zinc, copper and magnetic susceptibility (including fractional conversion analysis) was undertaken in two fields, Y Green and Fynwent Fawr, adjacent to Abbey.

Apart from zinc, all the properties measured displayed marked variability across the two fields which is very likely due to anthropogenic effects: phosphate enrichment probably associated with inputs of animal manure and/or human excrement (Fig. 20); lead (Fig. 21) and copper (Fig. 22) enrichment, particularly the former, possibly associated with metal working in areas with peak concentrations and more generally with human habitation (e.g. previous studies have linked low levels of heavy metal enrichment with inputs human cess); magnetic susceptibility (Fig. 23) indicative of in situ burning; and maximum potential susceptibility related to variations in iron content, which could reflect enrichment through human activity, natural variability in soil parent material and/or the effects of natural processes of iron movement and accumulation (e.g. hydromorphic panning) within soils.

Overall, the results have clearly demonstrated the potential of this type of analysis in providing insight into spatial patterns of activity at Strata Florida, in identifying specific locations for further investigation and in generating background data on variability in iron content which is critical to the interpretation of some of the geophysical survey plots.  Indeed one such location was tested by a small excavation in which a mass of burning and iron slag revealed the close proximity of an industrial area within the Inner Precinct of the Abbey. Further details, and an illustration of the ways in which soil analytical work can integrated within a wider landscape research project, are presented on the Strata Florida Project website.