Applications of micromorphology to understanding activity areas and site formation processes in experimental hut floors Rowena Y. Banerjea, Martin Bell, Wendy Matthews & Alex Brown Archaeological and Anthropological Sciences ISSN 1866-9557 Volume 7 Number 1 Archaeol Anthropol Sci (2015) 7:89-112 DOI 10.1007/s12520-013-0160-5 1 23 Your article is protected by copyright and all rights are held exclusively by SpringerVerlag Berlin Heidelberg. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”. 1 23 Author's personal copy Archaeol Anthropol Sci (2015) 7:89–112 DOI 10.1007/s12520-013-0160-5 ORIGINAL PAPER Applications of micromorphology to understanding activity areas and site formation processes in experimental hut floors Rowena Y. Banerjea & Martin Bell & Wendy Matthews & Alex Brown Received: 20 February 2013 / Accepted: 1 October 2013 / Published online: 6 December 2013 # Springer-Verlag Berlin Heidelberg 2013 Abstract Experimental buildings at Butser Ancient Farm and St. Fagans (UK) and Lejre (Denmark) were sampled to investigate micromorphology of known activity areas, to contribute to our understanding of the internal use of space in excavated buildings and formation processes of house floor deposits. The experimental buildings provided important information relating to activity residues and sediments over the 16 years that the buildings were in use. Specifically, these results contribute to our understanding of the routes and cycles for transportation of materials in occupation contexts, which can be used to inform archaeological studies. It has been possible to identify internal ‘hot spots’ within the buildings for the deposition of activity residues and for the formation of specific deposit types. Analysis also highlighted postdepositional alterations occurring in internal occupation deposits, which has provided a means of identifying roofed and unroofed spaces in the archaeological record. Keywords Experimental archaeology . Geoarchaeology . Micromorphology . Formation processes Introduction A key issue that confronts archaeologists working on settlements concerns the identification and interpretation of Electronic supplementary material The online version of this article (doi:10.1007/s12520-013-0160-5) contains supplementary material, which is available to authorized users. R. Y. Banerjea (*) Quaternary Scientific, School of Human and Environmental Sciences, University of Reading, Reading, Berkshire, UK e-mail: r.y.banerjea@reading.ac.uk M. Bell : W. Matthews : A. Brown Department of Archaeology, School of Human and Environmental Sciences, University of Reading, Reading, Berkshire, UK activity areas, and particularly the ability to identify stages in the life history of buildings (La Motta and Schiffer 1999) and the associated occupation deposits. In order to address this, archaeologists must understand the pre-depositional environment, the formation of archaeological deposits and the post-depositional processes that effect archaeological strata. Understanding these formation processes is central to interpreting the archaeological record (La Motta and Schiffer 1999; Schiffer 1987). Anthropogenic sediments within settlements have complex depositional and post-depositional formation processes, which provide challenges for geoarchaeologists in interpreting the origin of activity residues contained within them. Consequently, micromorphology has become an important tool in reconstructing the use of space and in interpreting formation processes within archaeological buildings (Matthews 1997). Experimental archaeology can play an important role in advancing such interpretations through creating a database of reference material from known activity areas and internal spaces, which can be used to provide more robust interpretations of the archaeological record. In this experimental research, buildings reconstructed from archaeological site plans at Butser Ancient Farm (Hants., UK), Lejre Historical and Archaeological Research Centre (Denmark), and St. Fagans, (National History Museum Cardiff, Wales) were subject to small-scale excavation and thin section micromorphology sampling to investigate the formation of the sedimentary record, in order to provide these comparative reference data. Most of the buildings investigated were constructed 16 years prior to sampling and have housed a range of activity spaces over their lifetime. These sites enable formation processes within buildings to be studied in a temperate climate in different geological settings, providing examples which will inform investigation and interpretation of activity traces in a range of settlement archaeological contexts, on a range of substrates. These experimental archaeological contexts enabled targeted examination of known activity areas, specific Author's personal copy 90 depositional processes and taphonomy within structures at the microstratigraphic scale, at a high chronological resolution. Specific processes such as dumping, trampling, decay and collapse were readily observed in the experimental buildings. The data from the experimental structures provide modern contextual analogues for archaeological research, supplementing data acquired from ethnoarchaeological research (Matthews et al. 2000; Milek 2012; Villagran et al. 2011) and previous experimental research (Canti et al. 2006; Macphail et al. 2004; Macphail et al. 2006; Rasmussen 2007). Examining the formation of occupation deposits within structures In many archaeological sites, few artefacts were left on the floors where they were used, and many end up in rubbish pits or middens, or were recycled (Nicholas and Kramer 2001). These artefact biographies have been documented ethnographically (Kent 1984; Kramer 1982; Schiffer 1987). Larger artefacts and bioarchaeological remains are often removed from primary accumulation contexts, either by sweeping (La Motta and Schiffer 1999; Metcalfe and Heath 1990; Schiffer 1987), or by levelling activities during refurbishment (Carver 1987). Within a building, refuse that has accumulated on the floor during primary deposition tends to consist of objects small enough to escape cleaning. Therefore, in regularly maintained areas, primary refuse will more than likely be small artefacts, as stressed by the McKellar principle (Schiffer 1987), or micro-refuse (La Motta and Schiffer 1999; Metcalfe and Heath 1990; Schiffer 1987). Previous research has highlighted a series of issues to address when reconstructing archaeological site-formation processes and thereby settlement spaces and their associated activity residues. For example, the importance of understanding the transportation mechanisms and pathways of plant remains (Greig 1982, p. 64; Matthews 2010), minerals and micro-artefacts (Schiffer 1987, p. 14; Rosen 1993) into occupation contexts, including processes such as transport by wind and introduction by trampling (Gé et al. 1993), has been shown. These processes can affect the identification of in situ activity areas. In addition, the taphonomy of plant microfossil assemblages in occupation contexts, such as the factors influencing the production and distribution of pollen and phytoliths (Tsartsidou et al. 2007; Tsartsidou et al. 2008) and their sources and catchments (Greig 1982, p. 64; Harvey and Fuller 2005; Macphail 1981; Shahack-Gross 2011) must be considered when interpreting assemblages. The differential preservation of biological materials also affects the interpretation of assemblages within the archaeological record (Boardman and Jones 1990; Robinson 2006; Shillito and Almond 2010; Stevens 2003; van der Veen 2007). Postdepositional processes such as bioturbation and decay can Archaeol Anthropol Sci (2015) 7:89–112 alter sediment/soil chemistry (Brady and Weil 2002; Breuning-Madsen et al. 2003; Canti 1999; Entwistle et al. 2000; Kabata-Pendias 2001) and rework stratigraphy and activity residues (Canti 2003; Canti 2007; Macphail 1994). Micromorphology enables investigation of the use of settlement space through identification of depositional pathways, through the study of micro-residues in situ within their sedimentary matrix and evaluation of their depositional and post-depositional histories (Jones et al. 2010; Matthews 1995; Matthews 2000; Matthews and Postgate 1994; Matthews et al. 1997; Macphail et al. 2004; Milek 2005, pp. 98–104; Milek and French 2007; Shahack-Gross et al. 2005; Simpson et al. 2006; Sveinbjarnardóttir et al. 2007), as well as chemical alterations to archaeological stratigraphy (Canti 1999; Canti 2003; Canti 2007; Courty et al. 1989). By using micromorphology to investigate modern occupation deposits in experimental buildings, this research aims to provide diagnostic sediment attributes to identify specific transportation mechanisms of materials within the archaeological record. Investigating the spatial distribution of processes and activities in structures This experimental research has also enabled horizontal sampling and spatial analysis of the composition, origin and deposition of activity residues in relation to the known activity areas. A range of agencies and processes can affect the use of space within a building and the final interpretation of archaeological artefact and biological assemblages. It is recognised that social and cultural considerations, agencies and contexts affect the selection, placement, deposition and post-depositional alterations of architectural materials and activity residues (Sillar and Tite 2000; Robb 2010; Boivin 2000; Matthews 2005). By using micromorphology on experimental contexts, this research will investigate the influence of the building superstructure such as upright supports and the location of doorways, structural modifications and the influence that the layout of internal furniture has on deposit type formation and deposit survival within buildings, in order to develop previous research concerning the production, placement and decay of construction materials (Goldberg and Macphail 2006; Matthews 1995). Materials and methods Experimental archaeology sites and sampling strategy At Butser, the Longbridge Deverill Cowdown roundhouse reconstruction, built in 1992, was excavated under rescue conditions in December 2006 because the building was Author's personal copy Archaeol Anthropol Sci (2015) 7:89–112 collapsing and replacement was imminent (Bell 2009). Consequently, the building and its deposits are not recorded in as much detail as those subsequently investigated. For the other buildings, a robust field methodology was developed to record sediments and processes in detail, and included details of construction materials, structural modifications, primary activities, impact activities, hearth usage, operational chains (‘chaîne opératoire’), intensity of use, vegetation and geology in the immediate vicinity, and wildlife infestations within the buildings (see Banerjea 2011 for full descriptions). The buildings from Lejre were sampled in August 2007 and recorded in more detail, as was the metalworking shed at Butser, sampled in December 2007. At St. Fagans, the collapsing Moel-y-Gaer roundhouse was excavated and recorded in detail by Professor Martin Bell and a team from University of Reading in 2009. Fieldwork at Butser, Lejre and St. Fagans experimental sites enabled sediment recording methods to be compared (Banerjea 2011), and presented an opportunity to collect samples from a range of occupation contexts and activity areas from buildings with different underlying substrates. A summary of the activities and processes that were targeted for sampling in each building is given in Table 1. Geology and soils of the study areas Butser Ancient Farm, Hants, UK (Fig. 1), lies on Upper Cretaceous Chalk overlaid by a thin patchy drift of soliflucted clay-with-flints, a sandy clay loam with gravel-sized flint inclusions on which a silty Ap plough rendzina with slightly alkaline pH 7.3–7.8 developed. Soils on the slope show evidence of past decalcification during stable episodes, cultivation increased chalk content and led to erosion events (Bell 1983). Construction of the experimental site followed an arable phase. The Lejre Historical and Archaeological Research Centre, near Roskilde, Denmark (Fig. 1) lies on Weichselian glacial till, the soil is an Alfisol, silty with sand and few stones an acidic pH 5.5, and a tendency to iron mobilisation (Breuning-Madsen et al. 2001). The glacial till sediment profile was recorded in a clay pit profile adjacent to Building 2. St. Fagans National History Museum, near Cardiff, Wales (Fig. 1) lies on Devensian glacial till with an acidic brown earth soil of the Radyar Series and ph 4.6–5.7. Longbridge Deverill Cowdown Roundhouse, Butser At Butser, most of the storm-damaged roof had been removed approximately 1 month before sampling, leaving only part of the west side of the roof giving partial protection to the remains of the floor on the west side of the building, providing an opportunity to compare a recently unroofed space with the original roofed space (Figs. 2 and 3). The northern half of the roundhouse was selected for sampling (Fig. 2) as this provided 91 the widest range of activities, materials and formation processes to be recorded and analysed (Table 1). The west half included the hearth, intact eaves and areas of both undisturbed and disturbed non-constructed floor surfaces (for definition see Table 2). The previous activities within the Longbridge Deverill Cowdown experimental roundhouse were mostly recalled from memory by the staff at Butser Ancient Farm. Two samples were collected covering storage locations of thatch, food preparation including some minor cereal processing, food cooking, lead working and bronze finishing (Table 1). In addition, daub and plaster that had eroded and fallen around the edges of the walls, was also sampled to characterise construction materials. Sample BLD1 was collected from the centre of the hearth to study fuel, concentrations of remaining hearth activity residues and heat effects on sediment. This had been exposed for 2 weeks prior to sampling as the roof was demolished. Sample BLD3 was collected from the semi-unroofed floor area to study postdepositional weathering effects and trampling. These are shown on Fig. 3. Metalworking workshop, Butser The metalworking workshop at Butser was a three-sided structure with an open frontage, and as a result, the area and internal deposits were exposed to weathering and erosion. The floor was a non-constructed/prepared surface that had formed through trampling of the Ap horizon. One sample, B14, was collected from the trampled silty clay loam Ap horizon (context 003) in the area of the doorway where ore crushing and bronze casting and moulding activities took place (Table 1). Building 2, Lejre Since 1974, Building 2 (Fig. 4, Table 1) had been inhabited by families recreating an ‘Iron Age life-style’, but only during the summer months and Danish Autumn school holiday. As a result the hearth, grindstone and fuel containers have been used, creating different activity areas within the building. The stable area of Building 2 housed animals between 1965 and the early 1980s and since then has been used as a storeroom for agricultural tools (Table 1). The entrance area houses have two fuel boxes located on the right side of both entrances and a grindstone on the left when exiting out of the northern door. In the living room, there are beds/benches on either side of the central hearth; limited cooking had taken place on the hearth. The modifications to Building 2 provided opportunities to target accumulations of residues for sampling. For example, the depressions caused by moving the upright posts in 1994 enabled both grinding residues and sweepings to accumulate, and the axial dung channel provided a section through the stable floor. Samples were collected from each activity context 92 Table 1 Activities and formation processes that were sampled within each building Site Butser Building Longbridge Deverill Cowdown Roundhouse Sample Location BLD1 Hearth/recently exposed space Food cooking Lead working Activities Lejre Bronze working Formation processes Trampling Weathering Burning Building 2 B14 Doorway/unroofed space Ore crushing Bronze casting L1 Stable Forge Herbivore penning Operational chains Bronze moulding Trampling Weathering Trampling Weathering Trampling Abrasion/erosion Semi-abandonment L9 Fuel basket/ post-depression Maintenance Operational chains L15 Grinding stone/ post-depression Crop processing Operational chains L39 Metal working area near hearth Iron working Food storage Stuctural modification Trampling Abrasion/erosion Structural modification Trampling Abrasion/erosion Storage of craft materials Trampling Semi-abandonment Lejre St. Fagans Building Sunken-shack Moel-y-Gaer Sample Location Activities L45 Doorway/roofed space Herbivore penning Bone working L51 Unroofed space Herbivore penning Bone working SF63 Hearth/roofed Food cooking Metal working SF68 Hearth/roofed Food cooking Metal working SF71 Doorway/roofed space Maintenance SF Wall edge Base of wall/roofed space Formation processes Trampling Abandonment Seconadry use Trampling Abandonment Seconadry use Collapse Soil development Burning Burning Trampling Abrasion/erosion Collapse Weathering Archaeol Anthropol Sci (2015) 7:89–112 Site Author's personal copy BLD3 Under eaves/semiunroofed space Thatch storage Minimal cereal processing Metalworking shed Author's personal copy Archaeol Anthropol Sci (2015) 7:89–112 93 Fig. 1 Map showing the location of the experimental sites (Butser Ancient Farm, Lejre Historical and Archaeological Research Centre, St. Fagans,National History Museum,Wales) within Building 2 (Fig. 4; Table 1): sample L1 from the stable (context 001); L9 from the sweeping residues adjacent to the fuel basket (context 004) which included the earthen floor surface (context 005); and L15 from grinding residues that accumulated in the upright post-depression (context 006) also including the earthen floor surface (context 005). strategically selected from an area which maximised the inclusion of activity residues, at equal distance from the hearth and anvil to include both ashes and hammerscale. L39 comprised the residues on the surface and the nonconstructed earthen floor (contexts 013 and 014 respectively). Analysis of sample L39 also enables the trampling effects on a non-constructed earthen floor surface to be studied (Table 1). Forge, Lejre Sunken-shack, Lejre The reconstructed forge in the Iron Age Village (Fig. 4) had been used for iron smithing activities between 1978 and 2002. Between 2003 and 2005, the central room was used to store craft materials for weaving and pottery activities and pots containing supplies of barley, wheat and horse beans. Since 2006, the building has been semi-abandoned, only utilised as a staff shelter (Table 1). At the time of sampling, the building served as a ‘display forge’ for visitors. One micromorphology sample, L39, was collected from the disused forge and was Fig. 2 Sampling area (A) and excavation sketch-plan (B) in the northern half of the Longbridge Deverill Cowdown roundhouse, Buster Ancient Farm. The following activities were recorded: storage of hay, timber, heather and reed matting under the eaves; lead-working and cooking (stews and meat) on the hearth; bronze-finishing, cereal processing, cheese-making, food preparation and spinning wool in the porch area The walls of the sunken-shack (Fig. 5) in the Viking village were assembled using a plank construction with rocks and turf stacked against the external sides. There was no internal rendering and the planks were untreated. The roof was also assembled using a plank construction; however, planks were missing due to decay and animal damage from area B (Fig. 5) despite repair in 2000. Area A (Fig. 5) remained partially turfed and water-proofed using tar paper (a glass fibre or polyester fleece impregnated Author's personal copy 94 Archaeol Anthropol Sci (2015) 7:89–112 Fig. 3 Location of samples BLD1 and BLD3 on the section drawing through the Longbridge Deverill Cowdown roundhouse, Butser. Deposit classifications as follows: LD002 is a non-constructed earthen floor; LD003 and LD004 are compacted trample deposits; LD005 and LD006 are in situ hearth ashes. Images A, B and C show the parallel orientation of plant material that is aligned parallel to the basal boundary with bituminous material) during August 2007 when the fieldwork was undertaken. Tar paper was used for water proofing when the use of the shack changed to house school children for activity demonstrations. The sunken-floor was created by digging a pit into the natural deposit of glacial till. Large granite cobbles were laid down as a floor surface on which occupation debris accumulated. The following activity contexts and stages in the life history of the building were studied: an occupation deposit, secondary use of space, abandonment and building collapse (Table 1). The changing use of space within the sunken-shack is relatively welldocumented in comparison with the other experimental buildings featured in this research. The sunken-shack was originally used for bone working during school visits during 1989–1996 (twice a week for a 10-week period each year). Then the sunken-shack was utilised as a summertime livestock shelter, firstly for goats (1996–2006) and then for sheep (2006–2007). Sweeping and both human and animal trampling were also documented for this structure. Previously undocumented, trampling and post-depositional soil development were recorded during the fieldwork as additional formation processes that had occurred prior to the fieldwork (contexts 016 and 017, respectively). Each of these processes is provisionally thought to be responsible for observed sedimentary differences between the two contexts. The clear differences between contexts 016 and 017 (Fig. 5) provided an opportunity to take comparative samples from sediment apparently deposited during the same events but which has undergone different post-depositional processes (trampling and soil development): micromorphology samples L45 (016) and L50 (017). Moel-y-Gaer roundhouse, St. Fagans The Moel-y-Gaer roundhouse (Fig. 6) was constructed by Dr P.J. Reynolds in 1992 using a circle of upright wooden posts that held up a thatched roof. Inside, the wattle and daub walls were coated with lime plaster. Four micromorphology samples were collected during the excavation of the collapsing roundhouse. Two samples, 63 and 68, were collected from the main excavation trench across the diameter of the roundhouse which truncated the central hearth. Sample 71 was collected from context 46 in the doorway section in order to study the effects of trampling and weathering (Table 1). Another sample was collected from a working section from the wall edge in order to study decay processes (Table 1) in an area where building materials had collapsed from the wall. Description Site Key micromorphological features Interpretation Primary (P) / Butser Lejre St. Fagans secondary (S) / tertiary (T) Pre-settlement horizon Sand size fraction is unsorted and unoriented. The silt-size quartz fraction is moderately sorted and moderately unoriented. All other inclusions have a random and unreferred distribution. P Sub-floor/hearth levelling/ packing Non-constructed earthen floor >8 cm in thickness, has a massive bedding structure, and comprises predominantly rock fragments and mineral with some sediment aggregates and plant remains. Unsorted. >9.5 cm in thickness. Loamy sand, sandy silt loam or silty clay loam particle size. Mid brown (PPL), dark brown (XPL). Embedded related distribution. Occasional sub-horizontal fissures in the microstructure. Anthropogenic detritus occurs at a depth of 9 cm. Sediment that formed before the location was used as a settlement/ experimental site. Sorting and orientation of inclusions indicative of cultivation in the field before construction. An earthworm sorted horizon that was also observed in the section adjacent to the earthwork at Butser (Bell 2009). Sediment that was deposited to create a level surface. P Constructed earthen floors Sandly clay loam or sandy loam particle size. Grey-mid brown (PPL), Grey-orange brown (XPL). Embedded related distribution. Anthropogenic detritus occurs in the upper 2 cm. Surfaces created from the existing ground surface rather than a floor created by the re-deposition of sediment from elsewhere. This floor creation process has been previously described as a beaten earth floor (Macphail et al. 2004). Trampling and bioturbation are responsible for any downward movement of anthropogenic debris. Surface may be more reactive than a constructed earthen floor. Using earth either in its unaltered form or with additional sand as a stabiliser, and/or plant remains to prevent cracking. This method of floor building is frequently used in earth building (Norton 1997; Houben and Guillaud 1994; Keefe 2005). Fabric used as earthen building material, such as daub or render, either in its unaltered form or with additional sand as a stabiliser, and/or plant remains to prevent cracking. This method of floor building is frequently used in earth building (Norton 1997; Houben and Guillaud 1994; Keefe 2005). Deposition of ‘clods’ of damp sediment which frequently form superimposed micro-laminations when deposited with downward compression onto a hard surface. Sediment containing high frequencies of anthropogenic material such as sweepings. Re-deposited away from their primary area of deposition by anthropogenic processes. Accumulation contexts from the experimental sites contain a very specific range of anthropogenic inclusions reflecting the activities recorded in the field. Material lies in close proximity to hearths. Rubified sediment aggregates, daub/furnace lining, charred organic remains, ash and fresh plant material have been moved from the primary place of deposition within the hearth itself. These ashes have accumulated in situ and therefore lie in their primary place of deposition. Earthen building material Compacted trample Soft materials oriented parallel to surface of the boundary below. Harder components are unoriented and unrelated. Discard deposits Unsorted. Inclusions are unoriented, unrelated, random and unreferred. Diverse range of components of geological source, and high frequencies of anthropogenic debris. Non-organic, sand size inclusions with a parallel orientation to the basal boundary are characteristic of these accumulation deposits. Sorting is often bimodal. Unsorted. Inclusions are unoriented, unrelated, random and unreferred. High frequencies of rubified sediment aggregates, daub/furnace lining, charred organic remains, ash and fresh plant material. Laminated bedding structures which contain microlenses of ashes, charred plant remains and rubified sediment aggregates which are orientated parallel to the basal boundary. Extensive weathering: very abundant evidence of mesofaunal bioturbation, (>20 %), occasional dusty impure clay coatings Accumulation deposits Rake-out material In situ hearth ashes X S X S X X X X X S S, T X X S P X X X X X S P P, S, T Author's personal copy Deposit type Archaeol Anthropol Sci (2015) 7:89–112 Table 2 Descriptions of deposit types and the experimental sites at which they were identified X X X X 95 Author's personal copy X Particles of rock, mineral and organic debris (dung) that have been (2–5 %), rare silty clay coatings (<2 %); rare iron reworked and transformed by post-depositional biological translocation (<2 %) and occasional vivianite neomineral processes to form a soil. formation (2–5 %). Post-depositional processes such as surface earthworm casts and vegetation growth and rounded earthworm granules, 20 % were also evident during excavation. Mixed compacted Accumulation processes are evident by both the orientation and Deposit which formed by both trampling and accumulation processes. S, T Thin lenses with strong parallel orientation and distribution of distribution of sand-size inclusions and laminated bedding trample and components generally suggest periodic accumulation and structures. Softer inclusions such as plant remains which have accumulation compaction over time (Goldberg and Macphail 2006). a parallel strong orientation aligned with the basal boundary are characteristic of deposition by trampling. Linked and coated lenses (1–2 mm in thickness) interspersed Embedding may have occurred when people stood here to empty P, S Mixed dump with embedded lenses. sweepings into the adjacent basket. deposit and accumulation X X Laboratory methodology Post-depositional soil formation/ 'dark earth' Key micromorphological features Deposit type Description Table 2 (continued) Interpretation Primary (P) / Butser Lejre St. Fagans secondary (S) / tertiary (T) Archaeol Anthropol Sci (2015) 7:89–112 Site 96 Micromorphology samples were oven dried at 40 °C, impregnated with epoxy resin and cured. The impregnated blocks were cut, mounted to slides and lapped to a standard geological thickness of 30 μm. Micromorphological investigation was carried out using a Leica DMEP polarising microscope at magnifications of ×40–×400 under plane polarised light (PPL), crossed polarised light (XPL) and oblique incident light (OIL). Thin section description was conducted using the identification and quantification criteria set out by Bullock et al. (1985) and Stoops (2003), with reference to Courty et al. (1989) for the related distribution and microstructure, Mackenzie and Adams (1994) and Mackenzie and Guilford (1980) for rock and mineral identification, and Fitzpatrick (1993) for further identification of clay coatings. Tables of results use the descriptions, inclusions and interpretations format used by Matthews (2000) and Simpson (1998). Post-depositional alterations were identified and quantified using a visual estimate (Bullock et al. 1985). Deposit type classification The depositional events are characterised by the following diagnostic sedimentary attributes: sorting, related distribution, orientation and distribution of the inclusions, and bedding structure (for full deposit type descriptions see Banerjea 2011). The range of deposit types that were identified from experimental sites are summarised in Table 2. To determine deposit type, each unit was grouped using diagnostic sedimentary attributes and inclusions to provide information concerning the origin of inclusions, transportation mechanisms and deposition processes. To assess the origin of sediment components, descriptions were made of particle size, shape and the composition of the coarse and fine fraction, particularly the frequency of rock, minerals and anthropogenic inclusions. Results and discussion Following observation and description, deposits were grouped into deposit types. Transportation mechanisms observed and described include wind and water transportation, trampling, construction and accumulation. Deposition ‘hotspots’ were identified as well as post-depositional alterations. These are discussed in detail below, and presented in Tables 3, 4 and 5. Transportation processes of materials in experimental huts Field observations identified the following routes and cycles of transportation of materials within the experimental huts: Author's personal copy Archaeol Anthropol Sci (2015) 7:89–112 97 Fig. 4 The Iron Age Village, Lejre Historical and Archaeological Research Centre. A: Photograph of Building 2. B: Sketch plan of the interior of Building 2. C: Photograph of the Forge. D: Sketch plan of the interior of the Forge wind and water/rain-induced transportation, trampling and transportation by human agencies such as maintenance and discard processes. Results of observations of accumulation processes are given in Table 3. Results of observations of trampled deposits are given in Table 4. Wind and water or rain-induced transportation In the external unroofed space outside the metalworking workshop at Butser, the accumulated deposit (context 003, sample B14) has a moderately sorted silt component which may have been transported by wind or rain. The bimodal sorting of poorly sorted/unsorted sand in this moderately sorted silt, suggests that specific activities contributed to the input of each these components. Similar bimodal sorting characterised internal accumulated occupation deposits elsewhere at Butser, including in situ ashes, and at Lejre. The moderately sorted silt component in deposits within internal spaces in the Forge at Lejre, may have been left behind on the floor after sweeping had removed larger sand-sized components (La Motta and Schiffer 1999; Metcalfe and Heath 1990), as percolated/ ’sieved’ sediments through mats (Matthews et al. 1997:289), or wind-blown sediments close to an entrance . Trampling as a depositional pathway of materials Compacted trampled deposits were identified in wet, open or partially open, experimental buildings at Butser, in samples BLD1 and BLD3, in an area where the roof had been partially removed (Fig. 3). They were also identified in mixed accumulation/trample deposits that occurred in doorways at Lejre in the sunken-shack (sample L45), and in the Moel-y-Gaer roundhouse St. Fagans (sample 71) (Fig. 6). Trampling acted as a depositional process (Table 2) transporting ‘clods’ of sediment from the soles of feet onto the floor surface. The parallel orientation of soft materials such as plant remains suggests that downward compression aligned these malleable inclusions parallel with the surface of the context below (Fig. 3). Harder materials such as rock fragments, minerals and metallurgical residues (Fig. 6a–c) are unoriented, randomly distributed and do not lie referred to any other components. The deposition of ‘clods’ of sediment from the soles of feet formed lenses of sediment when compressed during deposition on comparatively dry surfaces in roofed spaces (Fig. 3 samples BLD1 and BLD3, mixed trample/accumulation deposits in sample L45, Lejre and sample 71, St. Fagans); trampling in Author's personal copy 98 Archaeol Anthropol Sci (2015) 7:89–112 Fig. 5 Location of micromorphology samples collected within the sunken-shack, Viking Village, Lejre Historical and Archaeological Research Centre, Lejre: L45 from side A, the roofed space; L51 from side B, the unroofed space wet sediments can result in homogenous thick layers (Matthews 1995). Thin lenses with strong parallel orientation and distribution of components also generally suggest periodic accumulation and compaction over time (Goldberg and Macphail 2006, p221).This superimposition of lenses often results in the context having a laminated bedding structure. Compacted trample deposits can contain debris from primary and secondary, or even tertiary activities. At Butser, the same types of rock and minerals types in the compacted trample deposits were also present in the external presettlement construction horizon and the internal nonconstructed earthen floor, indicating that they could have been collected on the soles of feet from either inside or outside, the Longbridge Deverill roundhouse. At St. Fagans, the mixed trample/accumulation deposit in the doorway of the Moel-yGaer roundhouse contained metallurgical residues (metal fragments, <30 %, and slag, <15 %) that had most likely been trampled into the building from a nearby metalworking area, or perhaps from the hearth area within the building where a few metalworking residues (<15 %) occurred in primary contexts. Based on these observations, it is important to consider that, when studying the use of space within archaeological buildings, artefacts and biological remains within doorways may not only reflect the activities within the building, but also those activities that are taking place in open spaces and adjacent areas around buildings. In order for compacted trample deposits to form, this research has demonstrated that damp environmental conditions must be present. Damp conditions are crucial for the formation of compacted trampled deposits or deposits that contain a proportion of material which has been deposited by trampling. Damp conditions are evident by the concentrations of eroded building materials from the walls, clay coatings and chemical alterations such as neomineral formations (Table 4). Building collapse, or the partial removal of roofs, also played an integral role in the formation of internal deposits of compacted trample, by contributing higher densities of sediment and mud materials to locales, which were later frequented and trampled. A mixed trample/accumulation deposit developed within the Sunken Shack, Lejre, after the roof has failed (Fig. 5). At Butser, compacted trample deposits (contexts LD003 and LD004) had formed within the Author's personal copy Archaeol Anthropol Sci (2015) 7:89–112 99 Fig. 6 Location of micromorphology sample 71 on the section through the doorway of the Moel-y-Gaer roundhouse, St. Fagans (top left). Images A-E are microscopic residues from metalworking activities within the mixed trample/accumulation deposit in the doorway, of the roundhouse Longbridge Deverill roundhouse during a period of building collapse when the interior of the building became wet in the month or so before sampling, but the area was still visited. Chemical weathering has also been observed in compacted trample deposits from experimental buildings (Table 4). The identification of compacted trampled deposits has the potential to be a useful indicator of doorways or pathways in the archaeological record, particularly where there is little remaining archaeological evidence for the superstructure of the building. Transportation by human agency: inclusions within floor construction materials Earth materials were used in the construction of levelling and floor surfaces (Table 2) within the experimental buildings, and inclusions within floor surfaces can either be primary or secondary depositions. Deposits classified here as nonconstructed earthen floors (Table 2) refer to surfaces on the original ground surface, often including trampling of vegetation (Macphail et al. 2004), which therefore can be considered as a primary constituent of a floor surface. Table 3 Sediment attributes of experimental archaeology deposits that have formed by accumulation processes Deposit type number Sample Context Building number number Accumulation L39 013 B14 003 BLD1 LD005 In situ Ashes Location in Particle size building Sorting Inclusions: Orientation and Distribution Larger sand sized particles (>250 μm) are unorientated and unreferred. Others are have a linear and parallel distribution and are moderately orientated. Metalworking Open porch Sandy clay Bimodal: Unsorted sand Mostly Unorientated and unrealted. Random and unreferred. Sand-sized shed area loam size, moderatley inclusions have a linear and inclined sorted silt. distribution and moderately orientated. Hearth Coarse sandy Bimodal: Unsorted sand Unorientated and unrealted. Random and Longbridge unreferred. But most charcoal and plant clay loam size, moderatley Deverill fragments have a linear and inclined orted silt. Cowdown distribution and moderately orientated. R/H Forge Close to hearth Silt loam Bimodal: mod sorted silt in poorly sorted sand Author's personal copy Archaeol Anthropol Sci (2015) 7:89–112 d b c c c b c Organic staining Specific examples of this observed at the experimental sites include: the Longbridge Deverill Cowdown roundhouse, Butser, context BLD002 (Fig. 3); the disused forge, Lejre, (Fig. 4) context L014; and the Moel-y-Gaer roundhouse St. Fagans (Fig. 6), sample 68). Sub-floor levelling material and constructed earthen floors (Table 2) comprise sediment that has been transported from elsewhere and deliberately modified with aggregate to increase the strength of the material, or vegetable stabilisers to prevent cracking (Norton 1997; Matthews 1995; van der Veen 2007), as observed in Building 2, Lejre. In this case, these inclusions can be considered to be in a secondary context in floor materials. Primary and secondary materials in floor surfaces, therefore, were transported through very different depositional processes. Coarse sand aggregates increase the strength of an earthen building material (Berge 2000). This is evident at Lejre where sub-floor levelling deposits and constructed earthen floors within Building 2 were made from glacial till, quarried from a clay pit adjacent to Building 2. These floor surfaces are hard and compact. The glacial till floors included coarse flint rock components and minerals such as quartz. At Lejre, it was used in sub-floor levelling material (context 002, sample L1) and the original constructed earthen floor (context 005, sample L9) in Building 2 (Fig. 7), and is broadly similar to the nonconstructed earthen floor (sample L39), in the Forge, Lejre (Fig. 8), but with a greater frequency of flint inclusions, feldspars, amphiboles and chlorite minerals in Building 2 (Fig. 7) than the non-constructed floor (context 014) in the forge (Fig. 8). Context 020, Building 2, Lejre, has a different rock and mineral composition than the original constructed earthen floor (context 005), specifically the inclusion of chalk fragments (Fig. 7), which reflects the use of a different source material used to repair the floor. This ‘chalky clay’ was collected from a source some distance away (Hans Ol Hansen personal communication). It is not known whether higher frequencies of flint feldspars, amphiboles and chlorite minerals in the sub-floor levelling and original constructed earthen floor materials Building 2 relate to the addition of aggregate during preparation of the sub-floor levelling material and constructed earthen floors, or whether the variability reflects the exploitation of different seams of glacial till source material. As observed in the construction of modern brick making, the nature of the quarrying method can influence the nature of the source material. This depends on whether a uniform sediment seam was selected, or whether downwards excavation quarried and mixed different sediment seams (Prentice 1990). c a Vivianite neomineral formation b c LBD R/H, Butser LBD R/H, Butser Sunken-shack, Lejre R/H, St. Fagans Abundant 10–20 % d Rare <2 % Occasional 2–5 % Many 5–10 % c b a Mixed trample/ accumulation BLD1 BLD3 L45 SF71 Compacted Trample LD004 LD003 016 46 Sample number Deposit type Context number Building/site Dusty impure clay coatings: unlaminated d Dusty impure clay coatings: microlaminated Iron Chemical alteration Translocation Site and sample information Table 4 Weathering within experimental (Butser, Lejre and St. Fagans) compacted trample deposits and mixed trample/accumulation deposits Manganese neomineral formation Decay Spherical fungal spores 100 Transportation by human agency through accumulation processes Accumulation contexts from the experimental sites contain a very specific range of anthropogenic inclusions reflecting the Author's personal copy Archaeol Anthropol Sci (2015) 7:89–112 101 Table 5 Differences between the sediment properties (field descriptions) in the roofed and unroofed sides of the Sunken Shack, Lejre Sediment attribute Context (016) roofed Context (017) unroofed Thickness Bedding Colour Consistency and structure 1–5 cm Layered (lenses of dung) Brown black Strong. Platey (coarse material) and blocky (fine) structure. Non-plastic and non-sticky. Sand Trampling. Wind and rain bringing in leaves and acorns. 4–10 cm Massive Very dark brown Very weak. Crumb (fine–med) structure. Slightly plastic and slightly sticky. Sandy clay Trampling. Root activity. Soil development. Surface worm casts. Wind and rain. Unoriented and random Dung (R) <1 cm, 10 % Charcoal (A) 1–2 cm, 5 % Bone (A) 2–5 cm, 10 % Acorns (R) 2–3 cm, 5 % Leaves (R) 5–7 cm, 5 % Earthworm granules (R) 0.4 cm, 20 % Particle size Post-depositional alteration Inclusions: orientation/distribution Inclusions: composition, shape, size and abundance Parallel and random Dung (SR) <2 cm, 60 % Straw (A) <10 cm, 5–10 % Flint (A) <5 cm, 5 % Fur (R) 2–4 cm, <5 % Acorns (R) 2–3 cm, 5 % Leaves (R) 5–7 cm, 5 % activities recorded in the field. Accumulations have bimodal sorting, poorly/unsorted sand in moderately sorted silt, an embedded and coated related distribution, and the inclusions are moderately oriented and linear and parallel/inclined in distribution (Table 3). In comparison, discard deposits are unsorted, often have an intergrain aggregate and/or linked and coated related distribution, and inclusions are unoriented and randomly distributed, indicating higher energy in the deposition (Table 2); they also contain a greater diversity and frequency of inclusions than accumulation deposits. Fig. 7 Spatial distributions of rock and mineral inclusions within constructed earthen floors and sub-floor levelling material in Building 2, Lejre Author's personal copy 102 Archaeol Anthropol Sci (2015) 7:89–112 Fig. 8 Rock and mineral inclusions within the non-constructed earthen floor (context 014) and the overlying accumulation deposit (context 013) in the Forge, Lejre Compacted trample deposits also show linear and parallel distribution and local orientation of plant fragments (Table 2), whereas in accumulations, the linear and parallel distribution and local orientation occurs in the coarser harder materials (Table 3). Periodic, accumulation processes often result in the buildup of primary activity residues, attested in thin section (Table 3) by parallel orientation of coarse components aligned with the basal boundary (Goldberg and Macphail 2006). Examples of this from these experimental sites include: pellets of herbivore dung (context 016) in the Sunken Shack, Lejre; daub in Building 2, Lejre (context 004); and metalworking residues and charcoal within the accumulation context (003) outside the Metalworking shed, Butser. Micro-laminations, such as superimposed fine lenses of calcitic ash that formed in in situ hearth ashes (Fig. 3, sample BLD1 context LD006) within the central hearth, Longbridge Deverill roundhouse, Butser) often also indicate repeated, periodic accumulations (Goldberg and Macphail 2006). Within in situ hearth ashes, such as context LD006 Butser, charcoal and plant remains were observed to be moderately oriented, with a linear and inclined distribution, often lying referred to larger charcoal fragments which they have fallen against during deposition (Table 3). Identifying interior ‘hot spots’ of deposition Composition and Spatial Distribution of Discard Deposits Micromorphological observations from Building 2, Lejre (Fig. 9), demonstrate that the discard deposits in interior spaces may comprise materials that were transported and incorporated from around the entire building, as well as higher frequencies of a particular material where the catchment for the dump is close to a specific activity area. Within Building 2, discard deposits 006, 021 and mixed discard/accumulation Author's personal copy Archaeol Anthropol Sci (2015) 7:89–112 103 Fig. 9 Spatial distributions of building materials, artefacts and bone within occupation deposits in Building 2, Lejre 004 formed within depressions in the floor surface caused by the moving of upright roof support posts which are located close to the wall edge and external doorways in the living area. Fragments of daub and lime plaster building materials occurred in these dump deposits but not in the dump deposit that had formed within an axial dung channel in the former stable area (context 001) within Building 2 (Fig. 9). The occurrence of building materials in these locations may be due to the erosion by wind and rain, and abrasion by passing traffic, of daub and plaster from the walls in the area of the doorways that have then been trapped in the depressions during sweeping. Within the Moel-y-Gaer roundhouse at St. Fagans, the micromorphology observations of deposits adjacent to the edge of the wall showed that fragments of eroded lime plaster had accumulated in order of erosion from the wall face (Fig. 10). The earthen floor deposit below lens b contained unlaminated and laminated silty clay and clay coatings (2–5 %). Microlaminated clay/silty clay coatings exhibiting regular lamination and high birefringence indicate strong parallel orientation of fine particles as a result of slow aqueous deposition under calm conditions (Courty et al. 1989). It is possible that the alignment of clay and silt particles in this area is due to damp/periodic puddles, which may also have caused the erosion of plaster and daub from the wall. Higher frequencies of specific materials from localised activities within Building 2, Lejre, are present in dump deposits 001 and 006, proximate to the foci of these activities. In Building 2, context 001 contains dung representing the former use of the stable area (Fig. 11) and context 006 contains higher frequencies of fresh plant material as a result of the incorporation of sweepings from the grinding stone (Fig. 11). Post-depositional alterations to hut floor deposits The creation of new deposit types through post-depositional processes Micromorphology has demonstrated that post-depositional processes can create new re-deposited lenses through clear examples, firstly within the Moel-y-Gaer roundhouse, St. Fagans, and secondly within the sunken-shack, Lejre. In the Moel-y-Gaer roundhouse, micromorphological analysis shows that lens D (Fig. 10) was formed by the infilling of a mesofaunal channel with material from lens B and so this accumulated material is a secondary constituent. Ant and hornet activity were both observed within the Moel-y-Gaer roundhouse during excavation. The activity and effects of ants on the soil and archaeological deposits are less well Author's personal copy 104 Fig. 10 Micromorphology features within deposits by the edge of wall, Moel-y-Gaer roundhouse, St. Fagans. The sample was collected from the slot marked on the right of the excavation photograph (top left). This sample comprises lenses a-e. Lenses a, b and d are accumulation deposits, lens c is Archaeol Anthropol Sci (2015) 7:89–112 earthen building material, and lens e is a non-constructed earthen floor. A: Lens b, fresh plant material embedded within a plaster fragment (XPL). B: Lens e, finer quartz particle size and less rubified clay matrix than lens b (XPL). C: Lens c, coarse quartz particle size and more rubified than lens d (XPL) Fig. 11 Spatial distribution of plant remains within occupation deposits in Building 2, Lejre Author's personal copy Archaeol Anthropol Sci (2015) 7:89–112 105 understood than for earthworms or termites, for example Brady and Weil (2002). In thin section, there are clear traces of mesofaunal activity, most probably from ant activity and is represented by horizontal burrows that truncate the earthen building material (Fig. 10, lens C). Material from lens B has fallen through into a horizontal channel that the ants created between the earthen building material (lens C) and the nonconstructed earthen floor (lens E) forming lens D (Fig. 10). Failing roofs can lead to post-depositional alterations that radically transform the deposits within the building. The sunken-shack, Lejre, was first used to demonstrate boneworking craft techniques to school children, before a change in use when it was used to house goats. A subsequent decision was taken by staff at Lejre to use the sunken-shack as a shelter for sheep instead of goats, as the goats had destroyed the roof on one side, opening it up to the effects of weathering. Substantial post-depositional alterations had occurred within 2.5 years in the unroofed half of the building leading to soil formation. This sequence of use and post-depositional alterations is clearly attested in the microstratigraphic sequence. The differences between the roofed and unroofed side of the building were clearly visible in the field and from the initial field descriptions (Fig. 12; Table 5). Although the deposit across the floor of the sunken-shack had initially been the same stable floor deposit on both the roofed and unroofed sides of the building, micromorphology revealed the extent to which the unroofed side of the building had been radically transformed by exposure to weathering (Fig. 12), including for example, dissolution of faecal spherulites, probably by increased acidic conditions, below pH 7.7 when spherulites dissolve (Canti 1999), and there is less fresh plant material in 017 than 016 and phytoliths are present probably due to accelerated decay processes in the unroofed side of the building. Fig. 12 Comparative sediment features from the roofed and unroofed spaces within the sunken-shack, Lejre. Note the linear and parallel laminations of the dung lenses in sample L45 compared to the unoriented particles of dung in sample L51. Image A shows calcareous faecal spherulites that were not present in sample L51. Images B and C show bone fragments that did not occur in sample L45 Abrasion processes on floor surfaces within buildings Sweeping and trampling are major mechanisms in abrasion, disaggregation and transportation of floor materials and accumulated deposits. They are most probably the transportation mechanisms for some of the rock and mineral inclusions within discard deposits 006 and 021 from Building 2, Lejre. The fragments of granite most probably derive from abrasion through use of the adjacent granite quern stone and were subsequently redeposited in the process of sweeping, and/or may also originate from erosion of the granite cobbles in the doorway (Fig. 7). By comparing rock and mineral composition of constructed earthen floors and sub-floor levelling material in Building 2 (Fig. 7) with the composition of the overlying dump and accumulation deposits (Fig. 13) it is possible to infer that certain rock and minerals in dump deposits did not occur in the directly underlying constructed earthen floors, but were eroded from elsewhere in the building Author's personal copy 106 Archaeol Anthropol Sci (2015) 7:89–112 Fig. 13 Spatial distribution of rock and mineral inclusions in occupation deposits within Building 2, Lejre and transported by sweeping processes. For example, chalk fragments occur in dump deposit 004 in Building 2 (Fig. 13), but not in the underlying constructed earthen floor 005 (Fig. 7). However, chalk fragments do occur within a repaired patch of constructed earthen floor material, context 020, on the other side of Building 2 (Fig. 7). Granite does not occur in sub-floor levelling context 002 but does occur in overlying dump context 001 and elsewhere in Building 2. This suggests that granite fragments may have been transported by feet trampling across the granite cobbles and the area adjacent to the granite quern stone, and through the stable area rather than from incorporation due to the erosion of the underlying floor (Figs. 7 and 13). By comparing rock and mineral assemblages in floor materials and their overlying occupation deposits, sequences of activity and repair may be identified. The erosion of building materials The categories of building materials identified within secondary occupation deposits from Building 2 (Fig. 9) and primary occupation deposits in the Sunken Shack at Lejre (Fig. 14) in thin section correspond with those recorded in the field in the construction materials of the buildings. Micromorphological analysis suggests that their distribution is quite localised within the buildings. The original bitumen roofing had become weathered and fragmented and incorporated into 017 within the now unroofed side of the Sunken Shack, Lejre. Fragments of daub in occupation deposit 004 and daub and plaster within 006 and 021, Building 2 Lejre within 1 m of the wall. In thin section, the parallel orientation of the fragments to the basal boundary on 004, suggest that it was eroded daub from the walls. By comparison, the haphazard unoriented distribution of fragments of building materials in dumps 006 and 021 suggest that these were redeposited by sweeping. Weathering processes and trampling appear to have eroded granite floor cobbles in the Sunken Shack, Lejre, as micro-fragments of the cobbles became incorporated, most probably through bioturbation, into the overlying occupation deposits (Fig. 15) which had formerly been a mixed compacted trample/accumulation deposit (as on the roofed side of the building) but now post-depositional soil formation on the unroofed side of the building. Bioturbation (eg mixing by fauna) also introduced unburnt bone fragments which had been deposited during previous craft activities into post-depositional soil formation deposit (context 017) (Fig. 14). Author's personal copy Archaeol Anthropol Sci (2015) 7:89–112 107 Fig. 14 Comparative figure showing the frequency of building materials and artefacts before (context 016) and after (context 017) weathering in the sunken-shack, Lejre Localised chemical alterations within occupation deposits Experimental research has provided crucial observations concerning the effects of fluctuations in the oxidised/ reducing conditions within non-waterlogged occupation deposits from temperate sites (Fig. 16). Within occupation deposits inside buildings, chemical alterations can play a key role in the formation of silty clay coatings, in addition to turbulent hydraulic conditions and mixing and rotating of floor deposits inducing clay and silt translocation at temperate sites (Courty et al. 1989; Goldberg and Macphail 2006). Chemical alterations and changes in the redox (oxidation reduction processes) conditions can lead to the dispersal of silt and clay particles (Brammer 1971; French 2003) within highly localised areas, often in lenses or patches with decaying organics, in occupation deposits within roofed spaces. At Lejre, moderately or strongly oriented silty clay coatings in dump deposit 021 and mixed dump/accumulation deposit 004 are associated with areas of organic decay and staining (Fig. 16c and f). As this spaced is roofed and not open to the effects of wind and rain, it is possible that chemical changes caused by the decay of organic matter, and turbulent conditions Author's personal copy 108 Archaeol Anthropol Sci (2015) 7:89–112 Fig. 15 Comparative figure showing the frequency of rock fragments and minerals before (context 016) and after (context 017) weathering in the sunken-shack, Lejre from dumping processes and trampling, are causing the mobilisation of silty clay particles, rather than rainwater flowing though the profile. Within this context, strongly oriented clay coatings of voids and minerals formed in linear and parallel lenses occur horizontally across deposit 021 in clusters associated with areas of plant decay and the breakdown of daub. These areas of clay coatings are a different colour (orange PPL, dark orangey with occasional dark yellow XPL) from the surrounding matrix (dark brownish grey PPL, very dark greyish brown with hints of very dark brownish red XPL). Iron and manganese appear to have replaced organic material within the deposits LD005, Butser (Fig. 16a and d) and L021, Lejre (Fig. 16c and f) and there is also organic staining in context 001 (Fig. 16b) and context 021 (Fig. 16e) and similarly, also in deposit 004. The silty clay coatings may have been impregnated with iron. Iron and manganese replace the organic matter. At Buster, ashes within the hearth that had been left open to the effects of rain, have also begun to be replaced by manganese (Fig. 16a). Here, the silty clay coatings are impregnated with iron, and clay coatings within a decaying Author's personal copy Archaeol Anthropol Sci (2015) 7:89–112 109 Fig. 16 Micromorphological features resulting from localised redox processes on experimental sites: manganese replacement of ashes, PPL (A) and XPL (D), sample BLD1, context LD005, Butser; image B, decaying plant material with organic staining (top left ), manganese replaced plant remains (top centre), iron mottles (bottom), PPL, sample L1, context L001, Lejre; band of silty clay translocation directly below manganese replacement of decaying plant remains, PPL (D) and XPL (F). Sample L15, context L021, Lejre; image E, manganese replacement of decaying plant remains, PPL, sample L15, context L021, Lejre piece of wood have internal iron mottles. Within the in situ hearth ashes at Butser contexts LD005 and LD006, moderately oriented silty clay coatings are sometimes mixed with ash and are different in colour from surrounding ash matrix, suggesting that it has been incorporated from another source. They are a similar colour to material in overlaying trample deposit LD004, and may also be impregnated with organic staining. These characteristics resemble those from other studies, in which clay coatings which have a different colour from the surrounding sediment matrix are suggested to have been translocated from elsewhere and washed into the sediment profile (Brammer 1971; French 2003). Where clay coatings are the same colour as the surrounding matrix, as observed in some seasonally flooded sediments, Brammer (1971) and French (2003) have suggested that seasonal alterations of reduction and oxidation of topsoils may lead to the dispersal of fine material during the period when the iron present is strongly reduced thus, as suggested by French (2003), causing clay coatings to become impregnated with iron oxides and hydroxides. Where animal penning has taken place, animal trampling and inputs into the soil of organic matter-rich dung and liquid waste mobilises fulvic acid to produce dark reddish brown clay coatings (Macphail and Linderholm 2004; Macphail and Cruise 2001). It is apparent at both Butser and Lejre that chemical changes related to the decay of organic matter and dung and its replacement with iron and manganese (Fig. 16), in conjunction with anthropogenic and livestock disturbance, is causing fine silts and clays within the deposits to disperse. However, further research monitoring redox potential over time in nonwaterlogged occupation sediments in both roofed and unroofed contexts is required to understand these processes further. The timescales for these chemical changes at Butser and Lejre suggest that these processes can take place within months after deposition. These processes are more prolific at Lejre. The more alkaline calcareous environment at Butser, in conjunction with earthworm activity, may prevent localised chemical changes (Crowther et al. 1996), which require more acidic anaerobic conditions, from taking place to the extent which can be seen at Lejre, although manganese replacement of calcitic hearth ashes does occur (Fig. 16a). Conclusions Examination of field and micromorphological characteristics of architectural materials, surfaces and deposits at these experimental sites has produced significant observations that have further developed identifications of formation processes in the archaeological record, particularly the identification of trampling, the radical transformations that take place as a result of post-depositional events both in roofed and in poorly roofed spaces, and the timescales over which processes occur. Author's personal copy 110 The parallel orientation of soft materials such as plant remains suggests that downward compression aligned these malleable inclusions parallel with the surface of the context below. Harder materials such as rock fragments, minerals and metallurgical residues are unoriented, randomly distributed and do not lie referred to any other components. The deposition of ‘clods’ of sediment from the soles of feet formed lenses of sediment when compressed during deposition on comparatively dry surfaces in roofed spaces. Post-depositional processes have been shown to have the ability to transform stratigraphy to create entirely new deposittypes. At St. Fagans, ant activity created a new layer below those that had been previously deposited. At Lejre, it has been determined that failing, leaking roofs can radically transform occupation deposits within buildings and eventually lead to soil development, which may resemble a ‘dark earth’ (Macphail and Courty 1984; Macphail et al. 2000; Macphail et al. 2003), including anthropogenic debris from the period of building use. This could have significant implications for the identification of structures in the archaeological record, particularly when superstructure components such as walls and beam slots are not visible, and these deposits may be misinterpreted as garden soils in external areas. Localised redox processes can play an important role in the mobilisation of silts and clays as a result of weathering and decay processes within occupation deposits. However, it must be noted that these processes are difficult to relate to specific stages in the life of archaeological structures. This experimental analysis has shown that chemical alterations can occur within months of deposition. Experimental research has also demonstrated that geology will play an integral role in the formation of localised redox processes in occupation deposits. Additional influences are the types of source materials for building and the function of the area in terms of the inputs of residues. Certain locations within buildings have been identified where specific deposit types have both formed, and are more likely to survive. The occurrence of compacted trample deposits may be used to identify damper areas of buildings such as doorways or semi-open spaces in the archaeological record. The study of the spatial distribution of discard deposits within experimental buildings has demonstrated that their formation and survival is dependent on the occurrence of catchments that were formed by the modification of super-structural components causing depressions in the floor, and that the protection of residues by internal furniture leaves areas that escape sweeping such as the junction of the base of the wall and the edge of the floor. Building reconstructions at experimental sites can act as ‘working laboratories’ for archaeologists to study the formation of the archaeological record and life histories of buildings, provided basic constructional and activity information is regularly recorded. However, experimental sites are not ethnoarchaeological case studies as the buildings have often Archaeol Anthropol Sci (2015) 7:89–112 been utilised to different extents. The experimental sites at Butser, Lejre and St. Fagans have provided opportunities to spatially examine specific activity areas and formation processes within medium term experimental archaeology buildings. Spatial analysis within experimental buildings has highlighted the importance for archaeologists to devise sampling strategies for use in archaeological buildings with consideration to the possible layout of internal furniture, depressions in floors and in areas of structural modification, as these factors effect residue accumulation. Acknowledgements The authors would like to acknowledge the Arts and Humanities Research Council for funding Rowena Banerjea’s doctoral research, Lejre Historical and Archaeological Research Centre for a small research grant, the School of Human and Environmental Sciences, University of Reading, for funding the ‘Life-Histories of Buildings and Site Formation Processes’ research project, which formed part of this research, and the National Museum of Wales for funding the excavation and sampling of the Moel-y-Gaer roundhouse at St. Fagans. In addition, the authors would like to thank the staff at Butser, Lejre and St. Fagans, and all fieldwork team members for their assistance and contributions. Particular thanks go to Christine Shaw (Butser), Marianne Rasmussen (Lejre), Ken Brassil, Dr Adam Gwilt and Ian Daniels (St. Fagans), Nina Helt-Nielsen (University of Aarhus) and the following at the University of Reading: Professor Michael Fulford, Dr Rob Hosfield, Professor Stephen Nortcliff, Dr Jennifer Foster, Amy Poole and Christopher Speed. References Banerjea, R Y (2011) Characterising Urban Space: a case study from Insula IX, Silchester, Hants., UK. Unpublished PhD thesis, University of Reading Bell M (1983) Valley sediments as evidence of prehistoric land-use on the South Downs. Proc Prehist Soc 49:119–150 Bell M (2009) Experimental archaeology: changing science agendas and perceptual perspectives. In: Allen MJ, Sharples N, O’Connor T (eds) Land and People: papers in memory of John G Evans. Oxbow, Oxford, pp 31–45 Berge B (2000) The ecology of building materials. Architectural Press, Oxford Boardman S, Jones G (1990) Experiments on the effects of charring cereal components. J Archaeol Sci 17:1–11 Boivin NL (2000) Life rhythms and floor sequences: excavating time in Neolithic Catahoyuk. World Archaeol 31(3):367–388 Brady NC, Weil RR (2002) the nature and properties of soils. Prentice Hall, Upper Saddle River Brammer H (1971) Coatings in seasonally flooded soils. Geoderma 6:5–16 Breuning-Madsen H, Holst MK, Rasmussen M (2001) The chemical environment of a burial mound shortly after construction—an archaeological-pedological experiment. J Archaeol Sci 28:691–697 Breuning-Madsen H, Holst MK, Rasmussen M, Elberling B (2003) Preservation within long coffins before and after barrow construction. J Archaeol Sci 30:343–350 Bullock P, Fedoroff N, Jongerius A, Stoops G, Tursina T (1985) Handbook for thin section description. Waine Reasearch, Wolverhampton Canti MG (1999) The production and preservation of faecal spherulites: animals, environment and taphonomy. J Archaeol Sci 26:251–258 Canti MG (2003) Earthworm activity and archaeological stratigraphy: a review of products and processes. J Archaeol Sci 30:135–148 Author's personal copy Archaeol Anthropol Sci (2015) 7:89–112 Canti MG (2007) Deposition and taphonomy of earthworm granules in relation to their interpretive potential in Quaternary stratigraphy. J Quat Sci 22(2):111–118 Canti MG, Carter S, Davidson D, Limbrey S (2006) Problems of unscientific method and approach in ‘Archaeological soil and pollen analysis of experimental floor deposits; with special reference to Butser Ancient Farm. J Archaeol Sci 33:295–298 Carver, MJ (1987) The nature of urban deposits. In: Schofield J, Leech R (eds) Urban archaeology in Britain. Council for British Archaeology research report 61 Courty MA, Goldberg P, Macphail R (1989) Soils and micromorphology in archaeology. Cambridge University Press, Cambridge Crowther J, Macphail RI, Cruise G (1996) Short-term, post-burial change in a humic rendzina soil, Overton Down Experimental Earthwork, Wiltshire, England. Geoarchaeology 11(2):95–117 Entwistle JA, Abrahams PW, Dogshon RA (2000) The geochemical significance and spatial variability of a range of physical and chemical soil properties from a former habitation site, Isle of Skye. J Archaeol Sci 27:287–303 Fitzpatrick EA (1993) Soil microscopy and micromorphology. Wiley, Chichester French C (2003) Geoarchaeology in action. Routledge, New York Gé T, Courty MA, Matthews W, Wattez J (1993) Sedimentary formation processes of occupation deposits. In: Goldberg P, Nash DT, Petraglia MD (eds) Formation processes in archaeological context volume 17. Prehistory Press, Madison Goldberg P, Macphail RI, Matthews M (2006) Practical and theoretical geoarchaeology. Blackwell publishing, Malden Greig J (1982) The interpretation of the pollen spectra from urban archaeological deposits. In: Hall AR, Kenward HK (eds) Environmental archaeology in the urban context, vol 43, Council for British Archaeology research report., pp 47–65 Harvey EL, Fuller DQ (2005) Investigating crop processing using phytolith analysis: an example of rice and millets. J Archaeol Sci 32:739–752 Houben H, Guillaud H (1994) Earth construction: a comprehensive guide. Bourton-on-Dunsmore, ITDG Jones R, Challands A, French C, Card N, Downes J, Richards C (2010) Exploring the location and function of a Late Neolithic House at Crossiecrown, Orkney by geophysical, geochemical and soil micromorphological methods. Archaeol Prospect 17: 29–47 Kabata-Pendias, A (2001) Trace elements in soils and plants. CRC Press LLC, London Keefe L (2005) Earth building: methods and materials, repair and construction. Taylor Francis Group, London Kent, S (1984) Analysing activity areas. University of New Mexico Press Kramer, C (1982) Village ethnoarchaeology: rural Iran in archaeological perspective. Academic Press, New York La Motta, VM & Schiffer, MB (1999) Formation processes of house floor assemblages. In: Allison PM (ed) The archaeology of household activities. Routledge, London Macphail RI (1981) Soil and botanical studies of the dark earth. In: Jones M, Dimbleby GW (eds) The environment of man: the Iron Age to the Saxon period, vol 87, British Series. BAR, Oxford, pp 309–331 Macphail RI, Courty MA (1984) Interpretation and significance of urban deposits. Iskos 5:71–83 Macphail RI, Cruise G (2001) The soil micromorphologist as team player: a multi-analytical approach to the study of European microstratigraphy. In: Goldberg P, Holliday VT, Reid Ferring C (eds.) Earth sciences and archaeology. Kluwer Academic, New York Macphail RI, Cruise G, Allen MJ, Linderholm J (2006) A rebuttal of views expressed in ‘Problems of unscientific method and approach in Archaeological soil and pollen analysis of experimental floor deposits; with special reference to Butser Ancient Farm, Hants, UK. J Archaeol Sci 33:299–305 111 Macphail RI, Cruise G, Allen MJ, Linderholm J, Reynolds P (2004) Archaeological soil and pollen analysis of experimental floor deposits; with special reference to Butser Ancient Farm, Hants, UK. J Archaeol Sci 31:175–191 Macphail, R I, Cruise, G, Englemark, R & Linderholm, J (2000) Integrating soil micromorphology and rapid chemical survey methods: new developments in reconstructing past rural settlement and landscape organisation. In: Roskams S (ed) Interpreting stratigraphy: site evaluation, recording procedures and stratigraphic analysis. Archaeopress, Oxford, pp. 71–80 MacphailL, RI (1994) The reworking of urban stratigraphy by human and natural processes, in AR Hall & HK Kenward (eds.) Urban-Rural Connexions: Perspectives from Environmental Archaeology: Oxford: Oxbow, pp. 13–43 Macphail RI, Galinie H, Verhaeghe F (2003) A future for dark earth. Antiquity 349–358 Macphail RI, Linderholm J (2004) Neolithic landuse in south-east England: a brief review of the soil evidence. In: Cotton J, Field D (eds) Towards a New Stone Age, Research Report 137. CBA, York, pp 29–37 Mackenzie WS, Adams AE (1994) A colour atlas of rocks and minerals in thin section. Wiley, New York Mackenzie WS, Guilford C (1980) Atlas of rock forming minerals. Longman Group Ltd., Harlow Matthews W (1995) Micromorphological characteristics of occupation deposits and microstratigraphic sequences at Abu Salabikh, Southern Iraq. In: Barnham AJ, Macphail RI (eds) Archaeological sediments and soils: analysis, interpretation and management. Institute of Archaeology, University College, London, pp 41–76 Matthews, W (2000) Micromorphology of occupation sequences. In: Matthews R, Postgate JN (eds.) Contextual analysis of use of space at two Bronze Age sites. http://ads.ahds.ac.uk/catalogue Matthews, W (2005) Micromorphological and microstratigraphic traces of uses and concepts of space. In: Hodder I (ed) Inhabiting Çatalhöyük: reports from the 1995-99 seasons/by members of the Çatalhöyük teams.McDonald Institute for Archaeological Research, Cambridge; British Institute of Archaeology at Ankara, London Matthews W (2010) Geoarchaeology and taphonomy of plant remains and microarchaeological residues in early urban environments in the Ancient Near East. Quat Int 214:98–113 Matthews W, Postgate JN (1994) The imprint of living in a Mesopotamian city: questions and answers. In: Luff RM, RowleyConwy P (eds) Whither Environmental Archaeology? Oxbow Books, Oxford. Oxbow Monograph Series 3:171–212 Matthews W, French CAI, Lawrence T, Cutler DF, Jones MK (1997) Microstratigraphic traces of site formation processes and human activities. World Archaeol 29(2):281–308 Matthews W, Hastorf CA, Begums E (2000) Ethnoarchaeology: studies in local villages aimed at understanding aspects of the Neolithic site. In: Hodder I (ed) Towards reflexive method in archaeology: the example of Çatalhöyük. McDonald Institute for Archaeological Research and British Institute of Archaeology at Ankara, Cambridge Metcalfe D, Heath KM (1990) Microrefuse and site structure: the hearths and floors of the Hearthbreak Hotel. Am Antiq 55:781–813 Milek K (2005) Soil micromorphology. In: Sharples N (ed) A Norse Farmstead in the Outer Hebrides: excavations at mound 3, Bornais, South Uist. Oxbow, Oxford, pp 98–104 Milek KB (2012) Floor formation processes and the interpretation of site activity areas: an ethnoarchaeological study of turf buildings at Thverá, northeast Iceland. J Anthropol Archaeol 31:119–137 Milek KB, French CAI (2007) Soils and sediments in the settlement and harbour at Kaupang. In: Skre D (ed) Kaupang In Skiringssal. Aarhus University Press & the Kaupang Excavation Project, University of Oslo, pp: 321–361 Norton J (1997) Building with Earth: a handbook. ITDG, India Nicholas D, Kramer C (2001) Ethnoarchaeology in action. Cambridge University Press, New York Author's personal copy 112 Prentice JE (1990) The geology of construction materials. Chapman and Hall, London Rasmussen, M (2007) Building houses and building theories: archaeological experiments and house reconstruction. In: Rasmussen M (ed) Iron Age houses in flames: testing house reconstructions at Lejre. Studies in Ancient Technologies, volume 3. Historical-Archaeological Experimental Centre Robb J (2010) Beyond agency World Archaeol 42:493–520 Robinson, M (2006) The macroscopic plant remains. In: Fulford MG, Clarke A, Eckardt H. (eds) Life and labour in Late Roman Silchester: excavations In Insula IX Since 1997. Britannia Monograph Series 22, Society for the Promotion of Roman Studies Rosen AM (1993) Microartifacts as a reflection of cultural factors in site formation. In: Goldberg P, Nash DT, Petraglia MD (eds) Formation processes in archaeological context volume 17. Prehistory Press, Madison Schiffer, M B (1987) Formation processes of the archaeological record. University of Utah Press, Salt Lake Shahack-Gross R (2011) Herbivorous livestock dung: formation, taphonomy, methods for identification, and archaeological significance. J Archaeol Sci 38:205–218 Shahack-Gross R, Albert RM, Gilboa A, Nagar-Hillman O, Sharon I, Weiner S (2005) Geoarchaeology in an urban context: the uses of space in a Phoenician monumental building at Tel Dor (Israel). J Archaeol Sci 32:1417–1431 Shillito L-M, Almond MJ (2010) Comment on: fruit and seed biomineralization and its effect on preservation by E. Messager et al. Archaeol Anthropol Sci 2:25–34 Sillar B, Tite M (2000) The challenge of 'technological choices' for material science approaches in archaeology. Archaeometry 42(1): 2–20 Archaeol Anthropol Sci (2015) 7:89–112 Simpson I (1998) Early land management at Tofts Ness, Sanday, Orkney: the evidence of thin section micromorphology. In: Mills CM, Coles G (eds) Life on the edge: human settlement and marginality. Oxbow, Oxford Simpson IA, Guttmann EB, Cluett J, Shepherd A (2006) Characterizing anthropic sediments in North European Neolithic Settlements: an assessment from Skara Brase, Orkney. Geoarchaeol Int J 21(3):221– 235 Stevens CJ (2003) An investigation of agricultural and production models for Prehistoric and Roman Britain. Environ Archaeol 8:61–76 Stoops G (2003) Guidelines for analysis and description of soil thin sections. Soil Science Society of America, Madison Sveinbjarnardóttir G, Erlendsson E, Vickers K, McGovern T, Milek K, Edwards K, Simpson IA, Cook G (2007) Reykholt: the palaeoecology of a high status Icelandic farm. Environ Archaeol 12(2):187–206 Tsartsidou G, Lev-Yadun S, Albert R-M, Miller-Rosen A, Efstratiou N, Weiner S (2007) The phytolith archaeological record: strengths and weaknesses evaluated on a quantitative modern reference collection from Greece. J Archaeol Sci 34:1262–1275 Tsartsidou G, Lev-Yadun S, Efstratiou N, Weiner S (2008) Ethnoarchaeological study of phytolith assemblages from an agropastoral village in Northern Greece (Sarakini): development and application of a Phytolith Difference Index. J Archaeol Sci 35:600–613 Van der Veen M (2007) Formation processes of desiccated and carbonized plant remains-the identification of routine practice. J Archaeol Sci 34:968–990 Villagran XS, Balbo AL, Madella M, Vila A, Estevez J (2011) Experimental micromorphology in Tierra del Fuego (Argentina): building a reference collection for the study of shell middens in cold climates. J Archaeol Sci 38(3):588–604