THE ENVIRONMENTAL SAMPLES

Alexandra Schmidl, Örni Akeret and John Carrott

Summary

Most of the biological remains recovered from the ten sediment samples were plant remains well preserved by waterlogging, with a few also well preserved by charring. The concentrations of organic finds varied, with rather few being recorded from six of the deposits but the four others yielding interpretatively valuable assemblages of plant remains and three of these also giving useful numbers of invertebrate macrofossils. There were very few records for cultivated plant taxa – only occasional cereal remains. Overall, the larger plant and invertebrate assemblages recovered suggested that these deposits were principally composed of waste from the housing of domestic animals, together with subsequent colonizers of the cleared material. No identifiable microfossil remains other than of pollen grains/spores were detected.

Waterlogged silver birch timbers in pit 1247, Trench 12

METHODS

Seven sediment subsamples were processed by the excavators prior to delivery to PRS, and the unsorted ‘flots’ (or washovers) for analysis. The weights and volumes of the subsamples were recorded before being placed onto 500 micron nylon mesh in a sieving tank. The light organic fraction was washed over into a 500 micron sieve to collect the washover fractions.

In addition, three sediment subsamples were processed by PRS. The lithologies of these deposits were recorded using a standard pro forma and subsamples taken and processed, broadly following the techniques of Kenward et al. (1980; 1986), for the recovery of plant and invertebrate macrofossils. Plant and invertebrate remains in the processed subsample fractions (residues and flots) were recorded briefly by ‘scanning’ using a low-power microscope (x7 to x45), identifiable taxa and other biological and artefactual components being listed on paper.

Plant remains were identified by comparison with modern reference material at PRS and the use of published works (Cappers et al. 2006, Jacomet 2006). Identifiable taxa and other components being listed on paper. Nomenclature for plant taxa follows Stace (1997).

The flots were also examined for insect and other invertebrate remains using a low-power microscope (to x45). Unfortunately, the work required for full analysis of the invertebrate assemblages was far in excess of the financial resources available to the project and, consequently, these remains were only rather superficially recorded. Nomenclature for insect follows Kloet and Hincks (1964-77).

The three samples of sediment submitted were also examined for the eggs of intestinal parasitic nematodes using the ‘squash’ technique of Dainton (1992). Assessment slides were scanned at 150x magnification with 600x used where necessary. Although primarily for the detection of intestinal parasitic nematode eggs, the ‘squash’ technique routinely reveals other microfossil remains, and where present these have been noted.

RESULTS

The results are presented in context number order by trench. Archaeological information, provided by the excavator, is given in square brackets. A brief summary of the processing method follows in round brackets after the sample numbers. No unprocessed sediment remains from these samples. The plant remains recovered are listed in Table 1.

Click to view Table 1

TRENCH 1

Context 142 [possible 1st century AD buried turf]Sample 4/T (10 litres sieved to 500 microns with washover; processed by the excavator)

The small washover (2 g, dried) consisted mostly of modern rootlets and some charcoal (to 3 mm), with a few decayed wood fragments, unidentifiable plant fibres and sand, but there was also a very small number of waterlogged seeds and fruits including common nettle, orache/goosefoot and silver birch.

TRENCH 9

Context 908 [waterlogged fill of pit 909, 2nd century AD]Sample 2/T (10 litres sieved to 500 microns with washover; processed by the excavator)

There was a small washover (12 g, dried) mostly of ‘woody’ debris including decayed wood, bark and twig fragments, with some rootlets and charcoal (to 5 mm), and a few unidentifiable plant fibres and invertebrate remains. The last were restricted to occasional poorly preserved pieces of unidentified beetle sclerite (eroded and rather fragmented), however.

The deposit produced a moderate amount  of decayed seeds and fruits of plants representing various habitats, but dominated by taxa growing on heath/moor and grassland (especially damp), such as gypsywort, lesser/marsh stitchwort, lesser spearwort, marsh pennywort, sedge, silver birch, spike-rush and tormentil. There were also leaves of bogmoss which probably arrived with imported peat. Other botanical remains derived from waste/rough ground and included black nightshade, common nettle, hemlock, henbane, knotgrass, orache/goosefoot and white/red dead-nettle. In addition, a few remains of plant species growing in hedges (e.g. blackberry/raspberry, elder, hazel) were recorded and there was also a single charred grain of barley.

TRENCH 12

Context 1215 [fill of ditch 1224]Sample 16/T (20 litres sieved to 500 microns with washover; processed by the excavator)

The tiny washover (1 g, dried) contained sand, culm fragments (to 20 mm) and rootlets, with a little charcoal (to 2 mm), cinder (to 3 mm) and unidentifiable plant fibres. A rather small range of waterlogged seeds and fruits were recovered from this deposit. The most abundant plant taxon in this assemblage was common nettle, but there were also traces of other wild plant species such as blackberry/raspberry, elder, orache/goosefoot and selfheal. In addition, two charred caryopses of brome were found. 

Context 1220 [buried soil beneath track 1206]Sample 15/T (20 litres sieved to 500 microns with washover; processed by the excavator)

There was a small washover (2 g, dried) mostly of rootlets, with a few leaf fragments, and a little sand, charcoal (to 5 mm) and slag/cinder (to 3 mm). Identifiable botanical remains were restricted to a small number of waterlogged seeds and fruits, including chickweed, common nettle, dock, elder, gypsywort, orache/goosefoot and white/red dead-nettle, representing habitats such as waste ground and hedges. There were also two charred caryopses of brome in this assemblage.

Context 1222 [fill of ditch 1221]Sample 7/T (20 litres sieved to 500 microns with washover; processed by the excavator)

The small washover (4 g, dried) consisted mainly of slightly silted unidentified charcoal (to 15 mm), with some rootlets and a very little sand, but there were also a few waterlogged nuts and catkin-scales of silver birch and seeds of orache/goosefoot. Other identifiable remains were restricted to two charred caryopses of brome.

Context 1257 [‘peaty’ basal fill of circular pit 1247 within 2nd century Roman industrial building]Sample 11/T (2 kg/4.5 litres sieved to 300 microns with paraffin flotation; processed by PRS)

Moist, dark grey (externally) to light to mid orange-brown (internally), brittle, layered and compressed, slightly sandy slightly clay, fine and coarse herbaceous detritus and amorphous organic sediment. There were some areas where clay and silt formed a greater, but still minor, part of the deposit. Stones (2 to 6 mm) were present and ‘straw’ was abundant in the sample.

The fairly large wet residue (1.4 litres) and small flot (~25 ml) both consisted mostly of ‘straw-like’ material and small unidentifiable plant fibres, but there were also wood fragments, twiglets, bud scales, cereal pericarps, ‘stems’ and ‘leaves’ of mosses (Bryophyta), small undisaggregated lumps of possible herbivore dung, some charcoal and a few rootlets. Twigs and leaves of heather were the most commonly identified plant remains from this deposit. In addition, there were moderate numbers of very well preserved waterlogged seeds and fruits. However, the identifiable component of these remains was restricted to a rather small range of taxa: seeds and fruits of bent, sedge, clover, cornflower, dock, hemp-nettle, selfheal and wild basil. There were also a few charred utricles of sedge present. 

There were a few variably preserved invertebrate remains in the flot. Some beetle sclerites (e.g. of Cercyon analis (Paykull) and ?Monotoma picipes Herbst) were very well preserved and there were also some fragile remains present (e.g. wing fragments), whereas other macrofossils were heavily eroded (though fragmentation was generally quite low). Other remains present included sclerites of staphylinid beetles, a small weevil (probably Ceutorhynchus sp.), Helophorus ?flavipes (Fabricius), some fly puparia and a few mites (Acarina).

The microfossil ‘squash’ was mostly fragments of plant tissue, with some pollen grains/spores and a very small inorganic content. No eggs of intestinal parasitic nematodes were seen.

Context 1265 [burnt layer surrounding possible 2nd century hearth within 2nd to 3rd century industrial building]Sample 6/T (10 litres sieved to 500 microns with washover; processed by the excavator)

The small washover (3 g, dried) was mostly of rootlets and slightly silted charcoal (to 10 mm), with some sand, a single earthworm egg capsule and a few waterlogged nuts and catkin-scales of silver/downy birch.

Context 1278 [‘peaty’ fill of pit 1247 within late 2nd to 3rd century Roman industrial building]Sample 9/T (2 kg/3.8 litres sieved to 300 microns with paraffin flotation; processed by PRS)

Moist, mid to dark grey-brown to very dark grey (with some areas which were mid to dark orange-brown internally), brittle, layered and compressed, slightly sandy silty clay fine and coarse herbaceous detritus and amorphous organic sediment. There was a minor component of light to mid grey clay silt. Stones (2 to 20 mm), fly puparia, wood, ‘straw’ and rootlets were present in the sample.

There was a large wet residue (1.8 litres) and a small flot (~15 ml) both largely composed of tiny pieces of wood, ‘straw-like’ fragments and other unidentifiable plant fibres, with some undisaggregated lumps of possible herbivore dung. Again, a large number of the twigs and leaves were of heather, together with ‘leaves’ of mosses (Bryophyta) and bracken. Fruits and seeds were well preserved by waterlogging and included representatives of wild plant taxa such as brome, buttercup, chickweed, cornflower, foxtail, grass family, hemp-nettle, meadow-grass, mouse-ear, orache, ribwort plantain, rush, sedge, selfheal and tormentil. There were also a few remains of cultivated plant species, in the form of glume bases of spelt wheat, from this deposit.

The flot contained a substantial assemblage (forming perhaps 30% to 50% of the total volume) of highly variably preserved invertebrate remains. Many of the remains were very well preserved but still more were reduced to ‘filmy’ scraps of cuticle. There were numerous beetle remains, including those of Cercyon analisCryptopleurum ?minutum (Fabricius), various staphylinids (including Omalium rivulare (Paykull)), ?Monotoma picipesAnobium punctatum (Degeer) (the woodworm beetle), some Carabidae and weevils (Curculioindae, probably including Ceutorhynchus sp.). There were also many fly puparia and mites (Acari), some sculpted and other ants (Formicidae) and occasional remains of relatively fragile invertebrate body parts in the form of wing fragments.

The microfossil ‘squash’ was mostly fragments of plant tissue, with very many pollen grains/spores and a small inorganic content. No eggs of intestinal parasitic nematodes were seen.

Context 1280 [lens of black fibrous organic material in pit 1247 within late 2nd to 3rd century Roman building]Sample 12/T (1 kg/1.4 litres sieved to 300 microns with paraffin flotation; processed by PRS)

Wet, very dark grey-brown to black, unconsolidated to slightly sticky (working more or less soft), slightly sandy slightly clay silt. Stones (2 to 20 mm) and twigs were present and fine and coarse herbaceous detritus was abundant in the sample.

The fairly large wet residue (600 ml) and small flot (~10 ml) were, again, largely composed of plant remains. More than half of the plant material was charred or partially charred. This deposit consisted mostly of wood fragments, twiglets, bark, charcoal, ‘straw-like’ fragments, fibres, rootlets, bud scales, mosses and small lumps of possible herbivore dung. Identifiable charred botanical remains included brome, common knapweed, dock, hulled barley, rush (perianth), sedge and selfheal. In addition, waterlogged seeds and fruits of bedstraw, clover, corncockle, gypsywort, hawkbit and meadow-grass were noted.

The flot gave relatively few invertebrate remains most of which were of fly puparia. There were also some mites, ants and a few beetle remains (including Cercyon analis, staphylinids, ?Monotoma picipes and a weevil head). Preservation of the remains was, again, highly variable, with some being very well preserved and others no more than heavily eroded (‘filmy’) scraps – some of the remains appeared to be charred.

The microfossil ‘squash’ was mostly of charred (approximately two-thirds) and uncharred (most of the remaining third) organic detritus, with a small inorganic content. No eggs of intestinal parasitic nematodes were seen.

Context 1287 [fill of possible pre-Roman pit 1273]Sample 13/T (15 litres sieved to 500 microns with washover; processed by the excavator)

The tiny washover (~1 g, dried) was mostly modern rootlets and some sand, with a few leaf fragments and a little charcoal (to 3 mm). Identifiable botanical macrofossil remains were restricted to a few waterlogged seeds and fruits of common nettle, silver birch and small nettle.

DISCUSSION

Most of the plant remains recovered from the ten sediment samples were well preserved by waterlogging and a small amount of them by charring. The concentrations of finds varied between the deposits, being relatively low in Context 142 (Trench 1) and Contexts 1215, 1220, 1222, 1265 and 1287 (all from Trench 12), whereas the other four (Context 908 from Trench 9 and Contexts 1257, 1278 and 1280 from Trench 12) produced abundant seeds and fruits. The same pattern was seen for the diversities of the plant taxa represented (see Table 1).

The processed subsamples from Contexts 908, 1257, 1278 and 1280 gave interpretatively valuable assemblages of well preserved seeds and fruits and very variably preserved invertebrate remains. In Contexts 908, 1257 and 1278, all or almost all of the plant material was waterlogged, whereas in Context 1280 there was also a considerable proportion of charred remains. All of the plant remains recovered from the four contexts – whether preserved by waterlogging or charring – were in an excellent state of preservation and represented taxa of various habitats (see Table 1).

The plant assemblages were dominated by remains of wild taxa indicative of different habitats: heath and moor, wet or damp places (e.g. wet grassland, ponds, marsh, banksides), grassland, disturbed and rough ground, scrub and hedgerow (the abundant remains suggesting nearby, perhaps adjacent, hedges) and, to a lesser degree, areas of disturbed waste ground.

Overall, rather few finds of cultivated plants were found at this site – Contexts 908, 1222 and 1280 each gave a small number of charred hulled barley grains. The records of spelt wheat from Context 1278 accord well with the date of the building (late 2nd to 3rd century) in which the pit was situated, as this was an important cereal for the Roman period (Greig 1991, Van der Veen and O’Connor 1998). These food plant remains were presumably charred by accident during some stage of crop processing (e.g. parching, drying) or cooking, but were too few to be of any real interpretative value.

In Contexts 1257 and 1278 remains of heather (Calluna vulgaris (L.) Hull) were conspicuously frequent. This species is characteristic of heath land and may have been imported with turves, perhaps used for roofing or to serve as bedding (for animals or humans). It may conceivably also have been used as a dye plant (Kenward and Hall 1995) but, in the present archaeological context, the most likely utilisation would seem to be as bedding for animals.

Lumps of compacted and layered short plant fragments which are likely to be herbivore dung were found in Contexts 1257, 1278 and 1280. This, together with fragmentary cereal pericarp, grassland species and crop weeds, which might have been gathered with hay or eaten by animals grazing in pastures or meadows suggests that ‘stable manure’ (sensu Kenward and Hall 1997) was probably a major component of these sediments. Such ‘hay’-rich deposits have also been recorded from other Roman sites and especially from military establishments (see Hall and Huntley 2007).

In addition, a few Sphagnum ‘leaves’ were found in Context 908 (fill of pit 909) from Trench 9, together with other wetland species such as marsh pennywort, sedge, spike-rush and tormentil, which suggest an origin in imported peat, presumably brought to the site for use as fuel or, perhaps, in construction (see Hall 2003).

Invertebrate remains were present in the three samples for which paraffin flotation was employed (Contexts 1257, 1278 and 1280) and abundant in the flot from Context 1278 (Sample 9). Each of the assemblages exhibited highly variable preservation, but included some very well preserved remains. Remains of the beetle Cercyon analis were recovered from all three deposits. This species lives in decaying organic matter of various kinds (including dung and compost) and supports the suggestion that the lumps of compacted and layered short plant fragments seen were of herbivore dung. The presence of Omalium rivulare and Cryptopleurum ?minutum in Context 1278 lend further support, again indicating the presence of decaying organic material and specifically dung, respectively. The presence of a component of animal dung could also be one explanation for the presence of fly puparia in all three of these samples. Small numbers of poorly preserved and unidentified beetle sclerites were also seen in the washover from Context 908 but these were of no interpretative value.

Although some of the processed subsamples from Trench 12 were from contexts located within an industrial building, only traces of material indicative of industrial activity (i.e. cinder/slag) were found from just two deposits (Contexts 1215 and 1220). Overall, the larger plant and invertebrate assemblages recovered suggested that these deposits were principally composed of waste from the housing of domestic animals (e.g. soiled bedding material, traces of cereals – perhaps from fodder), together with subsequent colonizers of the cleared material.

The only identifiable microfossils recorded from the ‘squash’ subsamples were pollen grains/spores—numerous in Context 1278, present in Context 1257 and absent from Context 1280. In particular, no eggs of intestinal parasitic nematodes were seen.

REFERENCES

Cappers, R. T. J., Bekker, R. and Jans J. E. A. (2006). Digitale Zadenatlas van Nederland. Gronigen Archaeological Studies 4. Gronigen: Barkhuis Publishing & Gronigen University Library.

Dainton, M. (1992). A quick, semi quantitative method for recording nematode gut parasite eggs from archaeological deposits. Circaea, the Journal of the Association for Environmental Archaeology 9, 58 63.

Dobney, K., Hall, A. R., Kenward, H. K. and Milles, A. (1992). A working classification of sample types for environmental archaeology. Circaea, the Journal of the Association for Environmental Archaeology 9 (for 1991), 24-6.

Greig, J. R. A. (1991). The British Isles, pp. 299-334 in Van Zeist, W., Wasylikowa, K. and Behre, K.-E. (eds), Progress in Old World palaeoethnobotany. Rotterdam/Brookfield: Balkema.

Hall, A. (2003). Recognition and characterisation of turves in archaeological occupation deposits by means of macrofossil plant remains. Centre for Archaeology Report 16/2003. English Heritage.

Hall, A. R. and Huntley J. P. (2007). A Review of the Evidence for Macrofossil Plant Remains from Archaeological Deposits in Northern England. Research Department Report Series 87/2007. English Heritage.

Jacomet, S. (2006). Identification of cereal remains from archaeological sites – 2nd edition. published by the IPAS, Basel University.

Kenward, H. K., Hall, A. R. and Jones, A. K. G. (1980). A tested set of techniques for the extraction of plant and animal macrofossils from waterlogged archaeological deposits. Science and Archaeology 22, 3-15.

Kenward, H. K., Engleman, C., Robertson, A. and Large, F. (1986). Rapid scanning of urban archaeological deposits for insect remains. Circaea 3, 163–172.

Kenward, H. K. and Hall, A. R. (1995). Biological evidence from Anglo Scandinavian deposits at 16 22 Coppergate. The Archaeology of York 14 (7), 435-797 + xxii + loose figures. York: Council for British Archaeology.

Kenward, H. K. and Hall, A. R. (1997). Enhancing bioarchaeological interpretation using indicator groups: stable manure as a paradigm. Journal of Archaeological Science 24, 663-673.

Kloet, G. S. and Hincks, W. D. (1964-77). A check list of British Insects (2nd edition). London: Royal Entomological Society.

Stace, C. (1997). New Flora of the British Isles: second edition. Cambridge: Cambridge University Press.