Residues on stone artefacts: state of a scientific art.

Fullagar, Richard ; Furby, Judith ; Hardy, Bruce 等

At the 1996 Society for American Archaeology meeting in fabulous New Orleans, residues and functional analysis of stone artefacts were the specific focus of offerings scattered through the programme in posters (e.g.D. Lee, T. Murphy, L. Hooper and R. Donahue), workshop (N. Tuross, M. Wachowiak, R. Evershed, C. Kolman), at least two symposia (B. Hardy, B. Kimura & B. Hardy, T. Loy, D. Hyland, Charters et. al., Cummings et al., Tuross et al., C. Heron, H. Ceri and M. Newman) and several general sessions (E. Lohse, J. Furby & R. Fullagar, B. Williamson, M. Newman). We could not see them all, and we know some presenters listed in the programme did not show. A symposium (sponsored by the Conservation Analytical Laboratory at the Smithsonian Institution) and a workshop (sponsored by the Conservation Analytical Laboratory at the Smithsonian Institution and the National Center for Preservation and Training, National Park Service, Natchitoches, Louisiana) were dedicated to residue analysis. Both, organized by Noreen Tuross, were a success with lively debate. In dealing with much broader issues, the presentations seemed to portray conflicting views on the veracity of scientific techniques, and on the achievements of integrated functional analysis of stone artefacts. While papers presented in the general sessions assumed the validity of the methodologies being used, some in the sponsored workshop and symposium cast doubt on:

* the distinctiveness of microscopic residues;

* the integrity/reactivity of ancient bio-molecules;

* the viability of the current methodologies;

* and consequently on the reliability of taxonomic identifications on ancient stone artefacts.

Archaeologists could be forgiven for being both confused whether residues survive on stone tools, or whose technique is best. As archaeologists with one foot in the lab and one in the field, we were both stimulated and depressed by the workshop and sponsored symposium. We offer here our post-conference views on residues, on functional analysis of stone tools and on where consensus seems to lie.

Stone artefact function

Stone artefact function has been investigated by diverse lines of evidence in many theoretical contexts for over a century. Many articles summarize these notions and review lines of evidence, which include tool design, raw material characteristics, use-wear, residues, context, ethnography and the study of long-term trends (e.g. Hayden 1979; Kamminga 1982; van Gijn 1990; Yerkes & Kardulias 1994). In truth, artefact function rarely, if ever, is determined from a single line of evidence, although many claim use-wear or residues provide sufficient evidence of tool function. The rise and rise of residue studies can be correlated with the increased availability of specialized techniques in biochemistry for analysing low concentration samples. We think time-consuming techniques of observing and recording visible residues have been neglected, a casualty of the high-tech methods now available for targeting the particular species of animal or plant contact material.

Detailed microscopic observations and recording are essential pre-requisites to any residue analysis. In contrast to the impressions offered at the workshop, microscopically visible organic structures are common under optimal preservation conditions. Many residue analysts do 'whole tool' extractions in the process of analysis, a procedure which is often inappropriate for archaeological material as it is both destructive and does not target particular residues with their visible structures and associations with utilized edges.

Survival, detection and taxonomic identification of animal related residues

Blood, hair, bone and cartilage are the residues most often preserved on the surfaces of stone artefacts used in butchering animals. These and other traces have been documented by micro-wear analysts and in specific residue studies (Briuer 1976; Anderson 1980; Keeley 1980; Kamminga 1982; Loy 1983; Fullagar 1986; van Gijn 1991; Furby 1995). For example, blood forms characteristic films which are often visible at low magnification and clearly identifiable at high magnifications (Brown 1988; Furby 1995). In contrast, the sponsored workshop at the 1996 SAA sent the message that 'we can't separate residues from other stuff with any certainty' (Tuross & Wachowiak 1996).

Handling of the artefacts during excavation was another point raised at the sponsored symposium and workshop. There was a clear dichotomy between the recommended handling and storage methods of those examining pottery and those studying blood residues. It was recommended (by Tuross) that excavators wear plastic gloves during the recovery of artefacts, which may be a reasonable precaution if submitting material for DNA analysis. However, it is a complicating procedure for the level of analysis that most people are pursuing. Recent excavations at Cuddie Springs in central northern New South Wales were undertaken without the benefit of plastic gloves, yet the subsequent analysis of the stone artefact assemblage was successful, both in the identification and characterization of blood and in DNA analysis of the material (Furby 1995; Loy et al. submitted). The excavator's hands are covered in the enclosing sediments and the artefacts themselves are usually cloaked in sediment. With at least two layers of sediment between the excavator and the stone there is, in our experience, little or no detectable contamination.

Most interest in identifying residues associated with butchering animals has been in blood residue analysis, particularly in the determination of species of origin (Loy 1983; Newman & Julig 1989; Tuross & Dillehay 1995; Petraglia et al. 1996). It is difficult to reconcile the range of studies currently being reported which assume blood to be present in the absence of careful microscopic observations. The issue at stake appears to be whether residues can be characterized to determine species of origin. As blood residues are preserved on the surface of stone only under a limited number of depositional conditions and environments, it is imperative that the range of residues on the stone artefacts are documented before the artefact is tested. In our experience, where blood does survive, it is usually found in step fractures, at the edge of undetached flakes or in other protected areas, and in quantities significantly less than must have been deposited during use.

And what are the real costs of microscopy? While the sponsored workshop implied that tens of thousands of dollars are required for the visual identification, this is simply not true. Archaeologists can easily examine their assemblage and sample it for artefacts with visible signs of use and organic residues. While compound metallographic and biological microscopes can cost in excess of $10,000, they are commonly available in many university departments; more importantly, a hand lens or simple stereo-microscope can be used for examination of stone artefacts before they are submitted for analysis. Description and recording of the residues present can then be undertaken by a residue specialist.

A range of methods is available to characterize purported residues as blood (Newman & Julig 1989; Cattaneo et al. 1990; Hyland et al. 1990; Tuross & Dillehay 1995; Loy & Hardy 1992; Loy & Wood 1989). Constructing a hierarchy of evidence enables different lines of evidence to be addressed, as in the forensic sciences, to provide a confident level of interpretation. Loy & Hardy (1992) presented a range of techniques for identifying blood residues including the use of microscopy, Ames Hemastix and immunoassay using the 'dot-blot' method. At the SAA sessions, Loy indicated how false positives for Ames Hemastix can effectively be eliminated by the addition of 0.5M, pH 8 di-sodium EDTA which complexes metal ions in solution. These three steps have been demonstrated to provide a viable and reproducible method for characterizing blood residues (Furby et al. submitted).

An impression current at the residue workshop and symposium is that it will be at least another five years before blood residue analysis will be able to produce any meaningful results. Yet it is widely accepted that blood residues can be identified using relatively simple assays in series coupled with microscopic observation. On the other hand, species of origin determinations are problematic, under the methods described by Tuross, Newman and other workers; in many cases they may provide no additional useful information beyond that obtained from simple assays for Haem or Immunoglobulins. Haemoglobin crystallization, a simple method proposed by Loy (1983), has been replicated in the Australian Museum laboratory. The results are equivocal and this does not appear to be a technique that can yet be exploited by archaeologists in identifying species of origin (Garling 1994).

It is important to document the environmental conditions from which any stone assemblage is recovered the soil pH, particle size analysis, environmental setting etc the information that will eventually allow us to predict the most likely circumstances under which residues will survive. Loy (1990) has suggested that greater than 20% clay must be present for organic residues to be preserved, a figure related to the presence of certain clay minerals, e.g. montmorillonite.

As argued by many conference presenters, blood residue analysis needs to be critically evaluated by archaeologists themselves and questions asked as to what information they really need. In many cases, it would be sufficient to identify and characterize blood residues without the additional complication of species of origin analyses. Species of origin may not be critical to the research question, a point also made by Tuross et al. (1996). There was tacit criticism in New Orleans that a 'residue industry' is emerging prematurely, driven by commercial interests, with many analyses obtained just because the methodology is available. Presentations in New Orleans implied that some studies are not part of an integrated approach to the analysis of stone tool function. If these sentiments are true, study of residues is not to be split between laboratory technicians and archaeologists, but belongs in the context of a collaborative and integrated approach to the site as a whole.

Previously, Fiedel (1996) critically reviewed a range of blood residue analyses, arguing that results were frequently irreconcilable with contextual archaeological data. The same issues were raised in New Orleans. We emphasize three points. First, the criticisms are largely directed to the problem of cross-reactivity with CIEP (cross-over immuno-electrophoresis), and the lack of consistency between CIEP and other techniques of taxonomic identification. Newman's defence of CIEP at the conference cannot deny problems of cross-reactivity, though the ability to control for this and other variables does not appear to be unattainable. Second, as Fiedel points out, there is a need for further blind tests to assess more accurately the reliability, specificity and sensitivity of different techniques. Ceri, Newman and their colleagues seem keen to participate, as they have co-operated previously in print. Third, there seems little doubt about the survival of blood residues on utilized tool edges, based on presence of microscopically visible cracked plaques, erythrocytes, mammalian hair and consistent biochemical, immunological and DNA tests as demonstrated by Loy for the Toad River Canyon site in British Columbia (see references in Fiedel and presentations at the New Orleans conference). The lowest common denominator appears to be that blood residues do survive. Taxonomic identifications are limited by reference collections, cross-reactivity, protein concentrations and sensitivity of the technique used. Which technique is most reliable has yet to be ascertained, though DNA tests appear most promising (see below).

Plant residues

Plant residues were hardly mentioned in the workshop, but fared well in the symposium (with an excellent paper by Carl Heron) and in the general sessions where papers dealt with resins, starch grains, phytoliths and other microscopically visible structures. Plant processing has also been associated with highly visible polishes on stone tool surfaces (e.g. silica gloss), with some attempts to determine taxonomic identifications in conjunction with residues (Anderson 1980; Fullagar 1993; Loy et al. 1992). Given the presence of visible structures of likely taxonomic significance, it seems easier for archaeologists to accept the survival of plant residues on stone artefacts and their direct association with utilized tool edges. Nevertheless, in order to interpret residues like starch grains and phytoliths, it is essential to have both extensive reference collections of different parts of many plant species and also extractions from local sediments.

In addition to visible structures, much research has been undertaken to identify plants through biochemical and other means. At the symposium, Heron reported on long-term research to identify lipids using gas chromatography and mass spectrometry.

DNA: a final solution?

The recognition of morphologically identifiable residues on stone tools has naturally led to the application of DNA analysis to residue studies. The technique which has made this possible is the Polymerase Chain Reaction (PCR). From a minute amount of target DNA (theoretically as little as one molecule), PCR can create millions of copies, enough for analysis through cloning or sequencing. Determination of the base-pair sequence of the DNA can allow phylogenetic analysis and, in some cases, determination of the species of origin. Sequence analysis of DNA is a potentially more accurate way to determine species of origin of residues than the immunological techniques previously used. However, because PCR is capable of amplifying minute amounts of DNA, contamination by modern DNA is a significant hazard. In order to ensure that putative ancient DNAs are authentic, a number of informal precautions have been proposed including: extractions carried out in labs where the DNA of interest has not previously been analysed, no DNA control PCR reactions, dedicated pipettes, aerosol resistant or positive displacement pipettes, periodic UV irradiation of solutions and pipettes, etc. Strict adherence to these controls greatly reduces the possibility of contaminants. Because ancient DNA is highly degraded, fragments are usually less than 500 base pairs in length (e.g. Paabo 1989; Lindahl 1993). This general rule has been used to test the authenticity of putative ancient DNA fragments recovered, as those larger than 500 base pairs are suspect as modern contaminants. Ancient DNA studies in archaeological contexts have most often been applied to bone (e.g. Parr et al. 1996; Stone et al. 1996; Hoss et al. 1996) but DNA has also been recovered from preserved soft tissue (e.g. Hoss et al. 1996; Handt et al. 1994; Paabo 1989). As researchers improve their precautions against contamination, the results of ancient DNA studies have become more believable and valuable.

Residues on stone tools have recently been analysed for ancient DNA with the goal of attempting accurately to determine the species of origin of a residue (Hardy et al. in press; Loy, this symposium). Hardy et al. have recovered DNA from Middle Palaeolithic stone tools in excess of 35,000 years old from the site of La Quina, France, and have identified some to species. In the New Orleans symposium, Loy reported the recovery of DNA from stone tools of similar age and used sequence analysis to identify the DNA as diprotodont (an extinct Australian marsupial) in origin. Despite objections raised at the New Orleans symposium, use residues can clearly survive on stone tools and may retain recoverable DNA. DNA analysis can potentially provide species identification and phylogenetic information about the origin of the DNA. Furthermore, DNA analysis of residues is not limited to residues of animal origin, but can be expanded to include investigations of plant remains on stone tools as well. DNA analysis, not necessarily a final solution in residue investigations, is capable of providing precise identification of the species with which a stone tool was associated - when performed with the proper controls and precautions. By combining DNA analysis with other techniques, including use-wear and microscopic residue analysis, archaeologists can construct a much more complete picture of prehistoric tool function than is otherwise available.

Conclusion

Noreen Tuross is thanked for organizing the successful sponsored sessions, and we agree with her call not to accept residue analyses without scepticism, particularly the taxonomic identifications using CIEP. But we express surprise at how residues have been analysed with such a low level of basic microscopic observation and such little regard for other traces of use. In our experience, structurally distinct traces of plant and animal tissue, visible microscopically, undoubtedly survive over tens and probably hundreds of thousands years. Much can be done, at little expense, with careful observations.

The consensus at the workshop and symposium is that residues do survive on the surfaces of stone artefacts for considerable time periods. There is debate over the reliability and viability of various techniques for characterizing these residues. Few of the protagonists meet head-on in disputing the best techniques, and samples and artefacts are rarely treated in parallel. Blind testing is still a problem due to these variations in treatments; no one technique can yet be ruled out (see Fiedel 1996). In our experimental collections produced over 10 years ago, residues are still present, reactive and with structural material still clearly visible. A key issue, not discussed in the SAA sessions, is the process of recording and characterizing the visible residues.

A back-to-basics approach may be useful for archaeologists wishing to sample their collections for more sophisticated analyses. The debate over survival, reactivity and the most viable techniques has not yet been resolved. On the other hand, and in contrast to the impressions from the SAA Workshop, we believe that residue analysis is within the realm of the archaeologist, at least in the initial identification and observation stage. A positive point made at the SAA conference is that residue analysis needs to be planned in the context of meaningful and appropriate archaeological research questions which are achievable given the resources available. Archaeologists can continue, as they have for over a hundred years, examining artefacts and pursuing the different lines of evidence; an integrated approach to the identification of tool function. There is no magic bullet to determine artefact function, and no substitute for careful microscopic observations.

References

ANDERSON, P. 1980. A testimony of prehistoric tasks: diagnostic residues on stone tool working edges, World Archaeology 12: 181-94.

BRIUER, F.L. 1976. New clues to stone tool function: plant and animal residues, American Antiquity 41: 478-84.

BROWN, P.D. 1988. Residue analysis of stone artefacts from Yambon, West New Britain. Unpublished BA (Hons.) thesis, Department of Prehistoric & Historic Archaeology, University of Sydney.

CATTANEO, C., K. GELSTHORPE, P. PHILLIPS, R.J. SOKOL & D. SMILLIE. 1991. Identification of ancient blood and tissue - ELISA and DNA analysis, Antiquity 65: 878-81.

FULLAGAR, R.L.K. 1986. Use-wear and residues on stone tools. Functional analysis and its application to two southeastern Australian archaeological assemblages. Unpublished Ph.D thesis, Department of Archaeology, La Trobe University.

1993. Flaked stone tools and plant food production: a preliminary report on obsidian tools from West New Britain, Papua New Guinea, in P. Anderson et al. (ed.), Traces et fonction: les gestes retrouves: 331-7. Liege: University of Liege. ERAUL 50.

FURBY, J.H. 1995. Megafauna under the microscope: palaeo-environment and archaeology at Cuddie Springs. Ph.D thesis, School of Geography, University of New South Wales.

FURBY, J., T.H. LOY & R. FULLAGAR. Submitted. Reproducibility and viability of blood residue analyses; a case study from Cuddie Springs, Ancient Biomolecules.

GARLING, S. 1994. What was on the menu at Cuddie Springs? Hemoglobin crystallisation of blood residues on stone tools, Unpublished BA (Hons.) thesis, Department of Pre-historic and Historic Archaeology, University of Sydney.

GIJN, A.L. VAN 1990. The wear and tear of flint: principles of functional analysis applied to Dutch Neolithic assemblages. Leiden: University of Leiden. Analecta Prae-historica Leidensia 22.

HANDT, O. et al. 1994. Molecular genetic analyses of the Tyrolean ice man, Science 264: 1775-8.

HARDY, B.L., V. RAMAN & R.A. RAFF. In press. Recovery of DNA from Middle Palaeolithic stone tools, Journal of Archaeological Science.

HOSS, M., P. JARUGA, T.H. ZASTAWNY, M. DIXDAEOGLU & S. PAABO. 1996. DNA damage and DNA sequence retrieval from ancient tissues, Nucleic Acids Research 24(7): 1304-7.

HYLAND, D.C., J.M. TERSAK, J.M. ADOVASIO & M.I. SIEGEL. 1990. Identification of the species of origin of residual blood on lithic material, American Antiquity 55: 104-12.

KAMMINGA, J. 1982. Over the edge: functional analysis of Australian stone tools. Canberra: Australian National University. Occasional Papers in Prehistory.

KEELEY, L.H. 1980 Experimental determination of stone tool uses: a microwear analysis. Chicago (IL): University of Chicago Press.

LINDAHL, T. 1993. Instability and decay of the primary structure of DNA, Nature 262: 709-15.

LOY, T.H. 1983. Prehistoric blood residues: detection on tool surfaces and identification of species of origin, Science 220: 1269-71.

1990. Prehistoric organic residues: recent advances in identification, dating and their antiquity, in W. Wagner & H. Pernicka (ed.), Archaeometry '90: Proceedings of the 27th International Symposium on Archaeometry, Heidelberg: 645-56. Basel: BirkHauser-Verlag.

LOY, T.H. & B.G. HARDY. 1992. Blood residue analysis of 90,000-year-old-stone tools from Tabun Cave, Israel, Antiquity 60: 24-35.

LOY, T.H., K.I. MATTHEI, S. FOUNTAIN, D. BUCKLE, S. MCELROY & J. FURBY. Submitted. Species identification from 30,000-year-old Australian archaeological blood residues and 100,000-year-old marsupial bone using genomic 28s ribosamal DNA, Science (submitted July 1966).

LOY, T.H. & A.R. WOOD. 1989. Blood residue analysis at Cayonu Tepesi, Turkey, Journal of Field Archaeology 16: 451-60.

NEWMAN, M. & P. JULIG. 1989. The identification of protein residues on lithic artifacts from a stratified boreal forest site, Canadian Journal of Archaeology 13: 119-32.

PAABO, S. 1989. Ancient DNA: extraction characterisation, molecular cloning, and enzymatic amplification, Proceedings of the National Academy of Sciences of the USA 66:1939-43.

PARR, R.L., S.W. CARLYLE & D. O'ROURKE. 1996. Ancient DNA analysis of Fremont Amerindians of the Great Salt Lake Wetlands, American Journal of Physical Anthropology 99(4): 507-18.

PETRAGLIA, M., D. KNEPPER, P. GLUMAC, M. NEWMAN & C. SUSSMAN. 1996. Immunological and microwear analysis of chipped stone artifacts from piedmont contexts, American Antiquity 61: 127-35.

STONE, A.C., G. MILNER & S. PAABO. 1996. Sex determination of ancient human skeletons using DNA, American Journal of Physical Anthropology 99(2): 231-8.

TUROSS, N. & T. DILLEHAY. 1995. Mechanism of organic preservation at Monte Verde and one use of biomolecules in archaeological interpretation, Journal of Field Archaeology 22: 97-110.

TUROSS, N. & M. WACHOWIAK. 1996. Residue analysis: How much you can do it for and how much it will cost. Workshop at the Annual Meeting of the Society of American Archaeologist, New Orleans, 9-14 April 1996.

TUROSS, N., I. BARNES & R. POTTS. 1996. Protein identification of blood residues on experimental stone tools, Journal of Archaeological Science 23: 289-96.

YERKES, R. & P. KARDULIAS. 1994. Recent developments in the analysis of stone artefacts, Journal of Archaeological Research 1: 88-119.