image/svg+xml
199
XI/2/2020
INTERDISCIPLINARIA ARCHAEOLOGICA
NATURAL SCIENCES IN ARCHAEOLOGY
homepage: http://www.iansa.eu
Thematic review
Advances in Archaeological Soil Chemistry in Central Europe
Roderick B. Salisbury
a,b*
a
Department of Prehistoric and Historical Archaeology, University of Vienna, Franz-Klein-Gasse 1, 1190 Vienna, Austria
b
OREA Institute for Oriental and European Archaeology, Austrian Academy of Sciences, Hollandstrasse 11+13, 1020 Vienna, Austria
1. Introduction
Soils and sediments are archaeological materials,
archaeological resources that contain an archive of past
events and cultural behaviours. For nearly one hundred
years, archaeologists and soil scientists have accessed parts
of this archive through applications in archaeological soil
chemistry. Central Europe has been home to several key
developments in archaeological soil chemistry, from the
earliest applications of soil phosphate analysis to recent
practical application of portable X-ray Fluorescence
spectroscopy. Established methods, including phosphate
analyses and multi-element ICP-MS/OES, have provided
interpretations of the use of space within settlements and
houses, and the function of specifc archaeological features.
As archaeology continues to evolve as a discipline, micro-
remains are becoming increasingly important sources of
data about past human mobility, diversity, production,
consumption, and human ecodynamics. These data come
from a range of sources, including both human and animal
remains, primarily bones and teeth (proteomics, Charlton
et al.
, 2019; strontium isotopes, Giblin
et al.
, 2013), and
portable artefacts (
e.g.
starch grains on grinding stones and
ceramics, Duke
et al.
, 2018; ceramic residue analysis Dunne,
et al.
, 2019; residues on lithics, Rots
et al.
, 2015).
Chemistry of soils and sediments have been recognized,
and utilized, for a longer time. In addition to soil phosphate
analysis, soil chemical methods for archaeology traditionally
include multi-element chemistry, soil pH, magnetic
susceptibility, and soil organic carbon and nitrogen. Recently,
research in the Central European Palaeolithic have combined
these, for example, assessing carbon, nitrogen, and magnetic
susceptibility at the Pod Hradem cave to aid in interpreting
soil formation and potential climatic changes (Nejman
et al.
2018; 2020). Nevertheless, full exploitation of the soil
archive remains sporadic.
Cultural soilscapes (Salisbury, 2016; Wells, 2006)
hold ancient human and environmental DNA (Slon
et al.
,
2017; Willerslev
et al.
, 2003), biomarkers (Kovaleva and
Kovalev, 2015; Zocatelli
et al.
, 2017), traces of pollution
(Martínez Cortizas
et al.
, 2016; Veron
et al.
, 2014), evidence
of ecological changes and the environmental impact of
Volume XI ● Issue 2/2020 ● Pages 199–211
*Corresponding author. E-mail: roderick.salisbury@univie.ac.at
ARTICLE INFO
Article history:
Received: 9
th
September 2020
Accepted: 27
th
November 2020
DOI: http://dx.doi.org/10.24916/iansa.2020.2.5
Key words:
archaeological soil chemistry
archaeological prospection
settlement patterns
activity areas
Central Europe
ABSTRACT
Analytical technologies for the evaluation of archaeological soils have developed rapidly in recent
decades, and now support a range of innovative research and interpretations of archaeological sites
and landscapes. Established methods, including phosphates and multi-element ICP-MS/OES, have
provided interpretations of the use of space within settlements and houses, and the function of specifc
archaeological features. Recently, portable X-Ray Fluorescence has been introduced to archaeological
soil science, but published results have generated knowledge gaps. The correspondence between
archaeological geochemical anomalies and specifc human activities is partly dependent on geology
(including sediment type and relative acidity and permeability of the soil), topography, and formation
processes, as well as infuence of human activities. At the same time, which elements, and fractions
of elements, are measured is largely dependent on instrument parameters and extraction methods.
This paper provides an overview of archaeological soil chemistry in Central Europe, and the current
state-of-the-art, followed by an assessment of future developments in archaeological soil chemistry,
molecular biogeochemistry, and the signifcance of geoarchaeology in multi-disciplinary research.
image/svg+xml
IANSA 2020 ● XI/2 ● 199–211
Roderick B. Salisbury: Advances in Archaeological Soil Chemistry in Central Europe
200
cultural behaviour (Schumacher
et al.
, 2016; Sprafke, 2016),
archaeological site formation processes (Nicoll and Murphy,
2014), raw material provenance for components of ceramic
matrices (Riebe and Niziolek, 2015), and the contamination
of cultural heritage and artefacts by modern pollution (Nord
et al.
, 2005). Continued development of archaeological
biogeochemistry is beginning to recognize the archaeological
potential of proteomics, genomics, and other biomarkers in
the soil archive.
Several reviews of archaeological sol chemistry have been
published (Oonk
et al.
, 2009a; Wilson
et al.
, 2008); of especial
import was the detailed summary of soil phosphate methods
in archaeology by Holliday and Gartner (2007). A vast body
of literature exists for archaeological soil chemistry (Table 1).
This review of the history and development of archaeological
soil chemistry focuses on examples from Central Europe,
and the current state-of-the-art, followed by an assessment
of future developments in molecular biogeochemistry and
the signifcance of geoarchaeology in multi-disciplinary
research. Currently, trends are changing from the traditional
roles of archaeological soil chemistry – site prospection
and delimiting habitation areas – into a complex, multi-
disciplinary endeavour integrating various strands of
Table 1.
Selected references for archaeological soil chemistry, with focus on Central Europe (arranged chronologically).
Type of studyReferences
Soil chemistry & soil scienceFeigl, 1960; Bowen, 1979; Füleky, 1983; Kabata-Pendias and Pendias, 1984; Sposito, 1998;
Sparks 1996; 2003; Holliday 2004.
Early archaeological investigationsArrhenius, 1929; 1931; Lorch, 1930; 1940; 1941; 1951; Schnell, 1932; Christensen, 1935; Bandi,
1945; Stoye, 1950; Lutz, 1951; Dauncy, 1952; Pelikán, 1954; Dietz, 1957.
Methdological reviews and historiesSjoberg, 1976; Eidt, 1977; Keeley, 1981; Bethell and Máté, 1989; Zölitz and Heinrich, 1990;
Walker, 1992; Kondratiuk and Banaszuk, 1993; Bjelajac
et al.
, 1996; Aston
et al.
, 1998; Klamm
et al.
, 1998; Leonardi
et al.
, 1999; Haslam and Tibbett, 2004; Wells 2004a; Holliday and Gartner,
2007; Wilson
et al.
, 2008; Oonk
et al.
, 2009a; Pastor
et al.
, 2016.
Applications in Central Europe
PhosphateGundlach, 1961; Schwarz, 1967; Grimm, 1971; Sjöberg, 1976; Kiefmann, 1979; Zölitz, 1980;
1982; 1983; 1986; Gebhardt, 1982; Majer, 1984; Zimmermann, 1995; 2001; 2008; Stäuble and
Lüning, 1999; Majer, 2004; Sarris
et al.
, 2004; Ernée, 2005; Ernée and Majer, 2009; Hlavica
et al.
, 2011; Schreg and Behrendt, 2011; Petřík
et al.
, 2012; Salisbury, 2012; Lauer
et al.
, 2013;
Salisbury
et al.
, 2013; Weihrauch
et al.
, 2017; Weihrauch and Söder, 2020; Weihrauch
et al.
, 2020.
Multi-element
Hejcman
et al.
, 2011; 2013a; 2013b; Gauss
et al.
, 2013; Salisbury, 2013; Pető
et al.
, 2015; Lubos
et al.
, 2016; Salisbury, 2016; Dreibrodt
et al.
, 2017; Smejda
et al.
, 2017; 2018; Horák
et al.
, 2018;
Janovský and Horák, 2018; Pankowská
et al.
, 2018; Pető
et al.
, 2019; Dreslerová
et al.
, 2020.
Future directions
Lipid biomarkers
Bull
et al.
, 2000; Bull
et al.
, 2002; Schwark
et al.
, 2002; Killops and Killops, 2005; Zech
et al.
,
2010; Schatz
et al.
, 2011; Sistiaga
et al.
, 2014; Prost
et al.
, 2017; Zocatelli
et al.
, 2017; Harrault
et al.
, 2019; Schirrmacher
et al.
, 2019; Patalano
et al.
, 2020; Portillo
et al.
, 2020.
Isotopes
Bogaard
et al.
, 2007; D’Anjou
et al.
, 2012; Abell
et al.
, 2019; Bataille
et al.
, 2020; Snoeck
et al.
,
2020.
Sediment aDNA
Hebsgaard
et al.
, 2009; Giguet-Covex
et al.
, 2014; Madeja, 2015; Thomsen and Willerslev, 2015;
Parducci
et al.
, 2017; Slon
et al.
, 2017; Brunson and Reich, 2019; Epp
et al.
, 2019; Nejman
et al.
,
2020.
Proteomics
Oonk
et al.
, 2012.
geoarchaeology, bioarchaeology, and environmental studies.
This trajectory needs geoarchaeologists to contribute to
diachronic and synchronic examinations of archaeological
landscapes and human-ecodynamics.
2. Archaeological soil phosphates in Central Europe
The history of archaeological soil chemistry extends over
a century (Arrhenius, 1931) and across the globe, from
Alaska (Knudson
et al.
, 2004) to the Levant (Šmejda
et al.
, 2018) and to Australia (Fanning
et al.
, 2018). The
link between ancient human occupation and increased soil
phosphate content were frst noted in late 19
th
or early 20
th
century agricultural soil surveys (Russell, 1915). Publication
of methods and results began when Swedish agronomist
Olaf Arrhenius recognized and recorded the relationship
between surface fnds, human occupation, and soil phosphate
levels while conducting agricultural soil testing, and posited
a causal correlation between enhanced soil phosphates and
human settlements (Arrhenius, 1929; 1931). The method was
applied to Swedish habitation sites by Schnell (1932), who
sampled along transects radiating outward from an assumed
image/svg+xml
IANSA 2020 ● XI/2 ● 199–211
Roderick B. Salisbury: Advances in Archaeological Soil Chemistry in Central Europe
201
central point in a given site and constructed isopleth maps of
the phosphate values to delimit site boundaries. Following
this, archaeologists in Central Europe and elsewhere have
conducted soil phosphate analyses for archaeological survey
and site interpretations for nearly a century, with continuing
methodological developments.
Early archaeological soil chemistry focused primarily on
phosphates for four reasons: First, it was the frst element
explicitly connected to prehistoric human habitation,
and subsequent studies exploited this discovery. Second,
it is essential to life and therefore can be found in and
deposited by anything organic, as primary, secondary,
or
in situ
refuse. Inputs include food preparation wastes
such as bone, meat, fsh, and plants, wood ash, human or
animal burials, and urine and faeces of humans and animals.
Therefore, human habitation areas will have higher levels
of soil P than surrounding areas with no habitation. Conway
(1983) demonstrated that human occupation could result in
annual increases in concentrations of P from 1–10%. Third,
phosphates tend to accumulate quickly, have low solubility
and a strong ability to fx within the soil profle. In favourable
soil conditions, P remains stable and is likely to be retained,
even in disturbed soils, for millennia without appreciable
leaching. Fourth, in its most basic form, phosphate testing
is fast and inexpensive compared with almost every other
analytical technique, and can be conducted in the feld during
survey or excavation.
2.1 The frst 60 years of archaeological soil phosphate
analysis
Arrhenius’s methods were quickly applied to prehistoric and
medieval sites in Central Europe by German geographer
Walter Lorch (1940; 1941; 1951). Lorch sampled along
regular transects across and around ancient settlements,
used a laboratory colorimetric method to measure phosphate
content, and graphed the results. By comparing graphs
of density and distribution, and interpreting the diferent
profles, Lorch persuasively argued that variability in
phosphate was due to diferent subsistence economies in
the Palaeolithic, Neolithic, and metal ages. Other research
projects soon followed. Bandi (1945) used Lorch’s method
to locate a medieval site in Switzerland. Dietz (1957) used
a method of sulphuric acid in test tubes visually compared
to standards to examine a small plot of land with surface
material indicating prehistoric activity, looking for evidence
of organic waste deposits.
The next major advance in archaeological soil phosphate
studies came with the development of a spot-test, or ring-
chromatography test, by Friedrich “Fritz” Feigl in Vienna.
In his comprehensive two-volume compendium on chemical
spot-tests, Feigl (1960) recommended the highly toxic
and corrosive nitric acid (HNO
3
) to prepare an ammonium
molybdate solution, and the toxic and fammable benzidine
as a reducing agent to measure inorganic phosphate from
geological samples. Gundlach (1961) modifed Feigl’s
method to increase the speed and safety when testing soil from
boreholes at prehistoric sites. Gundlach retained the nitric
acid for digestion, but switched to the organic and relatively
harmless ascorbic acid (C
6
H
8
O
6
) for reduction, making it
more stable and eliminating the need for fame and ammonia
in feld conditions. In any acid-molybdate method, PO
4
reacts
with molybdate to form phosphomolybdic acid (yellowish
in colour); in Gundlach’s method, phosphomolybdic acid is
reduced by ascorbic acid to form a blue complex. Gundlach
conducted his tests in the feld, with the flter paper and drop
bottles attached to a pole stuck in the ground next to his
borehole (depicted in Gundlach, 1961 p.736, Figure 1), and
noted that the entire process takes approximately 90 seconds.
Gundlach ranked the results on a scale of 1–5, from none to
very high levels of P.
Schwarz (1967) used the Gundlach method and established
a feld methodology for sample collection, collecting samples
in plastic bags, collecting enough soil to run multiple tests
and to determine colour and grain size, and collecting
samples by layer. Schwarz conducted large-scale surveys
near the San Bernardino Pass in southern Switzerland, taking
samples at 30– 50 m intervals. Two suggestions regarding
research methodology are of note from the Schwarz’s paper.
First, he observed that some information about land-use,
previous archaeological investigations, and geomorphology
are essential to planning a chemical survey campaign.
Second, he suggested that phosphate surveys would be
more useful when complemented by geophysical surveys
or test trenching, and warned that carrion pits, where local
villagers dispose of diseased animals, should not be mistaken
for prehistoric habitations (Schwarz, 1967, pp.58–61).
Schwarz’s sampling approach has been criticized (Sjöberg,
1976, p.449) because of his suggestion that samples be taken
along natural lines, such as feld boundaries, rather than on
a regular grid or transect system.
Kiefmann (1979) conducted large-scale phosphate mapping
in East Holstein and arrived at two signifcant conclusions.
First, soil phosphate content is infuenced by both changing
land-use and pedogenesis. That is, archaeological soil
chemistry requires some understanding of both cultural and
non-cultural formation processes, including soil formation.
Second, Kiefmann found that diferent extraction methods
yielded diferent P concentrations and distributions.
Zölitz (1980) pioneered the use of variogram modelling
as a statistical method for analysing soil phosphate results.
Majer (1984) developed a 3-point relative scale for measuring
results, and noted that archaeologists were most interested
in the anthropogenic enhancement of soil phosphate, rather
than the total quantity of P. Phosphate surveys successfully
identifed residential areas in both prehistoric and medieval
sites (Grimm, 1971; Zölitz, 1982; 1983; 1986).
Eidt refned Gundlach’s method further, frst by replacing
nitric acid with hydrochloric acid (HCl) (Eidt, 1973) and
then by developing a bath to stop the chemical reaction,
so that results could be archived (Eidt, 1977). Eidt found
that HCl is superior to both nitric acid and sulphuric acid
when extracting P in laboratory tests, and observed that the
use of HCl for both feld and lab analyses would improve
comparability of results. He also noted that the question of
image/svg+xml
IANSA 2020 ● XI/2 ● 199–211
Roderick B. Salisbury: Advances in Archaeological Soil Chemistry in Central Europe
202