The Effect of Waste Disposal on Soils in and

Around Historic Small Towns

By Kirsty Ann Golding

Submitted to University of Stirling, March 2008

For the degree of Doctor of Philosophy

This research was undertaken at the School of Biological and Environmental Sciences,
University of Stirling, Stirling, UK

-1-
 Statement of Originality

I hereby confirm that this research was carried out by the undersigned alone and that all
research material has been duly referenced and cited.

Kirsty Ann Golding (28/03/08)

-2-
 Acknowledgements

This study was primarily funded by the Natural Environment Research Council (NERC)
with supplementary assistance from Scottish Natural Heritage (SNH).

I would like to thank Prof. Donald Davidson for having the chutzpah to take me on in the
first place and for his continued support throughout the course of this research. I would
also like to thank his ‘second in command’ Dr Clare Wilson who proved an invaluable
source of knowledge on all things anthropogenic in thin section. She also bought me
countless cocktails at conference dinners, for that I am appreciative. In addition
retrospective gratitude goes to Dr David Shimwell, formerly of Geography at Manchester
University, who paved the way for a future career in the realm of Geoarchaeology.

I would like to thank all past and present residents of Lauder, Pittenweem and Wigtown,
for depositing rubbish in the first place and allowing me to investigate it several hundred
years later. I am indebted to Dr Catherine Wearing and Andrew Collinge for help given
during field work especially in digging test pits, driving, dodging cattle and avoiding the
local crazies.

I would like to acknowledge the following people based within the School of Biological
and Environmental Sciences, University of Stirling; Prof Ian Simpson for advice during
early stages of research, Helen Ewen for assistance in the laboratories, George McLeod
for making thin sections, Scott Jackson and Bill Jamison for all things computer related,
and Kate Howie for guidance with statistical analyses.

Finally, thanks to the original proprietors of room 3a124a, you know who you are.

-3-
 Abstract

Soils in the urban environment are distinctive in that they are modified through waste
amendments. Consideration has been given to how urban soil properties reflect current
human influence; however, recent studies highlight their potential as historical archives.
The impact of waste disposal on the nature, properties and formation of urban soils is
significant, especially in historic small towns where the extent and complexity of refuse
management practices is only just emerging. This study uses a multi-method approach
to characterise and understand modes of urban anthrosol formation in three Scottish
burghs; Lauder, Pittenweem and Wigtown. The objectives of this study are threefold; to
establish the nature and diversity of urban anthrosols in and near to historic small towns,
to characterise and account for the multiplicity of urban anthrosols in and near to historic
small towns, and to elucidate the processes associated with waste management and
disposal in historic small towns.

Physical, chemical and micromorphological analysis of topsoil deposits indicate

sustained addition of past waste materials to soils within and near to historic small
towns. Soil characteristics were heterogeneous across burghs; however, distinct
patterns according to past functional zones were identified. The burgh core and burgh
acres are important areas of interest at all three burghs. Soil modification was most
pronounced within burgh cores resulting in the formation of hortic horizons. Soils within
burgh cores are characterised by neutral pH, increased organic matter content,
enhanced magnetic susceptibility and elevated elemental concentrations such as
calcium, phosphorus and potassium. In comparison the nature and extent of soil
modification within burgh acres is more varied. At Lauder hortic soils were identified in
the burgh acres suggesting pronounced soil modification through cultivation. Deepened
topsoil in the burgh acres at Pittenweem provided evidence for application of mineral rich
waste materials in the past. Moreover, magnetic and elemental enhancement (barium,
phosphorus, lead, zinc) within the burgh acres south of Wigtown revealed historic soils
based anthropogenic signal.

-4-
It is argued that changes in soil characteristics at Lauder, Pittenweem and Wigtown can
be explained through processes of waste management and disposal in the past.
Evidence from micromorphological analyses suggests that waste in burgh cores typically
comprised domestic waste, animal waste, building materials and fuel residues. These
materials were also identified within burgh acres, although it is noted that their
abundances were significantly lower. Variation in urban anthrosol characteristics
between burghs is attributed to differing industries and patterns of resource exploitation,
for example marine waste associated with fishing was only identified in coastal burghs.

The sustained addition of waste materials to soils within and near to historic small towns
was an effective waste management strategy. Waste disposal in burgh cores was likely
to be a combination of direct application and midden spreading in back gardens. This led
to enhanced soil fertility which was important in the development of urban horticulture;
particularly for poorer inhabitants who did not have access to arable farm land adjacent
to the burgh. Dunghills acted as temporary stores of waste in the main thoroughfares of
Lauder, Pittenweem and Wigtown. These dunghills were systematically transported to
the burgh acres for further use as a fertiliser; hence, an early form of urban composting.
Processes of waste disposal could not be deduced from soil characteristics alone;
however, likely methods include direct waste deposition, storage and redistribution of
midden waste, and storage and redistribution of dunghills.

The limitations of soil classification systems and mapping are highlighted, for example
urban soils are either omitted from soil maps or are misclassified. It is recommended that
urban soils in historic towns should be incorporated into future regional soil maps. Urban
soils represent a complex archive of past human behaviour not necessarily reflected in
archaeological excavation or documentary analysis. It is argued that soil and artefacts
are equally important, hence soil should be a consideration in urban heritage and
conservation strategies.

-5-
 Contents
Abstract……………………………………………………………………………. 5
Contents…………………………………………………………………………… 6
Figures…………………………………………………………………………….. 15
Tables……………………………………………………………………………… 29

Chapter 1 The importance of Heritage and Conservation in Historic Small

Towns………………………………………………………………… 33
1.1 Trends in Urban Conservation 33
1.1.1 Historic Small Towns 33
1.2 Thesis Structure 34

Chapter 2 The Historical Legacy of Urban Anthrosols with Specific Reference to

Scottish Royal Burghs……………………………………………. 35
2.1 Urban Anthrosols 35
2.1.1 Anthropogenic Soil Classification 35
2.1.2 Urban Soil Classification 36
2.1.3 Current Research into Urban Anthrosols 37
2.1.3.1 Heavy Metals 37
2.1.3.2 Organic Pollutants 38
2.1.4 The Historical Legacy of Urban Anthrosols 38
2.1.4.1 Occupational Sequences 38
2.1.4.2 Elemental Enhancement 39
2.1.4.3 Elemental Signatures 39
2.1.4.4 Soil Improvement 40
2.1.5 Summary 41
2.2 Waste Sources and Management in Royal Burghs 42
2.2.1 Medieval Origins of Urbanism 42
2.2.1.1 Morphology 42
2.2.2 Post-Medieval Urban Growth 44
2.2.3 Waste Sources 45
2.2.3.1 Human and Animal Excreta 45
2.2.3.2 Building Materials 46
2.2.3.3 Fuel 46
2.2.3.4 Industrial Wastes 46
2.2.4 Waste Management 47

-6-
 2.2.5 Summary 48
2.3 Evidence for Past Soil Modification in Royal Burghs 49
2.3.1 Deepened Soil Deposits 49
2.3.2 Backland Garden Soils 50
2.3.3 Stratigraphy and Dating 51
2.3.3.1 Relative Dating 51
2.3.3.2 Absolute Dating 52
2.3.4 Hinterland 53
2.3.5 Summary 54

Chapter 3 Investigating Urban Anthrosols in Historic Small Towns: A Research

Framework…………………………………………………………... 55
3.1 Archaeological Concepts of Rubbish 55
3.1.1 Defining Rubbish 55
3.1.2 Garbology 56
3.1.3 Classification of Rubbish Deposits 56
3.1.4 Refuse Cycle Flow Models 57
3.1.5 Summary 58
3.2 The Nature and Significance of Rural-Urban Interaction 60
3.2.1 Contemporary Research of Rural-Urban Interaction 60
3.2.1.1 Rural-Urban Fluxes 60
3.2.1.2 Rural-Urban Interaction Zones 61
3.2.1.3 Sectoral Interaction 61
3.2.2 Rural-Urban Interaction in Historic Small Towns 62
3.2.2.1. Hinterland Resource Flows 62
3.2.2.2 Evidence for Rural-Urban Zones and Sectoral
Interactions 62
3.2.3 Summary 63
3.3 Research Framework 64
3.3.1 Modelling Waste in Small Historic Towns 64
3.3.1.1 Town and Hinterland 64
3.3.1.2 Resource Flow 65
3.3.1.3 Waste Flow and Deposition 65
3.3.1.4 Model Limitations 66
3.3.2 Aim and Objectives 66
3.3.2.1 Aim 66
3.3.2.2 Objective 1 66

-7-
 3.3.2.3 Objective 2 67
3.3.2.4 Objective 3 67

Chapter 4 Materials and Methods……………………………………………. 69

4.1 Site Selection 69
4.1.1 Study Sites 70
4.1.1.1 Lauder 70
4.1.1.2 Pittenweem 71
4.1.1.3 Wigtown 73
4.2 Soil Survey 75
4.2.1 Survey Design 75
4.2.1.1 Stratified Grid 75
4.2.2 Field Methods 79
4.2.2.1 Auger Sampling 79
4.2.2.2 Soil Pit Sampling 82
4.3 Laboratory Analyses 84
4.3.1 pH 84
4.3.2 Loss on Ignition (% LOI) 85
4.3.3 Environmental Magnetism 86
4.3.3.1 Mass Dependant Magnetic Susceptibility 88
4.3.3.2 Frequency Dependant Magnetic Susceptibility 88
4.3.4 Elemental Concentrations 89
4.4 Micromorphological Analysis 91
4.4.1 Sample Preparation 91
4.4.2 Semi-Quantification of Anthropogenic Features 92
4.4.2.1 Revision of Coarse Organic Material Classification 99
4.5 Spatial Interpolation 100
4.6 Statistical Analyses 101
4.6.1 Zonal Delimitation 101
4.6.2 Determination of Significant Differences between Zones 103
4.6.2.1 Physical and Chemical Data 103
4.6.2.2 Elemental Data 104
4.6.2.3 Soil Micromorphology 104
4.6.3 Determination of Significant Differences between Depths 105
4.6.4 Determination of Significant Differences between Burghs 105
4.6.4.1 Physical and Chemical Data 105
4.6.4.2 Elemental Data 106

-8-
 4.6.4.3 Soil Micromorphology 106
4.6.5 Multivariate Analyses 107
4.6.5.1 Cluster Analysis 107
4.6.5.2 Discriminant Analysis 107

Chapter 5 Field Observations and Soil Micromorphology……………… 108

5.1 Soil Profile Characteristics 108
5.1.1 Variation within Burghs 108
5.1.2 Variation between Burghs 111
5.2 Coarse Mineral Material 112
5.2.1 Variation within Burghs 112
5.2.1.1 Zones 112
5.2.1.1 Hortic Horizons 114
5.2.2 Variation between Burghs 117
5.3 Summary of Results: Coarse Mineral Material 118
5.3.1 Variation within Burghs 118
5.3.1.1 Lauder 118
5.3.1.2 Pittenweem 118
5.3.1.3 Wigtown 118
5.3.1.4 Hortic Horizons 118
5.3.2 Variation between Burghs 120
5.4 Fine Mineral Material 121
5.4.1 Variation within Burghs 121
5.4.2 Variation between Burghs 121
5.5 Coarse Organic Material 122
5.5.1 Charcoal 122
5.5.1.1 Variation within Burghs 122
5.5.1.2 Variation between Burghs 122
5.5.2 Fuel Residue 1 (FR1) 123
5.5.2.1 Variation within Burghs 123
5.5.2.2 Variation between Burghs 124
5.5.3 Fuel Reside 2 (FR 2) 125
5.5.3.1 Variation within Burghs 125
5.5.3.2 Variation between Burghs 125
5.5.4 Fuel Residue 3 (FR 3) 126
5.5.4.1 Variation within Burghs 126
5.5.4.2 Variation between Burghs 127

-9-
 5.5.5 Fuel Residue 4 & 6 (FR 4 & 6) 128
5.5.5.1 Variation within Burghs 128
5.5.5.2 Variation between Burghs 129
5.5.6 Fuel Residue 5 & 7 (FR 5 & 7) 130
5.5.6.1 Variation within Burghs 130
5.5.6.2 Variation between Burghs 131
5.5.7 Fuel Residue 8 & 9 (FR 8 & 9) 131
5.5.7.1 Variation within Burghs 131
5.5.7.2 Variation between Burghs 132
5.5.8 Fuel Residue 10 (FR 10) 133
5.5.8.1 Variation within Burghs 133
5.5.8.2 Variation between Burghs 133
5.5.9 Fuel Residue 10-255µm (%) 134
5.5.9.1 Variation within Burghs 134
5.5.9.2 Variation between Burghs 136
5.5.10 Micromorphological Characteristics of Hortic Horizons 137
5.5.10.1 Variation within Burghs 137
5.5.10.2 Variation between Burghs 141
5.6 Summary of Results: Coarse Organic Material 142
5.6.1 Variation within Burghs 142
5.6.1.1 Lauder 142
5.6.1.2 Pittenweem 142
5.6.1.3 Wigtown 143
5.6.1.4 Hortic Horizons 143
5.6.2 Variation between Burghs 145
5.7 Fine Organic Material 147
5.7.1 Variation within Burghs 147
5.7.2 Variation between Burghs 148
5.8 Pedofeatures 149
5.8.1 Variation within Burghs 149
5.8.2 Variation between Burghs 150
5.9 Structure 151
5.9.1 Variation within Burghs 151
5.9.1 Variation between Burghs 152
5.10 Summary: Soil Micromorphology 153

- 10 -
Chapter 6 Soil Physical and Chemical Properties………………………… 154
6.1 Topsoil Depth 154
6.1.1 Variation within Burghs 155
6.1.2 Variation between Burghs 158
6.2 Soil pH 159
6.2.1 Variation within Burghs 161
6.2.1.1 Zone 161
6.2.1.2 Variation in pH with Depth 164
6.2.2 Variation between Burghs 167
6.3 Loss on Ignition 168
6.3.1 Variation within Burghs 170
6.3.1.1 Zone 170
6.3.1.2 Variation in LOI with Depth 174
6.3.2 Variation between Burghs 176
6.4 Magnetic Susceptibility 177
6.4.1 Mass Dependant Magnetic Susceptibility (Ҳ) 179
6.4.1.1 Variation within Burghs 179
6.4.1.2 Variation between Burghs 182
6.4.2 Frequency Dependant Magnetic Susceptibility (ҲFD) 183
6.4.2.1 Variation within Burghs 183
6.4.2.2 Variation between Burghs 186
6.4.3 Natural Variation of Magnetic Susceptibility 186
6.5 Summary of Results: Soil Physical and Chemical Properties 188
6.5.1 Variation within Burghs 188
6.5.1.1 Lauder 188
6.5.1.2 Pittenweem 188
6.5.1.3 Wigtown 188
6.5.1.4 Depth 188
6.5.2 Variation between Burghs 190

Chapter 7 Elemental Analyses……………………………………………….. 191

7.1 Elemental Concentrations 191
7.1.1 Barium 191
7.1.1.1 Variation within Burghs 192
7.1.1.2 Variation between Burghs 195
7.1.2 Calcium 195
7.1.2.1 Variation within Burghs 196

- 11 -
 7.1.2.2 Variation between Burghs 199
7.1.3 Lead 199
7.1.3.1 Variation within Burghs 200
7.1.3.2 Variation between Burghs 203
7.1.4 Potassium 203
7.1.4.1 Variation within Burghs 204
7.1.4.2 Variation between Burghs 207
7.1.5 Phosphorus 207
7.1.5.1 Variation within Burghs 208
7.1.5.2 Variation between Burghs 211
7.1.6 Strontium 212
7.1.6.1 Variation within Burghs 213
7.1.6.2 Variation between Burghs 216
7.1.7 Zinc 216
7.1.7.1 Variation within Burghs 217
7.1.7.2 Variation between Burghs 220
7.2 Summary of Results: Elemental Concentrations 221
7.2.1 Variation within Burghs 221
7.2.1.1 Lauder 221
7.2.1.2 Pittenweem 221
7.2.1.3 Wigtown 221
7.2.2 Variation between Burghs 223
7.3 Multiple Comparisons 224
7.3.1 Lauder 224
7.3.2 Pittenweem 228
7.3.3 Wigtown 231
7.4 Cluster Analysis 234
7.4.1 Lauder 234
7.4.2 Pittenweem 236
7.4.3 Wigtown 238
7.5 Discriminant Analysis 240
7.5.1 Lauder 240
7.5.2 Pittenweem 241
7.5.3 Wigtown 241
7.5.4 Discrimination between Burghs 242
7.5.4.1 High Street Zone 242
7.5.4.2 Hinterland Near Zone 243

- 12 -
 7.6 Summary of Results: Multi-Element Analyses 244
7.6.1 Variation within Burghs 244
7.6.1.1 Lauder 244
7.6.1.2 Pittenweem 245
7.6.4.3 Wigtown 245
7.6.2 Variation between Burghs 246

Chapter 8 The Impact of Waste Disposal on Soils in and Around Lauder,

Pittenweem and Wigtown………………………………………… 247
8.1 The Impact of Waste Disposal on Burgh Cores 247
8.1.1 Soil Modification 247
8.1.2 Elemental Signatures 248
8.1.3 Waste Materials 248
8.1.3.1 Harbour and High Street Zones 250
8.1.3.2 Topsoil Horizons 250
8.1.4 Fuel Residue 251
8.1.4.1 Harbour and High Street Zones 251
8.1.4.2 Topsoil Horizons 252
8.2 The Impact of Waste Disposal on Burgh Hinterlands 253
8.2.1 Soil Modification 255
8.2.1.1 Variation within Burgh Acres 256
8.2.2 Elemental Signatures 256
8.2.3 Waste Materials 257
8.2.4 Fuel Residue 258
8.3 Waste Disposal 259
8.3.1 Burgh Core 259
8.3.2 Hinterland 260

Chapter 9 The Wider Significance of Waste Disposal in and Around Historic

Small Towns………………………………………………………… 262
9.1 Study Findings 262
9.1.1 Objective 1 262
9.1.2 Objective 2 262
9.1.3 Objective 3 263
9.1.4 Model Revision 263
9.2 Rural-Urban Interaction 266
9.2.1 Modern Significance of Rural-Urban Interaction 266

- 13 -
 9.3 Soil Classification 268
9.3.1 Mapping Urban Soils 268
9.3.2 Classification of Soils in Historic Towns 270
9.4 Soil: An Archaeological Resource 273
9.4.1 Comparison of Soils and Archaeological Excavation 273
9.4.2 Comparison of Soils and Documentary Evidence 275
9.4.3 Conservation 275
9.5 Summary of Key Findings and Future Recommendations 276
9.5.1 Key Findings 276
9.5.2 Recommendations 277

Chapter 10 Bibliography………………………………………………………... 278

Appendices………………………………………………………………………… 293
Appendix 1 Soil Profile Descriptions 293
Appendix 2 Micromorphology Summary Tables 314
Appendix 3 Watching Brief: High Vennel, Wigtown 317
Appendix 4 Archaeological Evaluation of 64a High Street, Pittenweem 318

- 14 -
 Figures

Figure 1 Cross section of terrace P1 Pseira Island, Crete showing Minoan potsherd
scatters indicative of systematic spreading of household wastes across the
cultivated landscape as a fertiliser (Bullock et al., 2001: 227)
40

Figure 2 16th century AD map of the Royal Burgh of Dumfries, Dumfries and
Galloway shown as a single street of dwellings with burgage plots, taken
from the Pont Manuscripts (Image © National Library of Scotland, Licensor
www.scran.ac.uk)
43

Figure 3 Edinburgh c1460AD comprising Town (Royal Burgh, Burgh of Canongait),

School (Castle), and Cloister (Holyrood Abbey, St. Giles Cathedral) (Image
© Patrick Geddes Centre for Planning Studies, University of Edinburgh,
Licensor www.scran.ac.uk)
43

Figure 4 Medieval Street Patterns: Newburgh, Fifeshire 1855-56AD example of

single street system and burgage plots and Crail, Fifeshire 1855AD example
of parallel street system and burgage plots (Image produced from the
www.old-maps.co.uk service with permission of Landmark Information
Group Ltd. and Ordnance Survey)
44

Figure 5 Graphical Summaries of Potential Waste Sources in the Medieval/Post-

medieval Urban Environment (a) Reconstruction of market scene at St.
Johns Kirk, Perth (b) Reconstruction of Upperkirkgate, Aberdeen backlands
(c) Reconstruction of backland manufacturing in Meal Vennel, Perth (d)
Reconstruction of Upperkirkgate, Aberdeen backlands (Images © Aberdeen
City Council, Licensor www.scran.ac.uk)
45

Figure 6 Mid 17th century AD view of Canongate, City of Edinburgh. Note differences
in land use within and between burgage plots (Image © Royal Commission
on the Ancient and Historical Monuments of Scotland, Licensor
www.scran.ac.uk)
48

Figure 7 Crail from the air (a) Deepened anthropogenic sediments, west end of
Nethergate, (b) Shallower sediments, east end of Nethergate, (c) Deep soils
with medieval and mixed medieval and post-medieval assemblages,
attributed to medieval cultivation of the burghs infield and/or off loading
ballast by Dutch herring traders (Lowe CE, 2001, Image © Colin J M Martin,
Licensor www.scran.ac.uk)
50

Figure 8 Depth of topsoil in and near the Burgh of Nairn (left), distribution of rigs in
burgh acres 1790AD (right) (Davidson et al., 2006: 780)
53

- 15 -
Figure 9 ‘Rubbish Theory’ model showing cultural categories of objects and possible
transfers between them (Thompson M 2003: 322, 1979: 10)
55

Figure 10 Flow model showing artefact histories and proposed rubbish categories
(Schiffer 1972:162)
57

Figure 11 Flow model of adapted artefact life histories (LaMotta and Schiffer 2001:20)

58

Figure 12 Refuse cycle flow model for possible stages in artefact histories (Needham
and Spence 1997:78)
59

Figure 13 Flow model of waste movement in small historic towns

64

Figure 14 Project aim, objectives and hypotheses

68

Figure 15 Location of preliminary field visits undertaken between 12/11/04 and

04/01/05
69

Figure 16 Location and situation of Lauder, Scottish Borders

70

Figure 17 Location and situation of Pittenweem, Fife

72

Figure 18 Location and situation of Wigtown, Dumfries and Galloway

73

Figure 19 Location of survey area at Lauder (black rectangle). Red boundary delimits
1862AD urban extent taken from 1:10,560 scale Ordnance Survey map
inset © Landmark information group, www.old-maps.co.uk
76

Figure 20 Location of survey area at Pittenweem (orange rectangle). Red boundary

delimits 1855AD urban extent taken from 1:10,560 scale Ordnance Survey
map inset © Landmark information group, www.old-maps.co.uk
77

Figure 21 Location of survey area at Wigtown (orange rectangle). Red boundary

delimits 1850AD urban extent taken from 1:10,560 scale Ordnance Survey
map inset © Landmark information group, www.old-maps.co.uk
78

Figure 22 Stratified grid survey at Lauder showing auger sample locations (red
circles), soil pit locations (blue squares) and ‘reference’ soil pit locations
(green squares)
80

- 16 -
Figure 23 Stratified grid survey at Pittenweem (left) and Wigtown (right) showing auger
sample locations (red circles), soil pit locations (blue squares) and
‘reference’ soil pit locations (green squares)
81

Figure 24 Example of field sketch, profile description and soil sampling undertaken for
LA 2 (Lauder soil pit number 2)
83

Figure 25 Example of field sketch, profile description and soil sampling undertaken for
‘reference’ soil pit PT 1
83

Figure 26 Example images of coarse mineral anthropogenic features in thin section

from Lauder, Pittenweem and Wigtown. First row (left to right): pottery (OIL),
pottery (PPL), mortar (OIL), mortar (PPL). Second row (left to right): shell
(PPL), shell (XPL), bone (XPL), bone (PPL). Third row (left to right):
clinker/slag (PPL), heated mineral (OIL)
94

Figure 27 Example images of coarse organic anthropogenic features in thin section

from Lauder, Pittenweem and Wigtown. First row (PPL) (left to right):
Charcoal, FR4, FR 6, and FR 2. Second row (PPL) (left to right): FR1 (XPL),
FR7, FR5 and FR3. Third row (PPL) (left to right): FR 8, FR9, FR10 and FR
10 (XPL)
95

Figure 28 Left: slide prior to grid overlay, Right: slide with grid overlay, grid squares
selected through random number generation are highlighted with a blue
mark
96

Figure 29 Mean void space of Wigtown thin sections calculated incrementally (every
1%). Legend denotes town code (WG) and slide reference and context
number (1/1)
97

Figure 30 Mean % coarse material of Lauder thin sections calculated incrementally

(every 1%). Legend denotes town code (LA) and slide reference and context
number (1/1)
97

Figure 31 Mean % relative abundance of selected anthropogenic features for soil pit 4,
context 1 (4/1), Pittenweem
98

Figure 32 Mean % relative abundance of selected anthropogenic features for soil pit 4,
context 1, Wigtown
98

Figure 33 Spatial distributions of P (mg/Kg) (a) and Log P (mg/Kg) (b) at Pittenweem.
Red boundary delimits 1855AD urban extent
100

- 17 -
Figure 34 Delineation of zones at Lauder. Red boundary delimits 1862AD urban extent

101

Figure 35 Delineation of zones at Pittenweem (a) and Wigtown (b). Red boundary
delimits 1855AD and 1850AD urban extent respectively
102

Figure 36 Soil profile description of soil pit LA1 (see Appendix 1 for photographs)
109
Figure 37 Soil profile description of soil pit PT3 (see Appendix 1 for photographs)
109
Figure 38 Soil profile description of soil pit WG2 (see Appendix 1 for photographs)
110

Figure 39 Percentage abundance estimates of coarse mineral material within upper

(Aht 1) and lower (Aht 2) topsoil deposits in the High Street and Hinterland
Near zones at Lauder. Depth indicates the location of Kubiena tin samples
114

Figure 40 Percentage abundance estimates of coarse mineral material within upper

(Aht 1) and lower (Aht 2) topsoil deposits in the Harbour and High Street
zones at Pittenweem. Depth indicates the location of Kubiena tin samples
115

Figure 41 Percentage abundance estimates of coarse mineral material within upper

(Aht 1) and lower (Aht 2) topsoil deposits in the High Street zone at
Wigtown. Depth indicates the location of Kubiena tin samples
116

Figure 42 Kruskal-Wallis analysis with Dunn’s test multiple comparisons of FR 1 for

individual zones at Lauder (a) and Pittenweem (b), z values exceeding –Z or
Z are significant at p<0.05 confidence level
123

Figure 43 Kruskal-Wallis analysis with Dunn’s test multiple comparisons of FR 3 for

individual zones at Lauder (a) and Pittenweem (b), z values exceeding –Z or
Z are significant at p<0.05 confidence level
127

Figure 44 Kruskal-Wallis analysis with Dunn’s test multiple comparisons of FR 4 & 6

for individual zones at Lauder (a) and Pittenweem (b), z values exceeding –
Z or Z are significant at p<0.05 confidence level
129

Figure 45 Kruskal-Wallis analysis with Dunn’s test multiple comparisons of FR 5 & 7

for individual zones at Pittenweem, z values exceeding –Z or Z are
significant at p<0.05 confidence level
130

- 18 -
Figure 46 Kruskal-Wallis analysis with Dunn’s test multiple comparisons of FR 8 & 9
for individual zones at Lauder (a) and Pittenweem (b), z values exceeding –
Z or Z are significant at p<0.05 confidence level
132

Figure 47 % FR (10-255µm) within topsoil deposits in the High Street, Hinterland Near,
Hinterland Far and Thirlstane zones, and ‘reference’ soil at Lauder. Number
range refers to depth of Kubiena sample (cm)
134

Figure 48 % FR (10-255µm) within topsoil deposits in the Harbour, High Street,

Hinterland Near, and Hinterland Far zones, and ‘reference’ soil at
Pittenweem. Number range refers to depth of Kubiena sample (cm)
135

Figure 49 % FR (10-255µm) within topsoil deposits in the Harbour, High Street,

Hinterland Near, and Hinterland Far zones, and ‘reference’ soil at
Pittenweem. Number range refers to depth of Kubiena sample (cm)
136

Figure 50 Percentage abundance estimates of coarse organic material (Fuel Residue

categories) within upper (Aht 1) and lower (Aht 2) topsoil deposits in the
High Street and Hinterland Near zones at Lauder. Number range indicates
the depth (cm) of Kubiena tin samples
137

Figure 51 Percentage abundance estimates of coarse organic material (Fuel Residue

categories) within upper (Aht 1) and lower (Aht 2) topsoil deposits in the
Harbour and High Street zones at Pittenweem. Number range indicates the
depth (cm) of Kubiena tin samples
139

Figure 52 Percentage abundance estimates of coarse organic material (Fuel Residue

categories) within upper (Aht 1) and lower (Aht 2) topsoil deposits in the
High Street zone at Wigtown. Number range indicates the depth (cm) of
Kubiena tin samples
140

Figure 53 Distribution of topsoil depth at Lauder. Red boundary delimits 1862AD urban
extent
154

Figure 54 Spatial distribution of topsoil depth at Pittenweem (a) and Wigtown (b). Red
boundary delimits 1855AD and 1850AD urban extent respectively
155

Figure 55 (a) Boxplot of median topsoil depth for individual zones at Lauder;
boundaries of boxes represent interquartile range, green shading indicates
86.761% confidence interval for median, outliers identified as *, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Lauder, Z values exceeding –Z or Z are significant at p<0.05 confidence
level
156

- 19 -
Figure 56 (a) Boxplot of median topsoil depth for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, green shading indicates
86.761% confidence interval for median, outliers identified as *, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Pittenweem, Z values exceeding –Z or Z are significant at p<0.05
confidence level
157

Figure 57 (a) Boxplot of median topsoil depth for individual zones at Wigtown;
boundaries of boxes represent interquartile range, green shading indicates
80.529% confidence interval for median, outliers identified as *, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Pittenweem, Z values exceeding –Z or Z are significant at p<0.05
confidence level
157

Figure 58 Spatial distribution of soil pH at Lauder (a) 0-20cm and (b) 20-40cm depth.
Red boundary delimits 1862AD urban extent
159

Figure 59 Spatial distribution of soil pH at Pittenweem (a) 0-20cm, (b) 20-40cm, (c) 40-
60cm and (d) 60-80cm depth. Red boundary delimits 1855AD urban extent
160

Figure 60 Spatial distribution of soil pH at Wigtown (0-20cm depth). Red boundary

delimits 1850AD urban extent
161

Figure 61 (a) Boxplot of median pH for individual zones at Lauder; boundaries of

boxes represent interquartile range, blue shading indicates 86.761%
confidence interval for median, outliers identified as *, (b) Kruskal-Wallis
analysis with Dunn’s test multiple comparisons for individual zones at
Lauder (0-20cm depth), Z values exceeding –Z or Z are significant at p<0.05
confidence level
162

Figure 62 (a) Boxplot of median pH for individual zones at Pittenweem; boundaries of

boxes represent interquartile range, blue shading indicates 86.761%
confidence interval for median, outliers identified as *, (b) Kruskal-Wallis
analysis with Dunn’s test multiple comparisons for individual zones at
Pittenweem (0-20cm) using, Z values exceeding –Z or Z are significant at
p<0.05 confidence level
163

Figure 63 (a) Boxplot of median pH for individual zones at Wigtown; boundaries of

boxes represent interquartile range, blue shading indicates 80.529%
confidence interval for median, outliers identified as *, (b) Kruskal-Wallis
analysis with Dunn’s test multiple comparisons for individual zones at
Wigtown (0-20cm depth), Z values exceeding –Z or Z are significant at
p<0.05 confidence level
163

- 20 -
Figure 64 Boxplot of median pH for individual zones at Lauder for 0-20cm and 20-
40cm depth; boundaries of boxes represent interquartile range, outliers
identified as *
164

Figure 65 (a) Boxplot of median pH for 0-20cm, 20-40cm, 40-60cm and 60-80cm
depth at Pittenweem; boundaries of boxes represent interquartile range,
blue shading indicates 86.761% confidence interval for median, outliers
identified as *, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual depths at Pittenweem, Z values exceeding –Z or
Z are significant at p<0.05 confidence level
166

Figure 66 Boxplot of median pH for individual zones within 0-20cm, 20-40cm, 40-
60cm and 60-80cm depth at Pittenweem; boundaries of boxes represent
interquartile range, outliers identified as *
166

Figure 67 Spatial distribution of % LOI at Lauder (a) 0-20cm and (b) 20-40cm depth.
Red boundary delimits 1862AD urban extent
168

Figure 68 Spatial distribution of % LOI at Pittenweem (a) 0-20cm, (b) 20-40cm, (c) 40-
60cm and (d) 60-80cm depth. Red boundary delimits 1855AD urban extent

169

Figure 69 Spatial distribution of % LOI at Wigtown (0-20cm depth). Red boundary

delimits 1850AD urban extent
170

Figure 70 (a) Boxplot of median % LOI for individual zones at Lauder; boundaries of
boxes represent interquartile range, orange shading indicates 86.761%
confidence interval for median, outliers identified as *, (b) Kruskal-Wallis
analysis with Dunn’s test multiple comparisons for individual zones at
Lauder (0-20cm depth) , Z values exceeding –Z or Z are significant at
p<0.05 confidence level
172

Figure 71 (a) Boxplot of median % LOI for individual zones at Pittenweem; boundaries
of boxes represent interquartile range, orange shading indicates 86.761%
confidence interval for median, outliers identified as *, (b) Kruskal-Wallis
analysis with Dunn’s test multiple comparisons for individual zones at
Pittenweem (0-20cm depth), Z values exceeding –Z or Z are significant at
p<0.05 confidence level
173

Figure 72 Boxplot of median % LOI for individual zones at Wigtown; boundaries of

boxes represent interquartile range, outliers identified as *
173

Figure 73 Boxplot of median % LOI for individual zones at Lauder for 0-20cm and 20-
40cm depth; boundaries of boxes represent interquartile range, outliers
identified as *
174

- 21 -
Figure 74 Boxplot of median % LOI for individual zones at Pittenweem for 0-20cm, 20-
40cm, 40-60cm and 60-80cm depth; boundaries of boxes represent
interquartile range, outliers identified as *
175

Figure 75 Spatial distribution of (a) Ҳ (10-6m3Kg-1) and (b) ҲFD (10-6m3Kg-1) at Lauder.
Red boundary delimits 1862AD urban extent
177

Figure 76 Spatial distribution of (a) Ҳ (10-6m3Kg-1) and (b) ҲFD (10-6m3Kg-1) at

Pittenweem. Red boundary delimits 1855AD urban extent.
178

Figure 77 Spatial distribution of (a) Ҳ (10-6m3Kg-1) and (b) ҲFD (10-6m3Kg-1) at Wigtown.
Red boundary delimits 1850AD urban extent
179

Figure 78 (a) Boxplot of median Ҳ (10-6m3Kg-1) for individual zones at Lauder;

boundaries of boxes represent interquartile range, red shading indicates
86.761% confidence interval for median, outliers identified as *, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Lauder, Z values exceeding –Z or Z are significant at p<0.05 confidence
level
180

Figure 79 Boxplot of median and mean Ҳ (10-6m3Kg-1) for individual zones at

Pittenweem; bar indicates median, black circle denotes mean, boundaries of
boxes represent interquartile range, outliers identified as *
181

Figure 80 (a) Boxplot of median Ҳ (10-6m3Kg-1) for individual zones at Wigtown;

boundaries of boxes represent interquartile range, red shading indicates
80.529% confidence interval for median, outliers identified as *, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Wigtown, Z values exceeding –Z or Z are significant at p<0.05 confidence
level
181

Figure 81 (a) Boxplot of median ҲFD (10-6m3Kg-1) for individual zones at Lauder;
boundaries of boxes represent interquartile range, red shading indicates
86.761% confidence interval for median, outliers identified as *, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Lauder, Z values exceeding –Z or Z are significant at p<0.05 confidence
level
184

Figure 82 (a) Boxplot of median ҲFD (10-6m3Kg-1) for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, red shading indicates
86.761% confidence interval for median, outliers identified as *, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Pittenweem, Z values exceeding –Z or Z are significant at p<0.05
confidence level
185

- 22 -
Figure 83 (a) Boxplot of median Ҳ (10-6m3Kg-1) for individual zones at Wigtown;
boundaries of boxes represent interquartile range, red shading indicates
80.529% confidence interval for median, outliers identified as *, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Wigtown, Z values exceeding –Z or Z are significant at p<0.05 confidence
level
185

Figure 84 Scatterplot of Fe (mg/Kg) vs. Ҳ (10-6m3Kg-1) for Lauder, Pittenweem and

Wigtown
187

Figure 85 Spatial distribution of Log Ba (mg/Kg) at Lauder. Red boundary delimits

1862AD urban extent
191

Figure 86 Spatial distribution of Log Ba (mg/Kg) at Pittenweem (a) and Wigtown (b).
Red boundary delimits 1855AD and 1850AD urban extent respectively
192

Figure 87 Boxplot of median and mean Ba (mg/Kg) for individual zones at Lauder; bar
indicates median, black circle denotes mean; boundaries of boxes represent
interquartile range, outliers identified as *
193

Figure 88 (a) Boxplot of median Ba (mg/Kg) for individual zones at Pittenweem;

boundaries of boxes represent interquartile range, green shading indicates
86.761% confidence interval for median, outliers identified as*, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Pittenweem, -Z or Z are significant at p<0.05 confidence level
194

Figure 89 (a) Boxplot of median Ba (mg/Kg) for individual zones at Wigtown;

boundaries of boxes represent interquartile range, green shading indicates
80.529% confidence interval for median, outliers identified as*, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Wigtown, -Z or Z are significant at p<0.05 confidence level
194

Figure 90 Spatial distribution of Log Ca (mg/Kg) at Lauder. Red boundary delimits

1862AD urban extent
195

Figure 91 Spatial distribution of Log Ca (mg/Kg) at Pittenweem (a) and Wigtown (b).
Red boundary delimits 1855AD and 1850AD urban extent respectively
196

Figure 92 (a) Boxplot of median Ca (mg/Kg) for individual zones at Lauder; boundaries
of boxes represent interquartile range, yellow shading indicates 86.761%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis
analysis with Dunn’s test multiple comparisons for individual zones at
Lauder, -Z or Z are significant at p<0.05 confidence level
197

- 23 -
Figure 93 (a) Boxplot of median Ca (mg/Kg) for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, yellow shading indicates
86.761% confidence interval for median, outliers identified as*, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Pittenweem, -Z or Z are significant at p<0.05 confidence level
198

Figure 94 (a) Boxplot of median Ca (mg/Kg) for individual zones at Wigtown;

boundaries of boxes represent interquartile range, yellow shading indicates
80.529% confidence interval for median, outliers identified as*, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Wigtown, -Z or Z are significant at p<0.05 confidence level
198

Figure 95 Spatial distribution of Log Pb (mg/Kg) at Lauder. Red boundary delimits

1862AD urban extent
199

Figure 96 Spatial distribution of Log Pb (mg/Kg) at Pittenweem (a) and Wigtown (b).
Red boundary delimits 1855AD and 1850AD urban extent respectively
200

Figure 97 (a) Boxplot of median Pb (mg/Kg) for individual zones at Lauder; boundaries
of boxes represent interquartile range, green shading indicates 86.761%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis
analysis with Dunn’s test multiple comparisons for individual zones at
Lauder, -Z or Z are significant at p<0.05 confidence level
201

Figure 98 Boxplot of median and mean Pb (mg/Kg) for individual zones at Pittenweem;
bar indicates median, black circle denotes mean; boundaries of boxes
represent interquartile range, outliers identified as *
202

Figure 99 (a) Boxplot of median Pb (mg/Kg) for individual zones at Wigtown;

boundaries of boxes represent interquartile range, green shading indicates
80.529% confidence interval for median, outliers identified as*, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Wigtown, -Z or Z are significant at p<0.05 confidence level
202

Figure 100 Spatial distribution of Log K (mg/Kg) at Lauder. Red boundary delimits
1862AD urban extent
203

Figure 101 Spatial distribution of Log Pb (mg/Kg) at Pittenweem (a) and Wigtown (b).
Red boundary delimits 1855AD and 1850AD urban extent respectively
204

Figure 102 (a) Boxplot of median K (mg/Kg) for individual zones at Lauder; boundaries
of boxes represent interquartile range, pink shading indicates 86.761%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis
analysis with Dunn’s test multiple comparisons for individual zones at
Lauder, -Z or Z are significant at p<0.05 confidence level
205

- 24 -
Figure 103 (a) Boxplot of median K (mg/Kg) for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, pink shading indicates
86.761% confidence interval for median, outliers identified as*, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Pittenweem, -Z or Z are significant at p<0.05 confidence level
206

Figure 104 (a) Boxplot of median K (mg/Kg) for individual zones at Wigtown; boundaries
of boxes represent interquartile range, pink shading indicates 80.529%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis
analysis with Dunn’s test multiple comparisons for individual zones at
Wigtown, -Z or Z are significant at p<0.05 confidence level
206

Figure 105 Spatial distribution of Log P (mg/Kg) at Lauder. Red boundary delimits
1862AD urban extent
207

Figure 106 Spatial distribution of Log P (mg/Kg) at Pittenweem (a) and Wigtown (b).
Red boundary delimits 1855AD and 1850AD urban extent respectively
208

Figure 107 (a) Boxplot of median P (mg/Kg) for individual zones at Lauder; boundaries
of boxes represent interquartile range, purple shading indicates 86.761%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis
analysis with Dunn’s test multiple comparisons for individual zones at
Lauder, -Z or Z are significant at p<0.05 confidence level
209

Figure 108 (a) Boxplot of median P (mg/Kg) for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, purple shading indicates
86.761% confidence interval for median, outliers identified as*, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Pittenweem, -Z or Z are significant at p<0.05 confidence level
210

Figure 109 (a) Boxplot of median P (mg/Kg) for individual zones at Wigtown; boundaries
of boxes represent interquartile range, purple shading indicates 80.529%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis
analysis with Dunn’s test multiple comparisons for individual zones at
Wigtown, -Z or Z are significant at p<0.05 confidence level
210

Figure 110 Spatial distribution of Log Sr (mg/Kg) at Lauder. Red boundary delimits
1862AD urban extent
212

Figure 111 Spatial distribution of Log Sr (mg/Kg) at Pittenweem (a) and Wigtown (b).
Red boundary delimits 1855AD and 1850AD urban extent respectively
213

- 25 -
Figure 112 (a) Boxplot of median Sr (mg/Kg) for individual zones at Lauder; boundaries
of boxes represent interquartile range, blue shading indicates 86.761%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis
analysis with Dunn’s test multiple comparisons for individual zones at
Lauder, -Z or Z are significant at p<0.05 confidence level
214

Figure 113 (a) Boxplot of median Sr (mg/Kg) for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, blue shading indicates
86.761% confidence interval for median, outliers identified as*, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Pittenweem, -Z or Z are significant at p<0.05 confidence level
215

Figure 114 (a) Boxplot of median Sr (mg/Kg) for individual zones at Wigtown;
boundaries of boxes represent interquartile range, blue shading indicates
80.529% confidence interval for median, outliers identified as*, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Wigtown, -Z or Z are significant at p<0.05 confidence level
215

Figure 115 Spatial distribution of Log Zn (mg/Kg) at Lauder. Red boundary delimits
1862AD urban extent
216

Figure 116 Spatial distribution of Log Zn (mg/Kg) at Pittenweem (a) and Wigtown (b).
Red boundary delimits 1855AD and 1850AD urban extent respectively
217

Figure 117 (a) Boxplot of median Zn (mg/Kg) for individual zones at Lauder; boundaries
of boxes represent interquartile range, cyan shading indicates 86.761%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis
analysis with Dunn’s test multiple comparisons for individual zones at
Lauder, -Z or Z are significant at p<0.05 confidence level
218

Figure 118 (a) Boxplot of median Zn (mg/Kg) for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, cyan shading indicates
86.761% confidence interval for median, outliers identified as*, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Pittenweem, -Z or Z are significant at p<0.05 confidence level
219

Figure 119 (a) Boxplot of median Zn (mg/Kg) for individual zones at Wigtown;
boundaries of boxes represent interquartile range, cyan shading indicates
80.529% confidence interval for median, outliers identified as*, (b) Kruskal-
Wallis analysis with Dunn’s test multiple comparisons for individual zones at
Wigtown, -Z or Z are significant at p<0.05 confidence level
220

- 26 -
Figure 120 Dendrogram showing results of Cluster analysis for elemental data at
Lauder. Seven main groups are identified; cluster 1 (red), cluster 2 (green),
cluster 3 (blue), cluster 4 (orange), cluster 5 (magenta), cluster 6 (purple)
and cluster 7 (cyan)
235

Figure 121 Dendrogram showing results of Cluster analysis for elemental data at
Pittenweem. Five main groups are identified; cluster 1 (red), cluster 2
(yellow), cluster 3 (green), cluster 4 (blue) and cluster 5 (brown)
237

Figure 122 Dendrogram showing results of Cluster analysis for elemental data at
Wigtown. Six main groups are identified; cluster 1 (red), cluster 2 (green),
cluster 3 (blue), cluster 4 (coral), cluster 5 (brown) and cluster 5 (magenta)
192

Figure 123 Wigtown harbour (above) was used primarily as a fishing port between the
15th and 18th centuries (Simpson and Stevenson 1981b). The harbour fell
into disuse in the early 19th century due to increasing sedimentation and was
replaced by a new harbour 400m south of Wigtown (Image © Royal
Commission on the Ancient and Historical Monuments of Scotland, Licensor
www.scran.ac.uk)
249

Figure 124 The Old Boat Haven (left) was used as the main harbour of Pittenweem until
1541AD. It was carved into the rocky shoreline and its pier was used as a
natural outcrop with a road cut into it. Pittenweem harbour (right) was built in
1541AD to accommodate expanding fishing fleets. The inner harbour
comprises the west and east pier, both of which have been systematically
rebuilt over the 17th to 19th centuries, and the outer harbour is enclosed by
the south pier (Images © Royal Commission on the Ancient and Historical
Monuments of Scotland, Licensor www.scran.ac.uk)
249

Figure 125 Map of Lauder extracted from the 1747-1755AD William Roy Military Survey
of Scotland © The British Library Board, Licensor www.nls.uk/roy/style.html.
Red boundaries represent buildings and man-made structures and parallel
hatching indicates cultivated land. Burgh acres are represented by the
delineation of strips of land running parallel to burgage plots
253

Figure 126 Map of Pittenweem extracted from the 1747-1755AD William Roy Military
Survey of Scotland © The British Library Board, Licensor
www.nls.uk/roy/style.html. Red boundaries represent buildings and man-
made structures and parallel hatching indicates cultivated land
254

Figure 127 Map of Wigtown extracted from the 1747-1755AD William Roy Military
Survey of Scotland © The British Library Board, Licensor
www.nls.uk/roy/style.html. Red boundaries represent buildings and man-
made structures and parallel hatching indicates cultivated land. Burgh acres
are represented by the delineation of strips of land running parallel to
burgage plots north and south of Wigtown
254

- 27 -
Figure 128 Aerial Image of Lauder showing (a) burgh core, (b) Hinterland Near zone, (c)
Hinterland Far zone and (d) Thirlstane zone, image © Andrew Buchanan
2004, www.holy-cow.co.uk/
257

Figure 129 Invoice for the sale of ‘street manure’ at Pittenweem to a local farmer dating
1919 AD, dunghills were therefore a persistent feature in burgh cores
(Image © Scottish Life Archive, Licensor www.scran.ac.uk)
261

Figure 130 Photograph of a man with a cart load of seaweed in the early 20th century
Pittenweem. In previous centuries the right to collect seaweed was
associate with land ownership, hence burgesses had rights to seaweed on
particular stretches of beach (Image © National Museums Scotland,
Licensor www.scran.ac.uk)
261

Figure 131 Revised flow model of resource and waste movement in historic small towns

264

Figure 132 Soil map of Lauder © Soil Survey of Scotland 1959. Soils within and
adjacent to Lauder are mapped as a freely drained brown forest soil with low
base status belonging to the Lauder (LA) soil series
268

Figure 133 Soil map of Pittenweem © Soil Survey of Scotland 1975. Soils within
Pittenweem are not classified and adjacent soils are mapped as an
imperfectly drained brown forest soil belonging to the Quivox (QX) soil
series
269

Figure 134 Soil map of Wigtown © Soil Survey of Scotland 1971. Soils within Wigtown
are not classified and adjacent soils are mapped as a freely draining brown
forest soil belonging to the Linhope (LP1) soil series
269

Figure 135 Formation of the Urbic horizon (Nikolaevna and Vadimovna 2003)
272

Figure 136 Selected pottery finds from Pittenweem. First row (left to right), 16th century
AD Valencian Lustreware bowl (Spain), 17th AD century pottery from
Saintonge (south-west France), 17th century cooking pots (north Germany),
17th century AD earthenware skillets (Scotland). Second row (left to right),
17th century AD Green Glazed jars (Scotland), 17th century Loire Ware Jugs
(west France), 17th century AD Westerwald AD monochrome stoneware
(Germany), 17th century AD Westerwald polychrome stoneware (Germany)
(Images © Colin J Martin, Licensor www.scran.ac.uk)
274

- 28 -
 Tables

Table 1 Urban Percentage of total population of Scotland, 1500-1800AD and

percentage increase in urban population from previous date (from De Vries
1984:39)
44

Table 2 Percentage of total population in Scottish towns, mid 17th century to late
18th century AD (from Whyte 1989:28)
44

Table 3 Mineral Anthropogenic Features (PPL: Plane polarised light, XPL: Cross
polarised light, OIL: Oblique incident light)
92

Table 4 Organic Anthropogenic Features (PPL: Plane polarised light, XPL: Cross
polarised light, OIL: Oblique incident light), FR: Fuel Residue
93

Table 5 Summary of trends in coarse mineral material identified at Lauder,

Pittenweem and Wigtown, summary abundances are given in brackets
119

Table 6 Comparison of trends in the abundance of coarse mineral anthropogenic

material at Lauder, Pittenweem and Wigtown, summary abundances are
given in brackets
120

Table 7 Summary of trends in coarse organic inclusions identified at Lauder,

Pittenweem and Wigtown, summary abundances are given in brackets
144

Table 8 Summary of trends in the abundance of coarse organic anthropogenic

material between High Street zones at Lauder, Pittenweem and Wigtown as
indicated by Kruskal-Wallis analysis with Dunn’s test
146

Table 9 Summary of trends in key micromorphological analyses of topsoils within

Lauder, Pittenweem and Wigtown
153

Table 10 Summary of trends in soil physical and chemical properties identified at

Lauder, Pittenweem and Wigtown
189

Table 11 Comparison of trends in pH and LOI associated with depth at Lauder and
Pittenweem
190

Table 12 Comparison of trends in soil physical and chemical properties identified at

Lauder, Pittenweem and Wigtown
190

- 29 -
Table 13 Summary of trends in selected elemental concentrations identified at
Lauder, Pittenweem and Wigtown
222

Table 14 Comparison of trends in selected elemental concentrations identified at

Lauder, Pittenweem and Wigtown
223

Table 15 Average elemental concentrations (mg/Kg) of individual zones at Lauder

(median used for non-parametric data, mean used for parametric data). Test
indicates whether Kruskal-Wallis (KW) or One Way ANOVA (ANOVA) used
to compare elemental concentrations between zones
225

Table 16 Multiple comparisons for individual zones at Lauder (Dunn’s test used for
non-parametric data, Tukey-Kramer post-hoc analyses for parametric data)
226

Table 17 Average elemental concentrations (mg/Kg) of individual zones at

Pittenweem (median used for non-parametric data, mean used for
parametric data). Test indicates whether Kruskal-Wallis (KW) or One Way
ANOVA (ANOVA) used to compare elemental concentrations between
zones
228

Table 18 Multiple comparisons for individual zones at Pittenweem (Dunn’s test used
for non-parametric data, Tukey-Kramer post-hoc analyses for parametric
data)
229

Table 19 Average elemental concentrations (mg/Kg) for individual zones at Wigtown

(median used for non-parametric data, mean used for parametric data). Test
indicates whether Kruskal-Wallis (KW) or One Way ANOVA (ANOVA) used
to compare elemental concentrations between zones
231

Table 20 Multiple comparisons for individual zones at Wigtown (Dunn’s test used for
non-parametric data, Tukey-Kramer post-hoc analyses for parametric data)
232

Table 21 Summary of Cluster analysis amalgamation steps for elemental data at

Lauder
235

Table 22 Summary of Cluster analysis amalgamation steps for elemental data at

Pittenweem
237

Table 23 Summary of Cluster analysis amalgamation steps for elemental data at

Wigtown
239

- 30 -
Table 24 Summary of discriminant classification for individual zones within Lauder;
‘true zone’ indicates the location of observations (sample points), ‘predicted
zone’ indicates the predicted location of observations, ‘No. observations’
indicates the total number of observations per zone and ‘No. observations
correct’ signifies the number of observations correctly classified per zone
240

Table 25 Summary of classification for individual zones within Pittenweem; ‘true zone’
indicates the location of observations (sample points), ‘predicted zone’
indicates the predicted location of observations, ‘No. observations’ indicates
the total number of observations per zone and ‘No. observations correct’
signifies the number of observations correctly classified per zone
241

Table 26 Summary of classification for individual zones within Wigtown; ‘true zone’
indicates the location of observations (sample points), ‘predicted zone’
indicates the predicted location of observations, ‘No. observations’ indicates
the total number of observations per zone and ‘No. observations correct’
signifies the number of observations correctly classified per zone
242

Table 27 Summary of classification for the High Street zone at Lauder, Pittenweem
and Wigtown; ‘true burgh’ indicates the location of observations (sample
points), ‘predicted burgh’ indicates the predicted location of observations,
‘No. observations’ indicates the total number of observations within each
High Street zone and ‘No. observations correct’ signifies the number of
observations correctly classified per burgh
242

Table 28 Summary of classification for Hinterland Near zones at Lauder, Pittenweem

and Wigtown; ‘true burgh’ indicates the location of observations (sample
points), ‘predicted burgh’ indicates the predicted location of observations,
‘No. observations’ indicates the total number of observations within each
Hinterland Near zone and ‘No. observations correct’ signifies the number of
observations correctly classified per burgh
243

Table 29 Summary of elemental enhancement within zones at Lauder; pink indicates

enhancement within the High Street zone only, red indicates enhancement
within Hinterland Near zone only, orange indicates enhancement within High
Street and Hinterland Near zones, and blue indicates enhancement within
High Street, Hinterland Near and Thirlstane zones
244

Table 30 Summary of elemental enhancement within zones at Pittenweem; red

indicates enhancement within the Hinterland Near zone only, orange
indicates enhancement within the Harbour and High Street zones and blue
indicates enhancement within the Harbour, High Street and Hinterland Near
zones
245

Table 31 Summary of elemental enhancement within zones at Wigtown; pink

indicates enhancement within the High street zone only and orange
indicates enhancement within the High Street and Showfield zones
246

- 31 -
Table 32 Comparison of urban organic waste recycling practises in modern and
historic towns, modern examples taken from Cofie and Bradford (2006)
266

Table 33 Summary of soil classification terminology relevant to soils in historic towns

(FAO 2006, Gong 1999, Isbell 2002, Nikolaevna and Vadimovna 2003)
271

- 32 -
 1 The Importance of Heritage and Conservation in
Historic Small Towns

This chapter presents a brief introduction to previous and current trends in urban
heritage conservation. Specific attention is paid to the importance of soils within historic
small towns.

1.1 Trends in Urban Conservation

Historic small towns are complex urban archaeological sites characterised by the legacy
of successive inhabitants. Moreover, they are diverse in form, scale and location, and
have unique identities deep-rooted in local tradition, industries and agriculture (Evans
1999). Until recently little consideration has been given to conserving urban
archaeology. Appreciation of the cultural value of urban heritage was especially lacking
during the immediate post-war period in Britain when demands for new housing stock
and infrastructural development led to the irreplaceable loss of countless historic
buildings and urban deposits (Eydmann 1999).

The emergence of urban archaeology as an established sub-discipline in the 1970s

signified an increasing awareness of the need to investigate and preserve the cultural
and visual roots of historic urban environments (Owen 1999). Conservation in Britain
focussed largely on protecting surviving architecture and structural remains, for example
the 1972 Town and Country Planning Act gave statutory protection for listed buildings
(Lynch 1999). Moreover, in Scotland emphasis was placed on preserving ecclesiastical
and municipal buildings and monuments. However, there was increasing recognition that
such structures do not necessarily represent everyday life in past urban societies (Lynch
1999).

1.1.1 Historic Small Towns

In 1976 the Scottish Burgh Survey was established by Historic Scotland as a response
to concerns about increasing threats to the heritage of Scotland’s burghs (Dennison
1999a). These surveys were progressive in that they evaluated the potential of urban
archaeology both above and below ground (Coleman 1999). This approach viewed the
urban environment as a collective rather than a series of independent structures, hence
was useful for identifying sensitive zones and areas susceptible to future development.

- 33 -
Despite initial and ongoing work of the Scottish Burgh Survey, the legacy of past urban
life within Scotland’s historic burghs is threatened, for example re-development within
burgh cores and processes of peripheral expansion are leading to the loss of
characteristic medieval street layouts. This is especially resonant considering that urban
archaeological deposits located within these areas are vulnerable to damage and loss.
Moreover, deposits within smaller burghs are identified as particularly susceptible given
that they are often thinner and overlooked in favour of artefact rich sequences from
larger burghs (Owen 1999).

It is argued that soil deposits within historic burghs represent a heritage of everyday
urban life not readily documented or represented by surviving structures and
monuments. Moreover, it is expected that soils are heavily impacted by processes of
waste disposal and management given it is estimated that up to 182,000 litres of urine,
182,500 Kg of solid waste, 8100 Kg of ash from cooking and heating, and 36,500 Kg of
human faeces were produced annually per 100 households in pre-industrial societies
(Brothwell 1982). Investigation into the legacy of urban soils will enable characterisation
of a unique and until recently, largely ignored archaeological archive.

1.2 Thesis Structure

The following study investigates the effect of waste in and around soils in small historic
towns. A summary of the historical legacy of urban soils with specific reference to
Scottish royal burghs is provided in Chapter 2, followed by presentation of the research
framework including aims, objectives and hypotheses in Chapter 3. Chapter 4 outlines
materials and methods used in this study with specific reference to site selection, field
work, laboratory analyses, soil micromorphology and data analyses. Results soil
micromorphology, physical and chemical properties, and elemental concentrations in
Chapter 5, 6 and 7 respectively. Chapter 8 discusses the impact of waste disposal in
and around Lauder, Pittenweem and Wigtown. To conclude some reflections on the
wider significance of waste disposal in and around historic small towns are offered in
Chapter 9.

- 34 -
 2 The Historical Legacy of Urban Anthrosols with
Specific Reference to Scottish Royal Burghs

2.1 Urban Anthrosols

Soil classification systems no longer use soil formation processes to discriminate between
soil types, thus, the term ‘anthropogenic soils’ is technically redundant. However here it is
argued that there is a case for re-examination of anthropogenic soil classification with
specific reference to soils studied within this research (Dudal 2004).

2.1.1 Anthropogenic Soil Classification

Anthropogenic soils are accounted for at the highest order within the Soil Classification for
England and Wales as Man-made soils (Avery 1980). Similarly anthropogenic soils are
separated at the highest level within the World Reference Base for Soil Resources (WRB)
as the Anthrosol group (FAO 2006). Conversely, the North American based Soil
Taxonomy (Soil Survey Staff 1999) does not have a separate category. Consequently,
anthropogenic soils are separated at the sub-order level Anthrepts within Inceptisols using
Anthropic and Plaggen epipedon diagnostic surface horizons. Nevertheless these
systems are in agreement that anthropogenic soils occur as a result of human
modification over time and secondly, such changes are attributable to permanent
settlement and agricultural practises. Regarding the Soil Survey of Scotland (1984),
provision for anthropogenic soils is absent from soil classification categories although
discrete phases of deepened topsoil are mapped for Orkney.

Dudal (2004:3) offers an alternative proposal suggesting 6 main types of anthropogenic

soils namely; human induced changes of soil class, human made diagnostic horizons,
human induced new parent material, human induced deep soil disturbance, human
induced change of landform and human induced topsoil changes. The latter category,
human induced topsoil changes, is particularly resonant to this study given that it is the
impact of waste disposal on topsoil in historic Scottish Royal burghs which is of principal
interest. However the importance of topsoil characteristics in soil classification systems
has traditionally been marginalised with preference given to more stable subsurface
horizons (Dudal 2004). Within the scope of this study, it therefore seems particularly
important in the first instance to establish the range of physical and chemical
characteristics of these modified topsoils.

- 35 -
2.1.2 Urban Soil Classification

The soils studied in this research is located within the urban environment, specifically in
Scottish Royal burghs, known to be urban for over 600 years. Soils in the urban
environment are distinct from natural soils on account of their formation and location, in
particular the scale and intensity of human impacts on them (Bullock and Gregory 1991,
De Kimpe and Morel 2000). Despite the unique qualities of soils in the urban environment
knowledge regarding the nature, diversity and extent is currently lacking (De Kimpe and
Morel 2000, Hollis 1991. This is further reflected in the absence of sufficient provision for
urban soils in both the Soil Classification for England and Wales and Soil Taxonomy.
Regarding the WRB Soil Classification, an attempt at classifying urban soils has recently
been made through the inclusion in 2006 of a new Reference Soil Group (RSG),
Technosols. Technosols are defined as soils

“whose properties and pedogenesis are dominated by their technical origin. They contain
a significant amount of artefacts, or are sealed by technic hard rock. They include soils
from wastes (landfills, sludge, cinders, mine spoils and ashes), pavements with their
underlying unconsolidated materials, soils with geomembranes and constructed soils in
human–made materials”.
(FAO 2006:95)

Technosols appear to be based on contemporaneous urban environments. Consequently

it seems urban soils subject to human impacts over many centuries are largely ignored.
Potentially, soils which do not meet Technosol requirements should be classed as
Anthrosols. However the Anthrosol group primarily includes soils resulting from long
standing practises of agriculture for example, organic material addition, irrigation and
cultivation. Additionally, it seems that the Technosol RSG is not exclusive from the
Anthrosol RSG given that in principle Technosols fulfil the remit of Anthrosols as “soils
that have been modified profoundly through human activities” (FAO 2006:71).

As previously mentioned, soil studied in this research has formed an intrinsic part of the
urban landscape for many centuries and accordingly it is suggested that changes in
physical and chemical soil properties may reflect past processes specific to the urban
environment. It is therefore argued that these soils do not readily fit into the established
models of soil classification presented here. Consequently soils investigated within this
project are generically referred to as Urban Anthrosols.

- 36 -
2.1.3 Current Research into Urban Anthrosols

Urban soils act as a sink for a variety of anthropogenic activities such as burning, vehicle
emissions and industrial wastes resulting in soil contamination. These soils are often
contaminated with carcinogenic, mutagenic and toxic elements and compounds which can
be directly ingested, inhaled or consumed through home grown food (Alloway 2004,
Hursthouse et al., 2004, Mielke et al., 2001). Hence the main concern regarding polluted
urban soil is risk to human health (Chung Wong and Li 2004, Hough 2007, Liu et al.,
2005, Moller et al., 2006, Senesi et al., 1999, Yongming et al., 2006) . This problem has
globally attracted the attention of governments and regulatory bodies (Li et al., 2004,
Tijhuis et al., 2002) and is especially resonant in industrialising countries where
urbanisation is accelerating (Lu et al., 2003). The need to develop regulations and
standards to minimise subsequent risk is imperative (Ona et al., 2006). Consequently it is
not surprising that current Urban Anthrosol research is focussed on aspects of soil
contamination within modern towns and cities. These contaminants have largely been
considered in two distinct groups, namely heavy metals and organic pollutants.

2.1.3.1 Heavy Metals

Urban soils exhibit elevated levels of certain elements for example Moller et al., (2005)
comment that soils in Damascus city have significantly increased concentrations of Pb, Cu
and Zn compared the surrounding rural areas. Similarly Chirenje et al., (2003) found that
As is significantly greater in urban soils of Florida compared to non-urban areas.
Accordingly attempts have been made to distinguish discrete associations of elements
indicative of anthropogenic origin for example, Manta et al., (2002) discriminate between
anthropogenic contributions of Pb, Zn, Cu, Sn and Hg and lithogenic inputs of Co, Ni, Cr
and Mn in Palermo, Sicily. Likewise both Li et al., (2004) and Yongming et al., (2006)
separate anthropogenic elemental signals from natural associations in Hong Kong and
Xi’an, Central China, respectively. These anthropogenic groups routinely contain Pb, Zn,
Cu and Hg (Moller et al., 2005, Tijhuis et al., 2001). Additionally, it is recognised that
certain elements such as Cu, Zn and Pb are more mobile and bio-available in urban soils,
thus increasing the potential of groundwater contamination (Lu et al., 2003, Manta et al.,
2002).

- 37 -
Elemental concentrations typically vary across urban soils (Li et al., 2004, Ona et al.,
2006), however certain distribution patterns have been identified such as contamination
near industrial areas (Kelly et al., 1996, Moller et al., 2005, Tijhuis et al., 2001) and
pollution pathways associated with vehicular emissions (Chung Wong and Li 2004, Kelly
et al., 1996, Li et al., 2004, Manta et al., 2002).

2.1.3.2 Organic Pollutants

Similar to certain elements, organic pollutants such as Polycyclic Aromatic Hydrocarbons
(PAHs) and Polychlorinated Biphenyls (PCBs) have been found in elevated levels in
urban soils (Garcia-Alonso et al., 2003). Moreover, Mielke et al., (2001) identify a strong
association between PAHs and heavy metals in urban soils of New Orleans.
Consequently when assessing environmental quality it appears that urban soil
contaminants should not be examined in isolation. Additionally, attempts should be made
to integrate physical and biological properties when investigating urban soils, for instance
Scharenbroch et al., (2005) suggest time since initial disturbance is inversely related to
stability of physical, chemical and biological soil properties. This suggestion is supported
by Nora et al., (2005) who advocate investigating soil chemical and physical properties
collectively and linking those properties to the function of urban soils as plant, faunal and
microbial habitats.

2.1.4 The Historical Legacy of Urban Anthrosols

Since the advent of urbanisation humans have inadvertently altered physical and chemical
soil properties (Griffith 1980) the nature of which is determined by management and
accumulation of anthropogenic waste materials (Alexandrovskaya and Alexandrovskiy
2000, Rathje and Murphy 2001). Urban soils therefore provide a unique opportunity to
reconstruct past activities, which traditionally may have been overlooked owing to lack of
material evidence (Terry et al., 2004). Research on the historical legacy of urban
Anthrosols is limited in comparison to the overwhelming focus on contamination of
contemporary urban soils. However studies demonstrate the utility of urban soils in
reconstructing human activities in space (Bull et al., 2001, Wells 2004) and time
(Shahack-Gross et al., 2005, Alexandrovskaya and Panova 2003). Current knowledge
regarding past urban soil amendments is reviewed hereafter.

2.1.4.1 Occupational Sequences

Permanent human settlement modifies urban soil properties resulting in elemental
enhancement (Griffith 1980) and changes in certain physical properties such as colour,
depth and organic matter content (Aston 1998). Investigation of these changes has

- 38 -
enabled the identification of distinct phases of site activity such as population decline and
urban abandonment between the Roman and Medieval periods as inferred by ‘Dark Earth’
(Macphail et al., 2003, Ottaway 1992). Similarly Alexandrovskaya and Panova (2003) and
Alexandrovskaya and Alexandrovskiy (2000) reconstruct soil formation history through
characterisation of successive deposits in Moscow, for example the medieval habitation
layer is enriched in construction debris and waste materials such as charcoal, lime and
timber, and has maximum concentrations of Pb, Cu and As attributable to industrial
processes. Analysis of soil and sedimentary sequences has also been used to explore
how individual space within a site is used over time. Shahack-Gross et al., (2005) identify
successive ‘fill’ and ‘floor’ depositional units from Late Iron Age I to Early Iron Age II in a
public space at Tel Dor, Israel. The fill layers comprise domestic debris such as bone,
charcoal, plaster and shells and the floor layers reflect a change from fish processing to
livestock penning. These studies demonstrate the value of using anthropogenic waste
inputs in soil characterisation and for reconstructing past activities.

2.1.4.2 Elemental Enhancement

Aston (1998) proposes five ways elemental composition influenced by past activities;
human habitation, stalled animals, use of fires/hearths, metal working and other
processing activities. It is therefore recognised that multi-element analysis of urban soils
can provide information regarding the nature of past activities and the spatial distribution
of their associated deposits. This is especially resonant when preservation of
archaeological remains is limited, for example Cook et al., (2005) identify three elemental
hotspots in a Roman house complex, Silchester providing evidence for previously
unidentified working of copper alloys, gold and silver, and potentially lead. Where artefacts
have been recovered examination of elemental distributions serves to substantiate space
use patterning. Terry et al., (2004) associate midden refuse with elevated levels of P
indicating areas of food preparation, consumption and storage at Aguateca, Guatemala.
Moreover artefacts associated with crystal processing are linked to enhanced Fe
concentrations indicative of specific workshops.

2.1.4.3 Elemental Signatures

Elemental signatures can be used to distinguish between anthropogenic deposits. Wells
et al., (2000) differentiate between kitchen, workshop, craft and ceremonial middens using
specific elemental signatures of ancient Anthrosols in residential areas of Piedras Negras,
Guatemala. Wells (2004) also delineates manufacturing, ritual and domestic functional
areas using elemental signatures and links these areas to their associated middens at El
Coyote, Honduras. Analysis of elemental distributions in the urban environment, therefore,

- 39 -
provides a valuable tool for characterising and discriminating between deposits and for
hypothesising which activities led to the formation of certain Urban Anthrosols. Spatial
analysis of anthropogenic elements is identified as a necessary technique for more
detailed investigation of waste disposal and management in Scottish Royal burghs. This
recognition also follows Terry et al’s., (2004) recommendation that studies are needed in
regional contexts to evaluate the effect of different climatic conditions and local soil
properties on elemental concentrations.

2.1.4.4 Soil Improvement

The impact of past human activity is not confined to urban centres in isolation. Bull et al.,
(2001) present evidence to suggest that during the Minoan period domestic waste was
transported from the main site of occupation on Pseira Island, Crete and applied to nearby
agricultural terraces (Figure 1). Similarly the occurrence of Terra Mulata, an Amazonian
Black Earth, is attributable to intentional application of anthropogenic wastes to
agricultural fields surrounding high density pre-Columbian Amerindian settlements
(Sombroek et al., 2002). Analyses of urban environments and associated waste disposal
should therefore extend to surrounding areas likely to have been affected by central urban
activities.

Figure 1: Cross section of terrace P1 Pseira Island, Crete showing Minoan potsherd scatters
indicative of systematic spreading of household wastes across the cultivated landscape as
a fertiliser (Bullock et al., 2001: 227)

- 40 -
2.1.5 Summary

Urban Anthrosols are distinctive owing to their location and that their properties are
modified through waste amendments. Much consideration has been given to how these
properties reflect current human influence. However, recent studies indicate the potential
of investigating Urban Anthrosols to reconstruct past human activities in space and time. It
is therefore assumed that urban soils in and near to Scottish Royal burghs will exhibit a
historical legacy. Accordingly the following sections provide an overview of Royal burghs,
environmental conditions and existing evidence for soil modification.

- 41 -
2.2 Waste Sources and Management in Royal Burghs

2.2.1 Medieval Origins of Urbanism

Despite absence of evidence, it is likely that pre-urban nuclei existed in Scotland before
the 12th century AD (Dicks 1983). However, it was not until David I (1124-1153AD) whose
Anglo-Norman policies resulted in the founding of a series of Royal burghs in the central
belt and lowlands, that urbanism in Scotland was clearly identifiable (Adams 1978). Royal
burghs engaged in a reciprocal relationship with the crown, for example, property was
granted under permanent feudal tenure to burgesses in return for rents and personal
services for the security of public peace. Similarly charters detailing privileges such as the
right to engage in overseas trade and hold markets were granted in return for payment of
customs duties (Adams 1978). Other burghs founded increasingly during the Medieval
period include Baronial (burghs erected after royal petition from local barons) and
Ecclesiastical burghs (burghs created by the church). These burghs shared many
similarities with Royal burghs such as having planned towns and structured societies but
differed in that overseas and regional domestic trade was prohibited. Unsurprisingly
therefore, Royal burghs endured as the dominant economic force throughout the medieval
period (Whyte 1997).

2.2.1.1 Morphology
Burghs with medieval origins are morphologically distinct. The fundamental component of
urban planning was the burgage plot. Burgage plots were units ownership manifested in a
predetermined strip of land usually fronting onto the main street (Adams 1978, Hall 2002)
(Figure 2, Figure 3). The dense nature of these long and narrow plots meant access for a
maximum number of plots holders to the central main street. This ‘High Street’ often acted
as the market place whereby traders operated from stalls and booths attached to the front
of houses and workshops (Coleman and Smith 2004). Consequently most burgh plans
exhibit the single street system, although some, mostly east coast burghs, comprise two
parallel streets (Figure 4). Medieval planning is still clearly identifiable in many burghs
today, especially those subject to minimal commercial development and infilling.

- 42 -
Figure 2: 16th century AD map of the Royal Burgh of Dumfries, Dumfries and Galloway
shown as a single street of dwellings with burgage plots, taken from the Pont Manuscripts
(Image © National Library of Scotland, Licensor www.scran.ac.uk)

Figure 3: Edinburgh c1460AD comprising Town (Royal Burgh, Burgh of Canongait), School
(Castle), and Cloister (Holyrood Abbey, St. Giles Cathedral) (Image © Patrick Geddes Centre
for Planning Studies, University of Edinburgh, Licensor www.scran.ac.uk)

- 43 -
Figure 4: Medieval Street Patterns: Newburgh, Fifeshire 1855-56AD example of single street
system and burgage plots and Crail, Fifeshire 1855AD example of parallel street system and
burgage plots (Image produced from the www.old-maps.co.uk service with permission of
Landmark Information Group Ltd. and Ordnance Survey)

2.2.2 Post-Medieval Urban Growth

The post-medieval period, spanning 1500 to c1800AD (Darvill 2002, Starn 2002) is
characterised by unprecedented urban growth emergent from a stagnant population at the
turn of the 16th century AD (De Vries 1984, Whyte 1999) (see Table 1, Table 2).
Consequently existing towns expanded, which often meant site intensification as opposed
to outward growth (Whyte 1997, 1999) and increasing numbers of new Royal and Baronial
burghs were created (Devine 2000). The resultant urban hierarchy was dominated by an
abundance of small towns with comparatively few large centres and even less middle size
settlements (Whyte and Whyte 1991).

Table 1: Urban Percentage of total population of Scotland, 1500-1800AD and percentage

increase in urban population from previous date (from De Vries 1984:39)

Date AD 1500 1550 1600 1650 1700 1750 1800

% Urban Population 1.6 1.4 3.0 3.5 5.3 9.2 17.3
% Increase * 0 130 17 51 124 132

Table 2: Percentage of total population in Scottish towns, mid 17th century to late 18th
century AD (from Whyte 1989:28)

Date AD 1639 1690s 1755 1790s

Capital 2.7 4.5 4.5 5.6
Other towns >10,000 3.5 2.7 4.4 10.8
5000-9999 3.3 1.6 3.2 3.2
2000-4999 2.2 3.1 4.2 6.4
1000-1999 ? 3.5 ? 6.3
500-999 ? ? ? 3.4
Total in towns >2,000 11.7 11.9 16.3 26.0

- 44 -
Sewerage and municipal rubbish collection was absent in most towns until the 19th
century AD, hence alternative systems of waste management must have been in place to
accommodate waste and any subsequent rises in volume. Given that changes in urban
soil properties are determined by accumulation of anthropogenic wastes, it is proposed
that investigation of soils in and near to Royal burghs will provide evidence for past
processes associated with waste management.

2.2.3 Waste Sources

Royal burghs would have undoubtedly produced considerable volumes of rubbish from a
variety of potential sources (Figure 5). The following section summarises the principal
contributors.

a b
Domestic Fuel
Plaster and Mortar

Animal Excreta

Food Stuffs
Urban Livestock

c d

Thatch

Burning

Industrial Activities Wattle and Daub

Figure 5: Graphical Summaries of Potential Waste Sources in the Medieval/Post-medieval

Urban Environment (a) Reconstruction of market scene at St. Johns Kirk, Perth (b)
Reconstruction of Upperkirkgate, Aberdeen backlands (c) Reconstruction of backland
manufacturing in Meal Vennel, Perth (d) Reconstruction of Upperkirkgate, Aberdeen
backlands (Images © Aberdeen City Council, Licensor www.scran.ac.uk)

2.2.3.1 Human and Animal Excreta

Absence of sanitation coupled with widespread holding of urban livestock meant that
human and animal excrement was a significant source of organic urban waste (Croly
2003, Ewan 1990). Comparable accounts from 16th century Prescott, England suggest

- 45 -
that filth was a pervasive feature of historic small towns, where human excrement is
documented on pavements, in the church and piled property windows (King 1992).
Likewise in Scottish burghs excreta accumulated on the burghs’ front streets forming so
called dunghills (Croly 2003, Mair 1988). Annoyances pertaining to excreta were confined
to issues of smell and access given that links between disease and health had not yet
been established. Archaeological evidence for excreta and its management is limited
hence analysis of certain soil properties should reveal the legacy of this material in Royal
burghs.

2.2.3.2 Building Materials

Buildings typically comprised combinations of disposable materials such as wattle and
daub, timber, clay, stone, turf, heather, straw and rushes (Reid 1909, Stones 1987)
Accounts of early 18th century AD domestic dwellings at Monymusk, Aberdeenshire
indicate the nature of materials used in domestic dwellings; reference is made to walls
comprising “wooden supports packed with rough stones and clay topped with 1 to 2 feet of
turf and the roof consisting of branches overlain with thinly pared turf which was in turn
overlain with heather or thatch” (Hamilton 1945). Buildings were repaired and rebuilt on a
regular basis, for example earth built houses in Kiltearn in the late 18th century AD are
cited as “being razed to the ground once in every 5 or 7 years, when they are added to the
dunghill” (Robertson 1791-99:289) likewise, the earthy parts of houses in Criech are
referred to as being added to the dunghill and being rebuilt again (Rainy 1791-99:375-
376).

2.2.3.3 Fuel
Fuel was fundamental to daily life given that most domestic activity such as heating,
lighting and cooking, and many industrial processes were dependant on fire. Peat, wood,
moss and coal were principal sources however, the precise nature and abundance used
varied in accordance with proximity to local resources, access rights, resource pressure
and, with reference to coal, market forces (Bruce 1791-99, Molleson 1791-99, Ogilvy
1791-99, Oram 2006). Constant burning ultimately produced considerable quantities of
waste products including ash, charcoal and charred material, collectively referred to as
fuel residues.

2.2.3.4 Industrial Wastes

Royal burghs supported a variety of industrial activities; some were common to all burghs
especially those related to agriculture and processing animals (Spearman 1988) however,
by the 16th century AD economic diversification resulted in burghs taking on distinct
occupational structures (Whyte 1997). Accordingly both similarities and differences in

- 46 -
types of industrial waste are expected between towns. It is likely that processing and
disposal of primary and secondary materials affected soil properties, however the nature
and location of impacts remains largely unknown (Spearman 1988). It is proposed that
basic distinctions between burghs with differing principal economies can be made, for
example it is expected that coastal burghs generated higher levels of marine waste such
as fine fish bones and shell compared to inland burghs.

2.2.4 Waste Management

Historical evidence for waste management is largely focussed on the occurrence of

dunghills. Dunghills comprising human, animal, domestic and industrial waste were
ubiquitous along the streets of historic small towns acting as storage points of urban
rubbish (Croly 2003, Dillon 1953). These waste heaps were important sources of fertiliser
and intrinsically held monetary value. Accordingly they were bought, sold and auctioned to
farmers, exchanged for food by the poor and even confiscated by burgh councils to
generate revenue or for application to burgh lands without payment (Cook 1867, Mair
1988, Smout 2000). Application of dunghills to burgh acres, agricultural land belonging
directly to the burgh, was therefore a common practice (McAlpine and Rolland 1791-99,
Smout 2000, Stones 1987, Whyte 1997, Yeoman 1995).

Burgage plots to the rear of burgesses’ houses flanking the central main street were a key
area in terms of managing urban waste (Adams 1978, Coleman and Smith 2004, Ottaway
1992) (Figure 6). This ‘Backlands’ zone was distinctive in Royal burghs in that it was used
for a variety of activities including domestic and industrial accommodation, holding
livestock, small scale cultivation and waste disposal (Ewan 1990). The precise nature and
intensity of backland use differed within and between burghs, responding in space and
time to economic and demographic factors (Coleman and Smith 2004, Hall 2002, Stones
1987, Whyte 1997). Nevertheless garden cultivation was an enduring backland activity;
hence it is likely that dunghills were applied to open burgage plots. Moreover, mixtures of
straw and dung from adjacent yards and byres also provided an immediate and
convenient source of fertiliser (Coleman and Smith 2004, Hall 2002, Oram 2006).

- 47 -
Figure 6: Mid 17th century AD view of Canongate, City of Edinburgh. Note differences in land
use within and between burgage plots (Image © Royal Commission on the Ancient and
Historical Monuments of Scotland, Licensor www.scran.ac.uk)

2.2.5 Summary

The generation of waste was an integral part of urban life during the medieval and post-
medieval periods. Limited aspects of the nature and management of waste sources can
be extrapolated from documentary evidence and archaeological survey. However,
considering that soil within and near Royal burghs presents a record of past waste
disposal practises, it is proposed that investigation of soil properties will enable systematic
insight into the legacy of waste both within and between historic small towns. The
following section introduces existing evidence for urban soil modification in Scottish
burghs.

- 48 -
2.3 Evidence for Past Soil Modification in Royal Burghs

Understanding soil modification in Royal burghs is limited to subsidiary information gained

from urban excavation and a small number of recent studies integrating archaeology and
soil science (Carter 2004, 2001, Davidson et al., 2006, Golding and Davidson 2005).
Nevertheless, collaborative evidence indicates the existence of distinct changes in certain
soil properties attributable to past waste amendments.

2.3.1 Deepened Soil Deposits

Deepened soil deposits are a distinctive feature of Royal burghs occurring in both large
(Bowler 2004, Carter 2004, 2001, Cachart 2000, Coleman and Smith 2004, Rains, Hall
1997) and small (Dercon et al., 2005, Davidson et al., 2006, Golding and Davidson 2005,
Hall and Bowler 1997, Hall et al., 1998, Lowe 2001, Spearman 1982) towns. Phases of
deepened topsoil are characteristically varied resulting in pronounced spatial
heterogeneity, for example accumulated deposits are identified across Holyrood,
Edinburgh ranging from 0.8 to 2.2m (Carter 2004). Similarly, Lowe (2001) notes
anthropogenic soils ranging in depth from 0.5 to 1.5m at Crail, Fife (Figure 7). The extent
of topsoil depth spatial variability is further demonstrated at Nairn, Nairnshire (Davidson et
al., 2006) and Pittenweem, Fife and Lauder, Borders (Golding and Davidson 2005) where
topsoil depth survey results indicate stark differences both within and between burgage
plots.

It is proposed that deepened soil deposits result from sustained past waste material
addition. Carter (2004) attributes deepened topsoil at Holyrood, Edinburgh to continual
application of mineral material derived from ash and turf as indicated by the presence of
burnt coal, burnt sedimentary fragments and fine burnt residues. Furthermore, considering
the occurrence of pottery, bone, mortar and shell, it is argued that periodic spreading of
midden material onto cultivated burgage plots was the dominant mechanism of
accumulation. Aside from Carter (2001, 2004) there have been limited attempts to
systematically understand how and why these deposits formed. Accordingly, this study will
investigate the nature of deepened soil deposits spatially and through physical and
chemical characterisation.

- 49 -
 a) 1-1.8m

b) 0.4-0.8m

c) 1-1.5m

Figure 7: Crail from the air (a) Deepened anthropogenic sediments, west end of Nethergate,
(b) Shallower sediments, east end of Nethergate, (c) Deep soils with medieval and mixed
medieval and post-medieval assemblages, attributed to medieval cultivation of the burghs
infield and/or off loading ballast by Dutch herring traders (Lowe CE, 2001, Image © Colin J M
Martin, Licensor www.scran.ac.uk)

2.3.2 Backland Garden Soils

The term ‘garden soil’ refers to deep, homogenous, highly mixed soil resulting from
intensive cultivation (Carter 2001, Wordsworth and Clark 1997). Although garden soils are
recognised in many burghs, for example, Forfar, Angus (Spearman 1982), Elgin, Moray
(Hall et al., 1998) and North Berwick, East Lothian (Hall and Bowler 1997), they have
received most attention in St Andrews, Fife. It is accepted that garden soils represent
deliberate and sustained improvement in the backlands of St Andrews (Cachart 2000)
however, significant conjecture exists regarding processes leading to their accumulation.

Clark (1997) proposes that these deposits imply a major change in post-medieval land
use whereby backland activity ceased, land was turned over to cultivation and accordingly
material was imported for horticulture, hence giving rise to the widely accepted ‘Imported
Garden Soil Model’. Carter (2001) suggests an alternative ‘Occupation Deposit Model’
attributing accumulation of mineral sediments to continual replacement of building
materials associated with intensive occupation of the backlands. Considering that
differentiation between topsoil introduced for cultivation and turf used in building
construction is problematical, resolution of this issue seems unlikely. Moreover, given that

- 50 -
both scenarios could have produced similar deposits, a potential third hybrid model of the
two theories is plausible.

2.3.3 Stratigraphy and Dating

Garden soils comprising two distinct layers, a lower brown layer attributable to the
medieval period and an upper darker layer to the post-medieval period, have been noted
in several burghs (Cachart 2000, Carter 2001, Hall et al., 1998, Spearman 1982). Aside
from differences in colour, differentiation of soil properties between these layers has not
been elucidated. Accordingly this project will investigate differences in layer composition.
Despite this elementary distinction within some garden soils, urban soils associated with
Royal burghs do not exhibit definite stratification rendering relative and absolute dating
ineffective.

2.3.3.1 Relative Dating

Analysis of artefacts from garden soils indicates that deposit accumulation occurred
throughout the medieval and post-medieval periods, for instance, pottery dating to 14th-
15th centuries AD and clay pipe fragments typical of 17th-18th centuries AD have been
identified at the Market Street excavation, St. Andrews (Hall 1997). Similarly, at the
Rumford and Westgate sites in Crail, mixed assemblages of medieval and 17th century AD
pottery sherds have been recovered in association with deep soil, bone and glass (Lowe
2001). Nonetheless, reliable determination of deposit stratification is problematical, for
example although Ross and Clark (1997) propose successive phases of cultivation from
14-15th and 16-17th centuries AD at the Cinema House site, St. Andrews, a lack of
dateable material from primary contexts impedes interpretation. In addition to residual
finds, modern activities also hinder chronological establishment. Excavation at Castle
Street, St. Andrews revealed intrusions resulting from modern services coupled with small
quantities of modern pottery reworked into phase II soil deposits (Cox 1997). Moreover, it
is recognised that dates of artefact manufacture and deposition are separate entities. Hall
(1997) suggests that although phase V homogenous garden soil deposits at Market
Street, St. Andrews are dominated by 14-15th AD century pottery sherds, it is possible that
such artefacts may not relate to earlier occupation of the site if soil was imported and
dumped.

- 51 -
2.3.3.2 Absolute Dating
Utilisation of techniques such as radiocarbon (14C) and optically stimulated luminescence
(OSL) is problematic for chronological dating of soils formed over the medieval and post-
medieval periods. This is especially true for urban soils in Royal burghs where it is argued
that fundamental assumptions associated with these methods cannot be met. Primarily
radiocarbon dating relies on intact stratigraphy with minimal reworking and bioturbation
(Rapp and Hill 1998), yet presence of artefacts in residual contexts, modern intrusions
and evidence for past cultivation in garden soils (Hall 1997, Ross and Clark 1997) infer
high uncertainty as to whether this prerequisite can be met. Age contamination within
these soils is also a possible problem especially from fossil carbon sources such as
limestone associated with building mortar, paints and washes (Renfrew and Bahn 1997).
Moreover, turf material associated with building construction and fuel has an existing
radiocarbon age, hence soil would appear too old (Bokhorst et al., 2005). Radiocarbon
dating is further dismissed as a viable technique for dating post-medieval deposits given
its upper dating edge of 300 years BP (Taylor 2001).

Dating post-medieval deposits is possible using OSL dating which spans 100 to
100-200 000 years BP (Grϋn 2001). This method relies on the OSL signal of mineral
grains being reset to zero upon exposure to daylight prior to deposition. However
incorporation of material such as paired turf from building construction whose signal is not
necessarily reset at deposition may affect sample age, though Bokhorst et al., (2005)
suggest this effect is minor. The main concern using OSL to date post-medieval soils is
low precision (Cluett 2007, Feathers 2003, Sommerville et al., 2001). This is not
necessarily an issue when constructing age models over millennia however, dates with
errors of ± 100-300 years would result in considerable overlap over this shorter time
period. Aside from problems associated with absolute dating techniques, it is argued that
dating urban soils associated with Royal burghs is ultimately meaningless considering the
mixed and re-worked nature of deposits. Furthermore, given that Royal burghs originated
and developed throughout the medieval and post-medieval periods, it seems logical that
impacts on soil also date to those periods.

- 52 -
2.3.4 Hinterland

Deepened soil deposits are not exclusive to the backlands as demonstrated at Nairn,
Nairnshire (Davidson et al., 2006, Dercon et al., 2005), Pittenweem, Fife and Lauder,
Borders (Golding and Davidson 2005). Evidence from deepened phases at Nairn
indicates that these deposits are testament to sustained urban waste material addition.
Dercon et al., (2005) attribute finer material in the A horizon to mineral accumulation
through the addition of turves used in buildings, sand added to burgh dunghills and sand
in association with seaweed application. Deepened topsoil in Nairn’s hinterland is in
agreement with the location of burgh acres cultivated throughout the post-medieval
period, indicating that urban waste was intentionally used to enhance soil quality
(Davidson et al., 2006) (Figure 8) . Moreover, analysis of total phosphorus concentrations
revealed a peak in the AP3 horizon corresponding to the lowest old cultivated layer.
Despite these initial findings little is known about the occurrence, location and physical
and chemical legacy of deepened hinterland soil.

Figure 8: Depth of topsoil in and near the Burgh of Nairn (left), distribution of rigs in burgh
acres 1790AD (right) (Davidson et al., 2006: 780)

- 53 -
2.3.5 Summary

Results from archaeological excavation and geoarchaeological studies suggest that

management of urban waste has had long lasting impacts on soil both within and near to
Royal burghs. However, no attempt has been made to systematically identify, characterise
and compare deposits arising from waste material addition. Consequently the following
chapter provides a research framework detailing the project aims and objectives
addressing these knowledge gaps.

- 54 -
 3 Investigating Urban Anthrosols in Historic Small
Towns: A Research Framework

This chapter presents the adopted research framework for investigating urban anthrosols
in small historic towns and outlines the project’s aims, objectives and hypotheses. The
research framework is centred on a refuse flow model developed through a review of
archaeological concepts of rubbish, contemporary and historical perceptions of rural-
urban interaction and previously identified factors affecting material procurement,
management and disposal in Royal burghs.

3.1 Archaeological Concepts of Rubbish

3.1.1 Defining Rubbish

Rubbish can be defined socially as opposed to physically (Thompson 1979, 2003). This
idea is manifested in Rubbish Theory (Figure 9) where it is proposed that everything we
deal with can be divided into three categories, transient (things with an estimated lifetime
within which they continually lose value to the point where they have none), durable
(things that have an unlimited lifetime) and rubbish (all items that are neither transient nor
durable). To accommodate constantly shifting social definitions of rubbish, transfers
between categories are possible, for example from transient to rubbish and rubbish to
durable. Rubbish Theory incorporates social and economic fluidity hence is a useful tool
for conceptualising the status of material culture within past societies. Essentially
understanding what constituted waste in particular cultures and periods is dependant on
the ability to deduce past social perceptions however caution should be taken to avoid
superimposing contemporary values.

Transient Durable
Value decreases with time Value increases with time

Rubbish
No value, no time

Figure 9: ‘Rubbish Theory’ model showing cultural categories of objects and possible
transfers between them (Thompson 2003: 322, 1979: 10)

- 55 -
3.1.2 Garbology

The study of rubbish is a recognised archaeological sub-discipline. More specifically it can

be argued that garbology “the study of a community or culture analysing its refuse”
(Soanes and Stevenson 2006) is a fundamental approach in understanding the nature
and formation of archaeological records. Garbology was first pioneered in the 1970s by
William Rathje (1974) who argued that present societies should be systematically
investigated using archaeological enquiry to advance current understanding of past
cultures. This was in contradiction to traditional archaeological theory where antiquity is
understood using evidence from the past or correlates to the present. The University of
Arizona Garbage Project (Rathje 1974) was the first systematic attempt to study
contemporary discard behaviour. One of the key trends identified was that high income
households are associated with low rubbish generation. This finding challenges
established theories which assume conspicuous consumption in ancient societies is
related to social elites. Moreover, it was found that different racial and economic groups
exhibited different discard behaviour in terms of what and how much they threw away.

This novel approach to understanding waste has been largely adopted by social
anthropologists (for example, Edwards 2004, Harpet 2003, Pessel 2006, Rathje and
Murphy 2001b). Within archaeology the impact of garbology has been less pronounced.
Nevertheless, at a generic level its principal theories provoke enquiry into notions of race,
place and status and their role in waste management and discard behaviour in past
societies.

3.1.3 Classification of Rubbish Deposits

Concentrations of rubbish discovered on archaeological sites are usually referred to as

midden deposits regardless of their spatial characteristics and content. However, it is
argued that a basic distinction between middens and ‘refuse rich’ deposits exists. Middens
are spatially discrete features and have identifiable modes of construction, constituents,
resource roles and spatial associations (Needham and Spence 1997). In contrast,
indistinct areas of general waste accumulation can be described as “deposits rich in
refuse”. This basic distinction alludes to differences in discard behaviour.

A systematic attempt to classify rubbish disposal according to artefact use and disposal
histories is proposed by Schiffer (1972:156) in response to the consensus in traditional
archaeology that “spatial patterning of archaeological remains reflects the spatial
patterning of past activities”. Three categories of rubbish classification are proposed,

- 56 -
primary refuse (rubbish discarded at its location of use), secondary refuse (rubbish
discarded away from its location of use) and de facto refuse (material abandoned at the
use location but still having a perceived re-use value, for example large household items
left upon site abandonment) (Figure 10). In addition, Haydon and Cannon (1983) propose
a fourth addition to this system, provisional refuse (stored refuse having a perceived re-
use value) which differs from de facto refuse in that it infers intentional storage rather than
intentional abandonment. These categories are useful for understanding the social
perception, use and discard behaviour associated with material culture. Moreover, this
model provides a framework for investigating life cycle histories of material omitted from
the archaeological record.

System under analysis Systemic Archaeological

Context Context
Opportunity for storage and/or
transport

Discard Primary

Procurement Manufacture Use

Transport Secondary

de facto

Figure 10: Flow model showing artefact histories and proposed rubbish categories (Schiffer
1972:162)

3.1.4 Refuse Cycle Flow Models

It is argued that the key to understanding past societies is through the ability to relate
archaeological finds to systemic structures (Schiffer 1976). This has been realised in
refuse cycle flow models, where ‘refuse cycle’ is defined as a “conceptualisation of refuse
movement and management” (Needham and Spence 1997:77). These models make a
basic distinction between the systemic context (life cycles) and archaeological context
(incorporation and burial) (Figure 11 and Figure 12). Material flow moves generally left to
right from system processes to the archaeological record. However re-routing is possible,
for example recycling which refers to routing of an element at the completion of use to the
manufacture process of a different element (Schiffer 1972).

Alternatively, Rathje and Murphy (2001a) propose that rubbish management in antiquity
can be accounted for by four universal methods; dumping, burning, turning a material into

- 57 -
something that can be useful and source reduction (minimising the volume of material
goods that comes into existence in the first place). It is argued that all civilisations have
used these methods simultaneously to some extent. These proposals do not discount
refuse cycle models, conversely they reiterate fundamental concepts such as ‘dumping’
which is akin to discard processes and ‘turning material into something useful’ which is
analogous to re-use flows.

Systemic Archaeological
Context Context

Procurement Manufacture Use Discard Decay

Abandonment
Reuse Loss

Recycling

Reclamation

Figure 11: Flow model of adapted artefact life histories (LaMotta and Schiffer 2001:20)

3.1.5 Summary

Insight into the social and economic environment of historic towns is essential for
understanding the nature and management of waste materials. Schematic models are
identified as a useful tool for conceptualising and visualising material flow in small historic
towns, although it is noted that such models are not prescriptive representations of reality.

- 58 -
 Buried INCORPORATION/
BURIAL
In-deposit transformations

Re-use
(modified
USE CYCLE(S) or not)

Retrieval and Gross

Recycling disturbance
of deposit

Breakage
Disfunction
Incorporation ‘ultimate’ &
burial Archaeology/
Recovery
Discarding Visible contexts In-deposit transformations
Abandon ‘storage’
(conscious) abandonment
(conscious/ unconscious)
Manufacture Original
Procurement use (s) Remains on
surface, no Eventual
incorporation decay or
Loss reduction
on-site

REFUSE CYCLE(S)
Concealed/
Stowed
(not buried)

LIFE CYCLES(S)

Figure 12: Refuse cycle flow model for possible stages in artefact histories (Needham and Spence 1997:78)

- 59 -
3.2 The Nature and Significance of Rural-Urban Interaction

In developing a research framework for investigating urban anthrosols in small historic

towns it is important to consider what is meant by the terms rural and urban with specific
reference to their relationship with each other. Determining the nature and significance
of urban-rural interaction is vital in understanding resource flows, particularly for waste
materials in historic towns. The following section looks at key concepts associated with
rural-urban interaction derived from a brief review of contemporary and historical studies.

3.2.1 Contemporary Research of Rural-Urban interaction

Whilst the terms ‘rural’ and ‘urban’ serve as meaningful concepts for descriptive
purposes, their value in understanding complex spatial, sectoral and economic trends is
subject to debate (Lin 2001, Nivalainen and Schimdt-Thome 2003). Their comparability
on an international level is also questioned given that nations discriminate between rural
and urban according to varying population thresholds and densities (Tacoli 1998). It is
proposed that investigating the nature of rural-urban interaction is a more objective
approach than analysis of the spatial categories themselves (Gould 1987). Implicit within
rural-urban interaction is the notion of a spatial rural-urban continuum (Aguilar and Ward
2003). It is therefore argued that the idea of a rural-urban dichotomy in contemporary
and historic towns is a false divide.

3.2.1.1 Rural-Urban Fluxes

Fluxes of people, commodities, knowledge, assets and social transactions form the
basis of rural-urban interaction (Kaida and Maharjan 1990, Lin 2001). Determining the
nature and significance of these linkages is vital in evaluating society and economy in
towns both past and present. Current research focuses largely on population fluxes in
developing countries with specific reference to migration scale, migration direction,
gender, diaspora and age selectivity (Fekade 1995, Gould 1987, Lin 2002, Tacoli 1998,
Tanner 2003). It is argued that understanding people as agents of change in rural-urban
interaction is important considering their decisions economically, politically, socially and
culturally define the urban environment (Coppack 1998). Reconstruction of population
fluxes is problematic for historic towns considering the limited nature of past
documentation in terms of the number of records which diminishes with time and the
recognition that surviving sources do not necessarily record information relevant to
certain study questions.

- 60 -
3.2.1.2 Rural-Urban Interaction Zones
Physical manifestations of rural-urban interaction are noted in contemporary cities.
Some examples are discussed hereafter for a variety of urban scales ranging from mega
cities to local urban centres, whilst scale differs it should be noted that similarities exist in
the nature and extent of rural-urban interactions. Lin (2001) delineates a zone of rural-
urban interaction adjacent to and in between metropolitan centres in the Pearl River
delta, China. This area is characterised by functional duality comprising a mixture of
agriculture and industrial activities. Moreover Aguilar and Ward (2003) note two distinct
rural-urban interaction zones in Mexico City, Mexico collectively termed the peri-urban
hinterland. The inner peri-urban zone is functionally integrated with the city core and is
typified by a mixture of urban and rural land uses. In comparison the outer peri-urban
zone consists of more remote settlements outwith the metropolitan area but crucially
they remain contiguous to it. Urban-rural interaction zones are not specific to developing
or industrialising cities. Nivalainen and Schmidt-Thome (2003) estimate that up to one
quarter of Finland’s population live in distinct rural-urban interaction zones which are
defined as neighbourhoods of urban centres consisting of a high proportion of
commuters. Multi-agent system models of land-use/cover change (MAS/LUCC) are
identified as important tools for investigating complex spatial relationships associated
with rural-urban interaction. MAS/LUCC models are unique in that they integrate aspects
of the physical landscape with political, economic and cultural decision making
processes. Accordingly they can be applied to a variety of scenarios ranging from
natural resource management and agriculture to modern and historical settlements (for a
comprehensive review of MAS/LUCC models refer to Parker et al., 2003).

3.2.1.3 Sectoral Interaction

Principles of rural-urban interaction are also manifested in sectoral interactions. Sectoral
interactions are classed as rural activities taking place in urban zones and urban
activities occurring in rural zones (Tacoli 1998). One of the most prevalent cases of
urban-rural sectoral interaction is the recent increase of urban agriculture in developing
cities as a response to escalating poverty and food prices (Tacoli 1998). Likewise some
rural areas are becoming more urban in nature, for example Tanner (2003) notes how
competition among municipalities to present a more ‘modern’ and urban way of life has
resulted in infrastructural developments such as water and electricity in the north-
western province of Jujuy, Argentina.

- 61 -
3.2.2 Rural-Urban Interaction in Historic Towns

3.2.2.1 Hinterland Resource Flows

Investigation of rural-urban fluxes in historic towns is limited to analysis of surviving
physical remains. Accordingly most studies focus on flows of natural resources,
foodstuffs and manufactured commodities between hinterland and town, for example
bone evidence from medieval and post medieval Oslo, Norway suggests that cod was
transported as cargo stock from the north whereas fish like haddock, ling, eel and flatfish
were caught locally and brought in fresh (Schia 1994). Similarly archaeological and
documentary evidence from medieval and post medieval Oxford indicates that although
dairy herds were reared on the town’s common land, some cattle came from as far away
as Berkshire as a result of cattle droving from Wales and the west of England (Wilson
1994).

The case study of Oslo also presents evidence for changes in hinterland resource
supply and demand. Chemical analysis of 11-16th century AD iron suggests that
hinterland production sites fell from 4 to 1 during this period. This can be explained
through a decline in local bog iron sites coupled with increased dependence on Swedish
mountain iron (Schia 1994). Moreover, it is recognised that reconstructing patterns of
hinterland exploitation is useful for investigating society and economy of historic towns.
This is demonstrated at Medieval Lϋbeck, Germany where shifts from local resource
exploitation and agriculture to reliance on imported grain is explained by increased
pressure on available space caused by accelerating urbanisation (van Haaster 1994).

3.2.2.2 Evidence for Rural-Urban Zones and Sectoral Interactions

Given the importance of rural-urban resource fluxes the existence of historic rural-urban
interaction zones is not surprising, for example Wilson (1994) notes the importance of
animal rearing and marketing at sub-urban and sub-rural zones within Medieval and
Post-medieval Oxford. Moreover, Ciezar et al., (1994) identify a suburbanus on the
periphery of Roman and Medieval Paris. It is suggested that this zone represents an
urban food-belt which emerged as a response to declining or unstable rural food
networks. This theory is supported by archaeological excavation which revealed
habitation deposits, animal penning, animal bones and high soil phosphates.

Regarding sectoral interactions, organic soil rich in phytoliths at the Roman Deansway
Site, Worcester provide evidence for open air penning of herbivores and associated
burning of animal manure and trampled soil from animal pens (Macphail 1994).

- 62 -
Moreover, high concentrations of faecal material and coprolites indicative of systematic
dung spreading at 1st century AD Whittington Avenue Site, London infer that urban
horticulture was an important practise in Romano-British towns (Macphail 1994). Soil
management associated with urban horticulture is similarly recognised in late Medieval
Lϋbeck, Germany where urban gardens were being intensively fertilised through dung
application from as early as the 14th century AD (van Haaster 1994).

3.2.3 Summary

Evidence for resource fluxes, hinterland exploitation, suburbs and urban horticulture
supports the notion of a rural-urban continuum in historic towns. Moreover considering
that modern parallels are identified in contemporary towns and cities, it is argued that
such rural-urban interactions are inherent in the development and continued existence of
urban environments. The importance of appropriate nomenclature and provision for
rural-urban interactions in developing a waste flow model for small historic towns is
therefore recognised.

- 63 -
3.3 Research Framework

3.3.1 Modelling Waste in Small Historic Towns

The following section presents a flow model of waste movement in small historic towns
(Figure 13) and discusses its component features. The purpose of this model is to act as
a research framework for understanding management of different types of waste and to
identify variables affecting waste procurement, movement and disposal, within and
between historic small towns.

Figure 13: Flow model of waste movement in small historic towns

3.3.1.1 Town and Hinterland

In the first instance a distinction is made between town and hinterland. The town zone
corresponds to the built environment of historic towns hence is typically urban. In
addition this zone encompasses areas characterised by rural-urban interactions and
sectoral interactions accounting for activities traditionally classed as rural such as
cultivation and animal rearing. The hinterland is a theoretical area of resource
exploitation characterised by both rural-urban linkages and urban-urban linkages. The
hinterland can therefore be defined at different scales ranging from national and regional
to land immediately adjacent to the town. The town and hinterland are linked by rural-
urban fluxes of resources, goods and waste materials, hence are not mutually exclusive
zones. The interdependence of town and hinterland is represented in the model by
colour gradation as opposed to a definite boundary. Provision is therefore made for

- 64 -
shifting definitions of urban and rural, allowing reassessment of the nature and extent of
town-hinterland interactions.

3.3.1.2 Resource Flow

Acquisition represents movement of materials from the hinterland to town. These
materials range from locally available natural resources to manufactured trade goods.
The precise nature and abundance of incoming materials is largely dependant on
geographical location, proximity to local natural resources and industrial demand. These
variables are often inextricably linked, for example fishing and trade being the dominant
industries of coastal burghs. To a lesser extent materials can also be acquired from the
town itself including animal meat, vegetables and dairy products. Acquired resources are
incorporated into a cycle of use and re-use, for instance animal carcasses supply
primary material for fleshers, tanners, leather workers, horners and candle makers. This
is akin to LaMotta and Schiffer’s (2001) systemic context which features material reuse
and recycling as key flows.

3.3.1.3 Waste Flow and Deposition

Materials with no remaining economic or cultural value are classed as waste (Thompson
1979, 2003). Accordingly such materials transition to a cycle of waste management
which is characterised by processes of storage and redistribution. Storage refers to
temporarily holding waste at a given location, such as the use of dunghills. In contrast
redistribution is concerned with material movement. Redistribution of waste can occur
within the town itself, for instance to the backlands, and also in the hinterland.

Deposition of waste falls into two categories, point and diffuse. This distinction
discriminates between spatially discrete features such as middens and, deposits rich in
refuse. Potentially these practises can occur in both the town and hinterland, for
example town waste can be spread in town gardens and the burgh acres. Upon
deposition, waste is incorporated in the soil matrix. It is suggested that the impact and
magnitude of waste disposal diminishes with distance from the town core. The
hypothesised distance decay effect of the impact of waste disposal on soil properties is
reflected in graduation of colour from dark brown (town) to light brown (hinterland). After
deposition, continual reworking of waste is expected in association with system loss of
more mobile elements.

- 65 -
3.3.1.4 Model Limitations
It is acknowledged that some scenarios do not readily follow the model, for example
cattle have multiple roles such as producing milk, meat and as industrial raw materials.
However it can be debated whether cattle dung is a resource or waste material given
that it follows processes of storage, movement and deposition associated with waste
material flows. Conversely dung is known to have held an economic value hence its
place arguably belongs in the cycle of resource use and acquisition. Accordingly, it is
expected that model modifications will be proposed upon discussion of the project
findings.

3.3.2 Aim and Objectives

The project aim, objectives and hypotheses are stated below and are further presented
in a hierarchical flow model (Figure 14). The proposed hypotheses are broad and relate
to expected study outcomes.

3.3.2.1 Aim
Urban Anthrosols are distinctive owing to their location and that their properties are
modified through waste material amendments. Much consideration has been given to
how changes in soil physical and chemical properties reflect current human influence
(Chapter 2) however, their role as a historical archive is only just being recognised.
Recent studies indicate that urban soils reflect differences in past soil inputs over space,
time and location. There has been no systematic attempt to account for the diversity and
distribution of urban anthrosols either within or between historic towns. The aim of the
study therefore is to characterise and understand the modes of urban anthrosol
formation in historic small towns.

3.3.2.2 Objective 1
Discrete patterns in soil physical and chemical properties are identified both within and
adjacent to historic towns. This is especially true for Scottish royal burghs where
deepened phases of topsoil are identified in the medieval core and town hinterland. It is
proposed that such trends are the legacy of past waste material inputs. The first study
objective therefore is to establish the nature and diversity of urban anthrosols in and
near to historic small towns. Accordingly it is hypothesised that differences in soil
characteristics exist across burghs and their hinterland and that soil properties of urban
anthrosols in and near to burghs are determined by past waste material amendments.

- 66 -
3.3.2.3 Objective 2
Although trends in certain soil properties are replicated across historic towns, for
example dark earth deposits and garden soils, differences both within and between
historic towns are expected on account of past geographical, functional and economic
diversity. Consequently, the second study objective is to characterise and account for
the multiplicity of urban anthrosols in and near to historic small towns. It is expected that
different types of urban anthrosol exist, that distributions of different urban anthrosols are
related to functional zones and distributions of urban anthrosols are related to past burgh
functions and economies.

3.3.2.4 Objective 3
Urban anthrosols present a historical archive of human activities in towns. It is argued
that deposits formed through sustained application of waste materials reflect past waste
management practises. The nature and location of such practises are influenced by
burgh function and economy, and type, quantity and perceived utility of waste materials.
Consequently the third study objective is to elucidate the processes associated with
waste management and disposal in historic small towns. Additionally it is proposed that
processes of waste disposal can be deduced from urban anthrosol properties and that
processes of waste disposal were associated with economic and cultural functions of the
royal burgh.

- 67 -
 Aim
Characterise and understand the modes of urban anthrosol formation in historic small towns

Objective 1 Objective 2 Objective 3

Establish the nature and diversity of Characterise and account for the Elucidate the processes associated
urban anthrosols in and near to multiplicity of urban anthrosols in and with waste management and disposal
historic small towns near to historic small towns in historic small towns

Hypotheses Hypotheses Hypotheses

1.1 Differences in soil characteristics 2.1 Different types of urban anthrosol exist 3.1 Processes of waste disposal can be
exist across burghs and their deduced from urban anthrosol properties
hinterland 2.2 Distributions of different urban
anthrosols are related to functional zones 3.2 Processes of waste disposal were
1.2 Soil properties of urban associated with economic and cultural
anthrosols in and near to burghs are 2.3 Distributions of urban anthrosols are functions of burghs
determined by past waste material related to burgh function and economy
amendments

Figure 14: Project aim, objectives and hypotheses

- 68 -
 4 Materials and Methods

This chapter outlines the materials and methods used in this study with specific
reference to site selection, field work, laboratory analyses, soil micromorphology and
data analyses.

4.1 Site Selection

To facilitate selection of suitable study sites a series of practical and theoretical criteria
was developed. It was deemed important that potential burghs had limited modern urban
development infill and peripheral expansion given the need for abundant and relatively
unrestricted sampling opportunities both within and adjacent to medieval town cores.
Moreover, to enable comparison of past functional zones preference was given to sites
with characteristic medieval street layouts such as single street and parallel street
systems. Geographical and functional diversity were also identified as important factors,
for example it was desirable that towns had differing trading patterns, local resources
and principal economies. Twelve potential study sites were identified through desk-
based studies and a series of preliminary field visits (Figure 15).

Figure 15: Location of preliminary field visits undertaken between 12/11/04 and 04/01/05

- 69 -
Burghs with low sampling potential were discounted (Auchtermuchty, Ayr, Brechin, Crail,
Kirriemuir and Newburgh). Burghs without characteristic medieval street morphologies
were also excluded (Culross and Falkland). Of the remaining four burghs, Lauder,
Pittenweem and Wigtown were identified as the most geographically and functionally
diverse, hence New Galloway was eliminated from subsequent selection.

4.1.1 Study Sites

4.1.1.1 Lauder
The royal burgh of Lauder, Scottish Borders, lies immediately east of the A68 which runs
between Edinburgh and the English border. The town is located in a valley overlooked
by the Lammermuir hills and is situated on the right bank of the River Leader (Figure
16). The soil mapped at Lauder is a freely drained brown forest soil with low base status
belonging to the Lauder (LA) soil series. This series is part of the Lauder (LA)
association derived from conglomerates and sandstones of Upper Old Red Sandstone
(O.R.S) age (Soil Survey of Scotland 1959). In the 2001 census the population of Lauder
was recorded as 1087. The largest two employment sectors identified were retail and
wholesale, and health and social work accounting for 26% of employed persons,
followed by business services (12%), education (11%) and public administration (11%)
(General Register Office for Scotland 2007).

0.4 0 0.4 0.8 Kilometers

Figure 16: Location and situation of Lauder, Scottish Borders

- 70 -
The origins of Lauder are unclear although it is suggested that its foundation lay in the
12th century AD when Hugh de Moreville built a castle, church and mill complex for his
followers in the barony of Lauderdale (Simpson and Stevenson 1980). The first
documented existence of Lauder was in the late 13th century AD when Bishop William
Lamberton of St. Andrews noted ‘duo burgagia in villa de Lawder’. This recognition
indicates the presence of a settlement however; it does not necessarily confirm burghal
status at that time (Pryde 1965). Lauder was officially elevated to burgh status with
baronial privileges in the 14th century (1324x1325AD) and was subsequently granted
royal status in 1455AD after being forfeited by the Douglas family to the crown.

Throughout the medieval period the principal economy of Lauder was merchandising
and trade. Prosperity associated with these activities was largely dependant on external
factors, for example a 16th century AD burgh charter indicates the absence of foreign
trade on account of war, assaults and fire (Turner-Simpson and Stevenson 1980). By the
late 18th century AD Lauder held six yearly fairs and two weekly markets. Moreover,
growing political stability resulted in an increase in the town’s population, foreign trade
and traffic. In addition to trade, milling was a significant economic component in the
development of Lauder. This is confirmed in various 16th century AD references which
cite the existence of a corn mill, a waulk mill, a common Burn Mill belonging to town and
a Mill belonging to a Lauder family.

4.1.1.2 Pittenweem
The royal burgh of Pittenweem, Fife, lies on the A917 coastal road which runs between
Elie and St. Andrews on the East Neuk of Fife. The town is located 2 km south west of
Anstruther and is situated on a raised beach overlooking the northern North Sea (Figure
17). The soil mapped at Pittenweem is an imperfectly drained brown forest soil
belonging to the Quivox (QX) soil series. This series is part of the Dreghorn association
characterised by raised beach deposits derived mainly from Carboniferous sediments
(Soil Survey of Scotland 1975). In the 2001 census the population of Pittenweem was
recorded as 1747. The principal employment sectors identified were retail and wholesale
and health and social work accounting for 16% and 12% of employed persons
respectively. Other significant sectors included hotels and restaurants (10%), real estate
(10%) and education (10%) (General Register Office for Scotland 2007).

- 71 -
 0.4 0 0.4 0.8 Kilometers

Figure 17: Location and situation of Pittenweem, Fife

The origins of Pittenweem can be traced to the 12th century AD when its lands were
bestowed to monks resident on the Isle of May. This resulted in the transfer of the
Augustinian priory from the Isle of May to the site of the current parish church (Simpson
AT and Stevenson S 1981a). The development of the site over the following centuries is
uncertain, however, a charter dating to 1526 AD indicates that by the late 15th century
Pittenweem was a legitimate baronial burgh. In 1541 AD Pittenweem was subsequently
changed ‘in liberum burgum regalen’ to royal burgh status (Pryde 1965).

The principal economies of Pittenweem were fishing and trade, for example in 1537 AD
Pittenweem paid £205 16s duty on hides, cod, herring, malt and English goods in
contrast to nearby Kinghorn which paid £6 7s 4d (Simpson and Stevenson 1981a). The
financial dominance of Pittenweem in the 16th century AD was further demonstrated in
1542 AD when Pittenweem exported £432 19s in hides, salmon, cod, coal and herring
compared to £139 5s 4d by Perth. Civil war and plague during the 17th century AD had a
negative impact on Pittenweem’s population which in turn led to a decline in seafaring
and export revenues. Consequently a ruined pier and harbour, empty houses and an
absence of foreign trade were noted (Simpson and Stevenson 1981a). Nevertheless, by
the mid 18th century AD the town had started to recover. This was reflected in renewed
export of salt fish, herrings and malt to the Baltic, Mediterranean and Low Countries
(Horsburgh 1865).

- 72 -
4.1.1.3 Wigtown
The royal burgh of Wigtown, Dumfries and Galloway, lies on the A714 which connects
Girvan in west Scotland to the Machars peninsula in south west Scotland. The town is
located 11km south of Newton Stewart and is situated overlooking Wigtown Bay on the
Solway Firth (Figure 18). The soil mapped at Wigtown is a freely drained brown forest
soil belonging to the Linhope (LP1) soil series. This series is part of the Ettrick
association derived from Ordovician and Silurian greywackes and shales and their
associated drifts (Soil Survey of Scotland 1971). In the 2001 census the population of
Wigtown was recorded as 987. The principal employment sectors identified were retail
and wholesale (21%), construction (14%) and manufacturing (14%). Other important
sectors included health and social work, and agriculture, hunting and forestry accounting
for 17% of employed persons collectively (General Register Office for Scotland 2007).

Although it is thought that Wigtown existed as a burgh during the 12th century AD, it was
not until 1292AD that its status as a royal burgh was recognised by charter. However,
Wigtown’s position as a royal burgh was short lived. In 1341AD it was granted to Sir
Malcolm Fleming due to unpaid revenues (Pryde 1965). Ultimately Wigtown was
restored to its former royal status in 1455AD after it was forfeited by the Douglas family
to the crown.

0.4 0 0.4 0.8 Kilometers

Figure 18: Location and situation of Wigtown, Dumfries and Galloway

- 73 -
The principal economies of Wigtown were trade and agriculture. During the 15th century
AD Wigtown was one of the five principal west coast ports of Scotland including Ayr,
Irvine, Dumbarton and Kirkcudbright. Throughout the 16th and 17th centuries AD trade
steadily declined to the extent that in 1692AD foreign trade was absent and inland trade
was limited. The main reason for this demise was that Wigtown did not participate in the
herring trade which became a major export of rival west coat ports (Simpson and
Stevenson 1981b). Nevertheless, Wigtown remained a dominant agricultural centre. This
is attested to in the Old Statistical Account for Scotland where considerable
documentation is devoted to land, farming practises and agricultural revolution at
Wigtown (Duncan 1791-99). Towards the latter 18th century AD an increase in
population associated with the onset of industrialisation and Irish immigration is noted
along with the reinstatement of Wigtown’s markets (Simpson and Stevenson 1981b).

- 74 -
4.2 Soil Survey

Special-purpose soil survey was identified as a useful tool for systematically

investigating changes in soil properties across small historic towns (Bui 2004, Dent and
Young 1981, McRae 1988). The survey objective was to determine variation in soil
physical and chemical properties in historic urban landscapes. Furthermore, the intent of
soil mapping was to reveal differences in specific soil properties indicative of past
anthropogenic activities. The following section provides an overview of survey design
and subsequent field methods.

4.2.1 Survey Design

In the first instance the location and extent of the survey area was determined for
Lauder, Pittenweem and Wigtown (Figure 19, Figure 20 and Figure 21). The siting of
each survey area represents the most effective compromise between three factors;
historical land use, modern urban encroachment and abundance of potential sampling
opportunities. Each survey area included the medieval burgh core and its adjacent land,
whilst avoiding recent urban development. The survey area consisted of a 200 x 800 m
rectangle covering 160,000 m2 (0.16 km2). This was deemed large enough to reveal
identifiable changes in soil properties associated with different past functional zones.
Survey area dimensions were consistent for all burghs to maintain comparable sampling
densities. It is suggested that 100 observations per km2 should be made for very
detailed surveys (Bridges 1982, McRae 1988). This recommendation largely refers to
natural soil variation which is more homogenous than urban anthrosols. Therefore, the
adopted sampling density was 1 observation per 1600 m2, totalling 100 observations for
each 0.16 km2 survey area.

4.2.1.1 Stratified Grid

The use of a grid system is an effective mechanism for structuring observations in areas
with complex soil patterns (Dent and Young 1981). Moreover it is advantageous over
free sampling in that regular spaced sampling results in equal statistical representation
of soil variance (Bridges 1982). Grid surveys consist of four types, random survey,
square grid survey, stratified random survey and stratified grid survey. Random survey is
problematic for heterogeneous soil because observations are pre-determined through
random number generation which can result in large areas remaining un-sampled (Tan
1996).

- 75 -
Figure 19: Location of survey area at Lauder (black rectangle). Red boundary delimits 1862AD urban extent taken from 1:10,560 scale Ordnance
Survey map inset © Landmark information group, www.old-maps.co.uk

- 76 -
Figure 20: Location of survey area at Pittenweem (orange rectangle). Red boundary
delimits 1855AD urban extent taken from 1:10,560 scale Ordnance Survey map inset ©
Landmark information group, www.old-maps.co.uk

- 77 -
Figure 21: Location of survey area at Wigtown (orange rectangle). Red boundary delimits
1850AD urban extent taken from 1:10,560 scale Ordnance Survey map inset © Landmark
information group, www.old-maps.co.uk

- 78 -
Square grid survey is also inappropriate for this study due to inflexibility of observation
points and the potential that trends may go unnoticed if variations in soil properties
coincide with orientation and spacing of the grid (Rudeforth 1982). Stratified random
survey and stratified grid survey present a compromise between the latter approaches.
Stratified random and stratified grid survey use pre-defined grid squares to structure
observations. Observations are chosen at random in the former method and
systematically in the latter. Stratified grid survey was identified as the most appropriate
choice for this study given that provision is made for sampling obstacles associated with
urban environments, for example car parks, grave yards, roads and recreational ground.

4.2.2 Field Methods

4.2.2.1 Auger Sampling

Soil sample acquisition through auguring was chosen as the primary method of soil
survey because it is a relatively quick and easy technique for surveying both laterally
and vertically (Bridges 1982). In addition it enabled field assessment of soil variation
aiding subsequent location of representative soil pits. Auger sampling was structured
through stratified grid survey. The grid consisted of 40 x 40 m grid units totalling 100
potential observation locations per survey area. In some circumstances grid units were
not sampled, for instance when covered by an impermeable surface or modern housing.
Consequently, additional observations were made in emerging areas of interest. These
areas corresponded to past functional areas and exhibited pronounced variability in
certain field characteristics such as topsoil depth. The resulting sample distributions
comprised 117 observations at Lauder (Figure 22), 102 at Pittenweem and 102 at
Wigtown (Figure 23).

To compare soil properties over space and depth, soil samples were taken at 0-20 cm,
20-40 cm, 40-60 cm and 60-80 cm depth at each observation location or until the auger
could go no further. Samples were taken at depth increments rather than from discrete
horizons given that prior determination of horizon boundaries is impractical using a
Dutch auger and mixing between boundaries is an unavoidable consequence of the
barrelled auger tip. Soil samples were taken from the centre of the auger head at the
lowest depth of each increment range. Sample volume was dictated by the auger
headspace which equated to between 100-150 g of moist soil. In total 270, 377 and 158
loose bag samples were obtained from Lauder, Pittenweem and Wigtown respectively.

- 79 -
Figure 22: Stratified grid survey at Lauder showing auger sample locations (red circles), soil pit locations (blue squares) and ‘reference’ soil pit
locations (green squares)

- 80 -
Figure 23: Stratified grid survey at Pittenweem (left) and Wigtown (right) showing auger
sample locations (red circles), soil pit locations (blue squares) and ‘reference’ soil pit
locations (green squares)

Horizon depth(s), soil colour and presence of anthropogenic inclusions were also noted
at each observation location to provide context to subsequent results and supplement
field observations of soil characteristics.

- 81 -
4.2.2.2 Soil Pit Sampling
To enable further investigation of areas with distinct soil characteristics, samples were
taken from soil pits. The location and number of soil pits was determined by identification
of emergent patterns in soil variability in association with the need to investigate
comparable functional zones. Samples were also taken from so-called ‘reference’ soil
pits to provide a comparison of soil with lesser anthropogenic modification. It was
assumed that the impact and magnitude of waste disposal diminished with distance from
the town core (as discussed in section 3.3). Consequently, reference pits were located at
the far northern edge of each survey area with the exception of Lauder (Figure 22,
Figure 23). In the latter case there is evidence to suggest that burgess acres extended to
the rear of the survey area. Accordingly, topsoil samples were taken from a pit 1km
northwest of the survey area and C horizon samples extracted from an exposed profile
200m south of the survey area.

Prior to sampling a brief site description was made at each soil pit location. The following
information was recorded; profile identification number, grid reference, date, preceding
weather conditions, elevation (m), locality and vegetation/land use. Soil pits were dug to
a maximum of 1m or until the topsoil could be clearly distinguished from underlying
horizons. Field sketches were made at each location in association with a soil profile
description recording depth (cm), Munsell soil colour, macropores, mottling, stoniness,
roots and exotic inclusions (Hodgson 1974, Munsell Color ® 1994) (Figure 24).

To enable subsequent determination of physical and chemical soil properties, loose bag
samples were taken from each horizon. Bulk samples were extracted from a depth range
of 20cm allowing collection of sufficient material. To avoid incorporation of material from
adjacent horizons a 2cm buffer zone was imposed at each horizon boundary. In some
cases horizons were deeper than 44cm, hence, 20cm samples were taken in vertical
succession. Moreover if horizons were less than 24cm the utility of sampling was
decided on an individual basis. In total 18, 32 and 14 bulk samples were taken at
Lauder, Pittenweem and Wigtown respectively. In addition Kubiena tin samples were
taken for the purpose of micromorphological characterisation. Kubiena samples were
extracted in accordance with recommendations outlined in Goldberg and MacPhail
(1996: 328-333). Care was taken to ensure all Kubiena samples were contiguous with
bulk samples.

- 82 -
As reference soil pits, three profile faces were described and sampled instead of one
(Figure 25). This was done to increase confidence when characterising areas
representative of lesser anthropogenic influence. Alternatively, additional reference soil
pits could have been sampled; however, locations were limited in terms of space and
suitability.

Figure 24: Example of field sketch, profile description and soil sampling undertaken for LA
2 (Lauder soil pit number 2)

Figure 25: Example of field sketch, profile description and soil sampling undertaken for
‘reference’ soil pit PT 1

- 83 -
4.3 Laboratory Analyses

The following section presents a summary of laboratory analyses undertaken on bulk

soil samples obtained through auger and soil pit extraction. Prior to analysis all samples
were air dried for a minimum of two weeks until steady state mass was achieved.
Samples were then sieved to separate the fine fraction (<2mm) from the coarse fraction
(>2mm) to remove larger roots and stones. Exotic inclusions such as pottery sherds and
charcoal fragments were retained before discarding the coarse fraction. Considering
time constraints it was not possible to take replicate measurements of all soil samples,
therefore samples obtained through auger survey were analysed only once. Soil pit
samples were analysed three times to assess data variability. In all cases anomalous
values were systematically identified and re-checked.

4.3.1 pH

Soil pH is a measure of the concentration of hydrogen ions in a solution which is defined

as ‘-log (H+) where (H+) is the activity of hydrogen ions in solution’ (Rowell 1994:159).
Measurement of soil pH, therefore, allows determination of the degree of soil acidity or
alkalinity. In the natural environment soil pH is controlled by climate where rainfall and
temperature influence processes of nutrient leaching and weathering. Moreover
additional variables such as parent material, hydrology and vegetation further act to
influence pH at regional and local scales (Brady and Weil 1999).

Although soil pH is largely controlled by environmental factors, anthropogenic activities

can impact soil acidity, for example the use of ammonium based chemical fertilisers are
known to react with organic wastes resulting in nitric and sulphuric acid accumulation
(Brady and Weil 1999). Likewise, fossil fuel combustion and vehicular traffic are
identified as causes of increased acidity in urban soils (Bridges 1991). Certain
agricultural practices can result in an intentional reduction in soil acidity such as
application of lime, marl, shells and ground limestone. Similarly, decreased soil acidity
through addition of domestic refuse and fuel residues such as ash and soot is noted in
domestic urban gardens (Bridges 1991). Nonetheless increased soil alkalinity can be
problematic, for instance when salts drain insufficiently from soils irrigated with salty
water (Brady and Weil 1999, Rowell 1994).

Studies of soil pH in geoarchaeology are usually limited to issues of artefact preservation

(Berna et al., 2004, Matthiesen 2004). However, it seems logical that variations in soil
pH may reflect the historical legacy of past soil amendments such as shell, bone, fuel

- 84 -
residue and midden waste (Carter 2001, Cachart 2000, Davidson et al., 2006). Soil pH
is also an important supplementary measure when interpreting spatial patterns in
elemental concentrations, for example Entwistle et al., (1998) note a narrow pH range in
soils both on and off site at Greaulin, Isle of Skye suggesting that on site variation in
elemental concentrations is anthropogenic in origin.

Soil pH can be described as a ‘master variable’ affecting a multitude of chemical,

biological and physical soil properties (Brady and Weil 1999:252). Analysis of soil pH,
therefore, presents a relatively quick and easy method for obtaining information about a
multitude of soil properties including changes associated with past anthropogenic
activities. Accordingly, soil pH was measured using a MK11 Series Analogue meter in
accordance with procedures outlined by Rowell (1994:159-161). The meter was
calibrated using pH 4 and pH 7 buffer solutions and subsequently recalibrated after
every 40 readings to prevent instrumental drift.

4.3.2 Loss on Ignition (% LOI)

Loss on ignition (LOI) is a measure of the percentage mass of ‘soil organic matter (SOM)
lost from oven dry soil when heated to between 105oC and 500oC (Rowell 1994:48).
However, overestimation of SOM loss using the LOI method is possible, for example
through loss of structural water from clay minerals and loss of CO2 from carbonates in
calcareous soils. Ball (1964) proposes that error associated with CO2 loss can be
minimised through ignition in the lower temperature region (375oC as opposed to 850oC)
and suggests that variation in LOI is only 4-6% between soils with 5% clay and 50% clay
content. This level of accuracy is considered acceptable for most archaeological and
pedological studies (Ball 1964). LOI is therefore identified as a reliable and rapid
technique for approximation of SOM. LOI was determined in accordance with
procedures outlined in Rowell (1994:48) using the formula below. All results are
expressed as g per 100g of oven dry soil i.e. % LOI.

( mass oven dry soil ( g ) − mass ignited soil ( g ))

% LOI = 100 *
mass oven dry soil ( g )

In geoarchaeological studies LOI has been used in association with elemental analysis
to identify areas cultivated in the past, for example it is suggested that a moderate
positive correlation between LOI and phosphorus (P) at Knockaird, Isle of Lewis reflects
sustained application of manure rich in organic matter and P (Entwistle et al., 2000).
Areas with lower LOI levels compared to offsite control samples at Greaulin, Isle of Skye

- 85 -
are thought to reflect the practice of applying manure rich in nitrogen (N) to cultivated
land resulting in rapid microbial decomposition of SOM (Entwistle et al., 1998).
Cultivation and regular cropping also enhance SOM decomposition through changes in
soil conditions such as increased soil aeration and disaggregation. Elevated levels of
SOM in cultivated soil are likely to quickly disappear. It is therefore recognised that past
soil enrichment may not necessarily be reflected in higher LOI values.

4.3.3 Environmental Magnetism

Environmental magnetism refers to the study of rock-magnetic properties of natural

environmental materials such as soils, sediments, peat cores and ice cores, and
anthropogenic wastes arising from agricultural and industrial processes (Oldfield 1999,
Robinson 2002). All these materials exhibit magnetic behaviour which can be measured
and quantified allowing identification of processes associated with their formation and
recognition of component minerals (Dearing 1999). In pedological and archaeological
studies the most widely used measurements are low frequency magnetic susceptibility
(ҲLF) and frequency dependant magnetic susceptibility (ҲFD). These techniques differ
from most other mineral magnetic measurements in that they are determined in the
presence of an induced magnetic field. In comparison parameters such as isothermal
remanent magnetisation (IRM) and saturation isothermal remanence (SIRM) are a
measure of magnetisation retained after removal from an external magnetic field (Smith
1999).

Magnetic susceptibility can be defined as a measure of ‘the ability of a substance to be

magnetised’ (Rapp and Hill 1998:184). In the natural environment magnetic susceptibility
is dependant on concentrations of ferrimagnetic minerals such as magnetite (Fe3O4) and
maghaemite (Fe2O3) (Dearing 1999). Hence, magnetic susceptibility of soils and
sediments is largely controlled by the nature and abundance of iron oxides originating
from underlying parent material. Anthropogenic activities are known to enhance
magnetic susceptibility of soils, for example burning can result in formation of
ferrimagnetic minerals such as maghaemite and non-stoichiometric magnetite (Crowther
2003).

- 86 -
Frequency dependant magnetic susceptibility reflects the percentage of viscous-
superparamagnetic (VSPM) grains present in a mineral magnetic assemblage (Dearing
1999). This is based on the observation that relaxation times of superparamagnetic
grains differ at high and low magnetisation frequencies (Jordanova et al., 2001). Low
frequency (ҲLF) measures the combined response of stable single domain grains (SSD),
pseudo-single domain grains (PSD), multidomain grains (MD) and superparamagnetic
(SP) grains present within a sample. Conversely high frequency (ҲHF) measurements
exclude the contribution made by SP grains (for a full explanation see Dearing 1999:60-
62).

In geoarchaeological studies magnetic measurements have been used to discriminate

artefact origins, for example Jordanova et al., (2001) identify differences in magnetic
behaviour of burnt clay materials such as plaster and brick, and burnt soils between
archaeological sites in northern and southern Bulgaria. Recognition that anthropogenic
activity can lead to magnetic enhancement of soils has resulted in the use of magnetic
susceptibility in archaeological site prospectation, for instance spatial patterns derived
from underlying occupation and industrial sites are identified at Serra di Vaglio, Southern
Italy (Chianese et al., 2004). Moreover magnetic measurements can be used to
investigate use of individual dwellings, for instance Peters et al., (2000) associate hearth
residues with selected internal floor deposits and external midden material from Late
Iron Age houses at Galson site, Isle of Lewis.

Magnetic measurements have also been extensively used to identify past evidence of
burning including fuel residues. Dewar et al., (2002) use modern analogues to identify
fuel sources contributing to deposits from Old Scatness Broch, Shetland. Although
deposits comprised a mixture of minerals with differing magnetic properties, it was
possible to identify inputs from ash derived from turf and furnace residues. Similarly
Peters et al., (2001) demonstrate the utility of experimental data in investigating
archaeological deposits at Calanais Farm, Isle of Lewis where techniques such as high
temperature magnetic susceptibility were used to distinguish well humified peat and
wood from fibrous upper peat and peat turf.

Magnetic susceptibility and frequency dependant magnetic susceptibility are simple,

non-destructive and rapid techniques for investigating anthropogenic impact on soils with
particular reference to burning, fire histories and fuel residues.

- 87 -
4.3.3.1 Mass Dependant Magnetic Susceptibility
To determine mass dependant magnetic susceptibility soil samples were dried at 105oC
overnight to remove excess moisture prior to analysis. Measurements were undertaken
on a Bartington MS2 Magnetic Susceptibility Meter according to procedures outlined in
Bartington Instruments Ltd (1996). The chosen measurement interval was x0.1 which
provides an average reading every 10 seconds. This measurement interval is more
robust than taking individual readings and provides additional noise filtering. Corrections
for any thermally induced drift were made using the formula below.

corrected measuremen t ( Rk ) = sample measuremen t (10 −5 ) − air measuremen t (10 −5 ) / 2

Measurements for mass dependant magnetic susceptibility (Ҳ) were taken under low
frequency (0.4465 kHz) which measures the response of all magnetisable material in a
sample. Mass dependant magnetic susceptibility was calculated using the following
formula according to Dearing (1999:46).

χ (10 −6 m 3 Kg −1 ) = Rk (10 −5 ) / sample weight ( g ) / 10

4.3.3.2 Frequency Dependant Magnetic Susceptibility

The procedure outlined in section 3.3.3.1 was repeated under high frequency (4.65 kHz)
conditions to enable calculation of frequency dependant magnetic susceptibility (ҲFD).
Frequency dependant magnetic susceptibility was calculated using the following formula
according to Dearing (1999:47) where KLF is the corrected measurement under low
frequency and KHF under high frequency.

χ FD (10 −6 m 3 Kg −1 ) = (( K LF − K HF ) / sample weight ( g ) / 10)

- 88 -
4.3.4 Elemental Concentrations

Anthropogenic activities such as habitation, agriculture and industrial processing modify

soil chemical properties (Aston 1998, Wilson et al., 2006a). Accordingly, identification of
changes in elemental concentrations enable differentiation of former land use patterns in
space (Bull et al., 2001, Wells 2004) and time (Shahack-Gross et al., 2005,
Alexandrovskaya and Panova 2003). However, it should be noted that human induced
elemental changes are affected by elemental retention rates, elemental fixation, climatic
conditions and soil properties (Holliday and Gartner 2007).

Phosphorus (P) is routinely used in geoarchaeological studies as an indicator of human

activity considering its high soil retention rate and relative stability (Leonardi et al., 1999).
Increased phosphorus concentrations occur through deposition of anthropogenic
materials rich in organic matter including human and animal excreta, domestic refuse
and fuel residue (Holliday and Gartner 2007). Accordingly, spatial analysis of enhanced
phosphorus concentrations can be used to identify and delineate areas associated with
human habitation in the past (Schlezinger and Howes 2000). However, it must be noted
that although phosphorus can be used to identify areas of past human activity it cannot
differentiate between specific land use practises, for example accumulation of debris
from settlements and manuring practises both lead to elevated phosphorus
concentrations (Entwistle et al., 2000a, Entwistle et al., 2000b).

Conversely multi-element analysis enables investigation into a range of anthropogenic

activities using elemental signatures to identify patterns of enhancement, for example
Entwistle et al., (1998) identify four groups of elements representing differences in land
use at a Greaulin, Isle of Skye. Similarly Wilson et al., (2006a) discriminate a suite of
elements including barium (Ba), calcium (Ca), phosphorus (P), lead (Pb), strontium (Sr)
and zinc (Zn) which are consistently indicative of former activity areas across a range of
abandoned farms in Britain. Further it is suggested that emphasis should be placed on
interpreting relative enhancement patterns rather than absolute elemental
concentrations when comparing sites (Wilson et al., 2006a). This approach discounts the
influence of differing geology and soil properties between sites.

- 89 -
Multi-element analysis is identified as a relatively quick tool for investigating differences
in past land use at historic sites. Accordingly samples were prepared by nitric acid
digestion as described in Wilson et al., (2005:1095). Pseudo-total acid extraction was
preferred over weak acid or exchangeable fraction extraction considering Wilson et al’s.,
(2006b) finding that the latter methods impede the recovery of anthropogenic elements
associated with historic land use. Accordingly elemental concentrations were measured
with a Perkin Elmer Optima 3300RL Inductively Coupled Plasma-Atomic Emission
Spectrometer (ICP-AES) using the in-house traces programme at the NERC ICP-AES
Facility, Royal Holloway. Corrections for instrumental drift were subsequently made
using procedural blanks. Elemental concentrations (mg/Kg) for Ba, Co, Cr, Cu, Li, Ni,
Pb, Sc, Sr, V, Y and Zn were obtained using the following formula where the BCV is the
blank corrected values and the sample mass is 5±.01g.

elemental concentrat ion ( mg / Kg ) = BCV ( mg / Kg ) / sample mass (5 ± 0.01g )

The following formula was used to determine elemental concentrations (mg/Kg) for
oxides (Al2O3, Fe2O3, MgO, CaO, Na2O, MnO, K2O, TiO2 and P2O5). The abbreviation
BCV represents the raw blank corrected value. AMX stands for the total atomic mass of
the element in question and AMY the total atomic mass of oxygen.

elemental concentration (mg / Kg ) = ( BCV / 5 ± 0.01) * ( AM X / AM X + AM Y ) * 10,000

- 90 -
4.4 Micromorphological Analysis

Soil micromorphology is the ‘study of the structure of spatial patterns in natural and non-
natural layers, deposits and soils using microscopic techniques’ (Kooistra and Kooistra
2003:605). Micromorphology has been used since the 1950s as a routine analytical
method in archaeological science for investigation of palaeosols (Kemp 1998, Macphail
and Goldberg 1995), however, its use in geoarchaeological studies is still relatively
recent (Courty et al., 1989, Davidson et al., 1992, French 2003). Micromorphological
investigation into impacts of past human activities on soils and sediments is noted for a
range of applications including, soil formation processes (Simpson et al., 2006, Simpson
et al., 1998a), agriculture (Bryant and Davidson 1996, Davidson and Carter 1998,
French and Whitelaw 1999, Simpson 1997, Simpson et al., 1998b), occupation deposits
(Homsey and Capo 2006, Matthews 1995, Matthews et al., 1997) and cultural landscape
studies (Simpson et al., 2003, Simpson et al., 2005).

Micromorphological studies of urban contexts are limited although ‘dark earth’ deposits
(Macphail 1994, Macphail 2003, MacPhail et al., 2003, Ottaway 1992), garden soils
(Clark 1997, Carter 2001) and occupation sequences (Shahack-Gross et al., 2005) have
received some attention. It is recognised that soils are often neglected in urban
archaeology in favour of artefact recovery (Macphail and Goldberg 1995), however they
provide an important resource for investigating anthropogenic processes of deposition,
emplacement and modification (Courty et al., 1989). Anthropogenic features in thin
section largely consist of waste materials associated with human occupation, for
example fuel residues, domestic refuse and construction materials. Accordingly
micromorphology is a useful tool for characterising urban anthrosols and investigating
processes associated with waste management and disposal in historic small towns.

4.4.1 Sample Preparation

Thin sections were prepared from undisturbed soil samples at the Thin Section and
Micromorphology Laboratory, University of Stirling according to standard procedures
(http://www.thin.stir.ac.uk/methods.html, 2007). Samples were dried through solvent
exchange (acetone) in the vapour phase and impregnated under vacuum using an
epoxy resin (Araldite MY750 with hardener HY951). Subsequently impregnated blocks
were cut and precision lapped to 30µm thickness prior to mounting.

- 91 -
4.4.2 Semi-Quantification of Anthropogenic Features

There is a lack of an established protocol for analysing anthropogenic features in thin

section. It should therefore be noted that the methodology presented hereafter was self
evolutionary and guided by the study research objectives as recommended by Courty et
al., (1989).

In the first instance coarse mineral anthropogenic features were identified using a series
of reference images and descriptions (Courty et al., 1989, Stoops 2003). The ascribed
categories were verified by Dr Clare Wilson, University of Stirling. Key characteristics of
each mineral anthropogenic feature are listed in Table 3 and example images are
presented in Figure 26. Coarse organic anthropogenic features were categorised
according to differences in morphology and optical properties between black carbon
exotic inclusions termed ‘fuel residue’ (Table 4, Figure 27). In addition each category
was divided into two size classes, for example FR Type 1 10-255 µm and FR Type 1
255-500+ µm. Distinctions were based solely on descriptive differences in accordance
with recommendations made by Kemp (1998); it should therefore be noted that
classifications of fuel residue categories (FR) do not necessarily represent different fuel
sources.

Table 3: Mineral Anthropogenic Features (PPL: Plane polarised light, XPL: Cross polarised
light, OIL: Oblique incident light)

Feature Key Characteristics

Shell (Mollusc) Fibrous internal fabric (striations), very fine crystal size, high
interference colours (XPL)

Clinker/Slag Grey/brown/dirty yellow (PPL), bubbles in matrix, growth of some

minerals (crystalline, skeletal, dendritic)

Bone Fibrous internal fabric, presence of abundant Haversian Canals

Heated Mineral Red in (OIL), high interference (XPL), mineral matrix

Pottery/Brick Deep red/brown (PPL), diffuse strong/red (OIL), well sorted fabric,
high density

Mortar/Plaster Grey/green-grey (PPL), grey (OIL), inclusions of straw, brick,

bone, areas of cryptocrystalline calcite/calcitic fabric, well sorted
matrix

- 92 -
Table 4: Organic Anthropogenic Features (PPL: Plane polarised light, XPL: Cross
polarised light, OIL: Oblique incident light), FR: Fuel Residue

Feature Key Characteristics

Charcoal Complete cellular structure with little/limited decay, black (PPL),

black (XPL), morphology ranges from rounded to semi-angular.

FR Type 1 Morphology is angular/sub-angular, well defined perimeter, black

(PPL), black (XPL), limited-moderate evidence of internal
cracking/fracturing.

FR Type 2 Morphology is rounded/sub-rounded, defined perimeter, internal

degradation characterised by holes, black (PPL), black (XPL),
internal mineral content high interference (XPL).

FR Type 3 Morphology is angular/sub-angular, well defined perimeter, black

(PPL), black (XPL), limited- moderate evidence of internal
cracking/fracturing, strong red bands present (PPL).

FR Type 4 Morphology can be angular, sub-angular, sub-rounded or

rounded, black (PPL), black (XPL), pronounced internal
degradation characterised by large rounded/elongated holes in
close proximity, appearance almost ‘skeletal’ like, intermixture of
groundmass, organic material/excrements adjacent to perimeter.

FR Type 5 Morphology is angular/sub-angular, well defined perimeter, black

with areas of brown/red-brown (PPL), black (XPL), pronounced
internal degradation.

FR Type 6 Morphology can be angular, sub-angular, sub-rounded or

rounded, black (PPL), black (XPL), internal degradation
characterised by rounded/sub-rounded holes, organic
material/excrements adjacent to perimeter.

FR Type 7 Morphology is angular/sub-angular, well defined perimeter, black

(PPL), Black (XPL), pronounced internal degradation.

FR Type 8 Morphology sub-angular/sub-rounded, ‘ragged’ perimeter edge,

black (PPL), black (XPL), limited/moderate evidence of internal
cracking/fracturing.

FR Type 9 Morphology sub-angular/sub-rounded, ‘ragged’ perimeter edge,

black with areas of brown/red-brown (PPL), black (XPL),
limited/moderate evidence of internal cracking/fracturing.

FR Type 10 Morphology sub-angular, black (PPL), black with high interference

colours near perimeter (XPL), calcitic fabric near perimeter,
limited-moderate internal degradation.

- 93 -
Figure 26: Example images of coarse mineral anthropogenic features in thin section from Lauder, Pittenweem and Wigtown. First row (left to
right): pottery (OIL), pottery (PPL), mortar (OIL), mortar (PPL). Second row (left to right): shell (PPL), shell (XPL), bone (XPL), bone (PPL). Third row
(left to right): clinker/slag (PPL), heated mineral (OIL)

- 94 -
Figure 27: Example images of coarse organic, anthropogenic features in thin section from Lauder, Pittenweem and Wigtown. First row (PPL) (left
to right): Charcoal, FR4, FR 6, and FR 2. Second row (PPL) (left to right): FR1 (XPL), FR7, FR5 and FR3. Third row (PPL) (left to right): FR 8, FR9,
FR10 and FR 10 (XPL)

- 95 -
It is recommended that analysis of anthropogenic contexts should be holistic in approach
and undertaken by experienced well trained analysts (Macphail 1998, Macphail and
Goldberg 1995). However such interpretations often lack validation (Davidson and
Simpson 2001). Moreover, it is argued increasing priority in micromorphological studies
should be placed on experimental control, verification and quantification (Davidson and
Carter 2000). Accordingly, a semi-quantitative approach to slide description was adopted
using the methodology outlined hereafter.

An acetate grid consisting of 165 5x5mm squares was placed over the surface of each
slide covering 4125 mm2. A total of 66 numbers between 1 and 165 were independently
generated for each slide using a random number generator (Daniels 2001-2003).
Corresponding grid squares were subsequently highlighted with a blue transparent mark
(Figure 28). Percentage abundance estimates of coarse mineral (>10µm), coarse
organic (>10µm), pedofeatures, C:F ratio (coarse) and void space were made for each
individual square using absolute numbers, for example 2=<2%, 5=2-5%, 10=5-10%,
15=10-15%. This enabled calculation of a representative mean for each feature per
slide. Abundance estimates were made using an Olympus BX 51 petrological
microscope following procedures in Bullock et al, (1985) and Stoops (2003).
Furthermore a range of magnifications (x10-x400) and light sources (plane polarised
light, cross polarised light and oblique incident light) were utilised enabling feature
description and identification.

Figure 28: Left: slide prior to grid overlay, Right: slide with grid overlay, grid squares
selected through random number generation are highlighted with a blue mark

- 96 -
Abundance estimates were made for 1650mm2 of the total slide surface (4125mm2)
equating to 40% randomly distributed coverage. To verify whether 40% constitutes a
representative sub-sample of the total slide area, % mean void space was calculated
incrementally (1%) for Wigtown thin sections (Figure 29). Initially (1-15%) the data are
noisy and characterised by pronounced variation in estimated % mean void space.
However as % slide area increases the estimated % mean void space becomes
progressively steady resulting in nearly constant values between 30-40%. Similarly
continuous measurement of the C:F ratio for Lauder thin sections indicates distinct
variation in % coarse material between 0-20% slide area which becomes more uniform
as slide area increases (Figure 30).

60
W G 1/1 W G 1/2 W G 1/3 W G 2/1 W G 2/2 W G 3/1
55 W G 3/2 W G 4/1 W G 4/2 W G 5/1 W G 5/2 W G 5/3

50
M ean Void Space (%)

45

40

35

30

25

20

15

10
1 4 7 10 13 16 19 22 25 28 31 34 37 40

% Slide Area

Figure 29: Mean void space of Wigtown thin sections calculated incrementally (every 1%).
Legend denotes town code (WG) and slide reference and context number (1/1)

85
LA 1/1 LA 1/2 LA 2/1 LA 2/2 LA 3/1 LA 4/1 LA 4/2 LA 5/1
80
LA 6/1 LA 6/2 LA 7/1 LA 7/2 LA 7/3 LA 9/1 LA 9/2

75
C:F Ratio (% Mean Coarse)

70

65

60

55

50

45
1 4 7 10 13 16 19 22 25 28 31 34 37 40

% Slide Area

Figure 30: Mean % coarse material of Lauder thin sections calculated incrementally (every
1%). Legend denotes town code (LA) and slide reference and context number (1/1)

- 97 -
Analysis of changes in mean % relative abundance of anthropogenic features for slide
PT 4/1 (Figure 31) and WG 4/1 (Figure 32) also indicate relative stability in estimations
exceeding 30% of the slide surface area. Accordingly it is proposed that 40% is an
acceptable minimum representative area for anthropogenic features and soil
characteristics. Conversely, VandenBygaart and Protz (1999) suggest an alternative
method for determining the minimum representative area for quantitative analysis of
pedofeatures using image analysis systems. This approach is useful for determining
features readily identifiable when digitised, for example void space and soil pores.
However, it is argued that insufficient provision currently exists for accurate and
consistent identification of anthropogenic features using image separation techniques.

Pottery Mortar/Plaster FR 1 FR 4 FR 6 FR 8
14
13
12
Mean Relative Abundance (%)

11
10
9
8
7
6
5
4
3
2
1
0
1 4 7 10 13 16 19 22 25 28 31 34 37 40
% Slide Area

Figure 31: Mean % relative abundance of selected anthropogenic features for soil pit 4,
context 1 (4/1), Pittenweem

Bone FR 1 FR 4 FR 6
100
Mean Relative Abundance (%)

10

1
1 4 7 10 13 16 19 22 25 28 31 34 37 40
% Slide Area

Figure 32: Mean % relative abundance of selected anthropogenic features for soil pit 4,
context 1, Wigtown

- 98 -
4.4.2.1 Revision of Coarse Organic Material Classification
Size classes (10-255µm and 255-500µm) of coarse organic material were amalgamated
due to a lack of discernable trends within the lower size classification. To investigate size
differences in fuel residue more broadly the percentage of total fuel residue sized 10-
255µm was calculated for each thin section (% FR 10-255µm). In addition, certain
coarse organic material categories were merged to enable better detection of differences
in the abundance of fuel residue classes between zones and burghs. Categories FR 4
and FR 6 (FE 4 & 6) were merged owing to their similar morphological characteristics. It
is suggested that these two classes represent the same material albeit in different states
of degradation. Categories FR 5 and FR7 (FR 5 & 7) were also combined due to
similarities in their composition. The only difference between FR 5 and FR 7 is that FR 5
has areas which are brown/red-brown (PPL); however, it is suggested that this may
reflect variation caused through slide production rather than difference in the nature of
material. Variation in colour may occur due to variations slide thickness over the surface
area of the slide. Similarly, FR 8 and 9 (FR 8 & 9) were merged given the only difference
between these two categories was limited to colour. These revised categories produced
clearly identifiable patterns both within and between burghs. Results were summarised
in the form of ‘blobby table’ diagrams to enable comparison of data between zones.

- 99 -
4.5 Spatial Interpolation

Spatial distributions of topsoil depth, pH, LOI, mass dependant magnetic susceptibility
and frequency dependant magnetic susceptibility were plotted using the Inverse
Distance Weighted (IDW) spatial interpolation function available within ArcView 3.2. IDW
was identified as the most appropriate spatial interpolation method given that it is
suitable data with pronounced variation over very short distances (ArcView 3.2 GIS
2002). Consideration was given to the use of Spline interpolation; however, this
technique is more suitable for data with limited variation (ArcView 3.2 GIS 2002). Spatial
distributions of logged (LOG10) elemental concentrations (Ba, Ca, K, Sr, P, Pb and Zn)
were also plotted using the IDW spatial interpolation function. Logged elemental
concentrations were used to minimise the effect of geochemical ‘hotspots’ present in
natural data (Figure 33).
a) b)

Log P (mg/Kg)

Figure 33: Spatial distributions of P (mg/Kg) (a) and Log P (mg/Kg) (b) at Pittenweem. Red
boundary delimits 1855AD urban extent

- 100 -
4.6 Statistical Analyses

4.6.1 Zonal Delimitation

To enable statistical comparison of data both within and between burghs, each town was
divided into functional zones. Zones were delineated through a combination of historical
review (section 2.2, section 2.3), field experience and spatial analysis of selected soil
properties.

Zones at Lauder consist of; High Street, Hinterland Near, Hinterland Far and Hinterland
Thirlstane. The High Street zone is constrained by Lauder’s 1862 AD urban extent,
hence represents the location of the old burgh core. The Hinterland Near zone
corresponds to the immediate hinterland south of Lauder. Similarly the Hinterland
Thirlstane zone corresponds to the immediate hinterland north of Lauder. The Hinterland
Far zone represents the distant hinterland immediately south of Lauder.

Figure 34: Delineation of zones at Lauder. Red boundary delimits 1862AD urban extent

- 101 -
 a) b)

Figure 35: Delineation of zones at Pittenweem (a) and Wigtown (b). Red boundary delimits
1855AD and 1850AD urban extent respectively

Zones at Pittenweem comprise; Harbour, High Street, Hinterland Near and Hinterland
Far. The Harbour and High Street zones are constrained by Pittenweem’s 1855AD
urban extent, hence represent the location of the old burgh core. The Hinterland Near
and Hinterland Far zones correspond to the immediate and distant hinterland north of
Pittenweem.

Zones at Wigtown include; High Street, Hinterland Near and Hinterland Showfield. The
High Street zone is constrained by Wigtown’s 1850AD urban extent, hence represents
the location of the old burgh core. The Hinterland Near zone corresponds to the
immediate hinterland north of Wigtown and the Hinterland Showfield corresponds to the
immediate hinterland south of Wigtown.

- 102 -
4.6.2 Determination of Significant Differences between Zones

Minitab (15) statistical software was used for all statistical analyses described hereafter.
Statistical analyses of physical and chemical and elemental data were conducted on 0-
20cm depth soil data to enable direct comparison between burghs. Consideration was
given to comparing zones to ‘reference’ soil pits in addition to each other. However,
inclusion of ‘reference’ soil data was problematic given that it frequently overlapped with
zone data which is more variable in nature. This resulted in very few significant
differences between zones and ‘reference’ soil pits despite particular zones having
considerably larger average values. Statistical analysis of soil micromorphology data
was conducted on percentage abundance estimates of coarse organic material within
zones and ‘reference’ soil pits.

4.6.2.1 Physical and Chemical Data

Topsoil depth, pH, LOI, mass dependant magnetic susceptibility and frequency
dependant magnetic susceptibility data were tested for normality using the Anderson-
Darling normality test. In most cases data were non-normal and could not be
transformed using logarithmic or square root transformations.

Kruskal-Wallis analysis was identified as the most appropriate non-parametric method to

test for significant differences between zones. One of the limitations of Kruskal-Wallis
analysis is that it does not indicate which specific zones are different and in what way.
Consequently Kruskal-Wallis analysis was performed in collaboration with Dunn’s test.
Dunn’s test enables multiple comparisons to be made using non-parametric data with
unequal sample populations (Wheater and Cook 2000). Results of Kruskal-Wallis
analysis with Dunn’s test were presented in Pairwise Comparison diagrams. The test
statistic (z score) of each comparison was plotted and z scores exceeding –z or z were
classed as significant at p<0.05 (95%) confidence level. In cases where data were
normal a One Way Analysis of Variance (One-Way ANOVA) with Tukey-Kramer (95%
confidence) multiple comparisons was used to determine significant differences in
sample means between zones.

- 103 -
4.6.2.2 Elemental Data
All elemental concentrations were tested for normality using the Anderson Darling
normality test. Elemental data were predominantly non-normal and could not be
transformed using logarithmic or square root transformations. Kruskal-Wallis analysis in
association with Dunn’s test was therefore used to identify significant differences in
median elemental concentrations between zones. Results of Kruskal-Wallis analysis with
Dunn’s test were presented in Pairwise Comparison diagrams. The test statistic (z
score) of each comparison was plotted and z scores exceeding –z or z were classed as
significant at p<0.05 (95%) confidence level. In cases where data were normal a One
Way Analysis of Variance (One-Way ANOVA) with Tukey-Kramer (95% confidence)
multiple comparisons was used to determine significant differences in sample means
between zones.

4.6.2.3 Soil Micromorphology

Statistical investigation of significant differences between zones was limited to
comparison of coarse organic material given the limited abundance of coarse mineral
anthropogenic inclusions. Percentage abundance estimates of coarse organic
anthropogenic inclusions (Charcoal, FR 1, FR 2, FR 3, FR 4 & 6, FR 5 & 7, FR 8 & 9, FR
10 and % FR 10-255µm) were tested for normality using the Anderson-Darling normality
test. In most cases data were non-normal and could not be transformed using
logarithmic or square root transformations.

Kruskal-Wallis analysis was identified as the most appropriate non-parametric method to

test for significant differences between zones. One of the limitations of Kruskal-Wallis
analysis is that it does not indicate which specific zones are different and in what way.
Consequently Kruskal-Wallis analysis was performed in collaboration with Dunn’s test.
Dunn’s test enables multiple comparisons to be made using non-parametric data with
unequal sample populations (Wheater and Cook 2000). Results of Kruskal-Wallis
analysis with Dunn’s test were presented in Pairwise Comparison diagrams. The test
statistic (z score) of each comparison was plotted and z scores exceeding –z or z were
classed as significant at p<0.05 (95%) confidence level. In cases where data were
normal a One Way Analysis of Variance (One-Way ANOVA) with Tukey-Kramer (95%
confidence) multiple comparisons was used to determine significant differences in
sample means between zones.

- 104 -
4.6.3 Determination of Significant Differences between Depths

Soil pH and % LOI data for 0-20 and 20-40cm depth at Lauder, and 0-20, 20-40, 40-60
and 60-80cm depth at Pittenweem were tested for normality using the Anderson Darling
normality test. All data were non-normal and could not be collectively transformed using
logarithmic or square root transformations. Hence, Kruskal-Wallis analysis in association
with Dunn’s test was used to identify significant differences between depths. It should
be noted that Wigtown was omitted from this analysis due limited sample numbers at
lower depths.

Furthermore, to investigate the relationship between depth and zone a Two Way
Analysis of Variance with Tukey-Kramer multiple comparisons (Two Way ANOVA) was
undertaken using a General Linear Model (GLM). A GLM was used in favour of a
Balanced ANOVA because it does not assume equal sample populations. It is
recognised that a Two Way ANOVA is a parametric test, hence not necessarily
appropriate for non-normal data. However, there are no non-parametric analyses which
test for interaction between variables. Accordingly, results from this analysis are
discussed cautiously.

4.6.4 Determination of Significant Differences between Burghs

4.6.4.1 Physical and Chemical Data

To investigate differences in topsoil depth, pH, LOI, mass dependant magnetic
susceptibility and frequency dependant magnetic susceptibility between burghs a Two
Way Analysis of Variance with Tukey-Kramer multiple comparisons (Two Way ANOVA)
was undertaken using a General Linear Model (GLM). This analysis allows comparison
of sample means between burghs and sample means between zones in addition to
testing for interaction between burghs and zones. It should be noted that only the High
Street and Hinterland Near zones were used in the model given that these are the only
two zones which are present in all three burghs.

As stated in section 4.6.2.1, soil physical and chemical data were predominantly non-
normal. It could therefore be argued that a non-parametric test should have been used
to investigate data variation between burghs. However, there are no alternative non-
parametric methods which test for interaction between variables. The robustness of Two
Way ANOVA was confirmed though comparison of data with and without outliers.
Removal of outliers had minimal effect on the significance of data trends; hence outliers
were kept within datasets.

- 105 -
4.6.4.2 Elemental Data
To investigate differences in elemental concentrations between burghs consideration
was given to the use of Two Way Analysis of Variance with Tukey-Kramer multiple
comparisons (Two Way ANOVA). This technique allows investigation into differences
between burghs and zones, in addition to testing for interaction. However initial analyses
proved unreliable, for example the direction of relationships between multiple
comparisons was incorrect. This problem was attributed to the presence of outliers
which adversely skewed mean values. Consideration was given to excluding extreme
data points; however, it is argued that these values are an integral component of
elemental datasets. Therefore investigation into variation of elemental concentrations
between burghs focussed on comparing spatial trends identified using non-parametric
tests (section 4.6.2.2).

4.6.4.3 Soil Micromorphology

Statistical investigation of significant differences between burghs was limited to
comparison of coarse organic material between High Street zones given that the High
Street zone was the only area systematically sampled within each burgh. Percentage
abundance estimates of coarse organic anthropogenic inclusions were tested for
normality using the Anderson-Darling normality test. In all cases data were not normal
and could not be transformed using logarithmic or square root transformations.

Kruskal-Wallis analysis was identified as the most appropriate non-parametric method to

test for significant differences between High Street zones. One of the limitations of
Kruskal-Wallis analysis is that it does not indicate which specific zones are different and
in what way. Consequently Kruskal-Wallis analysis was performed in collaboration with
Dunn’s test. Dunn’s test enables multiple comparisons to be made using non-parametric
data with unequal sample populations (Wheater and Cook 2000). Results of Kruskal-
Wallis analysis with Dunn’s test were presented in Pairwise Comparison diagrams. The
test statistic (z score) of each comparison was plotted and z scores exceeding –z or z
were classed as significant at p<0.05 (95%) confidence level.

- 106 -
4.6.5 Multivariate Analyses

4.6.5.1 Cluster Analysis

Cluster analysis is as an appropriate method for simplifying elemental datasets and
grouping together elements with similar spatial distributions. Cluster analysis was
undertaken on elemental data for Lauder, Pittenweem and Wigtown using the Cluster
Variables function. This is an agglomerative hierarchical method which means that all
variables begin as independent clusters which are subsequently amalgamated in a
series of steps until there is only one remaining group.

Correlation was the preferred distance measure over absolute correlation because it
maximises the distance between positively correlated elements and negatively
correlated elements. The linkage method used was median linkage which calculates the
distance between clusters using the median distance between a variable in one cluster
and a variable in the other cluster, hence reducing the effect of outliers. The final number
of clusters for each burgh was determined through a combination of prior knowledge
relating to patterns in elemental concentrations and identification of an abrupt drop
between amalgamation steps. Results were displayed in the form of a dendrogram to
visualise similarities between clusters. It should be noted that the median linkage
method does not always produce a hierarchical dendrogram. In cases where
amalgamation distances do not increase with each step cluster joins that are both
upward and downward are produced.

4.6.5.2 Discriminant Analysis

Discriminant analysis is a useful tool for investigating whether predicted classifications
are similar to those observed, hence is an appropriate method to determine the
classification accuracy of zones delineated in section 4.6.1. Discriminant analysis was
undertaken on elemental data for Lauder, Pittenweem and Wigtown using cross
validation. The use of cross validation is important because it compensates for overly
optimistic classification given that the data being classified is the same as that used to
build the classification function. The summarised results were displayed in a
classification matrix table.

- 107 -
 5 Field Observations and Soil Micromorphology

This chapter presents results obtained from micromorphological analysis of topsoil

deposits including; semi-quantification of coarse mineral material, coarse organic
material and pedofeatures, and characterisation of fine mineral material and soil
structure.

5.1 Soil Profile Characteristics

Profile descriptions, field sketches and photographs of soil pits are presented in
Appendix 1 for Lauder, Pittenweem and Wigtown. Soil profiles are typically varied in all
three burghs; however, distinct horizons relating to certain zones are identified within
and between towns.

5.1.1 Variation within Burghs

The High Street and Hinterland Near zones within Lauder are characterised by two deep
dark horizons which comprise the topsoil. These layers are classed as hortic horizons,
hence are ascribed the notation Aht 1 (upper topsoil) and Aht 2 (lower topsoil). The Aht 1
horizon is characteristically black/dark brown in colour and ranges in depth from 30 to
50cm. The Aht 2 horizon is dark brown/brown in colour and varies in depth from 20cm to
50cm (Figure 36). The combined depth of horizons Aht 1 and Aht 2 result in topsoil
deposits which are consistently over 60cm. Moreover, both of these horizons contain
cultural debris such as fuel residue and pottery sherds. Topsoil within the Hinterland Far
and Thirlstane zones consists largely of a singular horizon which ranges from dark
reddish brown to brown in colour. This layer is ascribed the notation Ah on account of its
dark colour indicating high organic matter content. Similarly, ‘reference’ soil pit LA 7 is
characterised by an Ah horizon which is reddish brown in colour. In contrast to Aht
horizons, Ah horizons contain limited cultural debris.

Topsoil within the Harbour and High Street zones at Pittenweem comprises an upper
(Aht 1) and lower (Aht 2) hortic horizon (Figure 37). The Aht 1 horizon is black and
ranges in depth from 50-≥100cm in the Harbour zone and 30-40cm in the High Street
zone. The Aht 2 horizon is very dark brown/dark brown in colour and varies in depth
from 20-40cm. The combined depth of horizons Aht 1 and Aht 2 result in topsoil depths
which are consistently ≥80cm within the Harbour zone and over 60cm in the High Street
zone. Moreover, both of these horizons contain cultural debris such as fuel residue,

- 108 -
pottery sherds and shell. Topsoil within the Hinterland Near and Hinterland Far
comprises an Ah horizon which is consistently very dark brown in colour. Typically this
horizon ranges from 20 to 30cm in depth. The ‘reference’ soil pit PT 1 contains an Ah
horizon which is dark brown and 25cm deep. In contrast to Aht horizons, Ah horizons
contain limited cultural debris.

Figure 36: Soil profile description of soil pit LA1 (see Appendix 1 for photographs)

Figure 37: Soil profile description of soil pit PT3 (see Appendix 1 for photographs)

- 109 -
The High Street zone within Wigtown is characterised by topsoil consisting of an upper
(Aht 1) and lower (Aht 2) hortic horizon (Figure 38). The Aht 1 horizon is
characteristically black in colour and ranges in depth from 30 to 50cm. The Aht 2 horizon
is very dark brown/dark brown and is 30-50cm deep. The combined depth of Aht 1 and
Aht 2 horizons result in topsoils of over 70cm. Both of these horizons contain cultural
debris such as fuel residue, pottery sherds and shell remains. Topsoil within ‘reference’
soil pit WG 5 comprises an Ah horizon which is brown and 27cm deep. In contrast to the
High Street zone no cultural debris was identified within the ‘reference’ soil pit. Soil was
not described for the hinterland north of Wigtown given that the ‘reference’ soil pit is
located within the Hinterland Near zone. Accordingly it is suggested that the ‘reference’
soil pit provides a good indication of soil characteristics within the Hinterland Near zone.
Additionally, soil was not described for the Showfield zone due to problems of restricted
access. Nevertheless, soil descriptions made during the auger survey indicate the
presence of an Ah horizon within the Showfield zone which is typically dark brown in
colour and 15-25cm in depth.

Figure 38: Soil profile description of soil pit WG2 (see appendix 1 for photographs)

- 110 -
5.1.2 Variation between Burghs

Topsoil within the burgh core at Lauder, Pittenweem and Wigtown is characterised by
two hortic soil horizons. These horizons are typically dark (black/dark brown), deep and
contain clearly identifiable cultural debris such as fuel residue and pottery. Hortic topsoils
are limited to the burgh core at Pittenweem and Wigtown; however, they extend to the
immediate hinterland south of Lauder (Hinterland Near).

- 111 -
5.2 Coarse Mineral Material

The abundance of coarse mineral material within topsoil deposits at Lauder, Pittenweem
and Wigtown is summarised in tables presented in Appendix 2. Differences in the nature
and abundance of inclusions are identified and comparisons between zones, horizons
and burghs are made.

5.2.1 Variation within Burghs

5.2.1.1 Zones
Bone, pottery/brick, clinker/slag, mortar/plaster and heated mineral material are present
within the High Street zone at Lauder (<2%). There is considerable variation in the
occurrence of coarse mineral material between soil pits within this zone, for example
mortar/plaster is present within all soil pits whereas bone, pottery/brick and clinker/slag
are confined to soil pit LA 1. Heated mineral is present (<2%) within the Hinterland Near
zone and clinker/slag is identified within the Hinterland Far zone (<2%). Contrastingly
coarse mineral anthropogenic inclusions are absent within the Thirlstane zone and
‘reference’ soil pit LA 7. These results indicate addition of anthropogenic mineral
material to soils within the burgh core and hinterland south of Lauder (Hinterland Near,
Hinterland Far). This finding is significant considering there are no coarse mineral
inclusions within the ‘reference’ soil pit. Moreover, it is recognised that coarse mineral
material is greater in abundance and diversity within the burgh core.

Shell, bone, pottery/brick, clinker/slag, mortar/plaster and heated mineral are present
within topsoils in the Harbour zone at Pittenweem. These materials are typically trace in
abundance (<2%) although abundances of 2-5% are noted for shell and mortar within
soil pit PT 5, and heated mineral within soil pit PT 4. Shell, bone, pottery/brick,
clinker/slag, mortar/plaster and heated mineral are also present in topsoil deposits within
the High Street zone. Shell, bone and clinker/slag are present in trace abundances
(<2%) whereas pottery/brick and heated mineral inclusions vary in abundance from <2%
to 2-5%, and mortar/plaster ranges from <2% to 5-10%. The Harbour and High Street
zones contain the same classes of coarse mineral inclusions; however, shell is more
abundant within the Harbour zone and pottery/brick and mortar/plaster inclusions are
higher in number within the High Street zone.

- 112 -
There is a difference in the nature of coarse mineral material between the Hinterland Far
zone and Hinterland Near zone at Pittenweem, for example only heated mineral material
is present within the Hinterland Near zone (<2%) whereas pottery/brick, clinker/slag,
mortar/plaster and heated mineral are identified within the Hinterland Far zone (<2%).
Additionally, ‘reference’ soil pit PT 1 contains trace abundances (<2%) of pottery/brick,
mortar/plaster and heated mineral. These results indicate addition of anthropogenic
mineral material to soils within the burgh core and hinterland north of Pittenweem
(Hinterland Near, Hinterland Far). Coarse mineral material is most abundant and diverse
within the burgh core although subtle differences in shell, pottery/brick and
mortar/plaster are identified between the Harbour and High Street zones. Moreover, it
should be noted that inclusions of shell and bone are confined to zones within the burgh
core.

Shell, bone, pottery/brick, clinker/slag, mortar/plaster and heated mineral are present
within topsoils in the High Street zone at Wigtown. These materials are consistently
present within all High Street zone soil pits with the exceptions of pottery/brick and
heated mineral which are absent in soil pit WG 4, and mortar/plaster which is absent in
soil pit WG 3. Coarse mineral inclusions are typically trace in abundance (<2%) although
abundances of 2-5% are noted for mortar/plaster in soil pit WG 1 and bone in soil pit WG
4. In contrast coarse mineral inclusions are absent within ‘reference’ soil pit WG 5.
These results indicate addition of mineral material to soils within the burgh core. This
finding is significant given the lack of mineral inclusions within the ‘reference’ soil pit.
Moreover, considering the ‘reference’ soil pit is located within the Hinterland Near zone,
it is argued that coarse mineral inclusions are lacking in the hinterland north of Wigtown.

- 113 -
5.2.1.2 Hortic Horizons

Lauder
Differences in coarse mineral material classes are identified within Aht 1 horizons at
Lauder, for example bone and heated mineral inclusions are identified at 20-28cm within
soil pit LA 1 and pottery/brick and mortar/plaster inclusions are noted at 38-46cm (Figure
39). Likewise pottery/brick and mortar/plaster inclusions are limited to 25-33cm within
soil pit LA 6. Where materials are present throughout Aht 1 horizons their abundances
are shown to differ. Clinker/slag is <1% at 20-28cm within soil pit LA 1 and <2% at 38-
46cm. Similarly, heated mineral inclusions are <1% at 10-18cm within soil pit LA 6 and
<2% at 25-33cm. It is therefore argued that coarse mineral material varies in diversity
and abundance within Aht 1 horizons. Furthermore it is recognised that coarse mineral
material is limited to Aht 1 horizons at Lauder, for instance mineral inclusions are absent
from the Aht 2 horizon within soil pits LA 4, LA 6 and LA 9.

1.2
Bone
1.0 Pottery/Brick
Clinker/Slag
Abundance (%)

0.8 Mortar/Plaster
Heated Mineral
0.6

0.4

0.2

0.0
LA 1 LA 1 LA 4 LA 4 LA 6 LA 6 LA 6 LA 9/1 LA 9/2
20-28cm 38-46cm 9-17cm 42-50cm 10-18cm 25-33cm 57-65cm 10-18 40-48
Aht 1 Aht 1 Aht 1 Aht 2 Aht 1 Aht 1 Aht 2 Aht 1 Aht 2

High Street Hint Near

Figure 39: Percentage abundance estimates of coarse mineral material within upper (Aht
1) and lower (Aht 2) topsoil deposits in the High Street and Hinterland Near zones at
Lauder. Depth indicates the location of Kubiena tin samples

- 114 -
Pittenweem
Differences in coarse mineral material classes are identified within Aht 1 horizons at
Pittenweem, for example shell and bone inclusions are limited to 41-49cm within soil pit
PT 4. Likewise clinker/slag inclusions are exclusive to 62-70cm within soil pit PT 4
(Figure 40). In cases where materials are present throughout Aht 1 horizons it is
recognised that their abundances differ. Shell and mortar/plaster inclusions vary in
abundance from 2-5% at 20-28cm within soil pit PT 5 yet are <2% at 40-48cm. Similarly,
pottery/brick inclusions are <1% at 20-28cm within soil pit PT 5 but are <2% at 40-48cm.
Accordingly, it is suggested that coarse mineral material varies in diversity and
abundance within Aht 1 horizons at Pittenweem.

Differences in coarse mineral material classes are identified between Aht 1 and Aht 2
horizons, for instance bone inclusions are limited to the Aht 1 horizon within soil pit PT 3
and shell inclusions are exclusive to the Aht 2 horizon. In addition, differences in
abundances of coarse mineral anthropogenic material classes are recognised between
Aht 1 and Aht 2 horizons. Pottery/brick is 2-5% abundant within the Aht 1 horizon at soil
pit PT 3 and <2% within the Aht 2 horizon. Conversely mortar/plaster inclusions are <2%
abundant within the Aht 1 horizon at soil pit PT 3 in contrast to 5-10% within the Aht 2
horizon. Hence it is recognised that coarse mineral material varies in diversity and
abundance between Aht 1 and Aht 2 horizons at Pittenweem.

7.0 Shell
Bone
6.0 Pottery/Brick
Clinker/Slag
5.0
Abundance (%)

Mortar/Plaster
4.0 Heated Mineral

3.0

2.0

1.0

0.0
PT 4 PT 4 PT 5 PT 5 PT 3 PT 3 PT 6
41-49 cm 62-70cm 20-28cm 40-48cm 13-21cm 40-48cm 15-23cm
Aht 1 Aht 1 Aht 1 Aht 1 Aht 1 Aht 2 Aht 1

Harbour High Street

Figure 40: Percentage abundance estimates of coarse mineral material within upper (Aht
1) and lower (Aht 2) topsoil deposits in the Harbour and High Street zones at Pittenweem.
Depth indicates the location of Kubiena tin samples

- 115 -
 Wigtown
Differences in coarse mineral material classes are identified within Aht 1 horizons at
Wigtown, for example shell and bone are limited to 32-40cm within soil pit WG 1.
Likewise clinker/slag inclusions are exclusive to 12-20cm within soil pit WG 1 (Figure
41). In cases where materials are present throughout Aht 1 horizons it is recognised that
their abundances differ. Mortar/plaster inclusions vary from 2-5% at 12-20cm within soil
pit WG 1 and are <1% at 32-40cm. Similar to results obtained for Lauder and
Pittenweem, it is recognised that coarse mineral material varies in diversity and
abundance within Aht 1 horizons at Wigtown.

Differences in coarse mineral material classes are identified between Aht 1 and Aht 2
horizons, for instance shell, bone, pottery/brick, clinker/slag, mortar/plaster and heated
mineral inclusions are limited to the Aht 1 horizon within soil pit WG 1. Similarly,
pottery/brick and heated mineral are restricted to the Aht 1 horizon within soil pit WG 2.
In addition, differences in abundances of coarse mineral anthropogenic material classes
are recognised between Aht 1 and Aht 2 horizons. Bone is <1% within the Aht 1 horizon
at soil pit WG 4 in contrast to abundances of 2-5% within the Aht 2 horizon. Coarse
mineral material therefore varies in diversity and abundance between Aht 1 and Aht 2
horizons at Wigtown.

3.5 Shell
Bone
3.0 Pottery/Brick
Clinker/Slag
2.5 Mortar/Plaster
Heated Mineral
Abundance (%)

2.0

1.5

1.0

0.5

0.0
WG 1 WG 1 WG 1 WG 2 WG 2 WG 3 WG 3 WG 4 WG 4
12-20cm 32-40cm 55-63cm 13-21cm 38-46cm 10-18cm 40-48cm 12-20cm 45-53cm
Aht 1 Aht 1 Aht 2 Aht 1 Aht 2 Aht 1 Aht 2 Aht 1 Aht 2

Figure 41: Percentage abundance estimates of coarse mineral material within upper (Aht
1) and lower (Aht 2) topsoil deposits in the High Street zone at Wigtown. Depth indicates
the location of Kubiena tin samples

- 116 -
5.2.2 Variation between Burghs

Coarse mineral anthropogenic material is typically most diverse and abundant within the
burgh core at Lauder, Pittenweem and Wigtown. All three burghs have bone,
pottery/brick, clinker/slag, mortar/plaster and heated mineral inclusions in zones
corresponding to the burgh core (Harbour, High Street). In addition shell inclusions are
identified within the burgh core at Pittenweem and Wigtown.

Coarse mineral anthropogenic material is limited to the burgh core at Wigtown; however,
selected inclusions are present within hinterland topsoils at Lauder and Pittenweem.
Heated mineral and clinker/slag inclusions occur in the immediate hinterland south of
Lauder (Hinterland Near) and far hinterland (Hinterland Far) respectively. Similar to
results obtained for Lauder, topsoils within the immediate hinterland north of Pittenweem
(Hinterland Near) contain trace abundances of heated mineral inclusions. Moreover the
Hinterland Far zone and ‘reference’ soil pit at Pittenweem are characterised by
pottery/brick, clinker/slag and mortar/plaster inclusions. It is suggested that the
‘reference’ soil is of limited utility considering is location within the Hinterland Far zone.

Considerable variation in coarse mineral anthropogenic material is also apparent within

and between topsoil hortic horizons. Coarse mineral anthropogenic material classes vary
in diversity and abundance within Aht 1 horizons and between Aht 1 and Aht 2 horizons
at Lauder, Pittenweem and Wigtown. Differences within and between horizons appear
random with the exception of Lauder where coarse mineral anthropogenic inclusions are
absent from Aht 2 horizons.

- 117 -
5.3 Summary of Results: Coarse Mineral Material

The following section presents a brief summary of key trends in the nature and
abundance of coarse mineral material both within and between burghs. The significance
of these results is discussed in sections 8.1.3 and 8.2.3; in particular consideration is
given to differences in the nature and distribution of waste materials both within and
between burghs.

5.3.1 Variation within Burghs

5.3.1.1 Lauder
The main area of interest at Lauder is the High Street zone. This zone is characterised
by trace abundances of bone, pottery/brick, clinker/slag, mortar/plaster and heated
mineral inclusions. In addition, clinker/slag is present within the Hinterland Far zone and
heated mineral is identified within the Hinterland Near zone (Table 5).

5.3.1.2 Pittenweem
The Harbour, High Street, Hinterland Near and Hinterland Far zones and ‘reference’ soil
pit at Pittenweem contain coarse mineral inclusions. The burgh core is characterised by
shell, bone, pottery/brick, clinker/slag, mortar/plaster and heated mineral inclusions.
Abundances are typically <2% although it is noted that shell is most abundant within the
Harbour zone (2-5%), and pottery/brick and mortar/plaster are most abundant within the
High Street zone (2-5% to 5-10%) (Table 5). Coarse mineral material within the
Hinterland Near zone is limited to trace abundances of heated mineral. However,
inclusions within the Hinterland Far zone and ‘reference’ soil comprise trace abundances
of pottery/brick, clinker/slag, mortar/plaster and heated mineral.

5.3.1.3 Wigtown
The High Street zone is characterised by trace abundances of shell, bone, pottery/brick,
clinker/slag, mortar/plaster and heated mineral inclusions. Coarse mineral inclusions are
absent from the hinterland north of Wigtown.

5.3.1.4 Hortic Horizons

Results presented in section 5.2.1.2 indicate that coarse mineral anthropogenic material
classes vary in diversity and abundance within Aht 1 horizons and between Aht 1 and
Aht 2 horizons at Lauder, Pittenweem and Wigtown. Differences within and between
horizons appear random with the exception of Lauder where coarse mineral
anthropogenic inclusions are absent from Aht 2 horizons.

- 118 -
Table 5: Summary of trends in coarse mineral material identified at Lauder, Pittenweem
and Wigtown, summary abundances are given in brackets

Inclusion Lauder Pittenweem Wigtown

Shell Absent. Shell is present within Shell present is within
the burgh core the burgh core (<2%).
(Harbour <2% to 2-5%,
High Street <2%).
Bone Bone is present within Bone is present within Bone is present within
the burgh core (<2%). burgh core (<2%). the burgh core (<2%
to 2-5%).
Pottery Pottery/Brick is present Pottery/brick is present Pottery/Brick is
within the burgh core within the burgh core present within the
/Brick
(<2%). (Harbour <2%, High burgh core (<2%).
Street <2% to 2-5%)
and far hinterland
(<2%).
Clinker/ Clinker/Slag is present Clinker/Slag is present Clinker/Slag is present
within the burgh core within the burgh core within the burgh core
Slag
and far hinterland and far hinterland (<2%).
(Hinterland Far) (<2%). (<2%).
Mortar/ Mortar/Plaster is Mortar/Plaster is Mortar/Plaster is
present within the present within the present within the
Plaster
burgh core (<2%). burgh core (Harbour burgh core (<2% to 2-
<2% to 2-5%, High 5%)..
Street <2% to 5-10%)
and far hinterland
(<2%).
Heated Heated Mineral is Heated Mineral is Heated Mineral is
present within the present within the present within the
Mineral
burgh core and burgh core (2% to 2- burgh core (<2%).
immediate hinterland 5%) and hinterland
(Hinterland Near) (Hinterland Near,
(<2%). Hinterland Far <2%).

- 119 -
5.3.2 Variation between Burghs

All three burghs have bone, pottery/brick, clinker/slag, mortar/plaster and heated mineral
inclusions in zones corresponding to the burgh core (Harbour, High Street) (Table 6). In
contrast shell inclusions are limited to the burgh core at Pittenweem and Wigtown.
Coarse mineral material is most abundant within burgh cores. Moreover, coarse mineral
inclusions are most abundant within the burgh core at Pittenweem in comparison to
burgh cores at Lauder and Wigtown. Coarse mineral anthropogenic inclusions are
present in topsoils within the hinterland at Lauder and Pittenweem. Mineral material
within the hinterland at Lauder is limited to isolated traces of heated mineral and
clinker/slag; however, pottery/brick, clinker/slag, mortar/plaster and heated mineral
inclusions are consistent within the far hinterland at Pittenweem.

Table 6: Comparison of trends in the abundance of coarse mineral anthropogenic material

at Lauder, Pittenweem and Wigtown, summary abundances are given in brackets

Inclusion Comparison between Burghs

Shell Shell is present within the burgh core at Pittenweem (<2% to 2-5%)
and Wigtown (2%).
Bone Bone is present within the burgh core at all three towns (Lauder and
Pittenweem <2%, Wigtown <2% to 2-5%).
Pottery/Brick Pottery/Brick is present within the burgh core at all three towns (Lauder
and Wigtown <2%, Pittenweem <2% to 2-5%).
Pottery/Brick is present within the Hinterland Far zone and ‘reference’
soil (<2%) at Pittenweem.
Clinker/Slag Clinker/slag is present within the burgh core at all three towns (<2%).
Clinker/slag is present in the Hinterland Far zone at Lauder and
Pittenweem (<2%).
Mortar/Plaster Mortar/plaster is present within the burgh core at all three towns
(Lauder <2%, Pittenweem <2% to 5-10%, Wigtown <2% to 2-5%).
Mortar/plaster is present within the Hinterland Far zone and ‘reference’
soil (<2%) at Pittenweem.
Heated Heated mineral is present within the burgh core at all three towns
(Lauder and Wigtown <2%, Pittenweem <2% to 2-5%).
Mineral
Heated mineral is present in trace abundance (<2%) within the
immediate hinterland (Hinterland Near) at Lauder, and hinterland
(Hinterland Near and Hinterland Far zones/’reference’ soil) at
Pittenweem.

- 120 -
5.4 Fine Mineral Material

The colour and limpidity of fine mineral material within topsoil deposits at Lauder,
Pittenweem and Wigtown is summarised in tables presented in Appendix 2. Differences
in fine mineral material characteristics are identified and comparisons between zones
and burghs are made.

5.4.1 Variation within Burghs

Fine mineral material within the High street, Hinterland Near and Thirlstane zones at
Lauder is brown/dark brown and dotted. In contrast fine mineral material within the
Hinterland Far zone and ‘reference’ soil pit LA 7 is red/red brown and speckled, with the
exception of topsoil within soil pit LA 3 which is brown and dotted.

There is no difference in fine mineral material characteristics between zones at

Pittenweem. Fine mineral material within the Harbour, High Street, Hinterland Near and
Hinterland Far zones and ‘reference’ soil pit PT 1 is characteristically brown/dark brown
and dotted.

Fine mineral material within the High Street zone and ‘reference’ soil WG 5 at Wigtown
is brown/dark brown. Despite similarities in colour, it is recognised that limpidity of fine
mineral material differs between these areas. Fine mineral material within the High
Street zone is dotted in contrast to the ‘reference’ soil which is speckled.

5.4.2 Variation between Burghs

Fine mineral material within the burgh core at all three towns is characteristically
brown/dark brown and dotted. Similarities in the nature of fine mineral material are
identified within the burgh core and immediate hinterland north (Thirlstane) and south
(Hinterland Near) of Lauder. Additionally it is noted that there is no distinction in the
nature of fine mineral material between zones at Pittenweem. Fine mineral material
within the burgh core and hinterland (Hinterland Near/’reference’ soil) at Wigtown is
similar in colour; however, a marked difference in limpidity is apparent.

- 121 -
5.5 Coarse Organic Material

The abundance of coarse organic material within topsoil deposits at Lauder, Pittenweem
and Wigtown is summarised in tables presented in Appendix 2. Differences in the nature
and abundance of inclusions are identified and comparisons between zones, horizons
and burghs are made.

5.5.1 Charcoal

5.5.1.1 Variation within Burghs

Charcoal is present in trace abundances (<2%) within the High Street and Hinterland
Near zones at Lauder. In contrast charcoal is absent from the Hinterland Far and
Thirlstane zones, and ‘reference’ soil pit LA 7. Charcoal is therefore limited to the burgh
core and immediate hinterland south of the historical burgh limits (Hinterland Near).

Charcoal inclusions are <2% within the Harbour, High Street, Hinterland Near and
Hinterland Far zones at Pittenweem. In comparison charcoal is absent from ‘reference’
soil pit PT 1. Nevertheless these results indicate that there is no difference in the
abundance of charcoal between zones at Pittenweem.

Charcoal occurs as trace abundances (<2%) within the High Street zone at Wigtown.
However charcoal is absent from ‘reference’ soil pit WG 5. Charcoal inclusions are thus
limited to the burgh core at Wigtown.

5.5.1.2 Variation between Burghs

Charcoal is <2% within zones corresponding to the burgh core (Harbour and High
Street) at all three towns. Results of Kruskal-Wallis analysis with Dunn’s test multiple
comparisons indicate no statistical difference in the abundance of charcoal between
High Street zones at Lauder, Pittenweem and Wigtown. Charcoal inclusions are present
within the immediate hinterland south (Hinterland Near) of the historical burgh limits at
Lauder and across the hinterland (Hinterland Near, Hinterland Far) at Pittenweem. In
contrast charcoal is limited to the burgh core at Wigtown.

- 122 -
5.5.2 Fuel Residue 1 (FR 1)

5.5.2.1 Variation within Burghs

FR 1 inclusions are identified within the High Street, Hinterland Near, Hinterland Far and
Thirlstane zones, and ‘reference’ soil pit LA 7 at Lauder. Abundances of FR 1 within the
High Street zone range from <2% to 5-10%. Similarly FR 1 varies between <2% to 2-5%
within the Hinterland Near zone. FR 1 inclusions are <2% within the Hinterland Far and
Thirlstane zones and ‘reference’ soil pit. FR 1 is significantly more abundant within the
High Street zone than the Hinterland Near, Hinterland Far and Thirlstane zones, and
‘reference’ soil pit LA 7 (p<0.001) (Figure 42a). In addition FR 1 is significantly greater
within the Hinterland Near zone than the Hinterland Far zone (p<0.01). There is no
statistical difference in the abundance of FR 1 between the Hinterland Far and Thirlstane
zones and ‘reference’ soil pit. FR 1 inclusions are therefore greatest within the burgh
core followed in succession by the immediate hinterland south (Hinterland Near) of
Lauder.

a) b)
Pairw ise Comparisons (FR 1) Pairw ise Comparisons (FR 1)

High Street Hint Near Harbour High Street

Hint Far Hint Near

Thirlstane Hint Far

Ref erence Ref erence

Hint Near Hint Far High Street Hint Near

Thirlstane Hint Far

Ref erence Ref erence

Hint Far Thirlstane Hint Near Hint Far

Ref erence Ref erence

Thirlstane Ref erence Hint Far Ref erence

-Z 0 Z -Z 0 Z
|Bonf erroni Z-v alue|: 2.326 |Bonf erroni Z-v alue|: 2.326

Figure 42: Kruskal-Wallis analysis with Dunn’s test multiple comparisons of FR 1 for
individual zones at Lauder (a) and Pittenweem (b), z values exceeding –Z or Z are
significant at p<0.05 confidence level

- 123 -
FR 1 inclusions are present within the Harbour, High Street, Hinterland Near and
Hinterland Far zones, and ‘reference’ soil pit PT 1 at Pittenweem. Abundances of FR 1
within the Harbour and High Street zones range from 5-10% to 10-15%. FR 1 varies
between 5 to 10% within the Hinterland Near and Hinterland Far zones, and between 2
to 5% within the ‘reference’ soil. FR 1 is significantly more abundant within the Harbour
and High Street zones than the Hinterland Near zone and ‘reference’ soil pit (p<0.001)
(Figure 42b). In addition FR 1 is significantly greater within the Hinterland Far zone than
the Hinterland Near zone (p<0.01) and ‘reference’ soil pit (p<0.001). There is no
statistical difference in the abundance of FR 1 between the Harbour, High Street and
Hinterland Far zones. FR 1 inclusions are therefore greatest within the burgh core at far
hinterland (Hinterland Far) at Pittenweem.

FR 1 inclusions occur within the High Street zone and ‘reference’ soil pit WG 5 at
Wigtown. Abundances of FR 1 within the High Street zone range from <2% to 5-10%,
although values of 2-5% to 5-10% are typical. FR 1 varies between <2% to 2-5% within
the ‘reference’ soil pit. FR 1 is significantly greater within the High Street zone than the
‘reference’ soil pit (p<0.001). FR 1 inclusions are therefore greatest within the burgh core
at Wigtown.

5.5.2.2 Variation between Burghs

FR 1 inclusions are present within zones corresponding to the burgh core (Harbour and
High Street) and the hinterland at all three towns. FR 1 inclusions are greatest in
abundance within the burgh core at Lauder and Wigtown and within the burgh core and
far hinterland (Hinterland Far) at Pittenweem. Results of Kruskal-Wallis analysis with
Dunn’s test multiple comparisons indicate that FR 1 is significantly more abundant within
the High Street zone at Pittenweem than High Street zones at Lauder and Wigtown
(p<0.001). There is no statistical difference in the abundance of FR 1 between the High
Street zone at Lauder and Wigtown.

- 124 -
5.5.3 Fuel Reside 2 (FR 2)

5.5.3.1 Variation within Burghs

FR 2 inclusions are <2% within the High Street zone at Lauder. In contrast FR 2
inclusions are absent from the Hinterland Near, Hinterland Far and Thirlstane zones,
and ‘reference’ soil pit LA 7. FR 2 inclusions are therefore limited to the burgh core at
Lauder.

FR 2 inclusions occur as trace abundances (<2%) within the Harbour, High Street and
Hinterland Far zones, and ‘reference’ soil pit PT 1 at Pittenweem. FR 2 is absent from
the Hinterland Near zone. There is no difference in the abundance of FR 2 inclusions
between the burgh core and far hinterland (Hinterland Far/’reference’ soil) at
Pittenweem.

FR 2 inclusions are present in trace abundances (<2%) within the High Street zone at
Wigtown. In comparison FR 2 is absent from ‘reference’ soil pit WG 5. FR 2 inclusions
are therefore limited to the burgh core at Wigtown.

5.5.3.2 Variation between Burghs

FR 2 inclusions occur within zones corresponding to the burgh core (Harbour and High
Street) at all three towns and within the hinterland at Pittenweem. Results of Kruskal-
Wallis analysis with Dunn’s test multiple comparisons indicate no statistical difference in
the abundance of FR 2 between High Street zones at Lauder, Pittenweem and Wigtown.
FR 2 inclusions are limited to the burgh core at Lauder and Wigtown; however, FR 2 is
noted within both the burgh core and far hinterland (Hinterland Far/’reference’ soil) at
Pittenweem.

- 125 -
5.5.4 Fuel Residue 3 (FR 3)

5.5.4.1 Variation within Burghs

Abundances of FR 3 within the High Street zone at Lauder range from <2% to 2-5%. FR
3 inclusions are <2% within the Hinterland Near and Hinterland Far zones and
‘reference’ soil pit. FR 3 is significantly more abundant within the High Street zone than
the Hinterland Far (p<0.01) and Thirlstane (p<0.001) zones and ‘reference’ soil pit LA 7
(p<0.01) (Figure 43a). There is no statistical difference in the abundance of FR 3
between the High Street and Hinterland Near zones. In addition there is no significant
difference in FR 3 between the Hinterland Far and Thirlstane zones, and ‘reference’ soil
pit. FR 3 inclusions are therefore greatest within the burgh core followed in succession
by the immediate hinterland south (Hinterland Near) of Lauder.

Abundances of FR 3 within the Harbour and High Street zones at Pittenweem vary
between <2% and 2-5%. FR 3 inclusions are <2% within the Hinterland Near and
Hinterland Far zones and ‘reference’ soil pit. FR 3 is significantly more abundant within
the High Street zone than the Harbour (p<0.05), Hinterland Near (p<0.01) and
Hinterland Far zones (p<0.01), and ‘reference’ soil pit (p<0.001) (Figure 43b). There is
no statistical difference in the abundance of FR 3 between the Harbour, Hinterland Near
and Hinterland Far zones and ‘reference’ soil pit. FR 3 inclusions are therefore greatest
within the burgh core with specific reference to the High Street zone.

FR 3 inclusions occur as trace abundances (<2%) within the High Street zone at
Wigtown. In contrast FR 3 is absent from ‘reference’ soil pit WG 5. FR 3 inclusions are
therefore limited to the burgh core at Wigtown.

- 126 -
 a) b)
Pairw ise Comparisons (FR 3) Pairw ise Comparisons (FR 3)

High Street Hint Near Harbour High Street

Hint Far Hint Near

Thirlstane Hint Far

Ref erence Ref erence

Hint Near Hint Far High Street Hint Near

Thirlstane Hint Far

Ref erence Ref erence

Hint Far Thirlstane Hint Near Hint Far

Ref erence Ref erence

Thirlstane Ref erence Hint Far Ref erence

-Z 0 Z -Z 0 Z
|Bonf erroni Z-v alue|: 2.326 |Bonf erroni Z-v alue|: 2.326

Figure 43: Kruskal-Wallis analysis with Dunn’s test multiple comparisons of FR 3 for
individual zones at Lauder (a) and Pittenweem (b), z values exceeding –Z or Z are
significant at p<0.05 confidence level

5.5.4.2 Variation between Burghs

FR 3 inclusions are present within zones corresponding to the burgh core (Harbour and
High Street) at all three towns and within the hinterland at Lauder and Pittenweem. FR 3
inclusions are greatest in abundance within the burgh core (High Street zone) at Lauder,
Pittenweem and Wigtown. Results of Kruskal-Wallis analysis with Dunn’s test multiple
comparisons indicate that FR 3 is more abundant within the High Street zone at
Pittenweem in comparison to High Street zones at Lauder (p<0.05) and Wigtown
(p<0.001). FR 3 is also significantly greater within the High Street zone at Lauder than
the High Street zone at Wigtown (p<0.01). It is therefore recognised that FR 3 is greatest
in abundance within the High Street zone at Pittenweem and least abundant within the
High Street zone at Wigtown.

- 127 -
5.5.5 Fuel Residue 4 & 6 (FR 4 & 6)

5.5.5.1 Variation within Burghs

Abundances of FR 4 & 6 within the High Street zone at Lauder range from <2% to 5-
10%, although values between 2-5% and 5-10% are typical. FR 4 & 6 inclusions within
the Hinterland Near and Hinterland Far zones vary from <2% to 2-5%. In contrast FR 4 &
6 is <2% within the Thirlstane zone and ‘reference’ soil. FR 4 & 6 is significantly more
abundant within the High Street zone than the Hinterland Near, Hinterland Far and
Thirlstane zones and ‘reference’ soil pit (p<0.001) (Figure 44a). There is no statistical
difference in the abundance of FR 4 & 6 between the Hinterland Near, Hinterland Far,
and Thirlstane zones and ‘reference’ soil pit. FR 4 & 6 inclusions are therefore greatest
within the burgh core.

Abundances of FR 4 & 6 within the Harbour zone at Pittenweem range from 2-5% to 5-
10%. FR 4 & 6 inclusions within the High Street and Hinterland Far zones vary between
<2% to 2-5%. In contrast FR 4 & 6 is <2% within the Hinterland Near zone and
‘reference’ soil pit. FR 4 & 6 is significantly more abundant within the Harbour zone than
the Hinterland Near (p<0.01) and Hinterland Far (p<0.05) zones and reference soil
(P<0.001) (Figure 44b). In addition FR 4 & 6 is greater within the High Street and
Hinterland Far zones than the ‘reference’ soil pit (p<0.001). There is no statistical
difference in the abundance of FR 4 & 6 between the Harbour and High Street zones.
Moreover there is no significant difference in the abundance of FR 4 & 6 between the
High Street, Hinterland Near and Hinterland Far zones. FR 4 & 6 inclusions are therefore
greatest within the burgh core with specific reference to the Harbour zone.

Abundances of FR 4 & 6 within the High Street zone at Wigtown range from 5-10% to
10-15%. FR 4 & 6 inclusions within the ‘reference’ soil pit vary between <2% and 2-5%.
FR 4 & 6 is significantly more abundant within the High Street zone than the ‘reference’
soil pit (p<0.001). FR 4 & 6 inclusions are therefore greatest within the burgh core at
Wigtown.

- 128 -
 a) b)
Pairw ise Comparisons (FR 4 & 6) Pairw ise Comparisons (FR 4 & 6)

High Street Hint Near Harbour High Street

Hint Far Hint Near

Thirlstane Hint Far

Ref erence Ref erence

Hint Near Hint Far High Street Hint Near

Thirlstane Hint Far

Ref erence Ref erence

Hint Far Thirlstane Hint Near Hint Far

Ref erence Ref erence

Thirlstane Ref erence Hint Far Ref erence

-Z 0 Z -Z 0 Z
|Bonf erroni Z-v alue|: 2.326 |Bonf erroni Z-v alue|: 2.326

Figure 44: Kruskal-Wallis analysis with Dunn’s test multiple comparisons of FR 4 & 6 for
individual zones at Lauder (a) and Pittenweem (b), z values exceeding –Z or Z are
significant at p<0.05 confidence level

5.5.5.2 Variation between Burghs

FR 4 & 6 inclusions are identified within zones corresponding to the burgh core (Harbour
and High Street) and selected hinterland zones all three towns. It is noted that FR 4 & 6
inclusions are greatest in abundance within burgh cores at Lauder, Pittenweem and
Wigtown. Results of Kruskal-Wallis analysis with Dunn’s test multiple comparisons
indicate FR 4 & 6 is significantly more abundant within the High Street zone at Wigtown
in comparison to High Street zones at Lauder and Pittenweem (p<0.001). There is no
statistical difference in the abundance of FR 4 & 6 between the High Street zone at
Lauder and Pittenweem.

- 129 -
5.5.6 Fuel Residue 5 & 7 (FR 5 & 7)

5.5.6.1 Variation within Burghs

Abundances of FR 5 & 7 within the High Street, Hinterland Near, Hinterland Far and
Thirlstane zones and ‘reference’ soil pit LA 7 at Lauder are <2%. Hence, there is no
significant difference in the abundance of FR 5 & 7 between zones.

Abundances of FR 5 & 7 within the Harbour zone at Pittenweem range from 5-10% to
10-15%. In contrast FR 5 & 7 inclusions within the High Street, Hinterland Near and
Hinterland Far zones and ‘reference’ soil pit are <2%. FR 5 & 7 is significantly more
abundant within the Harbour zone than the High Street, Hinterland Near and Hinterland
Far zones and ‘reference’ soil pit (p<0.001) (Figure 45). In addition FR 5 & 7 is
significantly greater within the High Street zone compared to the ‘reference’ soil pit
(p<0.01). It is therefore identified that FR 5 & 7 inclusions are greatest within the
Harbour zone and to a lesser extent the High Street zone.

FR 5 & 7 occur as trace abundances (<2%) within the High Street zone and ‘reference’
soil pit WG 5 at Wigtown. There is no statistical difference in the abundance of FR 5 & 7
between the High Street zone and ‘reference’ soil pit.

Pairw ise Comparisons (FR 5 & 7)

Harbour High Street

Hint Near

Hint Far

Ref erence

High Street Hint Near

Hint Far

Ref erence

Hint Near Hint Far

Ref erence

Hint Far Ref erence

-Z 0 Z
|Bonf erroni Z-v alue|: 2.326

Figure 45: Kruskal-Wallis analysis with Dunn’s test multiple comparisons of FR 5 & 7 for
individual zones at Pittenweem, z values exceeding –Z or Z are significant at p<0.05
confidence level

- 130 -
5.5.6.2 Variation between Burghs
FR 5 & 7 inclusions are present within both the burgh core and hinterland at all three
burghs. There is no difference in the abundance of FR 5 & 7 between zones at Lauder
and Wigtown; however it is noted that FR 5 & 7 is greatest in abundance within the
burgh core at Pittenweem. Results of Kruskal-Wallis analysis with Dunn’s test multiple
comparisons reveal that FR 5 & 7 is significantly more abundant within the High Street
zone at Pittenweem than High Street zones at Lauder and Wigtown (p<0.001). There is
no statistical difference in the abundance of FR 5 & 7 between High Street zones at
Lauder and the High Street zone at Wigtown.

5.5.7 Fuel Residue 8 & 9 (FR 8 & 9)

5.5.7.1 Variation within Burghs

Abundances of FR 8 & 9 within the High Street and Hinterland Near zone at Lauder
range from <2% to 2-5%. FR 8 & 9 inclusions within the Hinterland Far and Thirlstane
zones and ‘reference’ soil pit are typically <2%. FR 8 & 9 is significantly more abundant
within the High Street zone than the Hinterland Far and Thirlstane zones and ‘reference’
soil pit (p<0.001) (Figure 46a). In addition FR 8 & 9 is significantly greater within the
Hinterland Near zone than the Hinterland Far and Thirlstane zones (p<0.05) and
‘reference’ soil pit (p<0.001). FR 8 & 9 is also significantly higher in the Hinterland Far
and Thirlstane zones than ‘reference’ soil pit (p<0.01). There is no difference in the
abundance of FR 8 & 9 between the High Street and Hinterland Near zones. FR 8 & 9 is
greatest in abundance within the burgh core and immediate hinterland south (Hinterland
Near) of Lauder.

Abundances of FR 8 & 9 within the Harbour and High Street zone at Pittenweem range
from <2% to 2-5%, although abundances within the Harbour zone are typically 2-5%. FR
8 & 9 inclusions within the Hinterland Far zone and ‘reference’ soil pit range from <2% to
2-5%. In contrast FR 8 & 9 is <2% within the Hinterland Near zone. FR 8 & 9 is
significantly more abundant within the Harbour than the High Street (p<0.01), Hinterland
Near (p<0.001) and Hinterland Far zones (p<0.001) and ‘reference’ soil pit (p<0.001).
FR 8 & 9 is significantly greater within the High Street zone than the ‘reference’ soil pit
(p<0.05) (Figure 46b). It is therefore recognised that FR 8 & 9 inclusions are greatest
within the Harbour zone and to a lesser extent the High Street zone.

- 131 -
Abundances of FR 8 & 9 within the High Street zone at Wigtown range from <2% to 5-
10%. In contrast FR 8 & 9 within the ‘reference’ soil is typically <2%. FR 8 & 9 is
significantly more abundant within the High Street zone than the ‘reference’ soil
(p<0.001). FR 8 & 9 inclusions are, therefore, greatest within the burgh core.

5.5.7.2 Variation between Burghs

The burgh core and hinterland at all three burghs contain FR 8 & 9 inclusions. FR 8 & 9
is greatest in abundance within the burgh core at Pittenweem and Wigtown and within
the burgh core and immediate hinterland south (Hinterland Near) of the historical burgh
limits at Lauder. Results of Kruskal-Wallis analysis with Dunn’s test multiple
comparisons indicate that FR 8 & 9 is significantly more abundant within the High Street
zone at Lauder in comparison to High Street zones at Pittenweem and Wigtown
(p<0.001). In addition FR 8 & 9 is significantly greater within the High Street zone at
Wigtown than the High Street zone at Pittenweem (p<0.001).

a) b)
Pairw ise Comparisons (FR 8 & 9) Pairw ise Comparisons (FR 8 & 9)

High Street Hint Near Harbour High Street

Hint Far Hint Near

Thirlstane Hint Far

Ref erence Ref erence

Hint Near Hint Far High Street Hint Near

Thirlstane Hint Far

Ref erence Ref erence

Hint Far Thirlstane Hint Near Hint Far

Ref erence Ref erence

Thirlstane Ref erence Hint Far Ref erence

-Z 0 Z -Z 0 Z
|Bonf erroni Z-v alue|: 2.326 |Bonf erroni Z-v alue|: 2.326

Figure 46: Kruskal-Wallis analysis with Dunn’s test multiple comparisons of FR 8 & 9 for
individual zones at Lauder (a) and Pittenweem (b), z values exceeding –Z or Z are
significant at p<0.05 confidence level

- 132 -
5.5.8 Fuel Residue 10 (FR 10)

5.5.8.1 Variation within Burghs

Trace abundances (<2%) of FR 10 inclusions are present within the High Street zone at
Lauder. In contrast FR 10 inclusions are absent from the Hinterland Near, Hinterland Far
and Thirlstane zones, and ‘reference’ soil pit LA 7. FR 10 inclusions are therefore limited
to the burgh core at Lauder.

FR 10 inclusions occur as trace abundances (<2%) within the Harbour, High Street and
Hinterland Far zones, and ‘reference’ soil pit PT 1 at Pittenweem. FR 10 is absent from
the Hinterland Near zone. There is no difference in the abundance of FR 10 inclusions
between the burgh core and far hinterland (Hinterland Far/’reference’ soil) at
Pittenweem.

The High Street zone at Wigtown contains trace abundances (<2%) of FR 10 inclusions.
However FR 10 inclusions are absent from ‘reference’ soil pit WG 5. FR 10 inclusions
are therefore limited to the burgh core at Wigtown.

5.5.8.2 Variation between Burghs

FR 10 inclusions are present within zones corresponding to the burgh core (Harbour and
High Street) at all three towns and within the hinterland at Pittenweem. Results of
Kruskal-Wallis analysis with Dunn’s test multiple comparisons indicate no statistical
difference in the abundance of FR 10 between High Street zones at Lauder, Pittenweem
and Wigtown. It is recognised that FR 10 inclusions are limited to the burgh core at
Lauder and Wigtown; however, FR 10 is noted within both the burgh core and far
hinterland (Hinterland Far/’reference’ soil) at Pittenweem.

- 133 -
 5.5.9 Fuel Residue 10-255 µm (%)

5.5.9.1 Variation within Burghs

The percentage of total fuel residue sized between 10 and 255µm within the High Street
zone at Lauder ranges from 10 to 18% (Figure 47). In comparison the % of total fuel
residue 10-255µm within Hinterland Near, Hinterland Far and Thirlstane zone varies
between 27 to 59%. Similarly, the % of total fuel residue 10-255µm within the ‘reference’
soil pit LA 7 varies between 28 and 52%. Results of One Way Analysis of Variance
(ANOVA) with Tukey-Kramer multiple comparisons indicate that the % of total fuel
residue 10-255µm is significantly lower within the High Street zone in comparison to the
Hinterland Near, Hinterland Far and Thirlstane zones and ‘reference’ soil pit LA 7
(p<0.001). The % of total fuel residue 10-255µm is therefore lowest within the burgh core
at Lauder.

70

60

50
% FR (10-255µm)

40

30

20

10

0
LA 1 LA 1 LA 4 LA 4 LA 6 LA 6 LA 6 LA 9 LA 9 LA 2 LA 2 LA 3 LA 5 LA 7 LA 7 LA 7
20-28 38-46 9-17 42-50 10-18 25-33 57-65 10-18 40-48 4-12 46-54 15-23 8-16 2-22 2-23 2-24
Aht 1 Aht 1 Aht 1 Aht 2 Aht 1 Aht 1 Aht 2 Aht 1 Aht 2 Ah 1 Ah 2 Ah Ah Ah Ah Ah

High Street Hint Near Hint Far Thirl. Reference

Figure 47: % FR (10-255µm) within topsoil deposits in the High Street, Hinterland Near,
Hinterland Far and Thirlstane zones, and ‘reference’ soil at Lauder. Number range refers to
depth of Kubiena sample (cm)

- 134 -
 The percentage of total fuel residue sized between 10 and 255µm within the Harbour
and High street zone at Pittenweem varies between 4 and 16% (Figure 48). In
comparison the % of total fuel residue 10-255µm within the Hinterland Near and
Hinterland Far zones ranges from 21 to 26% with the exception of soil pit PT 2 which
has a very low value of 7.7%. The % of total fuel residue 10- 255µm within ‘reference’
soil pit PT 1 has a mean value of 19%. Results of One Way Analysis of Variance
(ANOVA) reveal that there is no statistical difference in the % of total fuel 10-255µm
between zones at Pittenweem.

The mean percentage of total fuel residue sized between 10 and 255µm within the High
Street zone and ‘reference’ soil pit WG 5 at Wigtown is 12 and 18% respectively (Figure
49). These results indicate a lower % of total fuel residue 10-255µm within the burgh
core. However, results of One Way ANOVA reveal that there is no significant difference
in the % of total fuel residue 10-255µm between zones at Wigtown.

30

25

20
% FR (10-255µm)

15

10

0
PT 4 PT 4 PT 5 PT 5 PT 3 PT 3 PT 6 PT 9 PT 2 PT 7 PT 8 PT 1 PT 1 PT 1
41-49 62-70 20-28 40-48 13-21 40-48 15-23 10-18 22-30 18-24 11-19 6-14 6-14 6-14
Aht 1 Aht 1 Aht 1 Aht 1 Aht 1 Aht 2 Aht 1 Ah Ah Ah Ah Ah Ah Ah

Harbour High Street Hint Near Hint Far Reference

Figure 48: % FR (10-255µm) within topsoil deposits in the Harbour, High Street, Hinterland
Near, and Hinterland Far zones, and ‘reference’ soil at Pittenweem. Number range refers to
depth of Kubiena sample (cm)

- 135 -
 30

25

20
% FR (10-255µm)

15

10

0
WG 1 WG 1 WG 1 WG 2 WG 2 WG 3 WG 3 WG 4 WG 4 WG 5 WG 5 WG 5
12-20 32-40 55-63 13-21 38-46 10-18 40-48 12-20 45-53 5-13 10-18 12-20
Aht 1 Aht 1 Aht 2 Aht 1 Aht 2 Aht 1 Aht 2 Aht 1 Aht 2 Ah Ah Ah

High Street Reference

Figure 49: % FR (10-255µm) within topsoil deposits in the Harbour, High Street, Hinterland
Near, and Hinterland Far zones, and ‘reference’ soil at Pittenweem. Number range refers to
depth of Kubiena sample (cm)

5.5.9.2 Variation between Burghs

The percentage of total fuel residue sized between 10 and 255µm is variable within all
three burghs ranging from 10.6 to 58.6% at Lauder, 4 to 26.5% at Pittenweem and 4.7 to
28.5% at Wigtown. The % of total fuel residue 10-255µm is significantly lower within the
burgh core at Lauder; however, there is no difference in % FR 10-255µm between zones
at Pittenweem and Wigtown. Additionally results of One Way ANOVA indicate that there
is no statistical difference in the % of total fuel residue 10-255µm between High Street
zone at Lauder, Pittenweem and Wigtown.

- 136 -
5.5.10 Micromorphological Characteristics of Hortic Horizons

5.5.10.1 Variation within Burghs

Lauder
Similarities in fuel residue classes exist within Aht 1 horizons at Lauder, for example FR
1, FR 3, FR 4 & 6 and FR 8 & 9 are present at 20-28cm and 38-46cm within soil pit LA 1
(Figure 50). Likewise, FR 1, FR 3, FR 4 & 6 and FR 8 & 9 are recognised at 10-18cm
and 25-33cm within soil pit LA 6. It is noted that abundances of fuel residue classes are
consistent throughout Aht 1 horizons. The total abundance of fuel residue at 20-28cm
and 28-46cm within soil pit LA 1 is 15 and 15.5% respectively. Moreover, the total
abundance of fuel residue within soil pit LA 6 is 14% at successive depths. Aside from
trace abundances of infrequent fuel reside classes such as FR 2 and FR 10, it is argued
variation in the diversity and abundance of fuel residue classes within upper topsoil
hortic horizons is minimal.

20 Charcoal
18 FR 1
FR 2
16
FR 3
14
% Abundance

FR 4 & 6
12 FR 5 & 7
10 FR 8 & 9
8 FR 10
6
4
2
0
LA 1 LA 1 LA 4 LA 4 LA 6 LA 6 LA 6 LA 9 LA 9
20-28 38-46 9-17 42-50 10-18 25-33 57-65 10-18 40-48
Aht 1 Aht 1 Aht 1 Aht 2 Aht 1 Aht 1 Aht 2 Aht 1 Aht 2

High Street Hint Near

Figure 50: Percentage abundance estimates of coarse organic material (Fuel Residue
categories) within upper (Aht 1) and lower (Aht 2) topsoil deposits in the High Street and
Hinterland Near zones at Lauder. Number range indicates the depth (cm) of Kubiena tin
samples

- 137 -
Similarities in fuel residue classes are noted between Aht 1 and Aht 2 horizons, for
example charcoal, FR1, FR 3, FR 4 & 6, FR 5 & 7 and FR 8 & 9 are present within both
the Aht 1 and Aht 2 horizon within soil pit LA 4. Likewise FR1, FR 3, FR 4 & 6 and FR 8
& 9 are present within the Aht 1 and Aht 2 horizon in soil pit LA 6. Fuel residue classes
are less abundant within Aht 2 horizons than Aht 1 horizons. The total abundance of fuel
residue within the Aht 1 and Aht 2 horizon in soil pit LA 6 is 13.1 and 10.4% respectively.
Similarly, the total abundance of fuel residue within the Aht 1 horizon in soil pit LA 9 is
6.4% in contrast to 3.7% within the Aht 2 horizon. It is argued that variation in the
diversity of fuel residue classes between upper and lower topsoil hortic horizons is
minimal. Furthermore, is it suggested that fuel residue is less abundant within the lower
topsoil hortic horizon.

Pittenweem
Similarities in fuel residue classes are identified within Aht 1 horizons at Pittenweem, for
instance FR 1, FR 2, FR 4 & 6, FR 5 & 7 and FR 8 & 9 are present at 41-19 and 62-
70cm within soil pit PT 4. Similarly charcoal, FR 1, FR 3, FR 4 & 6, FR 5 & 7 and FR 8 &
9 and FR 10 are recognised at 20-28 and 40-48cm within soil pit PT 5 (Figure 51). It is
identified that abundances of fuel residue classes typically vary within Aht 1 horizons, for
example at FR 5 & 7 (5-10%) is the most abundant class at 20-28cm within soil pit PT 5
yet FR 1 (10-15%) is the most abundant class at 40-48cm. Additionally it is noted that
the total abundance of fuel residue within Aht 1 horizons is higher at lower depths.

The total abundance of fuel residue at 20-28cm and 40-48cm within soil pit PT 5 is 21.8
and 25.4% respectively. Similarly the total abundance of fuel residue at 41-49cm within
soil pit PT 4 is 17% in contrast to 27% at 62-70cm. It is argued variation in the diversity
of fuel residue classes within upper topsoil hortic horizons is minimal. Moreover, it is
recognised that abundances of fuel residue classes are characteristically varied within
upper topsoil hortic horizons; however, total fuel reside abundances are higher at lower
depths.

- 138 -
 Charcoal
30
FR 1
25 FR 2
FR 3
% Abundance

20 FR 4 & 6
FR 5 & 7
15 FR 8 & 9
FR 10
10

0
PT 4 PT 4 PT 5 PT 5 PT 3 PT 3 PT 6
41-49 62-70 20-28 40-48 13-21 40-48 15-23
Aht 1 Aht 1 Aht 1 Aht 1 Aht 1 Aht 2 Aht 1

Harbour High Street

Figure 51: Percentage abundance estimates of coarse organic material (Fuel Residue
categories) within upper (Aht 1) and lower (Aht 2) topsoil deposits in the Harbour and High
Street zones at Pittenweem. Number range indicates the depth (cm) of Kubiena tin
samples

Similarities in fuel residue classes are noted between Aht 1 and Aht 2 horizons, for
example FR 1, FR 2, FR 4 & 6, FR 5 & 7 and FR 8 & 9 are present within both the Aht 1
and Aht 2 horizon in soil pit PT 3. It is identified that abundances of fuel residue classes
typically vary between Aht 1 and Aht 2 horizons. However, it is recognised that the
abundance of individual fuel residue classes varies between horizons, for example FR 3
ranges from 2-5% within the Aht 1 horizon at soil pit 3 to <1% within the Aht 2 horizon
(Figure 51). Likewise FR 8 & 9 varies from 2-5% within the Aht 1 horizon to <2% within
the Aht 2 horizon at soil pit PT 3. Despite differences in the abundance of individual fuel
residue classes between the Aht 1 and Aht 2 horizons at soil pit PT 3, the total
abundance of fuel residue of both horizons is consistent at 14%. It is argued that
variation in the diversity of fuel reside classes between upper and lower topsoil horizons
is minimal. Moreover, it is recognised that abundances of fuel residue classes are
characteristically varied between upper and lower topsoil hortic horizons.

- 139 -
Wigtown
Similarities in fuel residue classes exist within Aht 1 horizons at Wigtown, for example
FR 1, FR 2, FR 3, FR 4 & 6, FR 5 & 7 and FR 8 & 9 are present at both 12-20 and 32-
40cm within soil pit WG 1. It is noted that abundances of fuel residue classes are
consistent throughout Aht 1 horizons. Abundances of FR 1 (5-10%), FR 4 & 6 (10-15%)
and FR 8 & 9 (2-5%) are the same at both sample depths within the Aht 1 horizon at soil
pit WG 1 (Figure 52). In addition, the total abundance of fuel residue at 12-20cm and 32-
40cm within soil pit WG 1 is 27 and 25.2% respectively. It is therefore concluded that
variation in the diversity and abundance of fuel residue classes within upper topsoil
hortic horizons is minimal.

30 Charcoal
FR 1
25 FR 2
FR 3
FR 4 & 6
20
% Abundance

FR 5 & 7
FR 8 & 9
15 FR 10

10

0
WG 1 WG 1 WG 1 WG 2 WG 2 WG 3 WG 3 WG 4 WG 4
12-20 32-40 55-63 13-21 38-46 10-18 40-48 12-20 45-53
Aht 1 Aht 1 Aht 2 Aht 1 Aht 2 Aht 1 Aht 2 Aht 1 Aht 2

Figure 52: Percentage abundance estimates of coarse organic material (Fuel Residue
categories) within upper (Aht 1) and lower (Aht 2) topsoil deposits in the High Street zone
at Wigtown. Number range indicates the depth (cm) of Kubiena tin samples

- 140 -
Similarities in fuel residue classes are noted between Aht 1 and Aht 2 horizons, for
example FR 1, FR 3, FR 4 & 6, and FR 8 & 9 are present within the Aht 1 and Aht 2
horizon at soil pit WG 3. Similarly FR 1, FR 4 & 6, FR 5 & 7 and FR 8 & 9 occur within
Aht 1 and Aht 2 horizons at soil pit WG 4. Fuel residue classes are less abundant within
Aht 2 horizons than Aht 1 horizons. The total abundance of fuel residue within the Aht 1
and Aht 2 horizon in soil pit WG 1 is 26.1 and 7.3% respectively. Similarly, the total
abundance of fuel residue within the Aht 1 horizon in soil pit WG 3 is 15.2% in contrast
to 5.9% within the Aht 2 horizon. However this trend is not observed at soil pit WG 2, for
instance the total % abundance of fuel residue within the Aht 1 and Aht 2 horizon is
19.6% and 19.4%. Nevertheless variation in the diversity of fuel residue classes between
upper and lower topsoil hortic horizons is minimal. Furthermore, is it suggested that fuel
residue is less abundant within the lower topsoil hortic horizon with the exception of soil
pit WG 2.

5.5.10.2 Variation between Burghs

There is minimal variation in the nature of coarse organic anthropogenic material classes
throughout Aht 1 horizons at Lauder, Pittenweem and Wigtown. Abundances of coarse
organic anthropogenic materials at Lauder and Wigtown are consistent throughout Aht 1
horizons. Conversely, the abundance of coarse organic anthropogenic materials within
Aht 1 horizons at Pittenweem is characteristically varied. In addition total fuel residue
abundances are higher at lower depths within Aht 1 horizons at Pittenweem. Similar
classes of coarse organic anthropogenic material are identified in Aht 1 and Aht 2
horizons at Lauder, Pittenweem and Wigtown. Abundances of coarse organic
anthropogenic material are characteristically less abundant within Aht 2 horizons at
Lauder and Wigtown. In comparison, the abundance of coarse organic anthropogenic
materials between Aht 1 and Aht 2 horizons at Pittenweem is typically varied. Moreover,
there is no difference in the total abundance of fuel residue between the Aht 1 and Aht 2
horizons at Pittenweem.

- 141 -
5.6 Summary of Results: Coarse Organic Material

The following section presents a brief summary of key trends in the nature and
abundance of coarse organic material both within and between burghs. The significance
of these results is discussed in sections 8.1.4 and 8.2.4; in particular’ consideration is
given to differences in the nature and distribution of fuel residue both within and between
burghs.

5.6.1 Variation within Burghs

5.6.1.1 Lauder
Coarse organic material is present within topsoils in the High Street, Hinterland Near,
Hinterland Far and Thirlstane zones at Lauder in addition to ‘reference’ soil pit LA 7
(Table 29). FR 2 and FR 10 inclusions are limited to the burgh core. In addition charcoal
is restricted to the burgh core and immediate hinterland south of Lauder (Hinterland
Near). FR 1, FR 3, FR 4 & 6, FR 5 & 7 and FR 8 & 9 are present within all zones at
Lauder; however, FR 4 & 6 is most abundant within the burgh core and FR 1, FR 3 and
FR 8 & 9 are highest within the burgh core and immediate hinterland south of the
historical burgh limits (Hinterland Near). In contrast there is no difference in the
abundance of FR 5 & 7 between zones at Lauder.

Accordingly two areas of interest are identified at Lauder, the High Street zone and the
Hinterland Near zone. The High Street zone is characterised by charcoal, FR 2 and FR
10 inclusions and significantly higher abundances of FR 4 & 6, FR 1, FR 3 and FR 8 & 9.
The Hinterland Near zone is characterised by charcoal inclusions and significantly higher
abundances of FR 1, FR 3 and FR 8 & 9.

5.6.1.2 Pittenweem
The Harbour, High Street, Hinterland Near and Hinterland Far zones at Pittenweem in
addition to ‘reference’ soil pit PT 1 contain coarse organic material (Table 29). Charcoal,
FR 1, FR 2, FR 3, FR 4 & 6, FR 5 & 7 and FR 8 & 9 are present within all zones at
Pittenweem. There is no difference in the abundance of charcoal and FR 2 inclusions
between zones. FR 3 is most abundant within the High Street zone and FR 4 & 6
inclusions are greatest within the Harbour zone. FR 5 & 7 and FR 8 & 9 are most
abundant within the burgh core. More specifically FR 5 & 7 and FR 8 & 9 are greatest
within the Harbour zone followed in succession by the High Street zone. In addition, FR
1 inclusions are significantly less abundant within the Hinterland Near zone. In contrast
FR 10 is absent from the Hinterland Near zone.

- 142 -
Accordingly two areas of interest are identified at Pittenweem; the Harbour zone and the
High Street zone. The Harbour zone is characterised by significantly higher abundances
of FR 4 & 6, FR 5 & 7 and FR 8 & 9 in comparison to the High Street, Hinterland Near
and Hinterland Far zones. The High Street zone is characterised by significantly higher
abundances of FR 3 in comparison to the Harbour, Hinterland Near and Hinterland Far
zones, and significantly more FR 5 & 7 and FR 8 & 9 inclusions than the Hinterland Near
and Hinterland Far zones.

5.6.1.3 Wigtown
Coarse organic material is present within both the High Street zone and ‘reference’ soil
pit WG 5 at Wigtown (Table 7). As previously stated in section 5.1.1, it is argued that the
‘reference’ soil provides a good indication of soil characteristics within the Hinterland
Near zone. Charcoal, FR 2, FR 3 and FR 10 inclusions are limited to the burgh core at
Wigtown. FR 1, FR 4 & 6, FR 5 & 7 and FR 8 & 9 are present with both the burgh core
and hinterland north of the historical burgh limits; however FR 1, FR 4 & 6 and FR 8 & 9
are significantly more abundant within the burgh core. In contrast there is no difference
in the abundance of FR 5 & 7 between the burgh core and hinterland.

The High Street zone is recognised as an important area of interest at Wigtown. The
High Street zone is characterised by charcoal, FR 2, FR 3 and FR 10 inclusions and
significantly greater abundances of FR 1, FR 4 & 6 and FR 8 & 9.

5.6.1.4 Hortic Horizons

Results presented in section 5.3.1.4 indicate that coarse organic material classes are
consistent throughout Aht 1 horizons at Lauder, Pittenweem and Wigtown. Abundances
of coarse organic material are relatively constant throughout Aht 1 horizons at Lauder
and Wigtown; however, abundances of individual classes are more varied at
Pittenweem. Similar classes of coarse organic material are also identified within Aht 1
and Aht 2 horizons at Lauder, Pittenweem and Wigtown. It should be noted that coarse
organic material is lower in abundance within Aht 2 horizons at Lauder and Wigtown. In
contrast, there is minimal difference in total abundances of coarse organic material
between Aht 1 and Aht 2 horizons at Pittenweem.

- 143 -
Table 7: Summary of trends in coarse organic inclusions identified at Lauder, Pittenweem
and Wigtown, summary abundances are given in brackets

Inclusion Lauder Pittenweem Wigtown

Charcoal Charcoal is present Charcoal is present Charcoal is present
within the burgh core within the burgh core within the burgh core
and immediate and hinterland (<2%).
hinterland (Hinterland (Hinterland Near,
Near (<2%). Hinterland Far) (<2%).
FR 1 FR 1 is present within FR 1 is present within FR 1 is present within
the burgh core (<2% to the burgh core (5-15%) the burgh core (<2%
5-10%) and hinterland and hinterland (2-10%). to 5-10%) and
(<2% to 2-5%). FR 1 is most abundant hinterland (Hinterland
FR 1 is most abundant within the burgh core Near/’reference’ soil)
within the burgh core (Harbour, High Street) (<2% to 2-5%).
and immediate and far hinterland FR 1 is most abundant
hinterland (Hinterland (Hinterland Far). in burgh core.
Near).
FR 2 FR 2 is present within FR 2 is present within FR 2 is present within
the burgh core (<2%). the burgh core and the burgh core (<2%).
hinterland (Hinterland
Near, Hinterland Far)
(<2%).
FR 3 FR 3 is present within FR 3 present within the FR 3 is present within
the burgh core (<2% to burgh core (<2% to 2- the burgh core (<2%).
2-5%) and hinterland 5%) and hinterland
(<2%). (<2%).
FR 3 is most abundant FR 3 is most abundant
within the burgh core in burgh core (High
and immediate Street).
hinterland (Hinterland
Near).
FR 4 & 6 FR 4 & 6 is present FR 4 & 6 is present FR 4 & 6 is present
within the burgh core within the burgh core within the burgh core
(<2 to 5-10%) and (2-5%, to 2-10%) and (5-10% to 10-15%)
hinterland (<2% to 2- hinterland (<2% to 2- and hinterland
5%). 5%). (Hinterland
FR 4 & 6 is most FR 4 & 6 is most Near/’reference’ soil)
abundant within the abundant within the (<2% to 2-5%).
burgh core. burgh core (Harbour). FR 4 & 6 is most
abundant within the
burgh core.
FR 5 & 7 FR 5 & 7 is present FR 5 & 7 is present FR 5 & 7 is present
within the burgh core within the burgh core within the burgh core
and hinterland (<2%). (Harbour 5-10% to 10- and hinterland
15%, High Street <2%) (Hinterland
and hinterland (<2%). Near/’reference’ soil)
FR 5 & 7 is most (<2%).
abundant in burgh core
(Harbour followed by
High Street).
FR 8 & 9 FR 8 & 9 is present FR 8 & 9 is present FR 8 & 9 is present
within the burgh core within the burgh core within the burgh core
(<2% to 2-5%) and and hinterland (<2% to (<2% to 5-10%) and

- 144 -
 hinterland (<2% to 2- 2-5%). hinterland (Hinterland
5%). FR 8 & 9 most Near/’reference’ soil)
FR 8 & 9 is most abundant in burgh core (<2%).
abundant within the (Harbour followed by FR 8 & 9 is most
burgh core. High Street) abundant within the
burgh core
FR 10 FR is 10 present within FR 10 is present within FR 10 is present
the burgh core (<2%) the burgh and far within the burgh core
hinterland (Hinterland (<2%).
Far) (<2%).

5.6.2 Variation between Burghs

All three burghs have charcoal, FR 1, FR 2, FR 3, FR 4 & 6, FR 5 & 7, FR 8 & 9 and FR

10 in zones corresponding to the burgh core (Harbour and High Street). Coarse organic
inclusions are also identified in the hinterland at Lauder, Pittenweem and Wigtown, for
example abundances of FR 1, FR 3 and FR 8 & 9 inclusions are comparable between
the burgh core and the immediate hinterland south of the historical burgh limits at Lauder
(Hinterland Near). Likewise, there is no difference in the abundance of FR 1 inclusions
between the burgh core and Hinterland Far zone at Pittenweem. Abundances of coarse
organic material are most abundant within burgh cores, although in the case of
Pittenweem differences in the nature of fuel residue classes between the Harbour and
High Street are noted.

It is proposed that differences in the abundance of selected coarse organic materials can
be used to discriminate between High Street zones at Lauder, Pittenweem and Wigtown.
FR 1, FR 3 and FR 5 & 7 inclusions are most abundant within the High Street zone at
Pittenweem (Table 8). These inclusions share similar morphologies characterised by
sharp, well defined angular/sub-angular perimeters (see Figure 27). FR 4 & 6 inclusions
are most abundant within the High Street zone at Wigtown. These inclusions are
typically sub-rounded with internally degraded holes. FR 8 & 9 inclusions are greatest
within the High Street zone at Lauder. These inclusions range from sub-angular to sub-
rounded but are distinct in that they have a ‘ragged’ perimeter edge. There is no
difference in the abundance of charcoal, FR 2 and FR 10 inclusions between High Street
zones.

- 145 -
Table 8: Summary of trends in the abundance of coarse organic anthropogenic material
between High Street zones at Lauder, Pittenweem and Wigtown as indicated by Kruskal-
Wallis analysis with Dunn’s test

Inclusion Comparison between High Street zones

Charcoal There is no difference in charcoal between High Street zones
FR 1 FR 1 is most abundant within Pittenweem (p<0.001)
FR 2 There is no difference in FR 2 between High Street zones
FR 3 FR 3 is most abundant within Pittenweem, followed in succession by Lauder
FR 4 & 6 FR 4 & 6 is most abundant within Wigtown
FR 5 & 7 FR 5 & 7 is most abundant within Pittenweem
FR 8 & 9 FR 8 & 9 is most abundant within Lauder, followed in succession by
Wigtown
FR 10 There is no difference in FR 10 between High Street zones

- 146 -
5.7 Fine Organic Material

The abundance of fine organic material within topsoil deposits at Lauder, Pittenweem
and Wigtown is summarised in tables presented in Appendix 2. Differences in the nature
and abundance of organic material are identified and comparisons between zones and
burghs are made.

5.7.1 Variation within Burghs

Fine organic material at Lauder consists of cell residues and amorphous red organic
material. Trace abundances (<2%) of cell residues are present within the High Street,
Hinterland Near, Hinterland Far and Thirlstane zones and ‘reference’ soil pit LA 7.
Hence, there is no difference in the abundance of cell residue between zones.
Amorphous red organic material is <2% within the Hinterland Near and is absent from
the High Street, Hinterland Far and Thirlstane zones and the ‘reference’ soil pit.
Amorphous red organic material is therefore limited to the burgh core.

Fine organic material at Pittenweem comprises cell residues, amorphous red organic
material and amorphous yellow organic material. Trace abundances (<2%) of cell
residues are identified within the Harbour, High Street, Hinterland Near and Hinterland
Far zones and ‘reference’ soil pit PT 1. Hence, there is no difference in the abundance of
cell residue between zones. Amorphous red organic material is <2% within Harbour,
High Street and Hinterland Far zones and ‘reference’ soil pit. In contrast amorphous red
amorphous organic material is absent from the Hinterland Near zone. There is no
difference in the abundance of amorphous red amorphous organic material between the
burgh core and far hinterland (Hinterland Far). Amorphous yellow organic material is
<2% within Harbour and High Street zones, yet is absent from the Hinterland Near and
Hinterland Far zones and the ‘reference’ soil pit. Amorphous yellow organic material is
therefore limited to the burgh core.

Fine organic material at Wigtown includes cell residues, amorphous red organic material
and amorphous yellow organic material. Trace abundances (<2%) of cell residues occur
within the High Street zone and ‘reference’ soil WG 5. There is no difference in the
abundance of cell residue between zones. Trace abundances (<2%) of amorphous red
and yellow organic material are confined to selected soil pits within the High Street zone.
Amorphous red and yellow organic material is therefore limited to the burgh core at
Wigtown.

- 147 -
5.7.2 Variation between Burghs

There is no difference in the abundance of cell residues between zones at all three
burghs. It is suggested that spatial distributions of red and yellow amorphous organic
material are more site specific. Amorphous red organic material is present within the
burgh core at Lauder, Pittenweem and Wigtown and within the immediate hinterland
south of Lauder (Hinterland Near) and within the hinterland at Pittenweem (Hinterland
Near/Hinterland Far). In contrast amorphous yellow organic material is restricted to
individual soil pits within the burgh core at Pittenweem and Wigtown.

- 148 -
5.8 Pedofeatures

The abundance of pedofeatures within topsoil deposits at Lauder, Pittenweem and

Wigtown is summarised in tables presented in Appendix 2. Differences in the nature and
abundance of pedofeatures are identified and comparisons between zones and burghs
are made.

5.8.1 Variation within Burghs

Trace abundances (<2%) of spheroidal excremental pedofeatures and iron impregnated

siltstone nodules are present within the High Street, Hinterland Near, Hinterland Far and
Thirlstane zones and ‘reference’ soil pit LA 7 at Lauder. Hence there is no difference in
the abundance of spheroidal excremental pedofeatures and iron impregnated siltstone
nodules between zones at Lauder. Mammilate excremental pedofeatures are identified
in trace abundances within the High Street and Hinterland Near zones (<2%) and are
absent from the Hinterland Far and Thirlstane zones and ‘reference’ soil pit. Mammilate
excremental pedofeatures are therefore limited to the burgh core and immediate
hinterland south (Hinterland Near) of Lauder. In addition it is recognised that trace
abundances (<2%) of iron impregnated amorphous organic material are restricted to the
burgh core (High Street).

Trace abundances (<2%) of spheroidal excremental pedofeatures, iron nodules and iron
impregnated siltstone nodules are present within the Harbour, High Street, Hinterland
Near and Hinterland Far zones, and ‘reference’ soil pit PT 1 at Pittenweem. Therefore
there is no difference in the abundance of spheroidal excremental pedofeatures, iron
nodules and iron impregnated siltstone nodules between zones at Pittenweem.
Mammilate excremental pedofeatures are identified in trace abundances within the High
Street and Hinterland Near zones (<2%) and are absent from the Harbour and
Hinterland Far zones at ‘reference’ soil. Mammilate excremental pedofeatures are
therefore limited to the burgh core (High Street) and immediate hinterland north of
Pittenweem (Hinterland Near). Iron impregnated amorphous organic material and
organic heated coatings are present in trace abundance (<2%) within the High Street
and Hinterland Far zones yet are absent from the Harbour and Hinterland Near zones
and ‘reference’ soil pit. These materials are therefore limited to the burgh core (High
Street) and far hinterland north of Pittenweem (Hinterland Far). In addition it is
recognised that monomorphic organic coatings are present in trace abundances (<2%)
within the Harbour and High Street zones and are absent from the Hinterland Near and
Hinterland Far zones and ‘reference’ soil pit.

- 149 -
Trace abundances (<2%) of spheroidal and mammillate excremental pedofeatures and
iron nodules are present within the High Street zone and ‘reference’ soil pit WG 5 at
Wigtown. Therefore there is no difference in the abundance of spheroidal and
mammillate excremental pedofeatures and iron nodules between the burgh core and
hinterland north of Wigtown. In contrast it is identified that iron impregnated amorphous
organic material (<2%) is limited to the ‘reference’ soil and heated organic coatings
(<2%) are limited to the High Street zone.

5.8.2 Variation between Burghs

Spheroidal and mammilate excremental pedofeatures, iron impregnated amorphous

organic material and organic coatings are identified within topsoils at all three burghs.
There is no difference in the abundance of spheroidal excremental pedofeatures
between zones at Lauder, Pittenweem and Wigtown. It is suggested that spatial
distributions of mammilate excremental pedofeatures, iron nodules, iron impregnated
siltstone nodules, iron impregnated amorphous organic material, monomorphic organic
coatings and heated organic coatings are more site specific.

- 150 -
5.9 Structure

5.9.1 Variation within Burghs

The microstructure of topsoils within the High Street zone is channel and chamber with
vughy elements. Similarly, topsoils within the Hinterland Near, Hinterland Far and
Thirlstane zones and ‘reference’ soil pit LA 7 are characterised by channel and chamber
microstructures. The coarse material arrangement of topsoils within all zones is random
and the groundmass (b fabric) is stipple-speckled. The C:F related distribution of topsoil
deposits at Lauder is close porphyric. The mean C:F ratio of the High Street, Hinterland
Near, Hinterland Far and Thirlstane zones and ‘reference’ soil is 58:42, 53:47, 56:44,
55:45 and 59:41 respectively. Results of One Way Analysis of Variance (ANOVA) reveal
that there is no statistical difference in the C:F ratio between zones. The mean % void
space of High Street, Hinterland Near, Hinterland Far and Thirlstane zones and
‘reference’ soil pit is 29, 21, 22, 28 and 27%. There is no statistical difference in void
space between the Harbour, High Street and Hinterland Near zones.

The microstructure of topsoils within the Harbour, High Street and Hinterland Far zones
and ‘reference’ soil PT 1 is channel and chamber with vughy elements. In contrast
topsoil within the Hinterland Near zone is predominantly vughy. The coarse material
arrangement of topsoils within all zones is random and the groundmass (b fabric) is
stipple-speckled. The C:F related distribution of topsoil deposits at Pittenweem is close
porphyric. The mean C:F ratio of Harbour, High Street, Hinterland Near and Hinterland
Far zones and ‘reference’ soil pit is 55:45, 57:43, 63:36, 54:46 and 55:45. Results of
One Way Analysis of Variance (ANOVA) reveal that there is no statistical difference in
the C:F ratio between zones. The mean % void space of Harbour, High Street,
Hinterland Near and Hinterland Far zones and ‘reference’ soil pit is 32, 32, 62, 22 and 16
%. Mean % void space is significantly higher within the Harbour and High street zones
than the ‘reference’ soil pit (p<0.01). In addition, mean % void space within the
Hinterland Near zone is greater than the Hinterland Far zone and ‘reference’ soil pit
(p<0.01). There is no statistical difference in void space between the Harbour, High
Street and Hinterland Near zones.

- 151 -
The microstructure of topsoils within High Street zone is channel and chamber with
vughy elements. In contrast the microstructure if topsoil within the ‘reference’ soil WG 5
(Hinterland Near) is channel and chamber. The coarse material arrangement of topsoils
within all zones is random and the groundmass (b fabric) is stipple-speckled. The C:F
related distribution of topsoil deposits at Wigtown is close porphyric. The mean C:F ratio
of both the High Street zone and ‘reference’ soil pit is 58%, hence there is no significant
difference. Similarly there is no significant difference in mean % void space between the
High Street zone and ‘reference’ soil which is 32% for both zones.

5.9.2 Variation between Burghs

The microstructure of topsoils within the burgh core at all three burghs is
characteristically channel and chamber with vughy elements. In contrast the
microstructure within the hinterland at Lauder, Pittenweem and Wigtown is channel and
chamber, with the exception of soil pit PT 9 at Pittenweem (Hinterland Near) which is
vughy. There is no difference in the coarse material arrangement, groundmass (b fabric),
C:F related distribution and C:F ratio of topsoils either within or between burghs. In
addition there is no difference in void space within topsoils at Lauder and Wigtown;
however, it is recognised that void space is significantly higher within topsoils in the
burgh core and immediate hinterland north of Pittenweem (Hinterland Near).

- 152 -
5.10 Summary: Soil Micromorphology

A succinct overview of key findings resulting from micromorphological analyses

presented in this chapter is provided in Table 9. The significance of these results is
discussed throughout chapter 8 in association with soil physical and chemical properties
and elemental concentrations.

Table 9: Summary of trends in key micromorphological analyses of topsoils within Lauder,

Pittenweem and Wigtown

Analysis Summary of Key Trends

Profile Characteristics All three burgh cores and immediate hinterland at Lauder
characterised by hortic topsoil horizons
Coarse Anthropogenic Coarse mineral material is most abundant and diverse in
burgh cores
Mineral Material
Coarse mineral material is present in selected hinterland
zones at Lauder and Pittenweem
Shell inclusions are limited to Pittenweem and Wigtown
Coarse mineral material is absent from Aht 2 horizons at
Lauder
Fine Mineral Material All three burgh cores are characterised by brown/dark
brown, dotted fine mineral material
Similarities in fine mineral material between burgh cores
and selected hinterland zones exist at all three burghs
Coarse Anthropogenic Coarse organic material is most abundant and diverse in
burgh cores
Organic Material
Coarse organic material is present within the hinterland at
all three burghs
There are differences in principal fuel residue types
between burghs
Coarse organic material is lower in abundance within Aht 2
horizons at Lauder and Wigtown
Fine Organic Material All three burghs contain amorphous red and yellow organic
material. Distributions of organic material are site specific
Pedofeatures The nature and distribution of pedofeatures is site specific
Structure All three burgh cores are characterised by channel and
chamber microstructures with vughy elements
There is no difference in coarse material arrangement,
groundmass, C:F distribution or C:F ratio between burghs

- 153 -
 6 Soil Physical and Chemical Properties

This chapter presents results obtained from selected soil physical and chemical
analyses including topsoil depth, pH, % loss on ignition, magnetic susceptibility and
frequency dependant magnetic susceptibility. Data are presented collectively for Lauder,
Pittenweem and Wigtown to facilitate visual and statistical comparisons.

6.1 Topsoil Depth

Spatial distributions of topsoil depth are presented for Lauder (Figure 53), Pittenweem
(Figure 54a) and Wigtown (Figure 54b). Pronounced spatial variability in topsoil depth is
observed for all three burghs; however specific patterns relating to distinct zones are
identified both within and between towns.

Figure 53: Distribution of topsoil depth at Lauder. Red boundary delimits 1862AD urban
extent

- 154 -
a) b)

Figure 54: Spatial distribution of topsoil depth at Pittenweem (a) and Wigtown (b). Red
boundary delimits 1855AD and 1850AD urban extent respectively

6.1.1 Variation within Burghs

There is clear evidence for topsoil deepening within the High Street and Hinterland Near
zones at Lauder. The median topsoil depth of the High Street zone is 75cm and 53cm
for the Hinterland Near zone. Statistically there is no significant difference between these
two zones, however both the High Street and Hinterland Near zones are significantly
deeper than the Hinterland Far and Thirlstane zones (p<0.05). The Hinterland Far and
Thirlstane zones have median topsoil depths of 38cm and 36cm respectively (Figure
55). Deepened topsoil deposits are, therefore, identified in both the burgh core and land
immediately adjacent to the historical burgh limits (Hinterland Near).

- 155 -
 a) b)
Boxplots w ith Sign Confidence Intervals Pairw ise Comparisons
Desired Conf idence: 86.761

80

70
High St

60
Topsoil Depth (cm)

50

Hint Near
40

30

20

Hint Far Thirlstane

10
Hint Far
Hint Near
0
High St Hint Near Hint Far Thirlstane -Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 55: (a) Boxplot of median topsoil depth for individual zones at Lauder; boundaries
of boxes represent interquartile range, green shading indicates 86.761% confidence
interval for median, outliers identified as *, (b) Kruskal-Wallis analysis with Dunn’s test
multiple comparisons for individual zones at Lauder, Z values exceeding –Z or Z are
significant at p<0.05 confidence level

The median topsoil depth of the Harbour, High Street and Hinterland Near zones at
Pittenweem is ≥80cm, ≥80cm and 72.5cm respectively. There is no significant difference
in topsoil depth between these zones, however the Harbour, High Street and Hinterland
Near zones are all significantly deeper than the Hinterland Far zone (p<0.05) (Figure
56b). Accordingly deepened topsoil deposits are identified within both the burgh core
and land immediately adjacent to the historical burgh limits (Hinterland Near).

The median topsoil depth of the High Street, Hinterland Near and Showfield zones at
Wigtown is 55cm, 18cm and 17cm respectively. The High Street zone is significantly
deeper than the Hinterland Near and Showfield zones (p<0.05). In addition it is evident
that the Hinterland Near and Showfield zones are spatially homogenous with minimal
variation in topsoil depth values (Figure 57a). A clear contrast in topsoil depth between
the burgh core and the hinterland is therefore noted.

- 156 -
 a) Boxplots w ith Sign Confidence Intervals
b) Pairw ise Comparisons
Desired Conf idence: 86.761

80

H arbour
70

60
Topsoil Depth (cm)

50 High S t

40

30

H int N ear
H int F ar
20
H int N ear
H igh S t
Harbour High St Hint Near Hint Far
-Z 0 Z
F amily A lpha: 0.2
Bonferroni Indiv idual A lpha: 0.033 |Bonferroni Z-v alue|: 2.128

Figure 56: (a) Boxplot of median topsoil depth for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, green shading indicates 86.761%
confidence interval for median, outliers identified as *, (b) Kruskal-Wallis analysis with
Dunn’s test multiple comparisons for individual zones at Pittenweem, Z values exceeding
–Z or Z are significant at p<0.05 confidence level

a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons

Desired Conf idence: 80.529

80

70

High St
60
Topsoil Depth (cm)

50

40

30

20
Hint Near

10

Showfield
0 Hint Near

High St Hint Near Showf ield

-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.067 |Bonf erroni Z-v alue|: 1.834

Figure 57: (a) Boxplot of median topsoil depth for individual zones at Wigtown; boundaries
of boxes represent interquartile range, green shading indicates 80.529% confidence
interval for median, outliers identified as *, (b) Kruskal-Wallis analysis with Dunn’s test
multiple comparisons for individual zones at Pittenweem, Z values exceeding –Z or Z are
significant at p<0.05 confidence level

- 157 -
6.1.2 Variation between Burghs

The results of a Two Way Analysis of Variance (ANOVA) undertaken using a General
Linear Model (GLM) indicate that mean topsoil depth is significantly different between
burghs (p<0.001). Tukey-Kramer (95% confidence intervals) multiple comparisons
reveal that the mean topsoil depth at Pittenweem is significantly greater than Lauder
(p<0.005) and Wigtown (p<0.001), and the mean topsoil depth at Lauder is significantly
greater than Wigtown (p<0.001). It is therefore argued that mean topsoil depth varies
between different burghs with Pittenweem having the deepest mean topsoil depth and
Wigtown the shallowest.

Results presented in section 6.1.1 show differences in topsoil depth between certain
zones within each burgh. These findings are supplemented by ANOVA which confirms a
significant difference in mean topsoil depth between the High Street and Hinterland Near
zones (p<0.001). Tukey-Kramer multiple comparisons indicate a greater mean topsoil
depth for the High Street zone compared to the Hinterland Near zone (p<0.001).
Conversely results from Kruskal Wallis pairwise comparisons suggest that there is no
significant difference between the High Street and Hinterland Near zones at Lauder and
Pittenweem. This discrepancy may reflect differences in measurement of central
tendency between parametric and non-parametric tests. Nevertheless, all three towns
have significantly deepened deposits corresponding to the burgh core and in the case of
Lauder and Pittenweem, significantly deepened deposits in the Hinterland Near zone.

Statistical analysis also indicates significant interaction between individual ANOVA test
factors, burgh and zone (p<0.001). This association implies that topsoil depth variation
cannot be fully explained by either burgh or zone as factors in isolation. Moreover it can
be argued the influence of one factor is dependant on the nature of the other factor.
Accordingly it is suggested that although topsoil depth is consistently enhanced within
the burgh core, the nature and extent of topsoil deepening in this area differs between
burghs.

- 158 -
6.2 Soil pH

Spatial distributions of soil pH are presented for Lauder (Figure 58), Pittenweem (Figure
59) and Wigtown (Figure 60) at selected depths. Patterns in soil pH associated with
discrete zones and depths are subsequently identified and comparisons between towns
are made.

a)

b)

Figure 58: Spatial distribution of soil pH at Lauder (a) 0-20cm and (b) 20-40cm depth. Red
boundary delimits 1862AD urban extent

- 159 -
a) b) c) d)

Figure 59: Spatial distribution of soil pH at Pittenweem (a) 0-20cm, (b) 20-40cm, (c) 40-60cm and (d) 60-80cm depth. Red boundary delimits 1855AD
urban extent

- 160 -
Figure 60: Spatial distribution of soil pH at Wigtown (0-20cm depth). Red boundary
delimits 1850AD urban extent

6.2.1 Variation within Burghs

6.2.1.1 Zone
There is clear evidence for enhanced soil pH within the High Street zone at Lauder. The
median soil pH of the High Street, Hinterland Near, Hinterland Far and Thirlstane zones
is 6.9, 5.4, 5.3 and 5.4 respectively (Figure 61). There is no difference in soil pH
between the Hinterland Near, Hinterland Far and Thirlstane zones, however these three
zones are all significantly lower than the High Street zone (p<0.001). A distinct contrast
between neutral soils associated with the burgh core and acidic hinterland soils is
therefore identified.

- 161 -
 a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761
8.0

High St Hint Near

7.5
Hint Far

7.0 Thirlstane

6.5
pH

6.0 Hint Near Hint Far

Thirlstane
5.5

5.0

Hint Far Thirlstane

4.5

High St Hint Near Hint Far Thirlstane

-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 61: (a) Boxplot of median pH for individual zones at Lauder; boundaries of boxes
represent interquartile range, blue shading indicates 86.761% confidence interval for
median, outliers identified as *, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Lauder (0-20cm depth), Z values exceeding –Z or Z
are significant at p<0.05 confidence level

The median soil pH of the Harbour, High Street, Hinterland Near and Hinterland Far
zones at Pittenweem is 7.2, 7.05, 5.4 and 5.3 respectively. There is no statistical
difference between the Harbour and High Street zones, however both these zones have
a significantly higher soil pH than the Hinterland Near and Hinterland Far zones
(p<0.001) (Figure 62b). Similar to results identified at Lauder, there is a clear distinction
between neutral soils in the burgh core with acidic soils in the hinterland.

There is a significant difference in soil pH between the High Street, Hinterland Near and
Showfield zones at Wigtown (p<0.001). The median soil pH of the High Street,
Hinterland Near and Showfield zones is 6.9, 5 and 4.7 respectively (Figure 63a). The
High Street has a neutral median soil pH which is significantly greater than the acidic
Hinterland Near and Showfield zones (p<0.001). Additionally there is a significant
difference between the Hinterland Near and Showfield zones (p<0.05), of which the
Hinterland Near zone has slightly higher pH values. Nevertheless the principal trend at
Wigtown is comparable to Lauder and Pittenweem in that there is a distinct contrast in
soil pH between the burgh core and surrounding hinterland.

- 162 -
 a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761

8.0

Harbour High St

7.5
Hint Near

7.0 Hint Far

6.5
pH

High St Hint Near

6.0

Hint Far
5.5

5.0

Hint Near Hint Far

4.5
Harbour High St Hint Near Hint Far
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 62: (a) Boxplot of median pH for individual zones at Pittenweem; boundaries of
boxes represent interquartile range, blue shading indicates 86.76% confidence interval for
median, outliers identified as *, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Pittenweem (0-20cm) using, Z values exceeding –Z or
Z are significant at p<0.05 confidence level

a)) Boxplots w ith Sign Confide nce Inte rvals b) Pairw is e Com parisons
Desired Conf idence: 80.529
7.5

7.0
High St Hint Near

6.5

Showf ield
6.0
pH

5.5

5.0

4.5 Hint Near Showf ield

4.0

High St Hint Near Showf ield

-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.067 |Bonf erroni Z-v alue|: 1.834

Figure 63: (a) Boxplot of median pH for individual zones at Wigtown; boundaries of boxes
represent interquartile range, blue shading indicates 80.52% confidence interval for
median, outliers identified as *, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Wigtown (0-20cm depth), Z values exceeding –Z or Z
are significant at p<0.05 confidence level

- 163 -
6.2.1.2 Variation in pH with Depth
At Lauder it is identified that soil pH increases with depth. Results of Kruskal-Wallis
statistical analysis indicate a significantly higher (p<0.001) median soil pH for 20-40cm
compared to 0-20cm depth. In addition it is recognised that spatial trends identified for 0-
20cm are apparent at 20-40cm depth (Figure 64), for example the High Street zone has
a significantly higher (p<0.001) soil pH than the Hinterland Near, Hinterland Far and
Thirlstane zones.

Results of a Two Way ANOVA undertaken using a GLM indicate a significant difference
(p<0.001) in pH between zones at Lauder, where the High Street has significantly higher
(p<0.001) pH values than the Hinterland Near, Hinterland Far and Thirlstane zones.
These results support spatial trends identified using the appropriate non-parametric
Kruskal-Wallis analysis and Dunn’s test in section 6.2.1.1. Moreover in supplementation
to results outlined in the above paragraph, 20-40cm is identified as having significantly
higher pH values compared to 0-20cm (p<0.001).

Statistical analysis also indicates that there is no significant interaction between zone
and depth at Lauder. It is therefore argued that while pH varies between depths, the
nature of this variation between zones is consistent.

8.0

7.5

7.0

6.5
pH

6.0

5.5

5.0

4.5

Depth (cm) 20 40 20 40 20 40 20 40
High Street Hint Near Hint Far Thirlstane

Figure 64: Boxplot of median pH for individual zones at Lauder for 0-20cm and 20-40cm
depth; boundaries of boxes represent interquartile range, outliers identified as *

- 164 -
Similar to results obtained for Lauder, soil pH increases with depth at Pittenweem.
Results of Kruskal-Wallis analysis and Dunn’s test indicate a significantly higher soil pH
for 20-40cm (p<0.05), 40-60cm (p<0.001) and 60-80cm depth (p<0.001) compared to 0-
20cm. However, there is no statistical difference in soil pH between 20-40cm, 40-60cm
and 60-80cm (Figure 65). In addition spatial trends noted for 0-20cm remain consistent
at subsequent depths (Figure 66), for example there is a distinct contrast between the
neutral burgh core and acidic hinterland zones at each successive depth.

Results of Two Way ANOVA undertaken using a GLM indicate that the Harbour and
High Street zones have a significantly greater pH than Hinterland Near and Hinterland
far zones. These results support spatial trends identified using the appropriate non-
parametric Kruskal-Wallis analysis and Dunn’s test in section 6.2.1.1. Moreover in
supplementation to results outlined in the above paragraph 0-20cm has lower pH values
than 20-40cm (p<0.005), 40-60cm (p<0.001) and 60-80cm (p<0.001).

Statistical analysis also indicates that there is no significant interaction between zone
and depth at Pittenweem. Similar to results obtained for Lauder it is suggested that
variation in pH between zones is consistent at successive depths.

- 165 -
 a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761

8.5
0-20 cm 20-40 cm
8.0
40-60 cm

7.5
60-80 cm

7.0

6.5
pH

20-40 cm 40-60 cm

6.0
60-80 cm

5.5

5.0
40-60 cm 60-80 cm

4.5
0-20 cm 20-40 cm 40-60 cm 60-80 cm
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 65: (a) Boxplot of median pH for 0-20cm, 20-40cm, 40-60cm and 60-80cm depth at
Pittenweem; boundaries of boxes represent interquartile range, blue shading indicates
86.76% confidence interval for median, outliers identified as *, (b) Kruskal-Wallis analysis
with Dunn’s test multiple comparisons for individual depths at Pittenweem, Z values
exceeding –Z or Z are significant at p<0.05 confidence level

8.5

8.0

7.5

7.0

6.5
pH

6.0

5.5

5.0

4.5
Depth (cm) 20 40 60 80 20 40 60 80 20 40 60 80 20 40 60 80

Harbour High Street Hint Near Hint Far

Figure 66: Boxplot of median pH for individual zones within 0-20cm, 20-40cm, 40-60cm
and 60-80cm depth at Pittenweem; boundaries of boxes represent interquartile range,
outliers identified as *

- 166 -
6.2.2 Variation between Burghs

The results of a Two Way ANOVA undertaken using a GLM indicate that mean soil pH is
significantly different between burghs (p<0.001). Tukey-Kramer (95% confidence
intervals) multiple comparisons reveal that the mean soil pH at Pittenweem is
significantly greater than Lauder (p<0.05) and Wigtown (p<0.001). It is therefore
recognised that mean soil pH differs between burghs.

Results presented in section 6.2.1.1 highlight a stark difference in soil pH between the
burgh core and surrounding hinterland zones at each town. These findings are
supported by ANOVA which verifies a significant difference in mean pH between the
High Street and Hinterland Near zones (p<0.001). Tukey-Kramer multiple comparisons
confirm a significantly higher mean pH for the High Street zone compared to Hinterland
Near zone (p<0.001). It is therefore argued that all three towns exhibit similar patterns in
spatial distributions of soil pH, where the burgh core is characterised by neutral pH
values in contrast to the hinterland which is acidic.

Statistical analysis suggests that there is no interaction between the individual ANOVA
test factors, burgh and zone. It is therefore argued that although differences in soil pH
are identified between the High Street and Hinterland Near zones, the nature and extent
of variation is consistent within each town.

- 167 -
6.3 Loss on Ignition

Spatial distributions of Loss on Ignition (% LOI) are presented for Lauder (Figure 67),
Pittenweem (Figure 68) and Wigtown (Figure 69) at selected depths. Trends in LOI
associated with discrete zones and depths are subsequently identified and comparisons
between towns are made.

a)

b)

Figure 67: Spatial distribution of % LOI at Lauder (a) 0-20cm and (b) 20-40cm depth. Red
boundary delimits 1862AD urban extent

- 168 -
 a) b) c) d)

Figure 68: Spatial distribution of % LOI at Pittenweem (a) 0-20cm, (b) 20-40cm, (c) 40-60cm and (d) 60-80cm depth. Red boundary delimits 1855AD
urban extent

- 169 -
Figure 69: Spatial distribution of % LOI at Wigtown (0-20cm depth). Red boundary delimits
1850AD urban extent

6.3.1 Variation within Burghs

6.3.1.1 Zones
The median % LOI of the High Street, Hinterland Near, Hinterland Far and Showfield
zones at Lauder is 13.88, 12.82, 10 and 11.37% respectively (Figure 70a). The High
Street zone has significantly higher % LOI than Hinterland Far (p<0.001) and Thirlstane
zones (p<0.01). Moreover the Hinterland Near zone has significantly higher % LOI
compared to the Hinterland Far (p<0.001) and Thirlstane zones (p<0.05). However,
there is no statistical difference between the High Street zone and Hinterland Near zone

- 170 -
(Figure 70b). These results provide clear evidence for increased levels of % LOI in the
burgh core and land immediately adjacent to the historical burgh limits (Hinterland Near).

The median LOI of the Harbour, High Street, Hinterland Near and Hinterland Far zones
at Pittenweem is 13.9, 12.8, 10.8 and 7.9% respectively (Figure 71a). The Harbour has
significantly greater % LOI than the Hinterland Near (p<0.01) and Hinterland Far zones
(p<0.001). Moreover, both the High Street and Hinterland Near zones have significantly
higher % LOI compared to the Hinterland Far zone (p<0.001) (Figure 71b). There is no
statistical difference in % LOI between the High Street and Harbour zones, and High
Street and Hinterland Near zones. These results indicate enhanced % LOI in the burgh
core and land immediately adjacent to the historical burgh limits (Hinterland Near).
Furthermore variation between enhanced zones is noted, for example although there is
no difference in median % LOI between the High Street and Harbour zones, and High
Street and Hinterland Near zones, there is a significant difference between the Harbour
and Hinterland Near zones. This supports the observation that the Harbour zone has the
highest median % LOI, followed by the High Street and Hinterland Near zone.

- 171 -
 Boxplots w ith Sign Confidence Intervals Pairw ise Comparisons
a) Desired Conf idence: 86.761
b)
25

High St Hint Near

Hint Far

20
Thirlstane
% LOI

15
Hint Near Hint Far

Thirlstane

10

Hint Far Thirlstane

5
High St Hint Near Hint Far Thirlstane
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 70: (a) Boxplot of median % LOI for individual zones at Lauder; boundaries of
boxes represent interquartile range, orange shading indicates 86.761% confidence interval
for median, outliers identified as *, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Lauder (0-20cm depth) , Z values exceeding –Z or Z
are significant at p<0.05 confidence level

The mean % LOI of the High Street, Hinterland Near and Hinterland Far zones at
Wigtown is 17.8, 15.6 and 16.1% respectively. Results of One Way ANOVA with Tukey-
Kramer multiple comparisons indicate that the High Street has a significantly higher
mean % LOI than the Hinterland Near zone (p<0.05). However, there is no statistical
difference in % LOI between the High Street and Showfield zones. A boxplot of median
LOI values at Wigtown is presented in Figure 72. Differences between median LOI
values suggest a greater contrast between the High Street (18.3%) and Showfield
(15.5%) zones in addition to minimal variation between the Hinterland Near and
Showfield zones. Similar to results obtained for soil pH in section 6.2.1.1, a distinct
contrast in % LOI is noted between the burgh core and surrounding hinterland.

- 172 -
 a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761

22.5

Harbour
20.0

17.5

15.0
% LO I

High St
12.5

10.0

7.5

Hint Near
5.0
Hint Far
Hint Near
High St
Harbour High St Hint Near Hint Far
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 71: (a) Boxplot of median % LOI for individual zones at Pittenweem; boundaries of
boxes represent interquartile range, orange shading indicates 86.761% confidence interval
for median, outliers identified as *, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Pittenweem (0-20cm depth), Z values exceeding –Z or
Z are significant at p<0.05 confidence level

30

25

20
% LOI

15

10

High Street Hinterland Showfield

Figure 72: Boxplot of median % LOI for individual zones at Wigtown; boundaries of boxes
represent interquartile range, outliers identified as *

- 173 -
6.3.1.2 Variation in LOI with Depth
It is identified that % LOI decreases with depth at Lauder. Results of Kruskal-Wallis
statistical analysis with Dunn’s test indicate a significantly higher (p<0.001) median %
LOI for 0-20cm compared to 20-40cm depth. In addition, it is recognised that spatial
trends identified at 0-20cm are not maintained at 20-40cm. The median % LOI for the
High Street, Hinterland Near, Hinterland Far and Thirlstane zones at 20-40cm is 8.83,
7.03, 8.21 and 10.71% respectively (Figure 73). In contrast to trends identified at 0-
20cm, there is no significant difference between the High Street and Hinterland Far
zone, High Street and Thirlstane zone, and Hinterland Near and Hinterland Far zone.
Moreover, at 20-40cm depth the Hinterland Near zone has a significantly higher % LOI
compared to Thirlstane zone (p<0.05).

Results of a Two Way ANOVA undertaken using a GLM indicate a significant difference
(p<0.001) in % LOI between zones at Lauder, where the High Street zone has a
significantly higher % LOI than the Hinterland Far zone (p<0.05). Notably results from
this analysis do not reveal a difference between the Hinterland Near zone with either the
Hinterland Far or Thirlstane zone as identified using the appropriate non-parametric
Kruskal-Wallis analysis and Dunn’s test in section 6.3.1.1. This may reflect differences in
the measurement of central tendency between parametric and non-parametric tests.
Moreover, in supplementation to results outlined in the above paragraph, 0-20cm is
identified has having a significantly higher % LOI compared to 20-40cm depth (p<0.001).

25

20

15
% LOI

10

0
Depth (cm) 20 40 20 40 20 40 20 40
High Street Hint Near Hint Far Thirlstane

Figure 73: Boxplot of median % LOI for individual zones at Lauder for 0-20cm and 20-40cm
depth; boundaries of boxes represent interquartile range, outliers identified as *

- 174 -
Statistical analysis indicates significant interaction (p<0.001) between the individual
ANOVA test factors, zone and depth. This association implies that variation in % LOI at
subsequent depth increments cannot be fully explained by zone as a factor in isolation.

The median % LOI for 0-20cm, 20-40cm, 40-60cm and 60-80cm depth at Pittenweem is
10.13, 7.86, 5.38 and 3.37% respectively. Results of Kruskal-Wallis analysis and Dunn’s
test indicate a significantly higher median % LOI for 0-20cm compared to each
successive depth (p<0.001). Similarly 20-40cm has a significantly higher % LOI than
both 40-60cm and 60-80cm (p<0.001). These results indicate a sustained decrease in %
LOI associated with depth. In addition it is recognised that spatial trends identified at 0-
20cm remain consistent at subsequent depths (Figure 74), for example at 40-60cm
depth the Harbour zone has a significantly higher % LOI compared to the Hinterland Far
zone (p<0.001). Moreover at 40-60cm depth both the High Street and Hinterland Near
zones have significantly higher % LOI than Hinterland Far zone (p<0.001).

25

20

15
% LOI

10

0
Depth (cm) 20 40 60 80 20 40 60 80 20 40 60 80 20 40 60 80
Harbour High Street Hint Near Hint Far

Figure 74: Boxplot of median % LOI for individual zones at Pittenweem for 0-20cm, 20-
40cm, 40-60cm and 60-80cm depth; boundaries of boxes represent interquartile range,
outliers identified as *

- 175 -
Results of a Two Way ANOVA undertaken using a GLM indicate that the Harbour, High
Street and Hinterland Near zones have significantly higher % LOI compared to the
Hinterland Far zone (p<0.001). These results support trends identified using the
appropriate non-parametric Kruskal-Wallis analysis and Dunn’s test in section 6.3.1.1.
Moreover in supplementation to results outlined in the previous paragraph, 0-20cm has a
significantly higher % LOI than 20-40, 40-60 and 60-80cm depths (p<0.001), 20-40cm
has a significantly greater LOI than 40-60cm and 60-80cm (p<0.001), and 40-60cm has
a significantly higher % LOI than 60-80cm (p<0.05).

In contrast to findings obtained at Lauder, statistical analysis using ANOVA indicates

that there is no significant interaction between zone and depth at Pittenweem. This
confirms the observed trend that variation in % LOI between zones is consistent at
successive depths.

6.3.2 Variation between Burghs

The results of a Two Way ANOVA undertaken using a GLM indicate that mean % LOI is
significantly different between burghs (p<0.001). Tukey-Kramer (95% confidence
intervals) multiple comparisons reveal that Wigtown has a significantly higher mean LOI
compared to both Lauder and Pittenweem (p<0.001). However, there is no statistical
difference in mean % LOI between Lauder and Pittenweem.

Results presented in section 6.3.1.1 indicate increased levels of % LOI within the burgh
core of each town, and in the case of Lauder and Pittenweem enhancement within the
Hinterland Near zone. These findings are supported by ANOVA which reveals a
significant difference in % LOI between the High Street and Hinterland Near zones
(p<0.001). Furthermore, Tukey-Kramer multiple comparisons confirm a significantly
higher mean % LOI for the High Street zone compared to Hinterland Near zone
(p<0.001). It is argued that spatial distribution of % LOI at Lauder and Pittenweem is
characterised by decreasing % LOI values in association with distance from the burgh
core.

Statistical analysis suggests that there is no interaction between the individual ANOVA
test factors, burgh and zone. It is, therefore, argued that although differences in % LOI
are identified between the High Street and Hinterland Near zones, the nature and extent
of this variation is consistent in each town.

- 176 -
6.4 Magnetic Susceptibility

Spatial distributions of mass dependant magnetic susceptibility (Ҳ) and frequency

dependant magnetic susceptibility (ҲFD) are presented for Lauder (Figure 75),
Pittenweem (Figure 76) and Wigtown (Figure 77).

-6 3 -1
a) Ҳ (10 m Kg )

-6 3 -1
b) ҲFD (10 m Kg )

Figure 75: Spatial distribution of (a) Ҳ (10-6m3Kg-1) and (b) ҲFD (10-6m3Kg-1) at Lauder. Red
boundary delimits 1862AD urban extent

- 177 -
 a) -6 3 -1
b) -6 3 -1
Ҳ (10 m Kg ) Ҳ FD (10 m Kg )

Figure 76: Spatial distribution of (a) Ҳ (10-6m3Kg-1) and (b) ҲFD (10-6m3Kg-1) at Pittenweem.
Red boundary delimits 1855AD urban extent

- 178 -
 a) b)

-6 3 -1 -6 3 -1
Ҳ (10 m Kg ) Ҳ FD (10 m Kg )

Figure 77: Spatial distribution of (a) Ҳ (10-6m3Kg-1) and (b) ҲFD (10-6m3Kg-1) at Wigtown. Red
boundary delimits 1850AD urban extent

6.4.1 Mass Dependant Magnetic Susceptibility (Ҳ)

6.4.1.1 Variation within Burghs

The median Ҳ of the High Street, Hinterland Near, Hinterland Far and Thirlstane zones
at Lauder is 2.29, 2.17, 0.54 and 0.58 (10-6m3Kg-1) respectively (Figure 78a). There is no
statistical difference in median Ҳ between the High Street zone and Hinterland Near
zone. However, both High Street and Hinterland Near zones have significantly higher Ҳ
than the Hinterland Far (p<0.001) and Thirlstane zones (p<0.001) (Figure 78b). These
results provide clear evidence for increased Ҳ within the urban core and land
immediately adjacent to the historical burgh limits (Hinterland Near).

- 179 -
 a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761
5

High Street
4

Hint Near

Hint Far

0 Thirlstane
Hint far
High S treet Hint N ear H int F ar Thirlstane Hint Near

-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 78: (a) Boxplot of median Ҳ (10-6m3Kg-1) for individual zones at Lauder; boundaries
of boxes represent interquartile range, red shading indicates 86.761% confidence interval
for median, outliers identified as *, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Lauder, Z values exceeding –Z or Z are significant at
p<0.05 confidence level

The mean Ҳ of the Harbour, High Street, Hinterland Near and Hinterland Far zones is
2.21, 2.03, 1.99 and 1.14 (10-6m3Kg-1) respectively (Figure 79). Results of One Way
ANOVA with Tukey-Kramer multiple comparisons indicate the Harbour, High Street and
Hinterland Near zones all have a significantly higher mean Ҳ than the Hinterland Far
zone (p<0.001). However, there is no statistical difference in mean Ҳ between the
Harbour, High Street and Hinterland Near zones. Similar to results obtained for Lauder,
elevated Ҳ is noted within both burgh core and in the immediate hinterland (Hinterland
Near).

There is a significant difference between the High Street, Hinterland Near and Showfield
zones at Wigtown (p<0.001). The median Ҳ of the High Street, Hinterland Near and
Showfield zones is 7.26, 1.71 and 1.95 (10-6m3Kg-1) respectively (Figure 80a). The High
Street zone has a significantly higher (p<0.001) Ҳ than both the Hinterland Near and
Showfield zones (Figure 80b). There is no significant difference between the Hinterland
Near and Showfield zones. These results indicate elevated Ҳ within the burgh core at
Wigtown.

- 180 -
 3.5

3.0

2.5

2.0

1.5

1.0

0.5

Harbour High Street Hint Near Hint Far

Figure 79: Boxplot of median and mean Ҳ (10-6m3Kg-1) for individual zones at Pittenweem;
bar indicates median, black circle denotes mean, boundaries of boxes represent
interquartile range, outliers identified as *

a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons

Desired Conf idence: 80.529
12

10 High Street Hint Near

8
Showf ield

2 Hint Near Showf ield

0
High Street Hint Near Showf ield
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.067 |Bonf erroni Z-v alue|: 1.834

Figure 80: (a) Boxplot of median Ҳ (10-6m3Kg-1) for individual zones at Wigtown;
boundaries of boxes represent interquartile range, red shading indicates 80.529%
confidence interval for median, outliers identified as *, (b) Kruskal-Wallis analysis with
Dunn’s test multiple comparisons for individual zones at Wigtown, Z values exceeding –Z
or Z are significant at p<0.05 confidence level

- 181 -
6.4.1.2 Variation between Burghs
Results of a Two Way ANOVA undertaken using a GLM indicate that mean Ҳ is
significantly different between towns (p<0.001). Tukey-Kramer (95% confidence
intervals) multiple comparisons reveal that mean Ҳ at Wigtown is significantly higher
than Lauder and Pittenweem (p<0.001). Additionally, there is no statistical difference in
Ҳ between Lauder and Wigtown. It should be noted that variation in mean Ҳ between
burghs does not necessarily reflect differences in the duration and intensity of
anthropogenic enhancement. It is more likely that absolute differences indicate a
contrast in underlying geology.

Results presented in section 6.4.1.1 indicate increased Ҳ within the burgh core of each
town, and in the case of Lauder and Pittenweem, enhancement within the Hinterland
Near zone. This finding is supplemented by ANOVA which confirms a significant
difference in mean Ҳ between the High Street zone and Hinterland Near zone (p<0.001).
Tukey-Kramer multiple comparisons indicate a higher mean Ҳ for the High Street zone
compared to the Hinterland Near zone (p<0.001). Conversely, results from Kruskal-
Wallis pairwise comparisons suggest that there is no significant difference between the
High Street and Hinterland Near zones at Lauder and Pittenweem. This discrepancy
may reflect differences in measurement of central tendency between parametric and
non-parametric tests. Nevertheless it is evident that all three towns have elevated Ҳ in
the burgh core, and in the case of Lauder and Pittenweem enhanced Ҳ in land
immediately adjacent to the historical burgh limits (Hinterland Near).

Statistical analysis indicates significant interaction (p<0.001) between the individual

ANOVA test factors, burgh and zone. This association implies that variation in Ҳ cannot
be explained by either burgh or zone as factors in isolation. It is therefore argued that
although a difference between the High Street and Hinterland Near zone is identified
using ANOVA, the nature and extent of this difference is not consistent for all three
burghs.

- 182 -
6.4.2 Frequency Dependant Magnetic Susceptibility (ҲFD)

6.4.2.1 Variation within Burghs

The median ҲFD of the High Street, Hinterland Near, Hinterland Far and Thirlstane zones
at Lauder is 0.12, 0.16, 0.05 and 0.04 (10-6m3Kg-1) respectively (Figure 81a). The High
Street has a significantly higher ҲFD than the Hinterland Far (p<0.001) and Thirlstane
zones (p<0.01). In addition, the Hinterland Near zone has a significantly greater median
ҲFD than the High Street (p<0.05), Hinterland Far (p<0.001) and Thirlstane zones
(p<0.001) (Figure 64b). There is no statistical difference between the Hinterland Far and
Thirlstane zones. These results reveal enhancement of ҲFD in both the burgh core and in
land adjacent to the historical burgh limits (Hinterland Near). Moreover, it is recognised
that ҲFD is significantly higher in the Hinterland Near zone in comparison to the burgh
core.

Similar to trends identified at Lauder, enhancement of ҲFD is noted within the burgh core
and Hinterland Near zone at Pittenweem. The median ҲFD of the Harbour, High Street,
Hinterland Near and Hinterland Far zones at Pittenweem is 0.06, 0.07, 0.09 and 0.04
(10-6m3Kg-1) respectively. The High Street has a significantly higher ҲFD than the
Hinterland Far zone (p<0.01). Moreover, the Hinterland Near zone has a significantly
higher ҲFD than both the High Street (p<0.01) and Hinterland Far zones (p<0.001)
(Figure 82b). There is no statistical difference in ҲFD between the Harbour zone with the
High Street, Hinterland Near or Hinterland Far zones. This is attributable to pronounced
data variability as indicated by a large interquartile range (Figure 82a). In addition, there
is considerable overlapping between the Harbour zone confidence interval with the High
Street, Hinterland Near and Hinterland Far zones. These results provide clear evidence
for enhanced ҲFD in the burgh core and land immediately adjacent to the historical burgh
limits (Hinterland Near). In addition, it should be noted that ҲFD is significantly higher in
the Hinterland Near zone compared to the burgh core.

- 183 -
 a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761

0.20 High Street Hint Near

Hint Far

0.15 Thirlstane

0.10
Hint Near Hint Far

Thirlstane

0.05

Hint Far Thirlstane

0.00

High Street Hint Near Hint Far Thirlstane

-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 81: (a) Boxplot of median ҲFD (10-6m3Kg-1) for individual zones at Lauder;
boundaries of boxes represent interquartile range, red shading indicates 86.761%
confidence interval for median, outliers identified as *, (b) Kruskal-Wallis analysis with
Dunn’s test multiple comparisons for individual zones at Lauder, Z values exceeding –Z or
Z are significant at p<0.05 confidence level

The median ҲFD of the High Street, Hinterland Near and Showfield zones at Wigtown is
0.47, 0.12 and 0.18 (10-6m3Kg-1) respectively Figure 83a). The High Street has a
significantly higher ҲFD than the Hinterland Near (p<0.001) and Showfield zones
(p<0.001) (Figure 83b). Additionally the Showfield zone has a significantly higher ҲFD
than the Hinterland Near zones (p<0.05). Despite the difference in ҲFD between the
Showfield and Hinterland Near zone, neither are enhanced in comparison to the High
Street zone. It is, therefore, argued that these results highlight a clear contrast in ҲFD
between the burgh core and its hinterland.

- 184 -
 Boxplots w ith Sign Confidence Intervals Pairw ise Comparisons
a) Desired Conf idence: 86.761 b)
0.14
Harbour High Street

0.12
Hint Near

0.10 Hint Far

0.08

High Street Hint Near

0.06

Hint Far
0.04

0.02
Hint Near Hint Far

0.00
Harbour High Street Hint Near Hint Far
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 82: (a) Boxplot of median ҲFD (10-6m3Kg-1) for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, red shading indicates 86.761%
confidence interval for median, outliers identified as *, (b) Kruskal-Wallis analysis with
Dunn’s test multiple comparisons for individual zones at Pittenweem, Z values exceeding
–Z or Z are significant at p<0.05 confidence level

Boxplots w ith Sign Confidence Intervals Pairw ise Comparisons

a) Desired Conf idence: 80.529
b)
0.8

0.7
High Street Hint Near

0.6

0.5 Showf ield

0.4

0.3

0.2

Hint Near Showf ield

0.1

0.0
High Street Hint Near Showf ield
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.067 |Bonf erroni Z-v alue|: 1.834

Figure 83: (a) Boxplot of median Ҳ (10-6m3Kg-1) for individual zones at Wigtown;
boundaries of boxes represent interquartile range, red shading indicates 80.529%
confidence interval for median, outliers identified as *, (b) Kruskal-Wallis analysis with
Dunn’s test multiple comparisons for individual zones at Wigtown, Z values exceeding –Z
or Z are significant at p<0.05 confidence level

- 185 -
6.4.2.2 Variation between Burghs
Results of a Two Way ANOVA undertaken using a GLM indicate that the mean ҲFD is
significantly different between burghs (p<0.001). Tukey-Kramer (95% confidence
intervals) multiple comparisons reveal that the mean ҲFD at Wigtown is significantly
higher than Lauder and Pittenweem (p<0.01). Furthermore, Lauder has a significantly
greater (p<0.01) mean ҲFD than Pittenweem.

Statistical analysis also reveals a significant difference between the High street and
Hinterland Near zone (p<0.001). Results of ANOVA indicate significantly higher ҲFD for
the High Street zone in comparison to the Hinterland Near zone (p<0.001). This finding
is contradictory to trends identified for Lauder and Pittenweem in section 6.4.2.1 which
are characterised by higher ҲFD in the Hinterland Near zone. This discrepancy arises
from the computation of ANOVA which collectively tests High Street and Hinterland Near
zone data for all three burghs. It is suggested that differences in the relationship
between the High Street and Hinterland Near zone data between burghs may produce
contradictory results to those obtained using Kruskal-Wallis analysis and Dunn’s test.

Statistical analysis indicates significant interaction between the individual ANOVA test
factors, burgh and zone. This finding supports results presented in 6.4.2.1 which reveal
the occurrence of higher ҲFD in the Hinterland Near zone at Lauder and Pittenweem, and
comparatively higher ҲFD in the High Street zone at Wigtown.

6.4.3 Natural Variation of Magnetic Susceptibility

Magnetic susceptibility of soil is influenced by a range of geological and pedological

factors which determine Iron (Fe) content and abundance of magnetic minerals
(Crowther 2003, Crowther and Barker 1995). Consequently areas of increased magnetic
susceptibility can be attributed to environmental factors such as a change in underlying
geology, and human activities. Correlation between Fe (mg/Kg) and Ҳ (10-6m3Kg-1)
(Figure 84) shows a distinction between Wigtown and Lauder, and Wigtown and
Pittenweem.

This finding supplements results presented in section 6.4.2.1, which indicate that there is
no significant difference in mean Ҳ between Lauder and Pittenweem, but do reveal a
significant difference in mean Ҳ between Wigtown with Lauder and Pittenweem. It is
therefore suggested that variation in mean Ҳ between burghs reflects differing geology.
In addition it is apparent that there are different phases in the relationship between Ҳ
and Fe at Lauder and Wigtown. Further analysis of data indicates that there is

- 186 -
separation of the High Street and Hinterland Near zones at Lauder, and High Street
zone at Wigtown. This is in agreement with results presented in section 6.4.1.1. In the
natural environment Ҳ is largely determined by concentrations of ferrimagnetic minerals.
It is therefore expected that variations in Ҳ are strongly correlated with Fe
concentrations. Results of Spearman Rank statistical analyses indicate a weak positive
correlation between Ҳ and Fe at Lauder (rs 0.253), Pittenweem (rs 0.206) and Wigtown
(rs 0.380). These results suggest that spatial distributions of Ҳ are not closely related to
Fe concentrations. It is therefore argued that trends in Ҳ and ҲFD identified in sections
6.4.1 and 6.4.2, can be partially attributed to factors other than natural variation.

Comparison between ‘reference’ soil profile Ҳ with functional zones in each burgh also
confirm enhancement exceeding natural conditions. At Lauder Ҳ is five times higher in
the High Street (2.29) and Hinterland Near (2.03) zone compared to reference profile LA
7 (0.33). At Pittenweem Ҳ is nearly double in the Harbour (2.04), High Street (2.21) and
Hinterland Near (1.99) zones compared to reference profile PT 1 (1.01). Similarly Ҳ is
four times higher in the High street (7.06) zone at Wigtown in comparison to reference
soil profile WG 5 (1.46).

12
Lauder
Pittenweem
10
Wigtown

10000 15000 20000 25000 30000 35000

Fe (mg/Kg)

Figure 84: Scatterplot of Fe (mg/Kg) vs. Ҳ (10-6m3Kg-1) for Lauder, Pittenweem and
Wigtown

- 187 -
6.5 Summary of Results: Soil Physical and Chemical Properties

The following section presents a brief summary of key trends in soil physical and
chemical properties both within and between burghs. The significance of these results is
discussed in section 8.1.1 and 8.2.1. Consideration is given to differences in soil
modification between burgh cores and their hinterland as a result of past waste disposal.

6.5.1 Variation within Burghs

6.5.1.1 Lauder
Two zones are identified as important areas of interest at Lauder; the High Street zone
and the Hinterland Near zone. Both of these zones are characterised by deepened
topsoils, higher % LOI and enhanced mass dependant magnetic susceptibility (Table
10). Enhancement in ҲFD is also noted within the High Street and Hinterland Near zones,
however ҲFD is comparatively higher in the Hinterland Near zone. Moreover a distinct
contrast in soil pH is noted between these two zones. The High Street has
characteristically neutral soils in comparison to Hinterland Near zone which is acidic.

6.5.1.2 Pittenweem
The Harbour, High Street and Hinterland Near zones are identified as important areas of
interest at Pittenweem. These zones are characterised by deepened topsoils, enhanced
mass dependant magnetic susceptibility and higher % LOI (Table 10). The Harbour zone
has the highest % LOI, followed in succession by the High Street and Hinterland Near
zones. Enhancement of ҲFD is identified in all three zones at Pittenweem, however it
should be noted that ҲFD is comparatively higher in the Hinterland Near zone. Moreover
a distinct contrast in soil pH is noted between the Hinterland Near zone which is acidic,
with the Harbour and High Street zones which have neutral soil pH.

6.5.1.3 Wigtown
The High Street zone is identified as an important area of interest at Wigtown.
Compared to the surrounding hinterland the High Street zone has deepened topsoils, a
neutral soil pH, higher % LOI, and enhanced magnetic susceptibility and ҲFD (Table 10).

6.5.1.4 Depth
Results presented in section 6.2.1.2 reveal a significant difference in soil pH between
depths at Lauder and Pittenweem (Table 11). In both cases soil pH increases with depth
whilst retaining spatial trends identified at 0-20cm, for example there is a consistent
distinction between neutral soil pH values in the burgh core with acidic hinterland soils.

- 188 -
Results presented in section 6.3.1.2 show a decrease in % LOI associated with depth at
both Lauder and Pittenweem, however, the relationship between zone and depth differs
between these two burghs. Given the absence of interaction between depth and zone at
Pittenweem it is argued that spatial trends identified at 0-20cm are consistent at
successive depths. Conversely statistical interaction between depth and zone at Lauder
signifies that variations in % LOI at subsequent depths are not fully attributable to
differenced associated with zones.

Table 10: Summary of trends in soil physical and chemical properties identified at Lauder,
Pittenweem and Wigtown

Measurement Lauder Pittenweem Wigtown

Topsoil Depth Deepened topsoil in Deepened topsoil in Deepened topsoil in
burgh core (High burgh core (Harbour burgh core (High
(cm)
Street zone) and and High Street Street zone)
immediate hinterland zones) and
(Hinterland Near immediate hinterland
zone) (Hinterland Near
zone)
pH Neutral pH in burgh Neutral pH in burgh Neutral pH in burgh
core (High Street core (Harbour and core (High Street
zone) and acidic pH High Street zones) zone) and acidic pH
in the hinterland and acidic pH in the in the hinterland
(Hinterland Near, hinterland (Hinterland (Hinterland Near and
Hinterland Far and Near and Hinterland Showfield zones)
Thirlstane zones) Far zones)
% LOI Enhanced % LOI in Enhanced % LOI in Enhanced % LOI in
burgh core (High burgh core (Harbour burgh core (High
Street zone) and and High Street Street zone)
immediate hinterland zones) and
(Hinterland Near immediate hinterland
zone) (Hinterland Near
zone)
Ҳ (10-6m3Kg-1) Enhanced Ҳ in burgh Enhanced Ҳ in burgh Enhanced Ҳ in
core (High Street core (High Street burgh core (High
zone) and immediate zone) and immediate Street zone)
hinterland (Hinterland hinterland (Hinterland
Near zone) Near zone)
ҲFD (10-6m3Kg-1) Enhanced ҲFD in Enhanced ҲFD in Enhanced ҲFD in
burgh core (High burgh core (High burgh core (High
Street zone) and Street zone) and Street zone)
immediate hinterland immediate hinterland
(Hinterland Near (Hinterland Near
zone). Enhancement zone). Enhancement
greatest in Hinterland greatest in Hinterland
Near zone Near zone

- 189 -
Table 11: Comparison of trends in pH and LOI associated with depth at Lauder and
Pittenweem

Measurement Lauder Pittenweem

pH Increase in pH associated with Increase in pH associated with
depth. Spatial trends present at depth. Spatial trends identified at
0-20cm occur at 20-40cm. 0-20cm are present at 20-40, 40-
No significant interaction 60 and 60-80cm depth. No
between zone and depth significant interaction between
zone and depth
% LOI Decrease in % LOI associated Decrease in % LOI associated
with depth. Spatial trends with depth. Spatial trends
identified at 0-20cm are not identified at 0-20cm are present
consistent at 20-40cm. at 20-40, 40-60 and 60-80cm
Significant interaction between depth. No significant interaction
zone and depth between zone and depth

6.5.2 Variation between Burghs

All three burghs have deepened topsoil, higher % LOI, and enhanced magnetic
susceptibility and ҲFD in zones corresponding to the burgh core (Harbour and High
Street). Trends identified in the burgh core at Lauder and Pittenweem are also replicated
in land immediately adjacent to the historical burgh limits (Hinterland Near), although it
should be noted that ҲFD is comparatively higher in the Hinterland Near zone (Table 12).
Moreover, a distinct contrast in soil pH is noted between the burgh core and Hinterland
Near zone at all three burghs. The burgh core has characteristically neutral soil in
contrast to the Hinterland Near zone which has typically acidic soils.

Table 12: Comparison of trends in soil physical and chemical properties identified at
Lauder, Pittenweem and Wigtown

Measurement Comparison between Burghs

Topsoil Depth All three burghs have deepened topsoil deposits in the burgh core.
Lauder and Pittenweem have deepened deposits in the immediate
(cm)
hinterland (Hinterland Near zone)
pH All three burghs have higher a pH in the burgh core.
All three burghs show a contrast between neutral pH in the old
burgh core and acidic pH in the hinterland
% LOI All three burghs have enhanced LOI in the burgh core.
Lauder and Pittenweem have enhanced LOI in the immediate
hinterland (Hinterland Near zone)
Ҳ (10-6m3Kg-1) All three burghs have higher Ҳ in the burgh core.
Lauder and Pittenweem have enhanced Ҳ in the immediate
hinterland (Hinterland Near zone)
ҲFD (10-6m3Kg-1) All three burghs have enhanced ҲFD in the burgh core.
Lauder and Pittenweem have enhanced ҲFD in the immediate
hinterland (Hinterland Near). Enhancement of ҲFD is most
pronounced in the Hinterland Near zone at Lauder and Pittenweem

- 190 -
 7 Elemental Analyses

This chapter presents the results of a series of elemental analyses including multiple
comparisons of anthropogencially significant elements, cluster analysis and discriminant
analysis. Data are presented collectively for Lauder, Pittenweem and Wigtown to
facilitate visual and statistical analysis.

7.1 Elemental Concentrations

The following section presents spatial distributions of selected elemental concentrations

including, barium (Ba), calcium (Ca), lead (Pb), potassium (K), phosphorus (P),
strontium (Sr) and zinc (Zn). These elements were selected for detailed analysis
because of their utility in characterising areas of past human activity (Entwistle et al.,
2000a, Entwistle et al., 2000b, Entwistle et al., 1998, Wilson et al., 2006a).

7.1.1 Barium

Spatial distributions of barium (Ba) are presented for Lauder (Figure 85), Pittenweem
(Figure 86a) and Wigtown (Figure 86b). Patterns in Ba associated with zones are
subsequently identified and comparisons between burghs are made.

Figure 85: Spatial distribution of Log Ba (mg/Kg) at Lauder. Red boundary delimits 1862AD
urban extent

- 191 -
 a) b)

Figure 86: Spatial distribution of Log Ba (mg/Kg) at Pittenweem (a) and Wigtown (b). Red
boundary delimits 1855AD and 1850AD urban extent respectively

7.1.1.1 Variation within Burghs

The mean Ba concentration of the High Street, Hinterland Near, Hinterland Far and
Thirlstane zones at Lauder is 256, 189, 151 and 228 mg/Kg respectively (Figure 87).
Results of One Way ANOVA with Tukey-Kramer multiple comparisons indicate that both
the High Street and Thirlstane zones have significantly higher Ba than the Hinterland
Near and Hinterland Far zones (p<0.001). Moreover, Ba is significantly higher in the
Hinterland Near zone compared to the Hinterland Far zone (p<0.001). There is no
significant difference in mean Ba concentration between the High Street and Thirlstane
zones. These results indicate similar patterns of Ba enhancement in the burgh core and
immediate hinterland north of Lauder (Thirlstane). To a lesser extent Ba is also
enhanced in the immediate hinterland south of Lauder (Hinterland Near).

- 192 -
 350

300

250
Ba (mg/Kg)

200

150

100

High Street Hint Near Hint Far Thirlstane

Figure 87: Boxplot of median and mean Ba (mg/Kg) for individual zones at Lauder; bar
indicates median, black circle denotes mean; boundaries of boxes represent interquartile
range, outliers identified as *

The median Ba concentration of the Harbour, High Street, Hinterland Near and
Hinterland Far zones at Pittenweem is 153.4, 147.7, 62.5 and 61.6 mg/Kg respectively
(Figure 88a). Both the Harbour and High Street zones have significantly higher Ba than
the Hinterland Near and Hinterland Far zones (p<0.001) (Figure 88b). There is no
significant difference in median Ba concentrations between the Harbour zone and High
street zones. Similarly, there is no statistical difference in Ba between the Hinterland
Near and Hinterland far zones. These results provide clear evidence of Ba enhancement
within the burgh core at Lauder.

The median Ba concentration for the High street, Hinterland Near and Showfield zones
at Wigtown is 236.6, 48.9 and 94.9 mg/Kg respectively (Figure 89a). The High Street
zone has significantly higher Ba than the Hinterland Near and Showfield zones
(p<0.001). In addition the Showfield zone has significantly higher Ba than the Hinterland
Near zone (p<0.001) (Figure 89b). These results indicate enhanced levels of Ba in the
burgh core and to a lesser extent the immediate hinterland south of Wigtown
(Showfield).

- 193 -
 a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761
300

H arbour
250

200
Ba (mg/Kg)

H igh S treet
150

100

Hint N ear
Hint Far
50
Hint Near
High Street
Harbour High Street Hint Near Hint Far
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonferroni Z-v alue|: 2.128

Figure 88: (a) Boxplot of median Ba (mg/Kg) for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, green shading indicates 86.761%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis analysis with
Dunn’s test multiple comparisons for individual zones at Pittenweem, -Z or Z are
significant at p<0.05 confidence level
Boxplots w ith Sign Confidence Intervals Pairw ise Comparisons
a) Desired Conf idence: 80.529
b)
500

High Street Hint Near

400

Showf ield
300
Ba (mg/Kg)

200

100
Hint Near Hint Near

0
High Street Hinterland Near Showf ield
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.067 |Bonf erroni Z-v alue|: 1.834

Figure 89: (a) Boxplot of median Ba (mg/Kg) for individual zones at Wigtown; boundaries
of boxes represent interquartile range, green shading indicates 80.529% confidence
interval for median, outliers identified as*, (b) Kruskal-Wallis analysis with Dunn’s test
multiple comparisons for individual zones at Wigtown, -Z or Z are significant at p<0.05
confidence level

- 194 -
7.1.1.2 Variation between Burghs
All three burghs have enhanced levels of Ba in zones corresponding to the burgh core
(Harbour and High Street). Enhancement of Ba is confined to Pittenweem’s burgh core;
however, both Lauder and Wigtown have additionally enhanced zones in the hinterland.
Ba is enhanced in the immediate hinterland north of Lauder (Thirlstane) to a level
comparable with the burgh core, and to a lesser extent in the immediate hinterland south
of the historical burgh limits (Hinterland Near). Similarly Ba is enhanced in the immediate
hinterland south of Wigtown (Showfield), albeit significantly less than the High Street
zone.

7.1.2 Calcium

Spatial distributions of calcium (Ca) are presented for Lauder (Figure 90), Pittenweem
(Figure 91a) and Wigtown (Figure 91b). Patterns in Ca associated with zones are
subsequently identified and comparisons between burghs are made.

Figure 90: Spatial distribution of Log Ca (mg/Kg) at Lauder. Red boundary delimits 1862AD
urban extent

- 195 -
a) b)

Figure 91: Spatial distribution of Log Ca (mg/Kg) at Pittenweem (a) and Wigtown (b). Red
boundary delimits 1855AD and 1850AD urban extent respectively

7.1.2.1 Variation within Burghs

The median Ca concentration of the High Street, Hinterland Near, Hinterland Far and
Thirlstane zones at Lauder is 9076, 2959, 1844 and 2458 mg/Kg respectively (Figure
92a). The High Street has significantly higher Ca than the Hinterland Near, Hinterland
Far and Thirlstane zones (p<0.001). The Hinterland Near zone has significantly higher
Ca than the Hinterland Far zone (p<0.001) (Figure 92b). However there is no statistical
difference in median Ca concentrations between the Hinterland Near and Thirlstane
zones. These results indicate enhancement of Ca within Lauder’s burgh core. Elevated
Ca levels are also identified in the immediate hinterland south of Lauder (Hinterland
Near); however it should be noted that Ca concentrations in the Hinterland Near zone
are significantly lower than the High Street zone.

- 196 -
 a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761

20000

High Street Hint Near

Hint Far

15000
Thirlstane
Ca (mg/Kg)

10000
Hint Near Hint Far

Thirlstane

5000

Hint Far Thirlstane

0
High Street Hint Near Hint Far Thirlstane
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 92: (a) Boxplot of median Ca (mg/Kg) for individual zones at Lauder; boundaries of
boxes represent interquartile range, yellow shading indicates 86.761% confidence interval
for median, outliers identified as*, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Lauder, -Z or Z are significant at p<0.05 confidence
level

The median Ca concentration of the Harbour, High Street, Hinterland Near and
Hinterland Far zones is at Pittenweem is 17414, 11756, 2580 and 2608.6 mg/Kg
respectively (Figure 93a). Both the Harbour and High Street zones have significantly
higher Ca than the Hinterland Near and Hinterland Far zones (p<0.001). There is no
significant difference in median Ca concentrations between the Harbour and High Street
zones. Moreover, there is no statistical difference in Ca concentrations between the
Hinterland Near and Hinterland Far zones (Figure 93b). These results provide a clear
indication of Ca enhancement within the burgh core at Pittenweem.

The median Ca concentration of the High Street, Hinterland Near and Showfield zones
at Wigtown is 17038, 1522 and 1143 mg/Kg respectively (Figure 94a). The High Street
has significantly higher Ca than the Hinterland Near and Showfield zones (p<0.001). In
addition, there is no significant difference in median Ca concentrations between the
Hinterland Near and Showfield zones. Similar to results identified at Pittenweem, there is
distinct enhancement of Ca within the burgh core at Wigtown.

- 197 -
 a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761
30000

Harbour
25000

20000
Ca (mg/Kg)

15000 High Street

10000

5000
Hint Near

Hint Far
Hint Near
0 High Street
Harbour High Street Hint Near Hint Far
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 93: (a) Boxplot of median Ca (mg/Kg) for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, yellow shading indicates 86.761%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis analysis with
Dunn’s test multiple comparisons for individual zones at Pittenweem, -Z or Z are
significant at p<0.05 confidence level

a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons

Desired Conf idence: 80.529

30000

25000 High Street Hint Near

20000
Showf ield
Ca (mg/Kg)

15000

10000

5000 Hint Near Showf ield

High Street Hint Near Showf ield

-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.067 |Bonf erroni Z-v alue|: 1.834

Figure 94: (a) Boxplot of median Ca (mg/Kg) for individual zones at Wigtown; boundaries
of boxes represent interquartile range, yellow shading indicates 80.529% confidence
interval for median, outliers identified as*, (b) Kruskal-Wallis analysis with Dunn’s test
multiple comparisons for individual zones at Wigtown, -Z or Z are significant at p<0.05
confidence level

- 198 -
7.1.2.2 Variation between Burghs
Similar to trends identified for Ba in section 7.1.1, all three burghs have significant
enhancement of Ca in zones within the burgh core (Harbour and High street). Increased
Ca concentrations are characteristically limited to the burgh core within Pittenweem and
Lauder. However, comparatively higher Ca levels are noted in the immediate hinterland
south of Lauder (Hinterland Near). Nevertheless it should be noted that this area has a
significantly lower median Ca concentration than the High Street zone.

7.1.3 Lead

Spatial distributions of lead (Pb) are presented for Lauder (Figure 95), Pittenweem
(Figure 96a) and Wigtown (Figure 96b). Patterns in Pb associated with zones are
subsequently identified and comparisons between burghs are made.

Figure 95: Spatial distribution of Log Pb (mg/Kg) at Lauder. Red boundary delimits 1862AD
urban extent

- 199 -
a) b)

Figure 96: Spatial distribution of Log Pb (mg/Kg) at Pittenweem (a) and Wigtown (b). Red
boundary delimits 1855AD and 1850AD urban extent respectively

7.1.3.1 Variation within Burghs

The median Pb concentration of the High Street, Hinterland Near, Hinterland Far and
Thirlstane zones at Lauder is 209.9, 77.7, 32.35 and 59.1 mg/Kg respectively (Figure
97a). The High Street zone has significantly higher Pb than Hinterland Near (p<0.05),
Hinterland Far (p<0.001) and Thirlstane (p<0.001) zones. The Hinterland Near zone has
significantly higher Pb than Hinterland Far (p<0.001) and Thirlstane zones (p<0.05).
Moreover, the Thirlstane zone has significantly higher Pb than the Hinterland Far zone
(p<0.001) (Figure 97). These results indicate significant enhancement of Pb within the
burgh core. To a lesser extent, elevated levels of Pb are also identified in the immediate
hinterland south (Hinterland Near) and north of Lauder (Thirlstane).

- 200 -
a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761
1800

High Street Hint Near

1600

Hint Far
1400

Thirlstane
1200
Pb (mg/Kg)

1000

800 Hint Near Hint Far

600 Thirlstane

400

200
Hint Far Thirlstane
0

High Street Hint Near Hint Far Thirlstane

-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 97: (a) Boxplot of median Pb (mg/Kg) for individual zones at Lauder; boundaries of
boxes represent interquartile range, green shading indicates 86.761% confidence interval
for median, outliers identified as*, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Lauder, -Z or Z are significant at p<0.05 confidence
level

The mean Pb concentration of the Harbour, High Street, Hinterland Near and Hinterland
Far zones at Pittenweem is 221, 182, 43.3 and 34.866 mg/Kg respectively (Figure 98).
Results of One Way ANOVA with Tukey-Kramer multiple comparisons indicate that Pb
levels are significantly higher in the Harbour zone compared to the Hinterland Near and
Hinterland Far zones (p<0.001). Likewise the High Street zone has significantly higher
Pb than the Hinterland Near and Hinterland Far zones (p<0.001). There is no statistical
difference between the Harbour and High Street zones, and the Hinterland Near and
Hinterland Far zones. These results provide clear evidence of enhanced Pb
concentrations within Pittenweem’s burgh core.

The median Pb concentration of the High Street, Hinterland Near and Showfield zones
at Wigtown is 373.9, 63.18 and 151.31 mg/Kg respectively (Figure 99a). The High Street
zone has significantly higher Pb than the Hinterland Near (p<0.001) and Showfield
zones (p<0.01). Moreover, the Showfield zone has a significantly higher median Pb
concentration compared to the Hinterland Near zone (p<0.001) (Figure 82b).

- 201 -
 1400

1200

1000
Pb (mg/Kg)

800

600

400

200

Harbour High Street Hint Near Hint Far

Figure 98: Boxplot of median and mean Pb (mg/Kg) for individual zones at Pittenweem;
bar indicates median, black circle denotes mean; boundaries of boxes represent
interquartile range, outliers identified as *

a) Boxplots w ith Sign Confidence Intervals

b) Pairw ise Comparisons
Desired Conf idence: 80.529

3500

3000
High Street Hint Near

2500

Showf ield
2000
Pb (mg/Kg)

1500

1000

500 Hint Near Showf ield

High Street Hint Near Showf ield

-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.067 |Bonf erroni Z-v alue|: 1.834

Figure 99: (a) Boxplot of median Pb (mg/Kg) for individual zones at Wigtown; boundaries
of boxes represent interquartile range, green shading indicates 80.529% confidence
interval for median, outliers identified as*, (b) Kruskal-Wallis analysis with Dunn’s test
multiple comparisons for individual zones at Wigtown, -Z or Z are significant at p<0.05
confidence level

- 202 -
These results indicate enhanced Pb levels within the burgh core at Wigtown. To a lesser
extent enhanced Pb levels are identified in the hinterland south of Wigtown (Showfield),
however it should be noted that median Pb concentrations are significantly lower
compared to the burgh core.

7.1.3.2 Variation between Burghs

All three burghs have significant enhancement of Pb concentrations in zones within the
burgh core (Harbour and High Streets). Enhancement of Pb at Pittenweem is confined to
the burgh core; however, both Lauder and Wigtown have additionally enhanced zones in
the hinterland. Pb levels are elevated in the immediate hinterland south (Hinterland
Near) and north of Lauder (Thirlstane). Similarly Pb is enhanced in the immediate
hinterland south of Wigtown (Showfield). It should be noted that Pb concentrations in
these hinterland areas are significantly less than the High Street zone.

7.1.4 Potassium

Spatial distributions of potassium (K) are presented for Lauder (Figure 100), Pittenweem
(Figure 101a) and Wigtown (Figure 101b). Patterns in K associated with zones are
subsequently identified and comparisons between burghs are made.

Figure 100: Spatial distribution of Log K (mg/Kg) at Lauder. Red boundary delimits 1862AD
urban extent

- 203 -
a) b)

Figure 101: Spatial distribution of Log Pb (mg/Kg) at Pittenweem (a) and Wigtown (b). Red
boundary delimits 1855AD and 1850AD urban extent respectively

7.1.4.1 Variation within Burghs

The median K concentration of the High Street, Hinterland Near, Hinterland Far and
Thirlstane zones at Lauder is 854.3, 474.36, 205.6 and 284.6 mg/Kg respectively (Figure
102a). The High Street zone has significantly higher K than the Hinterland Near,
Hinterland Far and Thirlstane zones (p<0.001). Moreover K is significantly higher in the
Hinterland Near zone compared to the Hinterland Far zone (p<0.001) (Figure 102b).
There is no statistical difference in K concentrations between the Hinterland Far and
Thirlstane zones. These results provide evidence for enhancement of K in the burgh
core. To a lesser extent K is also enhanced in the immediate hinterland south of Lauder
(Hinterland Near).

- 204 -
 a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761
1800

High Street Hint Near

1600

Hint Far
1400

Thirlstane
1200

1000
K (mg/Kg)

800
Hint Near Hint Far

600
Thirlstane

400

200

Hint Far Thirlstane

0

High Street Hint Near Hint Far Thirlstane

-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 102: (a) Boxplot of median K (mg/Kg) for individual zones at Lauder; boundaries of
boxes represent interquartile range, pink shading indicates 86.761% confidence interval
for median, outliers identified as*, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Lauder, -Z or Z are significant at p<0.05 confidence
level

The median K concentration of the Harbour, High Street, Hinterland Near and Hinterland
Far zones at Pittenweem is 636, 348, 79. and 126 mg/Kg respectively (Figure 103a).
The Harbour zone has significantly higher levels of K than the Hinterland Near and
Hinterland Far zones (p<0.001). Likewise the High Street zone has significantly higher K
than the Hinterland Near and Hinterland Far zones (p<0.001) (Figure 103b). There is no
significant difference in median K concentrations between the Harbour and High Street
zones, and the Hinterland Near and Hinterland Far zones. These results provide clear
evidence for enhancement of K within Pittenweem’s burgh core.

The median K concentration for High Street, Hinterland Near and Showfield zones at
Wigtown is 870.2, 284.8 and 300.6 mg/Kg respectively (Figure 104a). The High Street
zone has significantly higher K than the Hinterland Near and Showfield zones (p<0.001).
There is no significant difference in median K between the Hinterland Near and
Showfield zones (Figure 104b). Similar to results identified at Pittenweem, there is
distinct enhancement of K within the burgh core at Wigtown.

- 205 -
 a) Boxplots w ith Sign Confidence Intervals
b) Pairw ise Comparisons
Desired Conf idence: 86.761
1000

Harbour High Street

800 Hint Near

Hint Far

600
K (mg/Kg)

High Street Hint Near

400

Hint Far

200

Hint Near Hint Far

0

Harbour High Street Hint Near Hint Far

-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 103: (a) Boxplot of median K (mg/Kg) for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, pink shading indicates 86.761%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis analysis with
Dunn’s test multiple comparisons for individual zones at Pittenweem, -Z or Z are
significant at p<0.05 confidence level

a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons

Desired Conf idence: 80.529
1400

1200
High Street Hint Near

1000

Showf ield

800
K (mg/Kg)

600

400

Hint Near Showf ield

200

0
High Street Hint Near Showf ield
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.067 |Bonf erroni Z-v alue|: 1.834

Figure 104: (a) Boxplot of median K (mg/Kg) for individual zones at Wigtown; boundaries
of boxes represent interquartile range, pink shading indicates 80.529% confidence interval
for median, outliers identified as*, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Wigtown, -Z or Z are significant at p<0.05 confidence
level

- 206 -
7.1.4.2 Variation between Burghs
All three burghs have significant enhancement of K concentrations in zones within the
burgh core (Harbour and High Street). Increased concentrations of K are confined to the
burgh core in Pittenweem and Wigtown; however, elevated levels of K are identified in
the immediate hinterland south of Lauder (Hinterland Near). It should be noted
enhancement of K in the Hinterland Near zone is significantly less than the burgh core at
Lauder.

7.1.5 Phosphorus

Spatial distributions of phosphorus (P) are presented for Lauder (Figure 105),
Pittenweem (Figure 106a) and Wigtown (Figure 106b). Patterns in P associated with
zones are subsequently identified and comparisons between burghs are made.

Figure 105: Spatial distribution of Log P (mg/Kg) at Lauder. Red boundary delimits 1862AD
urban extent

- 207 -
a) b)

Figure 106: Spatial distribution of Log P (mg/Kg) at Pittenweem (a) and Wigtown (b). Red
boundary delimits 1855AD and 1850AD urban extent respectively

7.1.5.1 Variation within Burghs

The median P concentration of the High Street, Hinterland Near, Hinterland Far and
Thirlstane zones is at Lauder is 3129, 1798, 903.4 and 1658 mg/Kg respectively (Figure
107a). The High Street zone has significantly higher P than the Hinterland Near,
Hinterland Far and Thirlstane zones (p<0.001). The Hinterland Near and Thirlstane
zones have significantly higher median P concentrations than the Hinterland Far zone
(p<0.001). Moreover, there is no significant difference in P between the Hinterland Near
and Thirlstane zones (Figure 107b).

- 208 -
 a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761
5000

High Street Hint Near

Hint Far
4000

Thirlstane

3000
P (mg/Kg)

Hint Near Hint Far

2000
Thirlstane

1000

Hint Far Thirlstane

0
High Street Hint Near Hint Far Thirlstane
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 107: (a) Boxplot of median P (mg/Kg) for individual zones at Lauder; boundaries of
boxes represent interquartile range, purple shading indicates 86.761% confidence interval
for median, outliers identified as*, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Lauder, -Z or Z are significant at p<0.05 confidence
level

These results indicate significant enhancement of P within the burgh core. Areas with
elevated P are also identified in the immediate hinterland north (Thirlstane) and south
(Hinterland Near) of Lauder, however it should be noted that P levels in these two areas
are significantly lower than the High Street zone

The median P concentration of the Harbour, High Street, Hinterland Near and Hinterland
Far zones at Pittenweem is 1741, 2147, 964.5 and 960.1 mg/Kg respectively (Figure
108a). The Harbour zone has significantly higher P than the Hinterland Near zone
(p<0.01) and Hinterland Far zone (p<0.05). In addition, the High Street zone has
significantly higher P levels than the Hinterland Near and Hinterland Far zones
(p<0.001) (Figure 108b). There is no significant difference in median P concentrations
between the Harbour and High Street zones. These results provide clear evidence for
enhancement of P within Pittenweem’s burgh core.

- 209 -
 a) Boxplots w ith Sign Confidence Intervals
b) Pairw ise Comparisons
Desired Conf idence: 86.761

5000

Harbour High Street

Hint Near
4000
Hint Far
P (mg/Kg)

3000

High Street Hint Near

2000 Hint Far

1000
Hint Near Hint Far

Harbour High Street Hint Near Hint Far

-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 108: (a) Boxplot of median P (mg/Kg) for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, purple shading indicates 86.761%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis analysis with
Dunn’s test multiple comparisons for individual zones at Pittenweem, -Z or Z are
significant at p<0.05 confidence level

Boxplots w ith Sign Confidence Intervals Pairw ise Comparisons

a) Desired Conf idence: 80.529 b)
9000

8000
High Street Hint Near
7000

6000
Showf ield
P (mg/Kg)

5000

4000

3000

2000
Hint Near Showf ield
1000

0
High Street Hint Near Showf ield
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.067 |Bonf erroni Z-v alue|: 1.834

Figure 109: (a) Boxplot of median P (mg/Kg) for individual zones at Wigtown; boundaries
of boxes represent interquartile range, purple shading indicates 80.529% confidence
interval for median, outliers identified as*, (b) Kruskal-Wallis analysis with Dunn’s test
multiple comparisons for individual zones at Wigtown, -Z or Z are significant at p<0.05
confidence level

- 210 -
The median P concentration of the High Street, Hinterland Near and Showfield zones at
Wigtown is 5001, 1492.5 and 2553 mg/Kg respectively (Figure 109a). The High Street
has significantly higher P than the Hinterland Near (p<0.001) and Showfield zones
(p<0.01). In addition the Showfield zone has significantly higher P than the Hinterland
Near zone (p<0.001). These results indicate significant enhancement of P within the
burgh core. Moreover increased P is noted in the immediate hinterland south of Wigtown
(Showfield), although it should be noted that P concentrations in this area are
significantly lower than the High Street zone.

7.1.5.2 Variation between Burghs

All three burghs have significant enhancement of P concentrations in zones within their
burgh cores (Harbour and High Streets). Enhancement of P at Pittenweem is confined to
the burgh core; however, both Lauder and Wigtown have additionally enhanced zones in
the hinterland. P levels are elevated in the immediate hinterland south (Hinterland Near)
and north of Lauder (Thirlstane). Similarly P is enhanced in the immediate hinterland
south of Wigtown (Showfield). It should be noted that P concentrations in hinterland
areas within Lauder and Wigtown are significantly less than the High Street zone

- 211 -
7.1.6 Strontium

Spatial distributions of strontium (Sr) are presented for Lauder (Figure 110), Pittenweem
(Figure 111a) and Wigtown (Figure 111b). Patterns in Sr associated with zones are
subsequently identified and comparisons between burghs are made.

Figure 110: Spatial distribution of Log Sr (mg/Kg) at Lauder. Red boundary delimits
1862AD urban extent

- 212 -
a) b)

Figure 111: Spatial distribution of Log Sr (mg/Kg) at Pittenweem (a) and Wigtown (b). Red
boundary delimits 1855AD and 1850AD urban extent respectively

7.1.6.1 Variation within Burghs

The median Sr concentration of the High Street, Hinterland Near, Hinterland Far and
Thirlstane zones at Lauder is 78.4, 26.8, 12.1 and 13.4 mg/Kg respectively. The High
street zone has significantly higher Sr than the Hinterland Near zone (p<0.01) and the
Hinterland Far and Thirlstane zones (p<0.001) (Figure 112). In addition the Hinterland
Near zone has significantly higher Sr than the Hinterland Far and Thirlstane zones
(p<0.001). There is no significant difference between the Hinterland Far and Thirlstane
zones. These results indicate enhancement of Sr in the burgh core and land adjacent to
the historical burgh limits (Hinterland Near). However, it should be noted that
enhancement is comparatively higher within the burgh core.

- 213 -
 a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761
160

High Street Hint Near

140
Hint Far

120
Thirlstane

100
Sr (mg/Kg)

80
Hint Near Hint Far

60
Thirlstane

40

20
Hint Far Thirlstane

0
High Street Hint Near Hint Far Thirlstane
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 112: (a) Boxplot of median Sr (mg/Kg) for individual zones at Lauder; boundaries of
boxes represent interquartile range, blue shading indicates 86.761% confidence interval
for median, outliers identified as*, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Lauder, -Z or Z are significant at p<0.05 confidence
level

The median Sr concentration of the Harbour, High Street, Hinterland Near and
Hinterland Far zones at Pittenweem is 102.9, 91.5, 28.5 and 31.5 mg/Kg respectively
(Figure 113a). The Harbour and High Street zones have significantly higher Sr levels
than both the Hinterland Near and Hinterland Far zones (p<0.001). There is no
significant difference in median Sr concentrations between the Harbour and High Street
zones (Figure 113b). These results provide clear evidence for Sr enhancement within
the burgh core.

The median Sr concentration of the High Street, Hinterland Near and Showfield zones at
Wigtown is 153, 11.5 and 13.6 mg/Kg respectively (Figure 114a). The High Street has
significantly higher Sr than the Hinterland Near and Showfield zones (p<0.001). There is
no significant difference in median Sr concentrations between the Hinterland Near and
Showfield zones (Figure 114b). These results indicate significant enhancement of Sr in
Wigtown’s burgh core.

- 214 -
 Boxplots w ith Sign Confidence Intervals Pairw ise Comparisons
Desired Conf idence: 86.761
200

Harbour High Street

Hint Near

150
Hint Far
Sr (mg/Kg)

100
High Street Hint Near

Hint Far

50

Hint Near Hint Far

0
Harbour High Street Hint Near Hint Far
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 113: (a) Boxplot of median Sr (mg/Kg) for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, blue shading indicates 86.761%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis analysis with
Dunn’s test multiple comparisons for individual zones at Pittenweem, -Z or Z are
significant at p<0.05 confidence level
Boxplots w ith Sign Confidence Intervals Pairw ise Comparisons
Desired Conf idence: 80.529

200

High Street Hint Near

150

Showf ield
Sr (mg/Kg)

100

50

Hint Near Showf ield

High Street Hint Near Showf ield

-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.067 |Bonf erroni Z-v alue|: 1.834

Figure 114: (a) Boxplot of median Sr (mg/Kg) for individual zones at Wigtown; boundaries
of boxes represent interquartile range, blue shading indicates 80.529% confidence interval
for median, outliers identified as*, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Wigtown, -Z or Z are significant at p<0.05 confidence
level

- 215 -
7.1.6.2 Variation between Burghs
All three burghs have significant enhancement of Sr concentrations in zones within the
burgh core (Harbour and High Street). Elevated Sr levels are confined to the burgh core
in Pittenweem and Wigtown. However, an area with increased Sr concentrations is
identified in the immediate hinterland south of Lauder (Hinterland Near). It should be
noted that this area has a significantly lower median Sr concentration compared to the
High Street zone.

7.1.7 Zinc

Spatial distributions of zinc (Zn) are presented for Lauder (Figure 115), Pittenweem
(Figure 116a) and Wigtown (Figure 116b). Patterns in Zn associated with zones are
subsequently identified and comparisons between burghs are made.

Figure 115: Spatial distribution of Log Zn (mg/Kg) at Lauder. Red boundary delimits
1862AD urban extent

- 216 -
 a) b)

Figure 116: Spatial distribution of Log Zn (mg/Kg) at Pittenweem (a) and Wigtown (b). Red
boundary delimits 1855AD and 1850AD urban extent respectively

7.1.7.1 Variation within Burghs

The median Zn concentration of the High Street, Hinterland Near, Hinterland Far and
Thirlstane zones at Lauder is 267.8, 93.8, 45.32 and 53.2 mg/Kg respectively (Figure
117a). The High Street has significantly higher Zn than the Hinterland Near zone
(p<0.01) and Hinterland Far and Thirlstane zones (p<0.001). In addition the Hinterland
Near zone has significantly higher Zn than the Hinterland Far and Thirlstane zones
(p<0.001) (Figure 117b). There is no significant difference between the Hinterland Far
and Thirlstane zones. These results provide clear evidence for enhancement of Zn
within the burgh core. Elevated Zn levels are also identified in land immediately adjacent
to the historical burgh limits (Hinterland Near), although it should be noted that
enhancement is significantly higher within the burgh core.

- 217 -
 a) Boxplots w ith Sign Confidence Intervals b) Pairw ise Comparisons
Desired Conf idence: 86.761
700

High Street Hint Near

600
Hint Far

500 Thirlstane

400
Zn (mg/Kg)

Hint Near Hint Far

300

Thirlstane
200

100

Hint Far Thirlstane

0
High Street Hint Near Hint Far Thirlstane
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 117: (a) Boxplot of median Zn (mg/Kg) for individual zones at Lauder; boundaries of
boxes represent interquartile range, cyan shading indicates 86.761% confidence interval
for median, outliers identified as*, (b) Kruskal-Wallis analysis with Dunn’s test multiple
comparisons for individual zones at Lauder, -Z or Z are significant at p<0.05 confidence
level

At Pittenweem the median Zn concentration of the Harbour, High Street, Hinterland Near
and Hinterland Far zones is 205.3, 168.7, 49.3 and 41.7 mg/Kg respectively (Figure
118a). The Harbour zone has significantly higher Zn than the Hinterland Near (p<0.05)
and Hinterland Far zones (p<0.001). Similarly the High Street zone has significantly
higher Zn levels than the Hinterland Near and Hinterland Far zones (p<0.001) (Figure
118b). There is no statistical difference in median Zn concentrations between the
Harbour and High Street zones. These results provide clear evidence for Zn
enhancement within the burgh core. In addition, the Hinterland Near zone has
significantly higher Zn than the Hinterland Far zone (p<0.01). Elevated Zn levels are
therefore identified in land adjacent to the historical burgh limits at Pittenweem
(Hinterland Near), although it should be noted that enhancement is significantly higher
within the burgh core.

- 218 -
 a) Boxplots w ith Sign Confidence Intervals
b) Pairw ise Comparisons
Desired Conf idence: 86.761
500

Harbour High Street

Hint Near
400

Hint Far

300
Zn (mg/Kg)

High Street Hint Near

200
Hint Far

100

Hint Near Hint Far

0
Harbour High Street Hint Near Hint Far
-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.033 |Bonf erroni Z-v alue|: 2.128

Figure 118: (a) Boxplot of median Zn (mg/Kg) for individual zones at Pittenweem;
boundaries of boxes represent interquartile range, cyan shading indicates 86.761%
confidence interval for median, outliers identified as*, (b) Kruskal-Wallis analysis with
Dunn’s test multiple comparisons for individual zones at Pittenweem, -Z or Z are
significant at p<0.05 confidence level

The median Zn of the High Street, Hinterland Near and Showfield zones at Wigtown is
418, 52.9 and 84.8 mg/Kg respectively (Figure 119a). The High Street zone has
significantly higher Zn then the Hinterland Near and Hinterland Far zones (p<0.001).
Moreover, the Showfield zone has significantly greater Zn than the Hinterland Near zone
(p<0.01) (Figure 119b). These results indicate enhancement of Zn in Wigtown’s burgh
core and to a lesser extent the immediate hinterland south of Wigtown (Showfield).

- 219 -
 Boxplots w ith Sign Confidence Intervals Pairw ise Comparisons
Desired Conf idence: 80.529

4000

High Street Hint Near

3000
Showf ield
Zn (mg/Kg)

2000

1000

Hint Near Showf ield

High Street Hint Near Showf ield

-Z 0 Z
Family Alpha: 0.2
Bonf erroni Indiv idual Alpha: 0.067 |Bonf erroni Z-v alue|: 1.834

Figure 119: (a) Boxplot of median Zn (mg/Kg) for individual zones at Wigtown; boundaries
of boxes represent interquartile range, cyan shading indicates 80.529% confidence
interval for median, outliers identified as*, (b) Kruskal-Wallis analysis with Dunn’s test
multiple comparisons for individual zones at Wigtown, -Z or Z are significant at p<0.05
confidence level

7.1.7.2 Variation between Burghs

All three burghs have significant enhancement of Zn concentrations in zones within the
burgh core (Harbour and High Street). In addition, elevated Zn concentrations are
identified in land adjacent to the historical burgh limits at Lauder and Pittenweem
(Hinterland Near). Enhanced Zn levels are also identified in the immediate hinterland
south of Wigtown (Showfield). It should be noted that Zn enhancement in the
aforementioned hinterland areas is significantly lower compared to the burgh core within
each town.

- 220 -
7.2 Summary of Results: Elemental Concentrations

The following section presents a brief summary of key trends in selected elemental
concentrations (Ba, Ca, Pb, K, P, Sr, and Zn) both within and between burghs. The
significance of these results is discussed in section 8.1.2 and 8.2.2; specifically the utility
of elemental concentrations as indicators of past waste disposal is considered.

7.2.1 Variation within Burghs

7.2.1.1 Lauder
The High Street, Hinterland Near and Thirlstane zones are identified as important areas
of interest at Lauder. Both the High Street zone and Hinterland Near zone are
characterised by significantly enhanced levels of Ba, Ca, Pb, K, P, Sr and Zn (Table 8).
In addition, the Thirlstane zone has typically elevated concentrations of Ba, Pb and P.
Elemental enhancement is statistically greatest within the High Street zone, followed by
the Hinterland Near zone and Thirlstane zone in succession. These results indicate
distinct elemental enhancement within the burgh core and to a lesser extent in the
immediate hinterland south (Hinterland Near) and north (Thirlstane) of Lauder.

7.2.1.2 Pittenweem
Two zones are identified as important areas of interest at Pittenweem; the Harbour zone
and High Street zone. These zones are characterised by significantly elevated levels of
Ba, Ca, Pb, K, P, Sr and Zn (Table 8). There is no evidence for elemental enhancement
within the hinterland at Pittenweem, with the exception of Zn which is slightly elevated in
the Hinterland Near zone. These results provide clear evidence for elemental
enhancement within the burgh core.

7.2.1.3 Wigtown
The High Street zone and Showfield zone are identified as important areas of interest at
Wigtown. Similar to results obtained for Lauder and Pittenweem, the High Street zone is
characterised by significantly enhanced concentrations of Ba, Ca, Pb, K, P, Sr and Zn. In
addition, the Showfield zone has typically elevated levels of Ba, Pb, P and Zn (Table 13).
Considering elemental enhancement is statistically highest within the High Street zone,
these results indicate distinct elemental enhancement within the burgh core and to a
lesser extent in the immediate hinterland south (Showfield) of Wigtown.

- 221 -
Table 13: Summary of trends in selected elemental concentrations identified at Lauder,
Pittenweem and Wigtown

Element Lauder Pittenweem Wigtown

Ba Enhancement of Ba in Enhancement of Ba in Enhancement of Ba in
burgh core (High Street burgh core (Harbour burgh core (High Street
zone) and immediate and High Street zone) and immediate
hinterland (Hinterland zones). hinterland (Showfield
Near and Thirlstane zone). Enhancement
zones). Enhancement greatest in burgh core.
greatest in burgh core
and Thirlstane zone.
Ca Enhancement of Ca in Enhancement of Ca in Enhancement of Ca in
burgh core (High Street burgh core (Harbour burgh core (High Street
zone) and immediate and High Street zone).
hinterland (Hinterland zones).
Near). Enhancement
greatest in burgh core.
Pb Enhancement of Pb in Enhancement of Pb in Enhancement of Pb in
burgh core (High Street burgh core (Harbour burgh core (High Street
zone) and immediate and High Street zone) and immediate
hinterland (Hinterland zones). hinterland (Showfield
Near and Thirlstane zone). Enhancement
zones). Enhancement greatest in burgh core.
greatest in burgh core.
K Enhancement of K in Enhancement of K in Enhancement of K in
burgh core (High Street burgh core (Harbour burgh core (High Street
zone) and immediate and High Street zone).
hinterland (Hinterland zones).
Near). Enhancement
greatest in burgh core.
P Enhancement of P in Enhancement of P in Enhancement of P in
burgh core (High Street burgh core (Harbour burgh core (High Street
zone) and immediate and High Street zone) and immediate
hinterland (Hinterland zones). hinterland (Showfield
Near and Thirlstane zone). Enhancement
zones). Enhancement greatest in burgh core.
greatest in burgh core.
Sr Enhancement of Sr in Enhancement of Sr in Enhancement of Sr in
burgh core (High Street burgh core (Harbour burgh core (High Street
zone) and immediate and High Street zone).
hinterland (Hinterland zones).
Near). Enhancement
greatest in burgh core.
Zn Enhancement of Zn in Enhancement of Zn in Enhancement of Zn in
burgh core (High Street burgh core (Harbour burgh core (High Street
zone) and immediate and High Street zones) zone) and immediate
hinterland (Hinterland and immediate hinterland (Showfield
Near). Enhancement hinterland (Hinterland zone). Enhancement
greatest in burgh core. Near). Enhancement greatest in burgh core.
greatest in burgh core.

- 222 -
7.2.2 Variation between Burghs

All three burghs have enhanced concentrations of Ba, Ca, Pb, K, P, Sr and Zn in zones
corresponding to the burgh core (Harbour and High street). Elemental enhancement is
limited to the burgh core at Pittenweem however; trends in certain elements are
identified within the hinterland at Lauder and Wigtown, albeit to a lesser magnitude.
Elevated concentrations of Ba, Pb and P are noted in the immediate hinterland south
(Hinterland Near) and north (Thirlstane) of Lauder, and in the immediate hinterland
(Showfield) south of Wigtown. Moreover, enhanced concentrations of Ca, K and Sr are
identified within the Hinterland Near zone at Lauder.

Table 14: Comparison of trends in selected elemental concentrations identified at Lauder,

Pittenweem and Wigtown

Element Comparison between Burghs

Ba All three burghs have enhanced Ba in the burgh core
Lauder and Wigtown have enhanced Ba in the immediate hinterland
(Lauder: Hinterland Near and Thirlstane, Wigtown: Showfield).
Ca All three burghs have enhanced Ca in the burgh core
Lauder has enhanced Ca in the immediate hinterland (Hinterland Near).
Pb All three burghs have enhanced Pb in the burgh core
Lauder and Wigtown have enhanced Pb in the immediate hinterland
(Lauder: Hinterland Near and Thirlstane, Wigtown: Showfield) .
K All three burghs have enhanced K in the burgh core
Lauder has enhanced K in the immediate hinterland (Hinterland Near).
P All three burghs have enhanced P in the burgh core
Lauder and Wigtown have enhanced P in the immediate hinterland (Lauder:
Hinterland Near and Thirlstane, Wigtown: Showfield).
Sr All three burghs have enhanced Sr in the burgh core
Lauder has enhanced Sr in the immediate hinterland (Hinterland Near).
Zn All three burghs have enhanced Zn in the burgh core
Lauder and Pittenweem have enhanced Zn in immediate hinterland
(Hinterland Near), Wigtown also has enhanced Zn in immediate hinterland
(Showfield).

- 223 -
7.3 Multiple Comparisons

This section presents results obtained from a series of multi-element analyses including
multiple comparisons, cluster analysis and discriminant analysis. Multi-element analyses
were conducted for Lauder, Pittenweem and Wigtown using all 21 measured elemental
concentrations.

7.3.1 Lauder

Average elemental concentrations of individual zones within Lauder are presented in

Table 15. There is no significant difference in median Aluminium (Al) and Lithium (Li)
concentrations between the High Street, Hinterland Near, Hinterland Far and Thirlstane
zones. Accordingly these elements were excluded from subsequent multiple
comparisons. The results of multiple comparisons between the High Street, Hinterland
Near, Hinterland Far and Thirlstane zones at Lauder are summarised in Table 16. Zones
with significant enhancement are highlighted ( ) and elements with similar spatial
distributions are grouped accordingly.

- 224 -
Table 15: Average elemental concentrations (mg/Kg) of individual zones at Lauder (median
used for non-parametric data, mean used for parametric data). Test indicates whether
Kruskal-Wallis (KW) or One Way ANOVA (ANOVA) used to compare elemental
concentrations between zones

Element High Hinterland Hinterland Thirlstane Test Test P

Street Near Far Statistic Value
Al 10128 10324 9889 9921 KW H 6.95 0.073
Fe 16953 16324 15890 15149 KW H 9.4 0.024
Mg 3835 2882.5 3027.2 3327.2 KW H 29.21 <0.001
Ca 9076 2959 1844 2458 KW H 66.32 <0.001
Na 103.9 44.51 44.51 44.51 KW H 58.63 <0.001
K 854.3 474.6 205.6 284.8 KW H 56.03 <0.001
Ti 35.37 11.791 11.791 11.791 KW H 82.28 <0.001
P 3129 1798 903.4 1658 KW H 77.47 <0.001
Mn 518.9 789.9 534.4 588.6 ANOVA F 17.36 <0.001
Ba 257.6 191.8 144.9 225.4 ANOVA F 30.75 <0.001
Co 22.4 20.2 20 21.8 KW H 61.23 <0.001
Cr 46.7 25.2 13.3 16.8 KW H 22.51 <0.001
Cu 16.2 14.4 15.6 14.6 KW H 77.84 <0.001
Li 34.4 25.6 20.2 21.2 KW H 6.8 0.078
Ni 2.8 2.4 1.8 2 KW H 60.75 <0.001
Sc 78.4 26.8 12.1 13.4 KW H 57.34 <0.001
Sr 78.4 26.6 12.1 13.4 KW H 79.79 <0.001
V 23.3 20.8 17.8 21.8 KW H 25.89 <0.001
Y 7.7 5.4 4.3 5.6 KW H 65.27 <0.001
Zn 267.8 93.8 45.2 53.2 KW H 86.39 <0.001
Pb 209.9 77.7 33.25 59.1 KW H 72.81 <0.001

- 225 -
Table 16: Multiple comparisons for individual zones at Lauder (Dunn’s test used for non-
parametric data, Tukey-Kramer post-hoc analyses for parametric data)

Hinterland Near

Hinterland Far
High Street

Thirlstane
Element

Summary
Al No significant difference between zones

Li No significant difference between zones

Fe High Street has significantly higher Fe than Hinterland Far (p<0.05) and
Thirlstane (p<0.01)
Na High Street has significantly higher Na than Hinterland Near, Hinterland Far and
Thirlstane (p<0.001)
Mn Hinterland Near has significantly higher Mn than High Street, Hinterland Far and
Thirlstane (P<0.001)
Ca High Street has significantly higher Ca than Hinterland Near, Hinterland Far and
Thirlstane (P<0.001)
Hinterland Near has significantly higher Ca than Hinterland Far (p<0.001)
Co High Street has significantly higher Co than Hinterland Far and Thirlstane
(p<0.001)
Hinterland Near has significantly higher Co than Hinterland Far and Thirlstane
(p<0.01)
Cu High Street has significantly higher Cu than Hinterland Near (p<0.05), Hinterland
Far and Thirlstane (p<0.001)
Hinterland Near has significantly higher Cu than Hinterland Far and Thirlstane
(p<0.001)
K High Street has significantly higher K than Hinterland Near, Hinterland Far and
Thirlstane (p<0.001)
Hinterland Near has significantly higher K than Hinterland Far (p<0.001)
Ni High Street has significantly higher Ni than Hinterland Near (p<0.01), Hinterland
Far and Thirlstane (p<0.001)
Hinterland Near has significantly higher Ni than Hinterland Far and Thirlstane
(p<0.001)
Sr High Street has significantly higher Sr than Hinterland Near (p<0.01), Hinterland
Far and Thirlstane (p<0.001)
Hinterland Near has significantly higher Sr than Hinterland Far and Thirlstane
(p<0.001)
Ti High Street has significantly higher Ti than Hinterland Near, Hinterland Far and
Thirlstane (p<0.001)
Hinterland Near has significantly higher Ti than Hinterland Far (p<0.001)
Zn High Street has significantly higher Zn than Hinterland Near, Hinterland Far and
Thirlstane (p<0.001)
Hinterland Near has significantly higher Zn than Hinterland Far and Thirlstane
(p<0.001)
Mg High Street has significantly higher Mg than Hinterland Near (P<0.001) and
Hinterland Far (p<0.001)
Thirlstane has significantly higher Mg than Hinterland Near (p<0.01) and
Hinterland Far (p<0.05)
Cr High Street has significantly higher Cr than Hinterland Near and Hinterland Far
(p<0.001)
Hinterland Near has significantly higher Cr than Hinterland Far (p<0.001)
Thirlstane has significantly higher Cr than Hinterland Near and Hinterland Far
(p<0.001)
P High Street has significantly higher P than Hinterland Near, Hinterland Far and
Thirlstane (p<0.001)
Hinterland Near has significantly higher P than Hinterland Far (p<0.001)
Thirlstane has significantly higher P than Hinterland Far (p<0.001)
Pb High Street has significantly higher Pb than Hinterland Near (p<0.01), Hinterland
Far and Thirlstane (p<0.001)
Hinterland Near has significantly higher Pb than Hinterland Far (p<0.001) and

- 226 -
 Thirlstane (p<0.05)
Thirlstane has significantly higher Pb than Hinterland Far (p<0.001)
Sc High Street has significantly higher Sc than Hinterland Near (p<0.05), Hinterland
Far and Thirlstane (p<0.001)
Hinterland Near has significantly higher Sc than Hinterland Far (p<0.001) and
Thirlstane (p<0.05)
Thirlstane has significantly higher Sc than Hinterland Far (p<0.05)
Y High Street has significantly higher Y than Hinterland Near, Hinterland Far
(p<0.001) and Thirlstane (p<0.05)
Hinterland Near has significantly higher Y than Hinterland Far (p<0.001)
Thirlstane has significantly higher Y than Hinterland Far (p<0.001)
Ba High Street has significantly higher Ba than Hinterland Near, Hinterland Far and
Thirlstane (p<0.001)
Hinterland Near has significantly higher Ba than Hinterland Far (p<0.001)
Thirlstane has significantly higher Ba than Hinterland Far (p<0.001)
V High Street, Hinterland Near and Thirlstane all have significantly higher V than
Hinterland Far (p<0.001)

Eight groups of elements with distinct patterns of spatial enhancement are identified at
Lauder. The first group contains Al and Li which are relatively homogenous and show no
pattern of spatial enhancement within the sample area. The second group is
characterised by statistically higher levels of Fe and Na within the High Street zone.
Although Na is significantly enhanced compared to the other three zones, there is no
significant difference in Fe concentrations between the High Street and Hinterland Near
zones. The third group identified is dominated by elevated Mn concentrations within the
Hinterland Near zone. None of the other elements share a similar spatial distribution with
Mn. The fourth group contains Mg which is highest within the High Street zone followed
by the Thirlstane zone.

The fifth group comprises Ca, Co, Cu, K, Ni, Sr, Ti and Zn. Elements within this group
are primarily enhanced within the High Street zone and to a lesser extent within the
Hinterland Near zone. It can be argued that Ti does not necessarily belong in this group
considering median values for the Hinterland Near, Hinterland Far and Thirlstane zones
are all 11.791 mg/Kg. However results of multiple comparison analysis indicate a
significant difference between the Hinterland Near zone and Hinterland Far and
Thirlstane zones. This is because statistical analyses use all data within a zone not just
the median value.

The sixth group includes Cr, P, Pb, Sc and Y. These elements are enhanced primarily
within the High Street zone followed in succession by the Hinterland Near and Thirlstane
zones. Similarly the seventh and eighth groups which contain Ba and V respectively are
also enhanced within the High Street, Hinterland Near and Thirlstane zones. However,
Ba is highest within the High Street zone followed by the Thirlstane zone and Hinterland
Near zone, and V is statistically enhanced within all three zones.

- 227 -
7.3.2 Pittenweem

Average elemental concentrations of individual zones within Pittenweem are presented

in Table 17. There is no significant difference in median Li and V concentrations
between the Harbour, High Street, Hinterland Near and Hinterland Far zones.
Accordingly these elements were excluded from subsequent multiple comparisons. The
results of multiple comparisons between the Harbour, High Street, Hinterland Near and
Hinterland Far zones at Pittenweem are summarised in Table 18. Zones with significant
elemental enhancement are highlighted ( ) and elements with similar spatial
distributions are grouped accordingly.

Table 17: Average elemental concentrations (mg/Kg) of individual zones at Pittenweem

(median used for non-parametric data, mean used for parametric data). Test indicates
whether Kruskal-Wallis (KW) or One Way ANOVA (ANOVA) used to compare elemental
concentrations between zones

Element Harbour High Hinterland Hinterland Test Test P

Street Near Far Statistic Value
Al 5024 5659 6734 6544 KW H 18.76 <0.001
Fe 17268 14981 16016 15331 KW H 9.61 0.022
Mg 2708 2394 1887 1700.5 KW H 33.65 <0.001
Ca 17417 11756 2580 2608.6 KW H 53.07 <0.001
Na 267.1 244.8 89 59.35 KW H 60.45 <0.001
K 363.9 348.1 79.1 126.57 KW H 34.29 <0.001
Ti 64.9 82.5 70.75 47.17 KW H 37.44 <0.001
P 1741 2147 964.5 960.1 KW H 43.84 <0.001
Mn 441.4 395 480.2 294.3 KW H 58.47 <0.001
Ba 153.4 147.7 62.5 61.6 KW H 57.11 <0.001
Co 6.1 5.3 4.6 4.3 KW H 39.76 <0.001
Cr 16.1 16 15 16.6 KW H 15.48 0.001
Cu 44.2 44.1 23.5 23.4 KW H 44.33 <0.001
Li 6 5.6 6.4 6.4 KW H 1.29 0.732
Ni 27.5 24.7 19.6 18.8 KW H 37.34 <0.001
Sc 2.3 2.8 2.6 2.6 KW H 15.84 0.001
Sr 102.9 91.5 28.5 31.5 KW H 51.2 <0.001
V 23.6 24.4 27.6 25.6 KW H 6.76 0.080
Y 7.5 7.8 5.8 5.8 KW H 38 <0.001
Zn 205.3 168.7 49.3 41.7 KW H 66.29 <0.001
Pb 221 182 43.3 34.866 ANOVA F 14.85 <0.001

- 228 -
Table 18: Multiple comparisons for individual zones at Pittenweem (Dunn’s test used for
non-parametric data, Tukey-Kramer post-hoc analyses for parametric data)

Hinterland Near

Hinterland Far
High Street
Element

Harbour

Summary
Li No significant difference between zones

V No significant difference between zones

Mn Hinterland Near has significantly higher Mn than High Street and Hinterland Far
(p<0.001)
Harbour and High Street have significantly higher Mn than Hinterland Far
(p<0.001)
Ba Harbour and High Street have significantly higher Ba than Hinterland Near and
Hinterland Far (p<0.001)
Ca Harbour and High Street have significantly higher Ca than Hinterland Near and
Hinterland Far (p<0.001)
Cu Harbour and High Street have significantly higher Cu than Hinterland Near and
Hinterland Far (p<0.001)
K Harbour has significantly higher K than Hinterland Near (p<0.01) and Hinterland
Far (p<0.001)
High Street has significantly higher K than Hinterland Near and Hinterland Far
(p<0.001)
Ni Harbour and High Street have significantly higher Ni than Hinterland Near and
Hinterland Far (p<0.001)
P Harbour has significantly higher P than Hinterland Near (p<0.05) and Hinterland
Far (p<0.01)
High Street has significantly higher P than Hinterland Near and Hinterland Far
(p<0.001)
Pb Harbour and High Street have significantly higher Pb than Hinterland Near and
Hinterland Far (p<0.001)
Sr Harbour and High Street have significantly higher Sr than Hinterland Near and
Hinterland Far (p<0.001)
Y Harbour and High Street have significantly higher Y than Hinterland Near and
Hinterland Far (p<0.001)

Co Harbour has significantly greater Co than Hinterland Near (p<0.01) and

Hinterland Far (p<0.001)
High Street has significantly greater Co than Hinterland Far (p<0.001)
Hinterland Near has significantly greater Co than Hinterland Far (p<0.01)
Mg Harbour has significantly higher Mg than Hinterland Far (p<0.001)
High Street has significantly higher Mg than Hinterland Near (p<0.05) and
Hinterland Far (p<0.001)
Hinterland Near has significantly high Mg than Hinterland Far (p<0.001)
Na Harbour has significantly higher Na than Hinterland Near and Hinterland Far
(p<0.001)
High Street has significantly higher Na than Hinterland Near and Hinterland Far
(p<0.001)
Hinterland Near has significantly higher Na than Hinterland Far (p<0.001)
Ti High Street has significantly higher Ti than Hinterland Far (p<0.001)
Hinterland Near has significantly higher Ti than Hinterland far (p<0.001)
Zn Harbour has significantly higher Zn than Hinterland Near (p<0.01) and Hinterland
Far (p<0.001)
High Street has significantly higher Zn than Hinterland Near and Hinterland Far
(p<0.001)
Hinterland Near has significantly higher Zn than Hinterland Far (p<0.01)
Cr Harbour and High Street have significantly higher Cr than Hinterland Near
(p<0.05)
Hinterland Far has significantly higher Cr than Hinterland Near (p<0.001)
Sc Harbour has significantly higher Sc than Hinterland Near and Hinterland Far
(p<0.001)
High Street has significantly higher Sc than Hinterland Far (p<0.05)

- 229 -
Al Hinterland Near has significantly higher Al than Harbour (p<0.05) and High
Street (p<0.001)
Hinterland Far has significantly higher Al than Harbour (p<0.001) and High
Street (p<0.05)
Fe Harbour has significantly higher Fe than High Street (p<0.05)
Hinterland Near has significantly higher Fe than High Street (p<0.05)

Seven groups of elements with distinct patterns of spatial enhancement are identified at
Pittenweem. The first group contains Li and V which are relatively homogenous and
show no pattern of spatial enhancement within the sample area. The second group is
characterised by significantly higher levels of Mn within the Hinterland Near zone and to
a lesser extent in the Harbour and High Street zones. None of the other elements share
a similar spatial distribution with Mn.

The third group comprises Ba, Ca, Cu, K, Ni, P, Pb, Sr and Y. Elements within this group
are significantly enhanced within the Harbour and High Street zones. There is no
statistical difference in enhancement between these two zones.

The fourth group includes Co, Mg, Na, Ti and Zn. Typically these elements are primarily
enhanced within the Harbour and High Street zones and to a lesser extent are elevated
within the Hinterland Near zone. It could be argued that Ti should not be included within
this group considering the Harbour is not statistically enhanced comparative to the
Hinterland Near or Hinterland Far zones. However multiple comparisons indicate that
there is no significant difference in Ti concentrations between the Harbour and High
Street zone (Table 18).

The fifth group contains Cr and Sc. Statistical analysis indicates that Cr is enhanced
within the Harbour, High Street and Hinterland Far zones and Sc is elevated in the
Harbour and High Street zones. Nevertheless, median values of these two elements are
relatively comparable between zones (Table 17). It is therefore suggested that these two
elements do not show a distinct pattern of spatial enhancement within the survey area.
The sixth and seventh groups contain Al and Fe respectively, of which Al is enhanced
within the Hinterland Near and Hinterland far zones and Fe is elevated in the Harbour
and Hinterland Near zones.

- 230 -
7.3.3 Wigtown

Average elemental concentrations of individual zones within Wigtown are presented in

Table 19. There is no significant difference in median Al between the High Street,
Hinterland Near and Showfield zones. Accordingly Al was excluded from subsequent
multiple comparisons. The results of multiple comparisons between the High Street,
Hinterland Near and Showfield zones at Wigtown are summarised in Table 20. Zones
with significant elemental enhancement are highlighted ( ) and elements with similar
spatial distributions are grouped accordingly.

Table 19: Average elemental concentrations (mg/Kg) for individual zones at Wigtown
(median used for non-parametric data, mean used for parametric data). Test indicates
whether Kruskal-Wallis (KW) or One Way ANOVA (ANOVA) used to compare elemental
concentrations between zones

Element High Hinterland Showfield Test Test P

Street Near Statistic Value
Al 12928 12886 12870 KW H 1.84 0.399
Fe 22493 21479 22884 KW H 9.9 0.01
Mg 5535.8 4903 4529 KW H 36.35 <0.001
Ca 17038 1522 1143 KW H 65.68 <0.001
Na 222.6 59.35 66.77 KW H 65.82 <0.001
K 870.2 284.8 300.6 KW H 56.59 <0.001
Ti 165.08 53.06 47.17 KW H 61.15 <0.001
P 5001 1492.5 2553 KW H 79.29 <0.001
Mn 967.4 673.8 867.4 KW H 36.78 <0.001
Ba 236.6 48.9 94.9 KW H 74 <0.001
Co 8.4 4.6 4.6 KW H 58.49 <0.001
Cr 32.2 34.6 27 KW H 16.82 <0.001
Cu 78.4 18.2 32.5 KW H 73.20 <0.001
Li 7.4 7 5.3 KW H 14.93 <0.001
Ni 42.2 24.8 22.2 KW H 59.31 <0.001
Sc 3.2 1.8 2 KW H 68.05 <0.001
Sr 153 11.5 13.6 KW H 66.02 <0.001
V 37 30.8 22.1 KW H 28.34 <0.001
Y 10.6 4.2 5 KW H 66.08 <0.001
Zn 418 52.9 84.8 KW H 74.57 <0.001
Pb 373.9 63.18 151.31 KW H 76.51 <0.001

- 231 -
Table 20: Multiple comparisons for individual zones at Wigtown (Dunn’s test used for non-
parametric data, Tukey-Kramer post-hoc analyses for parametric data)

Hinterland Near
High Street

Showfield
Element

Summary
Al No significant difference between zones

Fe High Street has significantly higher Fe than Hinterland Near (p<0.01)

Showfield has significantly higher Fe than Hinterland Near (p<0.05)

Ca High Street has significantly higher Ca than Hinterland Near and Showfield (p<0.001)

Co High Street has significantly higher Co than Hinterland Near and Showfield (p<0.001)

K High Street has significantly higher K than Hinterland Near and Showfield (p<0.001)

Mg High Street has significantly higher Mg than Hinterland Near and Showfield (p<0.001)

Na High Street has significantly higher Na than Hinterland Near and Showfield (p<0.001)

Ni High Street has significantly higher Ni than Hinterland Near and Showfield (p<0.001)

Sr High Street has significantly higher Sr than Hinterland Near and Showfield (p<0.001)

Ti High Street has significantly higher Ti than Hinterland Near and Showfield (p<0.001)

Cr High Street has significantly higher Cr than Showfield (p<0.01)

Hinterland Near has significantly higher Cr than Showfield (P<0.001)
Li High Street has significantly higher Li than Showfield (p<0.001)
Hinterland Near has significantly higher Li than Showfield (P<0.01)
V High Street has significantly higher V than Hinterland Near (p<0.01) and Showfield
(p<0.001)
Hinterland Near has significantly higher V than Showfield (P<0.01)
Ba High Street has significantly higher P than Hinterland Near and Showfield (p<0.001)
Showfield has significantly P than Hinterland Near (p<0.001)
Cu High Street has significantly higher Cu than Hinterland Near and Showfield (p<0.001)
Showfield has significantly higher Cu than Showfield (p<0.001)
Mn High Street has significantly higher Mn than Hinterland Near (p<0.001)
Showfield has significantly higher Mn than Hinterland Near (p<0.001)
P High Street has significantly higher P than Hinterland Near (p<0.001) and Showfield
(p<0.01)
Showfield has significantly P than Hinterland Near (p<0.001)
Pb High Street has significantly higher Pb than Hinterland Near (p<0.01) and Showfield
(p<0.001)
Showfield has significantly higher Pb than Hinterland Near (P<0.001)
Sc High Street has significantly higher Sc than Hinterland Near and Showfield (p<0.001)
Showfield has significantly higher Sc than Hinterland Near (p<0.05)
Y High Street has significantly higher Y than Hinterland Near and Showfield (p<0.001)
Showfield has significantly higher Y than Hinterland Near (p<0.05)
Zn High Street has significantly higher Zn than Hinterland Near and Showfield (p<0.001)
Showfield has significantly higher Zn than Hinterland Near (p<0.01)

- 232 -
Four groups of elements with distinct patterns of spatial enhancement are identified at
Wigtown. The first group contains Al and Fe. Al concentrations are relatively
homogenous and show no pattern of spatial enhancement within the sample area. Fe is
statistically enhanced within the High Street and Showfield zones; however, median
values are relatively comparable between zones (Table 19). Accordingly it is suggested
that Fe concentrations do not exhibit a distinct spatial distribution at Wigtown. The
second group of elements comprises Ca, Co, K Mg, Na, Ni, Sr and Ti. These elements
are characteristically enhanced within the High Street zone.

The third group of elements contains Cr, Li and V. Elements within this group are
statistically enhanced within the High Street zone and Hinterland Near zone. There is no
significant difference in Cr and Li between the High Street and Hinterland Near zones,
though V is elevated to a lesser extent within the Hinterland Near zone. The fourth group
of elements includes Ba, Cu, Mn, P, Pb, Sc, Y and Zn. These elements are primarily
enhanced within the High Street zone and to a lesser extent are elevated within the
Showfield zone.

- 233 -
7.4 Cluster Analysis

Cluster analysis was undertaken on all elemental data for Lauder, Pittenweem and
Wigtown to simplify data sets and group together elements with similar spatial
distributions.

7.4.1 Lauder

Seven elemental clusters with statistically distinct patterns of spatial enhancement are
identified at Lauder (Table 21). The first cluster includes Al, Fe, Mg, Cr and Li (Figure
120). Although multiple comparisons indicate a significant difference in Fe, Mg and Cr
between particular zones (section 7.3.1) it is argued that variation in median values
between zones is relatively limited compared to other elements, for example Ca and P
(Table 15). Moreover, considering Mg and Cr are 89.02% similar to Al, Fe and Li it
suggested that these elements are characteristically heterogeneous with no real pattern
of spatial enhancement at Lauder

Cluster 2 consists of Ca, Na, K, Ti, P, Co, Cu, Ni, Sc, Sr and Zn. These elements are
significantly enhanced within the High Street zone and to a lesser extent are elevated
within the Hinterland Near zone. This cluster is directly comparable with group four
identified in section 6.3.1. Cluster 2 includes all elements present within group four and
additionally incorporates Na, P and Sc. Na has similar characteristics to Ca, Ti, Sr and
Zn (89% similarity), P is similar to K (88%) and Sc is similar to Co and Ni (94%). P and
Sc are primarily enhanced within the High Street zone and are slightly elevated within
the Hinterland Near zone. Conversely elevated Na levels are confined to the High Street
zone. Cluster 3 comprises Mn which is significantly enhanced within the Hinterland Near
zone. This cluster is comparable to group 3 identified in section 7.3.1.

Clusters 4 and 5 which contain Ba and V respectively are analogous to groups seven
and eight. Ba is enhanced within the High Street, Thirlstane and Hinterland Near zones
and is most similar to the cluster 2 element K (83% similarity). There is no significant
difference in V between the High Street, Hinterland Near and Thirlstane zones, hence is
80.69% similar to cluster 1 elements which are not spatially enhanced. However, it is
suggested that V was clustered independently on account of the High Street, Hinterland
Near and Thirlstane zones being significantly higher than the Hinterland Far zone.
Clusters 6 and 7 which contain Y and Pb respectively are primarily elevated within the
High Street zone and to a lesser extent the in Thirlstane and Hinterland Near zones. Y is
80% similar to Ba and Pb is 75% similar to Mn.

- 234 -
Table 21 Summary of Cluster analysis amalgamation steps for elemental data at Lauder

Step No. of Similarity Distance Clusters New No.

Clusters Level Level Joined Cluster Observations
1 20 96.4 0.071 20 17 2
2 19 95.2 0.095 15 11 2
3 18 94.9 0.100 17 7 3
4 17 94.7 0.105 16 11 3
5 16 91.7 0.165 14 2 2
6 15 91.3 0.172 7 4 4
7 14 91.1 0.176 2 1 3
8 13 89.8 0.202 5 4 5
9 12 89.1 0.217 11 4 8
10 11 89.9 0.200 13 4 9
11 10 91.8 0.163 8 4 10
12 9 88.3 0.233 6 4 11
13 8 88.3 0.233 12 3 2
14 7 89.0 0.219 3 1 5
15 6 80.6 0.386 18 1 6
16 5 80.6 0.386 19 10 2
17 4 83.1 0.337 10 4 13
18 3 84.5 0.308 4 1 19
19 2 75.2 0.494 21 1 20
20 1 66.9 0.661 9 1 21

66.94
Similarity (%)

77.96

88.98

100.00
Al Fe Li Mg Cr V Ca Ti Sr Zn Na Co Ni Sc Cu P K Ba Y Pb Mn

Element
Figure 120: Dendrogram showing results of Cluster analysis for elemental data at Lauder.
Seven main groups are identified; cluster 1 (red), cluster 2 (green), cluster 3 (blue), cluster
4 (orange), cluster 5 (magenta), cluster 6 (purple) and cluster 7 (cyan)

- 235 -
7.4.2 Pittenweem

Five elemental clusters with statistically distinct patterns of spatial enhancement are
identified at Lauder (Table 22). The first cluster includes Al, V, Cr and Sc (Figure 121).
Although multiple comparisons indicate a significant difference in Al, Cr and Sc between
particular zones (section 7.3.2) it is argued that variation in median values between
zones is relatively limited compared to other elements, for example Ba and Ca (Table
17). Moreover, considering Cr and Sc are 83 % similar to V and Al it is suggested that
these elements are characteristically heterogeneous with no real pattern of spatial
enhancement at Pittenweem.

Cluster 2 comprises Fe which is statistically elevated within the Harbour and Hinterland
Near zones. This cluster is comparable to group 7 identified in section 6.3.2.

Cluster 3 contains Mg, Na, Ca, Sr, P, Cu, Ba, Zn, Co, Ni, Y, K, Pb and Ti. These
elements are significantly enhanced within the Harbour and High Street zones. Moreover
Co, Mg, Na, Ti and Zn are elevated to a lesser extent within the Hinterland Near zone.
This cluster is directly comparable with groups three and four identified in section 7.3.2.
The strongest similarities between elements within this group exist between Co and Ni
(97%), Mg and Na (96%), Ca and Sr (94%), P and Cu (94%) and Ba and Zn (93%).

Cluster 4 comprises Mn which is significantly enhanced within the Hinterland Near zone.
This cluster is comparable to group 2 identified in section 7.3.2. Cluster 5 contains Li, of
which there is no distinct spatial enhancement pattern over the sample area. Li is 80 %
similar to Mn and 80 % to Fe.

- 236 -
Table 22: Summary of Cluster analysis amalgamation steps for elemental data at
Pittenweem

Step No. of Similarity Distance Clusters New No.

Clusters Level Level Joined Cluster Observations
1 20 97.3 0.052 15 11 2
2 19 96.2 0.075 5 3 2
3 18 94.2 0.114 17 4 2
4 17 94.0 0.118 13 8 2
5 16 93.5 0.129 19 11 3
6 15 93.3 0.132 20 10 2
7 14 93.6 0.126 10 8 4
8 13 90.8 0.183 11 8 7
9 12 88.1 0.236 8 4 9
10 11 87.3 0.252 6 4 10
11 10 84.5 0.308 18 1 2
12 9 82.6 0.347 21 4 11
13 8 81.1 0.376 4 3 13
14 7 85.2 0.295 7 3 14
15 6 80.0 0.398 12 1 3
16 5 83.4 0.330 16 1 4
17 4 79.3 0.412 3 1 18
18 3 80.0 0.399 2 1 19
19 2 80.6 0.386 9 1 20
20 1 65.8 0.682 14 1 21

65.87
Similarity (%)

77.25

88.62

100.00
Al V Cr Sc Mg Na Ca Sr P Cu Ba Zn Co Ni Y K Pb Ti Fe Mn Li
Element

Figure 121: Dendrogram showing results of Cluster analysis for elemental data at
Pittenweem. Five main groups are identified; cluster 1 (red), cluster 2 (yellow), cluster 3
(green), cluster 4 (blue) and cluster 5 (brown)

- 237 -
7.4.3 Wigtown

Six elemental clusters with statistically distinct patterns of spatial enhancement are
identified at Lauder (Table 23). The first cluster is directly comparable to group 1
identified in section 7.3.3 in that it contains Al and Fe (Figure 122). Although multiple
comparisons indicate a significant difference in Fe between particular zones (section
7.3.3) it is argued that variation in median values between zones is relatively limited
compared to other elements, for example Ba, Ca and P (Table 19). Moreover,
considering Fe is 81% similar to Al, it is suggested that these elements are
characteristically heterogeneous with no distinct pattern of spatial enhancement at
Wigtown.

The second cluster includes Mg, Ca, Na, K, Ti, P, Ba, Co, Cu, Ni, Sc, Sr, Y, Zn and Pb.
These elements are all significantly enhanced within the High Street zone. In addition
Ba, Cu, P, Pb, Sc, Y and Zn are elevated within the Showfield zone, albeit to a
significantly lesser extent. This cluster is directly comparable with groups 2 and 4
identified in section 7.3.3, with the exception of Mn. The strongest similarities between
elements within this cluster are Ca and Sr (98%), Sc and Y (98%), Co and Ni (96%) and
Zn and Pb (99%).

Cluster 3 contains Mn which is significantly enhanced within the High Street and
Showfield zones. Clusters 4, 5 and 6 comprise Cr, Li and V respectively. These
elements are statistically elevated within the High Street and Hinterland Near zones.
However, variation in median values is minimal especially when compared to elements
in cluster 2. Moreover, considering Cr and V are 76% and 77% similar to cluster 1
elements Al and Fe it is argued that clusters 4, 5 and 6 do not exhibit a distinct spatial
distribution at Wigtown.

- 238 -
Table 23: Summary of Cluster analysis amalgamation steps for elemental data at Wigtown

Step No. of Similarity Distance Clusters New No.

Clusters Level Level Joined Cluster Observations
1 20 99.0 0.019 21 20 2
2 19 98.3 0.033 19 16 2
3 18 98.0 0.038 17 4 2
4 17 97.0 0.058 16 7 3
5 16 96.7 0.065 15 11 2
6 15 96.5 0.069 5 4 3
7 14 95.9 0.080 7 4 6
8 13 95.8 0.083 10 4 7
9 12 94.8 0.103 8 4 8
10 11 92.0 0.158 6 4 9
11 10 91.0 0.179 13 11 3
12 9 88.7 0.225 11 4 12
13 8 81.6 0.367 20 4 14
14 7 83.0 0.339 4 3 15
15 6 81.1 0.376 2 1 2
16 5 76.6 0.467 12 1 3
17 4 77.2 0.454 18 1 4
18 3 77.8 0.443 3 1 19
19 2 79.2 0.415 14 1 20
20 1 68.1 0.636 9 1 21

68.16
Similarity (%)

78.78

89.39

100.00
Al Fe Cr V Mg Ca Sr Na Ti Sc Y Ba P K Co Ni Cu Zn Pb Li Mn
Element

Figure 122: Dendrogram showing results of Cluster analysis for elemental data at
Wigtown. Six main groups are identified; cluster 1 (red), cluster 2 (green), cluster 3 (blue),
cluster 4 (coral), cluster 5 (brown) and cluster 5 (magenta)

- 239 -
7.5 Discriminant Analysis

Discriminant analysis was undertaken on all elemental data for Lauder, Pittenweem and
Wigtown to investigate whether predicted classifications of sample points are similar to
those observed. It is expected that zones with spatially distinct patterns of elemental
enhancement will have a high percentage of observations correctly classified, for
example 22 sample points (observations) are located within the High Street zone at
Lauder (true zone), of which 21 (95.5%) were successfully predicted as belonging to the
High Street zone (predicted zone).

7.5.1 Lauder

The results of discriminant analysis for zones within Lauder are presented in Table 24.
The discriminant analysis identified the correct zone for 111 of 116 observations within
the survey area (95.7%). The Hinterland Far zone has the highest probability of correct
classification (97.2%) followed in succession by the Hinterland Near (96.8%), High
Street (95.5%) and Thirlstane (92.6%) zones. The high accuracy of classification at
Lauder suggests that zones can be successfully distinguished according to differences
in their elemental concentrations.

Table 24: Summary of discriminant classification for individual zones within Lauder; ‘true
zone’ indicates the location of observations (sample points), ‘predicted zone’ indicates the
predicted location of observations, ‘No. observations’ indicates the total number of
observations per zone and ‘No. observations correct’ signifies the number of observations
correctly classified per zone

True Zone
Predicted Zone High Street Hint Near Hint Far Thirlstane

High Street 21 0 0 0
Hint Near 0 30 1 1
Hint Far 1 1 35 1
Thirlstane 0 0 0 25
No. Observations 22 31 36 27
No. Observations Correct 21 30 35 25
% Observations Correct 95.5 96.8 97.2 92.6

- 240 -
7.5.2 Pittenweem

The results of discriminant analysis for zones within Pittenweem are presented in Table
25. The discriminant analysis identified the correct zone for 91 of 96 observations within
the survey area (94.8%). Similar to results obtained for Lauder, the Hinterland Far zone
has the highest probability of correct classification (100%) followed by the Hinterland
Near (78.6%) and High Street zones (70.8%). The high classification accuracy of
hinterland zones indicates that these zones have differing elemental characteristics.
Classification of the Harbour (50%) and High Street (70.8%) zones is less successful.
75% of misclassified observations within the harbour were assigned to the High Street
zone and 57% of misclassified observations within the High Street were assigned to the
Harbour Zone. This indicates that there are similarities in elemental concentrations
present within Harbour and High Street zones.

Table 25: Summary of classification for individual zones within Pittenweem; ‘true zone’
indicates the location of observations (sample points), ‘predicted zone’ indicates the
predicted location of observations, ‘No. observations’ indicates the total number of
observations per zone and ‘No. observations correct’ signifies the number of observations
correctly classified per zone

True Zone
Predicted Zone Harbour High Street Hint Near Hint Far

Harbour 4 4 0 0
High Street 3 17 2 0
Hint Near 1 3 22 0
Hint Far 0 0 4 36
No. Observations 8 24 28 36
No. Observations Correct 4 17 22 36
% Observations Correct 50 70.8 78.6 100

7.5.3 Wigtown

The results of discriminant analysis for zones within Wigtown are presented in Table 26.
The discriminant analysis identified the correct zone for 95 of 99 observations within the
survey area (96%). The High Street zone has the highest probability of correct
classification (97%) followed in succession by the Showfield (88.9%) and Hinterland
Near (85.4%) zones. The high accuracy of classification at Wigtown suggests that zones
can be successfully distinguished according to differences in their elemental
concentrations.

- 241 -
Table 26: Summary of classification for individual zones within Wigtown; ‘true zone’
indicates the location of observations (sample points), ‘predicted zone’ indicates the
predicted location of observations, ‘No. observations’ indicates the total number of
observations per zone and ‘No. observations correct’ signifies the number of observations
correctly classified per zone

True Zone
Predicted Zone High Street Hint Near Showfield

High Street 32 0 0
Hint Near 0 41 2
Showfield 1 7 16
No. Observations 33 48 18
No. Observations Correct 32 41 16
% Observations Correct 97 85.4 88.9

7.5.4 Discrimination between Burghs

7.5.4.1 High Street Zone

The results of discriminant analysis between High Street zones at Lauder, Pittenweem
and Wigtown are presented in Table 27. Observations within each High Street zone
were assigned to the correct burgh (100% accuracy) indicating that although similarities
in elemental enhancement patterns are identified between burghs, there are distinctive
differences in elemental concentrations between High Street zones.

Table 27: Summary of classification for the High Street zone at Lauder, Pittenweem and
Wigtown; ‘true burgh’ indicates the location of observations (sample points), ‘predicted
burgh’ indicates the predicted location of observations, ‘No. observations’ indicates the
total number of observations within each High Street zone and ‘No. observations correct’
signifies the number of observations correctly classified per burgh

True Burgh
Predicted Burgh Lauder Pittenweem Wigtown

Lauder 22 0 0
Pittenweem 0 24 0
Wigtown 0 0 33
No. Observations 22 24 33
No. Observations Correct 22 24 33
% Observations Correct 100 100 100

- 242 -
7.5.4.2 Hinterland Near Zone
The results of discriminant analysis between Hinterland Near zones at Lauder,
Pittenweem and Wigtown are presented in Table 28. The Hinterland Near zone at
Pittenweem has the highest percentage classification accuracy (100%) followed in
succession by Lauder (96.8%) and Wigtown (95.8%). The high accuracy of classification
suggests that there are distinctive differences in elemental concentrations between
Hinterland Near zones.

Table 28: Summary of classification for Hinterland Near zones at Lauder, Pittenweem and
Wigtown; ‘true burgh’ indicates the location of observations (sample points), ‘predicted
burgh’ indicates the predicted location of observations, ‘No. observations’ indicates the
total number of observations within each Hinterland Near zone and ‘No. observations
correct’ signifies the number of observations correctly classified per burgh

True Burgh
Predicted Burgh Lauder Pittenweem Wigtown

Lauder 30 0 1
Pittenweem 1 28 1
Wigtown 0 0 46
No. Observations 31 28 48
No. Observations Correct 30 28 46
% Observations Correct 96.8 100 95.8

- 243 -
7.6 Summary of Results: Multi-Element Analyses

The following section summarises the principal results of multi-elements analyses. The
significance of elemental associations within and between burghs is discussed in section
8.1.2 and 8.2.2.

7.6.1 Variation within Burghs

7.6.1.1 Lauder
The results of multi-element analyses for zones within Lauder are summarised in Table
29. Both the High Street and Hinterland Near zones are characterised by significantly
enhanced levels of Ca, K, Ti, Co, Cu, Ni, Sr, Zn, Ba, P, Pb, Sc and Y. In addition, the
Thirlstane zone has typically elevated concentrations of Ba, P, Pb, Sc and Y. Elemental
enhancement is statistically greatest within the High Street zone followed by the
Hinterland Near zone and Thirlstane zone in succession. There is no significant
enhancement within the Hinterland Far zone. These results indicate pronounced
elemental enhancement within the burgh core at Lauder. Moreover, similarity in the
range of elements enhanced in the High Street zone and Hinterland Near zone suggest
comparable enhancement patterns between the burgh core and immediate hinterland
south (Hinterland Near) of the historical burgh limits. There are two exceptions to this
generalisation, Na which is limited to the burgh core and Mn which is restricted to the
immediate hinterland south of Lauder (Hinterland Near). The immediate hinterland north
(Thirlstane) of Lauder is typically enhanced in a smaller number of elements and to a
lesser extent than the High Street and Hinterland Near zones.

Table 29: Summary of elemental enhancement within zones at Lauder; pink indicates
enhancement within the High Street zone only, green indicates enhancement within
Hinterland Near zone only, orange indicates enhancement within High Street and
Hinterland Near zones, and blue indicates enhancement within High Street, Hinterland
Near and Thirlstane zones

Elements
High Street Na Ca K Ti Co Cu Ni Sr Zn Ba P Pb Sc Y
Hinterland Near Mn Ca K Ti Co Cu Ni Sr Zn Ba P Pb Sc Y
Hinterland Far
Thirlstane Ba P Pb Sc Y

- 244 -
7.6.1.2 Pittenweem
The results of multi-element analyses for zones within Pittenweem are summarised in
Table 30. The Harbour and High Street zones are characterised by significantly
enhanced concentrations of Ba, Ca, Cu, K, Ni, O, Pb, Sr, Y, Co, Mg, Na, Ti and Zn.
There is no difference in the range of elemental concentrations or degree of
enhancement between these two zones. Enhancement of Mn is restricted to the
Hinterland Near zone. Moreover, the Hinterland Near zone has elevated concentrations
of Co, Mg, Na, Ti and Zn. Elemental enhancement is statistically greatest within the
Harbour and High Street zones followed by the Hinterland Near zone. There is no
significant enhancement within the Hinterland Far zone. These results indicate
pronounced elemental enhancement within the burgh core at Pittenweem. In addition,
the immediate hinterland north (Hinterland Near) of the historical burgh limits is typically
enhanced in a smaller range of elements and to a lesser extent that the Harbour and
High Street zones.

Table 30: Summary of elemental enhancement within zones at Pittenweem; green

indicates enhancement within the Hinterland Near zone only, orange indicates
enhancement within the Harbour and High Street zones and blue indicates enhancement
within the Harbour, High Street and Hinterland Near zones

Elements
Harbour Ba Ca Cu K Ni P Pb Sr Y Co Mg Na Ti Zn
High Street Ba Ca Cu K Ni P Pb Sr Y Co Mg Na Ti Zn
Hinterland Near Mn Co Mg Na Ti Zn
Hinterland Far

7.6.1.3 Wigtown
The results of multi-element analyses for zones within Wigtown are summarised in Table
31. The High Street zone is characterised by significantly enhanced concentrations of
Ca, Co, K, Mg, Na, Ni, Sr and Ti. These elements are restricted to the High Street zone;
however, Ba, Cu, Mn, P, Pb, Sc, Y and Zn are elevated within the High Street and
Showfield zones. Elemental enhancement is statistically greatest within the High Street
zone followed by the Showfield zone. There is no enhancement within the Hinterland
Near zone. These results indicate pronounced elemental enhancement within the burgh
core at Wigtown. In addition, the immediate hinterland south (Showfield) of the historical
burg limits is enhanced in a smaller range of elements and to a lesser extent that the
High Street zone.

- 245 -
Table 31: Summary of elemental enhancement within zones at Wigtown; pink indicates
enhancement within the High street zone only and orange indicates enhancement within
the High Street and Showfield zones

Elements
High Street Ca Co K Mg Na Ni Sr Ti Ba Cu Mn P Pb Sc Y Zn
Hinterland Near
Showfield Ba Cu Mn P Pb Sc Y Zn

7.6.2 Variation between Burghs

All three burghs have enhanced concentrations of Ba, Ca, Na, K, Ti, Co, Cu, Ni, P, Pb,
Sr Y and Zn within the High Street zone. This finding indicates comparable patterns of
elemental enhancement between burgh cores. In addition, there are no discernable
patterns in Al, Fe Cr, Li and V at Lauder, Pittenweem and Wigtown. Elemental
enhancement is most pronounced within burgh cores; however enhancement is noted
within certain hinterland zones, albeit to a lesser extent. Zones of enhancement are
identified in the immediate hinterland south (Hinterland Near) and north (Thirlstane) of
Lauder, in the immediate hinterland north of Pittenweem (Hinterland Near) and in the
immediate hinterland south of Wigtown (Showfield). The range of elemental
concentrations and magnitude of enhancement within these zones is characteristically
diverse. It should also be noted that all three burghs have one zone which is not
enhanced; namely the Hinterland Far zone at Lauder and Pittenweem, and the
Hinterland Near zone at Wigtown.

- 246 -
8 The Impact of Waste Disposal on Soils in and Around
Lauder, Pittenweem and Wigtown

This chapter examines the impact of waste disposal on urban anthrosols within and
adjacent to Lauder, Pittenweem and Wigtown. The following discussion is centred on
two contrasting areas of importance, the burgh core and its hinterland. The historical
legacy of waste disposal within these two areas is considered and comparisons between
burghs are made.

8.1 The Impact of Waste Disposal on Burgh Cores

8.1.1 Soil Modification

Deepened topsoils at Lauder, Pittenweem and Wigtown provide clear evidence for
sustained application of waste material within burgh cores. Accumulation of topsoils can
be explained through addition of waste rich in mineral material for example, turf used in
building construction, ash from hearths, sand used in byres and sand associated with
seaweed. Topsoils within burgh cores are characteristically neutral. It is suggested that
deposition of calcium rich materials such as mortar, plaster and lime washes used in
building construction, and shell and bone associated with kitchen waste account for the
transformation of topsoils from acidic to neutral pH. Furthermore, it seems likely that
these calcareous waste materials are reflected in enhanced concentrations of calcium
within burgh cores.

High organic matter content and enhanced concentrations of phosphorus within burgh
cores indicate addition of human and animal excreta, domestic refuse and fuel residues.
It is proposed that sustained application of these materials resulted in the formation of
hortic horizons within topsoils at Lauder, Pittenweem and Wigtown. According to the
World Reference Base (WRB) for soil classification Hortic horizons reflect processes of
deep cultivation, intensive fertilisation and/or sustained addition of human and animal
wastes and other organic residues (FAO 2006). Enhancement of topsoil magnetic
susceptibility within burgh cores signifies input of materials which have been heated to
high temperatures, for example charcoal, ash and charred remains (fuel residues)
resulting from burning of peat, wood, moss and coal in domestic and industrial hearths.
Moreover, it is suggested that sherds of fired ceramic materials such as pottery and brick
contribute to magnetic enhancement within burgh cores.

- 247 -
8.1.2 Elemental Signatures

Two elemental signatures are identified at Lauder, Pittenweem and Wigtown. Elements
within first group (Al, Fe, Cr, Li and V) do not exhibit distinct spatial distributions, hence
are not indicative of anthropogenic activity. It is proposed that elements within the
second group (Ba, Ca, Na, K, Co, Cu, Ni, P, Pb, Sr, Ti, Y and Zn) are related to past
anthropogenic activities given that they show consistent patterns of enhancement within
burgh cores. It is suggested that enhanced concentrations of phosphorus reflect burning
of woody material (Aston 1998), human and animal wastes, and domestic rubbish
(Wilson et al., 2008). However, considering the wide range of enhancement within burgh
cores it is difficult to assign individual elements to specific materials and activities.

Elemental enhancement within burgh cores can be attributed to multiple factors;

concentration of elements through combustion of fuel in domestic and industrial hearths,
industrial activities such as craft and metal working, deposition of building materials
including mortar, plaster and lime, and disposal of human and animal wastes. Industries
noted during the mid 16th century in Pittenweem brewing, fishing, butchery, tanning,
cloth production, and gold, silver and iron smithing (Simpson and Stevenson 1981a).
Certain industries such as tanning and metalworking produced wastes are associated
with distinct elemental compositions; however, it is argued that elemental signatures are
not readily identifiable due to pre-depositional integration of domestic and industrial
wastes and post–depositional mixing of elements within the soil matrix.

8.1.3 Waste Materials

Micromorphological analysis of topsoils at Lauder, Pittenweem and Wigtown confirms

that building materials, domestic rubbish and industrial wastes were routinely deposited
within burgh cores. Waste materials types are largely consistent between burghs, for
example bone, pottery, brick, clinker, slag, mortar, plaster and heated stones are present
within all three burgh cores. In contrast shell is limited to Pittenweem and Wigtown. It is
argued that shell reflects the historical legacy of fishing within these towns and their role
as ports (Figure 123, Figure 124). It is likely that shell remains constituted kitchen waste
given their location within back gardens and association with other domestic wastes. It is
also recognised that shell was used as a fertiliser on local farms within Wigtown, for
instance Duncan (1791-99) notes that shells ranging in price from 1s 4d per ton to 1s 6d
and of various qualities were sold at Wigtown harbour.

- 248 -
Figure 123: Wigtown harbour (above) was used primarily as a fishing port between the 15th
and 18th centuries (Simpson and Stevenson 1981b). The harbour fell into disuse in the
early 19th century due to increasing sedimentation and was replaced by a new harbour
400m south of Wigtown (Image © Royal Commission on the Ancient and Historical
Monuments of Scotland, Licensor www.scran.ac.uk)

Figure 124: The Old Boat Haven (left) was used as the main harbour of Pittenweem until
1541AD. It was carved into the rocky shoreline and its pier was used as a natural outcrop
with a road cut into it. Pittenweem harbour (right) was built in 1541AD to accommodate
expanding fishing fleets. The inner harbour comprises the west and east pier, both of
which have been systematically rebuilt over the 17th to 19th centuries, and the outer
harbour is enclosed by the south pier (Images © Royal Commission on the Ancient and
Historical Monuments of Scotland, Licensor www.scran.ac.uk)

- 249 -
8.1.3.1 Harbour and High Street zones
Waste material types are consistent between the Harbour and High Street zones at
Pittenweem; however, shell is more abundant within topsoils in the Harbour zone. It is
expected that industries associated with fishing were situated within the harbour
backlands. Cleaning and processing clams, mussels, cockles and whelks for local
markets and export would have led to considerable quantities of waste. Bait preparation
using shellfish may also explain why higher volumes of shell material occur within
topsoils in the harbour zone. In addition the practise of cleaning and drying fishing nets
suspended across the back gardens could have led to in-situ deposition of marine
derived waste.

In contrast pottery, brick, mortar and plaster materials are more concentrated within the
High Street zone. These materials became more important during the 17th century when
buildings were increasingly made of stone and lime mortar. Utilisation of these materials
would have depended on their availability and affordability. Houses within the High
Street zone were primarily occupied by burgesses who were considerably wealthier than
fisherman who rented cottages adjacent to the harbour (Horsburgh 1856). It is
suggested that higher abundances of building waste reflects the social and economic
disparity of inhabitants. Higher abundances of building waste can be attributed to earlier
and more intensive utilisation of materials such as lime mortar and brick within the High
Street zone. In contrast, the tradition of using turf and clay mortar for building repairs
may have endured longer within the Harbour zone.

8.1.3.2 Topsoil Horizons

Fuel residue is present in upper and lower topsoil horizons at Lauder indicating continual
deposition of waste materials produced through domestic and industrial combustion.
Absence of coarse mineral material from lower topsoil horizons provides clear evidence
for basic stratification of deposits within the burgh core. More recent inputs of domestic
and industrial wastes and building debris suggest a change in either land use within the
backlands or in the management and disposal of waste.

- 250 -
8.1.4 Fuel Residue

Fuel residue within topsoils at Lauder, Pittenweem and Wigtown provides clear evidence
for sustained application of materials resulting from domestic and industrial combustion
to soils within burgh cores. Fuel resources common to all three burghs include peat, turf
and coal. The importance of peat deposits are reflected in the royal burgh foundation
charter of Pittenweem (1546AD) which grants ownership of all surrounding moors to the
burgesses and inhabitants of the town. Turf and peat are cited as the principal fuel
resources at Lauder and Wigtown until the introduction of coal in the latter 18th century
(Duncan 1791-99, Ford 1791-99). Moreover, Nairne (1791-99) reports that coal was
progressively taken from seams underlying Pittenweem until industrialisation of coal
extraction in 1770AD. It is expected that locally available fuel resources such as timber
and mosses were exploited in addition to peat and coal at all three burghs despite
lacking documentary evidence.

Types of fuel residue waste are consistent between burghs, for example charcoal, FR1,
FR 2, FR 3, FR 4 & 6, FR 5 & 7, FR 8 & 9 and FR 10 inclusions are present within all
three burgh cores. This indicates that similar fuel sources were exploited at Lauder,
Pittenweem and Wigtown. However, results of micromorphological analysis reveal
differences in principal fuel residue types between burgh cores. It is suggested that
variation in fuel residues indicate preferential exploitation of certain fuel resources. It is
equally probable that fuel residue types are linked to differences in firing temperatures
and combustion processes associated with industrial activities. The effect of varying
temperature on ceramics (Tobert 2007) and archaeological sediments (Berna et al.,
2007) is well documented. It therefore seems reasonable to suggest that changes in
associated fuel residues are affected by similar processes. Furthermore, it is argued that
post-depositional processes are not responsible for selective preservation of fuel
residues given similarities in soil physical and chemical properties between burgh cores.

8.1.4.1 Harbour and High Street Zones

Fuel residue types are consistent between the Harbour and High Street zones at
Pittenweem; however, FR 4 & 6, FR 5 & 7 and FR 8 & 9 are more abundant within the
Harbour zone and FR 3 is greatest within the High Street zone. Differences in principal
fuel types may reflect the social and economic disparity of inhabitants, for example
burgesses within the High Street zone may have adopted the use of coal earlier than
poorer residents of the Harbour zone. Analogous evidence from Wigtown highlights the
relationship between social standing and fuel resources. Duncan (1791-99:485) notes
‘the common people, both in the town and country, burn peat, of which indeed there is

- 251 -
abundance within this parish’ and that ‘the better sort of inhabitants within the town,
though they used peat in their kitchens, burn coal in their rooms’. It is probable that
variation in fuel residue types are linked to differences in firing temperatures and
combustion processes associated with differing industrial activities between zones.

8.1.4.2 Topsoil Horizons

Fuel residue is present in upper and lower topsoil horizons within all three burghs. This
indicates continual disposal of waste materials associated with domestic and industrial
combustion. However, fuel residue is less abundant within lower topsoil horizons at
Lauder and Wigtown. Differences in the concentration of fuel residue may reflect
changing patterns in the availability and affordability of fuel resources, for example the
introduction of turnpike roads at Lauder led to widespread adoption of coal as the main
fuel resource during the latter 18th century (Ford 1791-99:77). It is also suggested that
an increase in fuel waste could be linked to population growth or intensification within
burgh cores. Ford (1791-99:74) notes a general increase in the population of Lauder
since the opening of turnpike roads. Moreover, at Wigtown the population grew from
1032 in 1755AD to 1350 in 1793AD (Duncan 1791-99:488). Increasing industrialisation
may account for increased deposition of fuel waste; however, industries such as coal,
lime and woollen manufacturers were still lacking at both Lauder and Wigtown during the
1790s (Duncan 1791-99:485, Ford 1971-99:76).

- 252 -
8.2 The Impact of Waste Disposal on Burgh Hinterlands

This section focuses on the impact of waste disposal within the Hinterland Near zone at
all three burghs, the Hinterland Far zone at Pittenweem and the Showfield zone at
Wigtown. These zones correspond to the location of the burgh acres which were strips
of arable land owned by burgesses (see section 2.3.4). The burgh acres, known as
burgess acres at Lauder, were located immediately beyond the burgh core and
extended to the common grazings (Romanes 1914, Johnston 1920) (Figure 125). There
is no direct reference to the location of the burgh acres at Pittenweem although their
existence is noted (Cook 1867, Leighton 1840, Webster 1819). It is suggested that the
burgh acres extended over the Hinterland Near and Far zones given the delineation of
land running parallel to burgage plots (Figure 126). This is supported by Leighton
(1840:108) who states that ‘the lands of Greendikes, Waterless and Coalfarm, lying to
the north-west and west of the burgh are the property of Sir Wyndham Carmichael
Anstruther, baronet’ and that ‘the other lands in the parish consist of burgh acres’.
Similar to Lauder and Pittenweem the burgh acres at Wigtown were located immediately
beyond the burgage plots within the burgh core (Duncan 1791-99:474, Brewster 1832:
521) (Figure 127).

Figure 125: Map of Lauder extracted from the 1747-1755AD William Roy Military Survey of
Scotland © The British Library Board, Licensor www.nls.uk/roy/style.html. Red boundaries
represent buildings and man-made structures and parallel hatching indicates cultivated
land. Burgh acres are represented by the delineation of strips of land running parallel to
burgage plots

- 253 -
Figure 126: Map of Pittenweem extracted from the 1747-1755AD William Roy Military
Survey of Scotland © The British Library Board, Licensor www.nls.uk/roy/style.html. Red
boundaries represent buildings and man-made structures and parallel hatching indicates
cultivated land

Figure 127: Map of Wigtown extracted from the 1747-1755AD William Roy Military Survey
of Scotland © The British Library Board, Licensor www.nls.uk/roy/style.html. Red
boundaries represent buildings and man-made structures and parallel hatching indicates
cultivated land. Burgh acres are represented by the delineation of strips of land running
parallel to burgage plots north and south of Wigtown

- 254 -
8.2.1 Soil Modification

Deepened topsoils adjacent to the historical burgh limits at Lauder and Pittenweem
provide clear evidence for sustained application of mineral material to the burgh acres.
Accumulation of topsoils can be explained through addition of waste rich in mineral
material, for example turf used in building construction, ash from domestic and industrial
fires, sand used in animal byres and in the case of Pittenweem, sand associated with
addition of seaweed. In contrast there is no evidence for topsoil deepening within the
Hinterland Near and Showfield zones at Wigtown. This finding indicates that waste rich
in mineral material was not routinely applied to the burgh acres. It is possible that waste
may have been preferentially deposited within the burgh core or within the burgh acres
beyond the extent of the survey area. Additionally, waste may have been sold for use as
fertiliser to other burghs.

High organic matter and enhanced concentrations of phosphorus within the Hinterland
Near zone at Lauder and Showfield zone at Wigtown indicate addition of human and
animal excreta, domestic refuse and fuel resides. In the case of Lauder sustained
application of such materials may explain the formation of hortic horizons within topsoils
in the burgh acres. There is no association between organic matter and phosphorus
within the Hinterland Near zone at Pittenweem. Considering its current use as managed
grassland, elevated organic matter within the Hinterland Near zone may be the result of
modern inputs such as decomposing plant matter. Phosphorus concentrations may have
decreased due to past cultivation, for example intensive cropping which exceeds
phosphorus deposition rates. Nevertheless it is possible that increased organic matter
content may reflect past inputs such as animal dung; even though it is argued more
modern sources of phosphorus are more likely.

Enhancement of topsoil magnetic susceptibility within the Hinterland Near zone at

Lauder and Pittenweem and Showfield zone at Wigtown indicates input of materials
which have been heated to high temperatures, for example charcoal, ash and charred
remains (fuel residues) resulting from burning peat, wood, moss and coal in domestic
and industrial hearths. Additionally sherds of fired ceramic materials such as pottery and
brick may contribute to magnetic enhancement within the burgh acres.

- 255 -
8.2.1.1 Variation within Burgh Acres
The median depth of topsoils within the Hinterland Near and Hinterland Far zones at
Pittenweem is greater than 50cm, thus providing evidence for the additional of mineral
waste within the burgh acres. Nevertheless there is a significant difference in topsoil
depth, organic matter and magnetic susceptibility between these two zones. It is
proposed that the impact of waste diminishes with distance from the burgh core resulting
in a distance decay effect across the burgh acres. Differences in the impact of waste
disposal are also apparent within the burgh acres at Wigtown, for example enhancement
of organic matter content and frequency dependant magnetic susceptibility is limited to
the Showfield zone. This may reflect selective deposition of wastes such as fuel residue,
domestic refuse and human and animal excreta to the burgh acres south of Wigtown.

8.2.2 Elemental Signatures

The Hinterland Near zone at Lauder is enhanced in elements associated with past
anthropogenic activity (Ba, Ca, K, P, Pb and Zn). There is no difference in the range of
elements enhanced between the burgh core and burgh acres thus indicating similarities
in the nature of waste material inputs. It is suggested that domestic and industrial waste
generated within the burgh core was deposited within the burgh acres at Lauder for the
purpose of soil improvement. Elements enhanced within the Thirlstane zone include Ba,
P and Pb; nevertheless, it is difficult to assign these elements to particular sources of
waste or anthropogenic activities given their limited range. Land within this zone is
associated with Thirlstane Castle (Figure 128). Thirlstane Castle was built in the 16th
century AD on the original site of Lauder Fort and has been successively occupied until
present day (Simpson and Stevenson 1981c). The grounds of Thirlstane Castle were
primarily used for recreation (Cosens 1834-45), although more recently additional uses
include keeping rare breed livestock and hosting horse trials.

Enhancement within the burgh acres at Pittenweem is limited to elevated concentrations

of Mn, Co, Mg and Zn within the Hinterland Near zone. This is in contrast to the burgh
core which is characterised by a range of anthropogencially significant elements (Ba,
Ca, Na, K, Co, Cu, Ni, P, Pb, Sr, Ti, Y and Zn). Differences between these areas may
reflect preferential deposition of certain wastes. It is possible that domestic and industrial
waste generated within the burgh core was deposited within the burgh acres at
Pittenweem; however, post-depositional processes such as leaching and past land
management within the burgh acres may adversely affect elemental retention rates.

- 256 -
 C
B
A

Thirlstane Castle

Figure 128: Aerial Image of Lauder showing (a) burgh core, (b) Hinterland Near zone, (c)
Hinterland Far zone and (d) Thirlstane zone, image © Andrew Buchanan 2004, www.holy-
cow.co.uk/

The Showfield zone at Wigtown is enhanced in a range of elements associated with past
anthropogenic activity including, Ba, Cu, Mn, P, Pb and Zn. Comparable patterns of
elemental enhancement within the burgh core and Showfield zone signify similar waste
material inputs. It is noted that Ca, K and Sr are not elevated within the Showfield zone;
this may be due to preferential leaching within the burgh acres. Moreover, absence of
elevated elemental concentrations within the Hinterland Near zone support the proposal
that waste was selectively applied to the burgh acres south of Wigtown (section 8.2.1.1).

8.2.3 Waste Materials

As discussed in section 8.1.3, building materials, domestic rubbish and industrial wastes
were routinely deposited within burgh cores. Evidence for these materials within the
burgh acres at Lauder and Wigtown is limited, suggesting selective application of waste
material. In contrast pottery, brick, clinker, slag, mortar, plaster and heated mineral
material occur within the burgh acres at Pittenweem. This indicates deposition of similar
types of waste materials within the burgh core and burgh acres. Higher abundances of
waste materials within the Hinterland Far zone may reflect preferential deposition of

- 257 -
certain wastes away from the burgh core. It is also possible that post-depositional
processes such as ploughing and tilling within the Hinterland Near zone resulted in
increased fragmentation of exotic inclusions.

8.2.4 Fuel Residue

Fuel residue within the burgh acres at Lauder, Pittenweem and Wigtown provides clear
evidence for addition of materials resulting from domestic and industrial combustion. In
addition, fuel residue is identified within the Hinterland Far and Thirlstane zone at Lauder
and Hinterland Far zone at Pittenweem. A distinction in fuel residue types between the
burgh core and hinterland is noted at Lauder and Wigtown. Inclusions of charcoal, FR 2
and FR 10 are absent from the hinterland at both of these burghs suggesting preferential
deposition of certain fuel residues within the burgh core. It is also possible that such
materials could have been deposited in the hinterland but have either been subject to
selective preservation or exist in abundances beyond the limit of detection. In contrast
there is no difference in the nature of fuel residue between the burgh core and hinterland
at Pittenweem. Fuel residue types are generally consistent between burghs, for example
FR1, FR 4 & 6, FR 5 & 7 and FR 8 & 9 are present within all hinterland zones. Moreover
there is no difference in the abundance of fuel residue types either within or between
burgh hinterlands.

- 258 -
8.3 Waste Disposal

8.3.1 Burgh Core

Based on the micromorphological and chemical results there are general similarities in
waste types between burgh cores, for example building materials, human and animal
excreta, kitchen refuse, industrial wastes and fuel residue. Nevertheless, differences in
waste materials are apparent between burghs. Principal fuel residue wastes vary
between burghs reflecting differing fuel resources and/or industrial processes. In
addition, the occurrence of shell at Pittenweem and Wigtown is indicative of industrial
processes related to fishing.

Sustained addition of waste materials in burgh cores led to improvement of topsoil,

facilitating widespread cultivation of garden soils. Production of crops was of economic
importance; crops were needed for the sustenance of livestock and for sale at market. In
addition, urban horticulture was socially significant. Given that backland plots were often
divided and sublet, poorer inhabitants were able to reliably produce foodstuffs within a
concentrated area thus enhancing food security.

Although it is accepted that soils in burgh cores reflect sustained deposition of waste
materials, it is not possible to attribute any one method to their formation. It is likely that
domestic refuse associated with individual households and mixtures of straw, sand and
dung from byres were applied to the backlands as a convenient source of fertiliser.
Considering the diversity of materials in burgh core topsoils, it is also suggested that
middens comprising domestic and/or industrial wastes may have been periodically
spread across the backlands. Potentially waste from dunghills may have been added to
burgage plots; however, it is argued that they chiefly acted as stores of urban waste prior
to redistribution within the hinterland.

This discussion has focussed on the role of cultivation in the formation of ‘garden soils’
within burgh cores. Nevertheless, it is contested that such deposits reflect accumulation
of sediments resulting from intensified occupation of the backlands (Carter 2001). Given
the formation of hortic horizons within burgh cores it is argued that soils at Lauder,
Pittenweem and Wigtown were formed predominantly through urban horticulture. In
addition, it is proposed that intensified backland occupation would have led to growing
demand for additional foodstuffs thus increasing cultivation rather than replacing it.

- 259 -
8.3.2 Hinterland

Sustained addition of waste materials to the hinterland at all three burghs resulted in
enhanced soil fertility within the burgh acres. The degree of soil improvement within the
burgh acres is reflected in its letting cost, for example Duncan (1791-99:479) notes that
arable land arable land at Wigtown is let at 10 to 20 shillings per acre rising to between
50 shillings and 3 pounds within the burgh acres. Similarly, the burgh acres at Lauder
are three to four times more expensive than arable land in the rest of the parish (Ford
1791-99:73). Crops provided an important source of revenue for burgesses and their
success was vital for meeting the consumption needs of inhabitants and their livestock.
The deliberate addition of burgh wastes would have improved the quality and yield of
crops, in addition to improving sustainability of cultivation.

Although not conclusive, it is proposed that dunghills were routinely applied to soils
within the burgh acres (see section 2.2.4). The application of dunghills would have
resolved problems associated with the formation of dunghills in burgh cores, in particular
their obstruction of thoroughfares (Cook 1867). In addition dunghills held an economic
value hence were often traded or sold to farmers or confiscated by the burgh authorities
(Figure 129). This is especially true at Pittenweem where legislation throughout the 17th
and 18th centuries repeatedly cites forfeiture of dunghills without compensation (Cook
1867: 86, 99, 152). It is most likely that dunghills were transferred to the burgh acres
using horse and cart although the additional use of labourers cannot be ruled out. In
addition to dunghills it is acknowledged additional sources of fertiliser could be sought
from the sea, for example shells of various qualities could be bought from Wigtown
harbour for manure on farms (Duncan 1791-99). Similarly seaweed was gathered and
collected from the beach at Pittenweem for sale to local farmers (Figure 130).

In agreement with Davidson et al., (2006), soil improvement within the burgh acres
reflects an early form of urban composting. This is particular resonant in light of
agricultural reform during the 18th and 19th centuries. Past land management systems
were considered ineffective; however, at Wigtown Duncan (1791-99:477) notes how
‘improvements to farms using manure have made it possible to raise bere barley. No
bere was previously raised in this parish except for the burgh acres’.

- 260 -
Figure 129 Invoice for the sale of ‘street manure’ at Pittenweem to a local farmer dating
1919 AD, dunghills were therefore a persistent feature in burgh cores (Image © Scottish
Life Archive, Licensor www.scran.ac.uk)

Figure 130 Photograph of a man with a cart load of seaweed in the early 20th century
Pittenweem. In previous centuries the right to collect seaweed was associate with land
ownership, hence burgesses had rights to seaweed on particular stretches of beach
(Image © National Museums Scotland, Licensor www.scran.ac.uk)

- 261 -
 9 The Wider Significance of Waste Disposal in and
Around Historic Small Towns

This chapter discusses the wider significance of waste disposal in and around historic
small towns. Key study findings are presented and recommendations for further work are
made accordingly. In addition, the significance of waste disposal in and around historic
towns is discussed in reference to urban soil classification, rural-urban interactions and
trends in archaeology.

9.1 Study Findings

9.1.1 Objective 1

The first objective of this study was to establish the nature and diversity of urban
anthrosols in and near to historic small towns. It was found that soil characteristics are
varied within and near to Lauder, Pittenweem and Wigtown (Hypothesis 1.1). Distinct
patterns in spatial distributions of topsoil depth, pH, organic matter content, magnetic
susceptibility and selected elemental concentrations are identified at all three burghs.
Soil properties within and near to Lauder, Pittenweem and Wigtown are determined by
sustained addition of waste materials (Hypothesis 1.2). Waste material inputs include
human and animal excreta, domestic waste, industrial waste, building materials and fuel
residues.

9.1.2 Objective 2

The second objective of this study was to characterise and account for the multiplicity of
urban anthrosols in and near to historic small towns. It was found that urban anthrosol
characteristics vary within burghs according to past functional zones (Hypothesis 2.1
and 2.2). The burgh core and burgh acres are important areas of interest at all three
burghs. Soil modification is most pronounced within burgh cores reflecting a greater
diversity and abundance of waste material inputs. Differences in urban anthrosol
characteristics are evident between burghs. The nature and extent of soil modification
differs between burgh acres. It is suggested that shell waste at Pittenweem and Wigtown
is linked to their past function as fishing ports (Hypothesis 2.3). In addition variation in
principal fuel residue types between burghs may indicate differences in resources
exploitation and/or combustion processes associated with industry. Further work is

- 262 -
needed to resolve this issue; it is recommended that visual and elemental comparisons
of key fuel residue types should be made with reference materials.

9.1.3 Objective 3

The third objective of this study was to elucidate the processes associated with waste
management and disposal in historic small towns. It is proposed that addition of waste
materials to soils within and near to Lauder, Pittenweem and Wigtown was an effective
waste management strategy. Waste disposal led to enhanced soil fertility within the
burgh core and to a lesser extent the burgh acres at all three towns. This is important
given that urban horticulture and arable farming was central to resource production in
historic small towns. It was hypothesised that processes of waste disposal could be
deduced from properties of soils (Hypothesis 3.1). This proposal was overly ambitious
considering similarities in waste management strategies between burghs in addition to
the impact of post-depositional processes. It is likely that direct waste deposition,
storage and redistribution of midden waste, and storage and redistribution of dunghills
were important modes of waste disposal at all three burghs. It is expected that
differences in the nature and modes of waste management existed between burghs;
however, this information could not be ascertained from soil properties alone
(Hypothesis 3.2). It is recommended that detailed documentary analysis may resolve this
issue.

9.1.4 Model Revision

A revised model of resource and waste material flow in historic towns is presented in
Figure 131. In the context of this research the town zone corresponds to the built
environment within burgh cores. This town encompasses areas characterised by urban-
rural interaction and sectoral interactions, for example backlands were used for urban
horticulture and animal rearing. The hinterland is a theoretical area of resource
exploitation which differs from town to town. All three towns are similar in that the
immediate hinterland encompasses the burgh acres, an area explicitly linked with urban
inhabitants through processes of fertilisation, crop/vegetable growth and animal rearing,
and procurement of foodstuffs. In addition each town exploited a wider area to acquire
fuel sources and trade items, for instance Pittenweem had direct trading links with the
continental Europe and Wigtown had contacts with Ireland and Cumbria. The burgh and
its hinterland are linked by rural-urban fluxes of resources, goods and waste materials;
hence, are not mutually exclusive zones.

- 263 -
Figure 131: Revised flow model of resource and waste movement in historic small towns

Acquisition represents movement of materials from the hinterland to the town. These
materials range from locally available natural resources to manufactured trade goods.
Fuel sources were obtained from the hinterland in all three burghs through activities
such as peat cutting, moss harvesting, coal mining and wood felling. The utilisation of
these materials varied from town to town, and was dependant on a range of geographic
and economic factors. Materials could also be acquired from within the town itself. There
is geoarchaeological evidence in the form of hortic topsoils to suggest urban horticulture
was an important source of foodstuffs within the burgh core at Lauder, Pittenweem and
Wigtown. Acquired resources are incorporated into a cycle of use and re-use until they
hold no remaining social or economic value.

Materials with no remaining economic or social value are classed as waste and
transition to a cycle of waste management characterise by processes of storage and re-
distribution. The two common storage mechanisms at all three burghs were middens in
the backlands and dunghills in the main thoroughfares. The constituents of middens
both within and between burghs remain unresolved given the homogeneous nature of
backland topsoil deposits. It is possible that specific middens were used for differing
waste types such as domestic and industrial rubbish. Dunghills are documented in all
three burghs with repeated legislation referring to their nuisance. These dunghills were
removed from the burgh either voluntarily through sale to burgesses or under forfeiture,

- 264 -
and applied to the burgh acres as fertiliser. Processes of midden spreading in the
backlands and dunghill movement to the burgh acres are classed as redistribution of
waste.

Deposition of waste falls into two categories; point and diffuse. At all thee burghs there is
evidence for point deposition of domestic/kitchen wastes in back gardens, for example
bones, shell and pottery fragments. Similarly there is evidence for diffuse deposition in
the form of hortic topsoil in burgh cores and modified topsoil in the burgh acres. Upon
deposition waste is incorporated into the soil matrix and subject to reworking and system
loss. One of the main limitation of the original flow model presented in chapter 3 (Figure
13) was that waste materials had no function upon incorporation into the soil matrix
aside from involvement in post-depositional processes. Given that waste was used for
soil improvement in the burgh acres it is argued that resource production and waste
management within and near to historic towns were inextricably linked. Accordingly flow
arrows linking soils to town and hinterland resource cycles are proposed as suitable
model amendments (Figure 131).

- 265 -
9.2 Rural-Urban Interaction

9.2.1 Modern Significance of Rural-Urban Interaction

Parallels in the nature and management of organic wastes are evident between past and
present urban environments. The main sources of organic waste in historic and modern
towns include kitchen refuse, human excreta, and horticultural and agricultural waste.
Similarities in urban organic waste management strategies are outlined in Table 32.
Organic waste in Lauder, Pittenweem and Wigtown, comprised domestic/kitchen refuse,
dung from urban livestock, human excreta and wastes from industrial processes. Whilst
there are broad similarities in the nature of organic waste between burghs the volume of
waste was likely to differ. The amount of waste produced would have depended on
population size and livestock numbers.

In modern towns organic waste is dealt with in a variety of ways including; organised
composting of solid and/or animal wastes prior to application, direct application of solid
waste onto soil, and application of animal wastes onto soil. It is evident from this study
that urban organic waste recycling is not a recent phenomenon. Waste management
strategies in historic towns similarly involved composting, for example animal wastes
and kitchen refuse in byres and middens.

Table 32: Comparison of urban organic waste recycling practises in modern and historic
towns, modern examples taken from Cofie and Bradford (2006)

Modern Towns Historic Towns

Use of fresh waste from vegetable Use of kitchen refuse, waste from markets
markets, restaurants, hotels and food and waste from horticulture as feed for
processing industries as feed for urban urban livestock
livestock
Direct application of solid waste on and Application of human and animal waste to
into the soil backlands and/or hinterland via dunghills
Mining of old waste dumps for application Use of landfills not a key feature of historic
as fertiliser on farmland towns
Application of animal manure such as Application of waste from stalled animals
poultry/pig manure and cow dung and other urban livestock to backlands
and/or hinterland via dunghills
Direct application of human excreta or bio- Application of human waste to backlands
solids to the soil and/or hinterland via dunghills
Organised composting of solid waste or Accumulation of human and animal waste
co-composting of solid waste with animal in middens and dunghills prior to soil
manure or human excreta application

- 266 -
Waste was applied directly to gardens in the burgh core and transported to the burgh
acres for use as fertiliser. Composting and soil improvement remain key strategies in
managing organic waste in modern towns; however, one of the key differences is the
utilisation of human wastes. In the past there was limited awareness of dangers
associated with using human faecal material as a source of compost. In contrast, human
waste in contemporary towns is processed and disposed by means of urban sewerage
systems which are subject to governmental regulations. Another difference in organic
waste management between historic and modern towns is the use of landfill sites (Table
32). It is suggested that landfill sites were not in burghs for disposal of organic wastes
given the economic and social value of waste as a fertiliser.

Urban horticulture is an activity deep-rooted in past urban environments. In Lauder,

Pittenweem and Wigtown the ability to grow food within the urban area was vital in the
supply of foodstuffs, especially for poorer inhabitants who could not afford to rent arable
land in the burgh acres. Similarly cultivation in developing cities is increasingly seen as a
key strategy for enhancing food security of present and future populations. This is
especially resonant in urban environments currently lacking an environmentally
sustainable solution to waste management of organic materials (Tixier and Bon 2006).

- 267 -
9.3 Soil Classification

This section discusses problems associated with mapping urban soils and classifying
soils in historic small towns with specific reference to key findings at Lauder, Pittenweem
and Wigtown.

9.3.1 Mapping Urban Soils

Conventional soil mapping is simplistic in its approach to urban soils, whereby soils are
either assigned to local soil series or are omitted from soil classification maps
completely. Soils in and near to Lauder are mapped as freely drained brown forest soil
with low base status belonging to the Lauder (LA) soils series (Figure 132). However,
results of this study indicate distinct changes in soil properties within and adjacent to
Lauder through processes of anthropogenic modification. It is argued that hortic soils in
the burgh core and burgh acres differ significantly from local soils. In Pittenweem and
Wigtown soils within the burgh core are not mapped (Figure 133, Figure 134). Results
from this study confirm the presence of hortic anthrosols at both burghs. It is suggested
that such soils have clearly identifiable physical and chemical soil properties and should
not be omitted from future classifications.

Figure 132: Soil map of Lauder © Soil Survey of Scotland 1959. Soils within and adjacent
to Lauder are mapped as a freely drained brown forest soil with low base status belonging
to the Lauder (LA) soil series

- 268 -
Figure 133: Soil map of Pittenweem © Soil Survey of Scotland 1975. Soils within
Pittenweem are not classified and adjacent soils are mapped as an imperfectly drained
brown forest soil belonging to the Quivox (QX) soil series

Figure 134: Soil map of Wigtown © Soil Survey of Scotland 1971. Soils within Wigtown are
not classified and adjacent soils are mapped as a freely draining brown forest soil
belonging to the Linhope (LP1) soil series

- 269 -
Soil within and near to Lauder, Pittenweem and Wigtown are modified through past
waste amendments; hence differ from surrounding ‘natural’ soils. It is therefore
recommended that soils associated with urban environments should be included in soil
maps. This would require a comprehensive system of classification distinguishing
between historic and modern soil modification (see section 9.3.2). Moreover, specific
attention should be paid to soils modified through past and present urban activities
which are not located in towns, for example historic improvement of arable land and
modern refuse dumping on landfill sites.

9.3.2 Classification of Soils in Historic Towns

Classification of soils in historic towns is problematic due to conflicting terminology both

within and between soil classification systems. Soils in and around Lauder, Pittenweem
and Wigtown are classified as Anthrosols within the World Reference Base for Soil
Resources (WRB) classification system despite the recent introduction of a Technosols
(Urban/Mine) group (Table 33). It is argued that Anthrosols is more appropriate for soils
in historic towns given they account for processes of waste disposal, soil improvement
and cultivation. In contrast Technosols are concerned with waste amendments resulting
from modern urban environments such as landfills, impermeable road surfaces and mine
spoil. The Australian Soil Classification System makes a distinction between Hortic,
Garbic and Urbic Anthroposols (Table 33). Under this system soils in burgh cores could
potentially be classed as Hortic or Garbic Anthroposols. Similar to the WRB classification
Urbic Anthroposols are not appropriate for classifying soils in historic towns. It is
suggested that conflicting terminology is a result of modern definitions of urban and rural
soil characteristics. In the past cultivation and associated improvements in soil fertility
were urban features; however, these are currently viewed as rural qualities. Likewise,
characteristics which are viewed as urban, such as industrial wastes and concrete
pavements are modern in origin.

- 270 -
Table 33: Summary of soil classification terminology relevant to soils in historic towns
(FAO 2006, Gong 1999, Isbell 2002, Nikolaevna and Vadimovna 2003)

System Term Definition

Australian Hortic Anthroposol Soils that have additions of organic residues
such as organic wastes, composts and
mulches that have been incorporated into
the soil and have obliterated pre-existing
pedological features.
Australian Urbic Anthroposol Mineral soil or regolithic materials that are
underlain by land fill of a predominantly
mineral nature.
Australian Garbic Anthroposol Mineral soil or regolithic materials that are
underlain by land fill of manufactured origin
and which is predominantly of an organic
nature. These materials may be of domestic
or industrial origin.
Chinese Anthropic Surface Diagnostic surface horizons that are the
Horizon result of agricultural activities that have
caused major changes in soil processes at
or near to the soil surface.
Chinese Cumulic Epipedon Formed by long-term cultivation, applying
manure or adding soil material rich in
organic matter or other mud’s to the soil.
Chinese Mellowic Epipedon Formed by planting vegetables and/or
adding night soil, organic trash or manure to
the soil under intensive cultivation and
frequent irrigation over a long period.
World Reference Anthrosols Soils that have been modified profoundly
Base for Soil through human activities, such as addition of
Resources materials or household wastes, irrigation and
cultivation.

World Reference Technosols Soils whose properties and pedogenesis are

Base for Soil (Urban/Mine Soils) dominated by their technical origin. They
Resources contain a significant amount of artefacts
(something in the soil recognizable made or
extracted from the earth by humans), or are
sealed by technic hard rock.
World Reference Kitchen Soils Anthrosols having a hortic horizon >50cm.
Base for Soil Horizon is thoroughly mixed with original
Resources strata usually not preserved. Artifacts and
cultural debris commonly occur.
Russian Urban Urbanozem A genetically individual soil which combines
Soil properties of natural soils in neighbouring
Classification areas and specific properties developed in
System the urban environment.

- 271 -
It is proposed that the term Urbanozem within the Russian Urban Soil Classification
System offers a suitable solution to soil classification in both historic and modern towns.
This system theorises that urban soils develop in three major ways; on loose artificial
deposits, on the cultural layer/anthropic material and by transforming natural soils
(Figure 135). Further, it is suggested that evolution and transformation of urban soils is
controlled by land-use, substrate type and time. The incorporation of time as a key factor
in urban soil formation accounts for differing intervals of soil modification in ancient,
historic and recent urban environments. Soils in and around Lauder, Pittenweem and
Wigtown are formed by transforming natural soils and through development of a clearly
identifiable cultural layer. These processes occurred over a time period spanning the
medieval and early modern periods, prior to the introduction of sewerage. In contrast
soils formed through recent activities such as urban landscaping are formed on loose
artificial deposits and consolidate over tens of years. The former historical example and
latter modern urban example are both accounted for under the term Urbanozem. It is
argued that a more subjective approach to soil classification in historic environments is
needed which takes into consideration site specific characteristics such as land use,
period of use/occupation and geographical location.

Figure 135 Formation of the Urbic horizon (Nikolaevna and Vadimovna 2003)

- 272 -
9.4 Soil: An Archaeological Resource

This section evaluates the contribution of soil as an archive of waste deposition and
management within historic small towns. More specifically evidence obtained from the
soils based cultural record is compared to information derived from archaeological
excavation and documentary evidence.

9.4.1 Comparison of Soils and Archaeological Excavation

Archaeological excavation is primarily concerned with recovery of artefacts. Accordingly

lack of artefacts is often interpreted as absence of archaeology, for example results of a
watching brief at High Vennel, Wigtown indicate that no deposits relating to the medieval
burgh core were located despite identification of deposits ranging in depth from 60 to
110 cm (Appendix 3). In contrast, results from this study suggest the occurrence of
deposits modified through past waste amendments within the burgh core at Wigtown.
The results of an archaeological evaluation at 64a High Street, Pittenweem infer
deposits have little significance aside from containing limited ceramic finds (Appendix 4).
Nevertheless analysis of soil pit PT 6 which is located in the garden of 64a High Street
provides clear evidence for sustained addition of domestic and industrial wastes.

Concentrations of rubbish discovered on archaeological sites are usually referred to as

middens regardless of their spatial characteristics and content. Misidentification of
middens can be attributed to the site specific nature of urban excavation, for example
excavations within the back garden at 5 Mid Shore, Pittenweem refer to a 17th century
midden containing pottery, clay pipes, animal bones and miscellaneous ironwork (Martin
1978, Martin 1979). In comparison, analysis of soils within the burgh core indicates the
occurrence of ‘refuse rich’ deposits. It is expected that spatially discrete middens are
present in addition to widespread refuse rich soils within burgh cores. Confusion
surrounding the use of these terms needs to be resolved, particularly in archaeology.

Both archaeology and soil analyses can be used to examine interaction between historic
towns and their hinterland. One of the key advantages of excavation it that it enables
recovery of artefacts which can be linked to source locations and assigned to specific
periods in time. Ceramic finds from Pittenweem infer trading connections with mainland
Europe in the 16th and 17th centuries (Figure 136). Moreover, soil analyses enable
investigation into the relationship between towns and their local hinterland through
resource and waste material flow.

- 273 -
Figure 136: Selected pottery finds from Pittenweem. First row (left to right), 16th century AD Valencian Lustreware bowl (Spain), 17th AD century
pottery from Saintonge (south-west France), 17th century cooking pots (north Germany), 17th century AD earthenware skillets (Scotland). Second
row (left to right), 17th century AD Green Glazed jars (Scotland), 17th century AD Loire Ware Jugs (west France), 17th century AD Westerwald AD
monochrome stoneware (Germany), 17th century AD Westerwald polychrome stoneware (Germany) (Images © Colin J Martin, Licensor
www.scran.ac.uk)

- 274 -
9.4.2 Comparison of Soils and Documentary Evidence

It can be argued that analysis of soils is unnecessary for investigating resource and
waste material flow in historic towns particularly in cases where documentary evidence
is abundant. However, historical sources are restricted to what contemporaries deemed
important at the time, for example legislation, taxation and issues relating to the nobility.
Consequently historical sources in this study were used to support discussions based on
soil analyses rather than as a primary research tool. It is suggested that the historical
legacy of soil enables insight into activities such as waste disposal and soil improvement
not generally accounted for by other means.

9.4.3 Conservation

Urban archaeology is rescue driven favouring maximum recovery of artefacts within a

limited timeframe; accordingly urban deposits are often overlooked. This study shows
that urban soils contain a wealth of information not readily available through excavation
or documentary analysis alone. It is proposed that sampling urban soils should be a
routine procedure in archaeological excavation to promote comparisons between sites
and supplement information derived from finds. Conservation in historic towns is heavily
centred on preserving surviving structures such as statues, churches and civic buildings;
however, it is argued that ‘monumentalism’ does not represent the daily experience of
urban life in the past. Conversely, soils represent a unique and valuable archive of
everyday life which has largely been ignored. It is argued that soil is as important as
individual artefacts and should be treated as such in respect to issues of heritage and
conservation. It is recognised that this is not an easy task considering soils lay beneath
existing communities. In addition, further discussion is needed to resolve preservation
and utilisation strategies.

- 275 -
9.5 Summary of Key Findings and Future Recommendations

9.5.1 Summary of Key Findings

Urban anthrosols in and around Lauder, Pittenweem and Wigtown have been modified
through sustained addition of waste. Soil modification is most pronounced in burgh cores
resulting in the formation of hortic soil horizons. Soils within all three burgh cores are
characterised by deepened topsoil, neutral pH, increased organic matter content,
enhanced magnetic susceptibility and elevated elemental concentrations such as
calcium, phosphorus and potassium. These amended soil physical and chemical
properties are attributable to a variety of wastes. Deepened topsoil indicates
accumulation of mineral rich materials such as turf, ash from hearths, sand added to
byre waste, and sand associated with seaweed. Increased organic matter content and
elevated phosphorus concentrations suggest addition of animal and human excreta, and
domestic rubbish. Modification of soil pH from acidic to neutral conditions imply
deposition of calcium rich materials such as mortar, plaster and lime washes used in
building construction, and shell and bone associated with kitchen waste. Furthermore
magnetic enhancement of topsoils indicates addition of materials heated to high
temperatures such as domestic and industrial fuel residues. The nature and extent of
soil modification within burgh acres is more varied. At Lauder hortic soils were identified
in the burgh acres suggesting pronounced soil modification through cultivation.
Deepened topsoil in the burgh acres at Pittenweem provided evidence for application of
mineral rich waste materials in the past. Moreover, magnetic and elemental
enhancement (barium, phosphorus, lead, zinc) within the burgh acres south of Wigtown
revealed historic soils based anthropogenic signal.

Although processes associated with waste disposal could not be deduced directly from
urban anthrosol properties, micromorphological analyses provided an insight into the
nature and distribution of deposited wastes. Waste materials were most abundant and
varied in burgh cores comprising domestic waste, animal waste, building materials and
fuel residues. These materials were found to varying extents in the burgh acres at all
three burghs; however, they were notably less abundant. Variation in urban anthrosol
characteristics between burghs is attributed to differing industries and patterns of
resource exploitation, for example marine waste associated with fishing was only
identified in coastal burghs. Similarly, variation in the abundance of marine waste and
building materials within the burgh core at Pittenweem may reflect differences in the
location of industries associated with fishing, and contrasting building traditions.

- 276 -
Sustained addition of waste materials to soils within and around Lauder, Pittenweem
and Wigtown was an effective waste management strategy. Waste disposal in burgh
cores was likely to be a combination of direct application and midden spreading in back
gardens. This led to enhanced soil fertility which was important in the development of
urban horticulture; particularly for poorer inhabitants who did not have access to arable
farm land adjacent to the burgh. Dunghills acted as temporary stores of waste in the
main thoroughfares of Lauder, Pittenweem and Wigtown. These dunghills were
systematically transported to the burgh acres for further use as a fertiliser; hence, an
early form of urban composting.

9.5.2 Recommendations

Discussions presented in sections 9.1, 9.2 and 9.3 highlight future areas of research in
which the findings of this study could be applied. In the first instance, there is a need for
re-evaluation of traditional soil classification and soil mapping in relation to historic and
urban soils. It is suggested that the Russian Urban Soil Classification outlined in section
9.3.2 could be used to better understand historic urban soils given that it encompasses
time and humans as a key factors in soil formation. It is also recommended that future
soil maps should include urban soils as distinct entities rather than omitting or
misclassifying them, although the outcome will largely be dependant on the soil
classification system used. Another significant contribution of this study is that it
highlights the role of soils in historic towns as unique archives of past human activities.
As discussed in section 9.4, soils based cultural records contain information not readily
available in documentary sources. Furthermore they provide an additional environmental
context to archaeological finds. It is argued that soil and artefacts are equally important,
hence exploration of soils based cultural records should be a future consideration in
urban heritage and conservation strategies. Implementation of this recommendation will
be dependant on the ability of individual archaeologists, planners and conservationists to
recognise the importance of soils as a record of the human past. It is anticipated that the
Soil Analysis Support System for Archaeologists (SASSA) developed at the University of
Stirling will serve as a useful introduction for archaeologists and curators with limited
geoarchaeological knowledge.

- 277 -
 10 Bibliography
Adams HI (1978) The Making of Urban Scotland London, Croom Helm

Aguilar AG and Ward PM (2003) Globalization, regional development, and mega-city

expansion in Latin America: Analyzing Mexico City's peri-urban hinterland Cities 20(1), 3-
21

Alexandrovskaya EI and Alexandrovskiy AL (2000) History of the Cultural Layer in

Moscow and Accumulation of Anthropogenic Substances in it Catena 41, 249-259

Alexandrovskaya E and Panova T (2003) History of the Soil, Cultural Layer, and People in
Medieval Moscow Revista Mexicana De Ciencias Geologicas 20(3), 289-294

Alloway BJ (2004). Contamination of Soils in Domestic Gardens and Allotments: A Brief

Overview Land Contamination and Reclamation 12(3), 179-187

ArcView GIS 3.2 (2002) ArcView Help Copyright © 1992-1996 ESRI, Inc.

Aston MA (1998). The Use of Heavy Metal Soil Analysis for Archaeological Surveying
Chemosphere 37(3), 465-477

Avery BW (1980) Soil Classification for England and Wales Technical Monograph No.14
Rothamsted Experimental Station, Harpenden

Ball DF (1964) Loss-on-ignition as an estimate of organic matter and organic carbon in

non-calcareous soils Journal of Soil Science 15(1), 84-92

Bartington Instruments Ltd (1996) Operation Manual for MS2 Magnetic Susceptibility
System Bartington Instruments Ltd, Oxford

Bellomo RV (1993) A methodological approach for identifying archaeological evidence of

fire resulting from human activities Journal of Archaeological Science 20, 525-553

Berna F, Matthews A and Weiner S (2004) Solubilities of bone mineral from

archaeological sites: the recrystallization window Journal of Archaeological Science
31,867-882

Berna F, Behar A, Shahack-Gross R, Berg J, Boaretto E, Gilboa A, Sharon I, Shalev S,

Shilstein S, Yahalom-Mack N, Zorn JR and Weiner S (2007) Sediments exposed to high
temperatures: reconstructing pyrotechnological processes in Late Bronze Age and Iron
Age Strata at Tel Dor (Isreal) Journal of Archaeological Science 34, 358-373

Bokhorst MP, Duller GAT and Van Mourik JM (2005) Optical dating of a Fimic Anthrosol in
the Southern Netherlands Journal of Archaeological Science 32, 547-553

Bowler DP (2004) Perth, The Archaeology and Development of a Scottish Burgh Perth,
Tayside and Fife Archaeological Committee

Brady NC and Weil RR (1999) Elements of the Nature and Properties of Soils New
Jersey, Prentice Hall

- 278 -
Brewster D (1832) The Edinburgh Encyclopaedia vol. 18 Philadelphia, Joseph Parker

Bridges EM (1991) Waste Materials in Urban Soils In Bullock P and Gregory PJ (editors)
Soils in the Urban Environment Oxford, Blackwell Scientific Publications

Bridges EM (1982) Techniques of modern soil survey In Bridges EM and Davidson DA

(editors) Principles and Applications of Soil Geography London, Longman

Brothwell D (1982) Linking Urban Man with his Urban Environment In Hall AR and
Kenward HK (editors) Environmental Archaeology in the Urban Context CBA Research
Report No 43 London, Council for British Archaeology

Bruce A (1791-99) Parish of Brechin In Sinclair J (editor) Statistical Account of Scotland

Vol 5 Edinburgh, University of Edinburgh

Bryant RG and Davidson DA (1996) The Use of Image Analysis in the Micromorphological
Study of Old Cultivated Soils: an Evaluation Based on Soils from the Island of Papa Stour,
Shetland Journal of Archaeological Science 23, 811-822

Buchanan A (2004) Aerial Image of Lauder Available Online at:

http://www.holy-cow.co.uk/htdocs/album/microlight/album.htm Accessed [18/03/08]

Bui EN (2004) Soil Survey as a Knowledge System Geoderma 120, 17-26

Bull ID, Betancourt PP and Evershed RP (2001) An Organic Geochemical Investigation of

the Practise of Manuring at a Minoan Site on Pseira Island, Crete Geoarchaeology: An
International Journal 16(2), 223-242

Bullock P and Gregory PJ (1991) Soils, A Neglected Resource in Urban Areas In Bullock
P and Gregory PJ (editors) Soils in the Urban Environment Oxford, Blackwell Scientific
Publications.

Cachart R (2000) Excavations at 106-110 South Street, St Andrews Tayside and Fife
Archaeological Journal 6,109-136

Carter S (2001) A Reassessment of the Origins of the St. Andrews 'Garden Soil' Tayside
and Fife Archaeological Journal 7:87-92

Carter S (2004) The Soils and Sediments Holyrood Archaeological Survey Edinburgh,
Headland Archaeology

Chianese D, D’Emilio M, Di Salvia S, Lapenna V, Ragosta M and Rizzo E (2004)

Magnetic mapping, ground penetrating radar surveys and magnetic susceptibility
measurements for the study of the archaeological site of Serra di Vaglio (southern Italy)
Journal of Archaeological Science 31, 633-643

Chirenje T, Ma LQ, Chen M and Zillioux EJ (2003) Comparison between Background

Concentrations of Arsenic in Urban and Non-Urban areas of Florida Advances in
Environmental Research 8, 137-146

Chung Wong SC and Li XD (2004) Pb Contamination and Isotopic Composition of Urban

Soils in Hong Kong The Science of the Total Environment 319, 185-195

- 279 -
Ciezar P, Gonzalez V, Pieters M, Rodet-Belarbi I and Van Ossel P (1994) In suburbano-
new data on the immediate surroundings of Roman and Early medieval Paris In Hall AR
and Kenward HK (editors) Urban-rural connexions: perspectives from environmental
archaeology, Symposia for the Association for Environmental Archaeology No.12 Oxford,
Oxford Monograph 47

Clark PR (1997) 29 North Street In Rains MJ and Hall DW (editors) Excavations in St

Andrews 1980-89 A Decade of Archaeology Glenrothes, Tayside and Fife Archaeological
Committee Monograph 1

Cofie O and Bradford AA (2006) Organic Waste Reuse for Urban Agriculture In
Veenhuizen R (editor) Cities Farming for the Future Urban Agriculture for Green and
Productive Cities International Institute of Rural Reconstruction and ETC Urban
Agriculture, Philippines

Coleman R (1999) The Scottish burgh Survey: An Archaeologists Perspective In

Dennison EP (editor) Conservation and Change in Historic Towns, Research Directions
for the Future, CBA Research Report 122 York, Council for British Archaeology

Coleman R and Smith C (2004) The Archaeology of Burgage Plots in Scottish Medieval
Towns: A Review Proceedings of the Society of Antiquaries of Scotland 134, 281-323

Cook D (1867) Annals of Pittenweem, Being Notes and Extracts from Ancient Records of
that Burgh 1526-1793 Anstruther, Lewis Russell

Cook SR, Clarke AS and Fulford MG (2005) Soil Geochemistry and Detection of Early
Roman Precious Metal and Copper Alloy Working at the Roman Town of Calleva
Atrebatum (Silchester, Hampshire, UK) Journal of Archaeological Science 32, 805-812

Coppack PM (1988) Reflections on the role of amenity in the evolution of urban field
Geografiska Annaler 70(B), 353-361

Cosens P (1834-45) Parish of Lauder Wigton In The New Statistical Account of Scotland
Vol 2 Edinburgh, William Blackwood

Courty MA, Goldberg P and Macphail R (1989) Soils and Micromorphology in

Archaeology Cambridge, Cambridge University Press

Cox A (1997) South Castle Street In Rains MJ and Hall DW (editors) Excavations in St
Andrews 1980-89 A Decade of Archaeology Glenrothes, Tayside and Fife Archaeological
Committee Monograph 1

Croly C (2003) Privies and Other Filthiness... The Environment of Late Medieval
Aberdeen c1399-1650 The Environment of Medieval Aberdeen Project Available Online
at:http://www.aberdeencity.gov.uk/ACCI/web/site/LocalHistory/NC/loc_MA_Historical_Res
earch.asp Accessed [20/02/08]

Crowther J (2003) Potential magnetic susceptibility and fractional conversion studies of

archaeological soils and sediments Archaeometry 45(4), 685-701

Crowther J and Barker P (1995) Magnetic Susceptibility: Distinguishing Anthropogenic

Effects from the Natural Archaeological Prospectation 2, 207-215

- 280 -
Daniels M (2001-2003) Random Number Generator Available Online at:
http://www.mdani.demon.co.uk/para/random.htm Accessed [08/12/07]

Darvill T (2002) The Concise Oxford Dictionary of Archaeology Oxford Reference Online,
Oxford University Press

Davidson DA and Carter SP (1998) Micromorphological Evidence of Past Agricultural

Practises in Cultivated Soils: The Impact of a Traditional Agricultural System on Soils in
Papa Stour, Shetland Journal of Archaeological Science 25, 827-838

Davidson DA and Simpson IA (2001) Archaeology and Soil Micromorphology In Brothwell

DR and Pollard AM (editors) Handbook of Archaeological Sciences Chichester, John
Wiley and Sons Ltd

Davidson DA, Dercon G, Stewart M and Watson F (2006) The Legacy of Past Urban
Waste Disposal on Local Soils Journal of Archaeological Science 33(6), 778-783

Davidson DA, Carter SP and Quine TA (1992) An Evaluation of micromorphology as an

aid to archaeological interpretation Geoarchaeology: An International Journal 7(1), 55-65

De Kimpe CR and Morel J (2000) Urban Soil Management: A Growing Concern Soil
Science 165(1), 31-40

De Vries J (1984) European Urbanization 1500-1800 London, Methuen

Dearing J (1999) Magnetic Susceptibility In Walden J, Oldfield F and Smith J (editors)

Environmental Magnetism: A Practical Guide Technical Guide No. 6 London, Quaternary
Research Association

Dent D and Young A (1981) Soil Survey and Land Evaluation London, Allen and Unwin

Dennison EP (1999a) The Scottish Burgh Survey and the Centre for Scottish Urban
History In Dennison EP (editor) Conservation and Change in Historic Towns, Research
Directions for the Future CBA Research Report 122 York, Council for British Archaeology

Dennison EP (1999b) Europe's historic towns: a future for their past? In Dennsion EP
(editor) Conservation and Change in Historic Towns, Research Directions for the Future
CBA Research Report 122 York, Council for British Archaeology

Dercon G, Davidson DA, Dalsgaard K, Simpson IA, Spek T and Thomas J (2005)
Formation of Sandy Anthropogenic Soils in NW Europe Identification of Inputs based on
Particle Size Distribution Catena 59,341-356

Devine TM (2000) The Scottish Nation 1700-2000 London, Penguin Books

Dewar I, Batt CM and Peters C (2002) A mineral magnetic investigation into fuel derived
deposits from Old Scatness Broch, Shetland Physics and Chemistry of the Earth 27,1343-
1348

Dicks B (1983) The Scottish Medieval Town A Search for Origins In George G and Dicks
B (editors) Scottish Urban History Aberdeen, Aberdeen University Press

- 281 -
Dillon WJ (1953) The Streets in Early Times In Dunlop AI (editor) The Royal Burgh of Ayr,
Seven Hundred and Fifty Years of History Edinburgh, Oliver and Boyd

Dudal R (2004) The Sixth Factor of Soil Formation International Conference on Soil
Classification Petrozavodsk, Russia 3-4th August 2004 Available Online at:
http://www.itc.nl/~rossiter/research/suitma/Dudal_6thFactor.pdf Accessed [04/02/08]

Duncan A (1791-99) Parish of Wigton In Sinclair J (editor) Statistical Account of Scotland

Vol 14 Edinburgh, University of Edinburgh

Edwards P (2004) The Garbage Crisis in Prehistory: artefact discard patterns of the early
Natufian site of Wadi Hammeh 27 and the origin of household refuse strategies Journal of
Anthropological Archaeology 23, 253-289

Entwistle JA, Abrahams PW and Dodgshon RA (1998) Multi-Element Analysis of Soils

from Scottish Historical Sites, Interpreting Land-Use History through Physical and
Geochemical Analysis of Soil Journal of Archaeological Science 25, 53-68

Entwistle JA, Abrahams PW and Dodgshon RA (2000a) The Geoarchaeological

Significance and Spatial Variability of a Range of Physical and Chemical Soil Properties
from a Former Habitation Site, Isle of Skye Journal of Archaeological Science 27, 287-303

Entwistle JA, Dodgson RA and Abrahams PW (2000b) An investigation of Former Land-

use Activity through the Physical and Chemical Analysis of Soils from the Isle of Lewis,
Outer Hebrides Archaeological Prospection 7,171-188

Evans B (1999) Millennial Burghs: Scotland’s Towns in the Future In Dennison EP (editor)
Conservation and Change in Historic Towns, Research Directions for the Future CBA
Research Report 122 York, Council for British Archaeology

Ewan E (1990) Townlife in Fourteenth Century Scotland Edinburgh, Edinburgh University

Press

Eydmann S (1999) Scotland’s Historic Towns: A Conservation Officer Reflects In

Dennison EP (editor) Conservation and Change in Historic Towns, Research Directions
for the Future CBA Research Report 122 York, Council for British Archaeology

FAO (2006) World Reference Base for Soil Resources 2006 World Soil Resources
Reports 103 Rome, Food and Agriculture Organization of the United Nations

Feathers JK (2003) Use of Luminescence dating in archaeology Measurement Science

and Technology 14, 1493-1509

Fekade T (1995) The role of small urban centres in the development of pastoral regions:
The case of Negele Town, Borena region, Southern Ethiopia Regional Development
Dialogue 16(2), 85-96

Ford J (1791-99) Parish of Lauder In Sinclair J (editor) Statistical Account of Scotland Vol
1 Edinburgh, University of Edinburgh

French CAI (2003) Geoarchaeology in action: Studies in soil micromorphology and

landscape evolution London, Routledge

- 282 -
French CAI (1999) Soil Erosion, Agricultural Terracing and Site Formation Processes at
Markiani, Amorgos, Greece: The Micromorphological Perspective Geoarchaeology: An
International Journal 14(2), 151-189

Garcia-Alonso S, Perez-Pastor RM and Sevillano-Castano ML (2003) Occurrence of

PCBs and PAHs in an Urban Soil of Madrid (Spain) Toxicology and Environmental
Chemistry 85(4-6), 193-202

General Register Office for Scotland (2007) Scotland's Census Results Online
http://www.scrol.gov.uk/scrol/common/home.jsp Accessed [23/08/2007]

Goldberg P and MacPhail RI (2006) Practical and Theoretical Geoarchaeology Oxford,

Blackwell Publishing

Golding K and Davidson DA (2005) The Effect of Past Waste Disposal on Soils Near to
Scottish Burghs Journal of the South East England Soils Discussion Group 16, 28-37

Gong ZT (1999) Chinese Soil Taxonomy: Theories, Methods, Practice Science Press,
Beijing

Gould WTS (1987) Urban bias, regional differentiation and rural-urban interaction in
Kenya African Urban Quarterly 2(2), 122-133

Griffith MA (1980). A Pedological Investigation of an Archaeological Site in Ontario,

Canada, I An Examination of the Soils in and Adjacent to a former Village Geoderma 24,
327-336

Grϋn R (2001) Trapped Charge Dating (ESR, TL, OSL) In Brothwell DR and Pollard AM
(editors) Handbook of Archaeological Sciences Chichester, John Wiley and Sons Ltd

Hall D (2002) Burgess, Merchant and Priest Edinburgh, Birlinn with Historic Scotland

Hall D and Bowler D (1997) North Berwick, East Lothian: Its Archaeology Revisited
Proceedings of the Society of Antiquaries of Scotland 127, 659-675

Hall DW (1997) 134 and120-4 Market Street In Rains MJ and Hall DW (editors)
Excavations in St Andrews 1980-89 A Decade of Archaeology Glenrothes, Tayside and
Fife Archaeological Committee Monograph 1

Hall DW, MacDonald ADS, Perry DR and Terry J (1998) The Archaeology of Elgin:
excavations on Ladyhill and in the High Street, with an overview of the archaeology of the
burgh Proceedings of the Society of Antiquaries of Scotland 128, 753-829

Hamilton H (1945) Selections from the Monymusk Papers (1713-1755) Edinburgh, T and
A Constable Ltd for the Scottish History Society

Harpet C (2003) From garbage dumps, Anthropology to an interdisciplinary research on

health risk exposure Natures Science Societes 11(2), 361-370

Hayden B and Cannon A (1983) Where the Garbage Goes: Refuse Disposal in the Maya
Highlands Journal of Anthropological Archaeology 2, 117-163

- 283 -
Hodgson JM (1974) Soil Survey Field Handbook Harpenden, Technical Monograph No. 5
Soil Survey of England and Wales

Holliday VT and Gartner WG (2007) Methods of soil P analysis in archaeology Journal of

Archaeological Science 34, 301-333

Hollis JM (1991) The Classification of Soils in Urban Areas In Bullock P and Gregory PJ
(editors) Soils in the Urban Environment Oxford, Blackwell Scientific Publications

Homsey LK and capo RC (2006) Integrating Geochemistry and Micromorphology to

Interpret Feature Use at Dust Cave, a Paleo-Indian Through Middle-Archaic Site in
Northwest Alabama Geoarchaeology: An International Journal 21(3), 237-269

Horsburgh JA (1865) A Glance at the Historical Traditions of Pittenweem During the Last
Two Centuries Pittenweem, James Scott

http://www.thin.stir.ac.uk/methods.html (2007) Methods of Thin Section Production,

Micromorphology Laboratory, University of Stirling, Accessed [19/11/07]

Hough RL (2007) Soil and human health: an epidemiological review European Journal of
Soil Science 58, 1200-1212

Hursthouse A, Tognarelli D, Tucker P, Marsan FA, Martini C, Madrid L, Madrid F and

Diaz-Barrientos E (2004) Metal Content of Surface Soils in Parks and Allotments from
three European Cities: Initial Pilot Study Results Land Contamination and Reclamation
12(3), 189-196

Isbell R (2002) The Australian Soil Classification, Australian Soil and Land Survey
Handbooks Series Volume 4, CSRIO Publishing, Australia Available online at
http://www.clw.csiro.au/aclep/asc_re_on_line/soilhome.htm, [Accessed 20/03/08]

James P (1999) Soil Variability in the Area of an Archaeological Site near Sparta, Greece
Journal of Archaeological Science 26, 1273-1288

Johnston T (1920) The Reviving of the Common Lands In The History of the Working
Classes in Scotland Available online at www.caledonia.org.uk/commonweal, [Accessed
11/03/08]

Jordanova N, Petrovsky E, Kovacheva M and Jordanova D (2001) Factors Determining

Magnetic Enhancement of Burnt Clay from Archaeological Sites Journal of Archaeological
Science 28, 1137-1148

Kaida Y and Maharjan KL (1990) Rural-Urban Interaction and Its Implication for Rural
Development in Bangladesh Southeast Asian Studies 28(3), 403-416

Kelly J, Thornton I and Simpson PR (1996) Urban Geochemistry: A Study of the Influence
of Anthropogenic Activity on the Heavy Metal Content of Soils in Traditionally Industrial
and Non-industrial areas of Britain Applied Geochemistry 11, 363-370

Kemp RA (1998) Role of Micromorphology in Paleopedological Research Quaternary

International 51/52, 133-141

- 284 -
King W (1992) How High is Too High? Disposing of Dung in Seventeenth-Century
Prescott Sixteenth Century Journal 23(3), 445-457

Kooistra MJ and Kooistra LI (2003) Integrated research in archaeology using soil

micromorphology and palynology Catena 54, 603-617

LaMotta VM and Schiffer MB (2001) Behavioural Archaeology: Towards a New Synthesis

In Hodder I (editor) Archaeological Theory Today Cambridge, Polity Press

Leighton JM (1840) History of the County of Fife from the Earliest Period to the Present
Time Glasgow, J Swan

Leonardi G, Miglavacca M and Nardi S (1999) Soil Phosphorus Analysis as an Integrative

Tool for Recognizing Buried Ancient Ploughsoils Journal of Archaeological Science 26,
343-352

Li X, Lee S, Wong S, Shi W and Thornton I (2004) The Study of Metal Contamination in
urban Soils of Hong Kong using a GIS-based Approach Environmental Pollution 129, 113-
124

Lin GCS (2002) Hong Kong and the globalisation of the Chinese Diaspora: a geographical
perspective Asia Pacific Viewpoint 43(1),63-91

Lin GCS (2001) Evolving Spatial Form of Urban-Rural Interaction in the Pearl River Delta,
China Professional Geographer 53(1), 56-70

Liu W, Zhao J, Ouyang Z, Soderlund L and Liu G (2005) Impacts of sewage irrigation on
heavy metal distribution and contamination in Beijing, China Environment International 31:
805-812

Lowe CE (2001) The Distribution and Survival of Medieval and Later Deposits in the
Medieval Burgh of Crail, Fife Tayside and Fife Archaeological Journal 7, 93-111

Lu Y, Gong Z, Zhang G and Burghardt W (2003) Concentrations and Chemical

Speciations of Cu, Zn, Pb and Cr of Urban Soils in Nanjing, China Geoderma 115, 101-
111

Lynch M (1999) Towns and Townspeople: Prospects for the Future In Dennison EP
(editor) Conservation and Change in Historic Towns, Research Directions for the Future
CBA Research Report 122 York, Council for British Archaeology

Macphail RI (2003) Soil Microstratigraphy: a micromorphological and chemical approach

In Cowan C (editor) Urban development in north-west Roman Southwark: excavations
1974-1990 London, MoLAS Monograph Series 16

Macphail RI (1998) A Reply to Carter and Davidson’s “An Evaluation of the Contribution of
Soil Micromorphology to the Study of Ancient Arable Agriculture” Geoarchaeology: An
International Journal 13(6), 549-564

Macphail RI (1994) The reworking of urban stratigraphy by human and natural processes
In Hall AR and Kenward HK (editors) Urban-rural connexions: perspectives from
environmental archaeology, Symposia for the Association for Environmental Archaeology
No.12 Oxford, Oxbow Monograph 47

- 285 -
Macphail RI, Galinie H and Verhaeghe F (2003) A Future for Dark Earth? Antiquity
77(296), 350-358

Macphail RI and Goldberg P (1995) Recent advances in micromorphological

interpretations of soils and sediments from archaeological sites In Barham AJ and
Macphail RI (editors) Archaeological sediments and soils: analysis, interpretation and
management: papers from the tenth anniversary conference of the Association for
Environmental Archaeology held at the Institute of Archaeology, UCL, July 1989 London,
Institute of Archaeology, UCL

Mair C (1988) Mercat Cross and Tolbooth: Understanding Scotland's Old Burghs
Edinburgh, John Donald Publishers Ltd

Manta DS, Angelone M, Bellanca A, Neri R and Sprovieri M (2002) Heavy Metals in Urban
Soils: A Case Study from the City of Palermo (Sicily), Italy The Science of the Total
Environment 300, 229-243

Matthews W (1995) Micromorphological characterisation and interpretation of occupation

deposits and microstratigraphic sequences at Abu Salabikh, Southern Iraq In Barham AJ
and MacPhail RI (editors) Archaeological sediments and soils: analysis, interpretation and
management: papers from the tenth anniversary conference of the Association for
Environmental Archaeology held at the Institute of Archaeology, UCL, July 1989, London,
Institute of Archaeology, UCL

Matthews W, French CAI, Lawrence T, Cutler DF and Jones MK (1997) Microstratigraphic

traces of site formation processes and human activities World Archaeology 29(2), 281-208

Matthiesen H (2004) In situ measurement of soil pH Journal of Archaeological Science

31,1373-1381

McAlpine W and Rolland R (1791-99) Parish of Culross In Sinclair J (editor) Statistical

Account of Scotland Vol 10 Edinburgh, University of Edinburgh

McRae SG (1988) Practical Pedology Studying Soils in the Field Chichester, Ellis
Harwood Ltd

Mielke HW, Wang G, Gonzales CR, Le B; Quanch VN and Mielke PW (2001) PAH and
Metal Mixtures in New Orleans Soils and Sediments The Science of the Total
Environment 281, 217-227

Moller A, Muller HW, Abdullah A, Abdelgawad G and Utermann J (2006) Urban Soil
Pollution in Damascus, Syria: Concentrations and Patterns of heavy Metals in the Soils of
the Damascus Ghouta Geoderma 124, 63-71

Molleson A (1791-99) Town and Parish of Montrose In Sinclair J (editor) Statistical

Account of Scotland Vol 5 Edinburgh, University of Edinburgh

Munsell Color ® (1994) Munsell ® Soil Color Charts Revised Edition New York

Nairne J (1791-99) Parish of Pittenweem (County of Fife) In Sinclair J (editor) Statistical

Account of Scotland Vol 4 Edinburgh, University of Edinburgh

- 286 -
Needham S and Spence T (1997) Refuse and the formation of middens Antiquity 71, 77-
90

Nikolaevna SM and Vadimovna PT (2003) City, Soil and Environment, Classification of

urban soils located within the taiga zone of Russia Available online at
http://www.soil.msu.ru/~ptv/city-eng.htm, Accessed [20/03/08]

Nivalainen S and Schmidt-Thome K (2003) Demographic development and the emerging

discourse on urban-rural interaction in Finland Geographia Polonica 76(1), 127-145

Ogilvy T (1791-99) Parish of Kirriemuir In Sinclair J (editor) Statistical Account of Scotland

Vol 12 Edinburgh, University of Edinburgh

Oldfield F (1999) Introduction In Walden J, Oldfield F and Smith J (editors) Environmental

Magnetism: A Practical Guide Technical Guide No. 6 London, Quaternary Research
Association

Ona LF, Alberto AMP, Rrudente JA and Sigua GC (2006) Levels of Lead in Urban Soils
from Selected Cities in a Central Region of the Philippines Environmental Science and
Pollution Research International 13(3), 177-183

Oram R (2006) Aspects of the Environmental History of Elgin c1150-1850

Ottaway P (1992) Archaeology in British Towns from the Emperor Claudius to the Black
Death London and New York, Routledge

Owen O (1999) Protecting the Archaeology of Scotland’s Historic Burghs In Dennison EP

(editor) Conservation and Change in Historic Towns, Research Directions for the Future
CBA Research Report 122 York, Council for British Archaeology

Parker DC, Manson SM, Jannssen, Janssen MA, Hoffmann MJ and Deadman P (2003)
Multi-Agent Systems for the Simulation of Land-Use and Land-Cover Change: A Review
Annals of the Association of American Geographers 93(2), 314-337

Pessel WK (2006) Rubbish as informants: a cultural contribution to Polish 'garbeology'

Anthropology Matters Journal 8(1) Available Online at:
http://www.anthropologymatters.com/journal/2006-1/index.htm, Accessed [10/06/07]

Peters C, Church MJ and Coles G (2000) Mineral Magnetism and Archaeology at Galson
on the Isle of Lewis, Scotland Physics and Chemistry of the earth (A) 25(5), 455-460

Peters C, Church MJ and Mitchell C (2001) Investigation of Fire Ash Residues Using
Mineral Magnetism Archaeological Prospectation 8,227-237

Pryde GCS (1965) The Burghs of Scotland A Critical List London, Oxford University Press

Rains MJ and Hall DW (editors) (1997) Excavations in St Andrews 1980-89 A Decade of

Archaeology Glenrothes, Tayside and Fife Archaeological Committee Monograph 1

Rainy G (1791-99) Parish of Criech In Sinclair J (editor) Statistical Account of Scotland

Vol 8 Edinburgh, University of Edinburgh

- 287 -
Rapp G Jr and Hill CL (1998) Geoarchaeology The Earth Science Approach to
Archaeological Interpretation London, Yale University Press

Rathje WL (1974) The Garbage Project: A New Way of Looking at the Problems of
Archaeology Archaeology 27, 236-241

Rathje WL and Murphy C (2001a) Rubbish: The Archaeology of Garbage Tucson,

University of Arizona Press

Rathje WL and Murphy C (2001b) Garbage Demographics In Whitten P (editor)

Anthropology: Contemporary Perspectives 8th edition London, Allyn and bacon

Reid A (1909) The Regality of Kirriemuir Edinburgh, Grant

Renfrew C and Bahn P (1996) Archaeology Theories, Methods and Practise 2nd edition
London, Thames and Hudson

Robertson H (1791-99) Parish of Kiltearn In Sinclair J (editor) Statistical Account of

Scotland Vol 1 Edinburgh, University of Edinburgh

Robertson JC (1990) Moving on from Holes and Corners: Recent Currents in Urban
Archaeology Urban History Yearbook 20, 1-13

Robinson SG (1997) Beginning Environmental Magnetism: An Introduction to the Theory

and Techniques of Applied Rock-Magnetism in Environmental Studies Manchester,
School of Geography, University of Manchester

Romanes JH (1914) The Village Economy of Lauder The English Historical Review 29
(155), 532-537

Ross L and Clark PR (1997) Cinema House In Rains MJ and Hall DW (editors)
Excavations in St Andrews 1980-89 A Decade of Archaeology Glenrothes, Tayside and
Fife Archaeological Committee Monograph 1

Rudeforth CC (1982) Handling Soil Survey Data In Bridges EM and Davidson DA (editors)
Principles and Applications of Soil Geography London, Longman

SASSA (2007) Soil Analysis Support System for Archaeologists Available Online at:
http://www.sassa.org.uk/index.php/Main_Page, Accessed [12/09/08]

Scharenbroch BC, Lloyd JE and Johnson-Maynard JL (2005) Distinguishing Urban Soils

with Physical, Chemical and Biological Properties Pedobiologia 49, 283-296

Schia E (1994) Urban Oslo and its relation to rural production in the hinterland-An
archaeological view In Hall AR and Kenward HK (editors) Urban-rural connexions:
perspectives from environmental archaeology, Symposia for the Association for
Environmental Archaeology No.12 Oxford, Oxbow Monograph 47

Schiffer MB (1976) Behavioural Archaeology Tucson, University of Arizona

Schiffer MB (1972) Archaeological and Systemic Context American Antiquity 37,156-165

- 288 -
Schlezinger DR and Howes BL (2000) Organic Phosphorus and Elemental Ratios as
Indicators of Prehistoric Human Occupation, 479-492 Journal of Archaeological Science
27, 479-492

Schofield J (1987) Recent Approaches in Urban Archaeology In Schofield J and Leech R

(editors) Urban Archaeology in Britain, CBA Research Report 61 London, Council for
British Archaeology

Senesi GS, Baldassarre G, Senesi N and Radina B (1999) Trace Element Inputs into
Soils by Anthropogenic Activites and Implications for Human Health Chemosphere 39(2),
343-377

Shahack-Gross R, Albert RM, Gilboa A, Nagar-Hilman O, Sharon I and Weiner S (2005)

Geoarchaeology in an Urban Context: the Uses of Space in a Phoenician Monumental
Building at Tel Dor (Israel) Journal of Archaeological Science 32, 1417-1431

Simpson AT and Stevenson S (1981a) Historic Pittenweem: The Archaeological

Implications of Development Glasgow, Scottish Burgh Survey

Simpson AT and Stevenson S (1981b) Historic Wigtown: The Archaeological Implications

of Development Glasgow University, Scottish Burgh Survey

Simpson AT and Stevenson S (1981c) Historic Lauder: The Archaeological Implications of

Development Glasgow University, Scottish Burgh Survey

Simpson IA (1997) Relict Properties of Anthropogenic Deep Top Soils as Indicators of

Infield Management in Marwick, West Mainland, Orkney Journal of Archaeological
Science 24,365-380

Simpson IA, Guttmann EB, Cluett J and Shepherd A (2006) Characterizing Anthropic
Sediments in North European Neolithic Settlements: An Assessment from Skara Brae,
Orkney Geoarchaeology: An International Journal 21(3), 221-235

Simpson IA, Barrett JH and Milek KB (2005) Interpreting the Viking Age to Medieval
Period Transition in Norse Orkney through Cultural Soil and Sediment Analyses
Geoarchaeology: An International Journal 20(4), 355-377

Simpson IA, Vẻsteinsson O, Adderley WP and McGovern TH (2003) Fuel resource

utilisation in landscapes of settlement Journal of Archaeological Science 30, 1401-1420

Simpson IA, Dockrill SJ, Bull ID and Evershed RP (1998a) Early Anthropogenic Soil
Formation at Tofts Ness, Sanday, Orkney Journal of Archaeological Science 25, 729-746

Simpson IA, Bryant RG and Tveraabak U (1998b) Relict Soils and Early Arable Land
Management in Lofoten, Norway Journal of Archaeological Science 25, 1185-1198

Smith J (1999) An introduction to the magnetic properties of natural materials In Walden

J, Oldfield F and Smith J (editors) Environmental Magnetism: A Practical Guide Technical
Guide No. 6 London, Quaternary Research Association

Smout TC (2000) Nature Contested, Environmental History in Scotland and Northern

England Since 1600 Edinburgh, Edinburgh University Press

- 289 -
Soanes C and Stevenson A (2006) The Concise Oxford English Dictionary 11th edition
Oxford Reference Online, Oxford University Press

Soil Survey of Scotland (1984) Organization and Methods of the 1:250 000 Soil Survey of
Scotland Aberdeen, The Macaulay Institute for Soil Research

Soil Survey of Scotland (1975) Soil Survey of Scotland Systematic Soil Survey: Kinross,
Elie and Edinburgh sheet 40 and parts of sheet 41 and 31 (soil) Aberdeen, The Macaulay
Institute for Soil Research

Soil Survey of Scotland (1971) Soil Survey of Scotland Systematic Soil Survey:
Kirkmaiden, Whithorn, Stranraer and Wigtown sheets 1,2,3,4 and parts of sheet 7 (soil)
Aberdeen, The Macaulay Institute for Soil Research

Soil Survey of Scotland (1959) Soil Survey of Scotland: Kelso sheet 25 Aberdeen, The
Macaulay Institute for Soil Research

Soil Survey Staff (1999) Soil Taxonomy: A Basic System of Soil Classification for Making
and Interpreting Soil Surveys Handbook No. 436 (2nd edition) Washington DC, US
Department of Agriculture

Sombroek W, Dirse K, Rodrigues T, Cravo MDS, Jarbas TC, Woods W and Glaser B
(2002) Terra Preta and Terra Mulata: pre-Columbian Amazonian Kitchen Middens and
Agricultural Fields, their Sustainability and their Replication 17th World Congress of Soil
Science Symposium 18 Paper 1935 Thailand.

Sommerville AA, Sanderson DCW, Hansom JD and Housley RA (2001) Luminescence

dating of Aeolian sands from archaeological sites in Northern Britain: a preliminary study
Quaternary Science Reviews 20, 913-919

Spearman RM (1988) Workshops, Materials and Debris- Evidence of Early Industries In

Lynch M, Spearman M and Stell G (editors) The Scottish Medieval Town Edinburgh, John
Donald Publishers Ltd

Spearman RM (1982) Excavation at 17-21 Castle Street, Forfar, 1979 Proceedings of the
Society of Antiquaries of Scotland 112, 455-464

Starn R (2002) The Early Modern Muddle Journal of Early Modern History 6(3), 296-307

Stones J (1987) A Tale of Two Burghs Aberdeen, Aberdeen Art Gallery and Museums

Stoops G (2003) Guidelines for Analysis and Description of Soil and Regolith Thin
Sections Wisconsin, Soil Science Society of America, Inc

Tacoli C (1998) Rural-urban interactions: a guide to the literature Environment and

Urbanisation 10(1), 147-166

Tan KH (1996) Soil Sampling, Preparation, and Analysis New York, Basel and Hong
Kong, Marcel Dekker Inc

Tanner T (2003) Peopling mountain environments: changing Andean livelihoods in north-

west Argentina The Geographical Journal 169(3), 205-214

- 290 -
Taylor RE (2001) Radiocarbon Dating In Brothwell DR and Pollard AM (editors) Handbook
of Archaeological Science Chichester, John Wiley and Sons Ltd

Terry RE, Fernandez FG, Parnell JJ and Inomata T (2004) The Story in the Floors:
Chemical Signatures of Ancient and Modern Maya Activities at Aguateca, Guatemala
Journal of Archaeological Science 31, 1237-1250

Thompson M (2003) Time's Square: Deriving Cultural Theory from Rubbish Theory
Innovation 16(4), 319-330

Thompson M (1979) Rubbish Theory Oxford, Oxford University Press

Tijhuis L, Brattli B and Saether OM (2002) A Geochemical Survey of topsoil in the City of
Oslo, Norway Environmental Geochemistry and Health 24, 67-94

Tixier P and Bon H (2006) Urban Horticulture In Veenhuizen R (editor) Cities Farming for
the Future Urban Agriculture for Green and Productive Cities International Institute of
Rural Reconstruction and ETC Urban Agriculture, Philippines

Tobert N (2007) Ethno-archaeology of pottery firing in Darfur, Sudan: Implications for

ceramic technology studies Oxford Journal of Archaeology 3(2), 141-156

Turner-Simpson A and Stevenson S (1980) Historic Lauder: the archaeological

implications of development Edinburgh, Scottish Burgh Survey

VandenBygaart AJ and Protz R (1999) The representative area (REA) in studies of

quantitative soil micromorphology Geoderma 89, 333-346

van Haaster H (1994) Plant resources and environment in late medieval Lϋbeck In Hall
AR and Kenward HK (editors) Urban-rural connexions: perspectives from environmental
archaeology, Symposia for the Association for Environmental Archaeology No.12 Oxford,
Oxbow Monograph 47

Walden J, Oldfield F and Smith J (1999) Environmental Magnetism: a practical guide

Cambridge, Quaternary Research Association

Webster D (1819) A Topographical Dictionary of Scotland Containing a General

Description of That Kingdom with a Particular Account of the Counties, Parishes, Islands,
Lakes, Rivers, Mountains, Valleys, Cities, Towns, Villages, Objects of Natural History,
Scenery of that Country, Mineralogy, Antiquities together with an Account of the Weights
and Measures of each County, the revenues of Towns, and of all the Religious Houses at
the Time of the Reformation Edinburgh, Baldwin, Craddock and Joy

Wells EC (2004) Investigating Activity Patterns in Pre-Hispanic Plazas: Weak Acid-

Extraction ICP-AES Analysis of Anthrosols at Classic Period El Coyote, North western
Honduras Archaeometry 46(1), 67-84

Wells EC, Terry RE, Parnell JJ, Hardin PJ, Jackson MW and Houston SD (2000)
Chemical Analyses of Ancient Anthrosols in Residential Areas at Piedras Negras,
Guatemala Journal of Archaeological Science 27, 449-462

Wheater CP and Cook PA (2000) Using Statistics to Understand the Environment London,
Routledge

- 291 -
Whyte ID (1999) The Historical Geography of Britain from AD 1500 In Hunter J and
Ralston I (eds) The Archaeology of Britain An Introduction from the Upper Palaeolithic to
the Industrial Revolution London, Routledge

Whyte ID (1997) Scotland’s Society and Economy in Transition 1500-1760 London,

Macmillan Press Ltd

Whyte ID (1989) Urbanization in Early-Modern Scotland: A Preliminary Analysis Scottish

Economic and Social History 9(1), 21-37

Whyte ID and Whyte K (1991) The Changing Scottish Landscape 1500-1800 London and
New York, Routledge

Wilson B (1994) Mortality patterns, animal husbandry and marketing in and around
medieval and post-medieval Oxford In Hall AR and Kenward HK (editors) Urban-rural
connexions: perspectives from environmental archaeology, Symposia for the Association
for Environmental Archaeology No.12 Oxford, Oxbow Monograph 47

Wilson CA, Cresser M and Davidson DA (2008) Multi-element soil analysis: an

assessment of its potential as an aid to archaeological interpretation Journal of
Archaeological Science 35, 412-424

Wilson CA, Cresser M and Davidson (2006a) Testing the Use of Multi-Element Soil
Contamination Data in Environmental Archaeology Proceedings of the 1st International
Conference of Environmental Science and Technology January 2005, New Orleans

Wilson CA, Cresser MS and Davidson DA (2006b) Sequential element extraction of soils
from abandoned farms: an investigation of the partitioning of anthropogenic element
inputs from historic land use Journal of Environmental Monitoring 8, 439-444

Wilson CA, Davidson DA and Cresser MS (2005) An evaluation of multielement analysis

of historic soil contamination to differentiate space use and former function in and around
abandoned farms The Holocene 15(7), 1094-1099

Wordsworth J and Clark PR (1997) Auction Hall In Rains MJ and Hall DW (eds)
Excavations in St Andrews 1980-89 A Decade of Archaeology Glenrothes, Tayside and
Fife Archaeological Committee Monograph 1

Yeoman P (1995) Medieval Scotland Edinburgh, Historic Scotland

Yongming H, Peixuan, Junji C and Posmentire ES (2006) Multivariate Analysis of Heavy

Metal Contamination in Urban Dusts of X’ian, Central China The Science of the Total
Environment 355, 176-186

- 292 -
 Appendix 1: Soil Profile Descriptions

Soil profile descriptions and field sketches are presented for soil pits at Lauder,
Pittenweem and Wigtown.

Lauder

High Street Zone

Profile description and photographs (knife inserted at 25cm) of soil pit LA 1

- 293 -
Profile description and photographs (knife inserted at 30cm) of soil pit LA 4

- 294 -
Profile description and photographs (knife inserted at 15cm) of soil pit LA 6

- 295 -
Hinterland Near Zone

Profile description and photographs (knife inserted at 25cm) of soil pit LA 9

- 296 -
Hinterland Far Zone

Profile description and photographs (knife inserted at 30cm) of soil pit LA 2

- 297 -
Profile description and photographs (knife inserted at 25cm) of soil pit LA 3

- 298 -
Thirlstane Zone

Profile description and photographs of soil pit LA 5

- 299 -
‘Reference’ Soil Profile

Profile description and photographs (knife inserted at 15cm) of soil pit LA 7

- 300 -
Pittenweem

Harbour Zone

Profile description and photographs (knife inserted at 50cm) of soil pit PT 4

- 301 -
Profile description and photographs (knife inserted at 50cm) of soil pit PT 5

- 302 -
High Street Zone

Profile description and photographs (knife inserted at 30cm) of soil pit PT 3

- 303 -
Profile description and photographs (knife inserted at 50cm) of soil pit PT 6

- 304 -
Hinterland Near Zone

Profile description and photographs (knife inserted at 15cm) of soil pit PT 9

- 305 -
Hinterland Far Zone

Profile description and photographs (knife inserted at 30cm) of soil pit PT 7

- 306 -
Profile description and photographs of soil pit PT 8

- 307 -
‘Reference’ Soil Profile

Profile description and photographs (knife inserted at 22cm) of soil pit PT 1

- 308 -
Wigtown

High Street Zone

Profile description and photographs (knife inserted at 30cm) of soil pit WG 1

- 309 -
Profile description and photographs (knife inserted at 30cm) of soil pit WG 2

- 310 -
Profile description and photographs (knife inserted at 30cm) of soil pit WG 3

- 311 -
Profile description and photographs (knife inserted at 30cm) of soil pit WG 4

- 312 -
‘Reference’ Soil Profile

Profile description and photographs of soil pit WG 5

- 313 -
 Appendix 2: Micromorphology Summary Tables

Thin section micromorphology descriptions of topsoil deposits in Lauder, percentage abundance estimates of coarse mineral and organic
anthropogenic material and pedofeatures are simplified as; <2% t (trace), 2-5% z, 5-10% zz and 10-15% zzz

- 314 -
Thin section micromorphology descriptions of topsoil deposits in Pittenweem, percentage abundance estimates of coarse mineral and organic
anthropogenic material and pedofeatures are simplified as; <2% t (trace), 2-5% z, 5-10% zz and 10-15% zzz

- 315 -
Thin section micromorphology descriptions of topsoil deposits in Wigtown, percentage abundance estimates of coarse mineral and organic
anthropogenic material and pedofeatures are simplified as; <2% t (trace), 2-5% z, 5-10% zz and 10-15% zzz

- 316 -
Appendix 3: Watching Brief at High Vennel, Wigtown
© SUAT Archaeology

- 317 -
 Appendix 4: Archaeological Evaluation of 64a High Street,
Pittenweem © Scotia Archaeology

The following report is an archaeological evaluation of 64a High Street Pittenweem

conducted by Scotia Archaeology, 11/11/04.

Introduction
This report describes the results of an archaeological evaluation undertaken prior to the
construction of a new domestic residence to the immediate east of 64 High Street,
Pittenweem, Fife. The evaluation was a condition attached by Fife Council to ensure that
any surviving archaeological record was taken into consideration during the proposed
development. It was understood that should features or deposits of archaeological
significance be uncovered during the evaluation a further phase of mitigation fieldwork
might be required by the planning authority. The field evaluation was carried out on 9th
November 2004 by John Lewis of Scotia Archaeology, on behalf of Mr James Martin.
Before fieldwork commenced, a desk assessment of the site was undertaken by John
Terry, also of Scotia Archaeology.

The Site
The site, which was centred on NO 5468 0248, lay at the western limit of the medieval
burgh and comprised a triangular area of ground measuring approximately 22m
east/west by a maximum 15m north/south. It was defined on its west by 64 High Street,
its north by a stone wall separating it from the High Street and its south-east by a wall on
West Wynd, a footpath leading down to the shore. The solid geology of the area consists
of calciferous sandstone measures of Carboniferous age, over which lie late glacial drift
deposits of alluvium, mostly sand and clay. At the time of the evaluation, the site was
covered with rough grass and other vegetation.

The Desk Assessment

Pittenweem is first mentioned in written records dating to c1143 when David I granted
Petenweme and Inverrin (St Monans) to the Priory of May (Lawrie 1905, 120) although it
did not achieve burgh status until 1526 (Pryde 1965, 57). The layout of Pittenweem’s
centre has changed little since the 16th century, many of the burgage plots still being
visible (Simpson & Stevenson 1981, 9). Recent developments have concentrated on the
east and west extremities of the village, beyond the limits of the medieval burgh.

- 318 -
Archaeological investigations in Pittenweem have been limited to excavation within the
garden of 5 Midshore (NO 548 025) where a 17th-century midden, containing Dutch,
French and local wares (DES 1978, 7), overlay a stone-built wynd and associated
structures believed to date to the early 16th century. Some 12th-century pottery was
recovered from levels below these features (DES 1979, 10).

A few archaeological sites have been identified within a short distance of the site: pits
and other features, possibly associated with coal extraction, are visible in aerial
photographs at NO 540 024; a former gasworks stood at NO 5452 0245; and two
Edward I pennies were retrieved by metal detector at NO 5435 0235.A study of early
editions of Ordnance Survey maps (1855, 1895, 1914, 1938 and 1965) showed no
changes within the site itself over the past 150 years.

The Evaluation
The principal aim of the evaluation was to determine whether there were any structures,
features or deposits relating to the medieval burgh or its post-medieval expansion within
the area of proposed development. This consisted of the excavation of three trenches,
covering a total area of 8m2 which comprised 5% of the site. All trenches were
excavated by hand. They were left open at the end of the excavation.

Trench 1
Trench 1, located near the centre of the site, measured 5m east/west by 1m wide. Below
the turf, was up to 0.6m of very dark, humic topsoil which overlay a thin layer of lighter,
more clayey soil, possibly the result of bioturbation, and very thin lenses of undisturbed
glacial clay. Bedrock was exposed at a depth of 0.5-0.8m, dipping slightly towards the
west but steeply towards the south. It was very friable and fractured easily, resembling
deliberately laid masonry in places although its natural origins were in no doubt. Three
sherds of pottery were retrieved from the base of the topsoil in Trench 1, two rim sherds
of post-medieval reduced ware and one body sherd of post-medieval oxidised ware.

Trench 2
This small trench was located towards the north-east corner of the site and measured
2m north/south by 1m wide. The dark humic topsoil was 0.35-0.40m deep and overlay
glacial deposits of coarse yellow sand and pale brown clay. Bedrock was not exposed in
Trench 2.

- 319 -
Trench 3
Trench 3, which measured 1m square, was located near the north boundary of the site.
Topsoil was 0.40-0.45m deep and overlay about 0.1-0.2m of rubble which included
fragments of brick and 19th-century pottery. Below the rubble lay glacially deposited
sand.

Conclusions
No evidence of any pre-19th-century structures or features was uncovered in any of the
trenches. Although three sherds of post-medieval pottery (perhaps dating to the 17th
century) were retrieved, they were found in garden soil and were clearly in a residual
context, there being nothing to suggest that they formed part of a midden. On this
evidence, it is believed that extending the investigation within the area adjacent to 64
High Street, Pittenweem would reveal little, if anything, of archaeological significance.

References
DES Discovery & Excavation in Scotland, Council for Scottish Archaeology. Edinburgh
Lawrie, A C 1905 Early Scottish Charters Prior to 1153. Glasgow
Pryde, G S 1965 The Burghs of Scotland. London
Simpson, A T & Stevenson, S 1981 Historic Pittenweem: the archaeological implications
of development (= Scottish Burgh Survey). Edinburgh

- 320 -