Finding Bosworth Battlefield: a multiproxy palaeoenvironmental investigation of lowland sediments from Dadlington, Leicestershire, England

Journal of Archaeological Science 37 (2010) 1579–1589

Contents lists available at ScienceDirect

Journal of Archaeological Science journal homepage: http://www.elsevier.com/locate/jas

Finding Bosworth Battlefield: a multiproxy palaeoenvironmental investigation of lowland sediments from Dadlington, Leicestershire, England Jane Wheeler a, *, Graeme T. Swindles a, Benjamin R. Gearey b a b

Division of Archaeological, Geographical & Environmental Sciences (AGES), University of Bradford, Bradford BD7 1DP, UK Birmingham Archaeo-Environmental, University of Birmingham, Edgbaston, Birmingham B15 2TT, UK

a r t i c l e i n f o

a b s t r a c t

Article history: Received 30 August 2009 Received in revised form 14 January 2010 Accepted 18 January 2010

This paper presents the results of palaeoenvironmental investigations in an area proposed to contain the site of Bosworth Battlefield, near Dadlington, Leicestershire. Polydore Vergil’s sixteenth century account is the only source, albeit secondary, that is referenced in histories and logistical interpretations of the battle. Antiquarians and historians repeatedly reference this account, citing its description of a ‘marsh’ which is believed to have been the central site of the meˆle´e. Two sites in the floodplain of the former River Tweed have been identified as containing organic deposits characteristic of wetland environments. High-resolution lithostratigraphic and palaeoenvironmental data from each site are used to critically evaluate if the deposits represent the marsh and thus define the battlefield as described by Vergil. These new multiproxy data consolidate the local chronology of vegetation, hydrology, and sedimentological dynamics at the site from the Neolithic to the Medieval period. Whilst a precise interpretation of ground conditions at the time of the battle in 1485 cannot be made, due to truncation of the record as a result of modern ploughing and floodplain processes, the results provide a wider landscape context and illustrate the presence of local wetlands in this area that existed into the Medieval period. Ó 2010 Elsevier Ltd. All rights reserved.

Keywords: Battle of Bosworth Pollen analysis Multiproxy Palaeoenvironment Medieval

1. Introduction The death of Richard III at Bosworth Field in August 1485 marked a crucial turning point in English history as the last of the Plantagenet kings was succeeded by the founder of the Tudor dynasty – Henry VII. In the absence of primary documentary source material, histories of the battle rely on Polydore Vergil’s sixteenth century account reproduced in The Anglica Historia (reproduced in Ellis, 1844). This biased account provides the only testimony to the strategy and the eventual outcome of the battle, albeit written some thirty years after the event by an author who was both absent from the battlefield, and patronised by Henry Tudor (Hay, 1950). The Anglica Historia has also influenced strategic hypotheses and theories pertinent to the chronology and logistics of the battle, although detail of the battlefield location, in the absence of any bona fide primary documentary evidence or concentrations of battlefield finds and/or human remains, remains speculative. One specific natural landscape feature which Vergil describes as a ‘marishe’ (marsh) is believed to have provided a defensive right flank to protect the army of Henry Tudor, creating ground conditions between the two armies that strategically disadvantaged and * Corresponding author. Tel.: þ44 1274 235396; fax: þ44 1274 235190. E-mail address: [email protected] (J. Wheeler). 0305-4403/$ – see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jas.2010.01.019

impeded the advance of the Yorkist forces, thus contributing to the Lancastrian victory (Brooke, 1857; Foss, 1990; Hutton, 1788; Potter, 1983; Rees, 1985; Rowse, 1968; Royle, 2009). The semantic of the sixteenth century term ‘marishe’ has not been questioned by historians in respect of the environmental context of Vergil’s description of the local environment. Interpretations of this reference follow standard etymological derivations from Old French (‘mareis’ and ‘maresche’) and Medieval Latin (‘mariscus’) meaning ‘a marsh’ or being ‘marshy’ (Fisher, 1997; Kirkpatrick, 1989; Latham, 1999; Partridge, 1990). Whilst this environmental description is generic, it would appear that the use of this term may be an appropriate reference to the adverse ground conditions which influenced the strategy and eventual outcome of the battle. If pinpointed, the ‘marsh’ has the potential to preserve organic deposits which may provide information regarding the landscape and the hydrology of the local terrain in the medieval period, and identify the location of this important battlefield. In this paper new lithostratigraphic and palaeoenvironmental data are presented from the former floodplain of the original course of the River Tweed to the west of Dadlington; to i) investigate if deposits representing former wetlands exist in this locale, ii) to determine the character of these wetlands to assess if they are representative of the marsh described by Vergil, and iii) to define the area most likely to be the site of the battlefield.

1580

J. Wheeler et al. / Journal of Archaeological Science 37 (2010) 1579–1589

2. Background The site of Bosworth Battlefield is believed to be situated to the west of the villages of Dadlington and Stoke Golding some 3.5 km south of Market Bosworth, Leicestershire (Fig. 1). The extent of the overall battlefield area including encampments (approximately 4.5 km2) has been calculated based on the limited scatter of archaeological artefacts with a military and fifteenth century provenance (Glenn Foard (The Battlefields Trust) and Richard Mackinder (Bosworth Battlefield Heritage Centre & Country Park) pers. comm. 2008), and logistical and strategic hypotheses proposed by various authors (Brooke, 1857; Foss, 1990; Hutton, 1788; Potter, 1983: Rees, 1985; Rowse, 1968; Royle, 2009). 3. Methodology 3.1. Fieldwork

e

Two major concentrations of auger surveys were undertaken by The Battlefields Trust and Bosworth Battlefield Heritage Centre between 1976 and 2008 (currently unpublished). These preliminary surveys identified two specific locations to the southwest of Ambion Hill (SK403003) (which had until recently been thought to be the site of the battle) (Gearey et al., 2008; Hill et al., 2006): Fen Hole (central grid reference SP381985) and Fen Meadow (central grid reference SP397982) to the immediate east. Both sites were potential locations for the marsh described by Vergil in The Anglica Historia. Where organic deposits had been located at Fen Meadow exploratory trenches were established to extract monolith samples (Gearey et al., 2008). Cross-cutting transects were then set up at Fen Hole and Fen Meadow where the organic deposits had been identified previously (Wheeler and Swindles, 2009). A lithostratigraphic survey was carried out at both sites using an Eijelkamp silt shoe auger. Where suitable records for further analyses were found,

a Russian chamber corer was used to extract continuous core sequences with minimum contamination (Jowsey, 1966). Each auger and core point was levelled to ordnance datum (m OD) and the sediments logged in the field using the Troels-Smith (1955) scheme. Representative cores were wrapped in Clingfilm and aluminium foil to minimise contamination. On return to the laboratory the cores were refrigerated at 4  C until removed for subsampling. 3.2. Laboratory analyses The cores were sub-sampled at a resolution of 4 cm and prepared for pollen analysis using non-acid extraction (after Hunt, 1985; Wheeler, 2007). A pollen sum of 500 total land pollen grains (TLP) was counted, excluding spores, to assess the representation of sub-fossil pollen at each site. Rare pollen types are quantified at 2%. Spores (including the Lycopodium ‘spike’ (cf. Stockmarr, 1971) which are an indicator of pollen concentration), and microscopic charcoal in fractions of <21 mm, 21–50 mm, and >50 mm, were counted in addition to TLP but not included in the total pollen sum. These data-sets are expressed as percentages of 500. Pollen was identified in accordance with keys in Moore et al. (1999), Beug (2004), supported by Reille (1999) and a modern pollen type-slide reference collection. Nomenclature follows Stace (2001). Non pollen palynomorphs (NPPs) (cf.van Geel et al., 1982/1983, 2003; cf. van Hoeve and Hendrikse, 1998) and testate amoebae (cf. Charman et al., 2000) were also identified and recorded to provide additional environmental information. Data are presented as pollen diagrams in Tilia 2.0.2 format (Grimm, 2004). Following initial pollen subsampling selected samples were extracted from the cores and submitted to the 14Chrono Laboratory (Queen’s University Belfast) for AMS radiocarbon dating. Samples submitted for radiocarbon dating were prepared using a standard acid-alkali-acid pre-treatment at 60  C followed by rinsing in deionised water until neutral,

Mill Lan

FEN MEADOW ne

La n

Dadlington

300

nes

line

n La

Fen

e

FEN HOLE

Ra

Lan e

Fox

ok

R

oa

d

Upt

on L

Ca na l

cov ert

St

Dadlington

ay

150 Kilometers

ilw

0

tto

Su

N

ane

ley

ck Hin

d

Roa

N Stoke Golding 50 m Fig. 1. Map showing the location of Fen Hole and Fen Meadow, near Dadlington, Leicestershire, England.

J. Wheeler et al. / Journal of Archaeological Science 37 (2010) 1579–1589

and then dried. Previous dates on samples from monolith sequences were dated at the NERC radiocarbon facility at East Kilbride, Scotland (Gearey et al., 2008). All radiocarbon dates were calibrated using INTCAL04 (Reimer et al., 2004).

4. Results 4.1. Lithostratigraphy Transects, auger, and coring points at Fen Hole and Fen Meadow are show in Fig. 2. The lithostratigraphies for Fen Hole (transects 1– 3) and Fen Meadow (transects 1–2) are presented in Figs. 3 and 4 respectively. The Fen Hole lithostratigraphy is characterised by basal sands, clays, and some gravel, which are overlain in one small area by a discontinuous silty well-humified herbaceous peat (approximately 40  30 m in area). The upper facies at this site consists of stiff, slightly organic alluvial silts and clays, capped with brown agricultural topsoil. The contacts between the alluvium and the silty peat are mostly very sharp suggesting an erosional interface leading to a stratigraphic non-sequence. The sporadic occurrence of silty peat deposits in neighbouring cores suggests localised erosion of the sequence related to fluvial processes. The silty peat deposits are thickest at the lowest altitudinal point (auger points FH14 and FH15) illustrating organic accumulation in a depression that formed a small floodplain mire – which may also explain the very limited extent of this feature. The basal lithostratigraphy at Fen Meadow is dominated by subangular gravel in a matrix of coarse sand and silt. In places this basal deposit is overlain by a dark brown silty well-humified peat

1581

similar to that found at Fen Hole. This organic deposit is locally variable in silt content, and sometimes contains sand and gravel. In addition, some macrofossil remains are present, including Eriophorum (cotton grass), Phragmites (reed), wood and twigs, and occasional charcoal fragments. This deposit forms one continuous unit, but is again of limited spatial extent (approximately 90  25 m in area). In certain areas the uppermost layers contain a high proportion of minerogenic material and may be classified as organic-rich silt. The deposits are overlain by alluvial sediments composed of stiff clays and silts, capped by brown agricultural topsoil. Similarly to Fen Hole, the contacts between the silty peat and the alluvial deposits are erosive in nature. The characteristics of the silty peat unit at Fen Meadow and topographic context suggest a small open pocket of floodplain mire. Extensive alluvial deposits with fluviatile gravels and coarse sand were present to the south of the silty peat unit, suggesting proximity to a former river channel (auger points FM1–FM4). 4.2. Chronology The radiocarbon dates presented in Table 1 show that the deposits from Fen Hole span the Late Iron Age/Early RomanoBritish period (core 14) to the Late Anglo-Saxon/Early Medieval transition (core 8). Organic accumulation at Fen Hole began before 49 BC–AD 56 (the underlying deposit is currently undated) with a transition to organic silts/clays dated to AD 54–221. The most recent date obtained for the organic deposits at Fen Hole is from the isolated unit in core 8 (AD 898–1024). Radiocarbon dates from Fen Meadow (core 18) are representative of the Earlier Neolithic to the Late Iron Age. A date of 3794– 3661 BC constrains the earliest accumulation of organic deposits in this core, and the transition to organic silts/clays is dated to 96 BC– AD 51. Radiocarbon dating was also carried out on extracted macrofossils from two monolith sequences (1a and 1c; Fig. 2) from the exploratory trench at Fen Meadow to investigate the variation in chronology of sediment accumulation at this location (Gearey et al., 2008). These results provide the earliest date for organic accumulation at Fen Meadow in sequence 1a (4240–3970 BC) with a transition to organic silts/clays dated to AD 610–770 and AD 600– 780. Further radiocarbon dating of this transition was carried out on the adjacent sequence (1c; Fig. 2) and revealed dates of AD 130– 350 and AD 530–660. The sharp sedimentary boundaries and age variation between the organic deposits and the overlying silts and clays suggest that the sequence has been affected by substantial erosion and/or post-depositional reworking. 4.3. Pollen analysis Two pollen diagrams from the sedimentary sequences at Fen Hole are presented in Figs. 5 and 6. Each diagram has the prefix ‘FH’. Diagram 1 has been divided into two local pollen assemblage zones (LPAZs). As there is little variability in the pollen assemblage throughout this profile the stratigraphic hiatus at 36 cm has been used as the visual boundary between FH1 and FH2. Diagram 2 comprises one LPAZ due to the shallow nature of the sediment and the radiocarbon date at 45–47 cm. The core lithostratigraphies are plotted on both pollen diagrams.

Fig. 2. Locations of the auger/core points at (A) Fen Hole; (B) Fen Meadow. Refer to Fig. 3 for the lithostratigraphic profiles. The monolith sequences (1a,1c) from the exploratory trench at Fen Meadow (Gearey et al., 2008) are also shown. Source of aerial images: Google Earth.

4.3.1. Fen hole 4.3.1.1. FH1 73–36cm: Late Iron Age/Romano-British transition (core 14). Radiocarbon dates at 47–49 cm and 37–39cm provide a Late Iron Age/Romano-British provenance for FH1. The pollen record shows the local landscape was cleared of trees by this period and comprised open wet grassland indicated by the major presence of Poaceae (grasses), Cyperaceae and Carex (sedges), and herbaceous rare

1582

J. Wheeler et al. / Journal of Archaeological Science 37 (2010) 1579–1589

m ASL

A

B

9

81.2

10

1

2

10

20

3

4 5 6

7

81.0 80.8 80.6 80.4 80.2 80.0 79.8 79.6 79.4 79.2 0

30

40

Distance (m)

C m ASL 81.0 19

18

17

16

15

D

* 1 *8 13 11 12 14

80.8 80.6 80.4 80.2 80.0 79.8 79.6 79.4 79.2 0

m ASL

5

10

15

20

30

35

40

45

50

Distance (m)

E

1

81.0

25

55 F

21

Legend

20

80.8

Sh

80.6

Tl

80.4

Th

80.2

Ld

80.0

Ag

79.8

As

79.6

Ga

79.4

Gg

79.2 79.0 0

2

4

6

8 10 Distance (m)

12

14

16

Fig. 3. Lithostratigraphic profiles of the Fen Hole transects (refer to Fig. 2 for the core locations). Shaded areas indicate the silty peat deposits. The cores analysed are indicated by an asterisk. Stratigraphic nomenclature follows Troels-Smith (1955).

J. Wheeler et al. / Journal of Archaeological Science 37 (2010) 1579–1589

A 1

m ASL 83.6

1583

B 2

3

4

5

6

7

8

9 19 10

11

12

83.4 83.2 83.0 82.8 82.6 82.4 82.2 82.0 81.8 81.6 81.4 81.2 0

50

m ASL

C

83.2

16

8

100

150 Distance (m)

200

D

*

18

17

14

15

13 Legend

83.0 Sh

82.8

Th

82.6

Ag As

82.4

Ga

82.2 Gg

82.0 81.8 81.6 0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

Distance (m) Fig. 4. Lithostratigraphic profiles of the Fen Meadow transects (refer to Fig. 2 for the core locations). Shaded areas indicate the silty peat deposits. The cores analysed are indicated by an asterisk. Stratigraphic nomenclature follows Troels-Smith (1955).

types (2%) including Ranunculaceae (Buttercups), Chenopodiaceae (Goosefoots), Plantaginaceae (Plantain family), Lactuceae and Taraxacum (Dandelions). Greater counts of Equisetum (Horsetails) between 73 and 53 cm reflect wetter ground conditions which correlates with the NPP presence. The influx of the cereal-type pollen Secale cereale (Rye) at 53 cm, albeit as a rare type (0.4%), is a reliable anthropogenic indicator for cultivation (Behre, 1981). Pollen counts for undifferentiated cereal-type pollen recorded at 41 cm (1%) and 37 cm (1.8%) also appear reflective of small-scale cultivation in the vicinity of the Fen Hole site. Minor abundances of Quercus (oak), Ulmus (elm), Alnus (Alder), Salix (Willow), Corylus avellana (Hazel) and Myrica gale (Bog-myrtle) throughout FH1 suggest a minor presence of arboreal and shrub taxa which may represent peripheral or more distant regional woodland in respect of arboreal species, and managed or grazed vegetation which may have prevented the regeneration of species such as C. avellana and M. gale, Alnus and Salix. If the latter species were cut or grazed on a regular and tight cycle this would effectively prevent flowering thus causing a reduction in pollen production and a subsequent decline in the pollen curve (Rasmussen, 1990).

The presence of microscopic charcoal particles <21 mm throughout the zone, in comparison to much lower counts of larger charcoal particulates (21–50 mm and >50 mm), appears indicative of airborne transportation from a regional source as opposed to localised burning. Yet the presence of reasonable quantities of microscopic charcoal particles 21–50 mm (43–24%) and >50 mm (0.8–4%) does suggest more localised burning, albeit peripheral, in the vicinity of Fen Hole. The failure of species such as C. avellana, M. gale, and Alnus to rejuvenate, all of which are relatively fire resistant species (Grau and Veblen, 2000), may reflect a combined regime of regular burning and/or tight cyclical cutting or grazing. The presence of the testate amoeba Arcella discoides-type throughout the zone is indicative of standing water conditions, whilst the species Arcella artrocrea is reflective of moderately wet conditions (Charman et al., 2000). Peaks in the counts of A. discoides-type at 73 cm (6%), 65 cm (8%), and between 57 and 49 cm (3%), indicate hydrological fluctuations which are most probably the result of regular flooding and poor drainage. The quantitative decline of A. discoides-type which begins to decrease at 45 cm indicates a shift to drier conditions, and is similarly reflected in

1584

J. Wheeler et al. / Journal of Archaeological Science 37 (2010) 1579–1589

Table 1 AMS 14C dates for the cores from Fen Hole and Fen Meadow and the monolith sequences (1a,1c) from the exploratory trench at Fen Meadow (Gearey et al., 2008). Dates were calibrated using INTCAL04 (Reimer et al., 2004). Site

Depth (cm)

Lab number

Material

14

Fen Hole

45–47 (core 8) 27–29 (core 14) 37–39 (core 14) 47–49 (core 14) 38–40 (core 18) 70–72 (core 18) 104–105 (core 18) 40–41 (sequence 1a) 40–41 (sequence 1a) 84–85 (sequence 1a) 37–38 (sequence 1c) 39–40 (sequence 1c)

UBA-10473 UBA-10474 UBA-10475 UBA-10476 UBA-10470 UBA-10471 UBA-10472 SUERC-16389 SUERC-16390 SUERC-16433 SUERC-17635 SUERC-17639

Peat Humin Peat Peat Peat Peat Peat Eriophorum seeds Carex seeds Salix sp. Eriophorum seeds Eriophorum seeds

1057 1880 1890 2004 2026 4436 4962 1360 1340 5265 1475 1780

Fen Meadow

Fen Meadow (Gearey et al., 2008)

the reduction of the pollen curve for Carex (45–37 cm). However, we note that it is possible that some of the testate amoebae may be allochthonous in nature and deposited by water. The quantitative dominance of the NPP Gloeotrichia Type 146b and the presence of fungal spores Type 729 and Globose Type 181 suggest continuous eutrophic and relatively warm conditions, whilst the presence of the chlamydospore Glomus cf. fasciculatum Type 207 indicates a phase of erosion (van Geel et al., 1982 et al., 1982/1983, 2003). Pollen preservation is good at 61–41 cm, 29 cm, and 21 cm, but lower levels of preservation were recorded at 73–65 cm, 37–33 cm, and 25 cm. Poorer preservation ratios of between 30 and 50% less pollen correspond with the NPP erosional indicator Glomus cf. fasciculatum Type 207 recorded in the three basal sub-samples, and the hiatus in the core lithostratigraphy at 36cm. Therefore it is likely that lower levels of pollen preservation are the result of corrosion and degradation caused by periodic aeration. 4.3.1.2. FH2 36–21 cm: Romano-British period (core 14). Zone FH2 is similar in terms of species composition to the underlying zone FH1. Of particular note is the influx of Poaceae cereal types which begins in the underlying zone at 41 cm, and despite their absence at 33 cm (which is most probably associated with the hiatus in the stratigraphy at this point) are present throughout FH2. The arrival of cereal-type pollen at 41cm in FH1 corresponds with a decline in Equisetum and the disappearance of Sphagnum moss and the testate amoeba A. discoides-type, indicating a shift to drier conditions which appear representative of the climatic upturn of the Romano-British period (Lamb, 1977). A temporary rise in the NPP Gloeotrichia Type 146b between 33 and 29 cm, and isolated peaks in fungal spores Type 729 and Valsaria variospora type ascospores Type 140 indicate a temporary period of eutrophication that corresponds with the slight rise in Cyperaceae at 29 cm (44%). The phase of burning which spans the horizon at 36 cm before declining at 29 cm suggests an intensification of more localised fire activity as indicated by the raised counts of microscopic charcoal 21–50 mm and >50 mm respectively. The lack of regeneration of arboreal and shrub species at 29 cm, particularly the failure of Alnus, Salix, and C. avellana, at the point where all microscopic charcoal fractions decline, may be the result of resource management and/or grazing as opposed to continued clearance or management by fire, as the decline in microscopic charcoal at 33 cm appears to have no associated impact on the pollen assemblage. Influx of the NPP Glomus cf. fasciculatum Type 207 at 25 cm (6%) corresponds with a further phase of erosional activity. Pollen preservation also appears reflective of the alluvial process with preservation levels 50% lower at 33 cm and 25 cm.

C age (BP)            

22 23 32 23 23 25 26 35 50 35 35 35

D13C

Calibrated age range (2s – 95.4%)

30.5 27.2 28.1 27.9 25.5 26.7 24.1 25.7 25.0 27.3 24.9 25.4

AD 898–1024 AD 71–215 AD 54–221 49 BC–AD 56 96 BC–AD 51 3327–2933 BC 3794–3661 BC AD 610–770 AD 600–780 4230–3980 BC AD 530–660 AD 130–350

Lower levels of preservation at these depths appear to be the result of periodic aeration and/or traumatic depositional and postdepositional activity which would correspond with alluvial processes. 4.3.1.3. FH3 54–42 cm: Late Anglo-Saxon/Early Medieval period (core 8). The shallow 12 cm pocket of silty peat obtained from core 8 has a Late Anglo-Saxon/Early Medieval date. It is unlikely that the overlying 3 cm of this silty peat deposit subsumes intact sediment with a later medieval provenance due to truncation and disturbance indicated by a sharp undulating contact with the overlying alluvium. The pollen assemblage shows that the local landscape was not vegetationally dissimilar to that of the Iron Age and Romano-British period, comprising open wet grassland. Similarly, NPP counts and the continual presence of the testate amoeba A. discoides-type indicate wet eutrophic conditions and erosional activity. The presence of coprophilous NPP ascospore types Arnium Type 261 and A. Imitans Type 262 (indicative of nitrogen- and dung-rich environments) may be attributable to grazing intensification during the Anglo-Saxon/Anglo-Scandinavian period. The absence of cereal-type pollen throughout this zone may reflect a transition to wetter conditions which resulted in the abandonment of tillage. Microscopic charcoal fractions show a gradual quantitative rise throughout the zone that does not correspond with any noticeable fluctuations in the pollen assemblage. Greater counts of microscopic charcoal particulates <21 mm and 21–50 mm suggest peripheral and more regional fire activity. Levels of pollen preservation (30–40%), in comparison to the two underlying zones, are quantitatively poor throughout FH3, particularly at 50 cm (54%). Consistently poor levels of preservation are likely to be associated with phases of aeration and sedimentary disturbance, most probably the result of alluvial processes. 4.3.2. Fen Meadow The pollen diagram from Fen Meadow (Fig. 7) has been visually divided into three LPAZs based on major fluctuations in the arboreal pollen assemblage, specifically the representation of Quercus and Alnus. The dates are therefore incidental in respect of zonation but provide a provenance for each LPAZ. FM1(3794–3661 BC) and FM2 (3327–2933 BC) both have Earlier Neolithic dates. The basal horizon of FM3 at 52 cm, whilst zoned in respect of the quantitative decline of arboreal species, correlates with the lithostratigraphic hiatus at this point. FM3 has a Late Iron Age/Early Romano-British date of 90 BC–AD 51. The sediment lithostratigraphy is plotted on the pollen diagram. Pollen preservation was consistently good in comparison to preservation levels recorded for the Fen Hole pollen cores.

1880 ± 23

1890 ± 32

2004 ± 23

Blue grey alluvium 20

Spores

20

4.3.2.1. FM1 105–88 cm: Earlier Neolithic (core 18). Whilst the arboreal presence in this zone is relatively low 10%, the presence of Tilia (Lime), Ulmus, and Quercus suggest peripheral woodland in an open landscape which may be the remnants of the original NPPs

20

20

40

20

21 -5 0 ha µm r Ly coa co l > po 50 di um µm Tr ee Sh s r H ubs er bs Aq ua tic s Zo ne

20

oa l

Aquatics

29

53

40 20

Testate Amoebae

29

53

20

Zo ne

Po ac Po ea ac e C ea ce yp e re er - S a l a c e -ty ea ca pe e le ce re

al e

Dwarf Shrubs

C

2004 ± 23 Shrubs

ha rc

1890 ± 32 C or y M lus yr a i v He ca ell de gal ana r Er a e ic h e a R ce lix an ae C u nc he u l Ru n op ace m o d ae As ex ia c ea te e Pl rac a n ea Pl tag e an o l Pl tag anc an o e o m So tag aj lat nc o m or a Fi hu ed lip s- ia t Ce e nd ype ra u la La stiu ct m Ta u ce -typ ra ae e Ap xa c ia um Po cea ac e ea e

C Ye De ars BP pt h (c m Li ) th ol og y Pi nu Be s tu Q la ue Ti rcu lia s U lm u Al s nu Fr s ax Sa inu lix s

14

Trees

C

1880 ± 23

Eq ui M se t yr um i Ty oph ph yll Pt a l um e r at i Po idiu folia ly m D po d ry iu o Po pte m l y ri s Ti stic lle hu t Sp ia s m h a ph D gn ag ip u n o m i G roth lo m ec So us a r rd cf hi z Sp ari . fa op i ro a T sc hil G gy yp icul a T lo ra e at yp 5 b u G ose Typ 5A m e 1 lo T y 43 bo Ty e 1 pe G se pe 30 lo 20 bo Ty 18 7 G se pe 1 lo eo Ty 18 tri pe 2 ch 1 As ia 84 co Ty M sp pe icr or 14 es o Al fo 6B T g a ss y pe i l l Fu sp Ty 1 ng o r pe 40 M al es T 30 ou sp y 7 B Ar ge o ore pe ce ti a s T 16 l y l Ar a cf p 7 ce art . p e 7 C lla roc unc 29 ha d re ta rc i sc a ta oa oi Ty l < de pe 21 s-t 31 y µm pe 3F

Ye D ars ep BP th (c m Li ) th ol og y C ar ex

C

14

J. Wheeler et al. / Journal of Archaeological Science 37 (2010) 1579–1589 1585

Herbs

21

25

FH2

33

37

41

45

49

57 FH1

61

65

69

73 40

Charcoal

21

25

33 FH2

37

41

45

49

57 FH1

61

65

69

73

20 40 60 80 100

Brown silty peat

Fig. 5. Percentage pollen and palynomorph (including testate amoebae) diagram for Fen Hole (core 14).

mixed oak forest (Vera, 2004). The dominance of the spore Dryopteris type (Male Fern) also indicates the presence of peripheral woodland. Low percentages of shrub pollen, specifically Salix and C. avellana, show little fluctuation which may be the result of

J. Wheeler et al. / Journal of Archaeological Science 37 (2010) 1579–1589 Herbs

Aquatics

yp er ac ea e

C

{

1057 ± 22

C

D

14

Disturbed

Eq u O ise t en um Ty ant ph he a -ty la pe t if Zo ol ne ia

Dwarf Shrubs

ar ex

Shrubs

Ye ar ep s B P th (c m Li ) th ol og Pi y nu Be s tu Q la ue Ti rcu lia s U lm u Al s nu Fr s ax Sa in u lix s C or y M lus yr a i v H ca g ell ed a an Er e ra le a ic he a R ce lix an ae u C nc he u la n R op ce um o ae d As ex ia c ea te ra e Pl c e an a e t Pl a g an o Pl tag lan a n o ce m So tag aj ola n o m or ta Fi chu ed lip s- ia C e nd type irs u i la C um er -t as yp Sc tiu e le m La ran -typ ct thu e Ta uce s-t ra ae ype Ap xa c ia um Po ce a ac e ea e

Trees

C

1586

42

X 46 FH3 50

54 20

NPPs

20

40

Testate Amoebae

20

Charcoal

14

C

Tr e Sh es r H ubs er b Aq s ua tic s Zo ne

Ye D ars ep BP th (c m Li ) th ol og y Se la Pt gin er el l Po idiu a ly m p D od r y iu o Po pte m ly ris Ti stic ll e h u Sp tia s m ha p h D gn ag ip u n o m i G rot lo he m c us a cf rhiz C .f o er as ph co ci ila So p cu T rd hor la y p tu e Ar ari a T m 1 ni a T yp u Ty 43 y A. m p e 1 pe Im Ty e 5 12 20 p 5 G ita e A 7 lo ns 2 bo T 61 G se y lo e T pe Al otr ype 26 ga ich 1 2 Eu l sp ia T 81 ry or y Ar cer es pe 1 ce cu Ty 4 Tr lla s c pe 6B ig dis f. 15 l o Am no coi am 0 e ph pyx des lla Ne itr is -ty tus a e b m r pe Ty Hy ela a cul pe a c a 72 o w r -t C losp llar ig h ype D ha h is tia nu rc en oa ia m l < su C ha 21 bf r µm lav c C o a ha al rc 21 Ly oa -5 co l > 0 po 5 µm di 0 µ um m

Spores

40

Disturbed 1057 ± 22

{

42

X 46 FH3 50

54 20

20

20

40

60

20 40 60 80 100

Brown silty peat Fig. 6. Percentage pollen and palynomorph (including testate amoebae) diagram for Fen Hole (core 8). The uppermost part of the organic deposit in Core 8 (Fen Hole) was disturbed, possibly by ploughing. An AMS 14C date was carried out where the sediment was intact (45–47 cm).

regular anthropogenic exploitation and/or grazing at the site. Herbaceous pollen is dominated by Poaceae and Cyperaceae, suggesting open wet grassland. The presence of the testate amoebae species A. discoides and A. artrocrea are indicative of wet conditions. Higher continuous counts of the NPP chlamydospore Glomus cf. fasciculatum Type 207 suggest elevated erosional activity. Counts of NPPs including Gloeotrichia Type 146b, Globose Type 181, and Spirogyra Type 130 indicate eutrophic conditions. Diporotheca rhizophila Type 143 ascospores represent nitrogen-rich and dung-rich deposits. The presence of microscopic charcoal from all fraction sizes suggests a relatively stable fire regime, but with a wider and more regional source as represented by the greater ratio of microscopic charcoal <21 mm. However, the peak in microscopic charcoal at 89 cm, particularly the noticeable rise in 21–50 mm and >50 mm fractions, suggests a phase of localised burning which may have triggered the immediate and overlying response of Alnus at 85 cm in the basal zone of FM2. This exacerbated incident of localised burning may be representative of either natural or anthropogenically induced fire, yet the disappearance of Dryopteris spores at 85 cm in association with the decline of arboreal species and shrubs

(specifically Salix and C. avellana) corresponds with a clearance phase. 4.3.2.2. FM2 88–52 cm: earlier Neolithic (core 18). The noticeable influx of Alnus at 85 cm and the overall decline of Tilia and Ulmus appear to have a strong correlation with the potential impact of burning (89 cm) which impacts on the arboreal presence at this point. The disappearance of Dryopteris is probably representative of the retreat of peripheral woodland as a consequence of fire. The peak in Alnus at 85cm (67%) is particularly interesting as it appears to be a direct response to this phase of burning, not only highlighting the resilience of the species to fire (Grau and Veblen, 2000) but also its ability to withstand wet ground conditions (Bennett and Birks, 1990) as indicated by the rise in the testate amoeba A. discoides-type (85–81 cm). The presence of algal spores Type 167, ascospores Type 307B, Gloeotrichia Type 146b, and Globose Type 181 at 85–81 cm indicate a eutrophic environment. The recovery of Quercus between 81 and 57 cm appears to be in response to the decline of Tilia and Ulmus, which (until the final large-scale burning episode defined at 81 cm) appears to have been suppressed by these shade-tolerant species (Vera, 2004). The influx

of M. gale at 81 cm corresponds with the peak in the testate amoeba A. discoides-type, suggesting a shift to wetter conditions. The relative stability of M. gale and C. avellana throughout this zone, particularly their failure to regenerate in comparison to Alnus, may

105

101

97

93

89

85

81

77

73

69

65

20

lia Ti

4962 ± 26

4436 ± 25

105

101

97

93

89

85

81

77

73

69

65

61

57

53

49

45

Brown silty peat

20

BP m) s lla m y ar h (c ne iu um og e gi len ridi ol Y ept a h l p e t 14 C D Li Se As Pt 37 2026 ± 23 41

4962 ± 26

4436 ± 25

61

57

53

49

45

BP m) s y c s ar h ( og la rcu s ol Ye ept th nu t u ue D 14 C Li Pi Be Q 37 2026 ± 23 41

40

20

Dw arf Shrubs

Herbs

Spores

7 20

NPPs

Humified organics

20

20

50

100

150

C

Testate Amoebae

3 14 e 0 14 pe Typ y T m pe B 29 y a il t u 12 va T 46 40 74 7B 67 7 a pe ic s fla ty p h u l a 1 A 30 62 1 2 ra 1 1 1 0 1 pe er en ub m zo cic pe 55 e 1 61 e 2 18 18 po ype pe pe e 3 ype T y ea esi h i c s µ p n rh fas Ty pe yp e 2 yp pe pe ios T Ty Ty Typ s T es cr oid is es ia 1 a ag a y T yp T Ty Ty ar hia es es il re or tro sc m ub en <2 v i s r s r p r ris h m ec cf. or T a he ia r ph oal te a sp gnu ot h us ph ria gyr m T itan se se ria tric po po f os spo l s a a a d s a p c a ll ll o o a o s s o l i o a r ll ug lo r y el llet ha ipo lom erc rd iro niu Im lob lob ls loe co co icr ga ng ce ce ce if fl ya ha Th Ti Sp D G C So Sp Ar A. G G Va G As As M Al Fu Ar Ar Ar D H C

60

Shrubs

a at . e ol r a is e ae i ff na e yp ar ea c e c e ajo edi p e m ix l la nd ae e yp -t n u g a i c l e u l i l e e v al e fo e e la m m -t y la p e e m-t h us e um a ce a e u la od ea a v a h i u a y u e go go go us d -t ea iu nt e s g ea u a ea c an ce ium rac ae c c s p a u x t a c a c u a n s n r q o i e er nta nta nta ch e pi tac as ra tu ax er c a lob se ce i n ix yl ric a a e ac a lun u n x i e l s n r p d m e r n t c l c r l a i l x e c i a a h u i a a a a l i re is e o y o rt ro Fr Sa C M R Il e H Er Er C R C R As Pl Pl Pl So Fi C C C Sc La Ta Va U Ep D Po

Silty peat with organic material

20

40

40

m iu ris od te yl p ryop Po D

20

us us lm n U Al

Trees

60

-5

m 0µ

40

1 l2 oa

20

rc ha

40

60

20

pe

20

40

µm 50 m l > diu a o o rc op ha c C Ly

20 40 60 80 100

FM1

FM2

FM3

cs s bs s ati ne ee hru erb qu r T S H A Zo

FM1

FM2

FM3

m olia m m ll u ti f tu onu phy la e ex is yg io h a ne ar u l y r p Zo C Eq Po M Ty

Charcoal

40

y l-t ea er e c a e ce ea ra a c yp e o P C

Aquatics

J. Wheeler et al. / Journal of Archaeological Science 37 (2010) 1579–1589 1587

Fig. 7. Percentage pollen and palynomorph (including testate amoebae) diagram for Fen Meadow (core 18).

suggest anthropogenic selection and exploitation of these species. The band of humified organics in the lithostratigraphy at 68–69 cm occurs at the time when fire activity declines, and coincides with a slight rise in the testate amoeba A. discoides type to 3.6%. Increases

1588

J. Wheeler et al. / Journal of Archaeological Science 37 (2010) 1579–1589

in the variety of herbaceous pollen at 57 cm, particularly the appearance of Plantaginaceae (which as a genus is the most important anthropogenic indicator (Behre, 1981), occurs at the point where arboreal and shrub species decline in response to burning – which may represent a change in agricultural regime. The presence of Pteridium (Bracken) throughout this zone, and its gradual rise following the cessation of burning at 61 cm infers a localised fire regime may have been practised in an attempt to limit or control its spread, or perhaps it was decreasingly exploited as the agricultural regime changed. 4.3.2.3. FM3 52–37 cm: Late Iron Age/Early Romano-British period. The basal horizon of FM3 (51 cm) has been introduced as a response to the quantitative decline of arboreal and shrub pollen and the stability of the pollen assemblage throughout the zone. The influx of Poaceae cereal-type pollen at 49 cm may be representative of tillage, whilst the general rise in Poaceae, Cyperaceae, and herbs, indicates a gradual expansion of open grassland. The decline in the presence of the testate amoeba A. discoides-type also implies a drier environment which may explain the arrival of cereal-type pollen, albeit as a rare type, in this zone. The introduction of tillage at this point corresponds with climatic and land management improvements generally associated with the Romano-British period. The influx of Pteridium at 49 cm, which steadily rises from 61 cm in the underlying zone, appears reflective of the invasive nature of the species as a consequence of the reduction or abandonment of fire activity indicated by the noticeable reduction in microscopic charcoal. The absence of coprophilous ascospores throughout the zone infers a reduction of livestock at the site, which appears to coincide with the establishment of meadowland (indicated by the stable presence of Poaceae) and the introduction of tillage (indicated by the arrival of cereal-type pollen at 49 cm), and raised counts of the NPP Glomus cf. fasciculatum Type 207 (indicative of erosional activity). This signature, particularly the disappearance of coprophilous ascospores in association with the presence of Glomus cf. fasciculatum Type 207, appears indicative of anthropogenically induced seasonal and cyclical flooding, which, due to the presence of cereal-type pollen, is most probably reflective of engineered water meadow management (Cook and Williamson, 2007). 5. Discussion The palaeoenvironmental data show that there is no evidence for a major peat-forming landscape feature in the local floodplain environment in the latter half of the fifteenth century. This finding is not unexpected as large bogs/fens are atypical of the Leicestershire lowlands in the present day. However, a floodplain environment was present in the Late Anglo-Saxon/Early Medieval period that appears to have comprised a seasonally flooded meadow system containing small isolated mires. The nature of the alluvium in this environment would have led to waterlogging which may have been particularly severe by the fifteenth century as climatic conditions deteriorated, becoming cooler and wetter at the start of the Little Ice Age (Lamb, 1977). Therefore it is probable that this expanse of waterlogged marshy ground corresponds with Polydor Vergil’s description of the fifteenth century landscape. No deposits were identified that are contemporaneous with the Battle of Bosworth due to truncation of the record by alluvial and agricultural processes in this highly dynamic environment. This may further support the hypothesis of a shift to increased wetness at this time. The radiocarbon dates from the uppermost organic deposits at Fen Meadow suggest uneven truncation, and it is possible that pockets of peat or organic silts of later date survive in this area. These data provide new evidence regarding the local landscape in the fifteenth century, which would have consisted of

open floodplain water meadows characterised by alluvial deposition from seasonal flooding of the River Tweed, which has since been replaced by the Ashby-de-la-Zouch Canal. The construction of the canal in the late eighteenth century and the introduction of modern field drainage have led to major hydrological change in this environment, specifically decreasing levels of soil moisture. Small mire units (in some cases reedswamp) may have been characteristic of a frequently waterlogged environment in the fifteenth century. The nature of the heavy clay alluvial sediments which dominated the floodplain and the mire units would have visibly indicated a waterlogged marshy environment. The advantage of higher and drier ground on the gentle slopes above the floodplain would have presented the strategic advantage to the army or battle group that did not need to cross this waterlogged environment in order to engage the enemy. Whilst Vergil’s description of the ‘marsh’ is meaningful, it seems likely that it does not refer to one specific peat landform or marsh, but to the complexities of a dynamic floodplain mosaic. 6. Conclusion The remit of this research project was to identify the conflict zone of the Battle of Bosworth using palaeoenvironmental techniques due to the absence of defining archaeological artefacts. Results highlight the problems of attempting to correlate palaeoenvironmental data from a complex sedimentary environment to target a precise historical date and known event. Whilst data from the Fen Meadow and Fen Hole sites provide a chronology of vegetational and sedimentological change which spans the Earlier Neolithic to the Late Anglo-Saxon/Early Medieval periods, thus enabling an environmental model to be proposed for the local environment at the time of the battle, the uneven truncation of the lithostratigraphy as a result of human activity and alluvial processes indicate that a single large unit of marsh (which in itself is a loose term, but one which reciprocates Vergil’s description) has not survived within the areas investigated. Therefore it is improbable that the extent of the conflict zone can be defined solely by the identification of a distinct area of marshland within the parameters of the proposed battlefield site. Vergil’s account provides no detail of the extent of this landscape feature or its role in respect of the topography of the battle, other than stating that it lay ‘betwixt both hostes’ (Ellis op. cit.). Although we have constrained the extent and chronology of organic deposits at Fen Hole and Fen Meadow, further investigations are needed to assess if any other similar organic units are present in the wider floodplain landscape. As there are limited archaeological finds with a determined fifteenth century military provenance from this area, a combination of precise artefact discrimination and GIS mapping linked with the proposed geoarchaeological study is now required to determine if an area of former marsh can be identified, and the sediment dated to the later medieval period. Whilst palaeoenvironmental data presented in this paper may support Vergil’s account of marshland in the landscape at the time of the battle and provides an environmental model for the Medieval period, sedimentological complexities, coupled with the current absence of concentrations or defined scatters of later Medieval military finds, prevent a definition of the contemporary area of marsh in relation to the site of the Battle of Bosworth from being isolated and identified. Acknowledgments We would like to thank Glenn Foard (The Battlefields Trust) and Richard Mackinder (Bosworth Battlefield Heritage Centre & Country Park) for their assistance throughout this investigation. We

J. Wheeler et al. / Journal of Archaeological Science 37 (2010) 1579–1589

are also grateful for the support of Ed Turner (University of Bradford) for his help and assistance in the field, and to Usha Gohil for her technical expertise in the laboratory. References Behre, K.-E., 1981. The interpretation of anthropogenic indicators in pollen diagrams. Pollen et Spores 23, 225–245. Bennett, K.D., Birks, H.J.B., 1990. Postglacial history of alder (Alnus glutinosa (L.) Gaertn.) in the British Isles. Journal of Quaternary Science 5 (2), 123–133. Beug, H.-J., 2004. Leitfaden der pollenbestimmung fu¨r mitteleuropa und angrenzende gebiete. Verlag Dr. Friedrich Pfeil, Mu¨nchen. Brooke, R., 1857. Visits to Fields of Battle in England of the Fifteenth Century. John Russell Smith: London/J. Mawdsley & Son, Liverpool. Charman, D.J., Hendon, D., Woodland, W.A., 2000. The Identification of Testate Amoebae (Protozoa: Rhizopoda) in Peats. Technical Guide No. 9. Quaternary Research Association, London. Cook, H., Williamson, T., 2007. Water Meadows: History, Ecology and Conservation. Windgather Press Ltd., Macclesfield. Ellis, H. (Ed.), 1844. Three Books of Polydore Vergil’s English History, Comprising the Reigns of Henry VI, Edward IV, and Richard III from an Early Translation Preserved Among the Mss. of the Old Royal Library in the British Museum. John Bowyer Nichols & Son, London available on-line at: http://www.questia.com. Fisher, J.L., 1997. A Medieval Farming Glossary of Latin and English Words. Essex Records Office Publications, Chelmsford. Foss, P., 1990. The Field of Redemore: The Battle of Bosworth 1485. Kairos Press, Leicester. Gearey, B., Hill, T., Howard, A., Marshall, P., 2008. Bosworth Fields. Palaeoenvironmental Survey and Assessment, Leicestershire (Unpublished report: Birmingham Archaeo-Environmental, The University of Birmingham). Grau, H.F., Veblen, T.T., 2000. Rainfall variability, fire and vegetation dynamics in neotropical montane ecosystems in north-western Argentina. Journal of Biogeography 27 (5), 1107–1121. Grimm, E.C., 2004. TGView 2.0.2. Illinois State Museum, IL 62703, USA. Hay, D. (Ed.), 1950. The Anglica Historia of Polydore Vergil A.D. 1485–1537. Offices of the Royal Society/Camden Series, vol. LXXIV (London). Hill, T., Howard, A., Gearey, B., 2006. Geoarchaeological Assessment of Bosworth Fields. Initial Results of Phase 1 Environmental Survey, Leicestershire (Unpublished report: Birmingham Archaeo-Environmental, The University of Birmingham). Hunt, C.O., 1985. Recent advances in pollen extraction techniques: a brief review. Oxford. In: Fieller, N.R.J., Gilbertson, D.D., Ralph, N.G.A. (Eds.), Palaeobiological Investigations: Research Design, Methods and Data Analysis. BAR International Series 266, pp. 181–188. Hutton, W., 1788. The Battle of Bosworth Field Between Richard the Third and Henry Earl of Richmond, August 22, 1485. Pearson & Rollason, Birmingham. Jowsey, P.C., 1966. An improved peat sampler. New Phytologist 64, 245–248. Kirkpatrick, B. (Ed.), 1989. The Cassell Concise English Dictionary. Cassell, London. Lamb, H.H., 1977. Climate: Present, Past and Future, vol. 2. Methuen and Co. Ltd., London.

1589

Latham, R.E., 1999. Revised Medieval Latin Word-list. Oxford University Press, London. Moore, P.D., Webb, J.A., Collinson, M.E., 1999. Pollen Analysis. Blackwell Science, Oxford. Partridge, E., 1990. Origins: An Etymological Dictionary of Modern English. Routledge, London. Potter, J., 1983. Good King Richard? An Account of Richard III and His Reputation. Constable, London. Rasmussen, P., 1990. Pollarding of trees in the Neolithic: often presumed – difficult to prove. In: Robinson, D.E. (Ed.), Experimentation and Reconstruction in Environmental Archaeology. Oxbow, Oxford, pp. 77–99. Rees, D., 1985. The Son of Prophecy: Henry Tudor’s Road to Bosworth. Black Raven Press, London. Reille, M., 1999. Pollen et spores d’Europe et d’Afrique du nord, second ed. Laboratoire de Botanique Historique et Palynologie, Marseille. Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Bertrand, C.J.H., Blackwell, P.G., Buck, C.E., Burr, G.S., Cutler, K.B., Damon, P.E., Edwards, R.L., Fairbanks, R.G., Friedrich, M., Guilderson, T.P., Hogg, A.G., Hughen, K.A., Kromer, B., McCormac, G., Manning, S., Bronk Ramsey, C., Reimer, R.W., Remmelle, S., Southon, J.R., Stuiver, M., Talamo, S., Taylor, F.W., van der Plicht, J., Weyhenmeyer, C.E., 2004. INTCAL04 terrestrial radiocarbon age calibration, 0– 26 CAL KYR BP. Radiocarbon 46 (3), 1029–1058. Rowse, A.L., 1968. Bosworth Field and the War of the Roses. Panther History, London. Royle, T., 2009. The Road to Bosworth Field: a New History of the War of the Roses. Little Brown, London. Stace, C., 2001. New Flora of the British Isles, second ed. Cambridge University Press, Cambridge. Stockmarr, J., 1971. Tablets with spores used in absolute pollen analysis. Pollen et Spores 13, 615–621. Troels-Smith, J., 1955. Karakterisering af løse jordarter Danmarks gliederung and definition der limnischen ledimente. Geologis-Geologiske Undersøgelse Series IV 3 (10), 73. van Geel, B., Hallewas, D.P., Pals, J.P., 1982/1983. A late Holocene deposit under the Westfriese Zeedijk near Enkhuizen (Prov. Of Noord-Holland, the Netherlands): palaeoecological and archaeological aspects. Review of Palaeobotany and Palynology 38, 269–335. van Geel, B., Buurman, J., Brinkkemper, O., Schelvis, J., Aptroot, A., van Reenen, G., Hakbijl, 2003. Environmental reconstruction of a Roman Period settlement site in Uitgeest (The Netherlands), with special reference to coprophilous fungi. Journal of Archaeological Science 30, 873–883. van Hoeve, M.L., Hendrikse, M. (Eds.) 1998. A study of Non-pollen Objects in Pollen Slides: the Types as Described by Dr Bas Van Geel and Colleagues. Unpublished report: Utrecht University Vera, F.W.M., 2004. Grazing Ecology and Forest History. CABI Publishing, Oxon. Wheeler, J., 2007. The Implications of Iron-Working on the Woodlands of Rievaulx and Bilsdale, North Yorkshire, UK: Historical, Palaeoecological and Palaeoenvironmental Perspectives, Circa 1068–2000. Unpublished PhD thesis, The University of Bradford. Wheeler, J., Swindles, G.T. 2009. Finding Bosworth Field: A multiproxy and palaeoenvironmental assessment of lowland sediments from the proposed sites of Bosworth Battlefield, Leicestershire. Unpublished report, The University of Bradford.