Content uploaded by Michel Dabas
Author content
All content in this area was uploaded by Michel Dabas on Nov 15, 2017
Content may be subject to copyright.
City Map of Ancient Epomanduodurum
(Mandeure–Mathay, Franche-Comté,
Eastern France): Contribution of Geophysical
Prospecting Techniques
G. BOSSUET
1
*,M. THIVET
1
, S. TRILLAUD
2
, E. MARMET
2
, C. LAPLAIGE
1
,
M. DABAS
2
, G. HULLIN
2
, A. FAVARD
2
, L. COMBE
2
, E. BARRES
2
, S. LACAZE
2
,
L. AUBRY
2
, M. CHASSANG
3
, A. MOUROT
3
AND C. CAMERLYNCK
3
1
UMR 6249-CNRS, Laboratoire de Chrono-environnement, Université de Franche-Comté, Besançon, France
2
Geocarta, 5 rue de la banque, Paris, France
3
UMR 7619-CNRS, Sisyphe, Université Pierre et Marie Curie, Paris VI, France
ABSTRACT This study presents the main results of the geophysical survey of ancient Epomanduodurum, at Mandeure–Mathay,
Eastern France. In northeastern Gaul, Epomanduodurum is a site of a major scientific interest for the understanding
of past settlements and territorial formation at the end of the Iron Age and during the Roman period. The site, including
a Roman urban centre and two suburbs containing workshops, occupies more than 500 ha inside and beyond a
meander of the River Doubs. From the beginning of survey in 2001, several methods (fast electrical imaging automatic
resistivity profiling system, ground penetrating radar, magnetic and electromagnetic mapping) have been performed on
a large scale in order to precisely identify the vast extent and structure of the Roman town. The interpretation of the
geophysical data was carried out using a combination of different data sources, including ancient maps and excavations
recently conducted on restricted areas. The overall organization of an artisan quarter of 8 ha was revealed along a main
Roman road leading to the left bank of the River Doubs. On the other bank, a river fortification and a vast monumental
religious complex were recognized in the southern part of the Roman town. The geophysical imaging shows that the
sacred area includes several temples or groups of temples radiating from the theatre. Other new buried structures
probably corresponding to annexes, chapels and altars were also detected within this sacred area of 10 ha bound on three
sides by a wall enclosure with monumental passages. In a surveyed area of 70 ha, geophysical prospecting techniques
provided a clear overall image of the Roman urbanism revealed through a rectangular road network and street system
delimiting insulae of variable sizes. The detection alongthe River Doubs of buildings similar to storehousesatteststo fluvial
transport and also suggests the likely existence of a port. More generally, the combination on a large scale of geophysical
prospecting, aerial photographs and excavations allowed a new plan of the urban structure of the ancient town of
Epomanduodurum to be produced for an area of over 300 ha. Copyright © 2012 John Wiley & Sons, Ltd.
Key words: Epomanduodurum; Mandeure–Mathay; geophysical prospecting; Roman urbanization; large-scale
archaeological survey; automatic resistivity profiling.
Introduction
Near-surface geophysical surveys have been available to
archaeologists for many decades (Scollar et al., 1990).
Advances in technology and practice over the past
decade allow geophysical surveys for archaeology to
produce maps of subsurface features over large areas
and in potentially great detail (Kvamme, 2003). A series
of case studies from northwestern Europe has shown that
geophysical surveys of Roman urban areas –Viroconium
(Wroxeter, England), Colonia Ulpia Traiana (Xanten,
Germany), Carnuntum (Vienna, Austria), Gisacum
(Viel-Evreux, France), Flavia Solva (Styria, Austria) –
can produce primary information suitable for the study
of site content, structure and organization (Buteux et al.,
2000; Doneus et al., 2001; Dabas et al., 2005; Neubauer
et al., 2009).
* Correspondence to: G. Bossuet, UMR 6249-CNRS, Laboratoire de
Chrono-environnement, UFR Sciences et techniques, 16 route de Gray,
25030 Besançon cedex, France. E-mail: gilles.bossuet@univ-fcomte.fr
Copyright © 2012 John Wiley & Sons, Ltd. Received 13 April 2012
Accepted 24 September 2012
Archaeological Prospection
Archaeol. Prospect. 19, 261–280 (2012)
Published online 6 November 2012 in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/arp.1433
In the same way, a multidisciplinary project was
performed between 2001 and 2011 on the site of
the ancient settlement of Epomanduodurum,modern
day Mandeure–Mathay, in Franche-Comté Region
northeastern France. Geophysical prospecting, aerial
photography, and archaeological investigations were
combined using a GIS system in order to make the site
exploration more efficient by mapping over a large area
and with high accuracy all the archaeological features
(Thivet, 2008). Today, the archaeological remains of the
Roman town occupy an area of approximately 500 ha
outside and inside a meander of the River Doubs.
Ancient Epomanduodurum is a particularly interesting
site for the understanding of past settlement and territorial
formation at the end of the Iron Age and during the
Roman period in northeastern Gaul ((Figure 1a; Barral
et al., 2007; Thivet et al., 2011b). Due to its large size,
its urban infrastructure, monuments and recognized
functions, it is considered to be the second city of the
Sequani territory, after the civitas Besançon-Vesontio
(Figure 1b; Jeannin, 1986; Frezouls, 1988).
This paper presents the main geophysical results for
the archaeological evaluation of the site, and the
creation of a city map of ancient Epomanduodurum.The
interpretation of the geophysical data was carried out
using a combination of different data sources including
ancient maps and excavations recently conducted on
restricted areas. Four functional areas of the town are
presented: an artisan’s quarter ‘Faubourg de Pont’on
the left bank of the River Doubs, a residential quarter
‘Champs des Combottes’within the loop of the meander,
a religious monumental complex in the southern part of
the ancient town and a peripheral occupation ‘Lomont
plateau’over the ancient theatre (Figure 2).
Geographical context
ThesiteofEpomanduodurum is located in the
Doubs valley, where the Alsace plain opens into the
marches of Burgundy, in a trading area between
the Vosges and the Jura mountains. The Saône and
Doubs valleys connect the Rhône valley to the Rhine
plain. (Millotte and Lambert, 1996; Figure1a). This
geographical situation was a significant factor in the
creation of the late Iron Age settlement, which later
turned into a major Roman town: an ancient road
system connected Pontarlier (Abiolica), Besançon
(Vesontio) and Bâle (Augst) (Joan, 2003; Figure1b).
In the floodplain of the River Doubs, Epomanduodurum
was formed as an urbanized centre with two suburban
areas used as craft centres, ‘Faubourg de Pont’and
‘l’Essarté’(Figure 2). The bottom of the valley is formed
on calcareous Pleistocene deposits and the 1000 m wide
alluvial plain was created by fluvial activity. Today, the
Roman town is partially covered by the modern
communities of Mandeure and Mathay. Elsewhere, the
ancient remains lie in open fields and vast meadows
which provide favourable conditions for large-scale
geophysical prospecting.
History of investigations
Archaeological research at Epomanduodurum started as
early as 1781: the first organized studies were the
‘Muraillebourg’excavations. In the nineteenth century,
the major monuments of the antique town were
uncovered: the theatre in 1820, the baths of ‘Courcelles’
in 1829, the portico and the monumental vestiges of
‘Muraillebourg’in 1830 and 1860, and finally the great
Roman sanctuary with peribolos of ‘Clos du Château’
in 1880 (Marc et al., 2007a). Ancient maps and detailed
plans of Roman ruins were also published during the
nineteenth century (Thivet et al., 2009).
However, during this period and until the 1950s, the
site suffered from damage caused by the search for
valuable objects and the reuse of building materials.
From 1950 to 1980, several urban works offered the
opportunity for carrying out limited excavations.
Those of the Roman theatre remain the main archaeo-
logical investigation during this period (Jeannin, 1986).
From the 1980s, the number of local interventions increased
thanks to new archaeological survey legislation and more
important excavations were opened, such as the artisans
quarter of ‘l’Essarté’(Lame and Mazimann, 1993)
and the Celtic necropolis of ‘Longues Raies’at Mathay
(Barral, 1996).
Geophysical prospecting
Geophysical survey at Epomanduodurum used a
combination of different methods and automatic
mapping. Three non-destructive techniques, electrical
resistivity, magnetic and ground-penetrating radar
(GPR) have been applied to recognize the extent and
organization of the Roman town.
Automatic resistivity profiling
For the exploration (depth 0–2 m) of architectural
remains a fast electrical imaging automatic resistivity
profiling (ARP) system was used. This system allows
vertical and lateral characterization of electrical resist-
ivity by continuous profiling (Dabas et al., 2002; Dabas
262 G. Bossuet et al.
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
Figure 1. (a) Regional orographic context of Franche-Comté, easter n France (from BD Alti I.G.N.; Thivet, 2008; Laplaige, 2011). (b) Main communications around the Doubs
valley at the Roman period (from Joan, 2003; Thivet, 2008; Laplaige, 2011). This figure is available in colour online at wileyonlinelibrary.com/journal/arp
263City Map of Ancient Epomanduodurum (Eastern France)
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
Figure 2. Mandeure–Mathay (Doubs). Areas investigated using geophysical methods. Background picture is an aerial photograph of Aerodata 2009 (Laplaige, 2011; PCR Mandeure).
This figure is available in colour online at wileyonlinelibrary.com/journal/arp
264 G. Bossuet et al.
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
and Tabbagh, 2003). The ARP device consists of metal-
lic wheels (probes) fastened to a frame and towed by a
quad bike. The multipole is made of four axles. The
dipole that is closer to the quad bike injects current to
the soil, while the three others measure the potential
resulting from the current flow. The three axles have
a respective width of 0.5 m, 1 m and 2 m, the depth of
investigation roughly being the shorter distance
between one voltage pole and one injection pole. Up to
150 000 measurements per hectare can be recorded by
using a 1 m interprofile spacing and resistivity values
collected every 0.2 m (Dabas, 2006). An area of 70 ha
was covered with the ARP system using this configur-
ation. Data processing included data reinterpolation on
an isotropic mesh of 0.25 m by 0.25 m and filtering
procedures (median, regional, spatial convolution) in
order to make anomalies of archaeological interest more
perceptible (Dabas, 2006).
Magnetic mapping
The magnetic survey was carried out using a G-858
Geometrics caesium magnetometer and a Ferex-CON60
Foerster fluxgate magnetometer, in gradiometer config-
uration, with the two magnetic probes set vertically
apart at a distance of 1 m, in such a way to automatically
remove the diurnal variations of the background mag-
netic field (Scollar et al., 1990; Tabbagh, 2003). The
magnetic measurements were collected on square grids
of either 50 m or 30 m, in a bidirectional pattern, by
walking both up and down lines of 1 m spacing. The sur-
vey was made in continuous mode with readings taken
at intervals of 0.1 s or 0.2 s cycle time and sensitivity of
0.01 and 0.3 nT.
The area prospected by magnetic mapping covered
40 ha. Wumap Software including digital and directional
filtering procedures was used for the processing of the
magnetic data (Tabbagh, 2001).
Ground-penetrating radar
A GPR survey was used occasionally to verify the nature
and structure of interrelated anthropogenic anomalies
(Neubauer et al., 2009). A GPR reflection survey was
carried out in the area south of the Roman theatre by
using PULSE EKKO 100 equipment and two 225 MHz
antennae with a separation of 0.5 m. The GPR data lines
were acquired every 0.4 m in a continuous mode in a
time window of 50 ns. The survey area of 0.5 ha was
covered by 250 parallel profiles of 50 m length following
a forward and reverse surveying pattern.
A series of time-slice and depth-slice maps was
created by using Ekko Mapper software. They showed
the evolution of reflective structures in accordance
with the time and depth. Computer data processing
had been carried out previously in order to remove
low-frequency wow in the data, to reduce high
frequency noise and to emphasize horizontal reflectors
(Sensors and Software, 2002).
Results and interpretation
The artisan quarter of ‘Faubourg de Pont’
This ARP survey of approximately 8ha revealed the
overall organization of an artisan quarter, divided in
several quadrangular 50-m sided blocks and separated
by streets (Figure 3). Buildings were located along a
main road (R1), which is believed to be one of the major
axes of the ancient town. A high resistive anomaly of
10m width marks the pavement, which is bounded for
a distance of 20m by the facades of buildings. The
comparison of the resistivity map with the cadastral
plan of the nineteenth century (Figure 3b) shows that
the artisan quarter divisions were perpetuated in the
land parcelling. A cross-section of the main R1 road
reveals that the pavement was raised several times
(Figure 3c; Mougin et al., 1997). A milestone, indicating
the distance to Vesontio was found at a crossroads (CR)
where another road of smaller dimensions was detected
as a resistive anomaly.
On both sides of R1 road, the artisan suburb area
appears bounded by resistive linear anomalies, which
correspond to a road (R4) and terrace walls already
recognized by a former archaeological survey (Mougin
et al., 1997). Former excavations showed that the streets
in the lower part of the suburb were lined mainly
by artisan workshops of quite diversified activity,
including pottery, metal industry, stone work and
smoked-meat production. The chronological indications
provide evidence of an early occupation beginning at
the first century BC (Augustan period) and continuing
until the third century AD.
After ‘l’Essarté’(Figure 2; Laplaige et al., 2011),
the quarter of ‘Faubourg de Pont’appeared as a
second complex entirely devoted to the production
and transformation activities on the left bank of
the River Doubs.
The residential quarter of ‘Champs des Combottes’
The ARP mapping (1 m depth) of 20 ha provided a clear
overall image of the Roman architectural remains buried
in sedimentary alluvial deposits (Figure 4). Several
anomalies of hydrographic events can also be identified,
265City Map of Ancient Epomanduodurum (Eastern France)
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
Figure 3. Mathay (Doubs). Left bank of the Doubs River. Exploration of the artisan quarter of ‘Faubourg de Pont’. (Acquisition and data processing: Chemin M, Favard A, Trillaud S,
Bossuet G and Thivet M; Geocarta/PCR Mandeure.) This figure is available in colour online at wileyonlinelibrary.com/journal/arp
266 G. Bossuet et al.
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
Figure 4. Mandeure (Doubs). Right bank of the Doubs River. Exploration of the residential area of ‘Champs des Combottes’. (Acquisition and data processing: Trillaud S, Marmet E, Dabas
M, Favard A, Thivet M, Bossuet G; Geocarta/PCR Mandeure.) This figure is available in colour online at wileyonlinelibrary.com/journal/arp
267City Map of Ancient Epomanduodurum (Eastern France)
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
such as prograding sand bars and palaeo-embankment
of the river (Thivet and Bossuet, 2008). The ancient town
is revealed through a rectangular road network and
street system (resistive anomalies R2, R12, R20, R21,
and R22) from 4 to 6 m of width, dividing the prospected
area into space of variable-size (60 m to 80 m wide by
150 m to 200 m long).
Filtering procedures and contrast enhancement have
allowed easy identification of the detailed organization
of insulae (Figure 4). In some places, discrete and linear
anomalies clearly evoke buildings with a columned
portico. However, their precise function (whether
public or private) cannot be determined with certainty.
Construction seems to be concentrated along the roads,
while the interior area of the insulae generally appears
to have been occupied by wide courtyards of variable
size. In the south part of the surveyed area, most of the
structures can be recognized as residential houses. In
plan they look similar to those of the domus (Figure 4b).
Indeed, some have ‘extra rooms’(extensions that could
have been used for craft or domestic activities) whereas
others seem completely empty of specific organization.
Furthermore, it is possible to identify rooms having
terrazzo grounds, mosaic and hypocaust (resistive
anomalies) on the one hand, and others with a mud
floor (rather conductive) on the other hand.
In the north ofthe area prospected, the resistive anom-
alies of buildings with narrow rooms in a row may be
interpreted as storehouses (horrea). Their location along
the ancient bank of the river suggests that they were
probably used for river transport (Figure 4d).
More generally, the internal organization of the
insulae does not submit to a strict plan: this may suggest
a slow progressive construction of the different areas,
the plan of the town evolving according to the urban
status of the agglomeration.
In this residential area of ‘Champs des Combottes’,
three new religious buildings were also detected by the
ARP mapping: they correspond to concentric regular
anomalies (rectangular and square) clearly evoking a
fanum (sanctuary) of native tradition (Figure 4c). The
first temple (NT) shows a square shape of 15 m a
side with a cella of 8 m a side. The south temple (ST),
of quadrangular shape, measuring 15 m by 13 m
includes a cella of about 5m a side. The last fanum
(CRT) of the same dimensions as temple ST is located
at the crossroads of R12 and R22. The semicircular
resistive anomaly partly surrounding the temple (CRT)
could be interpreted as a circulation area (Thivet, 2008;
Barral et al., 2011).
In the road network, the R12 causeway (10 m wide)
running parallel to the river is still preserved in the
topography of the flood plain as an elevation of 1.50 m.
This road could have played the role of a raised dyke
protecting the structures built within the loop from
the floods of the river. The alignment is detected for
more than 800 m by the resistivity mapping, but was
interrupted in a place where the R12 road takes another
direction (Figure 4). The detection of two temples at this
spot could well prove that we are in the presence of a
monumental quarter with a religious function, which
may have required at given times a particular
adjustment of the town planning (Barral et al., 2011).
Religious monumental complex in the southern part of
the ancient town
The monumental complex has been known at least since
the nineteenth century. Located in the south of the
ancient town, the theatre probably was built at the end
of first century AD (Flavian period). Its large dimensions
(frontage length of 142m) made it famous as the second
theatre of Gaules (Marc et al., 2007a). Before our
geophysical survey, the Roman sanctuary of ‘Cloux du
Château’was the only religious building known in
Epomanduodurum. Its location in front of the theatre
was similar to those of the neighbouring religious
complexes of ‘Cigognier’in Aventicum (Avenches) and
of ‘Schönbühl’in Augusta Raurica (Augst) (Joly, 2007;
Marc et al., 2007b). In the nineteenth and twentieth
centuries, the excavations produced a detailed plan of
the Roman sanctuary (Duvernoy, 1883). A Celtic
sanctuary lying under the Roman temple was attested
by the uncovering of two important sacred collections
of coins, bracelets and coloured glass rings. Study of
these artefacts dated the sanctuary from the Middle
Tène and shows a frequent use of the site during the
Final Tène (Barral et al., 2009).
The geophysical exploration of the surroundings
of the ancient theatre used a combination of three
methods (electric, magnetic and GPR), which produced
a plan showing a vast religious complex of approximately
10 ha in the southern part of the ancient town (Figure 5;
Thivet, 2008; Barral et al., 2011). Additionally, an extensive
view of the palaeohydrography was obtained in this area
by the mapping of abandoned meanders of the river
(Figure 5a; Thivet and Bossuet, 2008; Thivet et al., 2011a).
Buildings with a central plan at ‘Les Ouchottes’
In the eighteenth and nineteenth centuries, several
observations with no accurate location data noted the
presence of baths and various vestiges related to the
collection (well), distribution (aqueduct) and storage
of the water (basin) (Dunod, 1709; Morel-Macler,
1847; Duvernoy, 1875). In 2001 a group of structures
(temple, thermae?) were detected on the top of the
268 G. Bossuet et al.
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
Figure 5. Mandeure (Doubs). South quarter of the ancient town. Exploration of the monumental area of the religious complex. (Acquisition and data processing: Trillaud S, Hullin G, Sarro
L, Favard A, Bossuet G, Thivet M; Geocarta/PCR Mandeure.) This figure is available in colour online at wileyonlinelibrary.com/journal/arp
269City Map of Ancient Epomanduodurum (Eastern France)
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
alluvial terrace by magnetic prospecting. Later, elec-
tric mapping and GPR made the plan of these struc-
tures more precise (Figure 6) (Bossuet et al., 2007).
The remains of ‘Les Ouchottes’include buildings
with a central plan (1a in Figure 5b) surrounded by
polygonal enclosures probably delimiting the sacred
area perimeters (1b in Figure 5b). The rectangular
building (31 by 24 m) with semicircular exedres and
colonnaded portico (about 5 m wide) clearly evokes
the plan of a temple (T1) (Figure 6a and c). In the north,
a building (T2) of square shape and smaller dimension
(23 m each side) could correspond to a second temple
with a more classic plan. Some of the reflective and
resistive anomalies registered by GPR and ARP
mapping revealed the remains (T3) of another
underlying building (Figure 6b and d).
On the floodplain terrace a monumental polygonal
enclosure was identified as a resistive anomaly
(1b1 in Figure 5b). The wall enclosure appears
interrupted by three square anomalies of 7m aside
which can be interpreted as masonry piers.
A polygonal ditched enclosure (1b2) was detected
higher in the slope of the terrace.
A third enclosure (1b3) was identified as a straight
wall running from the frontage of the theatre to the top
of the alluvial terrace; the enclosure was interrupted by
an opening in front of magnetic and resistive anomalies
that were interpreted as pits, canalization and water
basin. Furthermore, a row of enigmatic conductive
anomalies (6a1) of large-dimension (4 m 4m) was
detected below the alluvial terrace.
The interpretation plan of geophysical data at ‘Les
Ouchottes’shows some similarities with the excavation
plan of the Lingon sanctuary of Mirebeau-sur-Bèze
(eastern Burgundy), which functioned from 200 BC to
AD150 (Joly and Barral, 2007).
The new sanctuary of ‘Champs des Fougères’
To the northwest of the theatre, at about 250 m from
the great Roman sanctuary of ‘Les Cloux du Château’,
magnetic and resistivity mapping have detected two
circular concentric anomalies interpreted as architectural
structures of a newly discovered religious building
(2a in Figure 5b; Bossuet et al., 2007). Indeed, several
fragments of colossal acrolith statues of Mars and
Bellona divinities were found in the area of ‘Champs
des Fougères’in 1889 (Bossuet et al., 2007). Recent
excavations confirmed our interpretation by uncovering
relics of a monumental sanctuary with oval peribolos,
partially destroyed by modern constructions (Thivet
and Nouvel, 2009). It was conserved as two parallel
walls defining an oval polygonal enclosure of 70 m by
64m. Six phases in the evolution of the sanctuary were
precisely characterized; they took place between the
Final Tène period, when sacred areas were first
delimited by fences, and the Flavian period, in the course
of which the peribolos was probably built. The religious
complex seems to have been abandoned around AD 350.
Underlying the sanctuary, the excavations uncovered
three kilns and a scattering of pits, respectively detected
as bipolar and positive magnetic anomalies (6a2 in
Figure 5b). The kilns with a furnace and central pillar,
dated from La Tène D2, correspond to the most ancient
artisan structures found on the site (Figure 7). They
attest to an early occupation, prior to the conquest in this
part of the Roman town (Thivet and Nouvel, 2009).
TheLateEmpirefortification of ‘Les Cloux du Château’
About 60 m west of the sanctuary at ‘Champs des
Fougères’, the remains of a military fortification super-
imposed on an earlier occupation (first century AD)were
detected by magnetic prospecting and ARP mapping
(Bossuet et al., 2007; Thivet, 2008) (Figure 8). The exist-
ence of a Late Empire fortification was deduced in
1960 by Y. Jeannin from the 1836 cadastral plan and
numerous objects from the fourth century AD found
over the area (Jeannin, 1986). Geophysical prospecting
revealed that the ditch enclosure (2b in Figure 5b)
and the surrounding wall of the castrum were built on
regular curves separated by a distance of 15m. On the
magnetic mapping, the interior of the fortification was
partly occupied by a square block of approximately
60 m by 60 m, probably built along a wide road
(Figure 8a). The excavations revealed a defensive
system comprising a ditched enclosure (12 m wide
and about 2.50 m deep) and a rampart foundation
(4 m wide). The re-employment of massive architectural
blocks for the basement construction showed that the
closest public monuments (sanctuaries, theatre) were
dismantled in this particular case.
The overall plan of the fortification can be traced
precisely from geophysical prospecting, excavations,
relief maps and nineteenth century plans (Figure 8b;
Bossuet et al., 2007; Kuhnle et al., 2007). Built on the right
bank of the River Doubs, the fortification stretches over a
large area (3 ha); its topography suggests a bell-shaped
structure similar to other river fortresses, such as at
Chalon-sur-Saône in France or Brugg, Olten or Solothurn
in Switzerland, the latter site being the closest parallel to
Mandeure ‘Les Cloux du Château’(Kuhnle et al., 2007;
Cramatte et al., 2012).
The vestiges of military thermal baths (legion Prima
Martia) and barracks built along the internal rampart
of the castrum were uncovered recently by excavations
(Cramatte et al., 2012). The rampart was preserved
as two semicircular towers as well as a gate on the
270 G. Bossuet et al.
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
Figure 6. Mandeure (Doubs). ‘Les Ouchottes’. Complementary geophysical methods used in the exploration of the southern Roman theatre. (Acquisition and data processing: Lacaze S,
Bossuet G, Valet J, Favard A, Chassang M, Thivet M, Aubry L; Geocarta/PCR Mandeure.) This figure is available in colour online at wileyonlinelibrary.com/journal/arp
271City Map of Ancient Epomanduodurum (Eastern France)
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
Figure 7. Mandeure (Doubs). ‘Champs des Fougères’. Exploration of the new sanctuary. (Acquisition and data processing: Barrès E, Bossuet G, Thivet M, Nouvel P; Geocarta/PCR
Mandeure.) This figure is available in colour online at wileyonlinelibrary.com/journal/arp
272 G. Bossuet et al.
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
Figure 8. Mandeure (Doubs). ‘Champs des Cloux du Château’. Exploration of the late Roman Empire fortification. (Acquisition and data processing: Dabas M, Aubry L, Bossuet G, Thivet
M, Kuhnle G; Geocarta/PCR Mandeure.) This figure is available in colour online at wileyonlinelibrary.com/journal/arp
273City Map of Ancient Epomanduodurum (Eastern France)
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
southern and eastern front of the fortification. This
configuration is similar to those of the western gate
of the castrum of Kaiseraugst (Argovie, Switzerland)
(Kuhnle et al., 2007; Cramatte et al., 2012). Within the
castrum area, a road (5d in Figure 5b) detected by
geophysical prospecting connects the front gate of the
fortification to a presumed ancient bridge on the River
Doubs; this road (5d in Figure 5b) was also identified
as an elevation in the relief map of the fortification
(Figure 8b).
The first buildings were built in the middle of the
fourth century AD (Kuhnle et al., 2007). The occupation
of the castrum area in the fifth century AD is also well
attested by ceramics. The underlying structures of artisan
and domestic activities were dated from the middle of
the first century AD to the beginning of the third century
AD. Evidence of Final Tène occupation was also identified
in this area (Cramatte et al., 2012).
The great Roman sanctuary of ‘Cloux du Château’
The great Roman sanctuary of ‘Cloux du Château’is
located 250 m in front of the theatre. At the present
time it is largely covered by woodland. The resistivity
survey provides new information about the monument
with regard to the descriptions and excavation plans of
the nineteenth century.
For example, the ARP mapping showed three parallel
high resistive anomalies (3a inFigure 5b) of 66m length,
which correspond to the triple wall of the south podium
frontage of the sanctuary. An area of circulation (3b in
Figure 5b) enclosing the oval peribolos of the sanctuary
was clearly revealed as a polygonal resistive anomaly
of 13m width. Moreover the lower resistive anomalies
(3c in Figure 5b) could correspond to the underlying
vestiges of the Latenian sanctuary recognized as dry-stone
walls at the end of the nineteenth century (Duvernoy,
1883). On the western front of the monument, the area
of circulation (3b in Figure 5b) was connected to a
crossroads leading to an ancient ford where the river
was easy to cross.
Further north, a quadrangular structure of 20 m by
20 m, closed on three sides by a double resistance
anomaly, suggests the plan of a building with a gallery
(4a in Figure 5b). This structure (a temple?) was built along
anEarlyEmpireroadalsodetectedasaresistiveanomaly
(5e in Figure 5b). On the east side of the sanctuary two
square conductive anomalies (4b in Figure 5b) of 15 m
by 15 m were interpreted as foundation trenches for
building construction: they could correspond to annexes,
chapels or altars.
The linear resistive anomalies (5b and 5c in Figure 5b)
were recognized as two parallel roads converging on
the eastern entrance of the sanctuary (Bossuet et al.,
2007). As for anomaly 5b, an exploration trench revealed
a Roman road superimposed on an underground
canalization leading to the promontory of the monument
(Figure 5c). The cross-section also showed that the road
(5c in Figure 5b) was the most ancient: as with the
structures (4c in Figure 5b) detected in front of the theatre,
this road was built over a palaeosol dated to 390–170 cal
yr BC by AMS radiocarbon dating (Bossuet et al., 2007).
Finally, a set of linear resistive anomalies (5a in
Figure 5b) delimiting a monumental area on three
sides was interpreted as the main wall enclosure of a
vast religious complex including several temples or
groups of temples radiating from the theatre. This
interpretation was proven by the excavations uncovering
the archbay of a monumental passage (7 in Figure 5)
built during the second century AD through the
surrounding wall (5a in Figure 5b). This monumental
passage connected public space in the west to another
large area containing sacred buildings and a theatre
to the east.
The peripheral occupation on the ‘Lomont Plateau’
Earlier observations and aerial photographs showed
that the ‘Lomont Plateau’occupied a large area running
up to the ancient theatre (Thivet, 2008; Laplaige, 2011).
Magnetic and ARP mapping were performed over an
area of about 5 ha (Figure 9).
To the north, two groups of quadrangular structures
were detected on the curving valley side. They were
separated by linear anomalies probably marking a road
oriented in a north–south direction. These buildings,
spreading out from the bottom to the top of the plateau,
were interpreted as storehouses and shops; they could
be connected to annexes recently uncovered behind
the theatre. Numerous jugs and tumblers as well as
a big dump of cattle rib bones suggest that these
annexes were used possibly for banquets celebrating
an imperial cult. The other resistive and conductive
anomalies detected in this northern area correspond
to other archaeological structures (walls, basin and
ditches), indicating that the level ground of the plateau
was also occupied.
Conductive anomalies detected in the south of the
prospected area are caused by the fracture of the calcar-
eous bedrock; but they are mixed with anthropogenic
anomalies, some of them corresponding to holes, ditches
and a road, this last one joining the double ditch of a
quadrangular enclosure. This enclosure (55 m by 63 m)
must be interpreted as a military camp, possibly a
cohort camp from the early Imperial period. Its plan
appears similar to those of Limberg near Sasbach and
Oedenburg (Bisheim) in the Alsatian plain (Reddé et al.,
274 G. Bossuet et al.
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
Figure 9. Mandeure (Doubs). Peripheral occupation of the ‘Lomont Plateau’. (Acquisition and data processing: Trillaud S, Laplaige C, Bossuet G; Geocarta/PCR Mandeure.) This figure is
available in colour online at wileyonlinelibrary.com/journal/arp
275City Map of Ancient Epomanduodurum (Eastern France)
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
2006). If the hypothesis of a camp is verified, it would be
the first permanent Roman presence identified on the site
of Epomanduodurum.
The urban plan of the ancient
Epomanduodurum
The geophysical survey results have profoundly
changed our knowledge of the plan and extent of the
ancient city of Epomanduodurum, revealing in full detail
the organization of the remains still buried, as we
can see on the general ARP mapping (Figure 10). The
combined interpretation of the geophysical prospecting,
aerial photographs and excavations produce a new
urban plan of the ancient town. It shows that
the ancient town occupied more than 500 ha
between Mathay ‘l’Essarté’and Mandeure ‘Courcelles’,
and half this space at least was densely occupied
(Figure 11).
Figure 10. Mandeure–Mathay (Doubs). Resistivity survey of a part of ancient Epomanduodurum. Automatic resistivity profiling system: 1 m deep, area of
60 ha. Adjusted resistivity scale: resistivity value in dark, conductivity value in light. (Acquisition and data processing: Geocarta/PCR Mandeure.) This
figure is available in colour online at wileyonlinelibrary.com/journal/arp
276 G. Bossuet et al.
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
Figure 11. Mandeure–Mathay (Doubs). Location of the remains of ancient Epomanduodurum. Background picture is a LiDAR hillshade 315_45 (2009). (Acquisition and data processing: Laplaige C,
Bossuet G, Thivet M; PCR Mandeure/LIEPPEC.) This figure is available in colour online at wileyonlinelibrary.com/journal/arp
277City Map of Ancient Epomanduodurum (Eastern France)
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
The urban network was structured according to three
major axes (R1, R2, and R3), with all three alignments
allowing for optimal use of the space inside the alluvial
plain. The first axis (R1) corresponds to a road of 20 m
width with a northwest–southeast orientation. In the
northern section, the road bends westward towards the
road leading to the civitas capital of Vesontio (Besançon)
through the Romont plateau. In its southern section,
the road (R1) leads to a ford where another road (R2)
converges on the right bank of the river. This road dated
from the early Roman Empire when the ancient town
was structured within the loop of the River Doubs.
In the northern part of the Castrum the urbanization
becomes denser and is organized around secondary
streets delimiting insulae of various dimensions. So
defined, the urban spread shows a homogeneous
extension within the loop. The excavations of the
ancient quarter of ‘La Récille’suggest that the town
stretched eastward to the thermae of Courcelles
(Kunhle et al., 2006) (Figure 11). Two perpendicular axes
(R17) and (R20) in the main road R2, connect the centre
of the town with the eastern suburb of ‘Courcelles’.
In the cadastral plan of 1836, these roads converge to
circular and rectilinear anomalies that remind us of the
plans of the two temples built in front of the theatre
(Thivet et al., 2009). The surrounding remains suggest
that the suburb of ‘Courcelles’could correspond to a
monumental area organized around a spring and a
religious and bath complex (Blin, 2009).
The Rhine road (R3) cuts the meander of the river
from west to east. This major axis was lined with streets
of south–north orientation. It allowed direct access to
‘Courcelles’from the theatre–sanctuary complex by
avoiding the urban centre of the Roman town.
The street network of the antique town extends to
the summit of the ‘Lomont Plateau’which overlooks
the ancient theatre. Further upstream on the alluvial
plain, the street network of the artisan quarter of
Mathay ‘l’Essarté’also showed the same north–south
orientation (Laplaige et al., 2011). In spite of the
evidence that an ancient road network connected
‘l’Essarté’to the suburb of ‘Faubourg de Pont’,no
urban occupation was identified between these two
settlements. This gap could be related to the swampy
character of this area, which was easily flooded
during the Roman period (Jeannin, 1986).
The urbanism scheme is organized according to two
preferential orientations: the first one (of northwest–
southeast direction) predominating in the loop of the
river and in the suburb of ‘Faubourg de Pont’,whereas
the second orientation (of north–south direction) is
centred mainly along the Rhine road (R3) and in the
artisan quarter of ‘l’Essarté’. At present, no chronological
information can be deduced from these different orienta-
tions (Jeannin, 1986; Frezouls, 1988; Mougin, 1991). They
could be contemporary and the difference in orientation
might be the result of topographic constraints.
Conclusion
At Epomanduodurum, geophysical investigation has
greatly revised our vision of the organization of the
Roman town, especially for the monumental complex
sanctuar–theatre. The detection of numerous unknown
structures, such as two new sanctuaries, provides
evidence of a major transformation of this area occurring
soon after the Roman conquest. The plan of this
monumental area has the same form as most religious
complexes in eastern Gaules, the closest examples
being Augusta Raurica (Augst, Switzerland), Aventicum
(Avenches, Switzerland) and Augusta Treverorum
(Trier, Germany) (Marc et al., 2007a; Thivet, 2008).
This religious monumental complex of 10 ha can be
considered as one of the most impressive of Gaules and
Germania (Blin, 2009). The status and influence of the
Latenian sanctuary located in the centre of this large
monumental complex from the early Roman Empire
seemstohaveplayedakeyroleintheemergenceofa
major settlement of Epomanduodurum (Barral et al.,
2009, 2011; Marc et al., 2007a).
On a large scale, the combination of different methods
has permitted us to map fully the extent and
organization of ancient Epomanduodurum over more than
five centuries of occupation. In this research, geophysical
prospecting has played a major part in bringing detailed
information to light on the dense urbanism scheme
of this ancient town: sanctuaries, temples, a castrum
fortification and a Roman fort, thermal baths, domus,
storehouses, workshops and a local system of roads
and streets network of insulae. All of the information
has enhanced the spatial pattern of buried remains,
which previously were deduced mainly from aerial
photographs and limited trial excavations.
The archaeological evaluation of Epomanduodurum
again demonstrates that geophysical prospecting
provides a suitable technique of investigation for the
study of complex and extensive sites. More generally
the use of an integrated approach has offered new
possibilities for the recording, visualization and analysis
of archaeological environments and landscapes in this
part of the Franche-Comté region. One of the benefits
of integrating different survey and remote sensing
methods (LiDAR, geophysical prospecting and aerial
photography) was to provide detailed information about
the drainage pattern, thereby providing a more
278 G. Bossuet et al.
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
comprehensive the study of the relationship between
the Roman urbanization and the dynamics of the
River Doubs.
Acknowledgements
Financial support for this research project was provided
by the French Ministry of Culture, the Regional Council
of Franche-Comté, the General Council of the Doubs
Department and Country Montbéliard Agglomération.
TheCollectiveResearchProject‘Interdisciplinary
research of Antique urban area of Epomanduodurum,
Mandeure–Mathay (Doubs). Archaeology, landscape
and environment (Coordinator Ph. Barral) started in
2001. The project was initiated by the University of
Franche-Comté (UMR Chronoenvironnement CNRS).
The authors express their sincere thanks to the research
groups from Universities of Franche-Comté, Strasbourg,
Lausanne, Paris IV Panthéon-Sorbonne, P. & M. Curie
Paris VI, and Bourgogne.
References
Barral P. 1996. La nécropole gauloise des Longues Raies à
Mathay (Doubs). Résultats de la campagne de 1995.
Bulletin de la Société d’Emulation de Montbéliard 119:
193–203.
Barral P, Bossuet G, Kuhnle G, et al. 2007. Epomanduo-
durum, une ville chez les Séquanes: bilan de quatre
années de recherche à Mandeure et Mathay (Doubs).
Gallia 64: 353–434, et pl. H. T. IV à XV.
Barral P, Bossuet G, Thivet M, Jacottey L, Pichot V. 2009. Les
dépôts du sanctuaire de Mandeure (Doubs). In: Actes de
la Table ronde internationale de Neuchâtel Le site de La
Tène: bilan des connaissances –état de la question
(1er –3 novembre 2007) sous la direction de Matthieu
Honegger, Denis Ramseyer, Gilbert Kaenel, Béat Arnold
et Marc-Antoine Kaeser. Archéologie neuchâteloise 43:
185–195.
Barral P, Thivet M, Bossuet G, et al. 2011. Les sanctuaires
de Mandeure. In Aspects de la romanisation dans l’Est de
la Gaule (publication dans le cadre du PPF ‘Les Sociétés
antiques dans l’Est de la Gaule: mutations politiques,
socio-économiques et environnementales à la finde
l’Âge du Fer et au début de l’époque romaine’), Reddé
M (director). Glux-en-Glenne: Bibracte (Coll. Bibracte)
2011, 21, Vol. 2: 557–567.
Blin S. 2009. Architecture et Société à Mandeure au Haut-
Empire. Bulletin de la Société d’Emulation de Montbéliard
131:23–48.
BossuetG,ThivetM,DabasM,et al. 2007. Evolution
morphosédimentaire de la plaine alluviale et
aménagements anthropiques aux abords du complexe
monumental sanctuaire-théâtre d’Epomanduodurum.In
EPOMANDUODURUM, une ville chez les Séquanes:
bilan de quatre années de recherche à Mandeure et
Mathay (Doubs). Gallia 64:360–371, plates H. T. IVand XV.
Buteux S, Gaffney V, White R, Van Leusen M. 2000. Wroxeter
hinterland project and geophysical survey at Wroxeter.
Archaeological Prospection (Special Issue) 7(2): 69–80.
Cramatte C, Glaus M, Mamin Y. 2012. Dossier: une église
du 5e siècle dans le castrum de Mandeure (F). Archéologie
suisse 35(1): 4–15.
Dabas M. 2006. Theory and practice of the new fast electrical
imaging system ARP. XV International Summer School
in Archaeology, Geophysics for Landscape Archaeology,
Grossetto, 10–18 July; 13 pp.
Dabas M, Tabbagh A. 2003. A comparison of EMI and DC
methods used in soil mapping –theoretical considerations
for precision agriculture. In Precision Agriculture,Stafford
J, Werner A (eds). Academic Publishers Muencheberg:
Wageningen ; 121–129.
Dabas M, Perrard C, Favard A, Rouiller D. 2002. A new
system for fast mapping of ER (A.R.P.): field application.
6th International Conference on Agriculture and Other
Precision Resources Management,14–17 July, Bloomington.
Dabas M, Guyart L, Lepert T. 2005. Gisacum revisité.
Croisement géophysique et archéologie. Les Dossiers
d’Archéologie 295:52–61.
Doneus M, Eder-Hinterleitner A, Neubnauer W. 2001.
Roman Carnuntum –prospecting the largest archaeo-
logical landscape in Austria. In Archaeological Prospection,
4th International Conference on Archaeological Prospection,
Doneus M, Eder-Hinterleitner A, Neubnauer W (Hrg.).
Austrian Academy of Sciences Press: 47–55.
Dunod J-M. 1709. La découverte entière de la villa d’An-
tre en Franche-Comté qui fait changer de face à l’histoire
ancienne, civile et ecclésiastique de la même province et
des provinces voisines.Amsterdam:ThomasLombrail;
175–181.
Duvernoy C. 1875. Notice sur le pays de Montbéliard
antérieurement à ses premiers comtes. Mémoire de la
Société d’Emulation de Montbéliard 4(2): 1–165.
Duvernoy C. 1883. Notice sur une enceinte récemment
découverte à Mandeure. Mémoire de la Société des
Antiquaires de France 4(5): 29–44.
Frezouls E. 1988. Mandeure antique. In Les villes antiques
de la France, AECR edn. Frezoul E (dir.). Germanie
Supérieure 1, Besançon–Dijon–Langres–Mandeure:
Strasbourg; 423–505.
Jeannin Y. 1986. Mandeure (Doubs). In Les agglomérations
secondaires en Franche-Comté romaine, Mangin M, Jacquet
B, Jacob J-P (dir.). Les Belles Lettres, Annales littéraires
de l’université de Besançon: Paris; 38–67.
Joan L. 2003. Le Doubs et le territoire de Belfort: 25 et 90. Carte
Archéologique de la Gaule. Académie des Inscriptions et
Belles Lettres: Paris; 561 pp.
Joly M. 2007. Le sanctuaire gallo-romain. In Epomanduo-
durum, une ville chez les Séquanes: bilan de quatre
années de recherche à Mandeure et Mathay (Doubs).
Gallia 64: 390–395.
Joly M, Barral P. 2007. Le sanctuaire de Mirebeau-sur-
Bèze (Côte-d’Or): bilan des recherches récentes. In L’âge
du Fer dans l’arc jurassien et ses marges. Dépôts, lieux
sacrés et territorialité à l’âge du Fer, Barral P, Daubigney
A, Dunning C, Kaenel G, Rouliere-Lambert M-J (dir.).
Actes du colloque international de l’AFEAF, Bienne,
5–8 mai 2005, Vol. 1. Presse universitaire de Franche-
Comté: Besançon; 55–72.
Kvamme KL. 2003. Geophysical surveys as landscape
archaeology. American Antiquity 68(3): 435–457.
KunhleG,HumbertS,FortB,et al. 2006. Mandeure
Rue de Récille: un quartier périurbain inédit
279City Map of Ancient Epomanduodurum (Eastern France)
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
d’Epomanduodurum.Bulletin de la Société d’Emulation de
Montbéliard 129:25–88.
Kuhnle G, Burgevin A, Fort B, Mazimann J-P, Mouton S,
Dard I, Gelot J, Mougin P. 2007. La fortification du
Bas-Empire in EPOMANDUODURUM, une ville chez
les Séquanes: bilan de quatre année de recherche à
Mandeure et Mathay (Doubs). Gallia 64: 407–426, plates
H. T. IV and XV.
Lame M, Mazimann J-P. 1993. L’atelier de potier du
‘Champ des Isles’à Mathay (Doubs) et sa production.
Revue Archéologique de l’Est 44(2): 429–469.
Laplaige C. 2011. Comparaison de signaux (géophysique,
LiDAR) utilisés dans l’étude des dynamiques anthropiques
et naturelles. Thèse de Doctorat, Besançon, Université de
Franche-Comté, 2 Vols.
Laplaige C, Fruchart C, Bossuet G, Opitz R. 2011. Apports
des prospections LiDAR, magnétiques et pédestres à la
connaissance du quartier artisanal antique de l’Essarté
(Mathay, Doubs). Revue archéologique de l’Est 60:523–535.
Marc J-Y, Andre N, Barral P, Blin S, Bossuet G, Joly M,
Mougin P, Thivet M. 2007a. Le complexe du sanctuaire
et du théâtre de Mandeure (Doubs, F), de ses origines
gauloises à sa monumentalisation romaine. In Premières
journées archéologiques frontalières de l’Arc jurassien. Actes.
Delle (F) –Boncourt (CH), 21–22 octobre 2005. Mandeure,
sa campagne et ses relations d’Avenches à Luxeuil et d’Augst
à Besançon. Actualités archéologiques régionales. PUFC,
Belet-Gonda C, Mazimann JP, Richard A, Schifferdecker F
(dirs). Besançon et Cahier d’archéologie jurassienne:
Porrentruy; 13–34.
Marc J-Y, Mougin P, Blin S. 2007b. Le théâtre antique. In
Epomanduodurum, une ville chez les Séquanes: bilan
de quatre années de recherche à Mandeure et Mathay
(Doubs). Gallia 64: 396–407.
Millotte J-P, Lambert GN. 1996. Le Pays de Montbéliard et
la Porte de Bourgogne et d’Alsace: voie de passage ou
frontière culturelle aux âges des métaux. Bulletin de la
Société d’Émulation de Montbéliard 119:3–50.
Morel-Macler F. 1847. Antiquités de Mandeure. Montbéliard:
Montbéliard; 22 pp.
Mougin P. 1991. Le site archéologique de Mandeure.
Fouilles récentes 1988–1990. Bulletin de la Société d’
Emulation de Montbéliard 114: 187–227.
Mougin P (dir.), Vaxellaire L, Watts D. 1997. Mathay
(Doubs) SHUNT, Franche-Comté, Document final de
synthèse. Diagnostic sur le tracé de la déviation nord de
Mathay. SRA Franche-Comté. S.I.V.U Mandeure Mathay.
A.F.A.N Grand Est: 89 pp.
Neubauer W, Seren S, Hinterleitner A, Löcker K, Melichar P.
2009. Archaeological interpretation of combined magnetic
andGPRsurveysoftheromantownFlaviaSolva,
Austria. In Archeo Sciences Revue d’archéométrie,GMPCA.
Supplément au n33-2009. Mémoire du sol, Espace des
hommes. Presses Universitaires de Rennes: Rennes;
225–228.
Reddé M, Brulet R, Fellmann R, Haalebos JK, von Schnurbein
S (dir.). 2006. Les fortifications militaires. In L’architecture de
la Gaule romaine. MSH-Ausonius, Documents d’archéolo-
gie française 100: Bordeaux; 477 pp.
Scollar I, Tabbagh A, Hesse A, Herzog I. 1990. Archaeo-
logical Prospecting and Remote Sensing. Cambridge
University Press: Cambridge; 674 pp.
Sensors and Software. 2002. EKKO_Mapper User’s Guide
Version 2.0, Mississauga; 48 pp.
Tabbagh J. 2001. Filtre directionnel permettant d’éliminer
les anomalies crées par le labour. Filtering optimisation
and modelling of Geophysical data. In Archaeological
Prospecting, Fondazione ing. Carlo Maurilio Lerici Politecnico
di Milano, 50th Anniversary;161–166.
Tabbagh J. 2003. Total field magnetic prospection: are
vertical gradiometer measurements preferable to
single sensor survey? Archaeological Prospection 10:
75–81.
Thivet M. 2008. Méthodes nouvelles de l’archéologie
appliquées au site antique de Mandeure- Mathay,
(Doubs). Reconnaissance spatiale, évolution chronologique,
statut urbain. Thèse de Doctorat, Besançon, Université
de Franche-Comté.
Thivet M, Bossuet G. 2008. Les aménagements des rives
du Doubs à Epomanduodurum. Impact de l’aléa fluvial
sur le développement d’une agglomération antique.
Archeopages, Rives et riverains Octobre:34–37.
Thivet M, Nouvel P. 2009. L’occupation laténienne du
sanctuaire des Champs des Fougères à Mandeure
(Doubs), Campagnes 2007 et 2008. Bulletin de
l’Association Française pour l’Etude de l’Age du Fer 27:
59–63.
Thivet M, Bossuet G, Mougin P. 2009. Mise en évidence de
l’impact de l’occupation antique dans la morphogenèse
d’un paysage de fond de vallée: le cas du site
d’Epomanduodurum (Mandeure-Mathay, Doubs). Revue
Archéologique de l’Est 58:417–437.
Thivet M, Bossuet G, Trillaud S, et al. 2011a. Mandeure:
impact de l’aléa fluvial sur le développement de
l’agglomération antique. In Aspects de la romanisation dans
l’Est de la Gaule (publication dans le cadre du PPF ‘Les
Sociétés antiques dans l’Est de la Gaule: mutations
politiques, socio-économiques et environnementales à la
findel’Âge du Fer et au début de l’époque romaine’),
Reddé M (director). Glux-en-Glenne: Bibracte (Coll.
Bibracte) 2011, 21, Vol. 1: 145–153.
Thivet M, Nouvel P, Barral P, Bossuet G. 2011b. Mandeure:
du site gaulois à la ville romaine. In Aspects de la
romanisation dans l’Est de la Gaule (publication dans le
cadre du PPF ‘Les Sociétés antiques dans l’Est de la
Gaule: mutations politiques, socio-économiques et
environnementales à la findel’Âge du Fer et au
début de l’époque romaine’), Reddé M (director).
Glux-en-Glenne: Bibracte (Coll. Bibracte) 2011, 21,
Vol. 1: 153–161.
280 G. Bossuet et al.
Copyright © 2012 John Wiley & Sons, Ltd. Archaeol. Prospect. 19, 261–280 (2012)
DOI: 10.1002/arp
