Tectonophysics 331 (2001) 8197

www.elsevier.com/locate/tecto

Gravity modellings of the Cadomian active margin of northern

Brittany
Catherine Truffert*, Emmanuel Egal 1, Elisabeth Le Goff 2, Gabriel Courrioux 1,
Pol Guennoc 1
BRGM, Avenue C. Guillemin, BP 6009, 45060 Orleans La Source, France
Received 1 April 1998; accepted 8 October 1999

Abstract
The geometrical relationships existing between the Saint-Brieuc, Saint-Malo and Guingamp units, which form the Cadomian
active margin of northern Brittany, have been determined by gravity analyses and 2.5D modelling, correlated with other
geophysical (magnetic, seismic, etc.) and geological observations. Automatic structural analyses were performed before
modelling, to identify lateral gravity discontinuities. Averaged density measurements were used for 2.5D gravity modelling
along six proles across the arcuate geological and gravity shape of the Cadomian basement. This study allowed the third
dimension to be obtained through correlating each prole with its neighbours. We propose that orthogneiss formations
equivalent to the Port-Morvan orthogneiss, constitute the Saint-Brieuc upper-crustal unit, below the Lanvollon Formation.
2.5D gravity modelling outlines downward geological differences in the western and eastern parts of the Ploufragan Complex,
as previously indicated by surface mapping. These results stress the necessity for distinguishing between the Guingamp and
Saint-Malo units. q 2001 Elsevier Science B.V. All rights reserved.
Keywords: Gravity modelling; Cadomian belt; volumic reconstruction

1. Introduction the best-preserved Cadomian block of Western

Europe. The Upper Proterozoic units constituting
The present study was carried out under the Armor this block were only weakly affected by later Variscan
project (GeoFrance 3D), the main objectives of which events. One can distinguish ve main units from north
were (i) to identify the geometry of the major tectonic to south: the Tregor batholith, the Saint-Brieuc Unit,
contacts bounding the Cadomian units of northern the Saint-Malo Unit, the Guingamp Unit, and the
Brittany, (ii) to evaluate the depth of the different Mancellian Domain (Fig. 1). Former geodynamic
units, and (iii) to dene the nature of their hypotheti- syntheses of the area (Chantraine et al., 1988; Dissler
cal basement. et al., 1988; Strachan et al., 1989, 1996; Rabu et al.,
The study area around Saint-Brieuc Bay presents 1990; Graviou, 1992; Egal et al., 1996; Chantraine
et al., 2001) have led to a general model, which
* Corresponding author. Fax: 133-2-38-64-33-34. emphasises an active margin in relation with a
E-mail addresses: [email protected] (C. Truffert), subduction process.
[email protected] (E. Egal), [email protected] (E. Le Goff),
(1) The Saint-Brieuc Unit consists of superimposed
[email protected](G.Courrioux),[email protected](P.Guennoc).
1
Fax: 133-2-38-64-33-34. formations comprising, from top to bottom (i.e. from
2
Fax: 133-2-38-64-36-52. north to south), volcano-sedimentary rocks (Paimpol
0040-1951/01/$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.
PII: S 0040-195 1(00)00237-7
82 C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197

Fig. 1. Geological map of the study area (after Chantraine et al., 2001) showing the location of the six 2.5D gravity model proles discussed in
this work. (1) Lanvollon basic metavolcanics; (2) Lanvollon acid metavolcanics; (3) Erquy volcano-sedimentary formation; (4) Binic meta-
sediments; (5) Port-Morvan orthogneiss; (6) Morieux Formation; (7) Fort La Latte tonalite; (9) Saint-Quay diorite; (8) Saint-Alban metagab-
bro; (11) Yfniac Complex; (10) Squifec metagabbros; (12) Leucogranites; (13) Anatexite and migmatite; (17) Lamballe schist; (14) Saint-
Brieuc diorite; (15) Saint-Malo migmatites; (16) Tregomar gabbro; (18) Syntectonic granites; (19) Chateau-Serein volcanics; (20) Brioverian
sediments; (21) Hercynian granites.

and La RocheDerrien Formations) and bimodal in the east it is represented only by the Yfniac
volcanic sequences (Lanvollon Formation) intruded Complex (Chantraine et al., 2001). The unit is
by numerous plutons and an orthogneiss complex bounded in the north and in the south by two major
(Port-Morvan and Morieux Formations; Egal et al., contacts, namely the Plouagat-Coetmieux and Belle-
1996; Chantraine et al., 2001). These formations are IsleLa Fresnaye Faults.
locally overlain by Brioverian clastic rocks, inter- (3) The Saint-Malo Unit comprises two main
preted as turbidite deposits (Binic Formation; Denis, formations the Lamballe metasedimentary
1988). This unit is bounded by the Losquemeau- complex and the Saint-Malo migmatite bounded
Lezardrieux Fault in the north and by the Plouagat- by the Belle-IsleLa Fresnaye and Plouer-Cancale
Coetmieux Fault in the south. Faults, the unit is offset by the North Armorican
(2) The Guingamp Unit contains magmatic Shear Zone (NASZ).
(migmatite and leucogranite) and basic (Yfniac (4) The Mancellian Domain, separated from the
Belle-Isle Complex) formations in the west, whereas Saint-Malo Unit by the Plouer-Cancale Fault, is
 C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197 83
22
limited in the south by the NASZ. Post-phtanitic sedi- around 0.5 km , except on the two selected targets
ments are the main constituents of this domain. (Binic and Erquy) where the distribution reaches
This paper examines the geometric relationships 2 km 22 due to the additional measurements. We
existing between the Saint-Brieuc, Saint-Malo and linked the newly acquired data to the IGSN71 refer-
Guingamp units, based on gravity analyses and ence system and reduced them with a density of
modelling, correlated with complementary geophysi- 2700 kg m 23 for the Bouguer correction which is the
cal (magneto-telluric, magnetic, seismic, etc.) and mean density of the mancellian formations. Latitudes
geological observations. 2.5D gravity modelling of and longitudes were converted to (x, y) coordinates
the study area, in association with geological using the Lambert II projection. For each target, the
analyses, 2D magnetic modelling, and seismic prole new data were integrated with existing data and inter-
interpretation, has permitted the construction of a 3D polated to a 0.25 km grid spacing. The composite
geometrical model (Brun et al., 2001). Bouguer anomaly map of the studied area (Fig. 2)
Previous authors have discussed the geometry of the was produced by interpolating the entire data (onshore
faults and Cadomian units at depth in relation to a and offshore) to a 1 km grid spacing. Several interpo-
general tectonic context, based on surface geological lation types were tested, but as these produced no
information. Bale and Brun (1983) have proposed that signicant differences in the results the minimal
the main faults have a south-verging thrust component curvature algorithm was used to set up the grids that
to the west of the Saint-Brieuc Bay (Plouagat-Coet- were used as a basis for later analyses.
mieux Fault) and a strike-slip component to the east Initial analysis of the Bouguer anomaly map deter-
(Belle-IsleLa Fresnaye and PlouerCancale Faults). mined the gravity signature of key features, such as
Other authors (Strachan et al., 1992) have suggested the geometry of the major faults, the structure of the
that the present geometry resulted from collapse of geological domains, and the location of the granitic
distinct terranes along sub-vertical faults. formations and sedimentary basins. The map (Fig. 2)
shows an arcuate shape for the areas with a general
increase in amplitude from south to north.
2. Gravity measurements In the west, the gravity gradient is superimposed on
the Plouagat-Coetmieux Fault, which separates the
2.1. Bouguer anomaly map Saint-Brieuc and Guingamp units, whereas in the
east the gravity gradient migrates towards the Belle-
Onshore and offshore gravity data were combined IsleLa Fresnaye Fault, which forms the boundary
to obtain the composite Bouguer anomaly map of the between the Guingamp and Saint-Malo units.
area under investigation (Fig. 2). In this study, we Field observations near the town of Saint-Brieuc
used the offshore data collected and compiled by the show an abrupt lateral change from the Guingamp
British Geological Survey (BGS) in the course of Unit to the Saint-Brieuc Unit. It was considered that
several marine surveys The data distribution is not 2.5D gravity modelling along several proles cutting
homogeneous, with a lack of information close to across these Cadomian units would indicate whether
the coast. Therefore, gravity interpretation is ambig- this lateral discontinuity affects the surface only or is
uous on the continental shelf. The onshore data used deeply rooted.
here come from the large French National Database of
land measurements, collected and maintained by the 2.2. Density measurements
BRGM, supplemented by two complementary
surveys carried out during the development of the As there is no unique solution in gravity modelling,
Armor project: the rst, over the Brioverian Binic a priori information concerning structures and density
Formation that cuts across the Plouagat-Coetmieux distributions is essential for limiting the number of
Fault (Binic target), and the second over the north- models that might explain the data. A model based
eastern part of the Saint-Brieuc Unit, which cuts on gravity data can never be proven to be the correct
across the Belle-IsleLa Fresnaye Fault (Erquy one, but a model can always be rejected if it does not
target). The distribution of land measurement is t with the analytical data.
 84
C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197
Fig. 2. Composite Bouguer anomaly map based on a 1 km grid spacing from offshore (density reduction 2200 kg m 23) and onshore data (density reduction 2700 kg m 23). The
distribution of land measurement is around 0.5 km 22, except on the two selected targets (Binic and Erquy close-up views located by white dashed rectangles) where the distribution
reaches 2 km 22 due to the additional measurements. The two boxes display close-up views based on these new data acquired at 0.25 km grid spacing. Geological formation
boundaries and faults from Fig. 1 are superimposed in black on the Bouguer anomaly map.
 C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197 85

Table 1
Mean density of rocks in the study area (densities in bold were obtained from sampled rocks)

Formation Rock type Density (kg m 23)

Saint-Brieuc Granodiorite 2650

Leucogranite 2580
Ploufragan Anatexite and migmatite 2630
Granite Granite 2580
Saint-Brieuc Unit
Erquy Volcano-sedimentary rocks 2750
Saint-Quay-Portrieux Gabbro and diorite 2850
Plouha and Fort La Latte Tonalite 2840
Plouvara Tonalite 2800
Squifec Metagabbro 2900
Saint-Alban Metagabbro 2940
Binic Turbiditic metasediment 2700
Lanvollon (acid member) Metavolcanics 2750
Lanvollon (basic member) Metavolcanics 2820
Hillion Trondhjemite 2770
Port-Morvan Orthogneiss 2770
Morieux Orthogneiss 2800
Guingamp Unit
Yfniac Orthogneiss 2870
Guingamp Migmatite 2620

Saint-Malo Unit
Tregomar Gabbro 2900
Lamballe (syntectonic) Granitoid 2580
Saint-Malo Migmatite 2620
Lamballe Metasediments 2700

Mancellian Domain
Mancellian Metasediments 2700

Rock samples were collected from the main geolo- reectors, which are attributed to the top part of the
gical formations. Considering that geological forma- layered crust (Bitri et al., 1997, 2001). This depth also
tions are homogeneous, we used the mean density roughly corresponds to the TLC reector, as
performed on these samples (bold values in Table 1). evidenced by refraction results gained farther to the
Density values for granitic formations were selected northeast (Grandjean et al., 2001).
according to the measurements by Vigneresse (1978). The thickness of the Lanvollon Formation was
The densities are bracketed between 2580 and estimated to be less than 6 km, according to geologi-
2940 kg m 23. cal observations made to the west of Saint-Brieuc
Measured densities combined with the average rock Bay. This assumption implies that an additional
composition provide density estimations for forma- body has to be added beneath each prole. As the
tions that we did not sample. density of this underlying body is unknown, we
A planar regional effect of 8 mgals attributed to the have assumed that density contrasts below the depth
Moho depth variations was removed from the of 13 km are low, according to seismic refraction data.
Bouguer anomaly for the 2.5D modelling. This shifts This basal body with a density of 2770 kg m 23 may
the zero value of the residual anomaly onto the main underlie the Lanvollon metavolcanics as well. This
gravity gradient (Belle-IsleLa Fresnaye Fault in the density is similar to that of the Port-Morvan orthog-
east and Plouagat-Coetmieux Fault in the west). At neiss, the closest rocks that are suitable to form the
13 km depth, vertical seismic data show the rst at upper crust in the Saint-Brieuc Domain.
 86
C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197
Fig. 3. Gravity discontinuities computed using gmipack software displayed along with geological boundaries (see Figs. 1 and 2).
 C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197 87

3. Interpretation and discussion sible for such a considerable discontinuity if their

contacts were steep. Further 2.5D modelling may
3.1. Automatic identication of elementary gravity reinforce this assumption.
features (2) The second-order gravity discontinuities are
related to specic geological formations, such as the
The automatic structural analysis capabilities of the Variscan granites that crop out in the southern part of
gmipack software (TOTAL/BRGM) were applied to the Saint-Malo and Guingamp units. For the two
our gridded data set (Castaing and Debeglia, 1992). westernmost granites (Quintin and Moncontour), the
This type of analysis is suitable for dening discrete second-order gravity discontinuities correspond
borders of causative bodies at depth. The vertical and roughly to their surface contours; the third granite
horizontal derivatives (in particular the vertical gradi- (Lamballe) however, seems to be more developed at
ent and the modulus of the horizontal gradient) were depth than the mapping would indicate, with a signif-
calculated using a Fourier transform of the gravity icant dip towards the north on its northern margin
eld (Gerard and Griveau, 1972). The vertical gravity revealed by the gravity discontinuities shift toward
gradient is effective for deciphering local and shallow the north. Gravity discontinuities do not outline the
features, whereas the modulus of the horizontal gradi- Plouer-Cancale Fault (which forms the boundary
ent emphasises contacts between geological units with between the Saint-Malo Unit and the Mancellian
contrasted densities. Where contacts are not vertical, Domain), except in this area where the granite gravity
the extremum of the modulus of the vertical gradient discontinuities and the fault merge on the map. This
may be shifted slightly towards the dip direction. In result suggests that the reector modelled on `shot
this study, only results from horizontal gradient analy- gathers' would correspond to the granite border
sis were effective. `Gravity discontinuities', corre- controlled by the Plouer-Cancale Fault (Bitri et al.,
sponding to variations of the modulus of horizontal 1997). In the east, gravity discontinuities related to
gradient, were characterised by their coordinates, their the Lamballe granite extend north of the Plouer-
directions and a parameter proportional to the gradient Cancale Fault, which suggests that the fault also
value. The elementary gravity discontinuities are dips to the north. Gravity discontinuities also indicate
presented by bars with a constant length and a thick- a deep extension of the Tregomar gabbro to the east. A
ness proportional to the parameter mentioned above few discontinuities also outline the Saint-Alban meta-
(Fig. 3). gabbro which, according to sample measurements and
Two kinds of discontinuities were distinguished 2.5D modelling, is the rock with the highest density in
according to their importance: the area. In the west, the Saint-Quay diorite presents a
(1) The rst kind emphasises the main gravity few discontinuities that outline a sub-circular part.
gradient of the area. From west to east, rst-order This observation is consistent with the 2D magnetic
elementary gravity discontinuities follow the Ploua- modelling and geological observations, which
gat-Coetmieux Fault, then disappear (as does the indicate a particular structure for this diorite with a
fault) close to the Ploufragan Complex, which was more magnetized core and a rim (Cauvin-Cayet et al.,
emplaced at the end of the Cadomian orogeny (see 2001). The density does not seem to be as homoge-
Table 1). To the east of the Ploufragan Complex, neous as the magnetization. A few gravity disconti-
rst-order discontinuities are well organised and nuities outline dolerite dikes in the east, where they
parallel the Belle-IsleLa Fresnaye Fault. are numerous according to the map of the magnetic
A rst-order gravity discontinuity is also present eld reduced to the pole (Galdeano et al., 2001).
within the Lanvollon metavolcanics in the south of
the Binic basin. It does not correspond to any mapped 3.2. 2.5D Gravity modelling
feature in Fig. 1, but a complementary geological
study of this area has revealed that the Lanvollon Modelling of the gravity data was performed with
metavolcanics are dominantly acid in the south, and the interactive computer program gmipack (TOTAL/
dominantly basic in the north. These two members, BRGM). The models were computed in 2.5D, which
which present a low density contrast, could be respon- means that extension of the bodies perpendicularly to
88 C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197

the proles can be limited. Seismic, magnetic, Gravity modelling supports such a geometry. The
magneto-telluric and geological constraints were highest gravity anomaly is attributed to the south-
taken into account in order to build gravity models dipping Saint-Alban metagabbro (Saint-Brieuc
that satised the observed anomaly. However, a Unit). The adjacent Fort-La-Latte tonalite causes a
unique solution still cannot be reached since a few gravity anomaly decrease. This formation seems to
ambiguities in the interpretation remain at depth. In have been emplaced along the north-dipping Ploua-
particular, no simple and unique relations exist gat-Coetmieux Fault (boundary of the Saint-Brieuc
between the various physical parameters, such as and Guingamp units). From 5 to 8 km, the Yfniac
density and velocity, and any attempt to integrate Complex is the only formation belonging to the Guin-
results from other geophysical methods will inevita- gamp Unit. From 8 to 12 km, the anomaly gradient is
bly require a discussion of plausible geological steep decreasing to 218 mgals. This is consistent with
interpretations. the north-dipping reectors revealed by shot gather
In the interpretation presented here, we assumed a modelling attributed to the Belle-IsleLa Fresnaye
density of 2700 kg m 23 for the surrounding basement, Fault. Gravity modelling shows that this fault is verti-
which corresponds to an average density for the upper cal from the surface down to 2 km depth, and then
continental crust. Gravity modelling was carried out dips to the north at about 358. The gravity anomaly
along six proles, perpendicular to the main tectonic slowly decreases from 12 km to the southern end of
features and crossing the Saint-Brieuc, Guingamp and the prole. The Saint-Malo Unit consists mainly of
Saint-Malo units (Figs. 1 and 2). The easternmost Lamballe metasediments and Saint-Malo migmatite.
transect (Prole 1) coincides with the vertical seismic Gravity modelling suggests that the Saint-Malo
prole AR1, surveyed in July 1995 (Bitri et al., 2001) migmatite extends beneath the Lamballe metasedi-
along which complementary gravity measurements ments down to a depth of 7 km. Such a geometry is
were also acquired (Erquy target). Prole 5 is located consistent with existing geological interpretations
in the Binic target area, where new gravity data were (Chantraine et al., 2001) which attribute the Saint-
collected west of Saint-Brieuc Bay. The four remain- Malo migmatite to fusion products of the Lamballe
ing proles are based on less dense gravity data, but metasediments. Shot gather reection modelling
they allow for structures to be correlated laterally and displays a north-dipping reector, which was rst
for the construction of the 3D model (Fig. 5). ascribed to the Plouer-Cancale Fault, the most signif-
icant tectonic contact in this area.
3.2.1. Prole 1 High-frequency positive anomalies occur in two
Running along the vertical seismic prole AR1, places of the gravity prole. The northern anomaly
(Fig. 4a), the prole starts in the northwest at the is centered on the Erquy and Saint-Alban Formations,
boundary with the Port-Morvan orthogneiss (Saint- the latter being characterised by the highest
Brieuc Unit), and then runs across the Guingamp density (2940 kg m 23) measured on our rock
and Saint-Malo units, and ends in the Lamballe gran- samples. This dense formation, together with the
ite in the Mancellian Domain. The residual anomaly Morieux Formation (density of 2800 kg m 23)
presented here was extracted from the detailed grid outcropping immediately westward, would explain
(Erquy). this relatively high anomaly. In this scenario, the
Between 2 and 8 km, the gravity anomaly decreases Morieux orthogneiss would underlie the Erquy sedi-
slowly from 20 to 15 mgals. `Shot gather' modelling mentary formation. This assumption has signicant
reveals a south-dipping reector, which may be corre- geological implications. Indeed, the Erquy Formation
lated to the Port-Morvan Fault (Bitri et al., 2001). was considered as a post-Cadomian unit on the basis

Fig. 4. af: 2.5D gravity models corresponding to cross-sections 16 (see Fig. 1 for location). Calculated anomalies (above) from the density
model (below) are compared with measured gravity anomalies. Density models are deduced from geological cross-sections (patterns corre-
spond to each geological formations described on left in labelled small boxes). Plouagat-Coetmieux (2), Belle-IsleLa Fresnaye (3) and
PlouerCancalle (4) Faults are displayed on density models as on Fig. 1. PM. F corresponds to Port-Morvan Fault mentioned in the text. Values
on the right of the small boxes correspond to densities in kg/m 3 and labels are the same as on Fig. 1.
C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197 89
90 C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197

Fig. 4. (continued)
C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197 91

Fig. 4. (continued)
92 C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197

of a Rb/Sr age obtained on volcanic rocks. However, present in this area, whereas the Ploufragan Complex
the characteristics of this formation, together with its is absent on Prole 1.
geological contours showing a complex boundary
with the Morieux Formation, would indicate rather 3.2.3. Prole 4
that the Erquy Formation is of Cadomian age. Conse- Prole 4 (Fig. 4d) runs northsouth along the
quently, these shallow rocks (volcanics and sedi- western edge of Saint-Brieuc Bay, crossing the
ments) would have been deposited onto a deeper Saint-Brieuc, Guingamp and Saint-Malo units. The
basement unit, namely the Morieux Complex as prole starts in the Saint-Quay diorite in the north,
suggested by gravity modelling. then crosses the metasedimentary rocks of the Binic
The other high-frequency positive anomalies may basin, the Lanvollon metavolcanics, the Saint-Brieuc
be attributed to the westward extension of the Trego- diorite and the Ploufragan anatexite, and ends within a
mar metagabbro, according to the results of automatic Variscan granite. A 2D magnetic modelling was
structural analysis (Fig. 3). performed along this prole. North of the Saint-
Quay diorite, the gravity data used are offshore, and
3.2.2. Prole 5 there are not enough measurements close to the coast
Prole 5 (Fig. 4e) begins in the north in the Saint- to constrain the model. However, magnetic modelling
Quay diorite and cuts across the Saint-Brieuc and gave good constraints on the southern boundary of the
Guingamp units. The gravity prole was extracted Saint-Quay diorite, which appears to be sub-vertical.
from the detailed grid (Binic target). The gravity The gravity anomaly prole is rather at with
anomaly along this prole is around 20 mgals in the values around 20 mgals, except for the Saint-Quay
north and decreases slowly, with some high-frequency diorite. The residual Bouguer anomaly ts well with
oscillations, as far as 16 km. From 16 to 23 km, the the data, when preliminary assumptions are respected.
anomaly decreases rapidly, then stabilises around The Squifec gabbro does not crop out, but seems to
220 mgals. This long wavelength anomaly is in be very close to the surface. We used a 2800 kg m 23
agreement with the geological data, when preliminary density for this gabbro in Prole 4 as well as Prole 5.
assumptions are respected. The gravity modelling Geological eld observations as well as magnetic
requires more attention southward of 20 km. modelling conrm the presence of the Lanvollon
The best t was obtained with a southward dip for acid member in the southern part of the Saint-Brieuc
the Plouagat-Coetmieux Fault (as well as an eastward domain; the gravity discontinuities in Fig. 3 mark its
dip on Prole 1), which juxtaposes the Squifec meta- boundary with the Lanvollon basic member. To the
gabbro and Ploufragan Complex (anatexite and south, the Saint-Brieuc diorite and the Ploufragan
migmatite). The measured density in the Squifec Formation overprint the Plouagat-Coetmieux Fault.
Formation (2900 kg m 23) is too high, even if we Hebert et al. (1993) have suggested that the unde-
suppose that this formation is very thin. A formed Saint-Brieuc diorite intruded the already
2800 kg m 23 density was nally used for this forma- cooled country rocks, represented by the Lanvollon
tion, the density change being explained by the metavolcanics, the Squifec metagabbro and the Yf-
presence of numerous pegmatitic dikes in this forma- niac Complex. The Saint-Brieuc and Ploufragan intru-
tion. The Squifec metagabbro and the Ploufragan sions were emplaced and crystallised at the same time.
Complex, belonging, respectively, to the Saint-Brieuc The Guingamp migmatite was previously correlated
and the Guingamp units, display a high density to the Saint-Malo migmatitic belt, but gravity model-
contrast, with values of 2800 and 2630 kg m 23, ling does not conrm a lateral continuity between the
respectively. First-order discontinuities may be due two migmatite formations. The presence of a 2-km-
to the high density contrast between the two forma- thick body with a density of 2630 kg m 23 can explain
tions, the Saint-Malo Unit on Prole 1 and the Guin- the observed anomaly; this density corresponds to
gamp Unit on this prole presenting similar that of the Ploufragan Complex. New geological
thicknesses of migmatitic rocks. Modelling of this observations (Chantraine et al., 2001), leading to
prole shows that the Yfniac Complex (Guingamp consideration of the Saint-Malo and Guingamp
Unit), characterised by a high density contrast, is not Formations as two distinct units, are consistent with
 C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197 93

gravity modelling that outlines the distinct characters sented by the Yfniac Complex, as is the case for
of these migmatites. A Variscan granite, well dened Prole 1.
by the automatic structural analysis, crops out in the The Plouagat-Coetmieux and Belle-IsleLa Fres-
south (Fig. 3). In this prole, as in Prole 5, the Yf- naye Faults dip to the north, as is the case farther
niac Formation is absent. west. In the Saint-Malo Unit, a 2-km-thick sequence
of migmatite underlies the Lamballe schist down to
9 km depth. The granite intrusion, which crops out
3.2.4. Prole 6
around 30 km and would have been emplaced during
Modelling along Prole 6 (Fig. 4f) aimed at den-
the last stages of the Cadomian orogeny, cannot alone
ing the northwestward evolution of the Plouagat-
explain the negative anomaly. Another granite body
Coetmieux Fault. The particularities of this prole
with the same density has to be added farther north.
are (1) the larger amount of acid volcanics in the
Both granite formations, which are, respectively,
Lanvollon Formation in the north, (2) the presence
Variscan and Cadomian in age, have similar densities,
of syntectonic extensive deformation in the Saint-
but the fact that the unknown body seems to be deeply
Quay diorite, and (3) the occurrence of a small
rooted down to 5 km depth, is more consistent with a
gabbroic lens attributed, on the basis of the geological
post-Cadomian emplacement. This Variscan granite,
map, to the Squifec Formation. Gravity discontinu-
which crops out in the west, may extend towards the
ities (Fig. 3), as well as magnetic measurements
east.
(Cauvin-Cayet et al., 2001) and petrological studies
(Fabries et al., 1984), contributed to distinguishing
this part of the Saint-Quay diorite from that crossed 3.2.6. Prole 2
by Prole 1. The small lens of gabbro, which does not Prole 2 (Fig. 4b), which is roughly parallel to
crop out in the prole (Fig. 3) is outlined by gravity Prole 1, starts in the offshore extension of the
discontinuities; the density value needed to model this Binic Basin. It cuts across the Lanvollon Formation,
lens would correspond to that of the Squifec meta- the Hillion trondhjemite (westward extension of the
gabbro, as measured on rock samples (2900 kg m 23). Port-Morvan trondhjemite), the Morieux and Yfniac
In the south, the Plouagat-Coetmieux Fault, which Formations, the ChateauSerein volcanics, the
forms the boundary between the Squifec metagabbro Lamballe schist and its syntectonic granitoid intru-
(Saint-Brieuc Unit) and the Ploufragan migmatite sions, and ends in a Variscan granite.
(Guingamp Unit), dips to the northeast. The Yfniac The best model tting the data of this prole
Complex is absent in this model, as it is in the two displays a sub-vertical southward-dip for the Port-
proles (Proles 4 and 5) to the east of the Ploufragan Morvan and Plouagat-Coetmieux Faults. Note that
Complex. This geometry is necessary to t the on Prole 1, farther east, these faults dip towards
observed residual anomaly and is consistent with the the south, whereas farther to the west they dip to the
two other models described above. The Ploufragan north. The Guingamp Unit is only represented by
migmatite is thicker than the anatexite farther east the Yfniac Complex as in Proles 1 and 3. In the
(Prole 4). This prole ends in leucogranite. Saint-Malo Unit, the thickness of the migmatite
underlying the Lamballe schist is similar to that of
Prole 1.
3.2.5. Prole 3
Located just 5 km west of Prole 2, Prole 3
(Fig. 4c) crosses the Plouagat-Coetmieux and Belle- 4. Conclusion
IsleLa Fresnaye Faults, where kinematic criteria
based on structural analyses point to a southwestward 2.5D gravity modelling is very consistent with new
thrust system (Bale and Brun, 1983; Hebert et al., geological interpretations by Chantraine et al. (2001),
1993). Note that the Morieux and Yfniac Formations and allows the construction of a 3D geometrical
are well represented down to a depth of 6 km on model using voronoi diagrams method (Courrioux
this prole, whereas they are absent on all of the et al., 2001) (Fig. 5). The Bouguer gravity contribu-
proles to the west. The Guingamp Unit is only repre- tion has been computed from the 3D model after
94 C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197

Fig. 5. 3D model reconstructs from gravity prole results using Voronoi diagrams method (Courrioux et al. 2001).

density attribution from Table 1 to each formation. anomalies. This will be done in a further step after
The result (Fig. 6) is close to the observed anomaly taking into account all geological and geophysical
(Fig. 2) except offshore where the observed Bouguer results.
anomaly has been reduced with a 2.2 density (rather The Saint-Brieuc Bay region shows the highest
than 2.7 for the calculated Bouguer anomaly). positive Bouguer anomaly observed in Proterozoic
Several questions remain unresolved concerning, Europe. The main gravity gradient is due to the juxta-
for example, the lateral composition change of the position of plutonic and metamorphic rocks in the
Guingamp Unit, and the disappearance of the Yfniac Tregor and Saint-Brieuc Units and partly the Guin-
and Morieux Formations beneath the Ploufragan gamp Unit, and of metasedimentary, migmatitic and
Complex and westwards. Forward modelling by trial leucogranitic rocks in the Saint-Malo Unit and partly
and error is necessary to reduce the difference in the Guingamp Unit.
between the observed and the calculated Bouguer Preliminary assumptions about the thickness of the
 C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197
Fig. 6. Gravity anomaly computed from the 3D model (Fig. 5) after attributing densities from Table 1 for each formation.

95
96 C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197

Lanvollon Formation and the density of the Lanvollon by an anonymous reviewer are greatly appreciated.
basement were made for the six models without intro- This work was supported by the GeoFrance 3D
ducing additional incoherent bodies. program. P. Ledru is gratefully acknowledged for
The gravity 2.5D modelling contributions can be his editorial help with the manuscript. GeoFrance
summarised as follows: 3D Publication 16.

The geometry of the Belle-IsleLa Fresnaye Fault

is sub-vertical from the surface down to 12 km References
depth in Proles 13, then attens and dips to the
north at depth. Gravity modelling was stopped at Bale, P., Brun, J.P., 1983. Les chevauchements cadomiens de la baie
13 km depth, and therefore never reached the de Saint-Brieuc (Massif Armoricain). C. R. Acad. Sci. Paris 297,
359362.
supposed horizontal plane of this structure. Bitri, A., Brun, J.P., Chantraine, J., Guennoc, P., Marquis, G.,
However seismic refraction reveals a at reector Marthelot, J.M., Perrin, J., Pivot, F., Tournerie, B., Truffert,
at 15 km depth northeast of the vertical seismic C., 1997. Structure crustale du bloc cadomien de Bretagne
prole (Grandjean et al., 2001), which can be Nord (France): sismique reexion verticale et sondage magne-
attributed to the Belle-IsleLa Fresnaye Fault. totellurique (projet Geofrance 3-D Armor). C. R. Acad. Sci.
Paris 325, 171177.
From Proles 4 to 6, the Belle-IsleLa Fresnaye Bitri, A., Brun, J.P., Truffert, C., Guennoc, P., 2001. Deep seismic
Fault deepens northward directly from the surface imaging of the Cadomian thrust wedge of Northern Brittany.
down to 13 km and deeper. This fault geometry is Tectonophysics 331 (12) 6580.
consistent with geological observations collected Brun, J.-P., Bale, P., 1990. Cadomian tectonics in northern
along the contact, which indicate a thrusting kine- Brittany. In: D'Lemos, R.S., Strachan, R.A., Topley, C.G.
(Eds.), The Cadomian Orogeny, Geol. Soc. Lond. Spec. Publ.
matics with a strike slip component (Brun and 51, 95114.
Bale, 1990). Brun, J.P., Guennoc, P., Truffert, C., Varion, J., and the
The Prole 1 model conrms the southward dip of Armor Working Group of the GeoFrance 3-D Program,
the Port-Morvan Fault indicated by the vertical 2001. Cadomian tectonism in Northern Brittany: a
seismic prole. The Morieux Formation (Saint- contribution of crustal-scale 3-D modelling. Tectonophysics
331 (12) 229246..
Brieuc Unit) displays also a southward dip in Castaing, C., Debeglia, N., 1992. A new method for combining
Proles 1 and 2. This trend may correspond to gravity and geological data. Tectonophysics 204, 151162.
thrusts with opposite vergences related to the Cauvin-Cayet, C., Galdeano, A., Egal, E., Pozzi, J.P., Truffert, C.,
main north-dipping feature (the Belle-IsleLa 2001. Magnetic modelling in the French Cadomian belt (North-
Fresnaye Fault) within a `pop-up' structure. ern Armorican Massif). Tectonophysics 331 (12) 123144.
Chantraine, J., Chauvel, J.J., Bale, P., Denis, E., Rabu, D., 1988. Le
The Guingamp and Saint-Malo migmatites are Brioverien (Proterozoque superieur a terminal) et l'orogenese
different, the Saint-Malo migmatite being thicker cadomienne en Bretagne (France). Bull. Soc. Geol. Fr. IV (5),
(around 23 km). Their densities are also distinct. 815829.
This observation is consistent with the new Chantraine, J., Egal, E., Thieblemont, D., Le Goff, E., Guerrot, C.,
unit distribution (Chantraine et al., 2001) that Ballevre, M., Guennoc, P., 2001. The Cadomian active margin
(North Armorican Massif, France): a segment of the North
considers the Guingamp and Saint-Malo Units to Atlantic Panafrican Belt. Tectonophysics 331 (12) 118.
be distinct. Courrioux, G., Nullans, S., Guillen, A., Boissonnat, J.D., Repus-
The Yfniac and Morieux Formations do not seau, Ph., Renaud, X., Thibaut, M., 2001. 3D volumetric model-
underlie the Ploufragan anatectic granite; they ling of Cadomian terranes (Northern Brittany, France): an
actually stop west of Prole 4. This observation automatic method using Vorono diagrams. Tectonophysics
331 (12) 181196.
permits an improved cartography of the Denis, E., 1988. Les sediments brioveriens (Proterozoque super-
Belle-IsleLa Fresnaye Fault by Chantraine et al. ieur) de Bretagne septentrionale et occidentale. These Univ.
(2001). Rennes, Mem. et Doc. CAESS 18, 223 pp.
Dissler, E., Dore, F., Dupret, L., Gresselin, F., Le Gall, J., 1988.
L'evolution geodynamique cadomienne du Nord-Est du Massif
Acknowledgements armoricain. Bull. Soc. Geol. Fr. IV (5), 801814.
Egal, E., Guerrot, C., Le Goff, E., Thieblemont, D., Chantraine, J.,
Comments by Roger Bayer and a thorough review 1996. The Cadomian orogeny revisited in northern Brittany
 C. Truffert, E. Egal / Tectonophysics 331 (2001) 8197 97

(France). Avalonian and Related peri-Gondwanan Terranes of Northern Armorican Massif (France). C. R. Acad. Sci. Paris
the Circum North Atlantic, Geol. Soc. Am. Spec. Paper 304, 316, 395401.
281318. Rabu, D., Chantraine, J., Chauvel, J.J., Denis, E., Bale, P., Bardy, P.,
Fabries, J., Conquere, F., Arnaud, G., 1984. The mac silicates in 1990. The Brioverian (Upper Proterozoic) and the Cadomian
the St-Quay-Portrieux gabbro-dioritic intrusion: crystallisation orogeny in the Armorican Massif. In: D'Lemos, R.S., Strachan,
conditions of a calk-alcaline pluton. Bull. Miner. 107, 715736. R.A., Topley, C.G. (Eds.). The Cadomian Orogeny. Geol. Soc.
Galdeano, A., Asrane, F., Truffert, C., Egal, E., Debeglia, N., London Spec. Publ. 51, 8194.
2001. The aeromagnetic map of the French Cadomian belt. Strachan, R.A., Treloar, P.J., Brown, M., D'Lemos, R.S., 1989.
Tectonophysics 331 (12) 99122. Cadomian terrane tectonics and magmatism in the Armorican
Gerard, A., Griveau, P., 1972. Interpretations quantitatives en grav- Massif. J. Geol. Soc. London 146, 423426.
imetrie et magnetisme a partir de cartes transformees de gradi- Strachan, R.A., Brown, M., Dallmeyer, R.D., D'Lemos, R.S., 1992.
ent vertical. Geophys. Prospect. 20, 459481. Discussion of the timing and kinematics of Cadomian deformation
Grandjean, G., Guennoc, P., Recq, M., Andreo, P., 2001. Refrac- (Northern Brittany): reply. J. Geol. Soc. Lond. 149, 151155.
tion/wide angle reection investigation of the Cadomian crust Strachan, R.A., D'Lemos, R.S., Dallmeyer, R.D., 1996. Neoproter-
between northern Brittany and the Channel Islands. Tectono- ozoic evolution of an active plate margin: North Armorican
physics 331 (12) 4564. Massif, France. Avalonian and Related peri-Gondwanan
Graviou, P., 1992. Reconnaissance d'une suture majeure au sein de Terranes of the Circum North Atlantic. Geol. Soc. Am.
la chane cadomienne. C. R. Acad. Sci. Paris 315, 17991802. Spec. Paper 304, 319332.
Hebert, R., Hallot, E., Guerrot, C., Chantraine, J., 1993. New struc- Vigneresse, J.L., 1978. Gravimetrie et granites armoricains. Struc-
tural, petrological and radiometric constraints within the Cado- ture et mise en place des granites hercyniens. These Univ.
mian belt: chronology of events in the baie de Saint-Brieuc, Rennes, 98 pp.