Journal of Structural Geology 28 (2006) 456–473

www.elsevier.com/locate/jsg

Late Paleozoic strike-slip faults and related vein arrays of

Cape Elizabeth, Maine
Mark T. Swanson
Department of Geosciences, University of Southern Maine, Gorham, ME 04038, USA
Received 25 May 2004; received in revised form 14 March 2005; accepted 2 December 2005
Available online 7 February 2006

Abstract

Strike-slip faults and related quartz vein arrays of Late Paleozoic-age cut gently-dipping metasedimentary rocks at Cape Elizabeth in southern
coastal Maine and formed in response to regional dextral shearing along the Norumbega fault system. Vertical quartz veins up to 20 m wide and
10s of meters long were emplaced orthogonal to the local shear zone-parallel elongation fabric, reflecting strain partitioning during transpression.
Earlier veins were reoriented by clockwise rotation toward this NE-trending regional shear direction. The later brittle strike-slip faults are oblique
to the regional shear direction and interpreted as a 10-km-scale R-shear array on the southeast flank of the Norumbega fault system. These left-
stepping en échelon fault zones consist of the three Two Lights fault zones (w200 m lengths and up to w5 m displacements) and the Richmond
Island fault zone (w1.6 km length and w40 m displacement). Displacements on these fault zones have developed fine-grained silicified,
obliquely-foliated and laminated cataclasites and locally, millimeter-thin pseudotachylyte fault and injection veins. Individual fault core zones are
up to 10s of centimeters thick as part of several complex anastamosing zones of faulting 10s of meters wide. Initial segments within each fault zone
are typically terminated with oblique extension fractures in horsetail configurations. The left-stepping en échelon relationships between these
segments led to dominantly contractional step-over zones where P-shear linkages created a through-going fault that truncated the ends of the
earlier-formed terminated segments. This linkage-growth model for fault zone evolution works toward larger scales and longer fault lengths as
displacement accumulates, within a limiting maximum displacement/length ratio characteristic of the host lithologies. Length–frequency data for
fault segments within these zones suggest a transition to linkage-dominated growth once fault segments were longer than w15 m. Continued
displacement was accommodated along the P-shear linked en échelon faults through imbrication, contractional duplexing and adhesive wear on
the outcrop-scale. Core zone processes on the micro-scale reflect cataclasis and frictional sliding during coseismic slip as well as cataclastic flow
and pressure solution during post-seismic creep. The development of foliated-to-laminated cataclasite was accompanied by pore volume collapse,
pressure solution and fluid expulsion that, in turn, triggered the development of the late fault-related quartz vein arrays.
q 2006 Elsevier Ltd. All rights reserved.

Keywords: Strike-slip faults; Quartz veins; Fault growth

1. Introduction The different shear and extension fractures found associated

with shallow crustal fault zones (Fig. 1) have been
Strike-slip faults are notably segmented at all scales and characterized through detailed field studies of exhumed fault
levels of exposure (Tchalenko, 1970; Segall and Pollard, 1980; structures (Tchalenko, 1970; Grocott, 1981; Hedgecoxe and
Barka and Kadinsky-Cade, 1988; Sylvester, 1988; Scholz, Johnson, 1986; Swanson, 1988, 1989, 1990, 2005; Cruikshank
1990; Vermilye and Scholz, 1999), typically in the form of en et al., 1990, 1991; Arboleya and Engelder, 1995; Davis et al.,
échelon shear fractures. The step-over zones of host rock 1999) and a history of experimental modeling (Riedel, 1929;
between the ends of en échelon shear fractures deform in order Morgenstern and Tchalenko, 1967; Tchalenko, 1970; Logan et
to accommodate continued displacement and eventual linkage al., 1979; Bartlett et al., 1981; Naylor et al., 1986; Arch et al.,
to a through-going fault zone (Fig. 1). The geometry of these 1988). Model studies using materials such as clay and sand
step-over zones and linking fault structures, in turn, controls (e.g. Riedel, 1929; Wilcox et al., 1973) have revealed a fairly
the overall style of fault zone development. consistent pattern of development in experimental fault zones.
Fault zones in clay, for example, usually begin with an early
fabric of small R-shears and minor R 0 -shears that evolves into
E-mail address: [email protected] larger en échelon R-shears. P-shear linkage, imbrication and
0191-8141/$ - see front matter q 2006 Elsevier Ltd. All rights reserved. duplexing (Fig. 1) combine to give a through-going but
doi:10.1016/j.jsg.2005.12.009 irregular fault zone as seen in natural examples
 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473 457

Fig. 1. Fault zone terminology includes en échelon R-shears, R 0 -shears and oblique extension-fracture terminations. Initial en échelon R-shear arrays develop P-shear
linkages, lenses and duplexes to form through-going fault zones.

and experimental studies (Riedel, 1929; Tchalenko, 1970; Lower Paleozoic Kittery Formation of the Merrimack Group of
Tchalenko and Ambraseys, 1970; Bartlett et al., 1981; Naylor southern Maine and New Hampshire (Hussey, 1988; Bothner
et al., 1986; Woodcock and Fischer, 1986; Arch et al., 1988; and Hussey, 1999). The exposures at Cape Elizabeth are
Cruikshank et al., 1991). separated from the rest of the highly sheared Casco Bay area by
Faulting in the upper crust has also been characterized the Broad Cove shear zone, considered to be one of the
(Sibson, 1989, 1990; Hickman et al., 1995; Janssen et al., 1997) as principal splays (Bothner and Hussey, 1999) of the Late
a seismo-structural process that involves stress cycling, Paleozoic-age Norumbega fault/shear zone system (Fig. 2).
coseismic displacements, post-seismic creep and the formation The restraining bend geometry in the Casco Bay area (Osberg
and collapse of fault-related dilatancy driving transient pulses of et al., 1985; Swanson, 1992, 1999a,b) contributed to the
hydrothermal fluids. Fault ‘valve’ models have been proposed formation of regional scale en échelon isoclinal upright F2
(Cox, 1995; Robert et al., 1995; Nguyen et al., 1998) where folds, oblique to the main Norumbega fault zone, which
increasing fluid pressures below sealed reverse faults period-
ically trigger displacements with subsequent enhanced fluid flow
and re-silicification. Fault ‘pump’ models have also been
proposed (Sibson et al., 1975), where coseismic collapse of
strain-induced micro-crack dilatancy within the host rock flushes
fluid through the broken fault zone rock (Matthai and Fisher,
1996). Post-seismic creep and compaction of newly formed
cataclasite (Sleep and Blanpied, 1994; Segall and Rice, 1995), in
turn, lead to pore volume reduction with fluid expulsion and vein
emplacement in the adjacent wall rocks (Renard et al., 2000).
The Cape Elizabeth fault zones studied here (Fig. 2) are
associated with pseudotachylyte as evidence of frictional
melting and coseismic slip as well as abundant quartz veins
as evidence of transient fluid flow, both characteristics of
earthquake faulting (Scholz, 1988, 1990; Sibson, 1989). The
mapping of these exhumed structures provides a detailed view
of these strike-slip fault/vein systems adding to our under-
standing of the processes of fault zone evolution and the
deformation mechanisms involved in earthquake generation.

2. Geology of the Cape Elizabeth study area

The strike-slip fault zones and related veins described in this

paper are exposed in coastal terraces in the vicinity of Two
Lights State Park and adjacent Richmond Island in Cape Fig. 2. Location and structural setting for the Cape Elizabeth fault zones. (a)
Elizabeth, Maine (Fig. 2), at the southern end of Casco Bay. Regional setting of the Norumbega fault system with the location of the study
area. (b) Specific fault zones in Cape Elizabeth include the Two Lights fault
zones (TL-1, TL-2 and TL-3) and Richmond Island fault zone (RI). Dashed and
2.1. Structural setting
dotted lines represent the host rock layering within and adjacent to the Broad
Cove shear zone, an outer splay to the Norumbega shear zone system. Strike
The host rocks at these localities consist of chloritized and dip symbols in (b) represent the gently SE-dipping fold limbs in the Two
garnet-bearing phyllitic quartzites that are correlated with the Lights exposures and the vertical flow layering in the Broad Cove shear zone.
458 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473

became reoriented and sheared into the main fault trace as the al., 1989) from the sheared metamorphic rocks on the southeast
transpressional deformation continued. side of the main Norumbega zone. These biotite ages indicate
This regional Norumbega-shearing is expressed in a wide that the onset of brittle faulting due to post-metamorphic uplift
variety of strain and kinematic indicators (Swanson, 1992, and cooling below the 300 8C closure temperature for biotite
1999a) throughout the oblique F2 folds on the southeast side of would have occurred at w290–295 my during the latest
the main fault zone. The steeply-dipping host rock layering Carboniferous. Early Permian K/Ar whole rock ages of
within this folded terrain helped to partition the transpression 270G32 and 277G16 my were also reported by Bothner and
into layer-parallel simple shear as the translation component, Hussey (1999) for pseudotachylyte from related(?) fault
and a horizontal layer-parallel elongation (and accompanying structures in the Fort Foster Brittle Zone (Swanson, 1988) of
layer-normal shortening) as the pure shear component to the southernmost Maine.
deformation. The local crustal thickening associated with this Pressure estimates of 2.3–3.0 kbars for Late Devonian peak
transpression led to uplift, erosional unroofing and cooling of metamorphism in the Saco-Harpswell sequence on the south-
the present-day crust through the brittle–plastic transition east side of the main Norumbega zone as reported by Dunn and
where progressive strain hardening promoted the development Lang (1988) represent maximum former crustal depths of w8–
of the brittle strike-slip faults. 10 km. Later retrograde chloritization commonly associated
The Two Lights and Richmond Island exposures southeast of with the Cape Elizabeth faults and veins as well as the regional
the Broad Cove shear zone and outside of the higher shear strain Norumbega fault system (Hussey, 1985) also indicates
zones in Casco Bay, reveal a gently southeast dipping planar progressively shallower crustal levels likely due to uplift and
S0/S1 anisotropy (related to recumbent F1 folds) with only open to erosion. Further erosion into the Mesozoic yields crustal depths
gentle upright F2 folds (Fig. 2b). The cataclastic strike-slip faults of 5.3–7.6 km and temperatures of only 170–200 8C (Dough-
in these exposures cut obliquely across this gently SE-dipping erty and Lyons, 1980) based on zircon-apatite ages for younger
lithological layering and are distributed along the southeast flank Jurassic plutons in nearby New Hampshire.
of the much larger NE-trending Broad Cove shear zone. The pseudotachylyte-bearing strike-slip faults in the Cape
Elizabeth area, then, are latest Carboniferous to Permian in age
2.2. Age and depth of exposure for the Cape Elizabeth (300–250 my) developed at w2.0 kbars, w250–300 8C, and
fault zones w7–8 km depth during the waning stages of oblique
convergence and associated dextral transpression. These
The age of these Late Paleozoic strike-slip faults can be conditions reflect former crustal depths for the exposed fault
constrained by 40Ar/39Ar ages on biotites (West, 1988; West et zones above the brittle–plastic transition (Scholz, 1990), where

Fig. 3. Two Lights shoreline exposure showing the orientation and distribution of selected structures developed during shearing and extension. (a) Latest undeformed
orthogonal vein array and small dihedral angle planar conjugate strike-slip faults with extension due to strain partitioning during transpression. (b) Variably
deformed quartz veins reoriented by clockwise rotation and antithetic (sinistral) slip due to dextral shearing. (c) Oblique dextral strike-slip faults including the TL-1,
TL-2, and TL-3 fault zones and vein associations with the youngest undeformed to slightly deformed, initially orthogonal-to-shear vein arrays.
 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473 459

deformation is dominated by brittle fracturing at high to the shear direction consistently cross-cut the older more
intermittent strain rates and diffusive mass transfer/solution deformed veins and boudin strings at lower angles to the shear
creep at low background strain rates (Knipe, 1989; Gratier and direction. The clockwise rotation of poles to veins as seen in
Gamond, 1990). stereogram (Fig. 4a and b) creates a great circle distribution
that defines a steeply-plunging rotation axis for the observed
dextral shear. A histogram of vein orientations relative to the
3. Quartz vein emplacement and deformation by dextral
shear direction (Fig. 5a) in the Two Lights exposures reveals a
shear
continuum of positions suggesting intermittent veining
throughout the period of dextral shear (see stereogram
The gently SE-dipping lithologies in the Cape Elizabeth
Fig. 4b). The apparent triple peak structure in the smoothed
exposures are just 1 km outside of the main Casco Bay shear
histogram (Fig. 5a) may suggest possible long-term fluctu-
zones but have still developed a distinctive NE-trending sub-
ations in fluid pressure during shear or variations in strain rate
horizontal stretching lineation that defines an elongation/shear
flow fabric parallel to the regional shear direction (Fig. 2b). during veining but it likely represents a bias in the distribution
NW-trending quartz–ankerite–chlorite-filled veins perpendicu- where the intense development of certain vein sets in localized
lar to the lineation in these outcrops (Fig. 3a) are typically only areas dominate the data.
a few centimeters thick and 10s of meters long but can be as While the strike-slip faults cut the earlier deformed and
much as 20 cm thick. Swanson (1992, 1999a) documented the reoriented veins and boudin strings, cross-cutting relations
initial emplacement of quartz veins orthogonal to this between the strike-slip faults and the youngest vein sets are
elongation/shear flow fabric (as the layer-parallel elongation contradictory. This suggests a broadly contemporaneous
component of the strain-partitioned transpression) with clock- relationship between the brittle strike-slip faulting and the
wise reorientation to lower angles (Figs. 3b and 4a and b) later part of a protracted history of episodic veining and
caused by continuous dextral shear. These syn-tectonic quartz
veins vary in orientation that can be related to age of
emplacement. Younger, less deformed veins at higher angles

Fig. 4. Structural data in stereograms for the Cape Elizabeth exposures. (a)
Poles to latest undeformed quartz veins orthogonal to shear zone-parallel
extension. (b) Poles to quartz veins deformed by regional dextral shear with
reorientation of initially orthogonal-to-shear undeformed veins by clockwise
rotation under a steeply-plunging rotation axis reconstructed from the great
circle distribution of poles to reoriented veins. (c) Poles to low angle dextral Fig. 5. Quartz vein distribution in the Two Lights shoreline exposures. (a)
shear bands and high-angle sinistral kink-bands with steeply-plunging rotation Histogram of vein orientations relative to the interpreted shear direction
axes that are interpreted as an R–R 0 conjugate shear set reflecting dextral strike- showing a continuum of trends with three possible clusters of vein orientations.
slip. (d) Fault striations and poles to dextral faults and termination veins show a (b) Profile line along the length of the Two Lights exposures. There is an
sub-horizontal great circle distribution characteristic of strike-slip. Arrows increase in vein intensity and vein extension associated with each of the TL-
indicate the NE–SW extension and NE-trending dextral strike-slip associated fault zones. High values between the TL-2 and TL-3 fault zones are due to
with the transpressional deformation. multiple faults with overlapping vein arrays.
460 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473

regional shear. There appear to be specific sets of veins shearing. These are in a conjugate relationship to the dextral
spatially associated with each of the main fault zones (Fig. 3c). R-shear bands where the acute bisector between the two sets
A profile line through the mapped Two Lights exposures reflects the paleostress orientation for dextral shearing. The
(Fig. 5b) shows an increase in vein intensity and extension larger 100C-m-long dextral cataclastic fault zones are similar
associated with the areas immediately surrounding each of the in orientation to the smaller scale shear band fabric elements
TL-fault zones. The vein arrays associated with the TL-1 and and are also interpreted as R-shears (Figs. 2 and 3c). This
TL-2 fault zones are similar in mineralogy and orientation but pattern of conjugate strike-slip structures with NNE-trending
are spatially separated by a gap in vein intensity and extension. dextral shear band fabrics and NNW-trending sinistral kink
At the eastern end of the Two Lights exposure the increase in bands is similar to the left lateral strike-slip faults and right-
extension and intensity between the TL-2 and TL-3 fault zones lateral monoclinal kink bands in the Sierra Nevada (Davies and
may be attributed to overlapping vein systems due to the Pollard, 1986), and to the conjugate arrays in proto-Riedel
proximity of the two faults. shear zones in Navajo sandstone of southern Utah described by
The latest extensional structures are conjugate sets of very Ahlgren (2001).
planar high-angle-to-shear brittle strike-slip faults that include
WNW-trending dextral and NNW-trending sinistral structures 4.2. Fault orientations and kinematics
(Fig. 3a) that indicate a similar NE–SW extension direction as
the orthogonal-to-shear quartz veins. Such late conjugate faults The Two Lights and Richmond Island fault zones have a
in shear zones have been described by Nguyen et al. (1998) and general 0808 strike and 758NW dip, oblique to the 0428 trend of
Poulsen and Robert (1989) and are similar to the X–X 0 faults of the regional F2 hinge-parallel lineation that defines the
Logan et al. (1979) and Swanson (1988). The change in elongation/shear flow fabric and interpreted regional shear
deformation response from pure shear extension fracturing to direction. Each of these oblique faults is composed of ENE-
the small dihedral angle, hybrid, extension/shear fracturing trending dextral strike-slip fault zones and NW-trending
could be related to variations in local fluid pressures and extension veins. The stereogram of fault striations and poles
differential stress at the time of deformation. to slip surfaces and veins within these fault zones (Fig. 4d)
shows a nearly-horizontal great circle distribution, representing
4. The Two Lights and Richmond Island Fault Zones the movement plane during faulting, and a steeply-plunging
kinematic rotation axis (668 towards 0558) that approximates
The exposure at Two Lights State Park (Figs. 2 and 3) is the intermediate principal stress direction for plane strain
1.2 km long and 50 m wide, is aligned parallel to the during the deformation. Rare centimeter-wide grooves in
interpreted regional shear direction and contains numerous cataclastic sheets along some of the main fault surfaces
strike slip fault zones. The three major Two Lights fault zones, (Fig. 4d) are also gently-plunging (w88 NE) indicating a
(TL-1, TL-2, and TL-3 from west to east) have displacements subhorizontal shear direction. NNW-trending sinistral strike-
in the 3–5 m range. The nearby Richmond Island shoreline slip kink-bands are also common within, and adjacent to, the
exposure is w600 m long aligned parallel to the Richmond individual dextral fault zones. The oblique orientation of these
Island fault zone (Fig. 2), which has a larger displacement of fault zones relative to the overall shear direction, their dextral
w40 m. shear sense and left-stepping en échelon geometry support their
The TL and RI fault zones were mapped using a variety of interpretation as outcrop-scale en échelon R-shear fault zones
techniques that included the use of measuring tape grids at a (Figs. 2 and 3c) in a kilometer-wide zone of dextral strike-slip
scale of 1:60, a 7 m camera pole and aerial photo mosaic, and a shearing.
plane table and alidade at a scale of 1:120. The maps of the
Two Lights and Richmond Island fault zones (Fig. 6) have been 4.3. Types of fault-related rocks and structures
reduced in size for publication so that considerable detail and
complexity cannot be shown. Detailed views of selected areas The Cape Elizabeth fault zones have produced quartz,
are provided in the subsequent figures to show the character- ankerite and chlorite veins (Figs. 7a and 8c), quartz- and
istic structures within each of the mapped faults. calcite-cemented dilation breccias (Figs. 7f and 8a), finer-
grained cataclasites (Fig. 7c and h), obliquely-foliated
4.1. Bulk dextral shear cataclasites (Figs. 7g and 8e and f), laminated cataclasites
(Fig. 8g and h) and local pseudotachylyte fault and injection
A late stage of dextral shear throughout these exposures is veins (Fig. 8b and d). The faults display narrow core zones
reflected in locally prominent centimeter-to-meter long dextral (Fig. 7c) with laminated ultracataclasite (Fig. 8g and h), and
shear bands (Figs. 4c and 7b) that cut obliquely across the flanking damage zones (Fig. 7e and f) with minor breccias
elongation/shear flow fabric (Swanson, 1992, 1999a), particu- (Figs. 7f and 8a and b), splay faults (Fig. 7d), shear band fabrics
larly in the more phyllitic layers. Centimeter-wide sinistral (Fig. 7b), kink bands (Fig. 7b) and extension veins (Fig. 7a).
strike-slip kink bands at high angles to the elongation/shear Individual cataclastic core zones vary in width depending on
flow fabrics also occur with steeply-plunging rotation axes (728 the magnitude of displacement but typically range up to w1–
towards 3208). These sinistral kink bands (Fig. 7b) are 2 cm for the Two Lights faults (Figs. 7c and 8g and h) and up to
interpreted as antithetic R 0 -shears related to regional dextral 10s of centimeters for the Richmond Island fault zone (Fig. 7h).
 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473
Fig. 6. Maps of the Two Lights and Richmond Island fault zones in the Cape Elizabeth exposures. (a) TL-1 fault as a through-going fault zone with wall rock lensing through a high-angle oblique vein and fracture
array. (b) TL-2 fault with low-angle initial en échelon array, discrete linkage structures and final through-going fault zone. (c) TL-3 fault zone with a broad zone of multiple sub-parallel fault surfaces, abundant
extension fractures and termination arrays. (d) RI fault zone with a distinctive linkage lens structure (note the change in scale).

461
462 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473

Fig. 7. Outcrop-scale characteristics of dextral strike-slip in the Cape Elizabeth fault zones. (a) Reorientation of quartz veins by clockwise rotation due to progressive
dextral shear with older, more-reoriented, elongated quartz vein cut by younger planar quartz vein showing only minimal clockwise rotation. (b) Conjugate geometry
between low angle-to-shear dextral shear band fabrics (as R-shears) and a single higher angle-to-shear sinistral kink band (as an R 0 -shear). (c) TL-1 fault zone
structure showing offset and drag of local host rock layering against the pseudotachylyte-bearing main fault core zone with damage zone fracturing in the wall rocks.
(d) Splay faults at the trailing end of an adhesive wear structure along the RI fault core zone. (e) Flanking quartz vein cut by later slip adjacent to the RI fault core
zone. (f) Local silicified breccia zone adjacent to a fault segment in the RI fault zone. (g) 5-cm-wide obliquely-foliated dextral cataclastic core zone with
recognizable clasts from the RI fault zone. (h) Larger, higher displacement core zone with more strongly-developed foliation. Arrows indicate dextral shear across
each photograph. For (d)–(f) the main fault core zone runs across the top of each image. The lens cap is 5.5 cm in diameter and the pencil is 14 cm long.

Pseudotachylyte is recognized in only a few places within (Fig. 8d), as well as flow layering within the fault veins and
these core zones as millimeter-wide fault-parallel veins of very larger reservoir structures.
fine-grained dark gray material, commonly with tapered Foliated cataclasites (Chester et al., 1985; Chester and
orthogonal injection veins extending only a few millimeters Logan, 1986) are also commonly developed within the
into the adjoining fault wall. In thin section the pseudotachy- larger, more evolved fault segments at Richmond Island
lyte is found to contain rounded clasts (Fig. 8b), and cuspate through clast rotation, clast alignment, pressure solution and
boundaries with the wall rocks due to contact melting effects cataclastic flow. The development of obliquely foliated
 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473 463

Fig. 8. Photomicrographs of dextral strike-slip fault rocks in thin section from the Cape Elizabeth fault zones. (a) Coarse-grained dilation breccia cemented with
calcite and minor quartz from a pull-apart breccia structure in the TL-2 fault zone. (b) Coarse-grained dilation breccia with pseudotachylyte matrix from the Two
Lights exposures. Note the rounded shapes to the included clasts due to resorption and thermal spalling. (c) Quartz veins and silicification adjacent to a segment of
the RI fault zone. (d) Pseudotachylyte fault vein from the Two Lights fault zones showing internal flow layering and a cuspate resorption contact with the adjacent
host rock. (e) A 0.5-cm-wide obliquely-foliated cataclasite zone in part of the RI fault zone. Several of the fragments within the cataclasite zone are finer grained
cataclasite from earlier faulting episodes. (f) Obliquely-foliated cataclasite texture with foliation surfaces as solution planes from the RI fault zone. (g) and (h) Two
contiguous thin sections of the TL-1 fault zone, showing the dark laminated core with reworked quartz lenses. Arrows indicate dextral shear across each
photomicrograph. White barsZ1 cm.

cataclasite (Fig. 8e and f) leads to laminated core formation The presence of pseudotachylyte within these narrow higher
(Fig. 8g and h) with fabric rotation and enhanced pressure strain core zones shows that some of these fault materials were
solution punctuated by frictional sliding and renewed produced during the rapid coseismic slip phase of the
cataclasis. Repetitive cycles of ultracataclasis and friction earthquake cycle. The accompanying cataclastic flow and
melting along with this cataclastic flow, pressure solution solution creep associated with the laminated core zones
and fabric rotation have produced distinctive fault-parallel represent a history of paleo-seismic activity (Knipe, 1989;
laminations within the narrow highly-strained core zone Sibson, 1989; Gratier and Gamond, 1990) preserved along
layers. these fault zones.
464 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473

4.4. Initial en échelon segmentation 4.5. Terminations

An initial en échelon segmentation structure can be recognized The isolated ends of fault segments in these exposures
within many of the Two Lights fault zones with segments w35– typically terminate in arrays of quartz–ankerite veins
50 m long although this is quite variable. Left-stepping geometries oblique to the adjoining strike-slip fault segments (Fig. 9d
have stepover widths up to w40 m with overlaps of up to w20 m. and e) in a horsetail configuration (after Granier, 1985).
In the TL-2 fault zone, one of the early R-shears can be recognized Individual veins in these terminations are up to 1 m long
as a 35-m-long segment with an isolated horsetail termination that with dozens of veins in individual termination arrays. A few
has been truncated and abandoned by P-shear linkage. Many of composite structures (Fig. 9b and c) show a P-shear splay
these early segments are, in turn, made up of their own smaller scale on the contractional side of the termination with a horsetail
segments w2–3 m long. The larger Richmond Island fault zone array on the extensional side of the termination. Larger-
also has an initial en échelon structure with three distinct segments, scale terminations with fractures several meters long are
w400–500 m long, with stepover widths of 10–40 m and overlaps developed in the TL-1 fault zone (Fig. 6c) at the ends of
of w10–100 m. Recognition of this early en échelon geometry is two main fault segments at a 40 m left step in the TL-1
key in reconstructing fault zone evolution. array.

Fig. 9. Maps of fault terminations with oblique extension fractures. (a) Horsetail termination from the TL-2 fault zone. (b) Fringe terminations suggesting successive
positions of fault tip. (c) Complex termination showing dextral and high-angle sinistral fault splays with contractional P-shears on the opposing side. (d) Extensional
linkages that developed as isolated or paired, linking termination arrays. (e) Modification of extensional linkages by forward (clockwise) rotation and induced
antithetic (sinistral) slip. (f) The development of fault terminations in relation to the contractional (C) and extensional (K) sides of a strike-slip fault tip and the
formation of oblique extension fractures in horsetail splays.
 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473 465

4.6. Linkages Multiple linking P-shears may create fault-bound lens

geometries known as strike-slip duplexes (Woodcock and
The initial en échelon geometry creates step-over zones Fischer, 1986; Swanson, 1988; Cruikshank et al., 1991), where
between the ends of adjoining fault segments. The rock within multiple closely-spaced splays form elongate horses imbri-
these step-over zones must deform in order to accommodate cated between overlapping en échelon segments (Fig. 10b and
continued displacement that leads to the formation of linking d). Internal deformation within duplexes takes the form of
faults and the evolution of a through-going fault zone. The multiple P-shears and P 0 -shears in the form of high-angle
step-over zones develop in contractional and extensional styles sinistral kink-bands (Fig. 10f). The formation of P-shear
that depend on the sense of slip and stepping direction of the en linkages will also isolate and abandon the outer sections of
échelon fault segments. Dextral en échelon R-shears, as seen in terminated segments in the initial array, as can be seen in
these exposures, develop predominantly left-steps and con- Fig. 10a, b and d as well as on the larger-scale in the TL-2 fault
tractional linkages. zone (Fig. 8). Deformation of the bounding fault surfaces by
slip along the linking P-shears can be seen in Fig. 10b, similar
4.6.1. Contractional linkages to deformation of roof thrusts in typical dip-slip duplexes
The linkage of adjacent fault segments in left-stepping (Boyer and Elliott, 1982) or in Fig. 10c where the overall
arrays (Fig. 10) is typically through the development of P-shear configuration is equivalent to ‘antiformal stacking’ commonly
splay faults that are considered contractional because they developed in thrust complexes (Boyer and Elliott, 1982). For
accommodate fault parallel shortening as shearing proceeds. en échelon faults with significant overlap and multiple linkages

Fig. 10. Maps of contractional P-shear linkages between left-stepping faults. (a) Small-scale linking duplex that isolates the ends of initial en échelon segments
marked by horsetail terminations. (b) P-shear imbrication showing deformation of ’upper’ fault surface during displacement. (c) Complex imbrication at an en
échelon offset that resembles antiformal stacking, common in thrust sequences. (d) Multiple slip surfaces developed during P-shear linkage that isolated former
segment terminations. (e) P-shear imbrication and development of a contractional duplex upon linkage. (f) Through-going fault zone that links across four
(numbered) en échelon fault segments (ladder symbols represent antithetic R 0 kink bands). (g) Typical development and evolution of contractional linkages and
linkage duplexes from stepping en échelon fault segments.
466 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473

the resulting linkage duplex is typically an asymmetric fault Wall rock isolated by splay faults to form asymmetric fault-
lens (Fig. 10a and e) with similar P-shear splay faults at both bounded lenses and slabs are referred to as sidewall ripouts
ends. (Swanson, 1989, 2005). A distinctive 10-m-long slab ripout in
Linkages across contractional stepovers that utilize anti- the TL-1 fault zone (Fig. 6a) may have developed along a
thetic minor faults at high angles to the bounding faults are less restraining P-shear linkage that formed across the initial array.
common. Such contractional linking structures have been The most prominent feature in the TL-3 fault zone (Fig. 6c) is a
described by McGarr et al. (1979), Hedgecoxe and Johnson series of nested ripouts that distort the main fault surface. The
(1986), Davis (1996), and Flodin and Aydin (2004). Continued larger RI fault zone (Fig. 6d) has numerous ripout lenses
displacement in these zones is accommodated by block rotation distributed throughout the structure but most are concentrated
with antithetic slip along the intervening high angle minor along the dominant P-shear linkage contributing to the overall
faults. This can be seen in the Cape Elizabeth exposures as anastamosing structure of this part of the residual through-
linking high angle R 0 shears or kink bands in contractional going zone.
stepovers and in the clockwise rotation of oblique extension
fractures with increasing antithetic slip common in the TL-3 5. Fault sequencing and the linkage-growth model for fault
fault zone (Fig. 9d and e). zone evolution

4.6.2. Extensional linkages Distinctive structures and geometries within the Two Lights
The dominant left-stepping geometry of the initial en and Richmond Island faults can be related to different stages in
échelon segments precludes the right-stepping dextral fault a linkage-growth model for fault zone evolution. In this model,
segments needed to form extensional linkages. Fault inter- larger faults coalesce from initial en échelon R-shear arrays
action within a wide complex zone of dextral shear may yield through the development of P-shear linkages to create a
the required stepping geometries for extensional linkages but through-going fault zone. This style of fault zone evolution
they are only locally developed in the exposed fault zones, where linking structures evolve from en échelon shear fractures
particularly in the TL-2 and TL-3 structures (Fig. 9). Oblique (e.g. Vermilye and Scholz, 1999; Ahlgren, 2001) is just one of
fractures form as relays between the en échelon fault segments several models available for fault zone development. Alternate
and, in some cases, represent the interaction of two opposing models include R-shear en échelon arrays with antithetic R 0 -
extensional horsetail termination arrays (after Granier, 1985), shear linkages (McGarr et al., 1979; Hedgecoxe and Johnson,
or zones of displacement transfer between active slip patches 1986; Davis, 1996) as well as forming fault zones from
along adjacent fault surfaces as described by Martel and preexisting joints or joint arrays (Martel and Pollard, 1989;
Pollard (1989). Similar extensional linkages have been Cruikshank et al., 1990; Mollema and Antonellini, 1999;
described by Cruikshank et al. (1990) for faulted joints in Pachell and Evans, 2002; Flodin and Aydin, 2004), from en
sandstone, in experimental shear of offset cracks by Lin and échelon veins (Willemse et al., 1997; Kelly et al., 1998), in
Logan (1991) and in regional fault zones by Sibson (1985, conjugate fault networks (Kelly et al., 1998; Flodin and Aydin,
1990). Gamond (1983, 1987) also described small-scale 2004) and from extension cracks in experimental studies (Cox
examples of extensional linkages and associated mineralization and Scholz, 1988).
in granite. In the TL-2 fault zone, extensional linkages occur in
the wider spaced step-over zone where non-linking P-shear 5.1. Fault growth by P-shear linkage of initial en échelon
splays across this intact step-over yield the needed right- segments
stepping geometries.
The sequences of faults illustrated in Figs. 11 and 12 for the
4.7. Fault zone modifications Richmond Island (RI) and Two Lights (TL-2) fault zones are
depicted in a three-stage model, from the interpreted initial en
Once the initial en échelon segments have linked through échelon strike-slip fault segments, through P-shear linkage
connecting splay faults the resulting through-going zone has an structures to a final through-going fault zone as mapped in
irregular trace consisting of alternating R-shear and P-shear these exposures. The RI fault zone has greater displacement
segments. The differences in orientation of these segments and represents a later stage in fault zone evolution than the
relative to the overall slip direction means that the P-shear smaller displacement Two Lights faults and thus has a more
linkages are in restraining orientations for continued displace- complex structure.
ments. Further fault zone evolution involves the modification The RI fault zone (Fig. 11) displays an initial en échelon
of these restraining sections through both abrasive and structure with three distinct segments of 400–500C m length
adhesive wear to form a more planar through-going fault along the main zone of faulting with stepover widths of w10
zone. Abrasive wear plays an integral role in the production of and 40 m. Multiple P-shear linkages have isolated two lenses of
the cataclastic core zone of the faults. Adhesive wear host rock as contractional duplexes. The smaller (60 m long,
(Swanson, 1989, 2005), referred to as scuffing, seizing or 10 m wide) linkage duplex at the western end of the mapped
galling in the engineering literature (Rabinowicz, 1965; fault has multiple internal imbrication that creates a series of
Bowman and Stachowiak, 1996), contributes significantly to contractional horses. The larger (300 m long, 40 m wide)
the outcrop-scale lens structure of the through-going zones. linkage duplex has few internal structures but has
 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473 467

more detail below) is 35 m long with a single horsetail

termination now truncated and abandoned by P-shear linkage.
The exposed fault zone contains several different P-shear
linkage structures on a variety of scales across variable width
steps in this complex en échelon zone. Several of the linkages
on the western end of the mapped structure merge to bring the
through-going slip surface right across the initial array. The
larger (6 m wide) stepover on the eastern end of the mapped
structure had not been completely breached and contains
several oblique P shears that have failed to connect across the
gap.
The dextral strike-slip fault zones in the Cape Elizabeth
exposures are themselves part of a 10-km-long zone of oblique
en échelon faults (Fig. 2b). Each kilometer-long fault within
the zone had reached the stage of through-going fault zone
development, with the linkage of smaller-scale R-shear fault
segments in the 10–100 m scale range. These segments were
formed from even smaller-scale segments in the 2–8 m range.
The P-shear linkages in these fault zones allowed growth to
Fig. 11. RI fault zone reconstruction. (a) Three initial en échelon segments and kilometer or near kilometer lengths and increased total
preexisting quartz veins. (b) P-shear linkage through prominent P-shear ramp.
(c) Through-going fault zone with imbrication to form larger linkage lens or
displacements. Fig. 13 illustrates the observed fault geometries
duplex with w40 m of displacement. in the Cape Elizabeth exposures at various scales from the 1 m
to 10 km field of view. The 1 m view shows the linked structure
accommodated most of the displacement along its northern of the truncated initial fault segment within the TL-2 fault
bounding P-shears. In the central section of the map, several zone. The 100 m view shows the complete linkage structure of
distinctive curved splays merge and extend across an intertidal the through-going TL-2 fault zone. Note that this through-
area (dashed in Fig. 8d) to define a 100-m-long ripout structure. going fault geometry persists through the 100 m view, but is
This larger ripout appears to have sheared off the restraining absent from the 1 and 10 km views that follow. Continued
P-shear/R-shear intersection of the larger linkage duplex. displacements on the kilometer-long Cape Elizabeth fault
The reconstruction of the TL-2 fault zone (Fig. 12) is based zones could have led to even larger-scale linkage structures on
on a well-constrained sequence, with initial en échelon what could have become a major regional brittle crustal fault
segments that link to a through-going fault. These initial zone accommodating continued regional strike-slip
segments are w35–50 m long with stepover widths up to deformation.
w6 m. One of these initial R-shear segments (examined in
6. Discussion

6.1. Displacement/length relationships

Detailed mapping of a 22 m section of the TL-2 initial

segment and termination described above also reveals a
smaller-scale internal lens and splay structure. Linked en
échelon segments on a smaller scale remain in the residual
geometry of this 35 m long ‘initial’ segment. Cross-cutting
preexisting deformed quartz veins and boudin strings allowed
displacement/length (d/l) profiling along this abandoned
segment (Fig. 14). This can pinpoint sections of low d/l values
typical for breached or linked offsets (Ellis and Dunlap, 1988;
Peacock, 1991; Pachell and Evans, 2002) that are now exposed
as terminated splay faults or asymmetric lenses. This ‘initial’
35-m-long fault segment has thus formed from its own smaller
scale en échelon segments with segment lengths on the order of
w2–8 m as indicated by the spacing of displacement minima in
the Fig. 14b profile.
A maximum displacement/length (dmax/1) ratio reflects the
Fig. 12. TL-2 fault zone reconstruction. (a) Initial en échelon R-shears. (b) P- elastic strain limit for the host rock under deformation
shear linkages isolate the outer terminated segments of the initial oblique R- (Muraoka and Kamata, 1983) before linkage structures
shears. (c) Through-going fault zone with w5 m of displacement. develop. Typical dmax/1 ratios for isolated small-scale normal
468 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473

Fig. 14. Internal structure of the TL-2 fault zone. (a) Detailed map of a
continuous 22 m section of the 35 m long initial fault segment in the TL-2 fault
Fig. 13. Multi-scale views of the Cape Elizabeth fault zones. (a) 1 m view of the zone where abandoned by P-shear linkage showing internal structure and
internal structure of an initial en échelon segment in the TL-2 fault zone termination. (b) Displacement/distance profile suggests early segmentation
showing its own internal linked segmentation structure. (b) 10 m view of the within the abandoned TL-2 segment. Arrows indicate displacement minima
full length of the abandoned R-shear section (see Fig. 15a) with horsetail and their positions along the mapped fault as possible linked offsets in an
termination cutoff by P-shear linkage. (c) 100 m view of the linkage-growth earlier, smaller-scale, en échelon array of strike-slip segments in the w2–8 m
structure of TL-2 fault zone with P-shear linkages across en échelon R-shear length range. Short dashed lines represent early deformed quartz veins offset by
segments. (d) 1 km view of the full Two Lights outcrop exposure of the TL-1, faults. Long dashed lines are later stage, planar quartz veins that cross-cut the
TL-2, and TL-3 faults from left to right. (e) 10 km view of the full Cape faults.
Elizabeth exposures with the TL and RI fault zones. Gray lines in (d) and (e)
represent outcrop outline and shoreline, respectively.
competent units (Muraoka and Kamata, 1983; Peacock, 1991).
faults in sedimentary rocks have been found to range from Many of the lower dmax/1 ratios may be from just-linked faults
0.002 to 0.04 (Muraoka and Kamata, 1983), with an average of ready to accommodate additional displacement (Cartwright et
w0.01. Ratios of w0.37–0.004 and an average of 0.021 were al., 1995).
reported for strike-slip faults in sandstone/shale turbidities An isolated 50-m-long dextral fault segment in the Two
(Peacock, 1991). Dawers and Anders (1995) reported dmax/1 Lights exposure with measured displacement of 0.75 m gives a
values of 0.009–0.014 for single segment normal faults in a 7- dmax/1 ratio of 0.015 and the early terminated segment
km-long fault zone. The longer fault zone developed through described above from the TL-2 fault zone gives a dmax/1
linkage of segments and yielded a dmax/1 value for the zone of ratio of 0.034. Using an average estimate of 0.025 for the
0.013. Lithological variations may account for most of the dmax/1 ratio in these lithologies, the Richmond Island fault zone
range of values, with incompetent units allowing higher strain with total (maximum?) displacements of w40 m must be at
gradients at the fault tips and higher dmax/1 ratios than more least 1.6 km long. Likewise, total (maximum?) displacements
 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473 469

of w5 m for the Two Lights fault zones suggest likely total The change in slope that commonly occurs at the left side of
fault zone lengths of w0.2 km. these length–frequency diagrams (Fig. 15) likely reflects the
Given the way in which these fault zones have evolved resolution of the mapping technique as a sampling bias where
through the development of a progression of multi-scale the more numerous smaller faults are progressively more
linkages, the total fault length and dmax/1 ratios must have also difficult to represent (Mansfield and Cartwright, 2001; Flodin
developed in a step-wise fashion (Cartwright et al., 1995; and Aydin, 2004). The right-hand side of many diagrams often
Mansfield and Cartwright, 2001). Interaction between adjacent involves a steepening of the lower part of the curve referred to
fault segments would build higher strain gradients at the fault as ‘rollover’ (Cladouhos and Marrett, 1996). This can be
tips (Peacock, 1991) to the limiting dmax/1 ratio and lead to attributed to recording apparent fault length values where the
eventual linkage. Linkage abruptly increases total fault length, longer fault lengths are underestimated due to the finite width
temporarily decreasing the dmax/1 ratio, allowing the accumu- of the available exposure. However, the length–frequency
lation of additional displacements and building toward the diagrams from the Cape Elizabeth fault zones show a
limiting dmax/1 ratio characteristic for that particular lithology. distinctive break in slope and a flattening of the lower portion
of the graph before the rollover for longer fault lengths
(Fig. 15; TL-3 and RI fault zones in particular). The lower
6.2. Fault length distribution
sections of the Cape Elizabeth diagrams have an average
power-law exponent of only 0.83 for average fault segment
Length–frequency diagrams for fault segments within each
lengths of w15 m and greater (compared with an average
of the fault zones illustrate the length distributions within the
power-law exponent of 1.65 for w1–15 m fault lengths in the
fault segment population developed during fault zone evolution
middle sections of the diagrams). Individually, this threshold
(Fig. 15). Straight line segments on log–log plots of cumulative
fault length at the break-in-slope varies between the different
fault length imply a power-law scaling where the negative
fault zones. The fault length threshold for the TL fault zones
slope of the line (as the exponent of the power-law equation) is
varies from 8 to 17 m while the value for the larger and higher
referred to as the fractal dimension (Mansfield and Cartwright,
displacement RI fault zone is at w28 m.
2001; Flodin and Aydin, 2004). Power-law exponents for
Such breaks in slope as well as associated gaps (Fig. 15; TL-
natural fault populations reported by Cladouhos and Marrett
1, TL-2 and RI fault zones) in the length–frequency diagrams
(1996) from a variety of locations and scale ranges were found
represent the abrupt development of longer lengths during
to vary between 0.67 and 2.07. Multiple generations of faults in
linkage that is reflected in a decrease in the power-law
a single location (Flodin and Aydin, 2004) show a range of
exponent or fractal dimension for the longer fault segment
power-law exponents from 1.2 to 1.68 but only for a restricted
populations (Cladouhos and Marrett, 1996; Mansfield and
scale range of 20–110 m fault lengths. The length–frequency Cartwright, 2001). Such a break in slope (and decrease in the
diagrams for the TL and RI fault zone segments all show linear power-law exponent) for longer fault lengths can be seen in
middle sections with an average negative slope or power-law experimental data (Cladouhos and Marrett, 1996) and is
exponent of 1.65 over an average scale range of w1–15 m fault apparent in only a few of the reported fault populations (Flodin
lengths. and Aydin, 2004; Fig. 15). As suggested by Flodin and Aydin
(2004) this implies that the power-law behavior of developing
fault populations is limited to restricted scale ranges. Several
different power-law functions develop for the different scale
ranges of the evolving population as a type of multi-fractal
behavior (Cowie et al., 1995). The Cape Elizabeth faults
display early power-law behavior over an apparent initial scale
range of w1–15 m with fault growth perhaps dominated by
propagation. This early development corresponds to the
formation of most of the damage zone structures and early
segments associated with these fault zones. Later linkage-
dominated growth and localization to a longer through-going
fault zone results in a switch to a lower slope and decreased
power-law exponent for the evolving fault segment population.
This correlates to the development of the fault core zone and
the accumulation of the bulk of the displacement.

Fig. 15. Length–frequency diagram for fault segments in the Cape Elizabeth 6.3. Fluid involvement in faulting
fault zones. The linear middle sections of all the plots have similar slopes that
give an average power-law exponent of 1.65 with power-law behavior for fault
The interaction between fluids and fracturing during the
zone development in the scale range of w1–15 m. A gap and break in slope in
the lower part of the curves reflects the onset of significant fault growth through earthquake cycle is a complex process and, in the Cape
linkage that gives a lower average power-law exponent of 0.83 for fault Elizabeth exposures, led to complicated field relationships
segments greater than w15 m. between faults and veins. The geometry and cross-cutting
470 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473

relationships of these faults and veins reflect sequential and

cyclic changes in fluid pressure, differential stress and stress
orientation that developed as regional shearing progressed and
fault zones initiated and evolved with increasing displacements
through multiple earthquake cycles.
The intermittent emplacement of vertical veins orthogonal
to the shear zone boundaries reflects strain partitioning during
early dextral shear and serves as a mechanism to relieve fluid
pressure associated with regional transpression. Rising pore
fluids and increasing fluid pressures were driven by localized
crustal thickening and subsequent dehydration in the lower
crust. In the Cape Elizabeth exposures, the gently-inclined
orientation of the phyllite–quartzite host rock layering created
a barrier to vertical fluid flow capable of building local fluid
pressures that, in turn, triggered intermittent shear and
extensional fracturing. The Cape Elizabeth strike-slip faults
also have steep to moderate NW dips and could, themselves,
have created a barrier to upward fluid flow similar to high-
angle reverse faults (Cox, 1995; Robert et al., 1995; Nguyen et
al., 1998), and hence acted as valves during their displacement
history. The emplacement of veins throughout the bulk of the
ductile shearing in these exposures (Fig. 16a and b) indicates
extensional fracturing as the active valving mechanism during
most of the transpression.
The association of faults and discrete vein arrays seen here
(Fig. 16c–e) is similar to vein relationships developed during
fault valve behavior on reverse faults in contractional and
transpressional environments from Canada and Australia (Cox,
1995; Robert et al., 1995; Nguyen et al., 1998). All of these
fault localities show similar vein lengths (up to 50 m), vein
thicknesses (up to w1 m) and are associated with similar
length faults (up to w1 km) with similar displacements (up to
10s of meters). All of these fault exposures show distinct vein
arrays associated with most fault segments where there is a
decrease in vein intensity, continuity and thickness away from
the faults (Cox, 1995) as can be seen in the Cape Elizabeth
exposures (Fig. 6c). Robert et al. (1995) note the spatial
association of veins and faults and conclude that the fault zones
must have been the source of the pressured fluids involved in Fig. 16. Sequence of veins and faults due to progressive shearing in the Cape
Elizabeth fault zone exposures. (a) Extensional valving and formation of
the veining event. initially orthogonal-to-shear quartz vein arrays during strain partitioning of
Nguyen et al. (1998) also note that early brittle faulting transpressional deformation. (b) Continued shearing results in clockwise
accompanied by ductile deformation and foliation develop- rotation of rocks and structures with antithetic vein-parallel sinistral slip. (c)
ment changed to brittle faulting with extensional fracturing as Strain hardening during shear forms en échelon dextral R-shear fault zones,
shearing continued. This extensional failure formed flanking with each zone made up of their own smaller-scale en échelon segments. (d)
Linkage of early en échelon segments to form through-going fault zones. (e)
vein arrays that included extensional veins and conjugate sets Cataclastic flow, compaction and fluid expulsion results in formation of
of ‘hybrid extensional shear veins’ or ‘oblique extensional adjoining quartz vein arrays. (f) Continued shearing along fault zones deforms
veins’ (Poulsen and Robert, 1989; Nguyen et al., 1998) with earlier veins and promotes larger-scale P-shear linkages. (g) Final stage
typically small dihedral angles. Similar vein arrays and late through-going fault zone with complex history of syntectonic veining.
conjugate small dihedral angle strike-slip faults have
developed in the Cape Elizabeth exposures (Fig. 6a) where pressures during early ductile shearing at greater depths likely
the associated elongation reflects continued strain partitioning drove repeated extensional fracturing and vein development
during regional transpression. that resulted in fluid flow and valving of local fluid pressure.
Such a change in deformation style from early ductile The later transition to brittle faulting as the valving mechanism
shearing and veining (Fig. 16a and b) to brittle faulting and correlates to more shallow crustal levels where strain
veining (Fig. 16c–e) reflects the erosional unroofing and hardening increased the level of differential stress prior to
progressive strain hardening characteristic of transpressional failure. Increasing fluid pressure under these higher differential
deformation zones. Low differential stresses and rising fluid stress conditions triggered the repeated brittle shear failure in
 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473 471

the fault zones where faulting and cataclasis allowed fluid flow, Acknowledgements
relieved local fluid pressures and led to silicification of the
cataclastic fault materials. Research for this project was supported by the U.S.
During the post-seismic period of creep deformation under Geological Survey, Department of the Interior under the
relatively low differential stress, compaction, pressure solution National Earthquake Hazards Reduction Program, Award
and cataclastic flow along the fault zone are accompanied by a nos. 14-08-0001-Gl395, 14-08-0001-Gl702 and 14-08-0001-
progressive decrease in porosity and permeability during core G2060 for l987, 1989 and 1991. The University of
zone formation. This will expel fluid (Sleep and Blanpied, Southern Maine provided released time from academic
1994) and eventually seal off the fault core zone forcing fluid duties in support of this project. Reviews and comments
flow through the flanking damage zone aureole (Evans et al., by David Peacock, Pierpaolo Guarnieri and Tom
1997). Both of these effects would contribute to the return of Blenkinsop helped to improve the manuscript. Thanks to
higher fluid pressures around the fault zones that, under low Leo Algeo for assistance in plane tabling efforts at
post-seismic differential stresses, would drive the formation of Richmond Island and to students in the Structural Geology
the late extensional veins associated with each of the fault class for help in analyzing the fault and vein arrays in
zones. With continued displacement, initial core zone these exposures.
formation through cataclastic flow, pressure solution, fluid
expulsion and wall rock veining (Fig. 16e) would be surpassed
References
by further strain localization, laminar cataclastic flow and
accumulating displacement within an established core structure
Ahlgren, S.G., 2001. The nucleation and evolution of Riedel shear zones in
that would cross-cut these earlier vein arrays (Fig. 16g). deformation bands in porous sandstone. Journal of Structural Geology 23,
1203–1214.
7. Conclusions Arboleya, M.L., Engelder, T., 1995. Concentrated slip zones with subsidiary
shears: their development on three scales in the Cerro Brass fault zone,
The Two Lights and Richmond Island fault zones of Cape Appalachian Valley and Ridge. Journal of Structural Geology 17, 519–532.
Arch, J., Maltman, A.J., Knipe, R.J., 1988. Shear zone geometries in
Elizabeth are part of a 10-km-long zone of en échelon R-shears experimentally deformed clays: the influence of water content, strain rate
related to regional dextral strike-slip deformation on the and primary fabric. Journal of Structural Geology 10, 91–99.
southeast flank of the Norumbega fault system. These Barka, A.A., Kadinsky-Cade, C., 1988. Strike-slip fault geometry in Turkey
kilometer-long R-shear segments show a sequence of and its influence on earthquake activity. Tectonics 7, 663–684.
development with increasing displacement from initial en Bartlett, W.L., Friedman, M., Logan, J.M., 1981. Experimental folding and
faulting of rocks under confining pressure. Part IX. Wrench faults in
échelon faulting through P-shear linkage, and structural
limestone layers. Tectonophysics 92, 667–686.
duplexing. Fault zone modification processes of adhesive Bothner, W., Hussey, A.M., 1999. Norumbega Connections: Casco Bay, Maine
wear along the through-going faults were dominant mechan- to Massachusetts?. In: Ludman, A., West, D.P. (Eds.), Norumbega Fault
isms during continued displacement, contributing to a more System of the Northern Appalachians Boulder Colorado. Geological
planar anastamosing zone of bulk cataclastic flow. Shallow Society of America Special Paper 331, pp. 59–72.
crustal fault rocks range from pseudotachylyte to cataclasite Bowman, W.F., Stachowiak, G.W., 1996. A review of scuffing models.
Tribology Letters 2, 113–131.
and can be related to a paleo-history of coseismic slip events.
Boyer, S.E., Elliott, D., 1982. Thrust systems. American Association of
Through-going cataclastic fault zones typically develop Petroleum Geologists Bulletin 66, 239–267.
obliquely-foliated to laminated cataclasite in well-developed Cartwright, J.A., Trudgill, B.D., Mansfield, C.S., 1995. Fault growth by
core structures dominated by cataclastic flow related to segment linkage: an explanation for scatter in maximum displacement and
interseismic creep. Compaction of the cataclasite during trace length data from Canyonlands Grabens SE Utah. Journal of Structural
shear led to a decrease in porosity, increased pressure solution Geology 17, 1319–1326.
Chester, F.M., Logan, J.M., 1986. Implications for the mechanical properties of
and fluid expulsion during fault core formation, which can be
brittle faults from observation of the Punchbowl fault zone, California. Pure
related to peripheral quartz veining at the outcrop scale. This and Applied Geophysics 124, 79–106.
fluid history coupled with a linkage-growth model implies that Chester, F.M., Friedman, M., Logan, J.M., 1985. Foliated cataclasites.
regional kilometer-scale fault zone evolution proceeded Tectonophysics 93, 139–146.
through a sequence of progressively larger-scale en échelon Cladouhos, T.T., Marrett, R., 1996. Are fault growth and linkage models
segments and linkages to form a series of fluid-expelling fault consistent with power-law distributions of fault length? Journal of
Structural Geology 18, 281–293.
core zones and peripheral vein arrays as displacement
Cowie, P.A., Sornette, D., Vanneste, C., 1995. Multifractal scaling properties of
accumulates. Linkage structures allow abrupt fault growth a growing fault population. Geophysical Journal International 122, 457–
with a jump in overall fault length, which, in turn, will allow 469.
larger displacements on these now longer faults. The onset of Cox, S.F., 1995. Faulting processes at high fluid pressures: an example of fault
linkage-dominated fault growth at w15 m fault lengths is valve behavior from the Wattle Gully Fault, Victoria, Australia. Journal of
reflected in the slope geometry of the length–frequency Geophysical Research 100, 12841–12859.
Cox, S.J.D., Scholz, C.H., 1988. On the formation and growth of faults: an
diagrams for the fault segment populations. The end result is
experimental study. Journal of Structural Geology 10, 413–430.
to maintain fault geometry below a limiting maximum Cruikshank, K.M., Zhao, G., Johnson, A.M., 1990. Analysis of minor fractures
displacement/length ratio reflecting a maximum elastic fault- associated with joints and faulted joints. Journal of Structural Geology 13,
parallel strain characteristic for the host rock lithology. 865–886.
472 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473

Cruikshank, K.M., Zhao, G., Johnson, A.M., 1991. Duplex structures Mansfield, C., Cartwright, J., 2001. Fault growth by linkage: observations and
connecting fault segments in Entrada Sandstone. Journal of Structural implications from analogue models. Journal of Structural Geology 23, 745–
Geology 13, 1185–1196. 763.
Davies, R.K., Pollard, D.D., 1986. Relations between left-lateral strike-slip faults Martel, S.J., Pollard, D.D., 1989. Mechanics of slip and fracture along small
and right-lateral monoclinal kink bands in granodiorite, Mt. Abbot quadrangle, faults and simple strike-slip fault zones in granitic rock. Journal of
Sierra Nevada, California. Pure Applied Geophysics 124, 177–201. Geophysical Research 94, 9417–9428.
Davis, G.H., 1996. “Riedel relays” in deformation band shear zones, Colorado Matthai, S.K., Fisher, G., 1996. Quantitative modeling of fault-fluid-discharge
Plateau, Utah. Geological Society of America Abstracts with Programs 28, and fault-dilation-induced fluid-pressure variations in the seismogenic
A-188. zone. Geology 24, 183–186.
Davis, G.H., Bump, A.P., Garcia, P.E., Ahlgren, S.G., 1999. Conjugate Riedel McGarr, A., Pollard, D., Gay, N.C., Ortlepp, W.D., 1979. Observations and
deformation band shear zones. Journal of Structural Geology 22, 169–190. analysis of structures in exhumed mine-induced faults. U.S. Geological
Dawers, N.H., Anders, M.H., 1995. Displacement–length scaling and fault Survey Open-File Report 79-1239, pp. 101–120.
linkage. Journal of Structural Geology 17, 607–614. Mollema, P.N., Antonellini, M., 1999. Development of strike-slip faults in the
Dougherty, J.T., Lyons, J.B., 1980. Mesozoic erosion rates in northern New dolomites of the Sella Group, Northern Italy. Journal of Structural Geology
England. American Association of Petroleum Geologists Bulletin 91, 16–20. 21, 273–292.
Dunn, G.R., Lang, H.M., 1988. Low pressure metamorphism in the Orrs Morgenstern, N.R., Tchalenko, J.S., 1967. Microscopic structures in kaolin
Island–Harpswell Neck area, Maine. Marine Sedimentation and Atlantic subjected to direct shear. Geotechnique 17, 309–328.
Geology 4, 257–266. Muraoka, H., Kamata, H., 1983. Displacement distribution along minor fault
Ellis, M., Dunlap, W.J., 1988. Displacement variation along thrust faults: traces. Journal of Structural Geology 5, 483–495.
implications for the development of large faults. Journal of Structural Naylor, M.A., Mandl, G., Sijpesteijn, C.H., 1986. Fault geometries in
Geology 10, 183–192. basement-induced wrench faulting under different initial stress states.
Evans, J.P., Forster, C.B., Goddard, J.V., 1997. Permeability of fault-related Journal of Structural Geology 8, 737–752.
rocks and implications for hydraulic structure of fault zones. Journal of Nguyen, P.T., Cox, S.F., Harris, L.B., Powell, C.M., 1998. Fault-valve behavior
Structural Geology 19, 1393–1404. in optimally oriented shear zones: an example at the Revenge Gold mine,
Flodin, E.A., Aydin, A., 2004. Evolution of a strike-slip fault network, Valley Kambalda, western Australia. Journal of Structural Geology 20, 1625–
of Fire State Park, southern Nevada. Geological Society of America 1640.
Bulletin 116, 42–59. Osberg, P.H., Hussey, A.M., Boone, G.M., 1985. Bedrock geologic map of
Gamond, J.F., 1983. Displacement features associated with fault zones: a Maine. Augusta, Maine, Maine Geological Survey, scale 1:500,000.
comparison between observed examples and experimental models. Journal Pachell, M.A., Evans, J.P., 2002. Growth, linkage and termination processes of
of Structural Geology 5, 33–45. a 10-km-long strike-slip fault in jointed granite: the Gemini Fault zone
Gamond, J.F., 1987. Bridge structures as sense of displacement criteria in Sierra Nevada, California. Journal of Structural Geology 24, 1903–1924.
brittle fault zones. Journal of Structural Geology 9, 602–620. Peacock, D.C.P., 1991. Displacements and segment linkage in strike-slip fault
Granier, T., 1985. Origin, damping and pattern of development of faults in zones. Journal of Structural Geology 13, 1025–1035.
granite. Tectonics 4, 7212–7227. Poulsen, K.H., Robert, F., 1989. Shear zones and gold: practical examples from
Gratier, J.P., Gamond, J.F., 1990. Transition between seismic and aseismic the southern Canadian Shield. In: Bursnall, J.T. (Ed.), Mineralization in
deformation in the upper crust. In: Knipe, R.J., Rutter, E.H. (Eds.), Shear Zones. Geological Association of Canada Short Course Notes 6,
Deformation, Rheology and Tectonics. Geological Society Special pp. 239–266.
Publication 54, pp. 461–473. Rabinowicz, E., 1965. Friction and Wear in Materials. John Wiley, New York.
Grocott, J., 1981. Fracture geometry of pseudotachylyte generation zones: a Renard, F., Gratier, J.-P., Jamtveit, B., 2000. Kinetics of crack-sealing,
study of shear fractures formed during seismic events. Journal of Structural intergranular pressure solution, and compaction around active faults.
Geology 3, 169–178. Journal of Structural Geology 22, 1395–1407.
Hedgecoxe, H.R., Johnson, B., 1986. Interaction between en echelon faults and Riedel, W., 1929. Zur Mechanik Geologischer Brucherscheinungen. Zentral-
formation of displacement transfer zones—examples from small fault blatt fur Mineralogie, Geologie und Paleontologie 1929B, 354–368.
systems in the Llano Uplift of central Texas. Geological Society of America Robert, F., Boullier, A., Firdaous, K., 1995. Gold-quartz veins in metamorphic
Abstracts with Programs 71, 95–116. terranes and their bearing on the role of fluids in faulting. Journal of
Hickman, S., Sibson, R., Bruhn, R., 1995. Introduction to the special issue: Geophysical Research 100, 12861–12879.
mechanical involvement of fluids in faulting. Journal of Geophysical Scholz, C.H., 1988. The brittle–plastic transition and the depth of seismic
Research 100, 12831–12840. faulting. Geologica Rundschau 77, 319–328.
Hussey, A.M., 1985. Bedrock geology of the Bath and Portland 2 degree Scholz, C.H., 1990. The Mechanics of Earthquakes and Faulting. Cambridge
map sheets. Maine Geological Survey Open File Report 85-87, 82. University Press, Cambridge. 439pp.
Hussey, A.M., 1988. Lithotectonic stratigraphy, deformation, plutonism and Segall, P., Pollard, D.D., 1980. Mechanics of discontinuous faults. Journal of
metamorphism, Greater Casco Bay region, southwestern Maine. In: Geophysical Research 85, 4337–4350.
Tucker, R.D., Marvinney, R.G. (Eds.), Studies in Maine Geology, Volume Segall, P., Rice, J.R., 1995. Dilatancy, compaction, and slip instability of a
1: Structure and Stratigraphy. Maine Geological Survey, pp. 17–34. fluid-infiltrated fault. Journal of Geophysical Research 100, 22155–22171.
Janssen, C., Michel, W., Bau, M., Luders, V., Muhle, K., 1997. The North Sibson, R.H., 1985. Stopping of earthquake ruptures at dilational fault jogs.
Anatolian Fault Zone and the role of fluids in seismogenic deformation. Nature 316, 248–251.
Journal of Geology 105, 387–403. Sibson, R.H., 1989. Earthquake faulting as a structural process. Journal of
Kelly, P.J., Sanderson, D.J., Peacock, D.C.P., 1998. Linkage and evolution of Structural Geology 11, 1–14.
conjugate strike-slip fault zones in limestones of Somerset and North- Sibson, R.H., 1990. Faulting and fluid flow. In: Nesbitt, B.E. (Ed.), Fluids in
umbria. Journal of Structural Geology 20, 1477–1493. Tectonically Active Regimes of the Continental Crust. Mineralogical
Knipe, R.J., 1989. Deformation mechanisms—recognition from natural Association of Canada, Short Course on Crustal Fluids, Handbook 18,
tectonites. Journal of Structural Geology 11, 127–146. pp. 93–132.
Lin, P., Logan, J.M., 1991. The interaction of two closely spaced cracks: a rock Sibson, R.H., Moore, J.M., Rankin, A.H., 1975. Seismic pumping—a
model study. Journal of Geophysical Research 96, 21667–21675. hydrothermal fluid transport mechanism. Journal of the Geological Society
Logan, J.M., Friedman, M., Higgs, N., Dengo, C., Shimamoto, T., 1979. of London 131, 653–659.
Experimental studies of simulated gouge and their application to studies of Sleep, N.H., Blanpied, M.L., 1994. Ductile creep and compaction: a
natural fault zones. U.S. Geological Survey Open-File Report 79-1239, pp. mechanism for transiently increasing fluid pressure in most sealed faults
305–343. zones. Pure and Applied Geophysics 143, 9–40.
 M.T. Swanson / Journal of Structural Geology 28 (2006) 456–473 473

Swanson, M.T., 1988. Pseudotachylyte-bearing strike-slip duplex structures in the Tchalenko, J.S., 1970. Similarities between shear zones of different magnitude.
Fort Foster Brittle Zone, S. Maine. Journal of Structural Geology 10, 813–828. Geological Society of America Bulletin 81, 1625–1640.
Swanson, M.T., 1989. Sidewall ripouts in strike-slip faults. Journal of Tchalenko, J.S., Ambraseys, N.N., 1970. Structural analysis of the Dasht-e
Structural Geology 11, 933–948. Bayaz (Iran) earthquake fractures. Geological Society of America Bulletin
Swanson, M.T., 1990. Extensional duplexing in the York Cliffs strike-slip fault 81, 41–60.
system, southern coastal Maine. Journal of Structural Geology 12, 499–512. Vermilye, J.M., Scholz, C.H., 1999. Fault propagation and segmentation:
Swanson, M.T., 1992. Late Acadian–Alleghenian transpressional deformation: insight from the microstructural examination of a small fault. Journal of
evidence from asymmetric boudinage in the Casco Bay area, coastal Maine. Structural Geology 21, 1623–1636.
Journal of Structural Geology 14, 323–342. West, D.P., 1988. 40Ar/39Ar mineral ages from southwestern Maine: evidence
Swanson, M.T., 1999a. Kinematic indicators for dextral shearing along the for Late Paleozoic metamorphism and Mesozoic faulting. M.S. Thesis,
Casco Bay section of the Norumbega fault zone, coastal Maine. In: University of Maine, Orono, Maine.
Ludman, A., West, D.P. (Eds.), Norumbega Fault System of the Northern West, D.P., Lux, D.R., Hussey, A.M., 1989. 40Ar/39Ar mineral ages from the
Appalachians Boulder Colorado. Geological Society of America Special
Casco Bay group, southwestern Maine. Geological Society of Americas
Paper 331, pp. 1–24.
Abstracts with Programs 20, 78.
Swanson, M.T., 1999b. Dextral transpression at the Casco Bay restraining
Wilcox, R.E., Harding, T.P., Seeley, D.R., 1973. Basic wrench tectonics.
bend, Norumbega fault zone, coastal Maine. In: Ludman, A., West, D.P.
American Association of Petroleum Geologists Bulletin 57, 74–96.
(Eds.), Norumbega Fault System of the Northern Appalachians Boulder
Willemse, E.J.M., Peacock, D.C.P., Aydin, A., 1997. Nucleation and growth of
Colorado. Geological Society of America Special Paper 331, pp. 85–104.
strike-slip faults in limestones from Somerset, U.K. Journal of Structural
Swanson, M.T., 2005. Geometry and kinematic of adhesive wear in brittle
strike-slip fault zones. Journal of Structural Geology 27, 871–887. Geology 19, 1461–1477.
Sylvester, A.J., 1988. Strike-slip faults. Geological Society of America Bulletin Woodcock, N.H., Fischer, M., 1986. Strike-slip duplexes. Journal of Structural
100, 1666–1703. Geology 8, 725–735.