Submarine canyons and channels in the Lower St. Lawrence Estuary (Eastern Canada): Morphology, classification and recent sediment dynamics

Geomorphology 241 (2015) 1–18

Contents lists available at ScienceDirect

Geomorphology journal homepage: www.elsevier.com/locate/geomorph

Submarine canyons and channels in the Lower St. Lawrence Estuary (Eastern Canada): Morphology, classification and recent sediment dynamics Alexandre Normandeau a,⁎, Patrick Lajeunesse a, Guillaume St-Onge b a b

Centre d'études nordiques, GEOTOP & Département de géographie, Université Laval, Québec, QC G1V 0A6, Canada Canada Research Chair in Marine Geology, GEOTOP & Institut des sciences de la mer de Rimouski, Université du Québec à Rimouski, Rimouski, Québec, Canada

a r t i c l e

i n f o

Article history: Received 1 December 2014 Received in revised form 27 March 2015 Accepted 28 March 2015 Available online 8 April 2015 Keywords: Submarine canyons Gravity flows St. Lawrence Estuary Seabed morphology Sedimentary processes

a b s t r a c t Series of submarine canyons and channels observed in the Lower St. Lawrence Estuary (LSLE) provide an opportunity to analyze in great detail the morphology, spatial distribution and modern activity of such systems in a relatively shallow (≤300 m) semi-enclosed basin. Based on their geomorphology and physical settings, the canyons and channels were classified into four categories according to their feeding sources (ancient or recent): glaciallyfed, river-fed, longshore drift-fed and sediment-starved systems. Their activity was interpreted based on geomorphological characteristics such as the presence of bedforms related to gravity flows, backscatter intensity, axial incision and the presence of rapidly deposited layers in surficial sediments. River-fed deltas were interpreted as inactive, mainly because suspended sediment concentrations at river mouths are low, preventing the generation of hyperpycnal currents or delta-lip failures related to high sediment supply. Longshore drift-fed canyons, present where the coastal shelf narrows, were found to be episodically active probably due to earthquakes or extreme storm events. Unlike other longshore drift-fed canyons observed elsewhere in the world, they are active infrequently because of the modern low rates of sediment supply to their heads. The most active canyons are the sediment-starved type and were observed near Pointe-des-Monts. Their activity is probably due to slope failures and to the presence of strong hydrodynamic processes. Therefore, sediment supply is not the main mechanism responsible for modern canyon and channel activity in the LSLE. However, sediment supply has been an important factor during the formation of the submarine channels and canyons. Their quasi-exclusive occurrence on the Québec North Shore is attributed to its larger watershed and important sedimentary delivery during deglaciation. The Québec North Shore watershed is 20 times greater than the Québec South Shore watershed, which favored the transport of greater volumes of sediment during the early-Holocene. Moreover, the slope proximity to the shore led to the formation of longshore-drift fed systems on the North Shore when sediment supplied to rivers were transferred on a narrow shelf. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Submarine canyons and channels incise most of the world's continental shelves and act as preferential routes for the transport of continental sediments to the deep-sea. In some cases, canyons incise the continental shelf up to a depth of b100 m and transfer coastal sediments to depths of more than 1000 m, mainly by gravity flows (e.g., Gaudin et al., 2006; Mulder et al., 2012). The activity of submarine canyons and channels depends largely on the rate of sediment supply at their head and on oceanographic processes able to remobilize these sediments (Puig et al., 2014). Many submarine canyons of the world are currently active in the present sea-level highstand (Boyd et al., 2008; Khripounoff et al., 2009; Xu et al., 2010; Biscara et al., 2011; Mulder et al., 2012; Paull et al., 2013). However, recent studies have demonstrated that canyon activity does not necessarily depend on ⁎ Corresponding author. Tel.: +1 418 656 2131x4508. E-mail address: [email protected] (A. Normandeau).

http://dx.doi.org/10.1016/j.geomorph.2015.03.023 0169-555X/© 2015 Elsevier B.V. All rights reserved.

relative sea-level (RSL) positions (Boyd et al., 2008; Ducassou et al., 2009), as previously suggested by sequence stratigraphic models (Vail et al., 1977). Tectonic and climatic factors, as well as the type and volume of sediments, play a major role in canyon activity (Stow et al., 1983; Bouma, 2004). For example, tectonic processes influence the width and slope of the continental shelf and the rock susceptibility to erosion and are known to have an important role in canyon activity in California (Covault et al., 2007) and the Mediterranean Sea (Migeon et al., 2006, 2012). The type and volume of sediment include the capacity of a material to be eroded and transported, while climate largely influences precipitations, which in turn influences river floods and sediment delivery to the coast. All these processes can have more impact than RSL positions on canyon activity (e.g., Ducassou et al., 2009). Deep-water submarine fans have been largely studied for their hydrocarbon potential (Mayall et al., 2006), whereas shallow submarine fans within continental shelves have been the focus of fewer studies (e.g., Conway et al., 2012; Normandeau et al., 2013, 2014; Warrick et al., 2013). Recent studies on shallow submarine canyons using

2

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

high-resolution multibeam echosounders have allowed imaging in great detail the recent and frequent sedimentary processes acting within them (Hill et al., 2008; Conway et al., 2012; Hughes Clarke et al., 2014). Such shallow submarine canyons or channels are often observed at the head of fjords up to a depth of 400 m, at the mouth of rivers supplying considerable amounts of sediments (e.g., Conway et al., 2012). More recently, submarine fans were also discovered in very shallow sectors of continental shelves at depths of less than 60 m (Normandeau et al., 2013; Warrick et al., 2013). More attention is now being drawn to these small-scale systems in shallow environments because they can be mapped at a higher resolution using higher frequency multibeam echosounders and can provide high-resolution visualization of gravitydriven processes. Similar shallow-water systems have been reported in the Lower St. Lawrence Estuary (LSLE) (Duchesne et al., 2003; Gagné et al., 2009; Pinet et al., 2011; Normandeau et al., 2014), one of the largest and deepest estuary in the world with depths reaching ~350 m. Submarine canyons are also known to transfer eroded coastal sediments to deeper marine basins (e.g., Romans et al., 2009; Budillon et al., 2011). Yoshikawa and Nemoto (2010) demonstrated that submarine canyons can profoundly affect coastal sediment dynamics because they often constitute the end of a littoral cell. In the LSLE, coastal erosion affects a large part of the coastline (Bernatchez and Dubois, 2004). It was first hypothesized that the numerous submarine canyons and channels present along the margin of the LSLE were still active and could contribute significantly to coastal erosion by transferring coastal sediments to the Laurentian Channel (Gagné et al., 2009; Pinet et al., 2011; St-Onge et al., 2011). However, evidence for this hypothesis at the LSLE scale is still lacking. Based on a high-resolution geomorphological mapping and surface sediment analysis (box cores) approach, this study aims to: 1) provide a classification of submarine canyons and channels present in the LSLE, where they were formed in a similar environment and in similar climatic and tectonic conditions, but with distinct morphologies; 2) define the factors controlling the distribution of canyons and channels in a semi-enclosed basin; and 3) assess the sedimentary processes currently shaping these systems. This paper highlights the heterogeneity of types of incisions in an otherwise relatively similar bathymetric environment. 2. Physical setting The LSLE is one of the largest estuaries in the world, reaching 200 km in length, 50 km in width and ≤ 350 m in depth (Fig. 1). It is located between the Grenville (north) and Appalachian (south) geological provinces, in eastern Canada. The Grenville geological province is mainly composed of hard igneous and metamorphic rocks (Franconi et al., 1975). The littoral is composed of many deltas, which are at the origin of the sand-size sediment found throughout the LSLE shoreline (Lessard and Dubois, 1984; Sala and Long, 1989; Dionne and Occhietti, 1996; Bernatchez, 2003; Jaegle, 2014). The base-level of the canyons and channels, which is the lowest point a particle can attain in the marine realm, is the Laurentian Channel, at ≤350 m (Loring and Nota, 1973). The Appalachian region bordering the LSLE is formed by the sedimentary rocks of the Gaspé Peninsula and is Paleozoic in age. The LSLE basin contains in some sectors ~400 m thick of sediments that recorded glaciation, deglaciation and the establishment of postglacial conditions (Cauchon-Voyer et al., 2008; St-Onge et al., 2008; Duchesne et al., 2010). Eight seismic units were identified in the LSLE, ranging from ice-contact sediments, to deglacial (ice-proximal and ice-distal) and recent sediment deposition. More local units are related to submarine mass movements and Holocene fans, located punctually on the Québec North and South shores. A surficial deposit map of the LSLE was realized by Pinet et al. (2011) and illustrates the silt-sand nature of the margins of the Laurentian Channel and the finer silts of the trough, where most sediments originate from the North Shore (Jeagle, 2014).

3. Methodology 3.1. Data and methods Multiple multibeam echosounder datasets were merged in order to produce the high-resolution bathymetric maps of the LSLE. Depths greater than 30 m were mapped by the Canadian Hydrographic Service (CHS) on board the CCGS Frederick G. Creed using a Kongsberg Simrad EM-1000 (95 kHz; ~ 10-m grid resolution; before 2005) and an EM-1002 (95 kHz; ~ 10-m grid resolution; since 2005) and on board the Guillemot using a Simrad EM-3000 (before 2005) and an EM-3002 (since 2005). New surveys were undertaken over submarine canyons in 2012 on board R/V Coriolis II, using a Kongsberg EM-2040 (300 kHz; 1-m grid resolution) to improve the horizontal resolution of the images. This system also allowed extracting backscatter intensities to produce reflectivity maps of the seafloor. Shallower areas of the LSLE were mapped by the Centre interdisciplinaire de développement en cartographie des océans (CIDCO) on board the R/V Bec-Scie and R/V Gabrielle C using a SEA SwathPlus-M (234 kHz; 2-m grid resolution) in 2009–2011 and our group on board the R/V Louis-Edmond-Hamelin in 2012 using a Reson Seabat 8101 (240 kHz; 3-m grid resolution). Seismic data were acquired using an Edgetech X-Star 2.1 Chirp (2–12 kHz; cmscale vertical resolution) subbottom profiler and an Applied Acoustics Squid 2000 sparker (2.4 kJ; m-scale vertical resolution). In total, more than 500 km of seismic lines were acquired during the 2012 expedition on board R/V Coriolis II. Box cores were collected during the 2012 Coriolis II cruise in different sectors of the LSLE and in 2006 in the Les Escoumins area (Gagné et al., 2009). The box cores were first opened, visually described and photographed. The cores were then analyzed through a Siemens somatom volume sensation CT-Scan that allowed a non-destructive visualization of longitudinal and transverse sections of cores using Xray attenuation. Gray levels vary mainly according to density (Fortin et al., 2013) and allow identifying sedimentary structures and establishing a high resolution stratigraphy (e.g., St-Onge and Long, 2009). Grainsize analysis was performed using a Horiba laser sizer. Sediments were diluted into a calgon solution for ≥3 h, shaken and then submitted to an ultra-sound bath in order to disaggregate the particles. At least three runs were averaged. Statistical parameters were obtained using the Gradistat software (Blott and Pye, 2001). The approximate age of the surficial sediments was determined by counting the activity of the radiogenic isotope 210Pb (half-life = 22.3 yrs) via gamma emission at the Centre d'études nordiques. Dating encompasses approximately a century.

3.2. Terminology Because the submarine canyons and channels studied in the LSLE are located in an inner-shelf shallow-water environment, the terminology used here differs slightly from the one used in deepwater settings. For example, Talling (2014) used the term canyon for deep incisions (N 100 m) formed primarily by erosion and gully was used for smaller incisions (b 100 m). If we used this terminology, all the incisions in the LSLE would be considered as gullies. We thus use a different terminology that is adapted to the bathymetric setting of the LSLE: (a) Canyon is used here for incisions (generally 10s of meters) formed primarily by erosion that reach the base-level, which is the bottom of the Laurentian Channel. Canyons are generally, but not exclusively, V-shaped. (b) Gullies are small-scale incisions (b10 m deep), that generally do not reach base-level. (c) Channels are defined as conduits that are formed primarily by depositional flows (Talling, 2014) and that reach the shelf edge or the baselevel. They are generally U-shaped and have wide and relatively flat floors.

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

3

A

B

C

Fig. 1. Location of the study area. A) Multibeam bathymetry of the Lower St. Lawrence Estuary (LSLE) with location of core COR1203-27BC shown in Fig. 6F. The location of the other cores are indicated on following figures specific to each turbidite system. B) Slope map of the LSLE with location of the different types of canyons and channels (see Discussion). C) Main direction of longshore drift and sediment composition of the shoreline along the North Shore of the LSLE. Sediment composition is derived from Dubois et al. (2005) where a) high volume of erodible sediments consists of clay, clay with blocks, gravel, organic matter, sand, sand with blocks, silt or silty sand; b) mid volume of erodible sediments consists of the latter sediments on bedrock; c) low volume of erodible sediment consists of till or till on bedrock, and d) bedrock.

4. Results 4.1. LSLE geomorphology The Laurentian Channel, a deep (N300 m) trough, extends over the entire LSLE region. On the Québec North Shore, the extent of the coastal shelf (from shoreline to slope break) varies between 0 and 15 km while on the South Shore, it varies between 5 and 20 km (Fig. 1). The width of the Laurentian Channel varies from Tadoussac to Pointe-des-Monts. Near the head of the Laurentian Channel (near Tadoussac), it is ≤10 km wide and gently becomes wider near Les Escoumins (N 15 km). Near

Manicouagan, it is ~50 km wide (Fig. 2). The slopes along the Laurentian Channel also vary from Tadoussac to Pointe-des-Monts. Near the head of the Laurentian Channel (Tadoussac), the margins are generally steep (≥10°). The slopes become gentler near Manicouagan and Godbout, at less than 10°. The northern slope generally has stronger gradients, especially in Portneuf and Pointe-des-Monts, where the slope is much steeper, reaching ≥20°, compared to the South Shore, where it is b5°. The LSLE receives sediments from many rivers along its North and South shores. The combined area of these watersheds totals ~ 183,000 km2 on the Québec North Shore (excluding the St. Lawrence River and Saguenay Fjord watersheds) and ~ 9000 km2 on the

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

Depth (m)

Outardes

e’

f’ Pointe-des-Monts

b

b’

20

-200

0

-400 0 c

c’

30

-200 -400 0

0 d

d’

10 -200 -400 0

0 e

e’

15 -200

0

-400 0

f

f’

25 -200

0

-400 0

10

20

30 Distance (km)

40

Gradient (°)

f

Godbout

0

-400 0

Gradient (°)

e

20 -200

Gradient (°)

d

Manicouagan d’

a’

Gradient (°)

Mitis

Depth (m)

Betsiamites

Depth (m)

Rimouski

c’ Portneuf

a

Gradient (°)

c

Depth (m)

b

Depth (m)

a

0

Gradient (°)

Tadoussac a’ N Les Escoumins b’

Depth (m)

4

50

Fig. 2. Bathymetric and gradient profiles of the Lower St. Lawrence Estuary in sectors incised by submarine canyons and channels.

Québec South Shore. A basic surficial geological map produced by the Geological Survey of Canada (Turner et al., 2003) was used to extract general information concerning the northern and southern watersheds (Fig. 3). In both watersheds, surficial geology consists mainly of rock, till, sand and gravel, and mud. On the North Shore, ~ 62% of the total area (~ 59,600 km2) consists of rock, ~ 33% (~ 31,200 km2) of till, and ~ 5% (~ 4750 km 2 ) of sand and gravel. On the Québec South Shore, ~ 72% of the total area (~ 6400 km2 ) consists of rock, ~ 15% (~ 1350 km2) of till, ~ 10% (~ 903 km2) of sand and gravel and ~ 3% (~ 250 km2) of mud. In these two watersheds, sand and gravel are mainly located on the shores of the LSLE while rock and till are mainly located inland. On the Québec North Shore, where most of the submarine canyons and channels are observed, the coastline totals ~ 400 km in length. Dubois et al. (2005) identified the type of sediments composing the shoreline (the data are unfortunately not available for the South Shore). These results were extracted in order to give an accurate representation of the northern shoreline. From Tadoussac to Pointedes-Monts, 44% of the shoreline consist of rock, 5% consist of low volume of erodible sediment (mainly till), 5% consist of mid volume of erodible sediment (mainly sand on till/bedrock), and 46% is composed of a high volume of erodible material (mainly sand and clay) (Fig. 1C). Therefore, ~ 50% of the coastline is susceptible to erosion and can be a source of sediment for submarine canyons and channels. The mid- and high-volume of erodible sediments are mainly located where land protrudes into the water, i.e., in deltaic settings.

Watershed Mud Peat Sand and Gravel Till Rock

Depth (m) 0 360

4.2. Tadoussac A series of submarine canyons are present in the Tadoussac sector. A first canyon, named here the Saguenay canyon, is located near the center of the LSLE and incises the head of the Laurentian Channel with an 85° azimuth (Fig. 4 and Table 1). The head of the canyon incises a chaotic seabed, at a depth of ~30 m, b 3 km from the coast. The canyon reaches ~4.5 km in length at a depth of 140 m but its fan progresses into

0

45

90

180 km

Fig. 3. Map of surficial geology and watersheds on the North and South Shores of the LSLE. Source of surficial geology map: Turner et al. (2003).

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

5

Fig. 4. Geomorphology of the Saguenay canyon: A) Multibeam bathymetry of the sector; B) Slope of the sector; C) Backscatter imagery of a dune field on the lower fan with presence of a wreck; D) Photograph of the seabed shown in A; and E) Matrix grain-size distribution of the samples collected over the submarine canyon and fan shown in A.

water depth of 200 m. Its width varies between 250 and 450 m and the mean slope is ~1–1.5°. Two mass movements disturb the surface morphology of the fan (Fig. 4A). Series of subaqueous dunes are also present

on the fan and at the head of the canyon (Fig. 4B,C). The dunes near the head of the canyon are located at a depth of ≤40 m and are 10–50 m long and less than ~ 3 m high. They correspond to large dunes

Table 1 Physical properties of submarine canyons and channels from the Lower St. Lawrence Estuary. Name

Azimuth (°)

Watershed (m2)

Distance from the coast (m)

Average thalweg length (km)

Average width (m)

Average upper slope (°)

Average lower slope (°)

Lobe?

Head water depth (m)

Feeding source

Saguenay Tadoussac Les Escoumins River Les Escoumins Portneuf Betsiamites Manicouagan Godbout Pointe-des-Monts Rimouski Mitis

85 120 60 150 90 115 100 170 180 350 325

– – 793 – 3085 18700 45908 1577 – 1600 1800

N3000 230–900 500 750–1250 b2700 – 4000–6000 500–1500 130–500 9000 7000

4.5 1–1.5 1 1.5 Buried Buried 5–20 8–9.5 2–3 5 4

250–450 300–400 200–300 150–200 Buried Buried 500–1500 400–900 100–300 300–400 200–300

1–1.5 6–11 18 15 – 3–3.5 1–5 2–4 15–20 – –

1–1.5 5 7 5 1 1 0.5–1.5 1 2–3 b1 1

Yes No Yes Yes Yes Yes Yes Yes Yes Yes Yes

30 6–20 5 5–8 Buried Buried 10–20 20–30 10–25 75 80

Ancient glacier Longshore drift River Longshore drift River River River River Starved River River

6

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

(Ashley, 1990). The dunes located on the fan (125 m deep) have wavelengths of ~30–40 m and heights of ~1 m. Where subaqueous dunes are absent, bottom photographs and sediment cores on the fan reveal rounded clasts and a sediment matrix with primary grain-size modes of ~200–1000 μm (Fig. 4D, E).

A series of canyons are located closer to the shoreline and are named here the Tadoussac canyons (Fig. 5 and Table 1). They are observed near the shore surrounding the Petite-Bergeronne River (watershed: ~570 km2) in a 120° azimuth. The shores are composed of erodible sediments, more specifically of sand and glacio-marine clays with boulders

A 360

B

C

D

Fig. 5. Geomorphology of the Tadoussac canyons: A) Multibeam bathymetry of the sector; B) Backscatter imagery draped on multibeam bathymetry revealing the homogeneous intensity over the submarine canyons; C) Slopes and sediment composition of the shore (see Fig. 1C for legend) of the sector with photograph of an onland mass movement; and D) Seismic profile of the base of the canyons revealing mass movement deposits (MMDs).

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

on till (Fig. 5C). The head of these canyons are all located close to the shoreline (between 230 and 900 m). They incise the margin at a depth of 6–20 m down to a depth of 160 m. The Tadoussac canyons are short, having length less than 1.5 km. The upper and lower slopes are 5–11°. They are characterized by numerous headscarps and amphitheater-shaped erosional scars (Fig. 5A,C). At the base of the

7

canyons, a seismic profile reveals the presence of numerous mass movement deposits (MMDs) within the seismic sequence (Fig. 5D). However, there is no apparent accumulation or fan-like deposits observed on the bathymetry data (Fig. 5A,C). The base of the canyons is rather characterized by an escarpment (Fig. 5), triangular facets and a relatively flat base level, leaving some of the canyons “hanging”

Fig. 6. Grain-size properties of cores A) 06BC (Tadoussac); B) 77BC (Les Escoumins); C) 76BC (Les Escoumins); D) 19BC (Godbout); E) 26BC (Pointe-des-Monts); and F) 27BC (Pointe-desMonts). The shaded areas illustrate fining-upward sequences, interpreted as gravity flow deposits.

8

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

(e.g., Harris et al., 2014). Furthermore, there is little variation of backscatter intensity over the canyons. The only two visible features on the backscatter map are two mass movements near the base of the canyons (Fig. 5B). A core collected at the mouth of one of the canyon (COR1203-06BC) reveals a fining-upward sequence with mud clasts and D50 (median grain size) decreasing from 600 to 100 μm (Fig. 6A). The base of the core is thus interpreted as a gravity flow deposit. 210Pb analysis reveals that it reaches its supported values around 13 cm (Fig. 6A), indicating that the gravity flow deposit is slightly older than 100 yrs. Mass movement processes can also be observed onshore, near the head of the canyons. A mass movement observed onshore was triggered during the winter of 2009–2010 (Fig. 5C). 4.3. Les Escoumins The first canyon observed in the Les Escoumins sector is located at the mouth of the Les Escoumins River (watershed: 793 km2; Fig. 7 and Table 1) in a 60° azimuth. The sandy shore surrounding the canyon head is 3 km long and is isolated between rocky shores (Fig. 7C). The head of the canyon is located b 1 km from the river mouth, at a depth of 5 m. Based on the rugosity of the multibeam imagery and the presence of rocky sediments and boulders in the intertidal zone, the coastal shelf in this sector appears to have low rates of modern sediment accumulation near the head of the canyon (Fig. 7A). The canyon is 1 km long and its fan reaches 275 m deep. The upper slope is ~ 18–20° and decreases to ≤ 10° downslope. The upper part of the canyon is composed of a terrace and a small axial channel (Fig. 7D). Further downslope, the main canyon axis is composed of small crescentic bedforms and the channel bifurcates right when arriving on the lobe (Fig. 7E). The bedforms are asymmetric downslope, have a wavelength of ~ 25–50 m and are ~ 1–4 m high. The canyon width is 200–300 m

while the small axial incision within the canyon is 50 m wide. The lobe and axial channel are characterized by high backscatter values (Fig. 7B) and the absence of acoustic penetration in sub-bottom profiles (Fig. 7F), particularly where the recent channel with crescentic bedforms is observed. Three other canyons are located east and were first described by Gagné et al. (2009) using lower resolution multibeam bathymetry (10 m grid). These canyons are not associated with a terrestrial drainage system. They are located where the coastal shelf narrows to b1 km, at a depth of 5–8 m and have a 150° azimuth (Fig. 8 and Table 1). The shores east of the canyons are mainly composed of high volumes of erodible sediment, such as sand and glaciomarine clay (Figs. 1C and 8C). New high-resolution multibeam bathymetry (1 m grid) allowed us to obtain a more detailed representation of the morphological character of these canyons. The three canyons are ~1.5 km long, ~200 m large and incise the margins of the Laurentian Channel down to a depth of ≥ 200 m. The mean slopes of the three canyons are ~ 15° decreasing downslope to ~ 5°. The westernmost canyon is the closest to the shoreline (~ 550 m) whereas the two others located east are farther from it (~ 1 km). However, all three canyons are located at a similar distance from the shore at low tide. The easternmost canyon is more deeply incised than the two others, as illustrated by higher slopes on its sidewalls (Fig. 8C). The upper section of the eastern canyon is characterized by a small axial channel and a terrace (Fig. 8E). It also contains crescentic bedforms in its thalweg, whereas the two other canyons are draped by fine silts. Crescentic bedforms are similar in size to those observed at the mouth of the Les Escoumins River. The backscatter intensity is high over the easternmost canyon and fan, and is relatively low over the center and westernmost fan and canyon (Fig. 8B). Sub-bottom profiles collected over the sector reveals an absence of acoustic penetration below the fans. A core collected on the eastern fan (COR0602-77BC)

A D

E

B

F

C

Fig. 7. Geomorphology of the Les Escoumins River canyon: A) Multibeam bathymetry of the sector; B) Backscatter imagery draped on multibeam bathymetry revealing the higher intensity over the submarine fan and canyon; C) Slopes and sediment composition of the shore (see Fig. 1C for legend) of the sector; D–E) Multibeam imagery of the canyon and fan illustrating a terrace, a small axial channel and crescentic bedforms (vertical exaggeration = ×6); and F) 3D fence diagram of Chirp sub-bottom profiles illustrating the canyon and fan system.

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

A

D

9

B

C

E

F

Fig. 8. Geomorphology of the Les Escoumins canyons: A) Multibeam bathymetry of the sector; B) Backscatter imagery draped on multibeam bathymetry revealing higher intensity over the eastern submarine canyon; C) Slopes and sediment composition of the shore (see Fig. 1C for legend) of the sector; D) 3D fence diagram of Chirp sub-bottom profiles illustrating the three submarine canyons; and E–F) Multibeam imagery of the canyon and fan illustrating a terrace, a small axial channel and crescentic bedforms (vertical exaggeration = ×6).

contains a fining-upward sequence that was interpreted as a turbidite (Fig. 6B) (Gagné et al., 2009). Over the turbidite, D50 values range between 200 and 300 μm with a coarsening-upward trend. 210Pb data

suggest that the turbidite was deposited during the 1860 AD RivièreOuelle earthquake (M ~6) or the 1870 AD (M ~6–6.5) Baie-Saint-Paul earthquake (Gagné et al., 2009). Conversely, a core collected over the

A

C

B

D

Fig. 9. Geomorphology of the Les Escoumins–Manicouagan shelf composed of submarine channels: A) Multibeam bathymetry of the sector; B) Slopes and sediment composition of the shore (see Fig. 1C for legend) of the sector; C) Seismic profiles collected over the Betsiamites channel illustrating its buried nature; and D) Multibeam imagery illustrating undulations and gullies at the mouth of the Betsiamites River.

10

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

western fan (COR0602-76BC) reveals silty to sandy sediments, with D50 values ranging between 10 and 40 μm at the base and coarsening up around 50–100 μm starting at 15 cm (Fig. 6C). 4.4. Les Escoumins–Manicouagan The coastal shelf gently becomes wider between Les Escoumins and Manicouagan, reaching 15 km off Forestville (Fig. 9). The deltas of three major rivers (Portneuf, Betsiamites and Outardes) are present in this sector and their submarine components are composed of channels (Table 1). These channels formed on the coastal shelf above the slope break of the Laurentian Channel (Fig. 9B). The upper slope of the shelf directly near the mouth of the Betsiamites River is 3.5° and decreases to less than 1.5° offshore. A seismic profile collected off the Betsiamites River reveals the buried nature of these channels (Fig. 9C). Gullies are observed near the mouth of the river (Fig. 9D). 4.5. Manicouagan The channels observed at the mouth of the Manicouagan River are located offshore the city of Baie-Comeau (Fig. 10 and Table 1). This river has the largest watershed of the North Shore (~45 900 km2). The head of the channels are located directly at the mouth of the main river channel but are still relatively far from the coast (~ 5 km). The head of the channels are observed at a depth of ~10–20 m. The upper slope is short and varies alongshore at ≤ 10°. The gradient then decreases to 1.5° downslope. These channels are the longest observed in the LSLE, with the southernmost reaching 20.5 km. The other channels vary between 5 and 10 km in length. All the channels vary between

500 and 1500 m in width. They all show varying levels of sinuosity. The southernmost has the highest incision gradients at a depth of 175 m (Fig. 10B). The two channels located at the river mouth show lower gradients on their margins and are characterized by sediment infilling (Fig. 10C, D). Sedimentary processes on the shelf mainly include longshore-drift related bedforms as indicated by the presence of transverse bars (Fig. 10D). 4.6. South Shore On the South Shore, two channels are observed offshore the Rimouski (watershed: ~1600 km2) and Mitis Rivers (watershed: ~1800 km2) (Fig. 11 and Table 1). They are both ~200–400 m large and have a relief of ≤10 m. The head of the channels are 70–80 m deep and are 7–10 km away from the shoreline. Their mean slope is 1°. The channels are ~4–5 km long and their fans extend in the Laurentian Channel but are hardly visible on the multibeam imagery. Both of them show a smooth morphology attributed to a sediment drape. 4.7. Godbout The Godbout canyons are located at the mouth of the Godbout River, which has a watershed of 1577 km2. The river mouth consists of sandy sediments that extend on 6 km between rocky shores (Fig. 12). Two other canyons located west of the river mouth probably relate to its ancient positions, 1500 m from the shoreline. The easternmost canyon incises the shelf to 500 m from the shoreline and its head is located 2 km east of the river mouth (Fig. 12 and Table 1), at a depth of 20–30 m. The three canyons vary in length between 8 and 10 km. The

A

C

B

D

Fig. 10. Geomorphology offshore the Manicouagan River mouth. A) Multibeam bathymetry of the sector; B) Slopes and sediment composition of the shore (see Fig. 1C for legend) of the sector; C) Seismic profile collected on a channel located at the mouth of the current Manicouagan River discharge; and D) Multibeam imagery of the channels at the mouth of the current river position illustrating longshore-related bedforms and the infilling of the channels.

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

11

Fig. 11. Geomorphology of the Rimouski and Mitis channels: A) Multibeam bathymetry of the sector; B) Slopes of the sector; C) Multibeam bathymetry illustrating the Rimouski channel and its draping morphology; and D) Multibeam imagery illustrating the Mitis channel and its partly buried nature.

slope of the upper part of the canyon is ~ 4° while the lower part decreases to ≤ 1°. Scarps are observed along the eastern canyon (Fig. 12A,B). A box core (COR1203-19BC) collected in the thalweg of the eastern canyon reveals silty-sand to sandy-silt sediment (Fig. 6D). No distinct facies are observed in the CT-Scan imagery. Sedimentary bedforms at the Godbout River mouth are characterized by undulations that are ~ 2 m high and 20–50 m long (Fig. 12C). These bedforms are mainly present on the shelf break but do not extend to the eastern canyon thalweg. They are localized in the first kilometer of the river mouth. A sub-bottom profile collected in the eastern thalweg reveals the absence of acoustic penetration over the canyon, whereas its base consists of ≤4 m of hemipelagic sediment (Fig. 12E). Furthermore, no bedforms are observed within the thalweg on the sub-bottom profiles, nor on the multibeam imagery (Fig. 12D). 4.8. Pointe-des-Monts Three distinct submarine canyons are observed in the Pointe-desMonts sector and were first described by Normandeau et al. (2014) (Fig. 13 and Table 1). The two canyons located to the east have similar physical settings. The shores surrounding the area and the heads of the canyons almost entirely consist of bedrock, except for small pocket beaches (Fig. 13C). These canyons have length of ≤ 3.5 km, down to ≤300 m deep. Their width varies greatly and ranges between 100 and 300 m. The upper slope is ~ 15–20° until ~ 125 m deep. This section is composed of numerous small gullies that feed the tributaries of the canyons (Fig. 13A). The canyon floor slope then decreases to 2–3°, 390 m from the head until the end of the fans located at a depth of

300 m. The canyons relief at their head varies between 10–15 m. Further downslope, it increases to 30 m and gently decreases down to 10 m before becoming unconfined. Numerous crescentic bedforms (Fig. 13E, F) are observed within the thalweg of the canyons. They are asymmetric downslope, have ≤ 60 m wavelengths and are ~1–3 m high. Sub-bottom profiles collected in the Pointe-des-Monts canyons show an absence of acoustic penetration. A hemipelagic drape is observed on each side of the canyons (Fig. 13D). Two cores were collected in the Pointe-des-Monts sector: a first in the thalweg of the eastern canyon (COR1203-26BC; Fig. 6E) and a second just outside the submarine fans (COR1203-27BC; Fig. 6F). The core collected in the thalweg of the eastern canyon is 10 cm long and consists of homogenous fine sand (D50 = 110–120 μm). 210Pb analysis reveals that this core did not attain its supported values, indicating recent sediment deposition (b100 yrs). The core collected just outside the submarine fans is 47 cm long and consists of fine to medium silt (D50 = 12–23 μm) but is composed of two modes (~10 and ~50 μm). 5. Discussion 5.1. Classification and modern sediment dynamics Four genetic types of submarine canyons and channels are observed in the LSLE. These systems were identified and classified based on their sediment source: 1) glacially-fed canyon; 2) longshore drift-fed canyons; 3) river-fed canyon and channels; and 4) sediment-starved canyons. A qualitative summary of their characteristics is provided in Table 2. A combined geomorphological and sedimentological approach

12

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

A

B

C

D

E

Fig. 12. Geomorphology of the Godbout canyons: A) Multibeam bathymetry of the sector; B) Slopes and sediment composition of the shore (see Fig. 1C for legend) of the sector; C) Multibeam bathymetry illustrating undulations at the river mouth; D) Multibeam imagery illustrating the absence of bedforms along the thalweg of the canyon; and E) Seismic profile revealing the absence of acoustic penetration in the thalweg of the canyon and the draping post-glacial sediments on the fan.

(e.g., Lastras et al., 2007; García-García et al., 2012; Hill, 2012; Babonneau et al., 2013; Obelcz et al., 2014) was used to describe the typical morphological characteristics of these systems and to assess their recent activity. Activity of submarine canyons is defined as a modification of canyon morphology by erosion and deposition in recent times as opposed to the occurrence of general background sedimentation (Mountjoy et al., 2014). It can also be defined as the frequency of particle-laden currents that are intrinsic to submarine canyons (Mulder et al., 2012). Here, activity is defined as a modification of submarine canyon and channel morphology by particle-laden currents directly related to these systems. If currents unrelated to the canyons erode the seafloor in a given region, the canyons are considered inactive. In order to assess their activity, the canyons located within the same geological and geomorphological context were grouped together. Therefore, the most active canyon or channel is used when defining the activity of each group. This way the inactivity of ancient channels that were abandoned to the profit of another are not accounted for.

5.1.1. Glacially-fed canyon The chaotic morphology of the seabed at the head of the Saguenay canyon (Fig. 4) is interpreted as the remains of a temporary stillstand of the Laurentide Ice Sheet margin, which is in accordance with the interpretation of marine charts by Dionne and Occhietti (1996). Therefore, the canyon that incises this morainic system is interpreted as a relic of a proglacial canyon formed during ice stabilization phases at the mouth of the Saguenay Fjord (Dionne and Occhietti, 1996; Locat and Sanfaçon, 2000). Submarine canyons and channels related to glacial outflow are often located off cross-shelf trough (e.g., Rise et al., 2013; Roger et al., 2013). Trough-mouth fans often comprise numerous gullies that are related to meltwater outflows (Dowdeswell et al., 2006; Pedrosa et al., 2011). Unlike other glacially-fed systems, the Saguenay fan is incised by only one canyon. This absence of widespread incisions on the fan could be due to the presence of the Laurentian Channel and strong tidal currents that confined the flows on a 5 km width. It can also be related to the stagnation of the glacier during its retreat which lasted less than 1000 yrs around 11 ka BP (Dionne and Occhietti, 1996).

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

A

13

B

D

E

C

F

Fig. 13. Geomorphology of the Pointe-des-Monts canyons: A) Multibeam bathymetry of the sector; B) Backscatter imagery draped on multibeam bathymetry revealing higher intensities over the submarine canyons; C) Slopes and sediment composition of the shore (see Fig. 1C for legend) of the sector illustrating the low volume of sediment; D) 3D fence diagram of Chirp sub-bottom profiles illustrating the three submarine canyons; and E–F) Multibeam imagery illustrating the high number of crescentic bedforms along the thalweg of the canyons (vertical exaggeration = ×6).

The incision of the Saguenay canyon into gravelly sediments indicates that it was formed by energetic gravity flows, which is most likely the result of hyperpycnal flows or mass movements related to high sediment discharge at the mouth of the glacier (e.g., Ó Cofaigh et al., 2004; Piper et al., 2007; Armitage et al., 2010; Roger et al., 2013). The coarse nature of the sediments composing the seafloor (Fig. 4D, E) supports this hypothesis, because such coarse material needs energetic flows to be transported. Today, strong currents due to upwelling at the head of the Laurentian Channel and internal tides limit the deposition of fine sediments in the sector (Ingram, 1983) and lead to a lag deposit on the seafloor. Strong currents are evidenced on the seafloor by the presence of a tidal scour (Fig. 1) similar to the one observed by Shaw et al. (2014; their Fig. 11) on the seabed of the Bay of Fundy, a region

known for its strong tidal currents. Numerous dune fields are present in areas of finer-grained sediments. These dunes are observed on the chaotic seafloor at the head of the Laurentian Channel, on the large mass movement located south-west of the canyon and on the distal part of the fan. They are not present within the canyon (Fig. 4A). These dunes are most likely associated with tidal currents or internal tides, as interpreted for similar dunes located upstream in the St. Lawrence Estuary (Bolduc and Duchesne, 2009), and are not related to recent sedimentary processes within the canyon. Therefore, there is no evidence for the recent passage of gravity flows in the canyon. The St. Lawrence River or the Saguenay Fjord cannot produce hyperpycnal flows that could currently be remolding the seafloor of the Saguenay canyon because sediment concentrations are too low (Mulder and

Table 2 Qualitative properties of submarine canyons and channels in the Lower St. Lawrence Estuary. Characteristics/classification

Glacially-fed

Longshore drift-fed

River-feda

Sediment-starved

Source of sediment

Glacially derived sediments 3 km

Remobilized fluvial and coastal erosion b1 km

Fluvial sediments 0–5 km

Sediments on the margins of the Laurentian Channel No sediment supply

3 km Hyperpycnal flows and slope failures No evidence 1–2° Saguenay

b1 km Slope failures and storm-induced gravity flows Evidence of episodic activity N5° Les Escoumins, Tadoussac

0–5 km Hyperpycnal flows and slope failures

b1 km Slope failures and internal tides (?)

No evidence 0–5° Betsiamites, Rimouski, Mitis, Portneuf, Godbout, Manicouagan

Evidence of frequent activity N10° Pointe-des-Monts

Proximity to a source of sediment supply Coastal shelf Main inferred processes Modern activity Slope Names a

Applies only to river-fed delta systems.

14

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

Syvitski, 1995). Moreover, the Saguenay Fjord is composed of three deep basins and a sill at its mouth, favoring sediment capture in the fjord rather than in the LSLE (Locat and Lévesque, 2009; St-Onge et al., 2012). This sector of the LSLE is rather today host of upwelling and internal tides that induce important bottom currents (Ingram, 1983). 5.1.2. Longshore drift-fed canyons Both the eastern canyons of Les Escoumins (Fig. 8) and the Tadoussac canyons (Fig. 5) are classified as longshore-drift fed submarine canyons. The morphology of these systems is characterized by a narrow shelf, a disconnection to a terrestrial drainage system and a steep slope. The sediment supplied to their heads is transported through longshore-drift. The narrow shelf allows sediments to be trapped between the Laurentian Channel margin and the shoreline. Sediments transported through longshore drift do not bypass the canyons. They are rather trapped at their head before being redistributed along the canyon floor and fan. The three eastern canyons of Les Escoumins have distinct morphological signatures. The western and central canyons are draped by fine hemipelagic sediment. The multibeam imagery does not reveal any bedforms along the axis of these two canyons. The presence of fine silts in core 76BC (Fig. 6C) indicates that sediment gravity flows did not occur in the recent past. Unlike the central and western canyons, the eastern canyon presents numerous bedforms along its thalweg and fan (Fig. 8E,F). Small crescentic bedforms and axial incision are observed along its axis. Its margins also show the highest slope gradients of the sector due to axial incision (Fig. 8C). According to Baztan et al. (2005), axial incision can be interpreted as the downcutting of the canyon thalweg during an erosion phase. These morphological properties suggest that gravity flows recently eroded the canyon bottom. Crescentic bedforms, possibly associated with supercritical currents (i.e., cyclic steps; Cartigny et al., 2011), are known to be present where gravity flows are frequent (Smith et al., 2007; Babonneau et al., 2013; Mazières et al., 2014). The higher values of backscatter intensity (Fig. 8B) and the absence of acoustic penetration (Fig. 8D) over the eastern canyon further support the presence of coarse particles related to the recent deposition of gravity flow deposits. Gagné et al. (2009) reported the presence of a turbidite at 30 cm depth on the fan (Fig. 6B), dated to ~ 1850. This date is in the range with two earthquakes that could have remobilized shelf sediments in 1860 (M ~ 6) and 1870 (M ~ 6–6.5) that occurred in the Charlevoix-Kamouraska seismic zone (CKSZ; Smith, 1962; Gouin, 2001). A mass movement triggered turbidite is thus the most likely explanation for this deposit. Core 77BC also shows no sign of fine hemiplegic sediments such as those observed in the neiboughring fan (Fig. 6B), but rather contains sand-size particles (200–300 μm) increasing towards the surface. This increase of sand suggests recent hydrodynamic processes related to the eastern canyon rather than hemipelagic sedimentation. However, there is no distinct sedimentary facies within this increase in grain-size related to a typical gravity flow deposit. The Tadoussac canyons were also classified as longshore drift-fed canyons even if their morphologies are not typical of other documented longshore drift-fed canyons. Unlike the Les Escoumins canyons, they do not incise the slopes where the shelf narrows but are nonetheless located on a narrow shelf. Further, their potential source of sediment could be related to longshore drift, because they are not directly located at the mouth of the Petite-Bergeronne River, but rather incise the shelf close to the shoreline. Sediment inputs from the Petite-Bergeronne River are redistributed along the coast and are likely trapped near the canyon heads. This redistribution is possibly limited because there has not been any erosion and accumulation along the coast during the 20th century (Dubois et al., 2005). Erosion and deposition along the coast are generally good indicators of active coastal processes (e.g., Yoshikawa and Nemoto, 2010). A gravity flow deposit is observed in a core collected at the mouth of one of the canyons (Fig. 6A). Based on 210 Pb dating, this deposit is older than 100 yrs and could also be related

to the AD 1860 (M ~ 6) or 1870 (M ~ 6–6.5) earthquakes in the Charlevoix-Kamouraska seismic zone (CKSZ). The amphitheatershaped scars at the head of the canyons suggest that the canyons evolved by retrogressive slope failures (e.g., Twitchell and Roberts, 1982; Farre et al., 1983). These slope failures are probably related to earthquakes because these canyons are located in the CKSZ (Lamontagne, 1987). The CKSZ is the most active seismic zone in Eastern Canada where five earthquakes of M ≥ 6 or more occurred during the last 350 years (Lamontagne, 2009). More than 100 mass movements were identified in the LSLE and the Saguenay Fjord (Duchesne et al., 2003; Locat et al., 2003; St-Onge et al., 2004, 2012; Cauchon-Voyer et al., 2008, 2011; Pinet et al., 2015) and most of them were interpreted to be the result of earthquakes from the CKSZ or the Saguenay area (Locat, 2011). St-Onge et al. (2004) proposed a recurrence of large earthquakes (M ≥ 6.75) of 300 years before 4 ka BP and of 2000 years after 4 ka BP for the Saguenay Fjord. However, smaller earthquakes could have caused smaller slope failures, such as proposed for slope failures observed in these longshore-drift fed submarine canyons. The absence of sediment accumulation at the base of these canyons suggests that slope failures were relatively infrequent during the late-Holocene which is in agreement with the earthquake recurrence rate of St-Onge et al. (2004). The absence of fan-like deposits could also be explained by the presence of an escarpment at the base of the canyons (Fig. 5C) that has all the morphological characteristics of a fault-scarp. The Tadoussac canyons are similar morphologically to hanging canyons described in Haida Gwaii, British Columbia (Harris et al., 2014). In both systems, an escarpment separates the slope from the base-level and there is absence of fan-like morphologies. Hanging canyons form where the rate of canyon erosion is lower than the vertical displacement of a fault, leaving the canyon mouth “hanging”. We speculate that the similar geomorphology observed in the Tadoussac canyons is attributed to vertical displacement along a fault due to lateQuaternary tectonic activity, forming a fault-scarp. Moreover, the presence of strong bottom currents in this sector can also mobilize sediments transported through the canyons and reduce the relief of their fans. In both longshore-drift fed submarine canyons of Tadoussac and Les Escoumins, sediments and bedforms indicate that recent sedimentary processes occurred along them during the last few centuries. The triggering mechanisms of these recent gravity flows are most likely related to mass movements along the head of the canyons, as observed in many other longshore-drift fed settings (Budillon et al., 2011; Biscara et al., 2013; Normandeau et al., 2013). Other processes such as advection of shelf sediments during major storms can play a role in sediment transfer, although there is currently no evidence for these processes in these sectors. These canyons most likely formed by a combination of upslope retrogressive failures and downslope eroding gravity flows. The Tadoussac canyons are a good example where upslope retrogressive failures are still ongoing. Numerous headscarps are slope-confined rather than shelf-indenting. These slope-confined headscarps suggest that the shelf-indenting canyons were formed by similar processes (e.g., Farre et al., 1983; Green et al., 2007). 5.1.3. River-fed channels and canyons The majority of the canyons and channels observed in the LSLE are located at the mouth of major rivers. These river-fed systems were subdivided into two categories: river-fed delta systems and river-fed canyon systems. The former were formed during the progradation of deltas in the LSLE. These systems are observed at the mouth of major rivers of the Québec North and South Shore (Portneuf, Betsiamites, Manicouagan, Godbout, Rimouski and Mitis), suggesting a formation by either hyperpycnal currents or slope failures related to high accumulation rates. During deglaciation, sedimentation rates and suspended sediment concentration were higher (Duchesne et al., 2010) and led to the formation of these deltas. Between Les Escoumins and Manicouagan, the channels formed mainly on the coastal shelf, while

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

their fan formed in the Laurentian Channel (Fig. 9). On the counterpart, the channels and canyons of the Manicouagan and Godbout deltas were mainly formed on the margins of the Laurentian Channel (Figs. 10 and 12). These two delta systems are characterized by relatively low mean slopes (less than 2°) and are the largest features in the LSLE. Bedforms or axial incisions were not observed within the channels and fans on any of the delta systems. In addition, box cores collected in the Godbout canyon did not reveal any sign of rapidly deposited layers (Fig. 6D). The majority of the channels in the LSLE are buried or draped by postglacial mud (Figs. 9, 10 and 11), most likely due to the lower modern rates of sediment supply by the rivers. According to the Inland Waters Directorate (1978), these rivers contain less than 50 mg L−1 of suspended sediment. This value is considerably less than the 35–45 kg m− 3 needed to generate a hyperpycnal flow (Mulder et al., 1998). Sediment outflowing from the rivers are mainly deposited near the river mouths and remobilized along the coast by longshoredrift (Fig. 10D). The absence of recent delta-slope failures such as observed in deltas and fjord-head deltas from British Columbia (Conway et al., 2012; Hill, 2012; Hughes Clarke et al., 2014) also indicates low accumulation rates at the mouth of the rivers while the longshore drift bedforms (transverse bars) indicate remobilization of sediment alongshore. Transport rates by longshore drift have thus outpaced the rate of sediment supply that could form large mass movements due to high rates of sedimentation. Where sediment supply is slightly higher, such as at the mouth of the Betsiamites and Godbout Rivers, undulations and gullies are observed on the slope. These undulations and gullies are restricted to the nearshore environment. The undulations are most likely due to local small instabilities since hyperpycnal currents cannot form due to low sediment suspended concentration in the rivers. The slightly higher rates of sediment supply at the mouth of these rivers can also be due to important coastal erosion at the mouth of the deltas (Bernatchez and Dubois, 2004) and more accumulation on the coastal shelf (Fig. 9D), rather than to higher supply from rivers. The river-fed canyon observed at the mouth of the Les Escoumins River is unrelated to the construction of a delta (Fig. 7). This canyon occurs on a steep slope on the shoulder of the Laurentian Channel. Its axial channel contains crescentic bedforms similar to those observed on the longshore drift-fed canyons of Les Escoumins (6 km apart) (Fig. 8). High backscatter values on the fan and thalweg also reveal the coarse nature of its surface sediments (Fig. 7B). Therefore, we suggest that it was also active during the 1860 or 1870 earthquake that likely led to the turbidite layer observed in the longshore-drift fed canyons. Although there is no sedimentological evidence for its episodic activity, its geomorphology shows many similarities with the longshore driftfed canyons of Les Escoumins. Their recent activity appears to be closely related. 5.1.4. Sediment-starved canyons Having no sediment input at their heads, the Pointe-des-Monts submarine canyons are the only sediment-starved canyons of the LSLE. However, as they are located on a steep slope, slope instabilities and hydrodynamic processes can remobilize their enclosing sediments, leading to active gravity flows. Normandeau et al. (2014) provided evidence that the Pointe-des-Monts canyons are currently active even without sediment supply at their heads through repeated multibeam bathymetry. The presence of crescentic bedforms that were displaced upslope between 2007 and 2012 suggest they are related to supercritical currents (i.e., cyclic steps; Cartigny et al., 2011; Hughes Clarke et al., 2014). These results, combined with higher backscatter values over the canyons indicate that sediments were transferred downslope recently. A core collected over the submarine canyons contains fine sands while a core collected just outside the fan contains the typical silty sediment of the Laurentian Channel (Fig. 6E,F). The two cores were collected at approximately the same depth (~300 m). The sandy sediments in the canyon thus reflect the presence of active gravity flows, possibly related to the migration of the bedforms. The Pointe-des-Monts canyons are the

15

only ones that show such well-defined crescentic bedforms along the canyons and that have evidence of very recent activity (≤5 years, see also Normandeau et al., 2014). Although sediment is not supplied to the canyons, crescentic bedform movements reveal that gravity flows are more frequent and recent within these canyons than in any other submarine canyon or channel in the LSLE. 5.2. Canyon and channel emplacement The majority of canyons and channels observed on the LSLE occur on the Québec North Shore while only two are observed on the Québec South Shore. Their predominant occurrence on the Québec North Shore and their rare occurrence on the Québec South Shore can be explained by: 1) the proximity of the slope to the shoreline on the Québec North Shore; 2) the steepness of the slopes bordering the Laurentian Channel on the Québec North Shore; and 3) the high volumes of sediment that were transported towards the LSLE during deglaciation on the Québec North Shore (Bernatchez, 2003; Dietrich et al., 2014). The longshore drift-fed submarine canyons have steep slopes (~10–15°) and are located less than 1 km from the shoreline. These systems can thus intercept sediment transported alongshore and remobilize them downslope. On the Québec South Shore, the slopes of the Laurentian Channel can also be steep (~10°) and could act as routes to remobilize sediments downward. However, canyons in this sector cannot form by intercepting longshore drift sediments because the shoreline is located more than 5 km away from the slope break. Therefore, longshore sediment transport occurs several kilometers away from the shelf break. The Québec South Shore is rather incised by mass movements at different sites (Pinet et al., 2015). Unlike the Tadoussac canyons, mass movements on the South Shore did not retrogress upslope to form canyon systems. On the Québec North Shore, between Les Escoumins and Manicouagan, the width of the coastal shelf is similar to the South Shore (≥ 5 km wide) and numerous submarine channels formed on the seafloor. In this sector, sediment supply during deglaciation most likely played a major role in forming the channels by hyperpycnal flows. The Québec South Shore has a watershed that reaches 9000 km2 while the Québec North Shore's watershed reaches 183,000 km2. The Manicouagan River by itself has a larger watershed (~45,000 km2) than the entire Québec South Shore. Two of the largest rivers on the Québec South Shore, the Mitis and Rimouski rivers, have 1600 and 1800 km2 watersheds respectively. Sediments from these rivers were capable of forming channels on the slope but they did not enlarge them as much as those of the Québec North Shore. The Québec South Shore rivers rather deposited their sediments on the gentle slopes of the coastal shelf and likely led to rarer gravity flows than on the Québec North Shore due to lower sediment concentrations in rivers during deglaciation. A similar watershed to the Mitis and Rimouski rivers is observed on the northern shore for the Godbout River. However, unlike the Rimouski and Mitis rivers, the Godbout River directly delivers sediment to the slope of the Laurentian Channel due to the absence of a coastal shelf, leading to a direct sediment transfer from land to the Laurentian Channel (e.g., Babonneau et al., 2013). 5.3. Controls on recent sedimentary processes Stow et al. (1983) and Bouma (2004) suggested that the activity of submarine canyons is influenced by tectonics, the type and volume of sediments, RSL changes and climate. In the LSLE, the proximity to the coast (tectonic setting) is important for the presence of a canyon or channel. However, there is no clear relationship between the proximity to the coast and the activity of gravity flows in the last centuries. The longshore drift-fed canyons (Tadoussac and Les Escoumins) show evidence of episodic activity while the sediment-starved canyons (Pointe-des-Monts) show evidence of frequent activity (less than 5 years). The head of these canyons are all located within 1 km from

16

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

A

B Escoumins (river) Escoumins (longshore drift) Pointe-des-Monts Tadoussac Godbout Betsiamites Saguenay Portneuf Manicouagan

0 -50

Depth (m)

-100 -150

0 No evidence of activity

-50

Evidence of episodic activity

-100

Frequently active

-150

-200

-200

-250

-250

-300

-300

-350

-350 0

5

10

15

20

25

0

5

10

15

20

25

Distance (km) Fig. 14. Plots of bathymetric profiles grouped by A) region and B) evidence of recent activity.

the shoreline but show different rates of activity and different volumes of sediments at their heads. Furthermore, the head of the channels in the river-fed delta systems are also located near the shoreline (less than 2 km in some cases) and do not show recent gravity flow deposits and bedforms. Therefore, a proximity to the sediment source cannot solely explain recent submarine gravity flow activity in the LSLE. Plotting the slopes of each channel and canyon system (Fig. 14) reveals that all the inactive channels are located on the lower slopes, while the canyons showing evidence of recent activity are located on the steepest ones. Slope is an important factor for the activity of submarine canyons (Piper and Normark, 2009). Therefore, the steeper the slope, the stronger are the probabilities that gravity flows will remobilize sediments. The importance of the slope compared to the proximity of a sediment supply in modern times is probably due to the fact that sediment supply is generally low in the LSLE. Compared to other regions of the world, the Laurentian Highlands (Québec North Shore) consist of very hard igneous rocks in a ~ 72% proportion, while sand and gravel only account for ~5% of the total Québec North Shore watershed. Without rivers delivering large volumes of suspended sediment to the coast, coastal erosion would be the main source of sediment supply today. Recent studies emphasize that the LSLE coastline is currently undergoing intense coastal erosion (Bernatchez and Dubois, 2004). However, coastal erosion does not appear to be sufficient to maintain frequent canyon activity as observed on the Californian coasts (Smith et al., 2005; Paull et al., 2013) or the Bay of Biscay coasts (Mulder et al., 2012; Mazières et al., 2014) where annual events of sediment transfer were recorded. Furthermore, eroded coastal sediments are probably redistributed along the deltas because there is no evidence for strong sediment transport near the canyons of Tadoussac, Les Escoumins and Pointe-desMonts. These low rates of sediment transport limit the activity of the Les Escoumins and Tadoussac canyons because sediment is not transported to their head as frequently as in other longshore drift-fed submarine canyons offshore Gabon (Biscara et al., 2011), California (Smith et al., 2005, 2007) or France (Mazières et al., 2014). Furthermore, the low abundance of sediment at the mouth of rivers flowing into the LSLE can explain the inactivity of the river-fed delta systems (Inland Waters Directorate, 1978). All the rivers of the Québec North Shore except the Godbout River have less than 50 mg L−1 of suspended sediment. Gravity flows cannot form if the concentration of suspended sediment is insufficient to create dense water that can flow by hyperpycnal currents. The channels are thus being buried by normal settling of suspended sediment or are remobilized by bottom currents. In summary, lower slopes need higher sediment supply to generate hyperpycnal currents or slumping whereas steeper slopes do not need as much sediment supply in order to transfer continental sediments to the Laurentian Channel. Slopes appear to be the main factor controlling the episodic activity of turbidite systems in the LSLE today. Sediment supply strongly limits the very recent activity of some submarine canyons and channels in the LSLE. However, an intriguing

characteristic of the LSLE is the activity of the sediment-starved submarine canyons. Normandeau et al. (2014) suggested that slope failures and internal tide-related currents could be responsible for the observed activity. However, no slope failures were observed during the 5-year repeated bathymetric surveys, and internal tide-related currents are not known for generating crescentic bedforms in other environments. The comparison of canyons and channels located in the LSLE also indicates that the Pointe-des-Monts canyons are the only frequently active systems despite the absence of sediment supply at their head, outlining the fact that unique and poorly understood processes are ongoing in this sector. There is a need to investigate the triggering mechanism for generating turbidity currents in the Pointe-des-Monts sector, when other sectors of the LSLE having similar physical settings and more sediment supply (Les Escoumins for example) reveal little or no sign of recent gravity flow activity. 6. Conclusions The LSLE is a unique Quaternary sedimentary system that allowed the initiation and development of morphologically contrasting submarine canyons and channels within a semi-enclosed basin. The diversity of types of canyons and channels located in a similar tectonic, eustatic, climatic and morphological setting allowed the observation of the different factors controlling their distribution and their modern sedimentary processes. The combined geomorphological and surface sediment analysis of the canyons and channels of the LSLE lead to the following conclusions: 1) Four genetic types of submarine canyons and channels are observed in the LSLE based on their feeding sources: glacially-fed, longshore drift-fed, river-fed and sediment-starved systems. The river-fed systems are further divided into two sub-categories, based on their relationship to a delta or not. 2) The larger river watersheds and the proximity of the slope break to the shore were shown to be responsible for the higher number of submarine canyons and channels on the Québec North Shore rather than on the Québec South Shore of the LSLE. 3) The genetic type of canyons and channels defines their modern sediment dynamics. While glacially-fed and delta related river-fed canyons and channels are currently inactive, longshore drift-fed canyons and sediment-starved canyons show evidence of episodic or frequent activity. 4) Unlike other marine deltas along the glaciated coastline of Canada, the deltas of the Québec North Shore cannot generate gravity flows today because sediment concentration from rivers is too low to produce hyperpycnal flows and accumulation rates are too low to produce frequent delta-slope failures. 5) Longshore drift-fed systems are episodically active today (N100 yrs recurrence?), probably due to the occurrence of large earthquakes

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

and exceptional storms advecting coastal sediments. This noncontinuous activity is attributed to limited sediment supply to their heads, unlike other longshore-drift fed systems in the world where sediment accumulation is important and frequent. Therefore, contrary to what was suggested in previous studies, the canyons in the LSLE do not appear to be contributing significantly to coastal erosion. 6) Sediment-starved submarine canyons are frequently active because slope failures and hydrodynamic processes do not rely solely on sediment supply. Therefore, these processes remobilize in-situ sediments within the canyons and transport them further downslope. Counter-intuitively, they are also the only canyons in the LSLE showing evidence of frequent activity, although sediment supply is absent, indicating that sediment supply is not necessarily an important contributor to gravity flow processes in estuaries and inner-shelf environments. 7) Steep slope has been an important factor for the occurrence of submarine canyon activity during the last 200 years. Canyons located on steep slopes all show evidence of relatively recent gravity flows, whether they are episodic or frequent; 8) In the absence of current-meter data, the high-resolution geomorphological approach used in this study proved to be an efficient method to assess modern submarine canyon and channel activity. The presence of bedforms, the analysis of surficial sediments, subbottom profiles and backscatter intensity values all allow us to define whether submarine canyons and channels are active or inactive. Acknowledgments We sincerely thank the captain, crew, and scientific participants of the COR1203 cruise on board the R/V Coriolis II. This study was supported by the Fonds québécois de la recherche sur la nature et les technologies and Geotop through a scholarship to A.N., by the Natural Sciences and Engineering Research Council of Canada through Discovery and Shiptime grants to P.L. G.S., and Jacques Locat (Université Laval), and by the Canadian Foundation for Innovation and the Ministère de l'éducation du Québec through equipment grants to P.L. We are grateful to Jacques Labrie and Gilles Desmeules for assistance in seismic data acquisition. Multibeam echosounder datasets were provided by the Canadian Hydrographic Service, the Ministère de la sécurité publique du Québec and the Centre interdisciplinaire de développement en cartographie des océans (CIDCO). We thank Pierre Francus, JeanFrançois Ghienne and Michel Allard as well as reviewers Pedro Cunha and Nathalie Babonneau and Editor Takashi Oguchi for providing constructive comments on this paper. References Armitage, D.A., Piper, D.J.W., McGee, D.T., Morris, W.R., 2010. Turbidite deposition on the glacially influenced, canyon-dominated Southwest Grand Banks Slope, Canada. Sedimentology 57, 1387–1408. Ashley, G.M., 1990. Classification of large-scale subaqueous bedforms: a new look at an old problem. J. Sediment. Petrol. 60, 160–172. Babonneau, N., Delacourt, C., Cancouët, R., Sisavath, E., Bachèlery, P., Mazuel, A., Jorry, S.J., Deschamps, A., Ammann, J., Villeneuve, N., 2013. Direct sediment transfer from land to deep-sea: insights into shallow multibeam bathymetry at La Réunion Island. Mar. Geol. 346, 47–57. Baztan, J., Berné, S., Olivet, J.L., Rabineau, M., Aslanian, D., Gaudin, M., Réhault, J.P., Canals, M., 2005. Axial incision: the key to understand submarine canyon evolution (in the western Gulf of Lion). Mar. Pet. Geol. 22, 805–826. Bernatchez, P., 2003. Évolution littorale holocène et actuelle des complexes deltaïques de Betsiamites et de Manicouagan-Outardes: Synthèse, processus, causes et perspectives. (Ph.D thesis), Dépatement de Géographie. Université Laval, Québec (531 pp.). Bernatchez, P., Dubois, J.-M.M., 2004. Bilan des connaissances de la dynamique de l'érosion des côtes du Québec maritime laurentien. Géog. Phys. Quatern. 58, 45–71. Biscara, L., Mulder, T., Martinez, P., Baudin, F., Etcheber, H., Jouanneau, J.-M., Garlan, T., 2011. Transport of terrestrial organic matter in the Ogooué deep sea turbidite system (Gabon). Mar. Pet. Geol. 28, 1061–1072. Biscara, L., Mulder, T., Hanquiez, V., Marieu, V., Crespin, J.-P., Braccini, E., Garlan, T., 2013. Morphological evolution of Cap Lopez Canyon (Gabon): illustration of lateral migration processes of a submarine canyon. Mar. Geol. 340, 49–56.

17

Blott, S.J., Pye, K., 2001. GRADISTAT: a grain size distribution and statistics package for the analysis of unconsolidated sediments. Earth Surf. Process. Landf. 26, 1237–1248. Bolduc, A., Duchesne, M.J., 2009. Découverte de mégadunes dans l'estuaire moyen du fleuve Saint-Laurent, Québec, Canada. Rev. Sci. Eau 22, 125–134. Bouma, A.H., 2004. Key controls on the characteristics of turbidite systems. In: Lomas, S.A., Joseph, P. (Eds.), Confined Turbidite Systems. Geological Society, London, pp. 9–22. Boyd, R., Ruming, K., Goodwin, I., Sandstrom, M., Schröder-Adams, C., 2008. Highstand transport of coastal sand to the deep ocean: a case study from Fraser Island, southeast Australia. Geology 36, 15–18. Budillon, F., Conforti, A., Tonielli, R., Falco, G., Martino, G., Innangi, S., Marsella, E., 2011. The Bulgheria canyon-fan: a small-scale proximal system in the eastern Tyrrhenian Sea (Italy). Mar. Geophys. Res. 32, 83–97. Cartigny, M.J.B., Postma, G., van den Berg, J.H., Mastbergen, D.R., 2011. A comparative study of sediment waves and cyclic steps based on geometries, internal structures and numerical modeling. Mar. Geol. 280, 40–56. Cauchon-Voyer, G., Locat, J., St-Onge, G., 2008. Late-Quaternary morpho-sedimentology and submarine mass movements of the Betsiamites area, Lower St. Lawrence Estuary, Quebec, Canada. Mar. Geol. 251, 233–252. Cauchon-Voyer, G., Locat, J., Leroueil, S., St-Onge, G., Demers, D., 2011. Large-scale subaerial and submarine Holocene and recent mass movements in the Betsiamites area, Quebec, Canada. Eng. Geol. 121, 28–45. Conway, K.W., Barrie, J.V., Picard, K., Bornhold, B.D., 2012. Submarine channel evolution: active channels in fjords, British Columbia, Canada. Geo-Mar. Lett. 32, 301–312. Covault, J.A., Normark, W.R., Romans, B.W., Graham, S.A., 2007. Highstand fans in the California borderland: the overlooked deep-water depositional systems. Geology 35, 783–786. Dietrich, P., Ghienne, J.F., Schuster, M., Deschamps, R., Lajeunesse, P., Nutz, A., Roquin, C., Duringer, P., 2014. Glacio-isostatically forced proglacial deltas: the examples of the Portneuf-Forestville delta, North Shore of the St-Lawrence estuary, Québec, Canada. 19th International Sedimentology Congress, Geneva, Switzerland. Dionne, J.-C., Occhietti, S., 1996. Aperçu du Quaternaire à l'embouchure du Saguenay, Québec. Géog. Phys. Quatern. 50, 5–34. Dowdeswell, J.A., Evans, J., O Cofaigh, C., Anderson, J.B., 2006. Morphology and sedimentary processes on the continental slope off Pine Island Bay, Amundsen Sea, West Antarctica. Geol. Soc. Am. Bull. 118, 606–619. Dubois, J.-M.M., Bernatchez, P., Bouchard, J.-D., Daigneault, B., Cayer, D., Dugas, S., 2005. Évaluation du risque d'érosion du littoral de la Côte-Nord du Saint-Laurent pour la période de 1996–2003. Conférence régionale des élus de la Côte-Nord (291 pp.). Ducassou, E., Migeon, S.B., Mulder, T., Murat, A., Capotondi, L., Bernasconi, S.M., Mascle, J., 2009. Evolution of the Nile deep-sea turbidite system during the Late Quaternary: influence of climate change on fan sedimentation. Sedimentology 56, 2061–2090. Duchesne, M.J., Long, B.F., Urgeles, R., Locat, J., 2003. New evidence of slope instability in the Outardes Bay delta area, Quebec, Canada. Geo-Mar. Lett. 22, 233–242. Duchesne, M.J., Pinet, N., Bédard, K., St-Onge, G., Lajeunesse, P., Campbell, D.C., Bolduc, A., 2010. Role of the bedrock topography in the Quaternary filling of a giant estuarine basin: the Lower St. Lawrence Estuary, Eastern Canada. Basin Res. 22, 933–951. Farre, J.A., McGregor, B.A., Ryan, W.B.F., Robb, J.M., 1983. Breaching the shelfbreak: passage from youthful to mature phase in submarine canyon evolution. In: Stanley, D.J., Moore, G.T. (Eds.), The shelfbreak: critical interface on continental margins. Society Of Economic Paleontologists And Mineralogists, Special Publication 33, pp. 25–39. Fortin, D., Francus, P., Gebhardt, A.C., Hahn, A., Kliem, P., Lisé-Pronovost, A., Roychowdhury, R., Labrie, J., St-Onge, G., 2013. PASADO science team, 2013. Destructive and non-destructive density determination: method comparison and evaluation from the Laguna Potrok Aike sedimentary record. Quat. Sci. Rev. 71, 147–153. Franconi, A., Sharma, K.N.H., Laurin, A., 1975. Région des rivières Betsiamites (Bersismis) et Moisie, Geological report. Ministère des Ressources naturelles du Québec (149 pp.). Gagné, H., Lajeunesse, P., St-Onge, G., Bolduc, A., 2009. Recent transfer of coastal sediments to the Laurentian Channel, Lower St. Lawrence Estuary (Eastern Canada), through submarine canyon and fan systems. Geo-Mar. Lett. 29, 191–200. García-García, A., Schoolmeester, T., Orange, D., Calafat, A., Fabres, J., Grossman, E., Field, M., Lorenson, T.D., Levey, M., Sansoucy, M., 2012. Recent sedimentary processes in the Cap de Creus canyon head and adjacent continental shelf, NE Spain: evidence from multibeam bathymetry, sub-bottom profiles and coring. In: Li, M.Z., Sherwood, C.R., Hill, P.R. (Eds.), Sediments, morphology and sedimentary processes on continental shelves. International Association Of Sedimentology Special Publication, pp. 71–97. Gaudin, M., Mulder, T., Cirac, P., Berné, S., Imbert, P., 2006. Past and present sedimentary activity in the Capbreton Canyon, southern Bay of Biscay. Geo-Mar. Lett. 26, 331–345. Gouin, P., 2001. Historical Earthquakes Felt In Quebec (From 1534 To March 1925). Guérin, Montréal (1491 pp.). Green, A.N., Goff, J.A., Uken, R., 2007. Geomorphological evidence for upslope canyonforming processes on the northern KwaZulu-Natal Shelf, SW Indian Ocean, South Africa. Geo-Mar. Lett. 27, 399–409. Harris, P.T., Vaughn Barrie, J., Conway, K.W., Greene, H.G., 2014. Hanging canyons of Haida Gwaii, British Columbia, Canada: fault control on submarine canyon geomorphology along active continental margins. Deep-Sea Res. II 104, 83–92. Hill, P.R., 2012. Changes in submarine channel morphology and slope sedimentation patterns from repeat multibeam surveys in the Fraser River delta, western Canada. In: Li, M.Z., Sherwood, C.R., Hill, P.R. (Eds.), International Association Of Sedimentology Special Publication, pp. 47–69. Hill, P.R., Conway, K., Lintern, D.G., Meule, S., Picard, K., Barrie, J.V., 2008. Sedimentary processes and sediment dispersal in the southern Strait of Georgia, BC, Canada. Mar. Environ. Res. 66, S39–S48. Hughes Clarke, J., Vidiera Marques, C.R., Pratomo, D., 2014. Imaging active mass-wasting and sediment flows on a fjord delta, Squamish, British Columbia. In: Krastel, S., Behrmann, J.-H., Völker, D., Stipp, M., Berndt, C., Urgeles, R., Chaytor, J., Huhn, K.,

18

A. Normandeau et al. / Geomorphology 241 (2015) 1–18

Strasser, M., Harbitz, C.B. (Eds.), Submarine Mass Movements And Their Consequences. Springer, pp. 249–260. Ingram, G.R., 1983. Vertical mixing at the head of the Laurentian Channel. Estuar. Coast. Shelf Sci. 16, 333–338. Inland Waters Directorate, 1978. Hydrographic Atlas Of Canada: Suspended Sediment Concentration. Department of Fisheries and the Environment, Government of Canada. Jeagle, M., 2014. Nature et origine des sédiments de surface de l'estuaire du Saint-Laurent. (M.Sc. dissertation), Institut des sciences de la mer de Rimouski (79 pp.). Khripounoff, A., Vangriesheim, A., Crassous, P., Etoubleau, J., 2009. High frequency of sediment gravity flow events in the Var submarine canyon (Mediterranean Sea). Mar. Geol. 263, 1–6. Lamontagne, M., 1987. Seismic activity and structural features in the Charlevoix region, Quebec. Can. J. Earth Sci. 24, 2118–2129. Lamontagne, M., 2009. Possible earthquake triggers of submarine landslides in the estuary of the St. Lawrence River: insights from the earthquake catalogue. Geological Survey of Canada. Open File 6007 (61 pp.). Lastras, G., Canals, M., Urgeles, R., Amblas, D., Ivanov, M., Droz, L., Dennielou, B., Fabrés, J., Schoolmeester, T., Akhmetzhanov, A., Orange, D., García-García, A., 2007. A walk down the Cap de Creus canyon, Northwestern Mediterranean Sea: Recent processes inferred from morphology and sediment bedforms. Mar. Geol. 246, 176–192. Lessard, G.L., Dubois, J.-M.M., 1984. Évolution littorale multitemporelle d'une côte récemment déglacée du nord du gole du Saint-Laurent. Rev. Géomorphol. Dynam. 33, 81–96. Locat, J., 2011. La localistion et la magnitude du séisme du 5 février 1663 (Charlevoix) revues à l'aide des mouvements de terrain. Can. Geotech. J. 48, 1266–1286. Locat, J., Levesque, C., 2009. Le fjord du Saguenay : une physiographie et un registre exceptionnels. Rev. Sci. Eau 22, 135–157. Locat, J., Sansfaçon, R., 2000. Multibeam surveys: a major tool for geosciences. Proceedings Of The Canadian Hydrographics Conference, Montréal (11 pp.). Locat, J., Martin, F., Lévesque, C., Locat, P., Leroueil, S., Konrad, J.-M., Urgeles, R., Canals, M., Duchesne, M.J., 2003. Submarine mass movements in the upper Saguenay Fjord, (Québec, Canada), triggered by the 1663 earthquake. In: Locat, J., Mienert, J. (Eds.), Submarine Mass Movements And Their Consequences. Kluwer Academic Publishers, pp. 509–519. Loring, D.H., Nota, D.J.G., 1973. Morphology and sediments of the Gulf of St. Lawrence. Bull. Fish. Res. Board Can. 182 (148 pp.). Mayall, M., Jones, E., Casey, M., 2006. Turbidite channel reservoirs—key elements in facies prediction and effective development. Mar. Pet. Geol. 23, 821–841. Mazières, A., Gillet, H., Castelle, B., Mulder, T., Guyot, C., Garlan, T., Mallet, C., 2014. Highresolution morphobathymetric analysis and evolution of Capbreton submarine canyon head (Southeast Bay of Biscay—French Atlantic Coast) over the last decade using descriptive and numerical modeling. Mar. Geol. 351, 1–12. Migeon, S., Mulder, T., Savoye, B., Sage, F., 2006. The Var turbidite system (Ligurian Sea, northwestern Mediterranean)—morphology, sediment supply, construction of turbidite levee and sediment waves: implications for hydrocarbon reservoirs. Geo-Mar. Lett. 26, 361–371. Migeon, S., Mulder, T., Savoye, B., Sage, F., 2012. Hydrodynamic processes, velocity structure and stratification in natural turbidity currents: Results inferred from field data in the Var turbidite system. Sediment. Geol. 245–246, 48–62. Mountjoy, J.J., Micallef, A., Stevens, C.L., Stirling, M.W., 2014. Holocene sedimentary activity in a non-terrestrially coupled submarine canyon: cook Strait Canyon system, New Zealand. Deep-Sea Res. II Top. Stud. Oceanogr. 104, 120–133. Mulder, T., Syvitski, J.P.M., 1995. Turbidity currents generated at river mouths during exceptional discharges to the world oceans. J. Geol. 103, 285–299. Mulder, T., Syvitski, J.P.M., Skene, K.I., 1998. Modeling of erosion and deposition by turbidity currents generated at river mouths. J. Sediment. Res. 68, 124–137. Mulder, T., Zaragosi, S., Garlan, T., Mavel, J., Cremer, M., Sottolichio, A., Sénéchal, N., Schmidt, S., 2012. Present deep-submarine canyons activity in the Bay of Biscay (NE Atlantic). Mar. Geol. 295–298, 113–127. Normandeau, A., Lajeunesse, P., St-Onge, G., 2013. Shallow-water longshore drift-fed submarine fan deposition (Moisie River Delta, Eastern Canada). Geo-Mar. Lett. 33, 391–403. Normandeau, A., Lajeunesse, P., St-Onge, G., Bourgault, D., St-Onge-Drouin, G., Senneville, S., Bélanger, S., 2014. Morphodynamics in sediment-starved inner-shelf submarine canyons (Lower St. Lawrence Estuary, Eastern Canada). Mar. Geol. 357, 243–255. Obelcz, J., Brothers, D., Chaytor, J., Brink, U.t., Ross, S.W., Brooke, S., 2014. Geomorphic characterization of four shelf-sourced submarine canyons along the U.S. MidAtlantic continental margin. Deep-Sea Res. II Top. Stud. Oceanogr. 104, 106–119. Ó Cofaigh, C., Dowdeswell, J.A., Evans, J., Kenyon, N.H., Taylor, J., Mienert, J., Wilken, M., 2004. Timing and significance of glacially influenced mass-wasting in the submarine channels of the Greenland Basin. Mar. Geol. 207, 39–54. Paull, C.K., Caress, D.W., Lundsten, E., Gwiazda, R., Anderson, K., McGann, M., Conrad, J., Edwards, B., Sumner, E.J., 2013. Anatomy of the La Jolla Submarine Canyon system; offshore southern California. Mar. Geol. 335, 16–34.

Pedrosa, M.T., Camerlenghi, A., De Mol, B., Urgeles, R., Rebesco, M., Lucchi, R.G., 2011. Seabed morphology and shallow sedimentary structure of the Storjorden and Kveithola trough-mouth fans (North West Barents Sea). Mar. Geol. 286, 65–81. Pinet, N., Brake, V., Campbell, D.C., Duchesne, M.J., 2011. Seafloor and shallow subsurface of the St. Lawrence River Estuary. Geosci. Can. 38, 31–40. Pinet, N., Brake, V., Campbell, C., Duchesne, M.J., 2015. Geomorphological characteristics and variability of Holocene mass-transport complexes, St. Lawrence River Estuary, Canada. Geomorphology 228, 286–302. Piper, D.J.W., Normark, W.R., 2009. Processes that initiate turbidity currents and their influence on turbidites: a marine geology perspective. J. Sediment. Res. 79, 347–362. Piper, D.J.W., Shaw, J., Skene, K.I., 2007. Stratigraphic and sedimentological evidence for late Wisconsinan sub-glacial outburst floods to Laurentian Fan. Palaeogeogr. Palaeoclimatol. Palaeoecol. 246, 101–119. Puig, P., Palanques, A., Martin, J., 2014. Contemporary sediment-transport processes in submarine canyons. Ann. Rev. Mar. Sci. 6, 1–5 (.25). Rise, L., Bøe, R., Riis, F., Bellec, V.K., Laberg, J.S., Eidvin, T., Elvenes, S., Thorsnes, T., 2013. The Lofoten–Vesterålen continental margin, North Norway: canyons and massmovement activity. Mar. Pet. Geol. 45, 134–149. Roger, J., Saint-Ange, F., Lajeunesse, P., Duchesne, M.J., St-Onge, G., 2013. Late Quaternary glacial history and meltwater discharges along the Northeastern Newfoundland Shelf. Can. J. Earth Sci. 50, 1178–1194. Romans, B.W., Normark, W.R., McGann, M.M., Covault, J.A., Graham, S.A., 2009. Coarsegrained sediment delivery and distribution in the Holocene Santa Monica Basin, California: implications for evaluating source-to-sink flux at millennial time scales. Geol. Soc. Am. Bull. 121, 1394–1408. Sala, M., Long, B.F., 1989. Évolution des structures deltaïques du delta de la rivière Natashquan, Québec. Géog. Phys. Quatern. 43, 311–323. Shaw, J., Todd, B.J., Li, M.Z., 2014. Geologic insights from multibeam bathymetry and seascape maps of the Bay of Fundy, Canada. Cont. Shelf Res. 83, 53–63. Smith, W., 1962. Earthquakes Of Eastern Canada And Adjacent Areas, 1534–1927. Publications of the Dominion Observatory, Ottawa 271–301. Smith, D.P., Ruiz, G., Kvitek, R., Iampietro, P.J., 2005. Semiannual patterns of erosion and deposition in upper Monterey Canyon from serial multibeam bathymetry. Geol. Soc. Am. Bull. 117, 1123–1133. Smith, D.P., Kvitek, R., Iampietro, P.J., Wong, K., 2007. Twenty-nine months of geomorphic change in upper Monterey Canyon (2002–2005). Mar. Geol. 236, 79–94. St-Onge, G., Long, B.F., 2009. CAT-scan analysis of sedimentary sequences: an ultrahighresolution paleoclimatic tool. Eng. Geol. 103, 127–133. St-Onge, G., Mulder, T., Piper, D.J.W., Hillaire-Marcel, C., Stoner, J.S., 2004. Earthquake and flood-induced turbidites in the Saguenay Fjord (Québec): a Holocene paleoseismicity record. Quat. Sci. Rev. 23, 283–294. St-Onge, G., Lajeunesse, P., Duchesne, M., Gagné, H., 2008. Identification and dating of a key Late Pleistocene stratigraphic unit in the St. Lawrence Estuary and Gulf (Eastern Canada). Quat. Sci. Rev. 27, 2390–2400. St-Onge, G., Duchesne, M.J., Lajeunesse, P., 2011. Marine geology of the St. Lawrence estuary. IOP Conference Series: Earth And Environmental Science 14. http://dx.doi. org/10.1088/1755-1315/14/1/012003. St-Onge, G., Chapron, E., Muslow, S., Salas, M., Viel, M., Debret, M., Foucher, A., Mulder, T., Winiarski, T., Desmet, M., Costa, P.J.M., Ghaleb, B., Jaouen, A., Locat, J., 2012. Comparison of earthquake-triggered turbidites from the Saguenay (Eastern Canada) and Reloncavi (Chilean margin) Fjords: implications for paleoseismicity and sedimentology. Sediment. Geol. 243–244, 89–107. Stow, D.A.V., Howell, D.G., Nelson, C.H., 1983. Sedimentary, tectonic, and sea-level controls on submarine fan and slope-apron turbidite systems. Geo-Mar. Lett. 3, 57–64. Talling, P.J., 2014. On the triggers, resulting flow types and frequencies of subaqueous sediment density flows in different settings. Mar. Geol. 352, 155–182. Turner, R.J.W., Clague, J.J., Hastings, N.L., 2003. Geoscape Canada: A map of Canada's earth materials. Geological Survey Of Canada, Miscellaneous Report 81. Twitchell, D.C., Roberts, D.G., 1982. Morphology, distribution and development of submarine canyons on the United States Atlantic continental slope between Hudson and Baltimore Canyons. Geology 10, 408–412. Vail, P.R., Mitchum, R.M., Thompson III, S., 1977. Seismic stratigraphy and global changes of sea level, part 3: relative changes of sea level from coastal onlap. In: Payton, C.E. (Ed.), Seismic Stratigraphy — Applications To Hydrocarbon Exploration 26. American Association of Petroleum Geologists Memoir, pp. 63–81. Warrick, J.A., Simms, A.R., Ritchie, A., Steel, E., Dartnell, P., Conrad, J.E., Finlayson, D.P., 2013. Hyperpycnal plume-derived fans in the Santa Barbara Channel, California. Geophys. Res. Lett. 40, 2081–2086. Xu, J.P., Swarzenski, P.W., Noble, M., Li, A.-C., 2010. Event-driven sediment flux in Hueneme and Mugu submarine canyons, southern California. Mar. Geol. 269, 74–88. Yoshikawa, S., Nemoto, K., 2010. Seasonal variations of sediment transport to a canyon and coastal erosion along the Shimizu coast, Suruga Bay, Japan. Mar. Geol. 271, 165–176.