Continental Shelf Research 27 (2007) 1779–1800 Long-term development and current status of the Barcelona continental shelf: A source-to-sink approach C. Liquete a , M. Canals a, , G. Lastras a , D. Amblas a , R. Urgeles a , B. De Mol a , M. De Batist b , J.E. Hughes-Clarke c a GRC Geocie`ncies Marines, Dept. d’Estratigrafia, Paleontologia i Geocie`ncies Marines, Facultat de Geologia, Universitat de Barcelona, E-08028 Barcelona, Spain b Renard Centre of Marine Geology, Universiteit Gent, Krijgslaan 281 S8, B-9000 Gent, Belgium c Ocean Mapping Group, Department of Geodesy and Geomatics Engineering, University of New Brunswick, Fredericton, New Brunswick, Canada E3B 5A3 Received 18 July 2006; received in revised form 16 January 2007; accepted 28 February 2007 Available online 19 March 2007 Abstract The Barcelona continental shelf, off the city of Barcelona (NE Spain), is a relatively narrow canyon-bounded shelf in the northwestern Mediterranean Sea. Three medium-size rivers (Tordera, Beso´ s and Llobregat) and several ephemeral rivulets flow into this margin. Two main domains have been recognized in the Barcelona shelf: (i) a modern, river-influenced area, and (ii) a relict, sediment depleted area, both affected by a variety of human impacts. A detailed geomorphologic study based on multibeam bathymetry and backscatter data, high resolution seismic profiles, and surface sediment samples allowed mapping and interpreting the main distinctive seafloor features on the Barcelona shelf. Modern sedimentary features reveal that the Llobregat River is the main sediment source of the Barcelona prodeltaic shelf. High-discharge fluvial events result in the formation of suspended sediment plumes and sediment waves on the shelf floor. Relict (late Pleistocene–Holocene) sedimentary features reflect that an important shift occurred in the seashore direction between MIS 4 and MIS 2, and that recent neotectonic reactivation has created a set of seafloor faults. The Barcelona inner and middle shelf is severely impacted by anthropogenic activities such as the enlargement works of the Port of Barcelona, sewage pipes, dredging, anchoring and trawling. r 2007 Elsevier Ltd. All rights reserved. Keywords: Shelf sedimentary environments; Source-to-sink; Prodelta; Multibeam; River basins; Seismic reflection; Neotectonics; Sediment waves; Late Quaternary sea level change; Human impacts; Northwestern Mediterranean Sea; Barcelona continental shelf; Spain 1. Introduction Most of the processes that shape land–ocean systems leave their imprint in the sediment record of the continental margins. Source-to-sink studies analyze the complex interplay of processes within entire sediment-dispersal systems at different tem- poral and spatial scales in order to understand the evolution of continental margins and its associated morphology (Driscoll and Nittrouer, 1999). The final aim of a source-to-sink analysis is to develop a quantitative approach to a given margin dispersal ARTICLE IN PRESS www.elsevier.com/locate/csr 0278-4343/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.csr.2007.02.007 Corresponding author. Tel.: +34 93 402 13 60; fax: +34 93 402 13 40. E-mail address: [email protected] (M. Canals).
Transcript
ARTICLE IN PRESS
0278-4343/$ - see
doi:10.1016/j.csr
�Correspondifax: +3493 402
E-mail addre
Continental Shelf Research 27 (2007) 1779–1800
www.elsevier.com/locate/csr
Long-term development and current status of the Barcelonacontinental shelf: A source-to-sink approach
C. Liquetea, M. Canalsa,�, G. Lastrasa, D. Amblasa, R. Urgelesa, B. De Mola,M. De Batistb, J.E. Hughes-Clarkec
aGRC Geociencies Marines, Dept. d’Estratigrafia, Paleontologia i Geociencies Marines, Facultat de Geologia, Universitat de Barcelona,
E-08028 Barcelona, SpainbRenard Centre of Marine Geology, Universiteit Gent, Krijgslaan 281 S8, B-9000 Gent, Belgium
cOcean Mapping Group, Department of Geodesy and Geomatics Engineering, University of New Brunswick, Fredericton, New Brunswick,
Canada E3B 5A3
Received 18 July 2006; received in revised form 16 January 2007; accepted 28 February 2007
Available online 19 March 2007
Abstract
The Barcelona continental shelf, off the city of Barcelona (NE Spain), is a relatively narrow canyon-bounded shelf in the
northwestern Mediterranean Sea. Three medium-size rivers (Tordera, Besos and Llobregat) and several ephemeral rivulets
flow into this margin. Two main domains have been recognized in the Barcelona shelf: (i) a modern, river-influenced area,
and (ii) a relict, sediment depleted area, both affected by a variety of human impacts. A detailed geomorphologic study
based on multibeam bathymetry and backscatter data, high resolution seismic profiles, and surface sediment samples
allowed mapping and interpreting the main distinctive seafloor features on the Barcelona shelf. Modern sedimentary
features reveal that the Llobregat River is the main sediment source of the Barcelona prodeltaic shelf. High-discharge
fluvial events result in the formation of suspended sediment plumes and sediment waves on the shelf floor. Relict (late
Pleistocene–Holocene) sedimentary features reflect that an important shift occurred in the seashore direction between MIS
4 and MIS 2, and that recent neotectonic reactivation has created a set of seafloor faults. The Barcelona inner and middle
shelf is severely impacted by anthropogenic activities such as the enlargement works of the Port of Barcelona, sewage
the continental margins. Source-to-sink studiesanalyze the complex interplay of processes withinentire sediment-dispersal systems at different tem-poral and spatial scales in order to understand theevolution of continental margins and its associatedmorphology (Driscoll and Nittrouer, 1999). Thefinal aim of a source-to-sink analysis is to develop aquantitative approach to a given margin dispersal
ARTICLE IN PRESSC. Liquete et al. / Continental Shelf Research 27 (2007) 1779–18001780
system, so that its response to natural or humanperturbations can be predictable.
The first requirement of a source-to-sink study isto compile a high-quality landscape and seascapedata set so that sediment erosion, transport andaccumulation can be analyzed as a whole. Highresolution swath bathymetry represented a funda-mental breakthrough for source-to-sink studies. Thefirst swath bathymetry map from a continentalmargin in the western Mediterranean Sea was theeastern Gulf of Lions multibeam map published byOrsolini et al. (1981–1982). During the followingyears, researchers from France, Italy and Spainadded new contributions (e.g. Berne et al., 2004;Canals et al., 2004b, Marani et al., 2004) to ageneral swath mapping effort that led to the recentmultibeam bathymetry compilation of the entirewestern Mediterranean Sea (Medimap Group,2005). However, continental shelves were mostlyleft out of such a mapping effort mainly because oftwo reasons: (i) the time consuming character ofswath mapping in shallow areas due to muchsmaller swath widths and (ii) the broadness of someof the continental shelves in the region, such as theGulf of Lions (�70 km) and the Ebro (�50 km)shelves. Still today only relatively small areas of thecontinental shelves in the western Mediterraneanhave been swath mapped and the results published.In this paper we present the results achieved so faron the Barcelona continental shelf where a compre-hensive data set made of multibeam data, highresolution seismic reflection profiles, sediment sam-ples and geographic and hydrological data is nowavailable. First, the sediment sources and the maindistinctive geomorphological features are presented,and then the origin of these features and the modernseafloor zonation is discussed, regarding aspectssuch as fluvial sedimentation, neotectonics, sea levelchanges, and human impacts.
This paper aims at providing an accuratemorphological description of the Barcelona con-tinental shelf in the northwestern MediterraneanSea following a source-to-sink perspective that linksfluvial inputs and physical oceanography processesto the development of the shelf morphosedimentaryfeatures. Mostly qualitative sedimentary analyseshave been developed at this stage due to the lack ofhomogeneous quantitative data onshore and off-shore. The time fork considered in our study is thelate Pleistocene–Holocene, when global sea leveloscillations had a profound impact on continentalshelf development. Therefore, our analysis deals
with modern and relict sedimentary features, andwith anthropogenic impacts.
2. General setting
The Catalan margin extends along 300 km fromthe Ebro Delta to the Cap of Creus peninsula(Fig. 1). It forms the western border of the northernValencia Trough, a 400� 300 km mid-Mioceneextensional basin in the western MediterraneanSea. The Catalan margin is underlain by a SW–NEoriented horst-and-graben structure (Maillard et al.,1992; Roca et al., 1999). Following the Mioceneextension, the Messinian closure of the straitbetween north Africa and the south of Iberia ledto an event that had a dramatic influence on theshaping of Mediterranean margins: the salinitycrisis and drying up of the basin between 5.96 and5.33 million years ago (Duggen et al., 2003). Duringthe initial stage of the Messinian event, the large fallin water level caused extreme erosion and incisionthat led to the development of large canyon systems.The physiography and the seascape of the modernCatalan margin are, therefore, strongly determinedby the Messinian crisis (Amblas et al., 2004). Thedevelopment of the modern terrigenous shelf andmargin started shortly after the Pliocene transgres-sion, being ultimately reshaped by Quaternaryglacioeustatic sea level changes (Bartrina et al.,1992).
The largest submarine canyons in the Catalanmargin, which clearly have Messinian ancestors,segment the margin in different sections. Thecontinental shelf is narrowest (3 km) where deeplyincised by submarine canyon heads, and reaches itsmaximum width off the Ebro Delta (50 km). TheBarcelona margin and shelf, off the city ofBarcelona, occupy the central stretch of the sectionbounded by the Foix Canyon to the south and theBlanes Canyon to the north, from 41180N to411320N (Fig. 1). The Barcelona continental shelfis 6–20 km wide with the shelf break at 110–120mdepth. Its average slope is 0.61 although it reachesup to 211 in the slope break off the Llobregat Rivermouth and at the edge of some prominent sandydeposits observed to the north of the Besos River.The Barcelona shelf consists of an inner, middle andouter shelf separated by the 30–40 and the 80misobaths, respectively (ITGE, 1989).
The Barcelona continental shelf is mainly fed bysediment inputs from the Llobregat and Besos rivers(Fig. 1), mostly draining Neogene-Quaternary and
ARTICLE IN PRESS
1°30'E 2°0'E 2°30'E
41°0
'N41°3
0'N
42°0
'N
400000 450000
4550000
4600000
4650000
Pyrenees
Montseny
Catalan Coastal Ranges
Barcelona
Masnou
2504m
1304m
1684m
642m
764m
646m
Bla
nes
C.
Are
nys C
.
Besó
s C
Fo
ix C
.
Berenguera C
.M
orrá
s C.
-50
-100-200
-1000
1
2
3
4
5
6
1 Tordera River2 Sant Cebrià Rivulet3 Argentona Rivulet4 Besós River5 Llobregat River6 Sant Perede Ribes Rivulet
8°W 4°W 0° 4°E 8°E
38
°N
42
°N
46
°N
VTSPAIN
FRANCE
ALGERIA
A
B
CD
E
F
Dams(letters after Table 1)
Barcelona continental
shelf
Cities mentioned inthe text
Northern Current
Littoral drift
Faults
Highest peak in eachriver basin
Canyon systems
Shelf break
CC
ED
Profiles in
Fig.4
else|csr|1585
Fig. 1. Physiographic and morphological setting of the study area. Blue lines and numbers illustrate the river courses, and bold white lines
mark the watersheds limits. The yellow area illustrates the Barcelona continental shelf. VT: Valencia Trough. CC: Cap of Creus. ED: Ebro
Delta. Faults after Bartrina et al. (1992). Map projection is UTM Zone 31N Datum WGS 1984. Both UTM and geographic coordinates
are shown for ease of reference.
C. Liquete et al. / Continental Shelf Research 27 (2007) 1779–1800 1781
ARTICLE IN PRESSC. Liquete et al. / Continental Shelf Research 27 (2007) 1779–18001782
Paleogene calcareous rocks, respectively, and theTordera River eroding a predominantly graniticmassif (SGC, 1992). The Llobregat river basin is themost regulated in the study area (Table 1). Otherephemeral rivulets or ‘‘rieras’’ draining graniticrocks and a soft Plio-Quaternary cover punctuallyadd to sediment inputs, essentially during violenteastern storms with heavy rains that mostly occur inlate summer and early fall, and are locally known as‘‘llevantades’’ (Martın Vide, 1982). The energeticwave regime associated to ‘‘llevantades’’ reinforcesthe southwards littoral drift, which becomes thenable to transport large volumes of sand that may belost from the coastal sediment cell either escaping todeeper sections on the shelf or getting trapped bysubmarine canyon heads.
The dominant oceanographic current in theGulf of Lions and the Catalan margin is theNorthern Current, also referred to as Liguro-Provenc-al Current, which flows south-westwardsover the continental slope reaching speeds of50 cm s�1 (Flexas et al., 2002) and occasionallyintruding and meandering over the shelf, especiallybetween December and May (Flexas et al., 2002;Arnau et al., 2004). The predominant currentacting near the coast is the south-westwards littoraldrift. Ports and man made coastal protectionworks severely interfere the drift current leadingto anthropogenically generated sheltered deposi-tional areas and erosional coastal stretches(DGPT, 2000). The losses of littoral sedimentsare evidenced by the increasing sand volumesrequired for beach nourishment. For instance,150 000m3 yr�1 were required for the 4 km long‘‘Olympic Zone’’ beaches in Barcelona during 1999and 2000, while the estimated littoral transportfor the same period was of some tens of thousandsof m3 yr�1 (DGPT, 2000). The fair-weatherwave-face has been established at 20m depth in
Table 1
Main characteristics of the dams built along the Tordera, Besos and L
River system Letter in Fig. 1 Reservoir Year of construc
Tordera A Santa Fe 1933
Besos B Vallforners 1989
Llobregat C Sant Ponc- 1957
D La Baells 1976
E Sant Martı de Tous 1997
F La Llosa del Cavall 1999
Letters allow identifying the location of each dam on Fig. 1. Dams lab
the study area (Calafat, 1986), while the majorstorm wave-face is about 30m depth (Sorribas et al.,1993).
3. Methodology and data set
Swath bathymetry and associated backscatter datawere acquired onboard the R/V Hesperides and the12m fishing inspection boat Arraix during theMARINADA (2002) and PRODELTA-1 (2004)surveys using multibeam echosounders Simrad EM-1002S and EM-3000D, respectively (Fig. 2). TheEM-1002S system operates at a frequency of 95 kHzwith a maximum ping rate of 10Hz and uses 111beams. The swath width is up to 7.4 times the waterdepth when working shallower than 200m depth,and 1500m wide on deeper areas. The EM-3000Dsystem operates at a frequency of 300 kHz with amaximum ping rate of 40Hz and uses 254 beams.The swath width is up to 10 times the water depth or200m from 1 to 150m water depth. Raw data werelogged using Simrad’s system and later processedusing the SwathEd software from the OceanMapping Group of the University of New Bruns-wick. This processing included editing of naviga-tion, both automatic filtering and manual beamediting, refraction correction and backscatter equal-ization. The total swath mapped area in this workcovers about 625 km2 (Fig. 2).
In total, 337 km of high-resolution seismic reflec-
tion profiles were acquired in July 2004 jointly withthe Renard Centre of Marine Geology from theUniversity of Ghent (Belgium) on board the fishinginspection boat Arraix (Fig. 2). The seismic sourcewas a SIG 300 J Sparker at a sampling frequency of6 kHz. About 200–2000Hz band-pass filtering andrelative position corrections preceded digital regis-tration onboard. Post-processing included severalband-pass filters and centered automatic gain
lobregat river basins
tion Maximum capacity (hm3) Dam height (m) Surface (ha)
1 23 6
2 62 16
24 60 139
115 102 367
1 32 15
80 122 300
eled B, D, E and F are also referred in Fig. 4.
ARTICLE IN PRESS
2°00'E 2°10'E 2°20'E 2°30'E
41°1
0'N
41°2
0'N
41°3
0'N
410000 420000 430000 440000 450000 460000
4560000
4570000
458
0000
4590000
-200
LlobregatRiver
BesósRiver
Shelfedge
LL2
LL5
LL6
LL9LL8
LL7LL4
LLB
LLD
LLA
LLC
LL1 0
LL11
llosp45
llosp38
llosp40
Swath mapped area during PRODELTA-1 survey
Seismic reflection profiles
Seismic lines in Figure 6 (llosp 38, 40 and 45)
Shelf edge
Sediment sample
Swath mapped area during MARINADA survey
Fig. 2. Marine data set used in this study, including multibeam coverage, location of sparker seismic reflection profiles, and position of
surface sediment samples. Map projection is UTM 31N WGS84.
C. Liquete et al. / Continental Shelf Research 27 (2007) 1779–1800 1783
control. The Sparker data have an approximateresolution of 10–15m horizontally and 0.4–1.0mvertically. A triple-line artifact, likely linked tooperational conditions, was occasionally observedoverlying the strongest reflectors, namely the sea-floor reflector. This reverberation is up to 5ms thick(two-way travel time, TWTT), thus yielding anuncertainty of up to four vertical meters.
Sediment samples were collected by means of a45 cm long multicoring system in September 2004on board the R/V Garcıa del Cid totaling 13sampling sites (Fig. 2). Subsamples were subse-quently described and analyzed for grain size,composition and organic matter content. Only thetopmost centimeter of each sampling site has beentaken into account for this study. Preparation forgrain size analysis included two hydrogen peroxideattacks for organic matter elimination, grain dis-
assembling with a dispersant agent and sieving ofthe 41mm fraction. Grain size measurements ofthe o1mm fraction were made with a LS100Coulter Counter. Compositional visual analyseswere carried out with binocular lens. Organic mattercontent was determined by measuring the weightloss after burning sediment samples at 560 1C in amuffle furnace.
Geographic and hydrological data from relevantriver systems have been gathered from the Spanish‘‘Instituto Geografico Nacional’’ digital topographyseries, ‘‘Centro de Estudios y Experimentacionde Obras Publicas’’ (CEDEX) and the compila-tion on Spanish Mediterranean watersheds byCanals et al. (2004a). This data set was inclusiveof watershed divides, morphometry, location ofgauging stations and liquid discharge time series.Fig. 3 shows the longitudinal profiles of the
ARTICLE IN PRESS
25 km
2000
-1000
0
1000
2000
-1000
0
1000
2000
-1000
0
1000
25 km
25 km
Ele
vation (
m)
Offshore
Tordera
Besós
Llobregat
Coastline
5.0º
4.6º
5.7º
Blanes Canyonhead
Fig. 3. Topographic profiles along the three main river systems of the study area, from headwaters to the continental margin segments
they open to. The slope of the uppermost river courses is indicated. See location on Fig. 1.
Table 2
Main morphologic and hydrological characteristics of the fluvial systems shown on Fig. 1
6 Sant Pere de Ribes 233.3 22.2 1.0 — 11.2 — — — —
Mean water discharge and runoff values are calculated after 33 years long water discharge time series for the Llobregat and Besos rivers
and 9 years long for the Tordera River. The small rivulets are non-monitored. Mean sediment flux and solid discharge values are derived
from punctual suspended sediment measurements along 7 years. Gauging stations are located less than 15 km upstream from the river
mouth. RM–SB: from river mouth to shelf break.
C. Liquete et al. / Continental Shelf Research 27 (2007) 1779–18001784
Llobregat (163 km long), Besos (52 km) and Tor-dera (59 km) rivers, including the respective adja-cent continental margin. The selected hydrologicaldata series are 33 years long for the Llobregat andBesos rivers and 9 years long for the Tordera River,including temporary gaps of variable duration. Thesmall rivulets in Fig. 1 are non-monitored. Totalsuspended sediment values of the Tordera, Besosand Llobregat river systems from the period1995–2002 have been gathered from ‘‘AgenciaCatalana de l’Aigua’’. Sediment flux (in g s�1) was
estimated by combination of total suspended sedi-ment measurements (in mg l�1) and the correspond-ing daily water discharge values (in m3 s�1) (Table 2).Water and solid discharge data were taken from thenearest gauging station to each river mouth that isless than 15 km far from the seashore. Informationrelated to damming history (Table 1) was extractedfrom the ‘‘Agencia Catalana de l’Aigua’’ database(ACA, 2000) and the Spanish ‘‘Ministerio de MedioAmbiente’’ data collection (MOPU, 1988; MOPT-MA, 1994).
ARTICLE IN PRESSC. Liquete et al. / Continental Shelf Research 27 (2007) 1779–1800 1785
4. Results
4.1. Sediment sources
The hydrological data show that the largestfluvial input both in terms of liquid and soliddischarge is the one from the Llobregat River,although the top runoff value corresponds to theTordera River (Fig. 4 and Table 2). The Torderaand Besos rivers show similar mean water dischargevalues, but this is largely because of the July 1999extraordinary flood event in the Tordera Basin.Under normal conditions, the Besos system hasconsiderably larger discharge values than theTordera. Water discharge regimes of the Tordera,Besos and Llobregat river systems are relativelyirregular, particularly in the Tordera River, thoughthe overall pattern is similar (Fig. 4A). Maximumwater discharge usually occurs between Novemberand January for the Besos and Tordera rivers, andin spring for the Llobregat River following snowmelting in the Pyrenees catchment area (Fig. 4B).
10
-67
10
-68
10
-69
10
-70
10
-71
10
-72
10
-73
10
-74
10
-75
10
-76
10
-77
10
-78
10
-79
10
-80
10
-81
10
-82
10
-83
120
100
80
60
40
20
0
Dis
ch
arg
e (
m3s
-1)
Dis
ch
arg
e (
m3s
-1)
Mon
25
20
15
10
5
0
M
OCT NOV DEC JAN FEB MAR
DTordera
Besós
Llobregat
Fig. 4. (A) Monthly water discharge time series from the Tordera (dark
The Tordera maximum discharge peak (out of scale) occurred in July 19
information on Table 1). (B) Mean monthly discharge histograms for
The largest average sediment flux to the sea issupplied by the Llobregat River, and the least by theTordera system. Cumulative sediment deposition atthe river mouth as reflected by delta areas is farmore significant in the Llobregat system (Table 2).The relatively high slope of the Tordera River andnearby ‘‘rieras’’ (rivulets), especially in the catch-ment areas (Fig. 3 and Table 2), and their strongseasonal regime favor large short-lived dischargeevents that are barely reflected by statisticalaverages or simply remain unrecorded. This maylead to underestimate the significance of thesefluvial systems as sediment vectors to the Barcelonacontinental shelf.
Damming in the Llobregat, Besos and Torderariver basins is essentially limited to upper courses(Fig. 1), with the main dams (La Baells, La Llosadel Cavall and Sant Ponc-) located in the Llobregatwatershed. However, the influence of these dams onthe Llobregat River final water discharge is notevident (Fig. 4A). Water extraction for agriculture,urban and industrial use is known to be especially
99. Letters point to dam constructions (see location on Fig. 1 and
the Tordera, Besos and Llobregat rivers.
ARTICLE IN PRESSC. Liquete et al. / Continental Shelf Research 27 (2007) 1779–18001786
high along the Llobregat low course, where severalpumping stations detract more than 60 hm3 of freshwater per year (ACA, 2005). In addition, theTordera, Besos and Llobregat rivers undergoanthropogenic water extractions of less than5 hm3 yr�1 in numerous sites along their uppercourses (ACA, 2005). It is to be noted that, relatedto the enlargement of the Port of Barcelona, thelowermost course of the Llobregat River wasdiverted in September 2004 shifting the river mouth2.5 km southwards.
4.2. Distinctive seafloor features
The detailed examination of the Barcelona con-tinental shelf led to the identification of its most
Shelf break
2°00'E 2°10'E
41°1
0'N
41°2
0'N
41°3
0'N
410000 420000 430000
5
2
6
5
Barcelona
C
Fig. 5. Swath bathymetry shaded relief map of the Barcelona continent
high density population of the area. White lines correspond to river cou
on Fig. 10. Numbers mark some examples of the distinctive seafloor fe
relevant distinctive seafloor features. This identifica-tion is based on the high resolution bathymetric andbackscatter data, the seismic reflection profiles, andthe sedimentological analyses. The location of someof the following seafloor features is shown in Fig. 5.
4.2.1. Abrupt slopes
1–5m high abrupt slopes are observed in thenorthern part of the Barcelona shelf. They appear asstraight steps on the seafloor bathymetry and asvertical slopes on the seismic data set (Fig. 6A andB). Their orientation is rather constant, rangingfrom N050-N075 with descending angles offshoreand southwards, except by one feature nearly N090oriented. Apart from the 5m high feature locatedoff the Besos River at around 55m depth, the rest of
Prodelta limit
2°20'E 2°30'E
440000 450000 460000
4560000
4570000
4580000
4590000
1 2
3
2
3
3
3
32
1
3
2
21
1 14
3
Masnou
DD
BA
Examples of distinct sea floor features:
1 - Abrupt slopes
2 - Narrow ridges
3 - Prograding sediment bodies
4 - Large submarine step
5 - Low back scatter areas (see Fig.7)
6 - Wavy prodelta front
al shelf at 10m resolution. The onland orthophotomap shows the
rses. Red boxes and capital letters point to the location of images
atures discussed in the text.
ARTICLE IN PRESS
Sharp ridge
Abrupt slopes
Line llosp40
1 km 2 km
50
70
130
90
110
Line llosp45
Prograding sediment body
1 km 2 km70
90
150
110
130
1 km 2 km
TW
TT
(m
s)
TW
TT
(m
s)
TW
TT
(m
s)
0
20
80
40
60
Line llosp38
NNW-SSE
NNW-SSE
NNW-SSE
Large submarine step
MultipleAbrupt slopes
10 m
Fig. 6. Sparker seismic reflection profiles illustrating the internal structure of several seafloor features discussed in the main text. Vertical
axes show the Two-Way Travel Time in milliseconds, corresponding approximately to the metric bar of the bottom right corner of line
llosp45. Vertical exaggeration is � 17. See location on Fig. 2.
C. Liquete et al. / Continental Shelf Research 27 (2007) 1779–1800 1787
the slopes are 1–3m high. The length of thesefeatures ranges from 150 to 8.5 km.
4.2.2. Narrow ridges
Discontinuous narrow ridges appear to the northof the study area. They are sharp concave-upstructures from tens of centimeters up to 2mhigh observed in the bathymetric and seismic data(Fig. 6B). The ridges are W–E to NNE–SSWoriented and their length ranges from 150m to10 km. Sample LL-B was collected from one of
these features, while samples LL-A and LL-Dwere taken in close proximity (less than 100m fromone of the ridges) (Fig. 2). The multicore systemcould not penetrate more than 5 cm in any of thethree sites. The grain size and composition analysesof the three samples showed relatively coarsematerial with a high percentage of bioclasts41mm diameter (13–41%) (Table 3). The meangrain size of sample LL-B was 0.34mm, while thoseof LL-A and LL-D were 0.46 and 0.64mm,respectively.
ARTICLE IN PRESS
Table 3
Water depth and grain size composition of the surface sediment
samples located on Fig. 2
Sample Depth (m) Clay % Silt % Sand % 41mm %
LL-A 72 9.4 20.6 43.3 26.7
LL-B 71 13.4 30.7 42.5 13.4
LL-C 132 6.7 23.4 69.1 0.9
LL-D 69 2.5 6.2 50.6 40.7
LL-2 69 12.4 53.7 29.4 4.5
LL-4 149 7.4 23.3 68.9 0.4
LL-5 96 18.0 62.9 14.9 4.2
LL-6 57 12.6 52.0 34.9 0.7
LL-7 30 13.7 51.3 35.0 0.0
LL-8 66 22.7 73.6 3.7 0.0
LL-9 85 35.9 63.5 0.0 0.5
LL-10 71 18.7 71.7 4.8 4.9
LL-11 92 6.1 19.6 68.2 6.1
Measurements of the o1mm fraction were made with a LS100
Coulter Counter. Grain size is expressed as weight percentage.
C. Liquete et al. / Continental Shelf Research 27 (2007) 1779–18001788
4.2.3. Prograding sediment bodies
The Barcelona shelf comprises several outcrop-ping and subcropping prograding sediment bodieslocated around 45, 50, 65, 75 and 105m depth. Theyare E–W to ENE–SWS oriented and have a clearseafloor expression. They show relatively highbackscatter values, most probably indicative ofrelatively coarse material. These sediment bodiescan reach 5 km in length and 4m in height, and theirseafloor slope ranges between 0.61 and 4.01.Internally, they show high-angle cross stratificationwith well-defined body base and top (Fig. 6C).
Another expression of these sedimentary featuresis a large, ENE–WSW oriented, thick sedimentbody limited towards deeper water by a steep slope,here called submarine step. It appears on thenorthern inner shelf near Masnou and extendsalong 11 km deepening northwards. Its height isup to 10m and its slope up to 211. At 41129.50N,around 20m depth, this large single step splits intothree smaller steps (Fig. 5). The deepest of thesesmaller steps is known to continue eastwardsbeyond the study area forming a 20 km long, 30mhigh, W–E oriented submarine step. It extends fromthe northern end of the study area to the BlanesCanyon, crossing the middle and outer shelf downto 110m depth (ITGE, 1989; Dıaz and Maldonado,1990; Liquete et al., 2004b). Sandy seafloor has beenreported from this large ENE–WSW sediment body(ITGE, 1989; Llamas, 2003), which is in agreementwith the observed high and homogeneous back-
scatter values (Fig. 7). The internal structure of thesediment prism forming the main step consists of7–121 tilted subparallel strata that get steeperseawards. This sediment body is bounded by anupper erosional truncation and a set of downlapingbasal reflectors (Fig. 6A).
4.2.4. Featureless seafloor with low backscatter
Two roughly elliptical patches showing the lowestbackscatter values are observed S–SW of the Besosand Llobregat river mouths (Figs. 5 and 7). Theirmain axes are parallel to the coast. The northernpatch is located on the inner-middle shelf, between25 and 50m depth, extending from the Besos Rivermouth to the Port of Barcelona. It is 6.5 km longand 3 km wide, and it is detached from the coastlineby 1 km. The larger southern patch is placed mostlyon the middle-outer shelf, between 25 and 85mdepth. It is 13 km long and 4 km wide, and theaverage distance from its shallowest boundary tothe shoreline is 2 km.
Low backscatter values are typically linked to finegrain sizes. Surface samples LL-9 and LL-10 werecollected from the southern patch (Fig. 7). Bothsamples are made up of relatively fine material,clayey silts and silts with very fine sands (Table 3)showing a mean grain size of 0.013 and 0.069mm,respectively. Their composition, inclusive of abun-dant mica, very fine-grained quartz and some plantdebris, reveals a strong fluvial influence.
4.2.5. Wavy prodelta front
The Llobregat prodelta front off the river mouthcomprises an elongated area of around 25 km2
(2.2� 12 km) located between 45 and 95m depthcharacterized by sediment undulations parallel tothe bathymetric contours (Fig. 5). The slope in thiswavy seafloor ranges from 0.31 to 31, and back-scatter values are relatively high. The prodeltasediment waves have a wavelength of 60–100mand are 30–80 cm high (Urgeles et al., 2006). Surfacesamples LL-6 and LL-8 were retrieved from theLlobregat prodelta front (Fig. 7). They containpoorly sorted sandy mud (Table 3), mainly ofcontinental origin, with average grain sizes of 0.076and 0.019mm, respectively. Their organic mattercontent is relatively high, exceeding 10% in weight.
4.2.6. The Port of Barcelona area
The ongoing works for the enlargement of thePort of Barcelona translate into dredging, loadingand infilling impacts that affect more than 5 km of
ARTICLE IN PRESS
2°00'E 2°10'E 2°20'E 2°30'E
41°1
0'N
41°2
0'N
41°3
0'N
410000 420000 430000 440000 450000 460000
4560000
4570000
4580000
4590000
-200
-500
LlobregatRiver
BesósRiver
Shelf
edge
10
7
89
5
6C4
2
11
D BA
Fig. 7. Backscatter map from the PRODELTA-1 survey at 5m resolution. High backscatter values (lighter tones) represent high seafloor
reflectivity, usually corresponding to coarse sediments. Low backscatter values (darker tones) generally correspond to fine soft sediments,
and are concentrated south-westwards from the Besos and Llobregat river mouths as discussed in the text. Black dots point to the
sediment sampling stations (see also Table 3). Map projection is UTM 31N WGS84.
C. Liquete et al. / Continental Shelf Research 27 (2007) 1779–1800 1789
shoreline and 2 km offshore, as deep as 25m depth(Figs. 5 and 8). The two main trenches to be filledwith concrete blocks are 200m wide and 2.1 and3.8 km long, respectively. The new port breakwatersembrace a future infill area of around 7.4 km2.
Independently of the current official anchoringareas off the Port of Barcelona, two zonescharacterized by a rough surface and high andirregular backscatter values have been interpreted asthe ones most affected by anchoring (Fig. 8). Singlemarks found on these areas reach up to 70 cm deep.The first of these zones spreads from the old jettydown to 30m depth, between 411200N and 411220N.The second one corresponds to the rather largeofficial anchoring area of the Barcelona harbor
(Fig. 8), with the strongest impacts on its north-eastern part between 40 and 70m depth.
4.2.7. Dredging marks
The bathymetric data show several dredge marksand trenches resulting from dredging operations.The largest of these dredging trenches found in thestudy area is about 400� 100m and is locatedaround 41115.50 2130E. Dredging for beach nour-ishment has been common during the last decadesfrom Masnou northwards usually between 10 and40m depth, and between 2180E and 2130E from 10 to25m depth (Fig. 8). These activities are assumed tohave a significant impact over the benthic commu-nities and over the water column turbidity.
ARTICLE IN PRESS
2°00'E 2°10'E 2°20'E 2°30'E
41
°10
'N4
1°2
0'N
41
°30
'N
410000 420000 430000 440000 450000 460000
45
60
00
04
57
00
00
45
80
00
04
59
00
00
-200
LlobregatRiver
BesósRiver
Shelfedge
Summer minimumtrawling depth (-65m)
Winter minimumtrawling depth (-50m)
Sewage pipes
Natural Llobregatlow course
rawl marks concentrations
Areas dredged for beach nourishment
Artificial reefs
Sea grass meadows protected areas
T
Port of Barcelona anchoring area
Observed anchoring impact
Port of Barcelona enlargement area
Fig. 8. Main human impacts on the Barcelona continental shelf. The Llobregat River lowermost course was diverted to the south in
September 2004. Map projection is UTM 31N WGS84.
C. Liquete et al. / Continental Shelf Research 27 (2007) 1779–18001790
4.2.8. Sewage pipes
Two 3-km-long submarine sewer pipes placednear the Llobregat and Besos river mouths (Fig. 8)exit from the ‘‘Baix Llobregat’’ and ‘‘El Besos’’water treatment plants that treat around 0.4 and0.5 hm3 per day, respectively, for a total populationof nearly 5 million (EMMA, 2003). Nearshore thepipes are placed within up to 2m deep narrowtrenches, while seawards the tubes emerge progres-sively till lying on the seafloor where they arecovered by a sediment pile of a few tens ofcentimeters. Leakages from the Besos pipe resultedin several mound-shaped features along its path.
4.2.9. Trawl marks
Fishing activities on the Barcelona margin aredecreasing but are still quite intense. Trawlersrepresent around 17% of the Barcelona fishing fleet
(DARP, 1997a). Trawling is banned shallower than50m depth during winter and shallower than 65mduring summer (Fig. 8). Most trawling marks onthe Barcelona shelf consist of straight furrows lessthan 1m deep and up to a few kilometers long.These marks concentrate in three areas: 411230N21160E between 50 and 70m depth, 411160N 217.50Ebetween 50 and 60m, and 41113.50N 214.50Ebetween 65 and 85m. Trawling activities havestrong effects on benthic communities, sedimentfabric and resuspension, and seafloor microtopo-graphy (e.g. Palanques et al., 2001; Liquete et al.,2004a).
4.2.10. Artificial reefs
Typical artificial reefs are concrete and steelblocks placed on the seafloor to offer shelter andhard substrata to plant and animal species in and
ARTICLE IN PRESSC. Liquete et al. / Continental Shelf Research 27 (2007) 1779–1800 1791
around the reef. Their form and size depend onwhether they have production or protection pur-poses. The production reef blocks weight around4T and are 3m in height, while the protectionblocks can be nearly double weighted and half sized(DARP, 1997b). The production reefs are intendedto ease the development of new or existing commu-nities. The protection reefs form barriers to defendweak benthic communities from trawling gears.
Off Barcelona, a reef complex of nearly 100production and protection artificial structures wasinstalled around 411230N 2112.50E at 15–30m depth,covering about 0.5 km2. The production blocks aregrouped in compact clusters while the protectionblocks are distributed following a sinuous line. Asecond artificial reef area locates north of the BesosRiver. It was created in 1992 and it is known as‘‘El Masnou’’. Some of the individual elements ofthat reef have been identified in our dataset.
2°00'E 2°20'E
4.2.11. Troughs and highs alignments
On the inner shelf north of the Besos River, twotroughs and highs alignments have been observed.The highs have elongated irregular forms thattranslate into abrupt bathymetric and backscatterchanges. The highs are up to 1m high and a few tensto 200m long. The deeper alignment starts off theBesos River mouth and follows the 20m contouralong 18 km. It continues northwards along the edgeof the large submarine step described within thesection 4.2.3. The shallower alignment initiates at411280N on the top of the same large step. It roughlyfollows the 12m contour northwards along 7 km.The highs and the area in between the twoalignments show relatively high backscatter, whichcould be indicative of coarse sediments but also ofother textural characteristics such as the presence ofsedimentary structures or bioturbation.
41
°20
’E
Shelf edge
Main shelf domains:
River-influenced domai n
Relict domain
Human impacted domain
Fig. 9. Sketch showing the general zonation of the Barcelona
shelf. It comprises a modern, river-influenced shelf domain; a
relict, sediment depleted shelf domain; and a severely anthro-
pogenically impacted area. Map projection is UTM 31NWGS84.
5. Discussion
This section attempts to: (1) identify the differentBarcelona shelf sedimentary environments; (2)interpret the main sedimentary features and theirorigin linked to fluvial or submarine processes;(3) analyze the recent evolution of the shelf fromobservable features related to sea level changes; and(4) emphasize the importance of the current humanimpacts on the continental shelf off Barcelona andtheir future increasing effects or mitigation.
5.1. Shelf floor zonation
From a morphosedimentary viewpoint, the multi-beam bathymetry and backscatter maps of theBarcelona shelf allow identifying two main do-mains: (i) a relict, sediment depleted shelf floor, and(ii) a modern, river-influenced shelf floor (Fig. 9). Avariety of anthropogenic impacts are overimposedwith different intensities on these two domains.
The relict shelf, markedly depleted of modernsediment accumulation, covers the southern end ofthe study area and most of the northern sector(Fig. 9). It comprises the widest part of theBarcelona shelf, reaching up to 20 km in width. Itsrelict character is evidenced by a relatively roughtopography and a number of paleorelieves andsandy deposits that are described in greater detail inthe following section. The relict shelf is also distin-guishable by high to medium backscatter intensitiescorresponding to coarse or cemented materials barelycovered by fine grained modern sediments.
The river-influenced shelf is directly affected bymodern fluvial inputs. The main sediment body isthe Llobregat and Besos joint prodelta that appearson the seismic data as a seaward thinning mudwedge up to 54m thick and 35 km long (Liquete etal., 2006). Off the Llobregat River mouth theprodelta practically reaches the shelf edge at about6.5 km, while off the Besos River mouth it reachesthe 50m isobath at about 2.5 km from the shoreline.The prodeltaic wedge extends to the southwest dueto the littoral drift and dominant currents in the
ARTICLE IN PRESSC. Liquete et al. / Continental Shelf Research 27 (2007) 1779–18001792
study area (see the section 2). North of the BesosRiver mouth, modern sedimentation is confined to anarrow alongshore band (Fig. 9) characterized bymodern littoral sandy prisms.
The anthropogenic overprint is particularly ap-parent along a 13 km long littoral fringe thatextends between the Besos and Llobregat rivermouths down to 50m depth (Fig. 9). Dredging andinfilling for the extension of the Port of Barcelona(still ongoing), sewage pipes placement and releases,and numerous coastline structures are the mainanthropogenic impacts in the area.
5.2. Origin of the sediment features
5.2.1. On the river-influenced shelf
The modern river-influenced continental shelf ischaracterized by a smooth seafloor and relativelylow backscatter values. This domain reaches itsmain development in the adjacent Llobregat andBesos prodeltas, which represent the principalHolocene sediment accumulation on the Barcelonacontinental shelf. Water and solid discharge valuesevidence that the Llobregat River is the mainmodern sediment supplier to the Barcelona shelf,despite the slightly decreasing trend of its meanwater and sediment discharge. In the Llobregatdelta/prodelta environment, accumulation ratesreached 12–25mmyr�1 between 4.5 and 0.4 kaBP(Gamez et al., 2005) and 0.7–1.5mmyr�1 during thepast century (Sanchez-Cabeza et al., 1999). Theprodeltas cover an area of about 190 km2 and theirlocation southwards of the parent river mouthsis determined by the south-westward dominantcoastal current.
The Tordera river, with a catchment area and amean water discharge similar to the Besos ones,does not show a recognizable typical prodelta. Thisis attributable to the lesser suspended sedimentconcentration of the Tordera, linked to the pre-dominantly granitic character of its basin, and to itsminor base water regime (i.e. when excludingmaximum discharge events). The Tordera systemshows a marked torrential character typical of thesmall rivulets in the study area. It is also likely thatsuspended sediment inputs by the Tordera Riverevolve into hyperpycnal flows (Mulder and Syvitski,1995; Mulder et al., 2003) that are trapped by thenearby Blanes submarine canyon head.
5.2.1.1. Large sediment prisms. The large submar-ine step to the north of Masnou (Figs. 5, 10B and
11B) has been ascribed to the modern fluvialdominated shelf as, according to the availableinformation, it is actively involved in the present-day littoral processes. It occurs at water depthsshallower than the wave-face depth of major storms(see the General Setting section). Similar submergedsandy bodies interpreted as coastal paleo-spits havebeen described at 5–10m depth off the TorderaRiver mouth (Serra et al., 2003). However, theorigin of the submarine step that appears in thenorthernmost part of the study area is likely relatedto the larger and deeper sediment prism that extends20 km north-eastwards beyond the Barcelona shelfas far as the Blanes Canyon (Calafat, 1986; ITGE,1989; Dıaz and Maldonado, 1990; Liquete et al.,2004b). The dimensions, morphology and sandycomposition of this submarine step suggest that itwas formed during an ancient lower-than-today sealevel stage.
5.2.1.2. Featureless seafloor with low backscatter.
Recent sediment accumulation over the modernriver-influenced shelf results in a relatively smoothseabed, except for the Llobregat wavy prodeltafront and the human impacted areas. The two lowbackscatter patches identified in this study areindicative of a fine sediment cover that overliespart of the Llobregat and Besos prodeltas. Bothpatches appear close to the Llobregat and Besosriver mouths, respectively, and extend south-westwards. The location and composition of thelow backscatter patches point to a fluvial origin.The larger size and distance to the coast of thesouthern patch is attributed to the greater input ofthe Llobregat River related to the Besos River,which is 2.5 times greater in water discharge and4.5 times greater in solid discharge. It is suggestedthat the low backscatter patches represent theaccumulation of the finest fluvial material thatsettled from short-lived sediment plumes follow-ing strong discharge events (Arnau et al., 2004)and spread south-westwards under the directinfluence of coastal currents.
5.2.1.3. Wavy prodelta front. Sediment-wave de-posits are relatively common on the seabed butmostly in deep water environments where they havebeen linked to turbidity currents or destabilizationprocesses (Alonso et al., 1995; Lee et al., 2002). Thesediment undulations that cover the Llobregatprodelta front (Fig. 10C) are relatively small sizedand consist of fluvial material. Shallow-water
ARTICLE IN PRESS
Fig. 10. 5m (A) and 1m (B, C, D) resolution zooms of bathymetric data showing specific features mentioned in the text. (A) Crossing
between older N090 beachrocks and younger N030 beachrocks, both used as paleo-coastal indicators. (B) Troughs and highs alignments
that could be interpreted either as biogenic constructions (sea grass meadows) or as dredging marks. (C) Trench and sewer pipe exiting
from the ‘‘Baix Llobregat’’ water treatment plant. At depths 440m the pipe lies directly on the sea bed. Large scale seafloor trenching for
the enlargement of the Port of Barcelona can be observed in the upper part of the image. The wavy character of the Llobregat prodelta
front is shown on the lower-right corner of the image. (D) Seafloor expression of tectonic and morphosedimentary features of the relict
Barcelona shelf. BR: Beachrocks. F: Faults. SB: Prograding sediment bodies. See location on Fig. 5. Map projection is UTM 31NWGS84.
C. Liquete et al. / Continental Shelf Research 27 (2007) 1779–1800 1793
undulations, often found in prodelta settings, havebeen associated to sediment deformation, or fluidexpulsion or bottom currents (Trincardi and Nor-mark, 1988; Bornhold and Prior, 1990; Chiocci etal., 1996; Lee et al., 2002; Cattaneo et al., 2004).Offshore decreasing wavelengths, internal parallelreflectors and the lack of headwall scarps, suggestthat the Llobregat prodelta front sediment wavesare the result of bottom currents possibly associatedto hyperpycnal flows during peak discharge events
by the Llobregat River (Urgeles et al., 2006).However, a genetic relationship with deformationprocesses of the gas-charged, soft prodeltaic sedi-ments cannot be totally ruled out.
5.2.2. On the relict shelf
Most of the study area has been classified as amodern sediment-starved shelf where relict mor-phological features crop out. The relict shelf owesits character to the limited fluvial inputs and to the
ARTICLE IN PRESS
2°20'E 2°30'E
41
°20
'N4
1°3
0'N
440000 450000 460000
45
80
00
04
59
00
00Besos
River
Shelfedge
Outer edge of sediment bodiesOutcropping faultsBeachrocks
Fig. 11. (A) Location of distinctive seafloor features identified on the northern, relict Barcelona shelf. (B) Distribution of the two sets of
beachrocks and prograding sediment bodies used as paleo-coastline indicators. arrows represent the hypothesized direction of the
shoreline landwards migration during the formation of N090 features (arrow ‘‘a’’, post-MIS 4) and N030 features (arrow ‘‘b’’, post-MIS
2). The inner shelf large submarine step to the north has been interpreted as active and, therefore, is not attributed to any of the paleo-
coastal sets. Map projection is UTM 31N WGS84.
C. Liquete et al. / Continental Shelf Research 27 (2007) 1779–18001794
focusing of sedimentation at the inner shelf andprodeltas. The southwards dominated littoral driftand Northern Current (see the General Settingsection) and the physiography of the margin favorsuch a situation. An unquantified part of the silt andclay drifting on the continental shelf is suggested tobe trapped by canyon heads (Granata et al., 1999;Sanchez-Cabeza et al., 1999; Puig et al., 2000). Theinterpretation of the morphosedimentary featureson the relict shelf domain brings to light newevidences about the sedimentary evolution of theBarcelona shelf during the late Pleistocene period.
5.2.2.1. Abrupt slopes. The straight pronouncedslopes of the relict shelf have been correlated inthe available crossing seismic profiles with gentlediscontinuities of the seabed and sub-surfacereflectors (Fig. 6A and B), taking into accountthat these gentle discontinuities can representseveral meters high breaks (see the section 3).The seismic representation of the abrupt slopesand their characteristic bathymetric expression leadus to interpret them as a possible set of submarinefaults with a rather constant NE–SW orientation(Figs. 10D and 11A).
Maillard and Mauffret (1999) and Roca et al.(1999), among others, confirmed the tectonicreactivation of the Catalan margin during theQuaternary. However, the tectonic regime respon-sible for this reactivation is disputed. Some authors(e.g. Dıaz del Rıo et al., 1986; Maillard andMauffret, 1999) propose an ongoing distensiveregime, while others (e.g. Goula et al., 1999) linkthe Catalan margin reactivation to compressionalstresses. These contrasting views leave open for thesubmarine faults identified in this work to be eithernormal or strike-slip structures.
The Barcelona region holds a moderate seismicactivity. Masana (1996) found morphological evi-dences of younger than 100 kaBP earthquakes witha magnitude 46.3 in the southern Catalan CoastalRanges. Also in the Catalan Coastal Ranges, Pereaet al. (2003) recognized two paleoseismic eventsbetween 125 and 34 kaBP and a third one youngerthan 13.5 kaBP, and suggested that earthquakes ofmagnitude 7 or above could have occurred in theregion. Two pre-instrumental epicenters dated on1427 and 1784 have been identified within a 150 kmdistance from Barcelona (NEIC, 2004). Four earth-quakes ranging between 4.5 and 5.5 magnitude, and
ARTICLE IN PRESSC. Liquete et al. / Continental Shelf Research 27 (2007) 1779–1800 1795
tens of them between 3.5 and 4.5 magnitude, havebeen instrumentally detected in the same area since1980 (NEIC, 2004). These seismic events maysupport the possibility to find active faults on theshelf.
Therefore, the abrupt slopes identified in thestudy area can be attributed to the seafloorexpression of a local fault set linked to thereactivation of the deep Neogene extensionalstructures (horst-and-graben) associated to tectonicdeformation and/or to local seismicity.
5.2.2.2. Narrow ridges. The sharp ridges on therelict domain are ascribed to cemented beachrocks(Figs. 10A and D, and 11A), most likely of marinevadose origin. Sediment samples LL-B, LL-A andLL-D support this interpretation as they consist ofbioclastic to siliciclastic-bioclastic, rounded to well-rounded, sands and gravels that are typical ofbeachrock formations (Dalongeville and Sanlaville,1984; Bernier and Dalongeville, 1996; Neumeier,1998). Similar structures, also interpreted as bea-chrocks, have been found on the inner shelf betweenArgentona and Sant Cebria rivulets (Fig. 1) (Serra,2002). The form of the identified beachrocks (sharp,irregular, rather symmetrical ridges) suggests thatthe not-cemented surrounding sediments must havebeen eroded while the considerable size of thebeachrocks likely favored their preservation.
Beachrocks are useful tools to identify ancientcoastlines. Their formation implies carbonate diag-enesis aided by biotic and abiotic factors in a stablesedimentary environment. Amongst promoting fac-tors the following have been cited: evaporation;rising temperatures; CO2-degassing of interstitialwater pushed into the sediment by waves and tides;lowering of PCO2
by algal photosynthesis; microbialactivity; and decreasing pH (Tucker and Wright,1990; Neumeier, 1998). The warm and confinedMediterranean Sea is a favorable place for beachcementation (e.g. Magaritz et al., 1979; Holail andRashed, 1992; Bernier and Dalongeville, 1996). Theusefulness of beachrocks to reconstruct the evolu-tion of the shoreline in the study area is discussedfarther below.
5.2.2.3. Prograding sediment bodies. Inclined(o101) cross-strata and parallel/sub-parallel lami-nae are usually indicative of intertidal or shallowsubtidal depositional facies, while higher anglestratification should correspond to eolian deposits.According to the criteria by Otvos (2000), the
setting and morphology of the prograding sedimentbodies on the Barcelona shelf fit with those of relictcoastal ridges resulting from transgressive orregressive histories. The prograding sediment bodieshave been interpreted as relict wave- or swash-builtbeach ridges (Figs. 10D and 11A). Therefore, suchsediment bodies of likely sandy nature mark theposition of ancient shorelines that are significantlyoblique to the present-day shoreline, but semi-parallel to the depth contours of the shelf.
5.3. Morphosedimentary evidences of sea level
change
We consider that the beachrocks and progradingsand bodies on the relict Barcelona shelf formedduring relatively short ancient sea level stillstands.Sea level changes during the late Quaternary arewell-documented around the western Mediterra-nean Sea (e.g. Alessio et al., 1996; Lambeck andBard, 2000; Lambeck et al., 2002; Antonioli et al.,2004). In contrast, there are much less publishedseafloor evidences on late Pleistocene stillstands.Collina-Girard (2002) identified up to seven relictshorelines between 11 and 100m depth in the NWMediterranean Basin, along the coasts of southernFrance, Corsica and the Italian Isle of Elba.
According to their orientation, the paleo-coastalindicators of the relict Barcelona shelf belong to twomain sets: (i) a N090 oriented set inclusive ofbeachrocks and prograding sediment bodies and (ii)a N030 oriented set made of well-defined beach-rocks (Figs. 10A and 11B). Landwards of 55mdepth, both sets tend to deflect towards a N060orientation subparallel to the present-day coastline.
(i) The main N090 paleo-coastal indicatorsappear at about 104m (prograding body), 70–77m(beachrock and prograding body), and 48–73m(large branched beachrock) below present sea level(mbpsl). These N090 morphosedimentary elementscut the dominant present-day bathymetric trend,which suggests that the landward migration(s) ofthe coastline during their formation followed a S-Ndirection, i.e. shifted 451 related to the present-dayshoreline (Fig. 11B).
(ii) The N030 paleo-coastal indicators are pre-served at 69–76, 54–55, 63–69, and 48–53 mbpsl.They likely formed as a result of a SE-NWadvancing shoreline pushed by one or more eventsof sea level rise (Fig. 11B).
In a first view, the N030 paleo-coastal indica-tors seem to correspond to a coastal regression
ARTICLE IN PRESSC. Liquete et al. / Continental Shelf Research 27 (2007) 1779–18001796
semi-parallel to the depth contours, while the N090prograding sediment bodies could match to sandybars in dynamic equilibrium with those N030beaches. However, the high angle between the twosets of paleo-coastal indicators, together with theintersection of N030 and N090 lithified beaches inthe northeastern part of the study area (Fig. 10A),make us suggest that both sets were developedduring different sea level cycles under different localconditions.
Late Quaternary sea level curves (Fig. 12) varyalong a pattern that is governed by ca. 100 ka high-amplitude (more than 100m) glacio-eustatic cyclesmodulated by higher frequency cyclicities (Imbrieet al., 1984). Orbital forcing, particularly Earth’seccentricity, obliquity and precession, have beenassumed as the triggers for these climatic and eustaticcycles (e.g. Imbrie et al., 1984; Shackelton, 2000).Higher frequency climatic cycles such as Heinrichevents or Dansgaard–Oeschger fluctuations influencesedimentological patterns from river basins to theseafloor (Bard et al., 2006). The two last glacio-eustatic lowstands, reaching 80–110 and110–120mbpsl, respectively, occurred during MarineIsotope Stages (MIS) 4 and 2 (Fig. 12). The rapid sealevel rises following these lowstands favored theflooding and preservation of morphosedimentaryfeatures along the former shorelines. Therefore wesuggest that the paleo-coastal indicators found on therelict Barcelona continental shelf were formed duringthe transgressions that followed MIS 4 and MIS 2.Sediment sampling and dating is required to confirmour interpretation. This would require rock coringdrilling of the hard beachrocks and the subcroppingand buried prograding sediment bodies.
MIMIS3 421
0 50 1
Age (
0
40
80
120Rela
tive s
ea level (m
bp
sl)
Waelbroeck et al., 2002SPECMAP curveSiddall et al., 2003
Fig. 12. Relative sea level curves estimated from d18O records. Vertical
Waelbroeck et al. (2002) curve corresponds to a global model; the SPEC
the Atlantic Ocean; and the Siddall et al. (2003) curve originates from
This interpretation of the paleo-coastal indicatorsimplies that between MIS 4 and MIS 2 thetopographic/bathymetric trend of the Barcelonacontinental shelf experienced significant directionalshift from S-N to SE-NW and so did the maindirection of the subsequent marine transgressions.We hypothesize that the causes for such directionalshift are (a) sedimentation changes involvingdepocenter migration due to modifications insediment point sources and oceanographic condi-tions; (b) the regional neotectonic activity stressingthe main NE-SW transfer faults, which have anintermediate orientation between the two sets ofpaleo-shoreline indicators; (c) a combination oftectonic events and entry points and depocentersmigration.
5.4. Importance of anthropogenic impacts
Man made actions leave their imprint over theshallow seafloor, as can be observed in theBarcelona shelf. Multibeam data are the perfecttool to control submarine human impacts and tomonitor their possible natural mitigation. Thefollowing section discusses the relative importanceand probable consequences of the main anthropo-genic actions on the Barcelona continental shelf.
The Port of Barcelona area. The Port ofBarcelona enlargement works have a strong impactover a large part of the Barcelona inner shelf(Fig. 10C) and the Llobregat prodelta depocenter.As the continental shelf is narrowest off theenlargement area, the new harbor area will compriseup to 35% of the Llobregat prodelta width and upto 29% of the entire continental shelf (Fig. 8).
MIS6S5
00 150 200
ka BP)
scale is in meters with respect to present sea level (zero value). The
MAP curve (Imbrie et al., 1984; Imbrie et al., 1990) derives from
the Red Sea. MIS: Marine Isotope Stages.
ARTICLE IN PRESSC. Liquete et al. / Continental Shelf Research 27 (2007) 1779–1800 1797
Major alterations are expected on coastal circula-tion and sediment-dispersal patterns. Furthermore,the diversion of the Llobregat lowermost course andmouth will likely cause the starvation of the wavyprodelta front and the establishment of a newsediment balance in the area.
Massive anchor ploughing leads to seafloorerosion and sediment resuspension. The ongoingchanges around the Port of Barcelona, involving theshifting of anchoring areas, provide an excellentopportunity to monitor the recovery of abandonedanchoring sites (like the one that extends off the oldbreakwater down to 30m depth) and the progres-sive disturbance of newly established anchoring sites(like the southern relatively small anchoring areaillustrated in Fig. 8).
Dredging impacts. Dredging for beach nourish-ment in the study area is mostly performed at shortdistance from the shoreline, at depths shallowerthan the major storm wave-face. Therefore, asdredging is performed at water depths undergoingactive sediment transport, the marks left on theseafloor tend to vanish in a matter of months or afew years. The impact of dredging on the benthicecosystem is punctually and locally remarkable.More than two years are required for dredged sandybottoms to restructure to their initial ecologicalconditions, as observed on the inner shelf to thenorth of the study area (Sarda et al., 2000). Whetheror not dredging for beach nourishment will continuein the future is essentially a matter of politicaldecision. However, the impact of coastal infrastruc-tures, as illustrated by the enlargement of the Portof Barcelona, will impose a strong need for infillingand coastline stabilization for which large volumesof sediment will be required.
Sewage pipes. The two long sewer pipes comingout from Barcelona efficiently dispose the sewagewater (Fig. 10C). However, the sedimentary signs ofleak away sites, long after the opening of a mudtreatment plant, indicate that either the system isnot efficient enough or some old sewer sedimentaccumulations have not been dispersed after years.
Trawling impacts. Trawl gears cause significantreworking and resuspension of the uppermostsediment cover, increasing the suspended sedimentconcentration in the water column and disturbingthe seafloor ecosystem. Trawling over the muddyLlobregat prodelta causes significant turbidityincreases lasting several days (Palanques et al.,2001). Trawling marks may remain on the seafloorfor years. Trawl marks have been observed very
close or even trespassing the shallowest legalboundary for this kind of activity, at short distancefrom artificial reefs and over the buried sewer pipes.
Artificial reefs. From the observations made,protection artificial reefs seem locally efficient inpreventing direct impacts on the benthic ecosystem,such as those from trawling, dredging or structureplacement, while their impact on sediment transportis negligible. When deployed on soft bottoms theytend to sink and get partially buried, thereforepotentially loosing part of its efficiency.
Troughs and highs alignments. The troughs andhighs alignments on the northern inner shelf(Fig. 10B) of the study area extend along severalkilometers. Their origin is a matter of debate. Thefirst hypothesis is that they result from beachnourishment dredging mostly performed duringthe 1980s. However, their large size, sharp bound-aries and location over sea grass meadows protectedareas (Fig. 8) points to a biogenic origin, likelyphanerogam upbuilts. The endemic MediterraneanPosidonia oceanica has been reported very close tothe northern limit of the study area down to 30mdepth (Manzanera and Cardell, 2002; Llamas,2003), while another phanerogam, Cymodocea
nodosa, has been identified within the study areaaround 15m depth (Castaneda, 1994). However, theupbuilding potential of C. nodosa is negligiblecompared to P. oceanica. Sea grass meadows donot spread southwards likely because of humanpressure, too turbid waters related to the Besos andLlobregat discharge, and the dominance of soft,prodelta bottoms.
6. Conclusions
The seafloor and subseafloor of the Barcelonacontinental shelf records its late Quaternary historyand the imprint of the factors controlling thissediment-dispersal system. Two main morphosedi-mentary domains have been recognized: (i) amodern, river-influenced area, and (ii) a relict,sediment depleted area. Both domains are affectedby a variety of anthropogenic impacts.
The main sediment source of the Barcelona shelfis the Llobregat river system. However, the con-tinental shelf reaches its minimum width in front ofthe Llobregat River mouth as most of the sedimentinputs are deviated southward by the dominantcoastal circulation. The modern, river-influencedshelf includes the Llobregat and Besos adjacentprodeltas, which represent the main Holocene
ARTICLE IN PRESSC. Liquete et al. / Continental Shelf Research 27 (2007) 1779–18001798
depocenter in the study area. The wavy nature ofthe Llobregat prodelta front may result either fromhyperpycnal and bottom currents or from softsediment destabilization. The finest fluvial materialis preferentially accumulated along two patchesshifted south-westwards of the parent river mouthas a result of the dominant littoral circulation.
A large part of the study area is sediment-depleted where relict morphosedimentary features,such a set of NE-SW faults, crop out. A number ofpaleo-coastal indicators consisting of beach ridgesand beachrocks are linked to short late Pleistoce-ne–Holocene stillstands. The paleo-coastal indica-tors point to a 451 shift of the seashore directionbetween MIS 4 and MIS 2. This directional shift istentatively attributed to tectonic activity along theNE-SW faulting system that involved modificationsof entry points and depocenter locations.
Anthropogenic impacts are widespread on theBarcelona inner and middle shelf, especially alongthe Besos-Llobregat coastal stretch. Coastal infra-structures already have a major impact on sedimentdispersal patterns, that will likely increase in thenear future. The Port of Barcelona directly affects alarge part of the Llobregat prodelta and has thepotential to disrupt the natural paths of sedimenttransport. Many human activities cause sedimentresuspension and disturb the benthic environment.The seafloor sediment cover is largely reworked bydredging, anchoring, and trawling activities. Prob-able sea grass meadows have been accuratelycharted.
Acknowledgements
The data shown in this paper come from thecollective effort of the whole ‘‘Arraix’’ ShipboardParty: Anna Sanchez-Vidal, Antonio Calafat,Carles, Cesar, Diana Zuniga, Jaime Frigola, JoanFabres, Jordi, Jose Luis Casamor, Koen De Rycker,Pedro Arnau, Sara Lafuerza, Sergi, Veronica Will-mott, Victor Centella and Wim Versteeg. Thankyou for your excellent support. The authorsacknowledge the information provided by ACAand CEDEX. This research has been funded by theSpanish PRODELTA project (REN2002-02323),the EU EURODELTA Concerned Action (EVK3-CT2001-20001), and the ‘‘Direccio General de Pescai Afers Maritims’’ from the ‘‘Generalitat deCatalunya’’ autonomous government. GRC Geo-ciencies Marines (GRCGM) is funded by ‘‘General-itat de Catalunya’’ excellence research grants
program (ref. 2005 SGR-00152). GRCGM alsoacknowledges the support received from LandmarkGraphics Corporation via the Landmark UniversityGrant Program, and from SMT Inc. via theeducational User License for Kingdom Suite inter-pretation software.
References
ACA, 2000. Embassaments. Web page of the ‘‘Agencia Catalana
de l’Aigua,’’ /http://mediambient.gencat.net/aca/ca//actua
cions/embassaments/inici.jspS. Last revision in 2006.
ACA, 2005. Caracteritzacio de masses d’aigua i analisi del risc
d’incompliment dels objetius de la Directiva Marc de l’Aigua
(2000/60/CE) a Catalunya (conques intra i intercomunitaries).
Agencia Catalana de l’Aigua, Generalitat de Catalunya,
860pp.
Alessio, M., Allegri, L., Antonioli, F., Belluomini, G., Improta,
S., Manfra, L., Preite, M., 1996. La curva di risalita del Mare
Tirreno negli ultimi 43 ka ricavata da datazioni su speleotemi
sommersi e dati archeologici. Memorie Descrittive del
Servizio Geologico Nazionale 52, 235–256.
Alonso, B., Canals, C., Palanques, A., Rehault, J.-P., 1995. A
Deep-sea Channel in the Northwestern Mediterranean Sea:
Morphology and seismic structure of the Valencia Channel
and its surroundings. Marine Geophysical Researches 17 (5),
469–484.
Amblas, D., Canals, M., Lastras, G., Berne, S., Loubrieu, B.,
2004. Imaging the seascapes of the Mediterranean. Oceano-
graphy 17 (4), 144–155.
Antonioli, F., Bard, E., Silenzi, S., Potter, E.K., Improta, S.,
2004. 215 kyr history of sea level based on submerged
speleothems. Global and Planetary Change 43, 57–78.
Arnau, P., Liquete, C., Canals, M., 2004. River mouth plume
events and their dispersal in the Northwestern Mediterranean
Sea. Oceanography 17 (3), 22–31.
Bard, E., Rostek, F., Menot-Combes, G., 2006. Chronologie des
variations climatiques rapides pendant la derniere periode
glaciaire. Comptes Rendus Palevol 5, in press.
Bartrina, M.T., Cabrera, L., Jurado, M.J., Guimera, J., Roca, E.,
1992. Evolution of the central Catalan margin of the Valencia
