Pockmarks, collapses and blind valleys in the Gulf of Cádiz

ORIGINAL

Pockmarks, collapses and blind valleys in the Gulf of Cádiz

Ricardo León & Luís Somoza & Teresa Medialdea &

Francisco Javier Hernández-Molina & Juan Tomás Vázquez & Victor Díaz-del-Rio &

Francisco Javier González

Received: 15 February 2009 /Accepted: 5 October 2009 /Published online: 21 October 2009# Springer-Verlag 2009

Abstract Herein we describe a suite of fluid escapedepression features, including pockmarks and collapsestructures, discovered in the Gulf of Cádiz (Spain) duringseveral recent cruises. We also establish an evolutionarymodel for these depressions and discuss the generation ofbottom undercurrent furrows from fluid-flow structures,considering the oceanographic and tectonic framework andgas expulsion mechanisms. We describe for the first timeblind valleys, which we define as giant, elongated (3 to10 km long), collapsed and complex fault-strike featurescomprising mega-collapses and mega-pockmarks, generat-ed in gas-venting areas and not associated to the collapse ofmud-volcano complexes. We detected the blind valleysabove diapiric structures. The collapse processes associatedto blind valleys result from fluid escape through migrationpathways which, in turn, are created by distension due todiapiric activity or to later tectonic reactivation of thesediapirs. The evolution of these blind valleys, and theirpresent-day morphology as furrows, derives from progres-sive fluid migration as well as from interaction ofMediterranean Outflow Water with the seafloor.

Introduction

Pockmarks are the most common depression structures inthe seafloor related to seabed fluid flow, and were firstreported on side scan records from the Scotian Shelf,Canada (King and MacLean 1970). These crater-likedepressions are commonly found in unconsolidated fine-grained sediments and are generated by fluid migrationthrough the sediments (Hovland and Judd 1988; Kelley etal. 1994). They are erosional structures which differ greatlyin shape and size, generally ranging from a few meters to800 m or more in diameter and from 1 to 100 m in depth(Judd and Hovland 2007). These important seabed featurescan provide information about fluid seepage in continentalmargins. The nature of the fluids expelled and theorganisation of pockmarks on the seafloor may beindicative of the depth of the reservoir and any potentialfluid migration pathways (Heggland 1998). Depression-likepockmarks occur in myriad physiographic domains (e.g.coastal, continental shelf, slope and rise, and abyssal plains)and geological contexts (e.g. convergent, divergent andtransform plate boundaries, as well as interplate settings;Judd and Hovland 2007). Nevertheless, their formation anddynamics remain poorly understood and, therefore, requirefurther study. In some cases, the development of pockmarkssuggests tectonic control related to structural surfaces alongbedrock (Shaw et al. 1997), salt diapirs (Schmuck and Paull1993; Taylor et al. 2000) or faults and faulted anticlines(Bøe et al. 1998; Eichhubl et al. 2000; Gay et al. 2004)which create pathways for fluid migration. Furthermore,gas hydrate dissociation can be crucial to pockmarkgeneration along the continental slope (Vogt et al. 1999;Paull et al. 2000; Wood et al. 2002; Hovland et al. 2005).

Several reports on the Gulf of Cádiz (Spain) haveanalysed seabed fluid flow along the continental slope by

R. León (*) : L. Somoza : T. Medialdea : F. J. GonzálezMarine Geology Division, Geological Survey of Spain, IGME,Rios Rosas 23,Madrid 28003, Spaine-mail: [email protected]

J. T. Vázquez :V. Díaz-del-RioCentro Oceanográfico de Málaga,Instituto Español de Oceanografía,Málaga 29640, Spain

F. J. Hernández-MolinaUniversidad de Vigo,Vigo 36200, Spain

Geo-Mar Lett (2010) 30:231–247DOI 10.1007/s00367-009-0169-z

evaluating different structures: pockmarks on the uppercontinental slope (Baraza and Ercilla 1996; Somoza et al.2000, 2003; León et al. 2001, 2006); mud volcanoes, someof which bear gas hydrates (Ivanov et al. 2000; Somoza etal. 2000, 2003; Gardner 2001; Stadnitskaia et al. 2001,Mazurenko et al. 2002); and hydrocarbon-derived authi-genic carbonates (HDACs; Díaz-del-Río et al. 2003;Magalhães 2007; González et al. 2009). These features areindicative of intense hydrocarbon-seep activity on both theIberian and African margins (Ivanov et al. 2000; Somoza etal. 2000; Gardner 2001). Among the aforementionedstructures, crater-like depressions have received littlescientific attention, despite the fact that their distributionreflects widespread fluid venting.

Herein we present new data of pockmarks andcollapses on the continental slope of the Gulf of Cádiz,in the TASYO field area (Somoza et al. 2003; Fig. 1)located in Sector 2 of the Contourite Depositional Systemdefined by Hernández-Molina et al. (2003) and Llave et al.(2007). This area of nearly 100 km contains intenseseafloor seepages and numerous diapiric structures(Somoza et al. 2002; Díaz-del-Río et al. 2003; Medialdeaet al. 2004, 2009). Based also on previously publishedgeophysical, core and dredge data, we characterise thedepressions, and then propose and discuss an evolutionarymodel for their development in the Gulf of Cádiz and theirrelation with contourite channels and furrows stemmingfrom Mediterranean Outflow Water (MOW), consideringthe regional tectonic framework and gas expulsionmechanisms.

Geological setting

The Gulf of Cádiz is situated at the African-Eurasian plateboundary and represents the westernmost part of theAlpine-Mediterranean orogenic belt. Its geological structureresults from the NW-SE African-Eurasian convergence, thedextral strike-slip motion of the Gloria transform fault zoneand the westward migration of the Betic-Rifean Arc(Fig. 1). The African-Eurasian convergence regime causedthe westward drift and collision of the Alborán Domainwith the North African and South Iberian margins in thelate Oligocene-middle Miocene, leading to the developmentof the Rif and Betic orogen (Platt et al. 2006) and,consequently, the emplacement of huge chaotic masses(Maldonado et al. 1999) named the AUGC (AllochthonousUnit of the Gulf of Cádiz; Medialdea et al. 2004; Fig. 1).The AUGC consists of Triassic evaporites and red bedswith blocks, mostly of Upper Cretaceous to Palaeogenelimestones and marlstones ranging from Aquitanian toTortonian (Maldonado et al. 1999).

The singular geodynamic evolution and nature of thegeological units of the Gulf of Cádiz has yielded anirregular physiography with a huge, smooth continentalslope, emplacement of the AUGC and development ofcomplex tectonics and diapirism, and occurrence ofseafloor features related to hydrocarbon seepages, suchas mud volcanoes, carbonate mounds, pockmarks andcollapses. On the continental slope, three physiographicdomains have been defined based on the gradient andon morphological features: the upper, middle and lowerslopes (Heezen and Johnson 1969; Maldonado andNelson 1999; Hernández-Molina et al. 2006). The upperslope occurs from 120–140 to 500 m water depth. Themiddle slope occurs at depths of roughly 500–1,200 m,with a maximum width of 100 km and low gradients of0.5–1°. The lower slope is 50–200 km wide andconnects with the abyssal plains at 4,300–4,800 m waterdepth.

A large Contourite Depositional System (CDS) generatedby the MOW has been characterised within the Gulf of Cádizmiddle slope region (Llave et al. 2001, 2007; Hernández-Molina et al. 2003, 2006). MOW flows below the AtlanticSurficial Water in the Strait of Gibraltar along the northernslope of the Gulf of Cádiz, warming the seafloor (Ochoa andBray 1991). After exiting the Strait, the MOW flows at1,200–1,400 m depth intercalated between the North AtlanticDeep Water and the North Atlantic Central Water, towardsSan Vicente Cape. Due to the Coriolis effect, the MOW turnsto the NW along the middle slope, separating into severalbranches channelled by tectonic structures (e.g. diapirs andbanks) and generating large contourite channels and furrows.MOW effects on the Gulf of Cádiz seafloor comprise thegeneration of a large contourite depositional system encom-passing five morphosedimentary sectors (see Hernández-Molina et al. 2003, 2006; Llave et al. 2007; García et al.2009); incision by bottom currents and the consequentformation of erosive features such as contourite channels,furrows and moats (Kenyon et al. 2000; Hernández-Molinaet al. 2003; García et al. 2009); triggering of gravitationalprocesses leading to the formation of, for example, sedimen-tary lobules and submarine fans (Kenyon et al. 2000;Habgood et al. 2003); and seafloor warming and subsequentalteration of the seafloor hydrate stability field (León et al.2009).

The major tectonic structures in the TASYO fieldcomprise thrust faults, extensional faults, strike-slip faultsand diapirs (Maldonado et al. 1999; Medialdea et al. 2004)which are closely related to fluid escape structures(Medialdea et al. 2009). This is an active seafloor fluid-venting area (León et al. 2001; Somoza et al. 2002, 2003)related to extensive mud diapirism triggered by the strongdeformation of the AUGC (Lowrie et al. 1999; Maldonado

232 Geo-Mar Lett (2010) 30:231–247

et al. 1999; Somoza et al. 1999, 2003; Fernández-Puga etal. 2007). Geochemical studies on gas, interstitial water,sediment and mud breccia clasts samples from coresrecovered from several mud volcanoes in the Gulf ofCádiz reveal a flow of deeply sourced fluids (e.g.Blinova and Stadnitskaia 2001; Stadnitskaia et al. 2001;Mazurenko et al. 2002; Hensen et al. 2007). Stadnitskaia etal. (2000) suggested two mechanisms of upward fluid flow:dispersed upward migration and focused fluid flow, thelatter being more important. Stadnitskaia et al. (2006)identified two distinct gas groups, from the Deep SouthPortuguese Field and the Western Moroccan Field. Bothrepresent allochthonous gas with high input of thermogeni-cally formed hydrocarbons, in a complex of secondarymigrated, microbially altered and mixed hydrocarbons mostlikely redeposited within olistostrome/Pliocene-Quaternarysedimentary units.

Materials and methods

The present study is based on a broad database obtainedduring the TASYO 2000, MOUNDFORCE07 andMVSEIS08 cruises. Multibeam data were acquired aboardthe R/V Hespérides using a hull-mounted Simrad EM-12S-120 (TASYO 2000 cruise) and Simrad EM-120 (MVSEIS08cruise) multibeam echo-sounder which enabled simultaneouscollection of high-resolution seafloor bathymetry and back-scatter (Fig. 2). The former operated at a frequency of13 kHz with 81 beams (swath width of 120º) at pulselengths of 2–10 ms and a vertical resolution of 0.6 m, thelatter at 13 kHz with 191 beams (swath width of 150º) atpulse lengths of 2–15 ms and a vertical resolution of 0.1–0.4 m. CTD and SIPPICAN depth-temperature measure-ments were made with a precision of ±0.15°C. Themultibeam data were processed with Simrad Neptune

Fig. 1 Geological and oceanographic settings of the Gulf of Cádiz,with geological and tectonic boundaries of the Allochthonous Unit ofthe Gulf of Cádiz (AUGC) according to Medialdea et al. (2004),

location of mud volcanoes after Medialdea et al. (2009), andoceanographic setting partially modified from Hernández-Molina etal. (2003)

Geo-Mar Lett (2010) 30:231–247 233

software and backscatter signal amplitudes were correctedwith Caraibes software.

Side scan sonar data were obtained by the U.S. NavalResearch Laboratory and Naval Oceanographic Office(NAVOCEANO) with the Seamap system. Furthermore,we made use of the SWIM multibeam compilationpublished by Zitellini et al. (2009) for the morphology ofthe Gulf of Cádiz.

Additional data are from gravity cores and dredges, as wellas a dense network of seismic reflection profiles collectedduring the TASYO 2000, Anastasya-99, Anastasya-00,MOUNDFORCE07 and MVSEIS08 cruises by means of amultichannel, high-resolution Sparker system and ultra-high-resolution topographic parametric sounder (TOPAS). Multi-channel seismic reflection (MCS) profiles were acquiredduring the TASYO and MOUNDFORCE07 surveys. Duringthe TASYO survey (2000), 1,728 km of MCS profiles wereobtained aboard the B/O Hespérides by means of a fiveBOLT airgun array (capacity: 22.45 and 34.8 l) and aTELEDYNE 96-channel streamer (length: 2.5 km), record-ing for 10 s at a 2-ms sampling rate with a shot interval of50 m. During the MOUNDFORCE07 survey, MCS datawere collected by means of two synchronized airgun arraysof 14–16 G.I. GUN and BOLT (total capacity: 56.5 and42.5 l), and a SERCEL 360-channel streamer (length:4.5 km) operating with 30 active sections and recording ata sampling rate of 2 ms. The shot interval was 30 s (75 m) atmaximum power and 20 s (50 m) at normal configuration.Standard data processing for both cruises was performedaboard the B/O Hespérides and at the Instituto Andaluz deCiencias de la Tierra of the Spanish National ResearchCouncil.

The Sparker system used an energy source ranging from3,500–7,000 Joules, recording length of 2 s two-way traveltime (TWT), and frequency range of 500 Hz to 4 kHz,providing medium resolution (1–10 m) and penetration of100 to 1,000 m. The TOPAS system worked with CHIRPwavelet, operating at two simultaneous primary frequenciesof 15 and 18 kHz and providing maximum penetration of100 m at 0.5–1 m resolution. Digital data were recordedusing Delph2 Triton-Elics software. Positioning was donewith a global positioning system (GPS) and a differentialGPS.

Results

Overview

Crater-like depressions are widespread in the Gulf of Cádiz,from the upper slope (cf. Baraza and Ercilla 1996) to themiddle-lower slope (cf. León et al. 2001, 2006; Somoza etal. 2002, 2003). The majority of fluid-flow depressions on

the middle continental slope are located in the TASYO fieldbetween 850 and 1,450 m water depth (Fig. 2), a vast fluid-venting area along the Iberian Margin of the Gulf of Cádizpresenting a NW-SE band of irregular relief with severalmud volcanoes, diapiric ridges and crater-like depressionssuch as pockmarks (Baraza and Ercilla 1996; León et al.2006) and collapses (Fig. 2) piercing the seafloor andconferring a smallpox-like facies. Most of the TASYO fieldis located within the theoretical limit of hydrate stabilityfield (León et al. 2009), and is bordered to the south by theCádiz Contourite Channel defined by Hernández-Molina etal. (2003) and to the north by the 35º36′N parallel, in the“mud-wave area” and “muddy sand wave area” defined byHabgood et al. (2003) in the “overflow sedimentary lobesector” (Hernández-Molina et al. 2003; Fig. 2), with a Plio-Quaternary sequence of silty sands and sandy silts (Hernández-Molina et al. 2003). The seismic profiles show goodstratification, with continuous and undulated high-amplitudereflections, patches of weak to blanking areas, and severaldiscontinuities. The contourite deposits are often deformed bydiapiric structures and pierced by diatremes and mud-volcanocones (e.g. the Hespérides, Faro and Cibeles mud volcanoes;Somoza et al. 2003).

Collapses and pockmarks are common and present innearly equal amounts. The highest density of crater-likedepressions is found in the southern part of the CádizChannel, at 1,095 to 1,137 m depth. The seafloor is spottedby several mud volcanoes and depressions, and crossed bychannels and furrows. The TASYO field is characterised byelongated and rectilinear depressions trending NE-SW,ENE-WSW and NW-SE (Figs. 2 and 3). Most of these areclosed depressions, lacking open ends.

The central sector of the TASYO field is crossed bythe Gil Eanes Channel. This sector is the deepest partof the middle slope (between 1,000 and 1,450 m depth)and forms a corridor approx. 54 km2 long and 15 km2

wide which terminates in a fan-shaped area (Fig. 2). Thedeposits correspond to sediments of the levees andsubmarine fan of the Gil Eanes Channel, part of the Gulfof Cádiz CDS (Hernández-Molina et al. 2003). Seismic

Fig. 2 Three-dimensional merged multibeam view of the TASYO 2000,MVSEIS08, MOUNDFORCE07 cruises and SWIM multibeam compi-lation (extracted from Zitellini et al. 2009), showing the location of themain fluid escape features of the TASYO field (white dotted line) incombination with swath bathymetry and side scan sonar data. In thisfigure, the spatial dimension of the fluid escape features of thesubsequent figures of the paper can be compared. The black dottedline shows the limit between the TASYO 2000 and SWIM multibeamcompilation. The side scan sonar mosaic of NAVOCEANO is showninside the black dashed line. The locations and directions of the figuresand seismic profiles in the text are marked. In the top right image, onlypockmarks with sub-circular or elliptical shapes have been plotted:circles: pockmark diameter versus pockmark depth; crosses: pockmarkdiameter versus water depth. BV Blind valley, MV Mud volcano

b

234 Geo-Mar Lett (2010) 30:231–247

Geo-Mar Lett (2010) 30:231–247 235

sections reveal dominant, transparent acoustic facies and,locally, acoustic masking probably due to gas-chargedsediments. Nonetheless, this sector is characterised byseabed fluid-flow structures such as the Aveiro mudvolcano, and freak sedimentary waves which maskelliptical and composite pockmarks (Fig. 2). A keycharacteristic of the central sector of the TASYO field(Gil Eanes levee) is that the crater-like depressions are ofsimilar size and shape (0.7 km long, 0.3 km wide and 35–55 m deep), and are more homogeneously distributed thanat other TASYO sites. Pockmarks are common and mainlyV-shaped, single elliptical features, NE-SW elongated, andconnected by sedimentary waves. By contrast, collapsesare rare.

Pockmarks

Pockmarks exist throughout the TASYO field, at waterdepths of 850 to 1,450 m. Pockmark reflectors are clearly

bent down and cut, indicating erosion due to fluidexpulsion (Figs. 4, 5 and 6).

Diameters vary widely, from 150 to 950 m butmostly between 100 and 300 m. The longest pockmarksare in the northern part of the TASYO field. Depressionsize generally correlates positively with water depth(Fig. 2): pockmarks located at water depths of 800 to950 m range from 300 to 500 m in diameter and are 20to 30 m deep, whereas those at water depths of 1,000 to1,100 m range from 500 to 900 m in diameter and are 30to 60 m deep. Locally, we detected mega-pockmarkslonger than 1 km; it is in these pockmarks that wemeasured the greatest depth between the edge and thepockmark floor (100–150 m).

The pockmarks are typically V-shaped and their slopesgenerally range from 15 to 25º (Fig. 6a). However, theirshapes vary from isolated, sub-circular forms to complexforms (where the pockmarks merge into one another);elongated asymmetrical morphologies are very common.

Fig. 3 Multichannel seismic profile belonging to line 1 of the MOUNDFORCE07 cruise, crossing the main blind valleys and pockmarks of theTASYO field. The seafloor has an “inverted relief”. Note that each depression feature is connected to a diapiric structure. BV Blind valley

236 Geo-Mar Lett (2010) 30:231–247

Isolated craters are scarce, usually appearing aligned instrings of pockmarks (Figs. 5 and 6). Pockmark-floorbackscatter varies from −17 to −19 dB, and usually has alow contrast with the surrounding backscatter of the CDS(from −20 to −21 dB). Nevertheless, the pockmarkssometimes show high backscatter (from −9 to −19 dB;Fig. 5).

In seismic profiles, elongated pockmarks longer than300 m are controlled by normal faults, generally overfocused seismic chimneys, and connected with diapiricstructures and dim zones (Figs. 3, 4 and 6). In high-resolution (TOPAS) and Sparker seismic profiles, thereflectors are clearly bent down and cut along focused fluidpathways (Fig. 6a). In multichannel seismic profiles, thepockmarks are located over focused seismic chimneysfollowing fault-like pathways, and are connected withdiapirs and/or areas of weak amplitude (dim zones; Figs. 4,6b and 7).

The pockmarks in the central sector of the TASYO fieldare mainly V-shaped, comprising single elliptical unitstrending NE-SW and connected by sedimentary waves. Theinterference between pockmarks and sedimentary wavesgives the seafloor an aspect of freak sedimentary waves.The composite pockmarks are generally formed by two

merged pockmarks. The pockmark flanks are sharp, with amean slope range of 7 to 14º, and maximum values of 22º.Backscatter values at the floor range from −10 to −15 dB(similar to those of the Aveiro mud volcano), contrastingwith a value of approx. −21 dB for the surrounding area.Locally, we detected high reflectivity values under cratersin TOPAS profiles (Fig. 8).

Collapses

Collapses are depressions delimited by normal faults.Craters of this type have been detected in the TASYOfield, associated with mud-volcano complexes (Somoza etal. 2003). Nevertheless, we also detected isolated, collapseddepressions disconnected from mud-extrusion structures.These vary greatly in size and shape, ranging from 500 m to5 km in diameter and from 20 to 248 m in depth, and withperimeter shapes ranging from isolated, sub-circular formsto complex, elongated forms (Figs. 2 and 4). Elliptical andelongated collapses are very common, often have oneopened flank, and never exceed 2 km in length. Theisolated, lobate or elliptical collapses are 50 to 80 m deepand have a width/length ratio (w/l) of less than 0.15. Theirlong axes range from 0.8 to 1.5 km, and their short axesfrom 0.3 to 0.5 km. The profiles vary from U-shaped toterraced, the former being more frequent. The slopes arealso highly variable, raging from 2 to 6º in smooth profilesand from 12 to 35º in sharp ones, with numerousgravitational instabilities.

In Sparker seismic profiles (Fig. 4), the sub-bottomreflectors appear clearly cut by normal faults, and chaoticand blanking acoustic facies are related to the depositsaffected by the gravitational processes. Generally, aninternal mound and/or a very irregular relief are foundinside the collapse. In multichannel seismic profiles, thecollapses are connected to diapiric structures via focusedseismic chimneys.

Some of the lobate structures appear lined up with othercollapses, forming a collective string. We detected locallyhuge, elongated collapsed depressions (ca. 10 km long),related to other fluid-flow structures such as pockmarks,carbonate-mud mounds and mud volcanoes. Most of thesecurrently serve as furrows and channel the MOW. Consid-ering their morphology, we have coined the term blindvalleys (BVs) to describe them (Figs. 2 and 5).

Blind valleys

We consider the blind valleys to be complex, collapsedand elongated mega-structures (up to 16 km long and0.5–2 km wide; Figs. 6, 8 and 9) typified by diversegeological features (e.g. pockmarks, collapses and slumps)and processes (e.g. gas erosion, mass wasting, gravita-

Fig. 4 Different types of crater-like depressions from the north sectorof the study area. The top left seismic profile belongs to line 28(Sparker) of the Anastasya 00/09 cruise. The top right multichannelseismic profile belongs to line 1 of the MOUNDFORCE07 cruise

Geo-Mar Lett (2010) 30:231–247 237

tional collapses, precipitation of authigenic carbonates,diapirism, and mud volcanism) related to seabed fluidflow. The BVs are gravitational features defined bycollapse structures (i.e. they start and finish in the flankof a collapse, thereby lacking an exit) and generallysurrounded by pockmarks, mounds or mud volcanoes(Fig. 5). Multichannel seismic profiles show that they

develop over normal faults formed at the crests of diapirs(Fig. 3). Seismic chimneys follow linear pathways alongnormal faults. Bright spots and dim zones over diapirsindicate the presence of fluids and of a focused fluid flowbelow the blind valleys (Figs. 3 and 8).

We characterised three main BVs in the TASYO field inthe overflow sedimentary lobe sector of the CDS: the

Fig. 5 Backscatter image and geomorphologic map of the vicinity of the Aquiles blind valley. Note the many mounds and pockmarks withmoderate-high backscatter close to the blind valley. MV Mud volcano

238 Geo-Mar Lett (2010) 30:231–247

Aquiles BV, Esperillas BV and Caleton BV, whichcorrespond to the large isolated furrows A1, A2 and A4respectively described by Habgood et al. (2003) andHanquiez et al. (2007).

The Aquiles BV is the most spectacular structure whichwe detected in the TASYO field (Figs. 3 and 5), anelongated collapse 15 km long, 1.5 km wide (w/l ratio<0.15) and with a maximum depth of 112 m. It trends NE-SW (N60E), approx. parallel to the Cádiz ContouriteChannel and the Gil Eanes Channel. The bottom isgenerally irregular and rough but flat in the widest areas.Its seafloor backscatter values are moderate and uniform(from −16 to −20 dB). The slope values vary strongly, witha mean range of 8 to 30º and a maximum value of 33º. Theextremes of the valley are surrounded by pockmarks andcollapses of smoother slopes (from 10 to 16º) and locallyhigh backscatter (from −7 to −12 dB). We detected minor

slumps on the southern flank, which confer it with aterraced, wrinkled and mounded morphology at its bottom(Fig. 5).

The Esperillas BV is an ENE-WSW trending (N82E)narrow valley which is 11 km long and 0.5 km wide, with amaximum depth of 50 m (Fig. 3). Its bottom is flat, exceptin three areas affected by sub-circular depressions of0.4 km2. It has a mean slope range of 6 to 9º, with amaximum value of 12º. As in the case of the Aquiles BV,both ends of the Esperillas BV are surrounded by an area ofpockmarks and collapses with locally high backscatter(from −8 to −14 dB).

The Caleton BV (Figs. 3, 8 and 9) is a NW-SE trending(N65W), complex collapsed structure which is 14 km longand lentil-shaped, its width varying from 1.7 km (in themiddle) to 300 m (at either end). The floor is shallow (meandepth: 15 to 20 m) and very irregular, due to collapses andpockmarks. We detected spots of high backscatter in somepockmarks (Fig. 8). This blind valley has a mean slopeangle of 7 to 10º, with a maximum value of 16º on thesouthwest flank. This flank follows a straight line, whereasthe opposite flank is less steep and more irregular, beinginterrupted by pockmarks, sedimentary waves and terraces.We also observed sedimentary waves beyond and around

Fig. 7 Multichannel seismic line of TASYO 8 at the western end ofthe Aquiles blind valley (see Fig. 2), showing pockmarks, furrows andblind valleys related with diapiric structures. Normal faults abovediapirs are related to fluid-flow pathways with dim zones and brightspots

Fig. 6 Different types of crater-like depressions in the vicinity of theAquiles blind valley. a Line 30 (TOPAS) of the TASYO 2000 cruiseacross the Aquiles blind valley. Note the U- and V-shaped morphol-ogies of the pockmarks. b Multichannel seismic profile belonging toline 1 of the MOUNDFORCE07 cruise showing fault-strike pock-marks connected to a diapir structure

Geo-Mar Lett (2010) 30:231–247 239

the valley, nucleated from pockmarks and collapses (someof which show high backscatter), masking them andgenerating freak sedimentary waves along sinuous stripesof low backscatter (Fig. 8).

Discussion

Role of tectonic and fluid-flow activity

Several authors have characterised large isolated furrowswithin the overflow sedimentary lobe sector of the Gulf ofCádiz Contourite Depositional System (Kenyon and Belderson

1973; Habgood et al. 2003; Hernández-Molina et al. 2003,2006; Mulder et al. 2003; Hanquiez et al. 2007; Llave et al.2007; García et al. 2009). These have interpreted furrows aserosive features caused principally by the MOW, and alsoaffected by gravitational processes. The relatively smalldimensions of the furrows indicate that they have beeneroded by flow filaments separate from the main current ofthe MOW. These filaments flow SW at current velocitieswhich decrease from 20 to ca. 10 cm/s, although furrowerosion must have occurred during periods of greater velocity(García et al. 2009). We contend that other factors must beconsidered when trying to establish how these depressionsdeveloped, such as the regional tectonics, oceanographic

Fig. 8 Morpho-structure of theCaleton blind valley. Top leftHill-shaded model of the sea-floor; middle backscatter mosa-ic; right morphological schema:blind valley (light grey), highbackscatter areas (dark grey),sedimentary waves (dotted line),pockmarks (continuous line).Bottom a Line 23 (TOPAS) ofthe TASYO 2000 cruise: dunesand mega-ripples nucleated fromcrater-like depressions of thenorth sector; b line 25 (Sparker)of the Anastasya 00/09 cruise

240 Geo-Mar Lett (2010) 30:231–247

setting and gas-hydrate stability, each of which could havecontributed to a different extent.

In the seismic profiles, pockmarks, collapses and blindvalleys have been observed above normal faults related todiapiric anticlines. These fault-strike features trend NW-SE,NE-SW and ENE-WSW. The NE-SW trend (Aquiles BV)could be related to diapiric anticline thrusts generated by theregional NW-SE compression (e.g. Guadalquivir DiapiricRidge and Cadiz Diapiric Ridge; see Fig. 2). Furthermore,the NW-SE trend (Caleton BV) could be related to normalfaults created by regional compression in the same direction.Finally, the ENE-WSW trend could be related to trans-pressive sinistral strike-slip faults and consequent diapiricdevelopment. However, all of these fault-strike features arelocated above the focused seismic chimneys which followlinear pathways, with bright spots and dim zones, along thenormal faults of diapiric anticlines. These seismic character-istics suggest that a focused fluid flow may have been activeunder the blind valleys, at least until very recent times.

Interestingly, the high backscatter values (from −9 to−19 dB) recorded above several pockmarks and collapses,and which are close to those recorded in mud volcanoes orcarbonate-mud mounds (from −6 to −10 dB) such asHespérides, Faro and Hormigas Ridge, could be related tothe precipitation of HDACs and to fluid flow which wasactive until recent times. In fact, the Hespérides, Faro andAlmazán mud volcanoes (see Figs. 2 and 5) are locatedabove the same diapiric anticlines as the Aquiles, Esperilasand Caleton BVs, but are located several kilometres awayfrom these. Therefore, indicators of upward fluid flux,HDACs and chemosynthetic fauna recovered in mudvolcanoes (Díaz-del-Río et al. 2003; Somoza et al. 2003;

Niemann et al. 2006) could be related to possible extrusionsfrom the seismic chimneys below the BVs. In the Faro mudvolcano, small patches of reduced sediments covered bygiant sulphide-oxidizing bacteria have been identified,indicating current anaerobic oxidation of methane andconsumption of hydrocarbons located in near-surface sedi-ments (Niemann et al. 2006). The gas chromatograph ofbitumen extracted from interstitial fluids in gravity coresand mineral precipitates in the Guadalquivir Diapiric Ridge(González 2008) revealed a high input of thermogenicallyformed hydrocarbons (e.g. phenanthrene and mature hydro-carbons, Ro<1.2%, derived from types II and III kerogen)from deep sources. Moreover, the 87Sr/86Sr ratios(0.70693–0.70993) and δ7Li isotopic values (as low as+11.9‰) obtained in Fe-Mn nodules and interstitial watersfrom local mud-carbonate mounds and mud volcanoes arein agreement with deep-rooted fluid migration from thereservoirs to the sea bottom, across the fault systems wherediagenetic and hydrothermal processes have occurred(Hensen et al. 2007; González 2008; Scholz et al. 2009).

Based on the descriptions of the gassy seismic faciescharacteristics and the tectonic structure (see Figs. 3, 5 and8) of some of these large isolated furrows (A1, A2 and A4)by Habgood et al. (2003) and Hanquiez et al. (2007), wepropose that these features—and, consequently, the BVs—arose from focused fluid flow from diapiric structures to theseafloor via seismic chimneys along faults. During thesubsequent evolution of the BVs, the constant influence ofthe MOW favoured their incision, thereby dictating themorphology of the furrows up to present times.

Therefore, the two interpretations (blind valleys andfurrows) are compatible. A blind valley would refer to theonset of the feature due to internal processes (i.e. faultsand fluid migration), whereas a furrow would refer to themorphology generated by the influence of bottom currentson the seafloor which, in the Gulf of Cádiz, is due toMOW circulation. Moreover, both BVs and furrows arefeasible within a complex system of interaction of warmbottom current (i.e. MOW) and fluid-venting areas.Herein, we propose a new origin for the onset ofdepressions which currently channel the MOW in theGulf of Cádiz. Bottom-current incision along a blindvalley might eventually open the lower end. Moreover,each of the two ends of a single blind valley could beopened and connected to others features such as channels.In this case, the lower end of the resulting compositefeature would then contain sedimentary structures relatedto submarine bottom-current activity.

Comparison with elongated structures in other settings

Different processes have also been inferred in the formationof similar closed, elongated depression-like pockmarks in

Fig. 9 Multichannel seismic line 10 of TASYO 2000 crossing theCaleton blind valley (see Fig. 2). Note that the blind valley is above adiapir structure. Bright spots, collapses and dim zones are related tofluid flow from the diapir to the seafloor

Geo-Mar Lett (2010) 30:231–247 241

other geological settings. The Cascadia and Hispaniolamargins are venting areas with several structures related tofocused fluid flow (Carson et al. 1990). Headless canyonsalong the Southern Cascadia Margin and headless gulliesalong the Northern Hispaniola Margin have been related toerosional processes due to incision of the seepage extru-sions downwards along the slope (Orange and Breen 1992).In Monterey Bay, isolated pockmarks forming elongatedstructures and headless canyons have been reported to havea similar origin (Orange et al. 1999, 2002). Along theequatorial West African continental margin, three types ofelongated depression-like mega-pockmarks have beenobserved. One comprises pockmarks and furrows whichare preferentially fed through a polygonal faults system.The pockmarks are located above a triple junction of faults,whereas the furrows open in the direction of the first-orderfaults of a hierarchical faults system, due to horizontalsliding of the sedimentary cover. Progressive loading ofsediments and compactional dewatering processes arerelated to the formation of the polygonal faults-furrowssystem (Gay et al. 2004). A second type comprisespockmark trains and gullies. Elongated mega-pockmarkshave been observed on the continental slope in areas ofslope instability and are associated with listric slump faultswhich root into parallel multiple-bedded detachment surfa-ces. Pockmarks are formed in the hanging wall of faults,and not at the surface scarp (Pilcher and Argent 2007). Athird type comprises pockmark meandering trains orsinuous pockmark belts developed over fluid-venting areasrelated to gas-bearing sediments in palaeochannel tracks(Gay et al. 2003).

None of the aforementioned fluid-flow features cancompletely describe the BVs of the Gulf of Cádiz,despite their similarities to these. Blind valleys are fault-strike features and are collapsed structures; they can notbe classified as erosive features resulting from fluidflow, such as pockmarks and headless canyons. Rather,they constitute a new type of complex, collapsed fluid-flow structure developed above diapiric anticlines,characterised by diapiric collapses along normal faults,and the occurrence of pockmarks, slumps and localHDAC precipitation.

Role of the contourite depositional system

The central sector of the TASYO field is covered bysandy-muddy deposits (Kenyon and Belderson 1973;Habgood et al. 2003; Mulder et al. 2003; Hanquiez et al.2007) located over the “muddy sand wave area (mediumvelocity zone)” (Habgood et al. 2003) developed in the“overflow sedimentary lobe sector” (Hernández-Molina etal. 2003, 2006), composed of medium- to fine-grainedsand which is moderately to well sorted, with up to 20%

silt and locally bioturbated clay-sized material, olive greysandy silt and clayey silt (Habgood et al. 2003). Numerousseepage morphologies are masked by these deposits, andcrater-like depressions nucleate sets of sedimentary waves.These could explain why the central sector containshomogeneous smallpox-like facies and lacks elongatedcollapses.

Nevertheless, this homogeneous morphology could beexplained by other tectonics and sedimentological factors.Diapiric anticlines have been detected in multichannel seismicprofiles, below the rectilinear tracks of the Gil Eanes Channel.Sediments from the central sector of the TASYO field couldcontain thermogenic gas-bearing sediments stemming fromvertical migrations from diapirs along normal faults (e.g.lateral migration from the Aveiro mud volcano; see Fig. 2).Sandy and muddy sediments could function as a gas trap,and lateral migrations could homogenise the gas in sedi-ments, transforming from focused, linear venting to asomewhat dispersed seabed fluid flow. This would in turnexplain why isolated lobate, crater-like depressions occurmore frequently than do more elongated depressions.

Dissociation of natural hydrates and influence of warmbottom currents

Natural gas and hydrates are known to exist in the Gulf ofCádiz (Kenyon et al. 2000; Mazurenko et al. 2002);however, there is controversy over their contribution tothe evolution and generation of pockmarks, as well as tothat of other structures related to hydrocarbon fluid venting(e.g. mud volcanoes and HDACs; León et al. 2006).Hydrates have been recovered only over mud volcanoes:disperse ones, in the Ginsburg mud volcano (Kenyon et al.2000), and laminar ones in the Bojardim mud volcano(Kenyon et al. 2001; Kopf and Cruise Participants 2004).Furthermore, no hydrates have been detected at any othersite under the continental slope of the Gulf of Cádiz.Moreover, in the Gulf of Cádiz the bottom simulatingreflector level is not continuous; along the MoroccanMargin, it has been observed only below mud volcanoes(Depreiter et al. 2005) and, along the Iberian Margin, it isonly semi-continuous below the upper continental slope(Somoza et al. 2000; Casas et al. 2003).

Nevertheless, blind valleys are areas of focused hydro-carbon fluid flow, where thermogenic gas migration ispiped by the main network faulting. The middle and lowercontinental slope is under hydrate stability field conditionsand some geological structures associated with hydrocarbonfluid emissions match with the theoretical base of thehydrate stability zone (HSZ), such as the bottom level ofsome collapses and pockmarks (León et al. 2006). TheCaleton BV is over a collapse structure rooted at thetheoretical base of the HSZ (León et al. 2009). This

242 Geo-Mar Lett (2010) 30:231–247

structure could be related to a massive dissociation ofhydrates. Indeed, the Caleton BV is on the track of animportant fault and is affected by a warm undercurrent.

Earthquakes and maximum glacials (subsequent changeof sea level and location of the MOW) could be consideredas the main triggers of massive hydrate dissociation in theTASYO field. Dissociation effects in the seafloor wouldinduce a network collapse of marked linear characterbecause (1) the main fluid-flow pathways are normal faultsrelated to diapirs; (2) tectonic reactivation of diapirs wouldinduce changes of pressure in the normal faults; and (3) thelocal channelling of filaments of the MOW along blindvalleys could accelerate the fluid flow and the collapseprocesses, due to warming of the seafloor and subsequenthydrate dissociation.

Model for the evolutionary origin of blind valleys

Geological study of the crater-like fluid-flow structures inthe Gulf of Cádiz enables establishment of an evolutionary

model for the generation of blind valleys. Herein wepropose a hierarchical description of the structures andprocesses active at different stages over the course of time,showing the function of fluid flow through the migrationpathways created by diapir activity. The structures varyfrom simple (e.g. conic or isolated pockmarks) to complex(e.g. blind valleys).

We envision three stages in the generation of the BVs:onset, growth and maturation (Fig. 10).

Onset

The onset stage (Fig. 10a) is characterised by thedevelopment of diapiric anticlines and normal faults.Seepage structures are little evolved and located oversemi-continuous focused fluid-flow pathways. These path-ways are driven by normal faults, and single V-shapedfault-strike features, such as isolated pockmarks, arecommon. Sometimes they can even feature strings. Diapiricdevelopment could create an elongated seafloor ridge. This

Fig. 10 Model proposed for the formation of blind valleys in the Gulfof Cádiz. a Onset stage: development of normal faults, and generationof isolated pockmarks and strings of fault-strike pockmarks. b Growthstage: growth of pockmarks, generating elongated shapes, andformation of isolated collapsed structures by lack of mass due to

fluid migration. c Mature stage: appearance of huge, elongated,collapsed areas, and formation of blind valleys. MediterraneanOutflow Water could flow along the blind valley and remodel into afurrow feature

Geo-Mar Lett (2010) 30:231–247 243

stage represents the start of the interaction and progressiveinfluence of Mediterranean Outflow Water with the seepagestructures.

Growth

The growth stage (Fig. 10b) is characterised by late growthof the fluid-flow structures along normal faults, andconsolidation of continuous, planar, focused fluid-flowpathways. The fluid-flow pathways appear on seismicprofiles as chaotic facies of seismic chimneys. Fluidpathways have local bright spots and dim zones and areconnected with diapir anticlines. Their most commonfeatures are elongated, complex mega-pockmarks andisolated collapses. Continuous fluid escape through themigration pathways increases the diameter of the crater-like depressions, and consequent merging with adjacentpockmarks generates elongated shapes. Isolated collapseprocesses occur at this stage due to the lack of massbelow the seepages, distension from diapir activity, andsubsequent tectonic reactivation of faults. Thus, pock-marks and collapses tend to align in the principal faultdirection, generating elongated, asymmetric, crater-likedepressions which encourage the flow of MediterraneanOutflow Water filaments. Moreover, the incision ofwarm bottom currents, such as Mediterranean OutflowWater, could contribute to the extension of elongatedcrater-like depressions with an open end. Furthermore,warming of the seafloor by bottom currents couldgenerate massive dissociations and accelerate the col-lapse processes.

Surficial slides and slumps are also crucial to thedevelopment of pockmarks and collapse processes.Rotated blocks sometimes develop along listric faultsrooted in the depression bottom and, therefore, give riseto features of various morphologies. These featureschiefly show pail or terraced profiles with irregular floorsand complex contours related to the merging of neigh-bouring craters; normal faulting associated with gravita-tional collapses and slumps; outcropping of a small diapiror mud volcano (Somoza et al. 2003; Fernández-Puga2004; León et al. 2006); and precipitation of authigeniccarbonates (Díaz-del-Río et al. 2003; Magalhães 2007;León et al. 2007).

Maturation

In the mature stage (Fig. 10c), the seafloor is affected bywidespread collapse processes, and is characterised by thegeneration of blind valleys. The fluid-flow pathways coverthe wide, extended area affected by the collapse, andcomprise a huge, elongated seafloor which is usuallyaffected by seepages. At the ends of the BVs, isolated

irregular depressions are very common. Moreover, in thesurroundings of the BVs, other seabed fluid-flow processes(e.g. mud volcanism or precipitation of HDACs) mayoccur.

In this stage, bottom currents could be channelled alongthe blind valleys. Thus, the behaviour as a furrow couldmodify the morphology of the valley and, by incision,would eventually open one end of a blind valley or connectit to other deep-water channels. In this case, seafloorstructures related to fluid flow would be partially maskedor deleted by the bottom-current activity.

Conclusions

The TASYO field is an area of intense seepages featuringseveral structures related to seabed fluid flow. Some ofthese, such as collapses and blind valleys, have beendescribed in this paper for the first time.

In the Gulf of Cádiz, pockmarks and collapses occur inhighest densities in the TASYO field, between 700 and1,400 m water depth. Furthermore, the diameter of thecrater-like depressions is positively correlated to waterdepth. Collapses are depressions which are limited bynormal faults, generally isolated and of lobate or ellipticalshape. They vary from 500 m to 5 km in diameter and from20 to 248 m in depth. Blind valleys are very elongated,complex, collapsed mega-structures ca.10 km in length.They are characterised by a lack of exits, and are limited bycollapse structures (i.e. they begin and end in the flank of acollapse), the occurrence of these structures being related toseabed fluid flow.

The distal contourite lobe of the Gil Eanes Channelaccounts for the singular morphology of the central sectorof the TASYO field.

Blind valleys develop above normal faults of diapirs.They represent the final stage in the evolution of a fluid-venting area controlled by linear fluid pathways. Thesepathways are driven directly by collapse processes causedby fluid escape through migration pathways which, in turn,are created by the distension resulting from diapir activityor from late tectonic reactivation of these diapirs.

Strings of pockmarks and collapses, blind valleys andnumerous minor deep-water channels, furrows and elon-gated depressions detected in the study area are alwaysassociated to (1) structural control of features along fault asfocused seismic chimneys, (2) the presence of diapiricstructures and (3) Mediterranean Outflow Water interaction(warming and erosion). Pockmarks, collapses and blindvalleys are strike-fault features related to normal faultsconnected to diapir anticlines. Focused seismic chimneyswith bright spots and dim zones follow linear pathwaysalong normal faults.

244 Geo-Mar Lett (2010) 30:231–247

In the Gulf of Cádiz, seabed fluid flow and gravitationalcollapses are crucial mechanisms for the onset of channel-ling bottom currents. Although the Caleton, Esperillas andAquiles blind valleys channel Mediterranean OutflowWater and function as furrows, their origin is linked togravitational collapse caused by seabed fluid flow.

Acknowledgements We thank Aurélien Gay and one anonymousreviewer for comments and suggestions which have improved thepaper. We are especially grateful to Cristino J. Dabrio González forsuggestions and support during the development of the concept ofblind valleys. We also thank Alan Judd for his input, the InstitutoEspañol de Oceanografía for its exemplary and productive collabora-tion, all those who participated in the research cruises of the TASYO2000 and MOUNDFORCE07 projects, as well as the captains,technicians and crews of the research vessels B/O Cornide deSaavedra, B/O Hespérides and R/V L’Atalante. This research wasfunded by the Spanish Marine Science and Technology Program underthe TASYO Project (MAR 98-0209) and the ESF EuroCORE-EuroMARGINS projects MVSEIS (O1-LEC-EMA24F, REN2002-11669-E-MAR) and MOUNDFORCE (O1-LEC-EMA06F,REN2002-11668-E-MAR), and is currently supported by the SpanishComisión Interministerial de Ciencia y Tecnología (CYCIT) projectCTM2008-06399-C04/MAR (CONTOURIBER).

References

Baraza J, Ercilla G (1996) Gas-charged sediments and largepockmark-like features on the Gulf of Cadiz slope (SW Spain).Mar Petrol Geol 13(2):253–261

Blinova V, Stadnitskaia A (2001) Composition and origin ofhydrocarbon gases from the Gulf of Cadiz mud volcanic area.IOC Worksh Rep 175:45–46

Bøe R, Rise L, Ottesen D (1998) Elongate depressions on the southernslope of the Norwegian trench (Skagerrak): morphology andevolution. Mar Geol 146:191–203

Carson B, Suess E, Strasser JC (1990) Fluid flow and mass fluxdeterminations at vent sites on the Cascadia margin accretionaryprism. J Geophys Res 95(B6):8891–8897

Casas D, Ercilla G, Baraza J (2003) Acoustic evidences of gas in thecontinental slope sediments of the Gula of Cadiz (E Atlantic). In:Woodside JM, Garrison RE, Moore JC, Kvenvolden KA (eds)Proc 7th Int Conf Gas in Marine Sediments, 7–12 October 2002,Baku, Azerbaijan. Geo-Mar Lett 23(3/4):300–310. doi:10.1007/s00367-003-0160-z

Depreiter D, Poort J, Van Rensbergen P, Henriet J (2005) Geophysicalevidence of gas hydrates in shallow submarine mud volcanoes onthe Moroccan margin. J Geophys Res 110:B10103. doi:10.1029/2005JB003622

Díaz-del-Río V, Somoza L, Martínez-Frías J, Mata MP, Delgado A,Hernández-Molina FJ, Lunar R, Martín-Rubí JA, Maestro A,Fernández-Puga MC, León R, Llave E, Medialdea T, Vázquez JT(2003) Vast fields of hydrocarbon-derived carbonate chimneysrelated to the accretionary wedge/olistostrome of the Gulf ofCadiz. Mar Geol 195:177–200

Eichhubl P, Greene HG, Naehr T, Maher N (2000) Structural controlof fluid flow: offshore seepage in the Santa Barbara Basin,California. J Geochem Explor 69(70):545–549

Fernández-Puga MC (2004) Diapirismo y estructuras de expulsión degases hidrocarburos en el talud continental del Golfo de Cádiz.PhD Thesis, Universidad de Cádiz

Fernández-Puga MC, Váquez JT, Somoza L, Díaz del Río V,Medialdea T, Mata MP, León R (2007) Gas-related morphologiesand diapirism in the Gulf of Cadiz. In: García-Gil S, Judd A (eds)Contrib 8th Int Conf Gas in Marine Sediments, 5–10 September2005, Vigo. Geo-Mar Lett 27(2/4):213–221. doi:10.1007/s00367-007-0076-0

García M, Hernández-Molina FJ, Llave E, Stow DAV, León R,Fernández-Puga MC, Diaz del Río V, Somoza L (2009)Contourite erosive features caused by the Mediterranean OutflowWater in the Gulf of Cadiz: quaternary tectonic and oceano-graphic implications. Mar Geol 257(1/4):24–40. doi:10.1016/j.margeo.2008.10.009

Gardner JM (2001) Mud volcanoes revealed and sampled on the WesternMoroccan continental margin. Geophys Res Lett 28(2):339–342

Gay A, Lopez M, Cochonat P, Sultan N, Cauquil E, Brigaud F (2003)Sinuous pockmark belt as indicator of a shallow buried turbiditicchannel on the lower slope of the Congo Basin, West AfricanMargin. In: Van Rensbergen P, Hillis RR, Maltman AJ, MorleyCK (eds) Subsurface sediment mobilization. Geol Soc Lond SpecPubl 216:173–189

Gay A, Lopez M, Cochonat P, Sermondadaz G (2004) Polygonalfaults-furrows system related to early stages of compaction—Upper Miocene to present sediments of the Lower Congo Basin.Basin Res 16:101–116

González FJ (2008) Fe-Mn nodules and crusts in the Gulf of Cadizand Antarctica: genesis and paleoceanographic implications (inSpanish). PhD Thesis, Complutense University of Madrid

González FJ, Somoza L, Lunar R, Martínez-Frías J, Martín Rubí JA,Torres T, Ortiz JE, Díaz-del-Río V, Pinheiro LM, Magalhães VH(2009) Hydrocarbon-derived ferromanganese nodules incarbonate-mud mounds from the Gulf of Cadiz: mud-brecciasediments and clasts as nucleation sites. Mar Geol 261:64–81

Habgood EL, Kenyon NH, Masson DG, Akhmetzhanov A, WeaverPPE, Gardner J, Mulder T (2003) Deep-water sediment wavefields, bottom current sand channels and gravity flow channel-lobesystems: Gulf of Cadiz, NE Atlantic. Sedimentology 50:483–510

Hanquiez V, Mulder T, Lecroart P, Gonthier E, Marchès E, Voisset M(2007) High resolution seafloor images in the Gulf of Cadiz,Iberian margin. Mar Geol 246:42–59

Heezen BC, Johnson GL (1969) Mediterranean undercurrent andmicrophysiography west of Gibraltar. Bull Inst OcéanogrMonaco 67(1382):1–95

Heggland R (1998) Gas seepage as an indicator of deeper prospectivereservoirs. A study based on exploration 3D seismic data. MarPetrol Geol 15:1–9

Hensen C, Nuzzo M, Hornibrook E, Pinheiro LM, Bock B, MagalhãesVH, Brückmann W (2007) Sources of mud volcano fluids in theGulf of Cadiz-indications for hydrothermal imprint. GeochimCosmochim Acta 71(5):1232–1248

Hernández-Molina FJ, Llave E, Somoza L, Fernández-Puga MC,Maestro A, León R, Medialdea T, Barnolas A, García M,Díaz-del-Río V, Fernández-Salas LM, Vázquez JT, Lobo FJ,Alveirinho-Dias JA, Rodero J, Gardner J (2003) Looking forclues to paleoceanographic imprints: a diagnosis of the Gulfof Cadiz contourite depositional systems. Geology 31(1):19–22

Hernández-Molina FJ, Llave E, Stow DAV, García M, Somoza L,Vázquez JT, Lobo FJ, Maestro A, Díaz del Río V, León R,Medialdea T, Gardner J (2006) The contourite depositionalsystem of the Gulf of Cadiz: a sedimentary model related to thebottom current activity of the Mediterranean outflow water andits interaction with the continental margin. Deep-Sea Res II53:1420–1463

Hovland M, Judd A (1988) Seabed pockmarks and seepages. Impactin geology, biology and the marine environment. Graham andTrotman, London

Geo-Mar Lett (2010) 30:231–247 245

Hovland M, Svensen H, Forsberg CF, Johansen H, Fichler C, FossaJH, Jonsson R, Rueslatten H (2005) Complex pockmarks withcarbonate-ridges off mid-Norway: products of sediment degass-ing. Mar Geol 218:191–206

Ivanov MK, Kenyon N, Nielsen T, Wheeler A, Monteiro H, GardnerJ, Comas M, Akhmanov A, Akhmetzhanov G (2000) Goals andprincipal results of the TTR-9 cruise. IOC/UNESCO Worksh Rep168:3–4

Judd A, Hovland M (2007) Seabed fluid flow. The impact on geology,biology and the marine environment. Cambridge University,Cambridge

Kelley JT, Dickson SM, Belknap DF, Barnhardt WA, Henderson M(1994) Giant sea-bed pockmarks: evidence for gas escape fromBelfast Bay. Mar Geol 22(1):59–62

Kenyon NH, Belderson RH (1973) Bed forms of the Mediterraneanundercurrent observed with side-scan sonar. Sediment Geol 9:77–99

Kenyon NH, Ivanov MK, Akhmetzhanov AM, Akhmanov GG (2000)Multidisciplinary study of geological processes on the North EastAtlantic and Western Mediterranean Margins. (Preliminaryresults of geological and geophysical investigations duringTTR-9 cruise of R/V Professor Logachev, June-July 1999).IOC/UNESCO Tech Ser 56

Kenyon NH, Ivanov MK, Akhmetzhanov AM, Akhmanov GG (2001)Interdisciplinary approaches to geoscience on the North EastAtlantic and Mid-Atlantic Ridge. IOC/UNESCO Tech Ser 60

King LH, MacLean B (1970) Pockmarks on the Scotian Shelf. GeolSoc Am Bull 81(10):3141–3148

Kopf A, Cruise Participants (2004) Report and preliminary results ofSONNE Cruise SO175, Miami-Bremerhaven, 12.11-30.12.2003.Berichte, Fachbereich Geowissenschaften, Universität Bremen,no 228

León R, Somoza L, Ivanov MK, Díaz-del-Río V, Lobato A,Hernández-Molina FJ, Fernández-Puga MC, Maestro A,Medialdea T, Alveirinho-Dias JM, Vázquez T (2001) Seabedmorphology and gas venting in the Gulf of Cadiz mudvolcanoarea: imagery of multibeam data and ultra-high resolutiondata. In: Akhmanov G, Suzyumov A (eds) Geologicalprocesses on deep-water European margins. IOC Worksh Rep175:43–45

León R, Somoza L, Medialdea T, Maestro A, Díaz-del-Río V,Fernández-Puga MC (2006) Classification of sea-floor featuresassociated with methane seeps along the Gulf of Cádizcontinental margin. Deep-Sea Res II 53:1464–1481

León R, Somoza L, Medialdea T, González FJ, Díaz-del-Río V,Fernández-Puga MC, Maestro A, Mata MP (2007) Sea-floorfeatures related to hydrocarbon seeps in deepwater carbonate-mud mounds of the Gulf of Cádiz: from mud flows to carbonateprecipitates. In: García-Gil S, Judd A (eds) Contrib 8th Int ConfGas in Marine Sediments, 5–10 September 2005, Vigo. Geo-MarLett 27(2/4):237–247. doi:10.1007/s00367-007-0074-2

León R, Somoza L, Gimenez-Moreno CJ, Dabrio CJ, Ercilla G,Praeg D, Díaz-del-Río V, Gomez-Delgado M (2009) Apredictive numerical model for potential mapping of gashydrate stability zone in the Gulf of Cadiz. Mar Petrol Geol26:1564–1579

Llave E, Hernández-Molina FJ, Somoza L, Díaz-del-Río V, StowDAV, Maestro A, Alveirinho Dias JM (2001) Seismic stackingpattern of the Faro-Albufeira contourite system (Gulf of Cadiz): aQuaternary record of paleoceanographic and tectonic influences.Mar Geophys Res 22(5/6):475–496

Llave E, Hernández-Molina FJ, Somoza L, Stow D, Díaz del Río V(2007) Quaternary evolution of the Contourite DepositionalSystem in the Gulf of Cádiz. In: Viana A, Rebesco M (eds)Economic and paleoceanographic importance of contourites.Geol Soc Lond Spec Publ 276:49–79

Lowrie A, Hamite R, Moffett S, Somoza L, Maestro A, Lerche I(1999) Potential pressure compartments sub-salt in the Gulf ofMexico and beneath massive debris flows in the Gulf of Cadiz.In: 19th Annu Bob. F. Perkins Conf Abstr, Advanced ReservoirCharacterization for the 21st Century. GCSSEPM Foundation,Houston, TX, pp 271–280

Magalhães VH (2007) Authigenic carbonates and fluid escapestructures in the Gulf of Cadiz. PhD Thesis, University of Aveiro

Maldonado A, Nelson CH (1999) Interaction of tectonic anddepositional processes that control the evolution of the IberianGulf of Cadiz margin. Mar Geol 155:217–242

Maldonado A, Somoza L, Pallarés L (1999) The Betic orogen and theIberian-African boundary in the Gulf of Cadiz: geologicalevolution (central North Atlantic). Mar Geol 155:9–43

Mazurenko LL, Soloviev VA, Belenkaya I, Ivanov MK, Pinheiro L(2002) Mud volcano gas hydrates in the Gulf of Cadiz. TerraNova 14:321–329

Medialdea T, Vegas R, Somoza L, Vázquez JT, Maldonado A, Díaz-del-Río V, Maestro A, Córdoba D, Fernández-Puga MC (2004)Structure and evolution of the “Olistostrome” complex of theGibraltar Arc in the Gulf of Cadiz (eastern Central Atlantic):evidence from two long seismic cross-sections. Mar Geol 209(1/4):173–198

Medialdea T, Somoza L, Pinheiro LM, Fernández-Puga MC, VázquezJT, León R, Ivanov MK, Magalhaes V, Díaz-del-Río V, Vegas R(2009) Tectonics and mud volcano development in the Gulf ofCádiz. Mar Geol 261:48–63

Mulder T, Voisset M, Lecroart P, Le Drezen E, Gonthier E, HanquiezV, Faugères JC, Habgood E, Hernández-Molina FJ, Estrada F,Llave-Barranco E, Poirier D, Gorine C, Fuchey Y, Voelker A,Freitas P, Lobo Sanchez F, Fernandez LM, Kenyon NH, Morel J(2003) The Gulf of Cadiz: an unstable giant contourite levee.Geo-Mar Lett 23(1):7–18. doi:10.1007/s00367-003-0119-0

Niemann H, Duarte J, Hensen C, Omoregie E, Magalhaes VH, ElvertM, Pinheiro LM, Kopf A, Boetius A (2006) Microbial methaneturnover at mud volcanoes of the Gulf of Cadiz. GeochimCosmochim Acta 70(21):5336–5355

Ochoa J, Bray NA (1991) Water mass exchange in the Gulf of Cadiz.Deep-Sea Res 38(1):S465–S503

Orange DL, Breen NA (1992) The effects of fluid escape onaccretionary wedges 2. Seepage force, slope failure, headlesssubmarine canyons, and vents. J Geophys Res 97(B6):9277–9295

Orange DL, Greene HG, Reed D, Martin JB, McHugh CM, RyanWBF, Maher N, Stakes D, Barry J (1999) Widespread fluidexpulsion on a translational continental margin: mud volcanoes,fault zones, headless canyons, and organic-rich substrate inMonterey Bay, California. GSA Bull 111:992–1009

Orange DL, Yun J, Maher N, Barry J, Greene G (2002) TrackingCalifornia seafloor seeps with bathymetry, backscatter andROVs. Cont Shelf Res 22:2273–2290

Paull CK, Ussler W III, Dillon WP (2000) Potential role of gashydrate decomposition in generating submarine slope failures. In:Max MD (ed) Natural gas hydrate in oceanic and permafrostenvironments. Kluwer, Boston, pp 149–156

Pilcher R, Argent J (2007) Mega-pockmarks and linear pockmark trainson the West African continental margin. Mar Geol 244:15–32

Platt JP, Anczkiewicz R, Soto JI, Kelley SP, Thirlwall M (2006) EarlyMiocene continental subduction and rapid exhumation in thewestern Mediterranean. Geology 34:981–984

Schmuck EA, Paull CK (1993) Evidence for gas accumulationassociated with diapirism and gas hydrates at the head of theCape Fear slide. Geo-Mar Lett 13(3):145–152. doi:10.1007/BF01593187

Scholz F, Hensen C, Reitz A, Romer RL, Liebetrau V, Meixner A,Weise SM, Haeckel M (2009) Isotopic evidence (87Sr/86Sr, δ7Li)

246 Geo-Mar Lett (2010) 30:231–247

for alteration of the oceanic crust at deep-rooted mud volcanoesin the Gulf of Cadiz, NE Atlantic Ocean. Geochim CosmochimActa (in press). doi:10.1016/j.gca.2009.06.004

Shaw J, Courtney RC, Currie JR (1997) Marine geology of St.George’s Bay, Newfoundland, as interpreted from multibeambathymetry and back-scatter data. Geo-Mar Lett. 17(3):188–194.doi:10.1007/s003670050025

Somoza L, Maestro A, Lowrie A (1999) Allochthonous blocks ashydrocarbon traps in the Gulf of Cadiz. Offshore TechnologyConference, OTC 10889:571–577

Somoza L, Díaz-del-Río V, Hernández-Molina FJ, León R, Lobato A,Alveirinho-Dias JM, Rodero J, TASYO Team (2000) Newdiscovery of a mud-volcano field related to gas venting in theGulf of Cadiz. Imagery of multibeam data and ultra-highresolution seismic. In: Ext Abstr 3rd Symp The Iberian AtlanticMargin, 25–27 September 2000, Faro, Portugal, pp 397–398

Somoza L, Gardner J, Díaz-del-Río V, Vázquez JT, Pinheiro LM,Hernández-Molina FJ, TASYO/ANASTASYA shipboard scien-tific parties (2002) Numerous methane gas-related sea floorstructures identified in the Gulf of Cadiz. Eos Trans AGU 83(47):541, 549

Somoza L, Díaz-del-Río V, León R, Ivanov M, Fernández-Puga MC,Gardner JM, Hernández-Molina FJ, Pinheiro LM, Rodero J,Lobato A, Maestro A, Vázquez JT, Medialdea T, Fernández-SalasLM (2003) Seabed morphology and hydrocarbon seepage in theGulf of Cadiz mud volcano area: acoustic imagery, multibeam andultrahigh resolution seismic data. Mar Geol 195:153–176

Stadnitskaia A, Ivanov M, Gardner J (2000) Hydrocarbon gascomposition and distribution in surface sediments of mud

volcanic province. Gulf of Cadiz. NE Atlantic. In: Abstr Vol VIInt Conf Gas in Marine Sediments, 5–9 September 2000, St.Petersburg, pp 126–129

Stadnitskaia A, Ivanov M, van Weering TCE, Sinninghe Damsté JS,Werne JP, Kreulen R, Blinova V (2001) Molecular and isotopiccharacterization of hydrocarbon gas and organic matter from mudvolcanoes of the Gulf of Cadiz, NE Atlantic. Final Proc Int ConfGeological Processes on Deep-Water European Margins, Mos-cow, Russia. UNESCO IOC Worksh Rep 175:47

Stadnitskaia A, Ivanov M, Blinova V, Kreulen R, Van Weering TCE(2006) Molecular and carbon isotopic variability of hydrocarbongases from mud volcanoes in the Gulf of Cadiz, NE Atlantic.Mar Petrol Geol 23:281–296

Taylor MH, Dillon WP, Pecher IA (2000) Trapping and migration ofmethane associated with the gas hydrate stability zone at the BlakeRidge Diapir—new insights from seismic data. Mar Geol 164:79–89

Vogt PR, Gardner J, Crane K, Sundvor E, Bowles F, Cherkashev G(1999) Ground-truthing 11- to 12-kHz side-scan sonar imagery inthe Norwegian-Greenland Sea. Part I. Pockmarks on the VestnesaRidge and Storegga slide margin. Geo-Mar Lett 19(1/2):97–110.doi:10.1007/s003670050098

Wood WT, Gettrust J, Chapman NR, Spence GD, Hyndman RD (2002)Decreased stability of methane hydrates in marine sediments dueto phase boundary roughness. Nature 420:656–660

Zitellini N, Gràcia E, Matias L, Terrinha P, Abreu MA, DeAlteriis G,Henriet JP, Dañobeitia JJ, Masson DG, Mulder T, Ramella R,Somoza L, Diez S (2009) The quest for the Africa-Eurasia plateboundary west of the Strait of Gibraltar. Earth Planet Sci Lett280:13–50

Geo-Mar Lett (2010) 30:231–247 247