Clinoform nucleation and growth in coarse-grained deltas, Loreto basin, Baja California Sur, Mexico:...

Clinoformnucleation and growth in coarse-graineddeltas,Loreto basin,Baja California Sur,Mexico: aresponse to episodic accelerations in faultdisplacementEstelle Mortimer,n Sanjeev Guptaw and Patience Cowien

nInstitute of Earth Science, School of GeoSciences, University of Edinburgh, Edinburgh, UKwDepartment of Earth Science and Engineering, Imperial College London, SouthKensington Campus, UK

ABSTRACT

We investigate the controls on the architecture of coarse-grained delta progradational units (PUs) inthePlioceneLoreto basin (BajaCaliforniaSur,Mexico), a half-graben located on thewesternmargin oftheGulf of California. Dorsey et al. (1997b) argued that delta progradation and transgression cycles inthe basin were driven by episodic fault-controlled subsidence along the basin-bounding Loreto fault.Here we test this hypothesis by a detailed analysis of the sedimentary architecture of11exceptionallywell-exposed, vertically arranged £uvio-deltaic PUs, each ofwhich shows lateral facies transition fromproximal alluvial facies palaeo-seaward into distal pro-delta facies. Of these11PUs, seven exhibit alateral transition from a shoal water to Gilbert-delta facies associations as they are traced palaeo-seaward.This transition is characterised by down-transport development of foresets, which grow inheight up to 35m. Foreset units thicken in a basinward direction, with initially an oblique topset^foreset geometry that becomes increasingly sigmoidal. Each delta is capped by a shell bed that recordsdrowning of the delta top.This systematic transition in delta architecture records increasing waterdepth through time during individual episodes of progradation. A mechanism that explains thistransition is an accelerating rate of fault-controlled subsidence during each PU.During episodes oflow slip rate, shoal-water deltas prograde across the submerged topography of the underlying deltaunit. As displacement rate accelerates, increasing bathymetry at the delta front leads to steepening offoresets and initiation of Gilbert deltas. Subsequent delta drowning results from sediment starvationat the shoreline at high slip rates because of sediment trapping upstream.The observed deltaarchitecture suggests that the long-term (4100kyr) history of slip on the Loreto fault wascharacterised by repetitive episodes of accelerating displacement accumulation. Such episodic faultbehaviour is most likely to be because ofvariations in temporal and spatial strain partitioning betweenthe Loreto fault and other faults in the Gulf of California. A physical explanation for the accelerationphenomenon involves evolving frictional properties on the episodically active Loreto fault.

INTRODUCTION

The sedimentary ¢ll of continental extensional basinsprovides the only record of allogenic forcing on basin evo-lution. A common stratigraphic feature of the ¢ll of thesebasins is the presence of vertically stacked delta prograda-tional units (PUs) in the hangingwall of the basin-bound-ing fault (Colella, 1988a, b). The depositional architecturehas been inferred to be the product of the interaction be-tween (1) tectonically controlled subsidence, (2) glacio-eu-static £uctuations, (3) sediment supply £uctuations to thedelta front (e.g. Colella, 1988a; Dart et al., 1994; Gawthorpe

et al., 1994; Gupta et al., 1999; Marr et al., 2000; Swensonet al., 2000). In this study we present new geometric con-straints from analysis of the depositional architecture,which alone can be used to discriminate between thesecontrols and to characterise and quantify variations in therate of tectonically controlled subsidence.

In a series of papers based upon stratigraphic observa-tions from the Pliocene Loreto basin, western margin Gulfof California, Dorsey and co-workers proposed a novel hy-pothesis for the generation ofvertically stacked deltaic suc-cessions.This basin contains a superbly exposed successionof syn-rift deltas preserved in the exhumed hangingwall ofthe Loreto fault. Based on evidence for the frequency ofPUs compared with eustatic curves, they suggested thatepisodic variations in the rate of fault-controlled subsi-dence on the basin-bounding Loreto fault controlled the

Correspondence:EstelleMortimer, Institute fˇrGeowissenschaf-ten, Universit�t Potsdam, Karl-Liebknecht Strasse, 14476, Pots-dam, Germany. E-mail: [email protected]

BasinResearch (2005) 17, 337–359, doi: 10.1111/j.1365-2117.2005.00273.x

r 2005 Blackwell Publishing Ltd 337

architecture of the deltas (Falk, 1996; Dorsey et al., 1997b;Dorsey & Umhoefer, 2000). They argued that the Loretofault behaved episodically, with periods of quiescence dur-ing which deltas prograded, punctuated by periods ofheightened activity during which the delta tops weredrowned.Their argument was based upon the recognitionof 16 units of delta progradation separated by correlatable£ooding surfaces, and temporally constrained by two datedtu¡ horizons to100� 80 kyr duration.This implies that in-dividual units of delta progradation and drowning have fre-quencies of 6250� 5000 years.They argued that this shortduration cannot be explained by Milankovitch (1930) bandglacio-eustatic sea-level £uctuations.

If successive units of delta progradation and transgres-sion are related to episodicity in displacement rates on theLoreto fault, it is likely that this signature is preserved inthe sedimentary architecture of the deltas. In this paper,we present new data on the facies architecture and stack-ing pattern of 11 PUs in the Loreto succession and re¢nethe temporal constraints with new ages on key tu¡ hori-zons.We document down-transport variations in the geo-metry of delta PUs, and quantify the length scales of faciesunits, major facies boundaries and key stratal surfaces.Weobserve a systematic variation in delta architecture, whichwe argue cannot be generated by sediment supply varia-tions or by eustatic sea-level £uctuations regardless oftheir periodicity.We believe that this recurring variationin architecture records a long-term displacement historyon the basin-bounding Loreto fault that is characterisedby episodic, accelerating displacement rates.

GEOLOGICAL SETTING

The transtensional Loreto basin (Fig. 1a) is located on thewestern margin of the Gulf of California, and is one of aseries of Pliocene basins that form the Gulf ExtensionalProvince (Lonsdale, 1989).The opening of the Gulf of Ca-lifornia commenced �12Ma (Lonsdale, 1989), and the‘modern’ Gulf con¢guration was achieved by 3.5Ma,although true oceanic spreading began only ca. 2.6Ma(Lonsdale, 1989). The oldest sediments preserved in theLoreto basin are ca. 3.5Ma (Umhoefer et al., 1994). SinceLate Pliocene times, the southern Loreto basin has beenuplifted and dissected to provide excellent exposure ofthe Pliocene sedimentary ¢ll (McLean, 1989).

The Loreto basin has a characteristic half-graben geo-metry (Fig.1b), bounded to thewest by theNNWtrending,35-km-long Loreto fault. Although the Loreto fault maypossibly have had some earlier movement, during theMiocene, the hangingwall motion during the Pliocenewas to the east under bulk extension (Umhoefer &Dorsey,1997).The Loreto fault is a dominantly dip-slip fault, withpossibly a very minor component of right slip (Umhoefer& Stone, 1996). To the east, the basin is bounded by theEastern Structural High, a block of Miocene volcanicrocks, uplifted as the hangingwall underwent westwardtilting during the Pliocene (Dorsey et al., 1995). The wes-

tern margin of the ESH is the point about which tiltinginto theLoreto fault has occurred and it represents the ap-proximate location of the half-graben fulcrum (Dorseyet al., 1997a).

The Pliocene sedimentary ¢ll in the Loreto basin is de-posited unconformably ontoMiocene volcanic rocks (Dor-sey et al., 1995, 1997a, b).The basin is divided into two sub-basins, separated by a NNE trending anticline (Dorsey &Umhoefer, 2000). The geometry of the half-graben ¢ll isasymmetric, with sediments thickening towards the westinto the Loreto fault (Dorsey et al., 1995).The sedimentary¢ll (Dorsey etal.,1995) in the immediate hangingwall of theLoreto fault is a narrow (2^3 km) belt of non-marine con-glomerates and sandstones.These deposits constitute theVinorama Conglomerates (Dorsey & Umhoefer, 2000),which inter¢nger eastwards (palaeo-seaward) with marinesandstones and conglomerates (Las Piedras Rodadas For-mation, Dorsey & Umhoefer, 2000), which in turn passinto more distal marine sandstones to the north and westof the basin (Las Piedras Rodadas Formation: Dorsey &Umhoefer, 2000).The uplift and subaerial exposure of theESH provides a source of carbonate material (Arroyo deArce limestone; Dorsey & Umhoefer, 2000) deposited insouthward prograding mixed carbonate and siliciclasticdeltas (Dorsey & Kidwell, 1999). The north and eastward¢ning non-marine andmarine sandstones and conglomer-ates constitute a series of vertically stacked, northwardprograding coarse-grained delta PUs (Falk, 1996; Dorseyetal., 1997a), whichwere deposited into the central sub-ba-sin during a period of heightened subsidence (Dorsey etal.,1997a).This set of deltas is the focus of our investigation.

METHODS

We present a high-resolution investigation into the faciesarchitecture of the Loreto basin.The stratigraphic and se-dimentological data come from an area of approximately6 km2 in the Loreto central sub-basin (Fig. 1c), commen-cing 1500m east of the Loreto fault and extending 2500meastward. The data comprise 61 measured sedimentarysections at a scale of 1 : 50, which are closely spaced (be-tween 20 and 320m) sub-parallel to the direction of pro-gradation of the depositional units. These sections covera two-dimensional section over 2000-m long and 400min the vertical. Analysis of the sedimentology of the suc-cession enabled the construction of a facies framework,which has been combinedwith ¢eld observations of faciesrelationships to interpret the depositional environments.We record17delta PUs (coarsening-up delta facies associa-tions separated by correlatable shell beds; see ‘Sedimen-tology of Loreto Basin Deltas’ for sedimentology), ofwhich the lower 11 are exceptionally well preserved in thecentral sub-basin. Our study focuses on the architectureof these lower 11, well-exposed, PUs, but where relevantincorporates observations of units above these.

Reconstruction of the depositional architecture wasachieved through correlation of facies boundaries and key

r 2005 Blackwell Publishing Ltd,Basin Research, 17, 337^359338

E. Mortimeret al.

stratal surfaces on the sedimentary logs. In particular, lat-erally extensive transgressive £ooding surfaces (shell beds),that are traceable for at least1km across the areawere usedin construction of correlation panels. The cross-sectionsthat are presented here have been generated by £atteningthe upper shell bed (£ooding surface) and correlating fa-cies boundaries between sections. Data were collected at aminimum of1.5 km east of the Loreto fault, ca. 6.5 kmwest

of the ESH. The maximum horizontal extent of correla-tions for any given PU is ca. 2.3 km.

AGEMODEL FOR THE LORETO BASINSTRATIGRAPHY

The Loreto basin has been temporally constrained by40Ar/39Ar dating of four tu¡ horizons (Umhoefer et al.,

Fig.1. (a) Inset: location map ofLoreto basin (L), BajaCalifornia Sur (BCS) along thewestern margin of theGulf of California (G ofC).The basin is part of the Gulf Extensional Province (GEP ^ highlighted in pale grey) comprising a series of Pliocene extensional basins.Main ¢gure: the geological map of theLoreto basin highlighting the Pliocene sedimentary ¢ll (sst, sandstone; cgl, conglomerate; sh.sst,shelly sandstone) bounded to the west by the Loreto fault, and to the east by the Eastern Structural High.The location of threeinterbedded tu¡s is marked,Tu¡s 2 and 3 have been datedwithin this study.The study area is highlightedwithin the box (after Dorseyet al., 1995). (b) Enlargement of the study area containing the investigated section withTu¡ 2 at its base (also located in (a)). Correlatableshell beds, which separate units of delta progradation are marked (grey), with measured sections positioned between them (very close-spaced sections (o30m) are not marked but are labelled ‘LORB’ sections in Figs 3 and 4 subsequently). (c) Schematic cross- sectionthrough the Loreto basin (see (a) for key) indicating the study area within the overall basin geometry (after Dorsey et al., 1997a).

r 2005 Blackwell Publishing Ltd,Basin Research, 17, 337^359 339

Clinoformnucleation and growth

1994; Dorsey et al., 1997b), of which the second and thirdbound the investigated section of stacked delta PUs and120m of overlying gypsiferous ¢ne-grained sandstones(Dorsey et al., 1997a).The hypothesis proposed by Dorseyet al. (1997b) argued that, based on the ages of the twobounding tu¡s, the duration of individual episodes of del-ta progradation was too small to be explained by glacio-eustatic sea-level £uctuations.They proposed an episodictectonic control model to explain the observed deposi-tional architecture. Accurately constraining the durationbetween the two tu¡s is therefore important to test the hy-pothesis of Dorsey et al. (1997a, b). The original geochro-nological study undertaken (Umhoefer et al., 1994) datedthe tu¡s using a multiple (10) crystal fusion technique.Here a new age model from a recent re-examination ofthe geochronology of the tu¡s using single-crystal40Ar/39Ar analysis (Mortimer, 2004) is presented.

The tu¡s were sampled at known localities within theLoreto basin stratigraphy.The two tu¡s are referred to asTu¡s 2 and 3 after Umhoefer et al. (1994) for the lower andupper tu¡, respectively.Tu¡ 2, a white to grey pumice-liketu¡with discrete crystals � 1mm of feldspar, hornblendeand biotite as well as varying quantities of glass, wassampled from the base of the section beneath PU 1.Tu¡ 3is located 320m above the stacked delta PUs within a suc-cession of gypsiferous sandstones and mudstones (Fig.1a;Dorsey &Umhoefer, 2000).Tu¡ 3 is a ¢nely laminated tu¡comprising amphibole, pyroxene, micas and feldspar.

Analysis of the tu¡s to obtain 40Ar/39Ar ages was under-taken at the Scottish Universities Environment ResearchCentre (SUERC), Scotland, UK. Feldspar crystals fromeach of the tu¡s were used for single-crystal, laser fusionanalysis. This single-crystal analysis yielded ages (to95% con¢dence) of 2.499� 0.036Ma for Tu¡ 2, and2.211� 0.046Ma forTu¡ 3 (Mortimer, 2004). These agesdi¡er from those previously attained by Umhoefer et al.(1994) of 2.46 � 0.177 and 2.36 � 0.02Ma forTu¡s 2 and 3,respectively. Statistically, this new age and that of Umhoe-fer et al. (1994) for Tu¡ 2 belong to the same population;however, there is a noticeable discrepancy in the age ofTu¡3.The single-crystal analysis of Tu¡ 3 recognises an olderpopulation of xenocrysts, which account for approxi-mately one- ¢fth of the total crystal population. Theseolder xenocrysts were not included in the new age calcula-tions.This explains the variation in ages between Umhoe-fer et al. (1994), who used heating of groups of crystals, andMortimer (2004) that uses single crystals. The newly ac-quired single-crystal method age is therefore a more accu-rate age of Tu¡ 3.

Using the new ages (Mortimer, 2004) of the two bound-ing tu¡s, the duration between the two tu¡s is0.287 � 0.117Myr to the 95% con¢dence. From this, anaverage duration of each PU between two shell beds of14.1 � 8.6 kyr is calculated (by dividing the total durationby the number of PUs), comparedwith the previously cal-culated 6.25� 5.00 kyr (Umhoefer et al., 1994; Dorsey et al.,1997b). If each of the PUs (section between two shell beds)is considered individually, an approximation for that dura-

tion of that unit can be achieved by dividing the durationbetween the two tu¡s by the proportion of the verticalthickness that that PU represents (i.e. duration single uni-t5 [vertical thickness of PU/total stratigraphic thicknessbetween tu¡s] � total duration between tu¡s).This calcu-lation is obviously an approximation, as it assumes con-stant sedimentation rate for the section; however, theresults give the duration for many units signi¢cantly lessthan the average (Table 1).This is primarily because of thethickness of PU1being greater that those above. If unit1 isexcluded, the average duration of a PU is 11.0 � 4.5 kyr.Furthermore, the calculated duration of each PU is likelyto be a maximum for two reasons: (1) having assumed aconstant sediment supply, it is important to consider thatthe sediment preserved include some time of depositionalhiatus or condensed section; (2) the upper portion of thesection beneathTu¡ 3 comprises ¢ne-grained sandstonesand mudstones the depositional rate of which is probablymuch lower than for the delta units, and therefore ac-counts for more time than the latter.

The duration of most of the PUs is therefore too small(by approximately a factor of 4) to be attributed to glacio-eustatic sea-level £uctuations, which were dominantly41kyr during the Pliocene (e.g. Naish et al., 1997; Raymoet al., 1998; Lisiecki & Raymo, 2005); or climatically con-trolled variations in sediment supply (Leeder et al., 1998;Allen&Densmore, 2000).Our re-examination of the geo-chronology leads us to propose a new age model for theLoreto basin stratigraphy, but nevertheless re-a⁄rms theargument of Dorsey et al. (1997b) that the duration of deltaprogradational episodes is too short to be explained byMilankovich-band glacio-eustatic sea-level £uctuations.Below we present geometric arguments against sea-level£uctuations and climatic variations in sediment supply asa control on delta architecture. Thus the geochronologicconstraints presented here, whereas supporting the tec-tonic control hypothesis, are of interest primarily for

Table1. Estimated duration of each progradational unit investi-gated in this study

Progradational unit Thickness (m) Duration (kyr)n � (kyr)

1 94 36 152w 51 20 83 42 16 74 18 7 35 26 10 46w 32 12 57w 39 15 68Aw 22 8 48w 13 5 29Aw 20 8 39w 21 8 3

The duration of many units is less than10 000 years.nBasedupon287kyr ( � 117kyr) and740m of sectionbetweenTu¡s2and3.wProgradational units containing a transition from shoal-water toGilbert deltas.


E. Mortimeret al.

quantifying variations in displacement rates, rather thandetermining the control.

SEDIMENTOLOGYOF LORETO BASINDELTAS

Our sedimentary facies analysis of the Loreto basin iden-ti¢es two distinct deltaic facies associations: (1) a shoal-water delta facies association (Postma, 1990, 1995) and; (2)aGilbert-delta facies association (e.g. Ethridge&Wescott,1984; Colella, 1988b; Postma, 1990).The shoal-water deltafacies association comprises £uvial facies that pass later-ally into proximal mouth bar, distal mouth bar and prodel-ta facies.The Gilbert-delta facies association comprises atripartite assemblage consisting of (1) planar-strati¢ed,£at-lying or shallowly seaward dipping topsets that passpalaeo-seaward into sub-aqueous steeply inclined fore-sets, and tangential bottomsets. The previous studies ofDorsey etal. (1995,1997b) andFalk&Dorsey (1998) have foc-used only on the identi¢cation ofGilbert-delta facies asso-ciation.

The proximal part of the shoal-waterdelta facies associa-tion is dominated by conglomerates forming channel-formunits with scoured basal surfaces. These conglomeratesinter¢nger with, and are laterally equivalent to, ¢ne-grained sandstones with rootlets. We interpret the con-glomerates as £uvial channel deposits, and the sandstonesas £oodplain deposits, all forming part of a subaerialcoastal plain. The submarine proximal mouth bar faciesof the shoal-water delta facies association represent de-position in a complex zone of interaction between thenon-marine and marine processes, especially during highsediment discharge events (Colella et al., 1997). Proximalmouth bar facies comprise interbedded pebble conglom-erates and sandstones, showing depositional dips of 9^161andwithvarying degrees of organisation and bioturbation.Deposition is interpreted to have occurred from rapid de-celeration of £ow (inertia-dominated £ows) and traction(friction-dominated £ows), resulting in rapid depositionof the bedload (Nemec, 1990). Proximal mouthbar faciesshow lateral transition into distal mouthbar facies whichcomprise ¢ne-grained sandstones with up to 5% of smallpebbles (1^3 cm) that are interpreted to have been depos-ited from more distal derivatives of density £ows and sus-pension (Lowe,1976,1982;Nemec,1990;Colella etal.,1997;Sohn, 1997).

TheGilbert-delta facies association comprises a tripar-tite assemblage. Gilbert-delta topset facies comprisechannel-form conglomerates and laterally equivalentrooted ¢ne-grained sandstones that are interpreted as al-luvial facies, as in the shoal-water delta facies association.There are, however, some notable di¡erences with the lat-ter. One distinct facies is only recognised in the Gilbert-delta association; this comprises planar bedded, wellstrati¢ed, ungraded to normally graded pebble^cobbleconglomerates containing abundant shelly material and,

in the upper PUs, broken algal material.The planar strati-¢cation of these beds re£ects variations in discharge anddiscontinuities in the accretion of the beds (Nemec &Steel, 1984). The sub-aqueously deposited Gilbert-deltaforesets are recognised by their steep depositional angle(191 and 281; up to 301 in some outcrops) that is diagnosticin its interpretation as a Gilbert delta (e.g. Gilbert, 1885;Nemec & Steel, 1984; Colella et al., 1987; Colella, 1988b;Postma, 1990; Sohn et al., 1997). Sedimentary processesdominating the foresets are grain £ows and avalanchingdown the foreset slope (e.g. Nemec & Steel, 1984; Postma& Roep, 1985; Colella et al., 1997; Sohn et al., 1997). Foresetdeposits pass laterally palaeo-seaward into planar strati-¢ed beds comprising conglomerates and ¢ne-to-med-ium-grained sandstones with occasional inter-strati¢eddisorganised pebble beds with abundant shell materialand bioturbation.These are interpreted as bottomset de-posits formed by rapid deceleration of £ows at the toe ofthe delta front (Nemec, 1990).

In both shoal-water and Gilbert-delta facies associa-tions, the most distal deposits represent deposition in aprodelta environment. These sediments comprise ¢ne-to-medium-grained, highly bioturbated sandstones withabundant marine fauna (including articulate bivalves). De-position occurs from suspension as the distal derivativesof density currents, and from background settling out ofsediment from suspension (Lowe, 1982; Nemec, 1990;Mulder & Alexander, 2001).

Each delta progradation unit is capped by a shell bed.These are laterally continuous, traceable for up to 2 kmparallel to progradation direction, and typically of the or-der of 0.5^4-m thickness. They appear as concentrated,densely packed beds comprised of pectens, oysters, gastro-pods, barnacles, coral, bryozoans and echinoderms invarying stages of preservation. Articulate, disarticulateand fragmented shells; shell hash; and moulds are pre-served. The distinct reduction in clastic sediments indi-cates a marine transgression, and a depositional hiatus orcondensed section on the delta top (Kidwell, 1989; Orpin& Reading, 1993; Abbott, 1997).

DEPOSITIONAL ARCHITECTURE OFDELTA PROGRADATIONAL UNITS

In our analysis, we de¢ne a delta progradational unit (PU)as the section of the stratigraphy that is bounded at itsupper and lower contacts by shell beds. The reconstruc-tions presented (Fig. 2) are correlations of the facies asso-ciations between measured logged sections using shellbeds as datum horizons. The sections are 2-D cross-sec-tions parallel to the direction of transport as determinedfrom palaeocurrent data (imbrication, cross-beds and fore-set orientation) to be towards the north to north-northeast.The reconstructions are therefore oriented south (proxi-mal)^north (palaeo-seaward). A total of 17 units of coarse-grained delta progradation separated by shell beds are re-corded betweenTu¡ 2 and shell bed 15; however, this study



Fig.2.

(a)C

orrelation

panelofthe

investigated

deltaicsuccession

intheLoretobasinachieved

throughthecorrelationofmeasuredsedimentary

sections

(verticalblack

lines)tolaterally

extensive

andtraceable£o

odingsurfaces(shellbeds

(SB),blue

lines).Progradation

alun

its(PU

)areboun

dedtopandbottom

byshellbeds.SB

9,7,6,4,3,2have

been

£attened

inthiscorrelation.Weob

servea

totalof17no

rth-northeastprograding

deltaun

its,14

ofwhich

containaG

ilbert-deltafaciesassociation.Thisstudy

focuseso

nthelower11un

itsb

etweenTu

¡2andSB

9.(b)D

epositionalarchitecture

ofthelower11PU

s.SB

9(upp

ersurface)£attened.T

heoverallgeometry

istypicalofanasym

metrichalf-graben.Thereisathickening

ofthealluvialsectiontowardtheL

oretofault,andthinning

away

from

it.Ofthese11un

its,8(PUbeneathSB

2,6,7,8A

,8,9A,9)con

tainGilb

ertd

eltasthatinitiated

asshoal-waterdeltas.A

lloftheseG

ilbertd

eltasexh

ibitab

asinwardthickening

offoresetu

nits(red)

andincrease

inforeseth

eight.


E. Mortimeret al.

Fig.2.

(b)C

ontinued



focuses on the exceptionally well-exposed lower sectionbetween Tu¡ 2 and shell bed 9.The exposure of the lowerPUs presented trends N^S and has been measured inwest-trending sections. Therefore, no geometric adjust-ment is required of the logged sections in order to con-struct the depositional architecture and progradationaldistances from the ¢eld data. A summary of the key quanti-tative data is provided inTable 2.

Overall depositional architecture

The Loreto basin contains a series of footwall-deriveddeltaic PUs that record episodes of shoreline regressionand transgression across a narrow regionwithin the hang-ingwall of the Loreto fault (Figs 2a and b, Table 2). Themost landward nucleation point for deltas is located1.3 km from the trace of the Loreto fault, while the mostseaward-located clinoform break point occurs 3 km fromthe fault.The landward extent of the shoreline is in manycases an estimate because of the lack of correlatable expo-sures more proximal to the Loreto fault. Mapping of themaximum seaward extent of shoreline progradation inPUs1^5 reveals an overall back-stepping trend; this is fol-lowed by fore-stepping of PUs 6^7; and backstepping be-tween PUs 7 and 8A. From PUs 8A to 9, the clinoformbreakpoint shows fore-stepping. When a comparison ismade between PUs, there is marked variation in theproportion of shoal-water and Gilbert deltas that occurwithin a single PU.

The reconstruction of the sedimentary architecture ofthe lower delta PUs (Fig. 2b) reveals11delta PUs separated

by correlatable shell beds. These show progradation ap-proximately north to north-east. All11units initially com-prise shoal-water delta facies association (showing lateraltransition from alluvial fan^proximal mouth bar^distalmouth bar^prodelta facies). In eight of the PUs a Gilbert-delta tripartite facies association is also recognised occur-ring within the same PU as a shoal-water delta. The bestexposed of these units containing both shoal-water andGilbert-delta facies associations are PUs 6 and 7.

PU 6

PU 6 (Fig. 3) is exceptionally well exposed in the Loretobasin initiating 1.5 km from the Loreto fault and a further1.5 km into the basin. The unit records �1.3 km of deltaprogradation. It comprises a single coarsening-up deltaprogradation unit that contains both a shoal-water andGil-bert-delta facies associations that are laterally displacedparallel to depositional dip.The proximal part of the deltaunit comprises typical shoal-water delta facies. Approxi-mately 600m palaeo-seaward of the delta initiation point,beds of the delta mouthbar facies show an increase in de-positional dip indicating an increase in clinoform dips.This indicates the initiation of distinct foreset facies, andthe nucleation of Gilbert-delta clinoforms.

Foreset facies are ¢rst noted in section LORB3 (Fig. 4),located2130mbasinward of theLoreto fault.The successionin this section shows a vertical transition from sandy distalmouth bar facies that coarsen-up into proximal mouth barsandstones and conglomerates. In this location, however,the depositional angle of the mouth bar deposits increases

Table 2. Summary of quantitative data for each delta progradation unit (PU) investigated in the Loreto basin

UnitTotalprogradation

S-Wprogradation

GDprogradation

D (maximum)LF-shoreline

MaxH. GDforesets

T.section

T.AF

D. LF- ¢rstsection

D. LF ^ GDnucleation Transgression

1 1472 1160 312 3000 23 94 69 1700 2688 9002 1060 960 100 3160 14 51 18 2000 3060 11603 406 406 0 2406 0 42 13 2000 0 4054 569 569 0 2570 0 18 5 2000 0 6705 156 156 0 2056 0 26 13 1500 0 5466 1265 610 655 2775 25 32 29 1500 2120 12757 1225 410 815 2725 26 39 19 1500 1910 12708A 700 281 419 2150 11.5 22 19 1600 1731 3808 640 211 329 2400 12 13 10 1600 2071 8839A 1082 670 412 2600 19 20 14 1600 2188 5129 762 250 512 2850 28 21 25 1600 2338 140010 N/A N/A N/A 2700 20 39 N/A UKN UKN UKN11 N/A N/A N/A 2900 10 30 N/A UKN UKN UKN12 N/A N/A N/A 3000 32 45 N/A UKN UKN UKN13 N/A N/A N/A 2900 29 42 N/A UKN UKN UKN14 N/A N/A N/A 3000 24 33 N/A UKN UKN UKN15 N/A N/A N/A 3200 30 52 N/A UKN UKN UKN

Shaded represents a minimumvalue.All values in metres unless otherwise stated.Each measured section commences at least 1500m basinward of the Loreto fault (LF).The maximum height of Gilbert foresets, and the distance ofclinoform nucleation from the LFare recorded for the lower11units (PU1-9) where present.This was not possible for units 10^15.D, distance;H, height;T, thickness; LF, Loreto Fault; GD, Gilbert delta; S-W, shoal-water delta; AF, alluvial fan; UKN, unknown.


E. Mortimeret al.

up in section from 161 to 191 immediately prior to the ¢rstforeset facies being identi¢ed. This observed increase indepositional angle and nucleation of aGilbert-delta foresetunit is entirely conformable.These initial clinoforms com-prise interbedded sandstones and conglomerates and have aforeset height of not more than 4m.The foresets are imme-diately overlain by planar strati¢ed conglomerates that areinterpreted as subaerial topset facies. The topset^foresetgeometry of the clinoforms is oblique.

Traced further palaeo-seaward, PU 6 shows continueddevelopment of a Gilbert-delta geometry. The height ofthe foreset unit increases to over 8mwithin 100m of pro-gradation; this increase in height continues palaeo-sea-ward reaching over 21m before the clinoform breakpoint, which occurs 2775m from theLoreto fault. Coupledwith the palaeo-seaward increase in foreset heightis an increase in the aggradational component on thetopsets, with the result that the clinoforms show a basin-ward transition from oblique to sigmoidal clinoformgeometry.

The topset of PU 6 is erosionally overlain by a pebbleconglomerate bed (shell bed 6) that shows an upward in-crease in shell material (shell bed 6) indicating drowningof the delta top.

PU 7

PU 7 (Fig. 3) commences 1.5 km palaeo-seaward of theLoreto fault, and has been reconstructed from 11 loggedsections spaced between 30 and 250m apart. It initiates inthe proximal section as shoal-water delta facies associa-tions that show a lateral transition intoGilbert-delta faciesassociations as it is traced palaeo-seaward. This faciestransition exhibits the same overall characteristics as PU6,with conformable nucleation of foresets from an initiallyshoal-water delta geometry.

Observations of this unit commence in the proximal re-gions with deposition of coarsening-up shoal-water deltafacies. Palaeo-seaward a transitional region in which thedepositional angle of the delta front increases in the prox-imal mouth bar facies of the shoal-water delta to 171 isnoted at locality LORB4, ca. 1670m palaeo-seaward of theLoreto fault. Above this increase in depositional angle ofthe shoal-water delta front, clinoforms nucleate with fore-set heights of not more than 5m.

The nucleation point ofGilbert-delta facies in PU7 hasbeen located through correlation of detailed and close-spaced sedimentary sections. As in PU 6, the foresetheights of this Gilbert-delta unit are small (4^5m) at itsnucleation point.These ¢rst clinoforms comprise foresetsof sandstone and conglomerate interbeds that are overlainby planar strati¢ed conglomerates of the sub-aerial topsetfacies. The topset-foreset geometry of the clinoforms isoblique. Palaeo-seaward the foresets increase in height toca. 10mwhere a sigmoidal clinoform geometry is ¢rst ob-served. Because some of the topset-foreset geometry can-not be determined, sigmoidal clinoforms are observed inthe more distal regions of this Gilbert-delta unit. Foreset

heights continue to increase seaward to 24m at the clino-form break point ca. 2.7 km palaeo-seaward of the Loretofault. PU 7 shows a minimum progradational distance of1.2 km, of which 410m consists of shoal-water delta faciesand a further 810m as a Gilbert-delta facies.

Summary of the depositional architecture

Of the11PUswe have studied in theLoreto basin, all com-mence with the progradation of a shoal-water delta (withalluvial fan^proximal mouth bar^distal mouth bar^pro-delta facies association). In eight of the PUs aGilbert-del-ta tripartite facies association is also recognised occurringwithin the same PU as a shoal-water delta. In all but one ofthese, PU1, the Gilbert delta initiates as a coarsening-upward, shoal-water delta. In these units, as shoal-waterdeltas are traced palaeo-seaward, clinoforms nucleate andevolve into Gilbert deltas.We term these units in whichGilbert-deltas nucleate from shoal-water deltas ‘transi-tional PUs’.

The best exposed ‘transitional’ PUs are PU 6 and 7,which have been described in detail above. PUs 2, 8A, 8,9A, and 9, however, exhibit similar geometry and sedimen-tary features. In those units that show both delta faciesassociations, deposition initiates with a coarsening-upshoal-water delta. As this shoal-water delta is traced pa-laeo-seaward there is an increase in the depositional angleat the delta front from151 to 201 as clinoforms nucleate. In-itially, the height of the foresets (i.e. the height from theclinoform break point to the clinoform toe associatedwiththat foreset) of the Gilbert delta is not more than 4^5moverlain by planar-strati¢ed conglomerate topsets withoblique topset^foreset geometry.

Subsequently, the height of the foresets increases asthey are traced palaeo-seaward to 10^12m (PU2, 8A, 8) orin excess of 20m (PU6, 7, 9A,9).The maximumheight of aclinoform systematically occurs at its maximumprograda-tional extent (Figs 5 & 6). Coupled with this observed pa-laeo-seaward increase in foreset unit height, the geometryof the clinoforms also changes from initially oblique top-set^foreset geometry to sigmoidal, indicating an increas-ing amount of aggradation with continued progradation.

PU 6 and 7 are the best constrained PUs in the Loretobasin section, and a plot of their foreset height with dis-tance from nucleation (Fig. 6) highlights this initialprogradation with little increase in foreset height, fol-lowed byamore rapid increase in foreset height.This morerapid increase in foreset height in PU 6 and 7 is related tothe transition from oblique to sigmoidal clinoform geo-metry. Data points are more limited for the other PUs,although qualitative observations of their geometry in¢eld exposures reveal a similar transition from oblique tosigmoidal clinoforms. Therefore, the shoreline trajectory(i.e. the trace of the clinoform break point in space) is initi-ally progradational but becomes increasingly aggrada-tional palaeo-seaward.That is, the PUs are characterisedby concave-up shoreline trajectories.



Fig.3.

(a)M

easuredsedimentary

sections

andfacies

associationinterpretation

sforP

U6andPU

7boun

dedby

shellbeds(SB

5and6;SB

6and7forP

U6and7respectiv

ely).(b)Depositional

architectureofPU

6andPU

7,basedon

correlationofcloselyspaced

sections

(verticalblack

lines)and

walking

outstratalboun

daries.Ineach

ofthesePU

sthereisadown-transporttransitionfrom

ashoal-waterdeltato

aGilb

ertd

elta.T

hepo

sition

ofclinoformnu

cleation

,and

clinoformbreakpoint

(¢nalp

ositionoftheshoreline);and

observed

clinoformgeom

etry

arerecorded.Inboth

units

foreseth

eightisinitially

notm

orethan

4.5m,and

clinoformgeom

etry

isobliq

ue.D

own-transportthe

foreseth

eightincreases

andthegeom

etry

becomes

sigm

oidal.


E. Mortimeret al.

Fig.3.

(b)C

ontinued



ALLUVIAL ARCHITECTURE

Thus far, this study has focussed on the depositionalarchitecture of PUs, the sub-aqueous delta front, andthe shoreline trajectory. It is, however important to con-sider how this observed architecture corresponds to the

sub-aerial depositional system. In the Loreto basin, theexposure of both sub-marine and alluvial componentspermits the two systems to be explicitly related. In a back-tilting basin, in which depositional architecture is con-trolled on the ¢rst order by variations in tectonicallycontrolled accommodation creation, such variations

Fig.4. Logged section LORB 3 (refer toFig. 3b for exact location) and ¢eld sketchthat document the ¢rst occurrence ofclinoforms within progradational unit 6.In this section, a facies association of adistal and then proximal shoal-water deltafront is documented, in which thedepositional angle of the beds increases.Immediately prior to the occurrence offoresets, a granular, matrix supportedsandstone is deposited. Above this,oblique clinoforms with a height of 3.5mare observed, overlain by cobble-rich,strati¢ed conglomerates of the topsets.

Fig. 5. Plot showing location of key palaeogeographic elements of each delta progradational unit with distance from the Loreto fault.The delta initiation point (circles), the nucleation point of Gilbert deltas (right pointing grey triangles), and the clinoform breakpoint(black triangles) are all indicated. For shoal water only deltas, the maximum extent of delta progradation (down facing grey triangles) isindicated. Dashed symbols represent minimumvalues because of the extent of data coverage. Inset highlights the Loreto basingeometry and position of the study area. SW, shoal-water delta; GD, Gilbert-delta progradation.


E. Mortimeret al.

might be anticipated to be re£ected in the palaeoslope andgrain-size of the alluvial system. For this reason apalaeoslope analysis of the most proximal accessible allu-vial section was undertaken. Unfortunately, the greatererosion and dissection of the exhumed Loreto basin inthe regions of preserved alluvial facies meant that this ana-lysis was possible only for PUs 6^9A.

Of paramount importance is the correlation betweenshell beds and the alluvial section. In some units (PU 6, 7and 9) the lateral extent of shell beds is good and either ashell bed is present, or a reworked conglomerate contain-ing abundant marine fauna is identi¢ed even in the mostproximal sections. In others (PU 8A, 8, 9A) the overlyingshell beds are more di⁄cult to correlate into the proximalarea. Instead, changes in stacking patterns, grainsize andsedimentary characteristics have been used to extrapolatethe £ooding surfaces associated with these shell beds intothe proximal region.

Alluvial palaeoslopes have been calculated using themethodology of Paola &Mohrig (1996).Two parameters arerequired for the palaeoslope analysis: the median grainsizeand the depth of the channel.Themedian grainsizewas cal-culated from grainsize data recorded by using a grid withsquares of 10 cm, and by measuring the exposed long axisof clasts located at each point of the grid.This exposed longaxis approximates across the outcrop to measuring the b-axis of the clasts (Paola &Mohrig, 1996).This was underta-ken for 100^200 clasts at each locality. The depth of thechannel was measured in at least four localities along thechannel and averaged. Care was taken to record depthswhere a ¢ning-up of the channel ¢ll was preserved. Theprinciple potential sources of error within the data includerecorded grain-size (reduced by measuring over 100 clastsand taking the median); and preservation of the channeldepth (reduced by measuring only where a ¢ning-up is pre-served). Paola & Mohrig (1996) suggest that typical errorsare a factor of 2 of the calculated palaeoslope value.

In this study, a vertical section between shell bed 5 and9A (Fig. 7) has been analysed. The resulting calculated

palaeoslopes range from 0.002 to 0.01, with the largestrecorded palaeoslope value in PU 8A. Generally, valuesrange between 0.002 and 0.006.The calculated slope whichoccurs directly above or below a shell bed is consistently0.002^0.003.These slope data are typical for gravely, allu-vial bed-load streams in modern river systems (Church &Rood, 1983; Andrews, 1984; Paola & Mohrig, 1996; Helleret al., 2003), andwater lain conglomerates close to sea level(Burns et al., 1997). Presented with the calculated palaeo-slope data are clast size (both average and median), andaverage channel depth data from each locality. Given thaterrors on such data can be a factor of 2, it is possible thatthere is no underlying pattern in the palaeoslope data;however, there is a notable recurring variation in the pa-laeoslope that is most pronounced in the lower two PUs(PU 6 and 7), which is associated with a grainsize trend.In this study, this variation is considered to be a real featurein the data that re£ects a variation in the dynamics of thealluvial section.

The change in palaeoslope andgrainsize is best observedin PUs 6 and 7. In these units there is an initial increase infollowed by a decrease in slope. In PU6, the slope increasesfrom 0.002 to 0.005, then decreases to 0.004, and subse-quently 0.003. In PU 7, the initial slope is not recorded,although it is likely to be in the region of 0.002^0.003 atthe shell bed (as is consistently recorded close to shell beds,and as is similar to the ¢nal slope in PU 6).The slope thenincreases to 0.004 and to 0.005, and subsequently decreasesonce more beneath shell bed 7.This variation is accompa-nied by a similar trend (i.e., increase then decrease) in grain-size. A similar pattern with an extremely high palaeoslopecompared with other data is recorded in PU 8A (betweenshell bed 7 and 8A). Above shell bed 8A the pattern is lessclear, and the data are less well constrained owing to a lackof good sites to perform such analyses.

It is possible to relate the alluvial and the submarine de-positional architecture.The relative position of the reduc-tion in alluvial slope in relation to the observed marinearchitecture has been considered. If the vertical heightabove the shell bed at which the peak alluvial slope occursis recorded as a percentage of the measuredvertical sectionbetween two shell beds, and this values is compared withthe position of clinoform nucleation in the marine section(by percent area shoal water vs. Gilbert delta) there is anapproximate correlation (Table 3). This implies that thetiming of clinoform nucleation and a reduction in deposi-tional slope of the alluvial fan occurs at a similar time, andcould potentially be related to the same control (Fig. 8).

DISCUSSION

Causes of clinoform nucleation

Our analysis of the facies architecture of the Loretocoarse-grained deltas demonstrates that the majority ofPUs are characterised by a distinctive down-transport var-iation in stratal geometry.Of the11PUs documented, eight

Fig. 6. Plot of foreset height and amount of progradation foreach of the transitional delta progradational units (PUs). PUs 6and 7 show a distinct point at which the foreset height increasesmore rapidly with distance, which is correlated to a change fromoblique to sigmoidal clinoform geometry.This records progra-dation to aggradation of the shoreline.The other units are lesswell constrained from direct ¢eld measurements, although asimilar change in clinoform geometry is observed.



show transition from shoal-water to Gilbert-delta archi-tecture within the same PU. Nucleation of Gilbert-deltaclinoforms is followed by continued palaeoseaward in-

crease in foreset height. This geometric transition is afunction of a change in bathymetry into which the deltasprograded. Shoal-water deltas are, byde¢nition, depositedin water depths of a few to several meters (Postma, 1990),whereas Gilbert-deltas form in deeper water, with theheight of the foresets recording the depth of water at thedelta front (Colella et al., 1987; Colella, 1988a; Postma,1990, 1995).

Foreset height is a function of palaeo-bathymetry at thedelta front; therefore, we can determine approximately thedepth of water into which the deltas progressively pro-graded for individual progradation units. The height ofthe foreset unit as measured from the topset to the bot-tomset for a single clinoform represents the depth of waterinto which that clinoform was deposited. At the start of aPU, deposition of a shoal-water delta requires that thewater depth was only a few meters (e.g. Postma, 1990).Thenucleation of clinoforms commences with oblique fore-set^topset geometry and foreset heights of not more than5m.This indicates awater depth of 5m intowhich theGil-bert-delta front prograded initially.The increase in foresetheight as Gilbert-delta foreset units are traced palaeo-

Table 3. Relationship between the magnitude of shoal-waterand Gilbert-delta progradation and the position of the maxi-mum slope achieved in the alluvial section

Cappingshell bed

Shoalwater%

Gilbert%

Maximumslope%

6 48 52 477 (min) 33 (max) 67 558A 40 60 378 3 97 819A 62 38 60

The percentage of shoal water andGilbert-delta deposition is given as apercentage of the total area (i.e. % shoal-water5 shoal-water area/(shoal-water area1Gilbert-delta area) of the submarine portion of the deltas.The maximum slope position is given as a percentage of the total heightof the measured alluvial section for that unit.This shows a reasonablecorrelation between the maximum slope and the nucleation of clino-forms, i.e. the palaeoslope begins to reduce (back-tilting signal) aroundthe time that clinoforms nucleate within a unit.

Fig.7. (a) A graphical representation of the palaeoslope data including the clast size and channel depth data.Median and averagegrainsize calculated from over100 clast measurements. Note how between shell beds 5^6; and 6^7 there is an increase and subsequentdecrease in both clast size andpalaeoslope.Above shell bed7 there is a greater variation in clast size (median) and calculated palaeoslope.These data are considered in relation to the correlated logged sections (b), and in the vertical measured sections between shell beds 5and 7 (c), and 7 and 9 (d).


E. Mortimeret al.

seawards records a progressive increase in the depth ofwater at the delta front, with depths reaching up to 25min PUs 6, 7, 9A and 9.

This recurring pattern of down-transport increase inforeset height, and clinoform nucleation from a shoal-water delta is di⁄cult to explain in a basin characterisedby a back-tilting subsidence pro¢le typical of half-grabenbasins. Two plausible explanations can be put forward:either that there is a spatial increase in bathymetry basin-ward; or that there is a temporal increase in bathymetry(accommodation creation) during progradation.

An increase in the water depth into which the Loretobasin deltas prograded can be generated spatially by: (1)progradation onto a palaeoseaward, or fore-tilting, basingeometry (Fig. 9a); (2) progradation across an intra-basinsyn-depositional normal fault that controls the landwardedge of a foreset unit (Fig. 9b); and (3) progradation acrossthe clinoform breakpoint of the preceding PU (Fig.9c).

Basin tilt geometry

Numerical models of clinoform growth in fore-tilting ba-sins typically show palaeo-seaward increase in clinoformheight because the depositional surface is tilted palaeo-seaward during progradation, that is water depth increasespalaeo-seaward caused by the tilt (e.g. Steckler et al., 1999;Swenson et al., 2000). In our example, deltas progradeacross the drowned delta-top topography of underlyingPUs, which are marked by transgressive shell bed hori-zons.These shell beds would have formed very low-gradi-ent seaward-inclined surfaces. The very low gradient ofthese surfaces would have been maintained, or even re-

versed, because incremental back-tilting in the half-gra-ben setting of the Loreto basin would have reduced thepalaeo-seaward slope of the upper surface of a shell bed.These observations imply that delta progradation acrossthe transgressed tops of underlying deltaic units is unlikelyto be able to generate seaward increase in foreset height.Indeed, the back-tilting subsidence geometry would pre-dict that delta PUs should show a palaeo-seaward thinningarchitecture.

Clinoform initiation by intra-basin faulting

Colella (1988b) proposed that the initiation of Gilbertforesets in the Crati basin required the presence of a syn-depositional normal fault scarp at the landward edge of theforeset units. Similarly, clinoform nucleation in the Cor-inth rift,Greece, has been explained bydelta progradationacross palaeo-sea£oor fault scarps (Dart et al., 1994;Gawthorpe et al., 1994). In the Loreto basin, however, it isclear that the clinoform nucleation point does not coincidewith any observed intra-basin normal faults.This is rein-forced by the observation that clinoform nucleation doesnot occur at any spatially persistent location along adepositional dip-oriented transect (Fig. 5).

Clinoform initiation over the clinoform breakpoint of anunderlying delta unit

A third possible spatial control on clinoform nucleationwas investigated by comparing the position of the nuclea-tion point of foreset units to the position of the clinoformbreakpoint of an underlying delta (Fig. 9c). An inheritedclinoform break point from a previous episode of delta

Fig. 8. A cartoon relating the calculated palaeoslopes and the observed depositional architecture in the Loreto basin, in particular forunits of progradation in which Gilbert-delta (GD) clinoforms nucleate from initially shoal-water (SW) deposition.We note that themaximumpalaeoslope is achieved prior to the nucleation of clinoforms in the basin.As the clinoforms prograde their geometry changesfrom oblique to sigmoidal.This change in geometry of the clinoforms occurs while the palaeoslopes are reducing.This reduction inpalaeoslope is associatedwith a change in the grainsize (decrease)more than it is to variations in channel depth (seeFig. 6).Therefore thetransition to GDs from shoal-water deltas, and the nucleation of clinoforms is associatedwith a reduction in palaeoslope and grainsizewithin the basin.This implies that back-tilting was occurring during the deposition of the GDs within the basin.



progradation is likely to provide a step in the bathymetrysu⁄cient to generate an abrupt seaward increase in waterdepth, and hence the necessary topographic condition forclinoform nucleation. Growth of shoal-water deltas intoGilbert deltas by this mechanism has been observed insyn-rift strata of the Suez rift (Gupta et al., 1999). There,Gilbert-delta clinoforms are observed to nucleate at thebreakpoint of the underlying delta progradation unit. Anew forest unit is constructed against the delta front facecreated during the previous episode of delta progradation.

A comparison of locations of Gilbert-delta nucleationpoint with breakpoint positions of the underlying deltaicunit for theLoreto deltas demonstrates that only in one ex-ample does this mechanism hold (Fig. 5). This rules outpre-existing delta topography as a control onGilbert -del-ta nucleation in the Loreto basin.We note, however, that

inherited topography does in£uence the continued pa-laeo-seaward increase in the height of foreset units wherethey prograde across and overstep the underlying clino-form breakpoint. In PU 9, Gilbert-delta foresets nucleateand increase in height palaeo-seaward of the clinoformbreakpoint of underlying PU 9A.

In summary, our analysis leads to the conclusion thatclinoform nucleation and delta architecture in the Loretobasin must be because of a temporally varying mechanismcontrolling the depth of water at the shoreline rather thana spatial control. It is di⁄cult to explain observed geome-tries by variations in sediment supply for a number of rea-sons. Firstly, an increase in water depth at the delta frontcannot be generated by variation in sediment supply with-out a signi¢cant reduction or cessation in sediment supplyand hence a hiatus. During such a time the delta would beeroded and subsequently re-deposited onto the underly-ing topography.The nucleation of clinoforms in theLoretobasin follows continuous sedimentation, with no uncon-formity, erosion, or hiatus. Secondly, palaeocurrents with-in and between cycles are consistent (also noted byDorseyet al., 1995, 1997b), such that lobe switching is unlikely, andagain the transition occurs conformably and within a sin-gle unit of progradation.

Clinoform nucleation and growth caused byaccelerations in fault slip-rate

Our observations lead us to propose an alternative and no-vel mechanism to explain the transition in delta geometry.We suggest that the systematic evolution of shoal water toGilbert-delta architectures and the palaeo-seaward in-crease in foreset height can only be explained by the deltafront encountering increasing water depth during progra-dation.This increase is from a fewmetres (shoal-water del-ta) to up to 25m (the maximum height of foresets recordedfor the Gilbert-delta beneath shell bed 9). Importantly,there is a change from oblique to sigmoidal clinoform geo-metry, which indicates an increasing degree of topsetaggradation at the shoreline, and the generation of a con-cave-up shoreline trajectory. Our analysis of the alluvialsection reveals that the reduction in the palaeoslope (i.e.back-tilting within the basin) occurs at approximately thesame time as clinoforms nucleate at the delta front.

We propose that the systematic change in the architec-ture of individual PUs can only be generated by varyingthe rate of tectonically controlled accommodation creationduring progradation.This is achieved by varying the dis-placement rate along the basin-bounding Loreto fault.In order to generate the continued increase in foresetheight, the increasingly aggradational component of theclinoforms, and the concave-up shoreline trajectory, theamount of accommodation creation must continue to in-crease during progradation of a delta unit.This can only beachieved tectonically if the rate of displacement along theLoreto fault continues to increase (i.e. accelerates) duringdeposition.

Fig.9. Cartoon to illustrate the nucleation of clinoforms andincrease in foreset height in response to: (a) A fore-tilting basin£oor topography (note that this does not apply for the Loretobasin as it is a back-tilting half graben). (b) The presence of anunderlying fault creating topography at the delta front (e.g.Colella, 1988a, b). For this to be the explanation a consistentspatial relationship for the nucleation of clinoforms, and theobservation of an intra-basin fault would be expected. (c)Progradation across the clinoform breakpoint topography of theunderlying delta clinoforms.This is the case for one delta in theLoreto basin (unit 8 ^ see also Fig. 4), and for enhancing theforeset height in another (unit 9), but for no others.


E. Mortimeret al.

If the back-tilting geometry of the Loreto basin is con-sidered then the amount of tectonically controlled accom-modation space created will always be greatest in theregion of the hangingwall closest to the Loreto fault andzero at the fulcrum (the ESH in the Loreto basin).

If we envisage the Loreto basin subsiding with aconstant rate of subsidence, and hence constant rate ofdisplacement on the Loreto fault (Fig. 10a), we wouldobserve a palaeo-seaward decrease in accommodationcreation typical of half-graben. If lines of constant time in-tervals are added during subsidence, then an inherent fea-

ture of a constant rate of backtilting is that the amountof accommodation created between isochrons is thesame for any speci¢c location within the basin. If adelta progrades into the basin under this scenariothen the amount of accommodation created seawardof the delta front decreases into the basin. This generatesa convex-up or relatively £at shoreline trajectory if a con-stant rate of progradation is assumed.Toward the fulcrumthe amount of accommodation created is decreasing,thus the height of foresets deposited also reduces palaeo-seaward.

Fig.10. Shoreline trajectories of progradational units (SW, shoal-water delta; GD, Gilbert delta) that result from (a) constant and (b)accelerating basin subsidence. (a) (i) In this scenario the rate of subsidence is constant, therefore the amount of displacement betweeneach time interval along the fault (t1^t6) is equal. (ii) Accommodation creation (y-axis) in relation to distance basinward in the Loretobasin (x-axis).The amount of space created during any single time interval (dashed lines) reduces toward the fulcrum.The amount ofaccommodation created for any given time interval (distance between two dashed lines) is the same as for any other interval of equal time(displacement rate constant).Whenwe consider the shoreline of a delta prograding into the Loreto basin then the amount of space at thedelta front as it moves basinward through time is less than the amount created during the previous time interval. If we then consider theheight of the foreset units (equal to the water depth at the delta front) then thesemust decrease as the deltaprogrades in this example.Constantdisplacement rate generates a convex-up trajectory. (b) (i) In this second scenario, the rate of subsidence along the Loreto fault accelerates suchthat the amount of displacement between each time interval (t1^t6) increases from one to the next. (ii) Accommodation creation (y-axis)in relation to distance basinward in the Loreto basin (x-axis).The amount of space created during any single time interval (dashed lines)reduces towards the fulcrum.The amount of accommodation created (distance between two dashed lines) at a speci¢c horizontaldistance from the fault increases with each progressive time interval (displacement rate accelerates).Whenwe consider the deltaprograding into the basin basinward and through time the amount of space created at the delta front is greater than the previous timeinterval despite the delta prograding toward the fulcrum.The e¡ect this has on foreset height is that they should increase basinward andtherefore aGDnucleates (clinoforms) fromSWdeltas as they prograde.Accelerating displacement rate generates concave-up shoreline trajectory.We have assumed a constant rate of progradation, however variation in sediment supply because of storage would reduce the rate ofprogradation as accommodation increases and therefore the deltawould in e¡ect slowdown or aggrade more rapidly (dashed grey line) inthe case of accelerating displacement.



Progressively palaeo-seawarddecreasing heights of fore-set units and convex-up shoreline trajectories are not typi-cally observed in the Loreto basin deltas thus discountingthis scenario. Only one example, PU1, exhibits an initiallyaggradational and later progradational trajectory.

A second scenario may be envisaged in which accelera-tion in the rate of tectonically controlled subsidence issuperimposed onto the back-tilting geometry of the Lore-to basin (Fig. 10b). In this scenario, the overall amount ofaccommodation created for any given isochron decreasestoward the fulcrum, as is anticipated for a back-tilting ba-sin.At anygiven point in the basin, however, the amount ofaccommodation created increases between isochronsthrough time (i.e., the acceleration in displacement rateleads to an overall increase in the amount of accommoda-tion creation through time for any given point in the ba-sin). A delta prograding in this scenario with (forsimplicity) a constant progradation rate will therefore en-counter an increase in the amount of accommodationcreation ahead of the delta front as the rate of subsidenceincreases. As a result, the water depth at the delta front in-creases through time, leading to an increase in the heightof the foreset units thus resulting in a concave-up shore-line trajectory. Only acceleration in the displacement ratecan generate the continued basinward increase in waterdepth ahead of the prograding delta front in a backtilting(i.e., accommodation creation reducing basinward) basin.In this second scenario, the predicted shoreline trajectoryand seaward increase in foreset heights generated are simi-lar to those observed in the Loreto basin.

Although not conclusive, the observed variation in bothgrainsize and palaeoslope of the alluvial section in PUs 6, 7and 8A also appear to be associated with variations in thedegree of backtilting during progradation of these units.Initially as a delta begins to prograde, the alluvial slope in-creases and coarse material is transported out to the deltafront. However, as backtilting increases during displace-ment rate acceleration, the palaeoslope is reduced and coar-ser material becomes trapped in more proximal regions.

In summary,we argue that the dominant control on del-ta architecture in theLoreto basin is variation in the rate oftectonically controlled subsidence. Speci¢cally, we pro-pose that repeated episodes of acceleration in displace-ment rate on the basin-bounding Loreto fault causedvariations in the rate of basin subsidence. The e¡ects ofeustatic £uctuations and/or tectonic uplift/subsidencecaused by neighbouring faults have not been included,but naturally mayhave modi¢ed this pattern of deposition.

Conceptualmodel for the development ofLoreto basin architecture

A conceptual model for the development of the deposi-tional architecture observed in the Loreto basin is illu-strated in Fig. 11. In this model, the transition in deltaarchitecture is a response to acceleration in the displace-ment rate on the basin-bounding Loreto fault during anepisode of delta progradation. Progradation commences

onto the drowned delta top of the preceding cycle.The un-derlying drowned delta top is marked by a shell bed depos-ited during maximum £ooding of the delta top, formedduring the maximum rate of back tilting (maximum rateof displacement on the Loreto fault).The initial prograda-tion represents a period during which there is a reductionin the displacement rate and an overall ‘recovery’of the de-positional system from the previous maximum displace-ment rate. This initial period of delta progradation ischaracterised by slight aggradation of the shoal-water deltafacies.This is interpreted to represent the trapping of se-diment in the proximal regions as space created in the im-mediate hangingwall must ¢rst be ¢lled before the alluvialfan slope can increase su⁄ciently to transport coarse sedi-ment.This initial phase is short-lived, and the shoal-waterdelta then begins to prograde. During this time it is envi-saged that the fault is active but that displacement rates arerelatively low.

Deposition of a shoal-water delta continues with gentlyinclined foresets of up to131 prograding into shallowwater(o5m) at the delta front. Progradation continues as sedi-ment is no longer trapped in the proximal regions andcoarse-grained sediments are supplied across the deltatop from the proximal regions. The alluvial slope mean-while increases toward a maximum. As the rate of subsi-dence, i.e., displacement rate on the Loreto fault, beginsto accelerate, the bathymetry at the delta front also in-creases, with the result that the steepness and height ofthe foreset units increases. At this point, shoal-water deltaforesets are no longer sustainable, the depositional angleincreases initially to 181, and subsequently to 20^281 asGilbert foresets nucleate.

As the displacement rate on the fault continues to in-crease, the amount and rate of accommodation createdahead of the Gilbert delta also increases with the conse-quence that the bathymetry intowhich the deltas progradealso deepens.This results in a progressive increase in theheight of the foreset units.This is coupledwith a reductionin the palaeoslope of the alluvial facies that is due to theincreased back-tilting and a greater amount of accommo-dation creation in the region close to the fault. Coarser-grained sediment begins to become trapped in the proxi-mal regions,which in turn causes a reduction in the supplyof coarse material to the delta front. Coupling these two ef-fects (increasing accommodation at the delta front, and areduction in sediment supply), results in the geometry ofthe clinoforms becoming increasingly aggradational, anda ¢ning-up on the foresets.

Eventually sediment supply to the delta front is over-come, and the delta top is drowned,with the shoreline mi-grating up to 2 km landward. This transgression ismost likely to occur at the maximum, or close to maxi-mum, displacement rate along the Loreto fault and leadsto a period of depositional hiatus or condensation and thedeposition of a shell bed that is often reworked. Subse-quently, the rate of displacement is reduced, and prograda-tion of the next unit of coarse-grained delta progradationcommences.


E. Mortimeret al.

In themodel ofDorsey etal. (1997b), deltas prograded onlyduring times that the fault was inactive. In our new model,deltas prograde during fault activity, and become drowned

during times of maximum displacement accumulation.Theobservation of repeated episodes of transition from shoalwater to Gilbert deltas in the stacked succession of Loreto

Fig.11. Cartoons illustrating the development of observed depositional architectures in theLoreto basin. (a)The study area falls withina region1500m basinward of the Loreto fault extending 2500m.The calculated amount of back-tilting required to explain the thicknessof the sectionwe have measured that comprises units 1^9 (378m) in the most proximal region is 3.31 total rotation about the fulcrum(assuming the fulcrum to be the ESH that is 6.5 km from the measured section). It is therefore clear that for an individual progradationalunit, the degree of rotation and hence back-tilting is small.This series of cartoons (b) therefore greatly exaggerates the amount of back-tilting in order to demonstrate the development of the architectures, with a vertical exaggeration of more than10 times (b, i^iii).Progradation is renewed onto the underlying depositional unit.The slope of the alluvial section is initially low but increases up sectionas the system‘recovers’ from the previous sediment storage and back-tilting. During these initial stages, this leads to a short episode ofaggradation of the shoal-water delta and then progradation as the sediment supply increases. (iv^v) The rate of subsidence increasealong the Loreto fault, generating an increase in water depth at the delta front.The height of the foresets increases as does theirdepositional angle in response to this increase in water depth.With continued increase in subsidence rate, the water depth at the deltafront increases to a point where shoal-water delta foresets cannot be sustained, a greater angle of deposition is reached, and clinoformsnucleate in the basin. During this time the alluvial slope has achieved its maximum, and begins to reduce as the rate of subsidenceincreases. (vi^viii) The rate of subsidence increases further, the rate of increase inwater depth at the delta front is therefore more rapid.Foreset heights increase and their geometry changes from oblique to clinoform as they aggrade to keep pacewith the creation of space atthe delta front.This aggradation is enhanced by the trapping and storage of sediment in the proximal region because of back-tilting, andhence a reduction in sediment supply to the delta front.This is re£ected also in a reduction in the slope of the alluvial system.Eventually,sediment storage in the proximal region, and the rate of increase in water depth at the delta front (both controlled by ever more rapidsubsidence) exceed the amount of sediment being transported to the delta front, and the delta top is transgressed.During this period ofdepositional hiatus a shell bed is deposited (ix) (c). Relationship between alluvial slope, sediment architecture, and tectonicallycontrolled subsidence in the Loreto basin.



deltas with each progradation unit showing progressive pa-laeo-seaward increase in foreset height leads us to proposethat delta progradation was governed by repeated episodesof accelerating displacement rate on theLoreto fault.

Quantifying variations in fault displacementrate

TheLoreto basin provides an excellent example of a tecto-nically controlled setting where it is possible to undertakea quantitative analysis of the longer-termvariation in faultdisplacement rates.There are obvious limitations to suchcalculations especially using assumptions of constant se-dimentation rates and, as has been undertaken here, areaof sediment as an approximation for the duration for thatdeposition. It is, however, an interesting undertaking be-cause it enables an, all be it crude, assessment of the varia-tion in displacement rate required to generate theobserved depositional architecture for a single unit of deltaprogradation.

In order to determine the approximate duration of de-position of the Gilbert and shoal-water delta componentsof progradation units, we determined the 2-D areas ofthese delta components (measured parallel to transportdirection).To approximate this duration, we used the pro-portion of the area that is a shoal-water andGilbert-deltasand divided this by the duration of the PU.The accommo-dation accumulation during a shoal-water delta is esti-mated from the shell bed to the nucleation of the Gilbert-delta foresets and is found to increase by no more than 3m(assuming up to 2-mwater depth of the shell bed, and ¢rstobserved Gilbert-delta foresets in water of ca. 5-m depth).The accommodation accumulation for the Gilbert deltasis calculated as the maximum height of the foresets.Table 4 provides a summary of quantitative data for each

of the transitional units of coarse-grained delta prograda-tion generated.These data reveal that the increase in dis-placement rate during the deposition of a Gilbert deltais ca. 1.5^10 times the displacement rate calculated duringthe deposition of a shoal-water delta in most PUs. Thedevelopment of shoal-water deltas that are subsequentlydrowned (i.e. PUs showing no transition in architecture)could either be the result of eustatic variations, or extre-mely rapid increases in the rate of tectonic subsidencesuch that the delta front is rapidly drowned and a Gilbertdelta not formed.

Implications for our understanding of faultbehaviour

We have argued that the development of transitional PUs,the nucleation of clinoforms, and the concave-up shore-line trajectories observed in delta PUs within the Loretobasin are controlled by variations in the rate of tectonicallycontrolled accommodation creation by the basin-bound-ing Loreto fault. Speci¢cally, our observations of the sedi-mentary architecture of the Loreto deltas require thatactivity on the Loreto fault was episodic, and that duringperiods of fault activity, the displacement rate accelerated.Variation in (a) the inferred duration (Table 1), (b) the pro-gradational extent (Fig. 5) and (c) the location of clinoformnucleation of particular PUs (Fig. 5), probably re£ects dif-ferences in the timing and character of displacement accu-mulation (e.g. the maximum displacement rate achieved)during each unit.

Episodic fault activity is predicted by numerical modelsthat simulate the nucleation, growth and elastic interac-tion of a population of extensional faults (Cowie, 1998). Inthis model, earthquake rupture on one fault modi¢es thesurrounding stress ¢eld and thus may either enhance or

Table 4. An approximation for the change in rate of accommodation relation during deposition of a transitional unit

Unit GD duration SW duration GD rate (mmyr�1) SWrate1 (mmyr�1) SWrate2 (mmyr�1) GD/SWrate1 GD/SWrate2

2 8210 11570 1.7 0.2 0.3 8.5 5.76 9100 3310 2.9 0.6 0.9 4.8 3.27 12120 3010 2.6 0.7 1.0 3.7 2.68A 5140 3390 2.2 0.6 0.9 3.7 2.48 4140 900 2.9 2.2 3.3 1.3 0.99A 3940 3820 4.8 0.5 0.8 9.6 6.09 7390 750 3.7 2.7 4.0 1.4 0.9

GD duration5 (area GD/total submarine area) � duration of unit (years.SW duration5 (area SW/total submarine area) � duration of unit (years).GD rate5GDmaximum foreset height/GD duration.SWrate15 2000mm/SWduration (see note).SWrate25 3000mm/SWduration (see note).Note:This is an approximation for either a 2 or 3m increase inwater depth at the delta front between a shoal-water delta being deposited and clinoformnucleation.This assumes deposition of the shoal-water delta to be initially into less than 5m ofwater (asGilbert foresets ¢rst occur at ca. 5mwater depth).The reader is reminded that units 8 and 9 are anomalous as the ¢nal height of the foreset units (PU 9) or nucleation (PU 8) is a¡ected by the underlyingdelta morphology. Rates are given at the delta front, andwould be greater at the fault; however, the change in rate is comparable.This approximation isbased upon the proportional area of each type of submarine delta facies association (Gilbert delta (GD) or shoal-water delta (SW)) as a proxy for theduration of progradation assuming constant sediment supply.Height of the foreset units is used to determine the increase inwater depth at the delta frontduring progradation, and therefore the amount of tectonically controlled accommodation creation at the delta front.When the SWandGD rates arecompared, we observe an increase of between two and nine times the rate of accommodation creation within a single unit of progradation.


E. Mortimeret al.

relax the stress loading neighbouring structures depend-ing on their relative location and orientation. The e¡ectof elastic interaction between faults successfully explainscumulative displacement pro¢les along faults (Willemseet al., 1996; Gupta & Scholz, 2000) as well as many exam-ples of earthquake rupture sequences (see King & Cocco,2001, for a review). In the context of fault growth over time,the elastic interaction model predicts temporal variationsin strain partitioning between neighbouring faults.Thesetemporal variations arise spontaneously within a popula-tion of growing faults because a complex history of stressinteraction emerges.Thus a fault may become inactive orbe re-activated at any stage in the evolution of the fault po-pulation because of its position and orientation relative toother active faults (Cowie, 1998).

Considering the structural setting of the Loreto fault inthe transtensional Gulf Extensional Province, Gulf of Ca-lifornia, there is obvious potential for interaction betweenthe Loreto fault and neighboring strike^slip faults as wellas extensional faults (Lewis & Stock, 1998); however, themost likely candidate for interaction with the Loreto faultis an extensional fault that lies to the east,which controlledthe growth of the Eastern Structural High (Dorsey et al.,1997a). There are two lines of evidence indicating thatESH fault (Fig. 1a) was active during the deposition of theLoreto basin sedimentary in¢ll. Firstly, south west pro-grading carbonate deltas, sourced from an uplifted andsub-aerially exposed ESH (Dorsey et al., 1997a), are timeequivalent to the upperPUs documented in this study.Thisimplies that the ESH was forming (although was not ex-posed) during deposition of the investigated PUs and thatthe fault bounding this structure was active during thistime. Secondly, palaeocurrent directions measuredthroughout the studied section show a consistent trendthat is not perfectly perpendicular to the Loreto fault (i.e.ENE) but NNE (Mortimer, 2004).This suggests that thedepositional low controlling sediment transport is pro-duced by an interaction between displacement on the Lor-eto fault and growth of the ESH footwall to the east. Atime-varying degree of partitioning of deformation be-tween these structures therefore o¡ers a potential me-chanism to explain episodic displacement addition on theLoreto fault. The inferred average duration for the PUs(Table 1) would suggest that, on average, activity switchedbetween these faults at ca.10 kyr intervals.

The acceleration that occurs on the Loreto fault duringa period of increased fault activity might be explained bythe frictional properties on the fault (see Scholz, 2002,Chapter 2, for a review). In this scenario, during episodesof quiescence the fault is inactive and undergoes ‘healing’.The static strength of the fault is increased, and is slightlymore than it is when the fault is active. As deformation issubsequently accommodated on the Loreto fault and sub-sidence resumes, slip on the Loreto fault occurs initiallyhaving to overcome a higher level of friction. As the fric-tional level evolves to a lower value, so the rate of slip onthe fault increases.The ¢nite time period required for theevolution of these frictional properties may explain the

gradually accelerating fault slip observed on the Loretofault.The duration of quiescence would a¡ect the ‘healingtime’, therefore variations between PUs could result fromdi¡ering frictional evolution during each episode of faultactivity.

SUMMARYAND CONCLUSIONS

The Loreto basin depositional architecture provides anexceptional setting in which to investigate whether varia-tions in tectonic displacement are recorded in the deposi-tional architecture. This study has documented a set ofstacked coarse-grained delta PUswith an average durationof 14.1 � 8.6 kyr that exhibit a recurring transition from ashoal-water to Gilbert-delta geometry within a singlePU, each showing a concave-up shoreline trajectory. Inthe eight PUs inwhich such transitions occur the durationof the units ranges between 8.5 � 3.5 and15.1 � 6.2 kyr.

This previously unrecorded transition in delta architec-ture and nucleation of clinoforms is most simply gener-ated by progressively increasing the water depth intowhich the deltas are prograding.This cannot be explainedin the Loreto basin by the basin’s backtilting subsidencepro¢le, or by the presence of syn-depositional intra-basinnormal faults or inherited topography created by an un-derlying clinoform break point.

We conclude that:

1. The frequency of PUs, the observed shoreline trajec-tory, and the systematic transition from shoal-water toGilbert-delta geometry within individual PUs is bestexplained by episodic variations in basin subsidencerate generated by repeated episodes of accelerating dis-placement accumulation on the basin-bounding Lore-to fault.

2. The depositional architecture, speci¢cally the nuclea-tion of clinoforms, can be used to distinguish a tectoniccontrol within the Loreto basin. In the absence of abasinward increase in bathymetry, recognition of thisgeometry could be used to invoke a tectonic control inother basins.

3. Clinoform nucleation in the Loreto basin requires thatdisplacement on the Loreto fault is episodic and accel-erates up to two to ¢ve times the rate at the start ofdeposition of a PU (fromo0.5^2.2 to1.7^4.8mmyr�1).

4. The long-term displacement history on theLoreto faultwas characterised by punctuated episodes of heigh-tened fault activity during which the displacement rateaccelerated.

ACKNOWLEDGEMENTS

EstelleMortimer was funded byNERC studentshipNER/S/A/2000/03349. PatienceCowie was funded by aRoyal So-ciety University Research Fellowship. Gupta was fundedby the Natural Environment Research Council GrantGR8/04397. Dating of tu¡ samples was funded by aNERC



‘Isotope Facilities Grant’ grant.We wholeheartedly thankR. Dorsey, for an introduction to the Loreto basin in the¢eld, for continued inspiring discussions, and for accessto unpublished ¢eld data.Our thanks are extended toMal-colm Pringle (MIT, formerly SUERC) for his assistancewith the dating of tu¡ samples and discussion of age inter-pretations. Many thanks to Douglas Paton for his assis-tance in the ¢eld.We are extremely grateful to C. Paola, R.Slingerland and N. Christie-Blick for their constructivereviews of this manuscript; and Ruth Robinson for valu-able discussions.

REFERENCES

Abbott, S.T. (1997) Mid-cycle condensed shell beds from mid-Pleistocene cyclotherms, New Zealand: implications for mid-cycle sequence architecture. Sedimentology, 44, 805^824.

Allen, P.A. & Densmore, A.L. (2000) Sediment £ux from anuplifting fault block.Basin Res., 12, 367^380.

Andrews, E.D. (1984) Bed-material entrainments and hydraulicgeometry of gravel-bed rivers inColorado.GSABull.,95, 371^378.

Burns,N.,Heller,P.L.,Marzo,M.&Paola,C. (1997)Fluvialresponse in a sequence stratigraphic framework: example fromtheMontserrat fan delta, Spain. J. Sedimen. Res., 67, 311^321.

Church,M. &Rood,K. (1983) Catalogue of Alluvial River Chan-nel Regime Data. Department of Geography,Vancouver, BC.

Colella, A. (1988a) Fault-controlled marine Gilbert-type fandeltas.Geology, 16, 1021^1034.

Colella, A. (1988b) Pliocene^Holocene fan deltas and braiddeltas in the Crati Basin, southern Italy: a consequence ofvarying tectonic conditions. In: Fan Deltas: Sedimentology andTectonic Setting (Ed. by W. Nemec & R.J. Steel), pp. 50^74.Blackie, London.

Colella,A.,De boer,P.L.&Nio,D. (1987)Sedimentology of amarine intrmontane Pleistocene Gilbert-type fan-delta com-plex in the Crati Basin, Calabria, southern Italy. Sedimentology,34, 721^736.

Colella,A.,De boer,P.&Nio,S.D. (1997) Sedimentology of amarine intermontane Pleistocene Gilbert-type fan-deltacomplex in the Crati Basin, Calabria, S. Italy. Sedimentology,34, 721^736.

Cowie, P.A. (1998) A healing-reloading feedback control on thegrowth rate of seismogenic faults. J.Struct.Geol., 20,1075^1087.

Dart, C., Collier, R., Gawthorpe, R., Keller, J. & Ni-

chols,G. (1994) Sequence stratigraphy of Pliocene^Quatern-ary synrift Gilbert-type fan deltas, northern Peloponnesos,Greece.Mar. Petrol.Geol., 11, 607^614.

Dorsey, R.J. & Kidwell, S.M. (1999) Mixed carbonate-silici-clastic sedimentation on a tectonically active margin: examplefrom the Pliocene of Baja California Sur, Mexico. Geology, 27,935^938.

Dorsey, R.J., Stone,K.A. &Umhoefer, P.J. (1997a) Stratigra-phy, sedimentology, and tectonic development of the south-eastern Pliocene Loreto Basin, Baja California Sur, Mexico.GSASpec. Pap., 318, 83^109.

Dorsey, R.J. & Umhoefer, P.J. (2000) Tectonic and eustaticcontrols on sequence stratigraphy of the Pliocene Loretobasin, Baja California Sur,Mexico.GSABull., 112, 177^199.

Dorsey, R.J.,Umhoefer, P.J. & Falk, P.D. (1997b) Earthquakeclustering inferred from Pliocene Gilbert-type fan deltas in

the Loreto basin, Baja California Sur, Mexico. Geology, 25,679^682.

Dorsey, R.J.,Umhoefer, P.J. & Renne, P.R. (1995) Rapid sub-sidence and stacked Gilbert-type fan deltas, Pliocene Loretobasin, BajaCalifornia Sur,Mexico.Sediment.Geol., 98,181^204.

Ethridge, F. & Wescott,W. (1984) Tectonic setting, recogni-tion and hydrocarbon potential of fan-delta deposits. Canad.Soc. Petrol. Geol.Mem., 10, 217^236.

Falk,P.D. (1996) Sedimentology, stratigraphy, andGilbert-deltasedimentation in the southern Pliocene Loreto basin, BajaCalifornia Sur,Mexico.MastersThesis, University of Oregon.

Falk, P.D. &Dorsey, R.J. (1998) Rapid development of gravellyhigh-density turbidity currents in marine Gilbert-type fandeltas, Loreto basin, Baja California Sur, Mexico. Sedimentol-ogy, 45, 331^349.

Gawthorpe, R., Fraser, A. & Collier, R. (1994) Sequencestratigraphy in active extensional basins: implications forthe interpretation of ancient basin ¢lls. Mar. Petrol. Geol., 11,642^658.

Gilbert, G.K. (1885) The topographic features of lake shores.USGeol. Survey, 69^123.

Gupta,A.A.&Scholz,C.H. (2000)Fault interactions and seis-mic hazard. J. Geodyn., 29, 459^467.

Gupta, S., Underhill, J.R., Sharp, I. & Gawthorpe, R.(1999)Role of fault interactions in controlling synrift sedimentdispersal patterns: Miocene Abu Alaqa Group, Suez Rift,Sinai, Egypt.Basin Res., 11, 167^190.

Heller, P.L.,Dueker,K. &Mcmillan,M.E. (2003) Post-Pa-laeozoic alluvial gravel transport as evidence of contintentaltilting in the U.S. Cordillera.GSABull., 115, 1122^1132.

Kidwell, S.M. (1989) Stratigraphic condensation of marinetransgressive records: origin of major shell deposits in theMiocene ofMaryland.Geology, 97, 1^24.

King, G.C.P. & Cocco, M. (2001) Fault interaction by elasticstress changes: new clues from earthquake sequences. Adv.Geophys., 44, 1^38.

Leeder,M.,Harris,T. &Kirkby,M.J. (1998) Sediment supplyand climate change: implications for basin stratigraphy. BasinRes., 10, 7^18.

Lewis, C. & Stock, J.M. (1998) Late Miocene to recent trans-tensional tectonics in the Sierra SanFermin, northeasternBajaCalifornia, Mexico. J. Struct. Geol., 20, 1043^1063.

Lisiecki, L.E. & Raymo,M.E. (2005) A Plio^Pleistocene stackof 57 globally distributed benthic d18O records. Paleoceanogra-phy, 20, PA1003, doi: 10.1029/2004PA001071.

Lonsdale, P. (1989) Geology and tectonic history of the Gulf ofCalifornia. In: Geology of North America, N (Ed. by E.L.e.a.Winterer), pp. 499^521. Geological Society of America,Boulder, CO.

Lowe,D.R. (1976)Grain £ow and grain £owdeposits.J.Sediment.Petrol., 46, 188^199.

Lowe,D.R. (1982)Sediment gravity £ows: II.Depositional mod-els with special reference to the deposits of high-densityturbidity currents. J. Sediment. Petrol., 52, 279^297.

Marr, J.G.,Swenson, J.B.,Paola,C.&Voller,V.R. (2000)A 2di¡usion model of £uvial stratigraphy in closed depositionalbasins.Basin Res., 12, 381^398.

Mclean (1989)Reconnaissance geological map of the Loreto and part ofthe San Javierquadrangles. Baja California sure,Mexico, USGS.

Milankovitch, M. (1930) Mathematische Kilmalehre and Austro-nomische Theorie der Klimaschwankungen Gebruder. Borntreger,Berlin, 176pp.


E. Mortimeret al.

Mortimer, E. (2004) Tectonic controls on the growth of coarse-grained delta clinoforms in the Pliocene Loreto Basin, BajaCalifornia Sur, Mexico. DoctorateThesis, University of Edin-burgh, Edinburgh, Scotland.

Mulder,T. & Alexander, J. (2001) The physical character ofsubaqueous sedimentary density £ows and their deposits.Sedimentology, 48, 269^299.

Naish,T.R.,Kamp, P.J.J. & Pillans, B. (1997) Recurring globalsea-level changes recorded in shelf deposits near theG/Mpo-larity transition,Wanganui basin, New Zealand: implicationsfor rede¢ning the Pliocene-Pleistocene boundary. Quat. Int.,40, 61^71.

Nemec,W. (1990) Aspects of sediment movement on steep deltaslopes. In:CoarseGrainedDeltas (Ed. byA.Colella &D.B. Prior),IASSpe. Publ, 10, 29^73.

Nemec,W. & Steel, R.J. (1984) Alluvial and coastal conglomer-ates: their signi¢cant features and some comments on gravellymass- £ow deposits. In:Sedimentology of Gravels and Conglomer-ates (Ed. by E.H. Koster & R.J. Steel),Memoir, 10, 1^31.

Orpin, G.J. & Reading, H.G. (1993) Variability of deltaic pro-cesses in terms of sediment supply, with particular emphasison grain size. Sedimentology, 43, 927^945.

Paola, C. & Mohrig, D. (1996) Palaeohydraulics revisited: pa-laeoslope estimation in coarse-grained braided rivers. BasinRes., 8, 73^90.

Postma,G. (1990) Depositional architectures and facies of riverand fan deltas: a synthesis. In: Coarse Grained Deltas (Ed. by A.Colella &D.B. Prior), IASSpec. Publ., 10, 13^27.

Postma, G. (1995) Sea-Level related architectural trends incoarse-grained delta complexes.Sedimen. Geol., 98, 3^12.

Postma, G. & Roep, T.B. (1985) Resedimented conglomeratesin the bottomsets of Gilbert-type gravel deltas. J. Sediment.Petrol., 55, 874^885.

Raymo, M.E., Ganley, K., Carter, S., Oppo, D.W. &Mcma-

nus, J. (1998) Millennial- scale climate instability during theearly Pleistocene epoch.Nature, 392, 71^98.

Scholz, C.H. (2002) Mechanics, Earthquakes and Faulting. Cam-

bridge University Press, Cambridge.Sohn, Y.K. (1997) On traction carpet sedimentation. Sediment.

Geol., 67, 502^509.Sohn,Y.K.,Kim, S.B.,Hwang, I.G., Bahk, J.J., Choe,M.Y. &

Chough, S.K. (1997) Characteristics and depositional pro-

cesses of large-scale gravellyGilbert-type foresets in theMio-

cene Doumsan fan delta, Pohang Basin, SE Korea. J. Sediment.Res., 67, 130^141.

Steckler,M.S.,Mountain,G.S.,Miller,K.G.& Christie-

Blick,N. (1999) Reconstruction of Tertiary progradation and

clinoform development on the New Jersey passive margin by

2-D backstripping.Mar. Geol., 154, 399^411.Swenson, J.B.,Voller,V.R., Paola, C., Parker, G. & Marr,

J.G. (2000) Fluvio-detlaic sedimentation: a generalisedStefan

problem. Eur. J. Appl. Math., 11, 433^452.Umhoefer, P.J. & Dorsey, R.J. (1997) Translation of terranes:

lessons from Baja California,Mexico.Geology, 25, 1007^1010.Umhoefer,P.J.,Dorsey,R.J.&Renne,P.R. (1994)Tectonics of

the Pliocene Loreto basin, Baja California Sur, Mexico, and

evolution of the Gulf of California.Geology, 22, 649^652.Umhoefer, P.J. & Stone, K.A. (1996) Description and kine-

matics of the SE Loreto basin fault array, Baja California Sur,

Mexico: a positive ¢eld test of oblique-rift models. J. Struct.

Geol., 18, 595^614.Willemse, E.J.M., Pollard, D.D. & Aydin, A. (1996) 3-di-

mensional. analyses of slip distributions on normal fault arrays

with consequences for fault scaling. J.Struct.Geol.,18, 295^309.

Manuscript received 10 November 2004; Manuscript accepted1July 2005.



Clinoform nucleation and growth in coarse-grained deltas, Loreto basin, Baja California Sur, Mexico:...

Documents

Transcript of Clinoform nucleation and growth in coarse-grained deltas, Loreto basin, Baja California Sur, Mexico:...