TECTONICS, VOL. 19, NO. 2, PAGES 314-334 APRIL 2000
Quaternary convergent margin tectonics of Costa Rica,
segmentation of the Cocos Plate, and Central American
volcanism
R. von Huene, C. R. Ranero, and W. Weinrebe
GEOMAR ResearchCenter for Marine Geosciences,Kiel, Germany
K. Hinz
Bundesanstalt
f'tirGeowissenschaften
und Rohstoffe,Hannover,Germany
Abstract. Along Costa Rica, new geophysical data
indicate considerable control of Quaternary convergent
margintectonicsby the subducting
lowerplate.Threetypesof
oceancrustenter the subductionzone: (1) CocosRidge with
its underlyingthick crust stands2 km high, (2) on its north
flank is normal crust covered 40% by seamounts,and (3)
along the adjacentNicoya margin the underthrustcrusthas a
smoothsea floor. A 3- to 10-km-wide base of slope frontal
prism varieslittle oppositedifferent subductingcrustsexcept
where subductingseamountseroded it. Once the breaching
seamounthas passedthe prism it is quickly restored. The
effect of oceaniccruston continentalmargin structureis most
evidentin the middle and upper slope. Where CocosRidge
and its flanking seamountssubduct,erosion is pronounced
relative to the stable slope where smooth lower plate
subducts. Aligned upper plate featuresabove lower plate
segmentboundariesextendmore than 120 km landwardof the
trench axis and correspondin varying degreeswith volcanic
arc segmentation. The offset of volcanoesacrossthe Costa
Rica/Nicaraguaborder correspondswith a changein crustal
structureand depth of the lava source. Subductedsediment
showslittle correlationwith the slab signal in volcanic arc
lavas but the magnitude of faulting associatedwith ocean
plate flexureadjacentto the trenchaxis parallelsit well. Thus
fluidsin oceancrustfracturesandboundwater in serpentinite
may have a recognizablegeochemicaleffectin arc lavas.
1. Introduction
In principle,upperplate tectonismappearsto be drivenby
the characterand relief of the subductionlower plate, but to
what degree? This questionis difficult to answer without
isolating a variable convergence rate, slab dip, and
sedimentation
from
variable
ocean crustal
character.
The
Middle Americamarginoff CostaRica andthat off Nicaragua
convergewith the CocosPlate at equivalentrates;both have
sparsetrenchsediment,but they vary in crustalstructureand
the subductingoceaniccrust. This situationis optimal for
studyingconvergent
plateinteraction.
Copyright2tX)0by theAmericanGeophysical
Union
betweenland and marinevariability is indicatedby current
data. High-resolutiongeophysicaldatashowan alignmentof
structuresthat continue from the deep ocean well into the
continent. Mapping with multibeam echo soundersreveals
the contrastingseamount-covered
Cocos Plate abuttingthe
ruggedcontinentalslopeof central Costa Rica and a smooth
seafloorabuttingthe linear continentalslopeoff the Nicoya
Peninsula[Shipleyand Moore, 1986; van Huene et al., 1995].
When swathmap soundingsin eachsweepwere editedbeam
by beam [Heeren, 1996, and this study], the resolutionof
morphologyin subsequentdisplayssurpassedthat of most
other surface ship data and provided a high-resolution
Quaternarytectonicmap (Plate 1). Thesedatawere combined
with seismicreflection data [Hinz et al., 1996], wide-angle
seismic data [Ye et al., 1996; Stavenhagenet al., 1998;
Christesan et al., 1999], and refined regional magnetic
compilations[Barckhausenet al., 1998]. They provide a
regionalcontextin which to place new insightsfrom drilling
during Ocean Drilling Program(ODP) Leg 170 [Kimura et
al., 1997].
Once recognized,segmentscan be comparedto studythe
effectsof subductingrelief on forearcstructure. At the scale
of swathmap resolutionthe subductedlower plate relief is
sufficientto influence upper plate tectonism,and a linear
relief associated
with subducted
plate segmentboundaries
can
be tracedfar landward. The frontal prism and volume of
trenchsedimentindicate little changein subductedsediment
volumes. Quaternarylavas of the volcanicarc alongCentral
America have a highly varying geochemistrythoughtto be
derivedfrom the subductingslab. The slab signalis very
high in Nicaraguaand decreasesto almostnothingin central
CostaRica [i.e., Carr et al., 1990]. One previousexplanation
for that geochemistrywas a variable volume of sediment
bypassingan accretionarywedge and reachingthe area of
lava generation[Plank and Langmuire, 1993;Leemahet al.,
1994]. However, observationsthat accretionoff CostaRica is
far lessthan earlier studiesindicated[van Huene et al., 1995;
Hinz et al., 1996; Kimura et al., 1997] encouragedus to study
the extent of accretion more thoroughly. We found
differences in sediment input to the subduction zone
insufficient to explain a geochemical variability, so we
exploredotherpossiblephysicalcauses.
Papernumber1999TC001143.
0278-7407/00/1999TC001143
The variable morphologyof the Cocos Plate and Pacific
convergentmargin off CostaRica (Figure 1) has been known
for morethan40 years[Fisher, 1961], andthe variationin arc
volcanismwas notedevenearlier [Sapper,1897]. A relation
$12.00
314
VON HUENE
105øW
ET AL.: COSTA RICA TECTONICS
AND
VOLCANISM
100"W
95"W
90"W
85øW
80"W
100"W
95"W
90"W
85"W
80"W
20øN
15øN
10øN
5øN
O
105"W
Figure 1. Topexbathymetryand topographyof CentralAmericashowingthe plate tectonicsetting of the
studyarea. Rectangleoutlinesareain Plate 1. DSDP, Deep SeaDrilling Project.
We presenta compilationof morphologyand seismicdata
to showthe Quaternarytectonicstructureof the margin. On
the basisof this compilation,we follow segmentation
of the
lower plate acrossthe upperplate along structuralfeatures
thattransectthe slopeand shelf. Thesesegmentboundaries
are consistentwith coastaluplift and roughlythe sameas
somegeochemicalvariationsalong the arc. The physical
correlationfits no singlegeochemicalprocessbut shouldbe
further investigatedin the contextof a multicomponent
north (rough-smooth
boundaryof Hey [1977]). With more
completemagneticcoverageit was discoveredthat near the
continenta previouslyunrecognizedwedge of older CNS
crustis insertedbetweenEPR crustand the rough-smooth
boundary[Barckhausenet al., 1998] (Plate 1). The older
CNS crust appearsless affected by Galapagoshot spot
volcanismand has a "smooth"morphologylike that of EPR
crust. Thereforethe rough-smooth
morphologicalproxy for
CNS andEPR crustsbreaksdownnearthecontinent
(Figure'1
system
and Plate 1).
2. General Setting and Previous Marine Studies
Off CostaRica andNicaraguathe CocosPlatewasformed
Three morphotectonic
segmentscorrespond
with the CNS
magnetic anomalies: (1) Cocos Ridge, (2) a seamountcoveredsegmentadjacentto the ridge, and (3) the smooth
crustoff the NicoyaPeninsula(FiguresI and2 and Plate 1).
These crustal domains flex differently into the Middle
alongtheeasttrendingCocos-Nazca
spreading
center(CNS)
andthe north-south
EastPacificRise(EPR) spreading
center America Trench, and therefore the trench axis is as little as 1
[Hey, 1977].
The trace of the CNS-EPR crustal weld
extendingto CostaRica (Figure 1) separates
a roughCNS
seafloor to the south from the smoother EPR seafloor on the
km deep at the crest of CocosRidge, increasesto 3.5-km
depththroughthe seamount
province,is 4 km deepopposite
the smooth ocean crust, and exceeds 5-km depth off
316
VON HUENE
ET AL.: COSTA RICA TECTONICS
AND
VOLCANISM
IIIll•
JlO"
00'N
!:
.........................
ß
9 ø 30'N
9 ø 00'N
8 ø 30'N
83 ø 30'W
Figure 2. Tectonicmap of the study areaon the swath map base gridded at 500 m. SB denotessegment
boundaries,and CNS denotesCocos-Nazcaspreadingcentercrust [Barckhausenet al., 1998]. Triple junction
denotesfossiltriplejunctionwhichmergeswith rough-smooth
boundary(Plate I inset). The extentof Figure
5a is indicatedby the line along the Nicoya Peninsula. Small unmappedfracturesare shownin perspective
diagramsof Figure 5.
Nicaragua. Trench sedimentponds where seamountspass
throughthe trench axis (Plate 1 and Figure 2). The frontal
prism disrupted by a seamount is restored quickly to its
formerwidth even as the seamountis producinga debristrail
farther up slope. We use the term frontal prism rather than
accretionaryprism becausemuch of the material may consist
of slope sedimentthat was kneadedinto the prism without
being transferred from the oceanic plate as described in
section5.1. In the ODP Leg 170 areas,only compactedslope
sedimentwas reportedfrom the prism althoughfrom seJ,
smic
data in other areas, a transfer of some trench sediment
(excludingslumpsfrom the landwardslope)is likely. We use
the term margin wedge for consolidated rock (seismic
velocity,3.5 km/s andgreater)that formsthe backstop.
The morphology of the continental slope reflects the
smooth or rough character on the adjacent ocean plate.
Morphology is regular opposite the smooth ocean crust,
ruggedoppositethe seamountdomain, and consistsof a short,
steep incline to a raised shelf where the Osa Peninsula is
uplifted over the subductingCocosRidge crest. The age of
the currentplate configurationcould be 5-6 Myr. [Collins et
al., 1995; deBoer et al., 1995; Griife, 1998] as indicatedby
deformation and uplift across Costa Rica above the
subductingCocosRidge.
Seismic reflection records [Hinz et al., 1996] indicate a
basic structural configuration of the margin from the Osa
Peninsula to the middle of the Nicoya Peninsula and
Nicaragua as well [Ranero et al., this issue]. This
configurationconsistsof a margin wedge coveredby slope
sediment,underthrustby trench sediment,and fronted by a
small frontal prism. This basic structureprobably existed
prior to the current plate tectonic configuration and was
modifiedby subductionof CocosRidge. This basicstructure
enduredthe onslaughtof many subductingseamountsbut not
withoutshort-termdamage.
The upper surfaceof the margin wedge (the rough surface
of Shipley et al. [1992] and the Base of Slope Sediment
(BOSS) reflectionof Kimura et al. [1997]) is buriedby slope
VON HUENE ET AL.: COSTA RICA TECTONICS AND VOLCANISM
317
sedimentand has been interpreteddifferently. However, al., 1998]. Ocean Bottom Hydrophone (OBH) refraction
sinceLeg 170, it is generallyagreedthat the margin wedge is measurementsmidway betweenthe crest and QueposPlateau
composedof older oceanigneousand associatedsedimentary show a 12-km-thick crust [Stavenhagenet al., 1998] which
rock as was found off Guatemala.
thins away from the crest principally in the upper layered
crustal section (Figure 3d). Reflection records display
considerablelayering in the upper ocean crust (Figure 3),
3. Information Developed Since Earlier Cruises
suggesting
many extensivelava flows.
During the R/V Sonne 81, 107, and 144 cruisesthe swathNear the continentalslopethe ridge is a 1.5-km-highbroad
mapped survey was extendedbeyond the Sonne 76 coverage swell upon which a local seafloorrelief (-0.5-1 km high)is
[von Hueneet al., 1995] to includethe crestalregionof Cocos superimposed.As the ridge entersthe convergencezone, its
Ridge, and the seawardmargin of Nicaragua. Editing of the seafloorflexes only half as much as the adjacentseamount
completedata set with the Mb systemprogram [Caress and segment,and essentiallyno trench parallel horst and graben
Chase, 1996] and in-house software [Heeren, 1996] allowed structure is observed.
Adjacent to Osa Peninsula, the highest topography on
griddingat 100 m ratherthan at the previous500-m intervals.
The magnetic data from Sonne 107 were added to the CocosRidge is formedby tilted blockswhich trendparallelto
compilation of Barckhausen et al. [1998] (Plate 1). the ridge axis (Plate 1 and Figure 3a). The blocks tilt away
Unprocessedseismic reflection data acrossthe ocean crust from the crestand form a crestalgrabenwhich is filled with
were worked up and interpreted [Flores-Hernandez,1996]. as much as 1 km of sediment some of which is accreted at the
Several wide-angle seismic lines [Ye et al., 1996;
baseof a 1.5-km-highslope.
Stavenhagenet al., 1998; SaYares et al., 1999; Christeson,
1999] constrainedcrustaldepths. Study of samplesfrom the 4.2. SeamountSegment
area is still in progress[Mrazek et al., 1996] and geochemical
The seamountsegmentextendsfrom Quepos Plateau to
analysesof rocks from oceanic areas have identified ocean FisherSeamountand Ridge (Figures2, 3b, and 3c). Oceanic
crust of Galapagoshot spot affinities [Werner et al., 1999]. crustin this segmentis -6-7 km thick [Ye et al., 1996] and has
From bottom-simulatingreflectors (BSRs) an inversion and an age between 15 and 20 Ma [Barckhausen et al., 1998].
analysis of heat flow over the accretionary prism were The seamountshave a Galapagosgeochemistryand are 13 conducted [Pecher et al, 1998].
Swath mapping off
14.5 Myr old adjacentto the margin [ Wernere! al., 1999].
Nicaragua [Mrazek et al., 1999] and reflection seismic data
QueposPlateau,an elongatevolcanicfeaturethat stands-2
[Raneroet al., this issue]are the first modem datato compare km abovethe surroundingseafloor,is the easternexposedend
with the extensive data off Costa Rica.
of a 280-km-long seamountchain (Figure 1). Its internal
structurein seismicrecordsconsistsof subhorizontallayering
(Figure 3b). The crest is flat, suggestingsurf zone erosion,
4. Rugosity of the Oceanic Crust
and this surfaceis punctuatedby many small conesshowing
Morphology along the seaward slope of the Middle waning volcanism after subsidencebelow sea level. The
America trench is resolved from the 3-km relief of Cocos
steep flanks merge into a locally hummocky elastic apron
Ridge,throughth• 2-km relief of seamounts
andridges,to the (Figure3b).
less than 0.1-km relief of normal fault scares. The seamounts
Fisher Seamountis-14 Myr old and has a Galapagos
are alignedroughlywith CocosRidge and have a petrologic geochemistry,
but its southwesttrendingridge is composedof
andgeochemical
affinitywith the Galapagos
hot spot[Mrazek enrichedmid-oceanridgebasalt(EMORB) that formedat -21
et al., 1996; Werner et al., 1999]. Faulting on the seaward Ma [Werner et al., 1999]. About 50 km of the Fisher Ridge
trenchslopeis influencedby crustalstructureinheritedfrom have been swath mapped (Figure 2). Satellite gravity
the Cocos-Nazcaspreadingridge. As the ocean crust is indicatesthat alignedseamountsextend 200 km farther west
depressedinto the Middle America Trench, its axis of flexure (Figure 1). An eastwardsubductedcontinuationof this trend
is commonlyat an angleto the strikeof magneticanomalies. is indicatedby a large underthrustingseamountlandwardof
Where that angle is not critically large, flexural stressis FisherSeamountthatdisruptsthe continentalslope(Figure2).
relieved by faulting along presumed zones of previous
structural weakness as observed between Fisher Seamount
4.3. Smooth Segment
and QueposPlateau(Plate 1 and Figure2). Where the angle
The smoothsegmentmorphologicaldomainextendsfrom
betweenmagneticanomalyand the trench axis is nearly Fisher Seamountand Ridge to the northwesternpart of the
normal,faultscrossthe trendof magneticanomaliesas off the
Nicoya Peninsula. However, thick crustof CocosRidge is
faulted very little parallel to the axis of flexure at the
subductionzone. Extensionalstructurein Cocos Ridge
parallelstheridgeaxisasexplainedin section4.1.
4.1. CocosRidge
The northwest flank of Cocos Ridge extends from its
crestalarea off Osa Peninsulato the QueposPlateau(Figure
2). In our surveyarea, CocosRidge is a hot spot trace that
trends35i from the CNS magneticanomalies[Barckhausenet
Nicoya Peninsula(Figure 2). The oceancrust is from 20 to
25 Myr old on the basisof magneticanomaliesand is 5-7 km
thick. In contrastto the sharprough-smooth
boundaryto the
westthat showsup well in satellitegravity anomalies(Figure
1), the boundary to the northwest between CNS and EPR
crustshas little morphologicalcontrastbut occurs along a
small escarpment(Figures 1 and 2 and Plate 1) that extends
discontinuouslyacross the oceanic plate and across the
continentalslope.
Thecrustbreaksintonormalfaultstrendingsubparallel
to
the axisof flexureandessentially
perpendicular
to the CNS
318
VON HUENE ET AL.: COSTA RICA TECTONICS AND VOLCANISM
jjj
o
[w>t]41daa
[s] •w!/•eM-OM/
[S]•W!.L.•eM-OM/
Is] aW!.L.•eM-OM/
VON HUENE ET AL.: COSTA RICA TECTONICS AND VOLCANISM
magneticanomaliesbut vertical displacementis lessthan the
thicknessof sedimenton the CocosPlate. A subtlechangein
fault pattern occurs across the low escarpment that
corresponds
to the changefrom CNS-I to EPR crusts(Figure
5a).
5. Character of the Continental Slope
319
cross-sectionalarea of the prism with sediment input, and
assumingall of it is sedimenttransferredfrom the subducting
plate, indicatesthat more than 95% of the sedimentremains
on the subductingplate passing the backstop [Hinz et al.,
1996]. For at least 350 km along the trenchthe frontal prism
has a common morphology and structure across all three
tectonicsegments.Along CocosRidge, wherethick sediment
in the crestal graben enters the subductionzone, the frontal
prism is clearly accretionary(Plate 1 and Figures2 and 5c).
The 1-km-thick sedimentin this graben and sedimentponds
along the trench axis (Figure 2) indicate a substantial
The relief along boundariessegmentingthe ocean crust
continues up the continental slope as aligned transversetrending subdued features. Continuations of the Fisher
sediment
source
and
efficient
sediment
Seamountand QueposPlateausegmentboundariesdivide the continental
slope into three segments.The large subductedseamount subduction.
landward of Fisher Seamountis aligned with Cabo Blanco
Gaps in the prism, formed as seamountstransitthroughit,
and the southernmargin of the Nicoya Peninsula(Figures 2, appear to heal rapidly by local tectonic thickening and
5a and b, and 7c). The subductedextension of Quepos accretion(Figure 6). The rate of this accretionwas estimated
Plateauis aligned with a ridge acrossthe continentalslope,a from the length of adjacenttracks left by two seamountsthat
low dome on the shelf, and the northwestend of the Terraba entered the subduction zone at different times. The seamount
deformedbelt onshore(Figure 2 and Plate 1). Betweenthese subductingnear QueposPlateau(Plate 1, Line 3) hasmigrated
boundariesthe middle and upper slope structuresdiffer in 23 km from the deformationfront through the frontal prism
character. The lower slopefrontal prism seemslittle affected and margin wedge. Here the prism is still completely
The next subducted seamount to the northwest
by various morphologies except where destroyed by a breached.
subducting
ridge or seamount.
(Plate 1, Line 13) has migrated 35 km, and here a 6-km-wide
prism fills the breach. At a plate convergenceof 88 mm/yr
5.1. Lower Slope-Frontal Prism
the time differenceof migrationpast the deformationbetween
the two seamounts is 140 kyr.
During this time,
As mentionedin section2, we use the term frontal prism to
underthrusting
of
the
complete
oceanic
section,
accretion of
include tectonically incorporated slope sediment as was
trench
fill,
and
incorporation
of
slope
sediment
have
restored
discoveredduring ODP Leg 170 drilling. Like much of the
Middle America Trench, the sparsetrench fill off Costa Rica the prism, indicating that frontal prism constructioncan be
is commonlysubducted[Shipleyand Moore, 1986; Shipley et very rapid when conditionsare favorable. There alsoappears
al., 1990, 1992; Hinz et al., 1996; Ranero et al., this issue]. to be a maximum limit beyond which growth stops. The
The term frontal prism helps distinguish between two prism is probably stableat the presenttaper (-10i) and 3-10
processes: (1) elevating pore pressureand reducing basal
friction (facilitating sedimentunderthrusting)and (2) storing
materialtransferredfrom the lower plate (accretion).
The frontalprism displayedin seismicreflectionimagesis
a mappableunit in our morphology (Figure 2). In seismic
imageswe considerthe prism the reflective sequenceseaward
of the margin wedge or seaward of a recognizable BOSS
reflection and a semicoherentslope sedimentsection(Figure
4). The prism correspondsto the "outerwedge" of Langseth
and Silver [1996] and the sedimentarywedge of Sites 1040
and 1043, Leg 170 [Kimura et al., 1997]. The contact
betweenthe prism and the margin wedge is associatedwith a
rapid landward increase in seismic velocity in both wideangle and normal incidencereflection seismicdata [Ye et al.,
1996]. The backstopis not a sharpinterfacebut a transition
betweenpresumedmoderatelyconsolidated(velocity 2 km/s)
and well-consolidated(velocity 4 km/s) rocks. In reflection
images the base of the prism is commonly defined by a
detachmentreflection. Thick underthrustsedimentsequences
cut by normal faults are commonlyimagedbeneaththe prism
andthe detachment(Figure4 bottom).
The frontal prism morphologyresemblesthe large slide
adjacentto Fisher Seamountbut with lineamentsparallel to
the strike (Plate 1 and Figures 2 and 5a). The landward
boundaryof the frontalprismis commonlya low escarpment.
The prismis 5-8 km wide narrowingto 3 km wide over some
partsof CocosRidge and is temporarilyremovedacross10 to
20-km gapsby subductingrelief (Figure 5b). Balancingthe
km width.
The uniform width and rapid restoration of the frontal
prism despitecontrastingsubductingplate morphologiesshow
a strong dispositiontoward a stable configuration. Prism
growth seemsto be a self-limiting process. A slopeor trench
source of its material
and variable
abundance
of trench fill is
apparently less important than achieving a critical size and
elevating pore fluid pressureto allow sedimentsubduction.
The frontal prism off Costa Rica stores an insignificant
volume of trench sediment compared to the Cocos Plate
sediment input, and an elevation of pore fluid pressure
appearsto be its primary function. However, this prism is not
tectonicallystatic as indicatedby the terminationof canyons
at its landward limit (Figure 2). Canyons have developed
across seamounttrails some Kiloyears after the seamounts
passed. Consideringthis developmenttime, the absenceof
canyonsindicatestectonicactivity in the prismof similarage.
5.2. Middle and Upper Slope, Smooth and Seamount
Segments
The smoothsegmentmorphology is characterizedby an
upper slope with many canyonsand a gentler middle slope
with shallower canyons and tracts smoothed by recent
sedimentation (Figures 5a and 7 and Plate 1). Only a
representativepart of the smooth segment morphology is
illustratedas a perspectivediagramhere(Figure5a). Near the
ODP Leg 170 drill area, major canyonsare absentacrossthe
middle slope,but large canyonsoccurjust upslopeof the drill
320
VON HUENE
ET AL.: COSTA RICA TECTONICS
AND
VOLCANISM
Z
0
"0
Is] OW!/•,eM-OMI
[S]OW!/J;BM-OM/
VON HUENE
ET AL.: COSTA RICA TECTONICS
AND VOLCANISM
321
subducted
(a)
.......................
....
•-_
smt
':......
'.................
'::'•:-'
..............
: ":•
..........................
"::'•' •: •
: ,.. ::•:•'-:.:':%:
:::•':•
=4.-•
"•• .• .........
-:. ':"' :"'
-:..3
•':':..•':';-:.:.-.'
.:'"'-::..
• .::•::...•
.....
:' .,
•': ::::'::
......
•'-.:.. -::..::.'
•
:
...•
.......
..
.........
:•
...................
.....•....
•.-..
•.:..•:.
::::
•..-•:'•--•:•:"'•-::::'
::•:
........:::
<:,:.:..•
........
...........
.•..:::.•
.....
:::::::::::::::::::::::
:::::::•:•:
::•
::•::. '...........
.:':•"'•
-:-::
%:..
:::::::?:.:
..•.•.':• •.:'::
...........
•?:.::: ....
'• :: :<•
•.•
....
:...........
::•:•.-.. • ..... :- :•::-j:•': ß
•:..•:
-
....
..........
...
. ........
-•: ::::::::::::::::::::::::::::::::::::::::::
<,•
:-•.---.................
:
'
•:•:-:•:•:::
-----•:::&•;:•::-•::
•.•
...................
60
...••<•:.•,•.•.•....•.•.,•
90
........
'-
.•,.••,,
3o
120
.:x,
•:::
• :"•
.......
............
•.:=•::;:<<.•.
::•:::::::,<:::
..•...=,,.
..•/•.•.:.:.•`•::*•*••.•.::•:•:::::::•:•::::•::::.••::.•:•:•.•::.•.••"•:•:•:•:::•:•
6o
.•:
,..:::::::::::
=:•......
:::::::?.:
•.<
<-::----•,'"•:<'
'•
[km]
=:::':<::'
:•:::.:::.::..:•
,:::::::-:: ....
......
_:<•:•........:.:
.
::::
..............
:::::::::::::::::::::::::::::::::::::::::::::::::::
:•:•:-o
::...::•.
======================
3C
.:::.•:•
....
:::.: ':............
..............................
9o
:.•..•:•.•::...•:•:::::•.
•:::•:•::.-.--••••••:.•;:•,•
•2o
......• ...................
•..•:::.::::•
........................
...::::
........................
...........
•
Figure 5. Perspectivedia•ams of the studyarea morphology illustratin• detail: (a) diagramextendin• from
the Le• 170 area to Fisher Seamount(smt) showin• the adjacent slide and the trailin• flank of a large
subducti• seamount,(b) diagram extendin• from Fisher Seamountto the flank of Cocos Ridge, and (c)
diagram1ookin• down the trench axis and extendin• over the crestof Cocos Ridge. In Figure 5a and b, a
•ractumpatte• over the crest o• the seamountsindicates extension o• the sediment section. Many short
nodal •aults extend acrossthe middle slope and resemblethe crevassesin a flowing •lacier. In Figure 5b,
notete•ination o• canyonsat the •rontal prism. In Figure 5c, note the lower slope retreatover the ridge crest
and the shar• juncture whcm tilted block ridges arc subducted.
•m]
322
VON
HUENE
ET AL.: COSTA
TECTONICS
AND
VOLCANISM
two-dimensional morphology, faulting of Quaternary
sediment involves only small (_< 100 m) vertical
-' ?..:::•'
"'......
'%?•:•::?•:•½':?':"•'"';;•':•:::•
...................
:•::;i'
displacements
mostly on extensionalfaults [Mcintoshet al.,
1993]. Extensional fractures are resolved in perspective
diagramsacrossmost of the middle slopeseafloor(Figure 5)
just downslopeof the canyons.
The smooth segment is not complicated by recent
:.g;::' ::..•::•:.-,:?
.........
&"::'::.:eL.-.4::.'
:' '-•.:.<&::'
'½r
-:-5:•::(Z.3.'
z':'•--:•:.::::•-:-,7•i•;4
....
subductionof lower plate seafloorrelief and CocosRidge. Its
........
•..:..:?.:.,.•..•::::::::::::::::::::::::::::::::::::::::::
........
..........
-',*:•*
.........
:-,.:...:,½.%;::;..•&%..
•--.----,
upper slope has been sufficiently stablefor developmentof
deep canyons. The middle slope is at critical dip, consistent
with creepin the upper40-80 m of the section[Baltucket al.,
'•..
..:7
;•:-•:;.-.:
...........................
:•...,.:.
....
1985]. The underlying margin wedge (basement)transfers
:•-,
'"•'•'•'•;;•'•'"'•½?;;';•:½•?';'•'•'
little interplatecompressionthat affects Quaternarytectonics
'• •.•},•."??• -:..• "
:•';•-•
';::7?";%
..•
.;;•.:; "e
:• .............
....•....,
............. ;:;:;::.½-.--?;;•;*
,• :'•h•::•:..½,ff•½•-..::
................ .::•
ß. q;.-½,;•
;.- •.
RICA
:,•:::--..;;.•::•::•.•
-•':-•-:•,.-
.%•:::•:;....v,.•
.... -.......
:;.
•:½:-:-.-.--:.
.... ::•<::•.,..:.:.....c
in the sediment section.
.•.. -•;
Subductedseamountsare indicated by rounded uplifts or
domesmost clearly observedon the middle slope(Plate I and
Figures2 and 5). The seamountsare 1.5-2.5 km high on the
oceanic plate and 1.5 km high where seismically imaged
beneaththe margin wedge (Figure 7) [Ye et al., 1996]. In the
mid-slope area where the crust is •5 km thick, the domes
above 1.5-km-highunderthrustseamountsare clearlyresolved
.......ß;.................
.:.•...;";:½'":"•':'-'-;:•
•';::.':'i'.
.............
:;'.
................
'"'"'""::•"'•4•4:•'•';':'•*•;"g•;}•-(:'.
""•'"'"•"x;
....':". ......
' ....
'-%--.
-•"
::'"::...
':•'1'5•:;•'•'%"-'::;•5};;•:::•::';...
ß*•;•?•:•::';%
:":"½J•&•,•:
' .......
• .'.......
'"''""-:--.:-•:;.;-•¾'
:::'..':•
::":*'t:•.•
in swath map data (Plate 1 and Figures 2 and 5b). Above
-•..•;;:.•
•&...........
'•:: -'.:::•;::-:?.'
:::::.:-::• ;:• *"• ...,•..,
%'•
subductedseamountsthe upper sedimentsectionfails on the
oversteepened
seawardor trailing slopewith as little as 3 i-4i
of addedtilt. Despite substantialalbeit temporaryuplift, the
margin wedge along a seamount track remains coherent
,-*•
-•-••••t:•,,,..½cc:•..i_-.-----',-'-•w•.•$
:•
-...:
....a*. ,,'•.....
==========================
.........
-•.•?.
,;•-,:
...'•"•,:::--:.•*•••,S.m.-:-,--•_:::.;;•.:.•
;.• •.•
*:•**
,:..-:•:•;•,:.•..<
<,•4•:,•%.;½,.:,::..
•....
(Figures 6 and 7), and mainly the frontal prism is breached.
..::.::.-*
...,. i:::;.7:,::i•,•,•:::::4:;
-::
..... •
-.•,
."-'..;:•&,...
.::...:.
'.,::":.:..:.':•;1:,:::;•:.,:½,½;•,::.%;.;:
.
-•-•:'*:::::...
,•:..' ;•'--:;'*"'4'.:.:'
'-'.* ....
':***:::,:•:'-•**-.::•
........
•
2•,•
:•::½•½•':•:•,-:½;"•...
'
%......
:• ........
..
'"
'•
....•.• .:•....
..........,..:•:•{:•;:•½;•:.
.....::'.. ß..........
.:-,-.::•:•
..
. . '................
:•'
":A
The features
'•.;':T?:.:5•-,..•.::::;•
•;;;:.-;" ":::•::•.............
,
:.:.•,:'.:•:::.;?:::.::::;%*•:.:::.'•'"•':•:,,•
' •*,•:::•-:--;•::,:•A:•:•.
":'•:•'A
•
:.:%.....-:
.....
"*'(*:*:a::::**::::*
:*•*• ":"*'*"-'q•g;;'T•:;*:'**½S½:****
......
....
..•::.•
"•. :::•'
..............
''•½.:;:;..':L•%
•..•
'• ..•............
•.....
g
ß'.::•'
'½
...
.
ent
..•'•.'.
:'"'•..
•
•:•:....:...;::.
•"-'•$:•:'½•?;•.•.
...........
• f•"•
:•::• -.•'.:•:'
::•
•:.: :'--:
..........
•?•.•.w
:::•:•,.•!,:•:•..:•::.:
:::•:•,:.•.:..
•:....
:•.:.•.:.::•.•
o
*
. ...•,•...:..,.,,.•s•..
.....................
.•..:•.:.;....::..
::.• .....:..:.:•.,:.•.•,..,
•.
3o
[km]
Figure 5. (continued)
_
sites[McAdooet al., 1996]. In themiddleslopea 40- to 80m-thick creeping sediment blanket was documentedwith
microfauna
transported from shallower
water,
paleomagnetics,
and sedimentreworkingin coresfrom Deep
Sea Drilling Project(DSDP) Leg 84, Site 565 [Bahucket al.,
1985]. The difficulty in clearly imaging strata in slope
sedimentin many seismicreflectionsections[i.e., Shipleyet
al., 1992] is consistentwith stratal disruption during
downslopecreep. Drilling at Sites565 and 1041 documentsa
changein tectonismat '-,280-m depth from a lower more
deformed to an upper less disturbedand less consolidated
sediment. This boundary within the late Miocene may
correspondto an unconformitylocally observedin seismic
records.Precludingmoreprecisedating,we notethe possible
coevalarrivalof CocosRidgeandtheunconformity.
Slope stratarest on the BOSS reflection, which has a relief
thatis commonlygreatest(500 m high)fromthemiddleslope
to the frontalprism. Althoughthe uppersurfacehasa rough
associated
with
seamount
subduction
off Costa
Rica were studied with analog modeling [Dominguezet al.,
1998] and convincingly replicated including the surface
fracturesand slopefailuresapparentin Figure5.
Three seismic images along seamounttracks (Figure 6)
show how rapidly margin structurecan be restoredafter a
seamounthas tunneledbeneaththe upperplate. After the first
10-15 km of underthrustingthe upperplate is uplifted but not
severelydismembered
(Figure6b). Only the upperhalf of the
slopesedimentsectionfails as shownin both the dip and
strike lines across a seamount 25 km from the trench axis
(Figure7a,km 90-95). Oncetheseamount
trailhasaged•0.5
Myr, the seismic image and morphologybecomenearly
indistinguishablefrom an undisturbedseismicsectionacross
the frontal prism. Seafloormaterialflux from erosionby
singleseamounts
is small and difficult to quantifywithout
three-dimensional
images.
The removalof materialfrom the undersideof the upper
plateby underthrust
seamounts
is resolvedwith difficultyin
two-dimensional
seismicdip lines (Figure 6c). Along the
strike line (Figure 7, Line 6, km 55-75), •500-700 m of
thinningis observed
just forwardof thepeakof an underthrust
seamount (see depth section of Ranero and von Huene
[2000]). Material flux is unclearin time-migratedimages
above other seamounts.
Fromthe southeast
half of the Nicoya Peninsulato the Osa
platform,the upperslopedisplaysthreebroad(50, 45, and70
km wide) embaymentsseparatedby seawardprotruding
headlands.The shelfbreakwhichoutlinestheseembayments
is commonlyan erosionalscarpat the edge of the swathmappingsurvey(Plate 1 and Figures4 and 5). Two of the
embaymentsin the central part of the surveyedarea are
oppositethe seamountdomain, and a third is off the southern
Nicoya Peninsula,in part, oppositea currentsmoothoceanic
VON HUENE ET AL.: COSTA RICA TECTONICS AND VOLCANISM
323
sw
0
5
,
,
,
,
10
I
i
,
,
,
15
I
I
,
I
,
I
20
,
,
I
,
NE
25
I
,
,
,
,
I
ISonne
81 Line111
30 km
,
,
,
•
I
,
,
uplift rough
surface
2-
subducted
'
seem
ount
"•"-•'
•••••
'/'
-"'••••
'•"••"'
•'.......
••'
ne m•pe•ag•c
t re nc h
.... ,..
' -- '- '•••
.."•••••••••
.:-.
,.,,,,,,:;.•_-•
;,,,.•
Pela g ic
.•. :.....
.......
.
..........
6
. ....
•.-.•.-
.....
•:• .•,..•::.:•;-•:'•
.... :•a, ......
• . ...........
.......................
..•..,•..:,•-.:•.... :-.••....,•
,
.-•
.....,
:•......
-......
•-:•.--•--•,.
.- .....
-. -...
,• •--..... • .:,.•.-:•.•;•
. • ......
• ......
•-..•-:•
.........
• ,•:
ß.• •.............. •,•..:_•,• .• ;•
•:•.5•Z•,-*•••;•,•.•--
=•.$• ,, •----........ -• ....
•.•
•* .•-•2 •
,• ?,•.•.:•.•
..
•..•...•
.:.•--'•;..• - ß
....
•=:ß .,•,
.
.'-•-:4t
......
,•..,"•'."•
........
--.
....:•.•
'••' ' • .-.......
"•-::•-•',
'••.•cam
.:•-*-•.•.-.-•'
.......
'•• '.::..•..,.•
........
'•"•
•-'
.................
•;.• .........
• ......
-,•½•
.......
:47•',-'
.....
'•'x•'••
......
-".::•
........
'
............
.':/..•..,-:::•
•' ,z :..•-•
.....
-••••/••T•:"•"•
.....
•,*:..•ß::.-.,:•4t'-",
.=•.::".'..-:-..,'.',.--.
'•-:"•-•••••:•:• •..+.•-,•:'*•-':
.:..•:,.•;•.•,..
t•4•;=•.•:•C,• ..,•.• •,•,,•...•...•..•..•=..•.•....•..,
•.•.?,:.•;•..:.•..•,:....,•
....................
.• ...•. •"•..•
112oo
116oo
0 km
5
J • , , , I , ,
ISonne 81 L•ne
ß 13
__[
'"-"-""
'• .......
•
,
•
10
I ,
12ooo
•
,
J
3I
/
..
,
15
I ,
,
,
,
20
I ,
CM P
,
,
slide
deposits
frontalpnsm
\•_
•
?-:.:..•..-:..-.•
.•,.. _
.
124oo
25
30
I • , , , I ,
roughsurface
slid
--•'e
- ----t---:-.-•'•
•.,..-..-.......
."::[..,.
'....
"-.::..:•-:'
.....
"
•"'••-'-"'"'-•"-"•'••••••
•--•••:•:•"•••••
---•,,,,•
"'•'•
'••'
'"-'-'
'*'"'
'-'"'
..
decolie...........
.-.3•
"•'•'•"•,•,•.•.•..:•::;•-,:;
'•?"-"•j•
;,-.-•,,.--'-w••-..
...•,,..,,••
• .......•ß.........
.,:....--•,,
•-.-,,,,,,---.•••••
._••,•,,_••,•,•••,
-_
. . -•,-*-•.
••.._.
•,•---•----•.-.-..,
......•,,..•
• _....,•••••..••••••
.......
-•.
- -'
..:.•"•.•
....
•c-,---....•.-'-•••.,..•.'•
.....•.-...*
r_•
....
...:,•
......
....-::...... .. . .-.N m.,e_.nt..-:
..-.----.
...••••:.•.•......:•
.--_•-.,,.•__,,•,•.•.._•.......
_
......
o
-'.*-'-'••'-,•"•-••••-.ß
'.-,-•-'
-••"
"•'''
' -'• '- '-••
CMP
,
....
.i
,
..•
....•;•:-:•,•
;.• •%:•....•:.½•,•,•:*,_.,.,.:•,.,..
..-,•:::•.......
=•"•••',.•-.•.•,-.=.-.
..-•..=,.•
...........
.,..•..::.•.
.....
•, ...........................
.,:::,••,•,•.......
•:•.;.•.•.,.s.:•-•:.:•....:...•-:•.-..•
......"••..•,,
12400
30
I
I
,
,
,
...,,.•-.-•
.•.•..'..........
•'--'"--.-•••-':•-..•
-'•""•'"-""•'
12800
25
20
I
ß
.......
13200
15
10
3 Sonne
81Line
12a]
E 4
'""•:•• '*•t:.•
5
km
0
rough
surface
. . -.._:.•
frontal
pr,sm
.....
normal
•.•.•..•.•.•..•.•?=
.....
............................
.........
......................
......................................................
:• _. •,1 ,.•
•._
.••.••••.•..•
.......
• .:,•:='•'--•
-••'••'•'•:.•.•
•:4• ;:.• :.•4•:•
ß
.......
;•-•
........••'•"'•'•=" •
'•••4•
•.._..
..........
.•,•.
'--•' :.-•:'•••
.... -.•. • .......' '-•*•'-•."•
............
½:'.-.:-•.'•-.-m":
'--...:•';.-::•:•
ß....• .....• ....
'
'••c ' '• .-•.'"•
'-..-'•.'•=•-..---'.-"•
-'-•":....••...•
:"•'"•
•;•/•
••-'••"••"•"••:':g%'
'•'"
"•-"-'•••"•m•'"•••••
'•••'.'-'--•.,.:-•H•te
,•_ ••.-':--•'m•-•-•
..:•-?•
• •,•r-•.•:•••.:.,•,' '•-•-:-:•
• ...............
•--'"'
"•••••4• •"'•'-•
•.=:' '•*••
....
--?.
'•:• ".'........--'
,..,:.-.--.
........
.......
::.-..-.•-•..•,.
.........
•,:•:-:'•:•;:•½:-?•:•-;'•Z'
.•'••m,
'•'•½'•'•,•::
...................................................
I
11200
10800
10400
i
CMP
10000
Figure 6. Seismicreflectionimagesshowingstagesin subduction
of seamounts:
(a) Line 11, whereseamount
hasbreached
the frontalprismand its trailing flank slidedebrisarereachingthe trench,(b) Line 13, where
seamount
hasmigrated
•25 km fromthetrenchaxisandthestructure
of frontalprismis restored
(seeFigure4),
and(c) Line 12a,wherelowerslopehasbeenrestored•0.5 Myr afterseamount
wasunderthrust
andits trackcan
no longerbe distinguished
in a seismicrecordacrossthe marginfront. SeePlate I for locations.
324
VON HUENE ET AL.: COSTA RICA TECTONICS AND VOLCANISM
I
I
eo•pns LifinoJ
o
o
•
ø
[s] ew!j./[eM-OMj.
[s] ew!J./[eM-OMJ.
.
o
• •
327
VON HUENE
ET AL.: COSTA RICA TECTONICS AND
segment. The latter embaymentwas probably oppositelower
plate relief as indicatednot only by failed slope sedimentand
transversefaults (Figures 2 and 5a) but also by seamounts
seawardof our study area (Figure 1) and in the direction of
plate convergence.
Removal of material from the embayedmargin might be
inferred from a comparison of seismic records across
embaymentsand the headlandsbetween them (Figure 4)
[Hinz et al., 1996, Plate 1]. Across each headland, structure
of sedimentstrataindicatethe edge of the shelf basin near or
just beyond the current shelf edge (Figure 4). If prior to
subductionof Cocos Ridge, the margin configurationalong
all of the Costa Rica margin was like that off northwest
Nicoya Peninsula [cf. Shipley et al., 1992], the volume of
materialmissingfrom the threeembaymentsis 783, 1085, and
1543km3 fromnorthwest
to soufheast,
respectively.
The
latter volume may include some nonerosional features.
Althoughmasswastingand downslopematerial transportare
indicated,the volumeof missingmaterialis largerthan can be
explained with surface erosion of only slope sediment.
Removalof material alongthe plate boundaryin the embayed
areais likely, but its volumeis not well constrained.
5.3. CocosRidge
Above the crestof CocosRidgethe continentalsloperising
to the Osa platform is short,steep,and without canyons(Plate
1 and Figures 2 and 5c). The slope and the shelf edge are
indented landward over the ridge crest, indicating erosion
accompanyingsubductionof Cocos Ridge. The continental
slopeaccommodatessubductingridges and seamountson the
ridgecrestwith only minorupperplatedisruption.
Osa Peninsulais the emergentpart of an equally extensive
uplifted area comprisingOsa platform (Figure 2). Seafloor
erosionof the platform sectiondisplayedin seismicrecord
P1600 has removed -1 km of strata (Figure 8, km 50-65).
Erosionis greatestnear the currentshelf edgeand decreases
landward. Archingover the crestof the ridge,presumedfrom
studieson land [cf. Fisher et al., 1998; Kolarsky et al., 1995;
Gardner et al., 1992; Corrigan et al., 1990], was confirmed
over the northwest flank in strike line 6 (Figure 7a). The
seafloorrises 1.5 km, the plate boundaryshoals-1 km, and
the slope sedimentsection is folded. The wedge has a 10i
taperin the 25 km beneaththe continentalslope[Stavenhagen
et al., 1998], which is 2i-3 i lessthan the taper where normal
oceanic crust subducts. Above the ridge crest the slope is
indented ~15 km. Beneath the slope, landward dipping
reflectionsare truncatedat the plate boundary(Figure 8, km
35-45). Thus, over CocosRidge, wave base erosionat the
seafloor, retreat of the slope, and narrow taper accompany
erosion.
The highestridge parallel normal fault scarpin the crestal
grabenmeasures1 km at the seafloor and extends0.8 km
beneaththe sedimentpond (Figure 3a). Its trend continues
into a transverse offset of the slope, a partly surveyed
elevation at the shelf edge, and an indentation in the coast
which is alignedwith a ridge acrossthe Osa Peninsula(Plate
1). That ridge is located near an area of rapidly uplifted
marine shoreline features on the Peninsula [Gardner et al.,
1992]. A ~7-km-thick upperplate beneaththe Osa Peninsula
[Stavenhagenet al., 1998] has apparently not completely
VOLCANISM
attenuatedsurfaceuplift over local relief on the lower plate
thatis-1.8 km high on the exposedcrestof the ridge.
A temporalrelation is documentedbetweensubductionof
Cocos Ridge and development of the Terraba Basin and
Cordillera Talamancauplifts (Figure 2) [Grafe, 1998; deBoer
et al., 1995], a gap in the Quaternary volcanic arc, and
tectonismin the back arc [Collins et al., 1995; Kolarsl•T et al.,
1995]. The uplifted QuaternaryTerrabaBasinsedimentwas
compressionally thickening during the past 5 Myr,
exhumationof the Talamancabegan between4 and 3 Ma. A
causeand effect relation between deformation in the upper
plate and subductionof Cocos Ridge is generally accepted,
but the tectonic mechanismis unexplained. Subductionof
Cocos Ridge resulted in considerably greater permanent
deformation than observed where the Nazca Ridge [Hsu,
1988] subducts. Lack of pronounceddeformationalong the
Peruvian coast may be attributed to migration of the
subductingNazca Ridge along the margin whereas Cocos
Ridge and its flanking seamountshave subductedin the same
generalareaduringperhaps5 Myr.
6. Basal Friction
The evidencethat some lower plate morphologyremains
intact as far as the coast and into the seismogeniczone is a
general sign of relatively weak coupling along the plate
boundary. Off Osa Peninsula the seemingly passive
subductionof the high tilted block ridges on Cocos Ridge
shows unhindered accommodation. ODP Leg 170 drilling
and seismic data along the margin show subduction of
essentiallythe entiresedimentsection,which indicatesa basal
frictionalongthe detachmentnearthe shearstrengthof trench
sediment.
In the first 5-10 km landward of the trench axis in
Line 13 (Figure 6) the underthrustsedimentsectionremains
essentially intact indicating interplate friction close to or
lower than the sedimentsectionshearstrength. Fartherdown
the subduction
zonewheresubductingseamounts
~2 km high
are buried beneath a 5-km-thick upper plate, their peaks
appear essentially intact, and no unusual seismicity is
observed(Figures4 and 7) despiteerosionof the upperplate.
The upper plate domes up over subductingseamountsand
ridges without the dismembermentexpected from highfriction shear zones, but material flux is locally observed
[Ranero and von Huene, 2000]. Sufficient plate boundary
friction to produceteleseismicearthquakesbegins35-45 km
landwardof the trenchoff CostaRica, wherethe upperplate
is 7-12 km thick. The clusteringof earthquakesbeneaththe
shelf and the alignment of these clusters with the trend of
seamountchains (Plate 1) allude to seamountsas physical
asperitiesin the seismogeniczone. Uplift patternsalongthe
coastare attributedto subductedlower plate relief [Marshall
and Anderson, 1995; Fisher et al., 1998]. It is inferred that
seamounts
may be detachedthereto explainthickeningof the
crust. Alternatively, the uplifts might last only during the
passingof the seamounts
beneaththe coast.
The physical evidence of friction at the level of sediment
shearstrengthto 5-km depthand friction at or below the shear
strengthof a seamountto depthsof 25 km beneaththe coastis
reminiscentof the San Andreas weak fault paradox [i.e.,
Zoback and Beroza, 1993]. The physical evidence is
consistent with inferences of weak coupling along the
328
SW
VON HUENE ET AL.: COSTA RICA TECTONICS
20
25
30
35
40
45
50
55
AND VOLCANISM
60
65
70
75
NE
85 km
80
1
800
1200
1600
2000
2400
2800
3200
CM P 3600
Figure 8. Migrated time sectionof Shell InternationalePetroleumMij.B.V. Line 1600 acrossthe Osa platform.
See Plate 1 for location.
Erosion
of shelf sediment
strata occurs between
km 50 and 70.
Cascadiamargin[Wanget al., 1995] and the northernJapan component decreasing steeply in southern Nicaragua but
gently in Costa Rica. Herrstrom et al. favored a combination
margin[MageeandZoback,1993].
of increasedslab flux and changesin the compositionof the
mantle to explain along-strikemagma variations. Although
7. Segmentation
no author argues forcefully for differences in subducted
We have followed the morphologicalproxy for boundaries terrigeneous sediment volumes as a major control for
variability, the point is often discussedinformally and in oral
that separatedifferent oceaniccrustalsegmentsfor 70-110 km
presentations.
from the trench axis acrossthe Costa Rica margin. The two
Indicators of subducted terrigeneous or upper oceanic
with a pronouncedseafloorrelief seawardof the trench axis
are also coincident with concentrations of earthquakes crustal componentsare Ba/La ratios [Cart eta!., 1990], Be
beneath the shelf. Each type of ocean crust modulates isotopes[Morris, 1991], and U/Th ratios [i.e., Patino et al.,
2000]. High 10Be/9Be,U/Th, and B/La ratiosin Nicaraguan
material input to the subduction zone and has different
lavas and the very low onesin central Costa Rica have led to
intensitiesof normal faulting impartedby crustalflexure into
the trench. We now relate marine tectonic observations to the
some investigators to speculate that much of the trench
characterof the volcanic arc noting that if initial subduction sediment was subductedin the former compared with the
of Cocos Ridge began at 5 Ma the geochemistryof recenI latter. However, our observation of little variation in size of
lavas developed in a tectonic system similar to the current the frontalprism along CostaRica and a similar frontalprism
one.
off Nicaragua [Ranero eta!., this issue] indicate little
modulation of subducted sediment flux from accretion.
The
Although it is clear that ocean crust or the slab
contributes material to volcanic arc magmas, a lesser searchfor anotherrelationbetweeninput at the trenchand arc
lava geochemistryrecalls the opinion of Leeman and Cart
consensusrelates to the terrigeneousclastic material that
travels with the subducting plate. On a global scale, [1995]: the lateral heterogeneousgeochemistry probably
subducted sediment volumes and elemental fluxes appear reflects a combination of different ages and evolution of
important [Plank and Langmuire, 1993]. However, along subductingoceaniccrust, its thermal state, configurationof
only Middle America, the casefor this relation has not been the subduction zone, as well as subducted terrigeneous
made strongly. The case for along-strikevariability in the material. Patino et al. [2000] presentevidencethat hydrous
fluids dominatethe transfer of componentsfrom the slab to
volcanic arc was clearly argueda decadeago by Cart eta!.
[1990]. They explainedthis variability on the basisof slab lavas and speculatethat more fluids might be releasedwhere
dip and its effects on focusing the distribution of fluids the subductingcrustwashighly fracturedduringflexure.
releasedfrom oceaniccrust. Leemanet al. [ 1994] arguedthat
slabtemperaturewas important. The fluids would be released
earlier from a hot slab relative to a cool one. However, the
7.1. Physical Segmentation
Three segmentboundariesdivide the Cocos Plate and the
elevated temperature associated with Cocos Ridge and continentalslope (Figure 9). A transversetrend recognized
abnormallycold temperatureoff Nicoya [Langsethand Silver, by early investigators consists of the successivefeatures
1996] createa large northwestcooling temperaturegradient, alignedwith Fisher Seamountand the associatedridge jump
which is inconsistent with currently known southeast between older CNS-1 and younger CNS-2 crusts
decreasingslab signal gradient. These gradientswere well
[Barckhauseneta!., 1998]. Landward of Fisher Seamount,
displayedby Herrstrom et al. [1995], showing a subduction this trend includesa large subducteds•amount(Plate 1 and
VON HUENE ET AL.: COSTA RICA TECTONICS AND VOLCANISM
87øW
12øN ß
86øW
329
83øW
85øW
,
.:
.
•:
VOLCANO OFFSET
1 IøN
10øN
9•N
8•N
.• ....•
=
i
•
!
I•
Nicaragua
Costa
Rica
•Nicaragua
I ' II Costa
I Rica
I 60• "
2.5 .'•,
"
100 -
•'
I
•
fracture
zone
width
2.0 // •'•
/
'E
verticaldisplacement
• •,
offracture
zone
•
•O
•
40
•
•
•-
50.
I
I
,,.
•--
I..
•
Distance-
•
• .....i,,,
100 Km Intervals
0.6
0.4 ......
17
!
I .
!
•.
.......
I.....
-
1
•, 10
r-;
!
-,
I ....... I
0
I
Distance - 100 Km Intervals
Figure 9. Top. A diagramshowingoceancrust segmentboundariesoff Costa Rica and Nicaragua. Solid
trianglesshowthe locationof Quaternaryvolcanoes.Opentriangle(labeledEX) is a 2 Myr old dome site. TU,
Turrialba;IR, Irazfi; BA, Barba;PO, Pofis;AR, Arenal; CH, Chopo; RV, Rinc6n; CO, Concepti6n. Bottom. (a)
Regionalvariationsin the Ba/La traceelementratio alongthe arc,which indicatesvariation in the slab signal
[from Patino et al., 2000]. (b) Width of the flexed and fracturedoceancrust (dashedline) and the vertical
displacementof the seaflooracrossthe flexed zone (solid line), which display a trend similar to the slab
signal. The width of the frontalprism(dashedline) varieslittle alongNicaraguaand CostaRica. Approximate
locationof segmentboundaries
shownby the followinglines: VO, volcanicoffset;CNS-EPR,crustaljunction
along fossil triple junction; FS, Fisher Seamount.
Figures2, 5, and 9), coastaluplift at Cabo Blanco [Marshall
and Anderson, 1995], the linear southeastcoastof the Nicoya
Peninsulaincluding the Cobano earthquakecluster (Plate 1)
and the left lateral offset and gap betweenPlatanarand Arenal
volcanoes,which occursover the QuesadaSharpContortion
ofProtti et al. [1995]. The volcanoeson either side of this
boundary are very different in morphology and large-ion
lithophile (LIL) elements [Patino et al., 2000]. This
transverse trend is more obvious than those on either side.
QueposPlateau is a morphologicalproxy marking the
transitionto normalcrustfrom the thick CocosRidge crust
(Figure 3). Once subducted,the transverseboundary
330
VON HUENEETAL.: COSTARICATECTONICS
AND VOLCANISM
smt
FISHER
province
•
SMT.
....
--•;. • ?....•
(a)
,;:•.•.......
CENTRAL
--
•::.:•:•.--.::..:•
........
".'-•:-•,•,2.'%.•-•
COSTA RICA
.' -•
•::::::::½"
;' '
...:.•-::.-:.•
½'
.-..%::•...
.......7.......::.•
•,'"%•?}::
•
./.:.,,).;
..?'•7"
..:::/.57"'
•.":
•:.•:
..?•
....
.
q':•
ß
.::,•
;.
;'• **,•:•½•' smooth
.'• * :.•.?'
,.•
province
METRIC
PROFILES
5oom
1
km
....
...
......
trench axis
:•:::%•.
intense normal
faulting
(c)
CENTRAL
NICARAGUA
Figure
10.Nicaragua
andCosta
Rica
normal
faulting
adjacent
totrench
incomparative
perspective
views:
(a)
offcentral
Costa
Rica;
(b)clustered
bathymetric
profiles
(vertical
exaggeration,
VE--15:l)for
theseamount,
thesmooth,
andtheNicaraguan
segments;
and(c)offNicaragua
(a 100-kin-wide
section).
Notedifference
in
fault spacing,displacement,
andwidth of flexure.
continues
into a headlandmarkingthe northwestendof the
Osaplatform,an earthquake
clusterandassociated
low dome
on the shelf (Plate 1), the end of the Valle General and
Terrababelt [Kolarskyet al., 1995],andperhapsthe endof
significant Quaternary volcanism. In the Cordillera
Talamanca
atRioLorinearCerroDurika,
2 Myroldandesitic
todacitic
domes
occur
ona projection
oftheQuepos
segment
boundary
[Abratis,
1998].Seismicity
aligned
withtheQuepos
Plateausegmentboundaryextendsacrossthe continental
slope
tothe2 Myroldvolcanic
arc(Plate1) andis sharply
VON HUENE ET AL.: COSTA RICA TECTONICS
definedin the local earthquakes
recordedsince 1995 by the
Borucanetwork [Arroyo, 1999]. Togetherwith the Fisher
Seamount
segmentboundary,
the Queposlineamentbounds
thecentralCostaRicaearthquake
rupturezoneof Protti et al.
[1995].
A third transverseboundaryon ocean crust is found
between the CNS and EPR lithospheresoff the Nicoya
Peninsula(Plate 1 and Figures2, 5a, 9, and 10). The
boundaryis a fossil triplejunctionexceptadjacentto the
trenchwhereit separates
crustof differentages(Plate1 inset).
It has a morphologicalproxy in a discontinuouslow
AND VOLCANISM
331
The offset in volcanoes along the Fisher lineament is
accompanied
by step-likechangein slab components:Ba/La
and Sr-Nd isotopicsystematicsare discontinuous
here. In SrNd plotsthe slab signalfrom Rinconto Arenal disappearsin
centralCostaRica [Patino et al., 2000].
The CNS-EPR boundaryis aligned with the changein Pb
isotopes noted between Chopo and Tenorio volcanoes
[Feigensonet al., 1996]. A GalapagosPb signaturein lavas
oppositethe CNS crust is missing oppositethe EPR crust,
consistentwith a subductedoceancrustalboundaryhere.
The Tertiary volcanichistory of Costa Rica is completely
transverse
escarpment
on the oceancrustandits subducted different from that of Nicaragua and the rest of northern
extensionis markedby uplift and sedimentfailure up the Central America [Cart and Stoiber, 1990]. Volumes of lava
slope. At theshorethe convexcoastof the northernNicoya in northernCosta Rica are about twice that in Nicaragua per
Peninsula becomes concave seaward to form a broad
equivalent length of arc [Alvarado et al., 1992]. The
embayment
that parallelsthe embayment
in the continental geochemical slab signal increase in central Nicaragua is
summarizedelsewhere[Cart' and Stoiber, 1990; Patino et al.,
slope(Figure9).
A majoroffsetandlong-standing
divisionin the arcoccurs
betweenCostaRica andNicaragua[Cart' and Stoiber, 1990].
2000].
Unusualis the positionof the arcwith relationto the dip of
the BenioffZone. Off Nicaragua,wherethe BenioffZone is
verysteep,thearcis landwardandovera deeperpartof the
7.3. Relation of Marine Geologyto GeochemicalTrends
zone than in Costa Rica where the Benioff Zone is shallow.
This condition predates the beginning of upper plate
deformationassociatedwith the subductingCocos Ridge.
The Miocene arc in Costa Rica was nearer to the coast and
hasmigratedlandward.Conversely,
themiddleMiocenearc
in Nicaragua
andnorthwestward
wasfartherlandward
andhas
migratedseaward[Weinberg,
1992;Balzer, 1999],sothe arc
offset was formerly greater. A fundamentaldifferencein
crustal structure is observed between Costa Rica and
Nicaragua[Raneroet al., this issue;Walther et al., 2000;
Christesonet al., 1999; Sallareset al., 1999], a changethat
corresponds
with a terranesutureof pre-Miocene
age. This
suture appearsinactive as indicatedby lack of active
transversestructureacrossthe continentalslope and absence
of a corresponding
boundary
across
theoceancrustin swath
mappingoppositethe international
border[Raneroet al.,
1999]. However, the morphologicalcharacterof the
continentalslope,changesnoticeablyon eithersideof this
upperplatesegment
boundary.Thereforethe volcanooffset
is perhapsmore closelyassociated
with the differencein
upperplatecrustalstructure
ratherthana currentboundary
in
the subductingoceanicplate.
7.2. GeochemicalSegmentation
The currentphysical trends from Cocos Ridge toward
Nicaraguadevelopedwith subductionof the ridge beneath
Costa Rica. These trends include a northwest decrease of
ocean crustalthickness,subductionerosion, and thicknessof
ocean basin sediment. Conversely, slab dip, lower plate
flexure, and normal faulting adjacentto the trenchincrease
northwestward. Geochemical variation thus parallels the
intensityof faultingin theoceancrust(Figures9 and 10) and
a steepening
of dip of the seismogenic
zone [Protti et al.,
1995].
Sedimentinput to the subductionzone is fairly well
constrainedalong Costa Rica but the Nicaraguan swath
bathymetryis supportedby a singleseismicsection(Ranero
et al., this issue). Off Nicaraguathe sedimentcoveron ocean
crustis 250-300 m thick whereasin the Leg 170 areait is 400
m thick. Transportnorthwestalong the trench axis from
CocosRidge towardNicaraguadistributestrenchsediment
inputbut eachseamount
enteringthe trenchoff centralCosta
Rica blocks axial sediment flow for-0.5
Myr.
Frequent
slides from the middle slope of Nicaragua (Figure 10)
[Raneroet al., this issue;Mrazek et al., 1999] could also
redeposit
theupperlayersof slopesedimentacrossthetrench
axis. So thereis no indicationof a pronounceddifferencein
sedimentinput to the subductionzone that parallels the
variation in elements of the arc lavas.
The possibleconnectionbetweenintensityof fractureand
variationin elementratiosmay be relatedto hydrationof the
subducting
lowerplate. In the pastattentionhasbeendirected
mainly towardthe fluid flow in accretedsediment. During
accretion,fluid first drainsfrom the upper sedimentlayers
abovethe decollement
andventsat the seafloor,but drainage
from lower layersis impededas shownquantitativelyalong
an Alaska seismictransect[von Huene et al., 1998]. Below
the decollementat ODP Site 1043, 1 km from the Costa Rica
showa plume-related
isotope
signature
by their2ø6pb/2ø4pb
deformationfront, the upper75 m had compacted37%, the
ratioswhich are higherthanthe calc-alkalinemagmaticrocks next 55 m had compactedonly 9%, andthe remaining185 m
had not compactedat all [Moritz et al., 2000]. Fluid in
of the older arc. Large-ion lithophile element (LILE)
enrichmentis variable in the adakites,but they containamong subductedsedimentbelow the decollementhas a drainage
the highestconcentrationsof LILE, for example, Ba (1500paththatlengthens
with increasing
subduction.This drainage
2000 ppm), and the highestBa/Th ratios(900-1500) sampled patternslowsfluid dischargewhich may thuscontinuetensof
The subductingoceancrustalsegmentation
corresponds
in
varying degrees to geochemical changes along the arc.
Landwardof the Queposlineament,2 Myr old andesiticto
daciticdomeshave a traceelementsignature(heavyrare earth
element (HREE) depletion) characterizingthem as partial
meltingproductsof the subductedbasalticoceaniccrustwith
garnetin the residue(adakites)[Abratis, 1998]. Both alkaline
back arc magmaticrocks and adakitesin southernCosta Rica
in the area.
kilometers down the subduction zone.
The mud volcanoes
332
VON HUENE ET AL.: COSTA RICA TECTONICS AND VOLCANISM
nearthe upperslopeoff the Nicoya Peninsula[Shipleyet al.,
1990; McAdoo et al., 1996] indicatefocuseddrainageof fluid
nearly 40 km from the deformationfront. Pore fluid in the
subductingoceanic crust will be releasedeven farther from
the trench,becauseit must first penetratea sedimentlayer that
has a low permeabilityuntil it is sufficientlyfractured. Just
wherea fracturepermeabilitydevelopsis unknownas is fluid
flow to a detachment which probably forms a principal
drainagepath updip. Normal ocean crust contains 10-15%
fluid [Fisher, 1998], and the additional fluid taken into the
faulted zone of flexure adjacentto a trench is unknownbut
mustvary with intensityof fracturing. Off Nicaragua,ocean
crustallayer 2a has a notably low seismicvelocity (3.3-3.4
km/s) whereasthe globalaveragefor 20 Ma crustis 4.5 km/s
[Walther et al., 2000]. The low velocity is attributed to
fracturingand increasedfluid in the fractureporosity. In the
zone of flexure and faulting the entirecrustand uppermantle
are brittle becausethe oceanlithosphereis more than 13 Myr
old [Werner et al., 1999; Morgan et al., 1994]. Normal faults
displacing the ocean plate seafloor off Nicaragua are
associatedwith dipping intrabasementreflections[Ranero et
al., 2000] interpreted as deep penetratingfaults. A few
dippingreflectionsare also visible beneathnormal faults in
the oceanplate off Costa Rica (Figure 4). Thus, when the
lithosphereis flexed,brittlefaultingmay allow the entirecrust
to becomehydrated. The 5.5-km-thick oceancrustalso has a
low-velocityzone at the baseof layer 3. This could indicate
alteration
of the lower crust but the thickness of the low-
velocityzone is poorly constrained. Alternatively,alteration
may have occurredat the spreadingridgeprior to oceancrust
flexure[Waltheret al., 2000].
Since the zone of flexure
is 3 times as wide off central
Nicaraguaas off CostaRica and its verticaldisplacementis 3
timesas great(Figure9), fracturezonesoff Nicaraguawill be
exposedto the ocean3 times longer(0.6 Myr) than off Costa
Rica. The sealing of basementacquifersby hemipelagic
sediment, as documented at the Juan de Fuca Ridge
[Giambalvoet al., 2000], will increasethat contrast. Greater
vertical displacementalong the faults off Nicaragua fully
ruptures the sediment but it does not off Costa Rica.
Thereforepore fluid and the fluids bound in altered ocean
crustenteringthe subductionzone are probablymuchgreater
off Nicaraguathan off CostaRica. However, we are not in a
positionto discussthe up-takeof trace elementsin hydrated
and serpentinizedoceancrustand thesegeochemicalaspects
are more fully covered by Patino et al., [2000], who also
relate
ocean
crust
fracture
to
the
variation
in
arc
geochemistry.
As seawateris admittedand migratesdownwardinto the
oceancrust,it reacheselevatedtemperatures
and mafic rock
that serpentinizes easily. Subducted water bound in
serpentine
will be releasedby dehydration
at stageslaterthan
the initial drainage of pore waters from fractures and
subductedclastic materials. Serpentinizedlower ocean crust
and upper mantle are proposedto generatemuch of the
volatile flux along or above subductionzones [cf. Ernst,
1999]. Thereforethe intensityof fracturingas the oceancrust
is flexed into the trench,by its modeof fluid storageandlater
release, could provide a controlling factor in frictional
behaviorand formationof arc magmas.
8. Summary
The resemblancebetween the continental slope and the
adjacentsubductinglower plate morphologiesalong Costa
Rica is clearly resolved with high-resolutionreprocessed
swathbathymetry. Its integrationwith seismicdataprovides
a regional structuralpicture. Differences in ocean crust
thickness, seafloor relief, and in flexure of the Cocos Plate
producestructuraldifferenceson the adjacentmiddleto upper
continentalslopewherethe relativelythin upperplatebeneath
the continentalslopeactsas a coverthroughwhichthe muted
lower platerelief is visible. Subductedseamounts
elevatethe
upper plate into broad seafloordomes that migrate to the
shelf. Uplift from a -2-km-high seamountwasresolvedwell
wheretheupperplateis 7-10 km thick. On the trailingflanks
of thesemigratingseamountdomes,oversteepened
sediment
failureleavesa trail-like slumpscarto the shelfedge. Current
coastaluplift alongprojectionsof seamountchainshasbeen
attributedto migratingseamounts[Marshall and Anderson,
1995; Fisher et al., 1998]. The assembledevidenceindicates
that seamountscan remain attachedto the subductingplate at
least30-60 km into the seismogenic
zoneandto depthsof-25
km.
The frontalprismgrowsrapidly after beingdestroyedby a
subductingseamount,but growth ceaseswhen it reaches510-km width. We speculatethat the prism grows until it
elevatespore pressureto a level where the reducedfriction
allowsmostincomingsedimentto be subducted.The narrow
prism alongthe Middle America Trench storesonly a minor
amountof sediment. Two broad embaymentsin the middle
and upper slope are anomalousto Middle America margin
morphology. Their position oppositea rugged subducting
lowerplatesuggests
an erosionalorigin. In additionto frontal
erosion,erosionalong the undersideof the upper plate is
required.
The near linear
continuation
of the transverse trend of
aligned but perhaps en echelon segmentsfrom Fisher
Seamountto the Quesda contortion of Protti et al. [1995]
corresponds
to an offsetin the volcanicarc. The lower plate
transit time from the trench to the arc is 2 Myr. The more
than 150-km-longsubductedextensionfrom FisherSeamount
to the arc together with the 90-km-long exposed
volcanotectoniclineament is indeed long but of a length
similar to that of the exposedQuepostransverselineament
(Plate1). Despiteits length,the Quepaslineamenthasa less
apparenteffect on continentalstructures. The 15i angle
betweenthe convergencevector and the Queposlineament
causesmigrationof its pointof impactat the trench. With no
boundarysubductedin one place over time, upper plate
deformation
fromlowerplateasperities
maybe dispersed.
The CNS-EPRjunctionalsohas a smallerapparenteffect
despiteits nearaligr •entwith the directionof convergence.
Althoughthisjunctic corresponds
with a majoroceancrustal
boundary,an oceanfloor relief is not obviousat the scaleof
satellitegravity anomalies. Little is known regardingthe
petrologyon either side of the boundary. Clearlythereis a
plate tectonicbreak in the oceaniccrustalongthe CNS-EPR
junction, but not a pronounced break in volcano
geochemistry.
The changein volcanismacrossthe CostaRica-Nicaragua
bordercorresponds
with an older inactivecrustalboundary.
VON HUENE ET AL.: COSTA RICA TECTONICS
AND VOLCANISM
333
upperslopethe basalfriction equalsor doesnot exceedthat of
a seamounts peak or outer layers. Where stick-slipbehavior
begins, friction does not yet appear to exceed the shear
strengthof the seamountmain body.
processes.
The degree to which subductingocean crust affects the
Cocos Plate physical segmentation offers some upperplate is perhapsunusualoff Costa Rica. Unique to the
explanationsfor geochemicaldifferencesin arc lavas and has
Costa Rica margin is the extent of a segmentboundaryalong
been consideredin various comprehensivegeochemical a trend that subductsessentiallynormal to the margin and
studies. The low Ba/La ratio in central Costa Rican lavas and
therefore in one area for a significantperiod. Cocos Plate
the very high ratiosof Nicaragua[Cart et al., 1990] are not
segmentationis emphasizedmorphologicallyat the seafloor
explainedby accretionversusnonaccretion
or by a subduction by hot spot volcanic construction,and this morphology
erosion input (i.e., dilution of subductedsediment with a
facilitates geophysicaldetectionthrough the upper plate
fluid-richslurryerodedfromthebaseof theupperplate). The despitedeepsubduction.
The Nicaraguaseismogenic
zonesteepens,
and volcanoesare
locatedabovea deeperpart of the subduction
zonerelativeto
CostaRica. Here crustalconfigurationratherthanoceancrust
compositional differences may dominate segmentation
intensityof normalfaultingthatallowschargingof the upper
oceanic
crust
with
water
offers
another
constraint
in
explainingthe differencebetweenthe high slab signal in
Nicaragua and the low one of central Costa Rica. Circulation
of water along faults, serpentinization,and later release of
boundwater deep in the subduction
zone duringserpentine
dehydration
areintriguingsubjects
for furtherstudy.
Basalfrictionin the aseismicpartof the subduction
zone
canbe derivedfrom the persistence
of structures
in subducted
sedimenton the lower plate. Below the frontalprism,basal
frictionis locallyequalto or lessthanthe shearstrengthof the
sediment section on the Cocos Plate. Beneath the middle and
Acknowledgments.The ResearchVessel Sonne,supportedby
the Bundesministeriumfr Forschungund Technologie (German
Federal Researchand TechnologyAgency), is the platform from
whichmostof the datain this studywere acquired.Discussions
with
M. Carr and K. Hoernle were particularlyhelpful. We thank M.
Protti, who providedus with an electronicfile of his earthquake
epicenters,and Harry Doust of Shell InternationalePetroleum
Mij.B.V., whoprovidedthefieldtapesof Line P1600. We alsothank
StewartSmithof the ScrippsInstituteof Oceanography,
Geological
Data Center,who provideddatafrom the RogerRevelleacceptance
cruise. Reviewsby Terry Plank,an anonymous
reviewer,and Gaku
Kimurahelpedgreatlyin improvingan earlierversion,especiallythe
geochemicalaspects.
References
Abratis, M., Geochemicalvariationsin magmatic
rocks of southernCosta Rica as a consequence
of Cocos Ridge subductionand uplift of the
Cordillera de Talamanca,Ph.D. thesis, 189 pp.,
G ettingenUniv., G ettingen,Germany,1998.
Alvarado, G.E., S. Kussmaul, S. Chiesa, P.-Y.
Gillot, H. Appel, G. W6rner, and C. Rundle,
Resumen cronoestratigr fico de las rocas
igneas de Costa Rica basado en dataciones
radiom tricas, S. Am. J. Earth Sci., 6, 151-168,
1992.
Arroyo, I., The seismicityin Costa Rica, editedby
Bialas, J., E.R. Flueh, and G. Bohrmann, Sonne
Cruise SO144 l&2, PAGANINI, GEOMAR
Report 94, 16-22, 1999.
Baltuck, M., E. Taylor, and K. McDougall, Mass
movementalong the inner wall of the Middle
America Trench, Costa Rica, in Initial Rep.
Deep SeaDrill. Proj., 84, 551-570, 1985.
Balzer, V.G., The late Miocene history of sediment
subductionand recycling as recorded in the
Nicaraguavolcanic arc, M.S. thesis,Univ. of
Kansas, Lawrence, 1999.
Barckhausen,U., H.A. Roeser, and R. von Huene,
Magnetic signature of upper plate structures
and subductingseamountsat the convergent
margin off Costa Rica, J. Geophys.Res., 103,
7079-7093, 1998.
Carr, M.J., and R.E. Stoiber, Volcanism, in The
Geology of North America, vol. H, The
CaribbeanRegion,editedby G. Dengoand J.E.
Case,pp. 375-391, Geol. Soc.of Am., Boulder,
Colo., 1990.
Carr, M.J., M.D. Feigenson, and E.A. Bennett,
Incompatibleelementand isotopicevidencefor
tectonic control of source mixing and melt
extraction along the Central American Arc,
Contrib. Mineral. Petrol., 105, 369-380, 1990.
Caress, D.W., and D.N. Chase, Improved
processingof HydrosweepDS multibeamdata
on the RV Maurice Ewing, Mar. Geophys.
Res., 18, 631-650, 1996.
Christeson, G.L., K.D. Mcintosh, T.H. Shipley,
E.R. Flueh, and H. Goedde, Structure of the
CostaRica convergent
margin,offshoreNicoya
Peninsula,J. Geophys.Res., 104, 25, 443-25,
468, 1999.
Collins, L.S., A.G. Coates,J.B.C. Jackson,and J.
Obando,Timing andratesof emergenceof the
Flores-Hernandez,
A.,
Processing and
interpretation of multichannel seismic
reflection
data on continental
and oceanic crust
of the Pacific margin of Costa Rica, diploma
thesis, 87 pp., GEOMAR Forsch. fr Mar.
Limon and Bocas del Toro basins: Caribbean
Geowissenschaften
effectsof Cocosridge subduction?,
in Geologic
and TectonicDevelopmentqf the Caribbean
Plate Boundary in SouthernCentral America,
editedby P. Mann, Spec.Pap. Geol. Soc.Am.,
Univ. zu Kiel, Kiel, Germany, 1996.
295,263-289, 1995.
Corrigan, J., P. Mann, and J.C. Ingle, Forearc
responseto subductionof the CocosRidge,
Panama-Costa
Rica,Geol.Soc.Am.Bull.,102,
628-652, 1990.
deBoer,
J.Z.,M.S. Drummond,
M.J.Bordelon,
H.
Bellon,andR.C. Maury,Cenozoic
magmatic
phases
of theCostaRicanislandarc(Cordillera
de Talamanca),Spec.Pap., Geol.Soc.Am.,
Gardner, T.W.,
der
Christian-Albrechts
D. Verdonck,
N. Pinter, R.
Slingerland,K. Furlong, T.F. Bullard, and S.G.
Wells, Quaternary uplift astride the aseismic
Cocos Ridge, Pacific coastof Costa Rica, Geol.
Soc. Am. Bull., 104, 219-232, 1992.
Giambalvo, E.R., A.T. Fisher, J.T. Martin, L.
Darty, and R.P. Lowell, Origin of elevated
sediment permeability in a hydrothermal
seepagezone, easternflank of the Juan de Fuca
Ridge, and implications for transportof fluid
andheat,J. Geophys.Res., 105, 913-928, 2000.
Gr•ife, K., Exhumation and thermal evolution of
the Cordillera de Talamanca (Costa Rica):
Constraintsfrom fission track analysis,40 Ar39 Ar, and 87 Rb-87 Sr chronology,Tbinger
Geowissenschafiliche
Arb., Set'.A, Vol. 39, 113
pp., 1998.
Tectonophysics,
293, 207-224, 1998.
Ernst, W.G., Hornblende,the continentmaker: Heeren, F., Erweiterung der HydrosweepDatenverarbeitung und Anwendung auf
Evolution of H20 during circum-Pacific
morphotektonische Untersuchunger der
subduction
versus
continent
collision,
Geology,
Subduktionszone
vor CostaRica, Diplomarbeit,
27, 675-678, 1999.
69 pp., GEOMAR, Christian-Albrechts-Univ.
Feigenson, M.D., M.J. Carr, L.C. Patino, S.
zu Kiel, Kiel, Germany, 1996.
Marahaj, and S. Juliano,Isotopicidentification Herrstrom, E.A., M.K. Reagan, and J.D. Morris,
of distinct mantle domains beneath Central
Variations in lava compositionassociatedwith
America, Geol. Soc.Am., Abstr. Programs,p.
flow of asthenosphere
beneathsouthernCentral
A152, 1996.
America, Geology,23, 617-620, 1995.
Fisher, A.T., Permeability within basaltic oceanic Hey, R.N., A new class of pseudofaultsand their
crust,Rev. Geophys.,36(2), 143-182, 1998.
bearing on plate tectonics: A propagatingrift
295, 35-55, 1995.
Dominguez,
S., S.E.Lallemand,
J. Malavieille,and
R. von Huene, Upper plate deformation
associated with seamount subduction,
Fisher, D.M., T.W. Gardner, J.S. Marshall, P.B.
Sak, and M. Protti, Effect of subductingseafloor roughnesson fore-arckinematics,Pacific
coast, Costa Rica, Geology, 26(5), 467-470,
1998.
Fisher,R.L., Middle America Trench: Topography
and structure, Geol. Soc. Am. Bull., 72, 703720, 1961.
model, Earth Planet. Sci. Lett., 37, 321-325,
1977.
Hinz, K., R. von Huene, C. R. Ranero, and the
PACOMAR
Working Group, Tectonic
structure of the convergent Pacific margin
offshore Costa Rica from multichannel
seismic
reflection data, Tectonics,15, 54-66, 1996.
Hsu, J.T., Emerged Quaternarymarine terracesin
334
VON HUENE ET AL.: COSTA RICA TECTONICS AND VOLCANISM
southern Peru:
Sea level changes and
continental margin tectonics over the
subductingNazca Ridge, Ph.D. thesis,309 pp.,
Cornell Univ., Ithaca, N.Y., 1988.
Kimura, G., et al., Proceedings qf the Ocean
Drilling Program Initial Reports, vol. 170,
OceanDrill. Program,College Station,Tex.,
1997.
Kolarsky, R.A., P. Mann, and W. Montero, Island
arc response to shallow subduction of the
Cocos ridge, CostaRica, Spec. Pap. Geol. Soc.
Am., 2 95, 235-262,
1995.
Langseth, M.G., and E.A. Silver, The Nicoya
convergentmargin: A region of exceptionally
low heat flow, Geophys. Res. Lett., 23, 891894, 1996.
Leemah, W.P.,
and
M.J.
Carr,
Geochemical
constraints on subduction processes in the
CentralAmericanVolcanic Arc: Implications of
boron geochemistry,Spec.Pap. Geol. Soc. Am.,
295, 57-73, 1995.
Leeman, W.P., M.J. Carr, and J.D. Morris, Boron
geochemistryof the Central Americanvolcanic
arc: Constraintson the genesisof subductionrelatedmagmas,Geochim. Cosmochim.Acta, 58,
149-168,
1994.
Magee, M.E.. and M.D. Zoback, Evidence for a
weak interplatethrust fault along the northern
Japansubductionzone and implicationsfor the
mechanics of thrust faulting and fluid
expulsion, Geology,21, 809-812, 1993.
Marshall,J.S., and ILS. Anderson,Quaternaryuplift
and seismic cycle deformation, Peninsula de
Nicoya, Costa Rica, Geol. Soc. Am. Bull., v.
10 7, 463-473,
199 5.
McAdoo, B.G., D.L. Orange, E.A. Silver, K.
Mcintosh, L. Abbott, J. Galewsky, L. Kahn, and
M. Protti, Seafloor structural observations, Costa
Rica accretionaryprism,Geophys.Res.Lett., 2 3,
883-886,
1996.
Mcintosh, K., E. Silver, and T. Shipley, Evidence
and mechanisms
for
forearc
extension
at the
accretionary Costa Rica convergent margin,
Tectonics, 12, 1380-1392,
1993.
Morgan, J.P., A. Harding, J. Orcutt, G. Kent, and
Y.J. Chen,
An
observational
and theoretical
synthesis of magma chamber geometry and
crustal genesis along a mid-ocean ridge
spreadingcenter in Magmatic Systems,edited
by M.P. Ryan, Chap. 7, pp. 139-178. Academic
Press,Inc., San Diego, Calif., 1994.
Moritz, E., Bornholdt, S., Westphal, H., and
Meschede,M., Neural network interpretationof
LWD data (ODP Leg 170) confirmscomplete
sediment
subduction
at
the
Costa
Rica
Convergent Margin, Earth and Platernary
Science Letters, 174,301-312,
2000
Morris,J.D., Applications
of cosmogenic
løBeto
problemsin the earthsciences,Annu. Re,,. Earth
Planet. Sci., 19, 313-350,
1991.
Mrazek, J., T. Spangenberg,and R. yon Huene,
Geologische
und
Geophysikalische
Untersuchung vor Costa Rica und Nicaragua-
Beitrfige zum Vers/indnis des aktiven
ostpazifischenKontinentalrandes,FS Sonne
Fahrtbericht
SO
Balboa/Panama-San
14.04,
107,
Pacomar
3,
Francisco/USA,
10.03.-
Ernst-Moritz-Arndt Univ.
Greifswald,
Germany,and GEOMAR, Kiel, Germany,1996.
Mrazek,J., D. Korich, T. Spangenberg,K. Kreisel,
and N. Stroncik-Treue, Geologische und
Geophysikalische
Untersuchungen
in
Seegebietenvor Costa Rica und NicaraguaBeitr•ige zum Verstfindnis des aktiven
Ostpazifischen
Kontinentalrandes
(ForschungsfahrtSO 107), 182 pp., EmstMoritz-Arndt Univ., Greifswald,Germany,1999.
Patino, L.C., M.J. Cam and M.D. Feigenson, Local
and regional variations in Central American
lavas controlled by variations in subducted
sedimentinput, Contrib. Mineral. Petrol., in
press,2000.
Pecher, I.A., C.R.
Ranero, R.
yon
Huene, T.A.
Minshull, and S.C. Singh, The natureand
distributionof bottom simulatingreflectorsat
the Costa Rican convergentmargin, Geophys.
J. Int., 133, 219-229, 1998.
Plank, T., and C.H. Langmuire, Tracing trace
elements from sediment input to volcanic
output at subductionzones,Nature, 362, 739742, 1993.
Protti, M., F. Gundel, and K. McNally, Correlation
betweenthe age of the subductingCocosplate
and the geometry of the Wadati-Benioff zone
under Nicaragua and Costa Rica, Spec. Pap.
Geol. Soc. Am., 295, 309-326, 1995.
Ranero, C.R., and R. von Huene, Subduction
erosionalong the Middle America Convergent
Margin, Nature, in press,2000.
Ranero, C.R., R. von Huene, E. Flueh, M. Duarte,
D. Baca, and K. Mcintosh, A cross section of
the convergentPacific margin of Nicaragua,
Tectonics,this issue.
Ranero, C.R., R. von Huene, H. Steiner, and
watchkeepers, Multibeam swath mapping,
edited by Bialas, J., E.R. Flueh, and G.
Bohrmann,
PAGANINI,
Sonne
Cruise
GEOMAR
SO144
l&2,
Report 94, 88-93,
1999.
Sailares, V., J.J. Danobeitia, E.R. Flueh, and G.
Leandro,Seismicvelocity structureacrossthe
middle Americanlandbridgein northernCosta
Rica, Geodynamics,27, 327-344, 1999.
Sapper, K., Uber die raumliche Anordnungder
millelamerikanischen Vulkane, Z. Dtsch.Geol.
Ges., 49, 672-682, 1897.
Shipley, T.H., and G.F. Moore, Sediment
accretion,subduction,and dewatering at the
base of the trench slope off Costa Rica: A
seismic reflection view of the d collement, ar.
Geophys.Res.,91, 2019-2028, 1986.
Shipley, T.H., P.L. Stoffa, and D.F. Dean,
Underthrustsediments,fluid migrationpaths,
and
mud
volcanoes
associated
with
the
accretionarywedge off Costa Rica: Middle
America Trench, J. Geophys.Res., 95, 87438752, 1990.
Shipley, T.H., K.D. Mcintosh, E.A. Silver, and
P.L. Stoffa, Three-dimensionalseismicimaging
of the CostaRica accretionaryprism:Structural
diversityin a small volume of the lower slope,
J. Geophys.Res.,97, 4439-4459, 1992.
Stavenhagen,A. U., E.R. Flueh, C. Ranero,K. D.
Mcintosh, T. Shipley, G. Leandro,A. Schulze,
and J.J. Dafiobeitia,
Seismic wide-angle
investigations
in CostaRica: A crustalvelocity
model from the Pacific to the Caribbean Coast,
ZBL. Geol. Palaontol., Part I, 393-408, 1998.
von Huene, R., D. Klaeschen,M.-A. Gutscher,and
J. Fruehn,Mass and fluid flux duringaccretion
at the AlaskaMargin, Geol. Soc.Am. Bull., 110
(4), 468-482, 1998.
von Huene, R., et al., Morphotectonics of the
Pacific convergentmargin of Costa Rica, in
Geologic and Tectonic Development of the
CaribbeanPlate Boundaryin SouthernCentral
America, edited by P. Mann, Spec.Pap. Geol.
Soc.Am., 295, 291-308, 1995.
Walther, C.H.E., E.R. Flueh, C.R. Ranero, and R.
von Huene, Seismic investigations at the
Pacific Margin of Nicaragua, Geophys.d. Int.,
in press,2000.
Wang, K., T. Mulder, G. Rogers,andR. Hyndman,
Case for very low coupling stress on the
Cascadia subduction fault, J. Geophys. Res.,
100, 12,907-12,918, 1995.
Weinberg,R., Neotectonicdevelopment
of western
Nicaragua,Tectonics,11, 1010-1017,1992.
Werner, R., K. Hoernle, P. van den Bogaard,C.R.
Ranero, R. von Huene,
and D.
Korich,
Drowned 14 m.y. old Galapagosarchipelago
off the coast of Costa Rica: Implications for
tectonicand evolutionarymodels,Geology,27,
499-502, 1999.
Ye, S., J. Bialas, E.R. Iqueh, A. Stavenhagen,R
yon Huene, G. Leandro, and K. Hinz,
structure
of the
Middle
American
Crustal
Trench
off
Costa Rica from wide-angle seismic data,
Tectonics, 15, 1006-1021,
1996.
Zoback, M.D., and G.C. Beroza, Evidence for near-
frictionlessfaulting in the 1989 (M6.9) Loma
Prieta,California, earthquakeand its aftershocks,
Geology,21, 5, 181-185, 1993.
K. Hinz,
Bundesanstalt far Geowissenschaften
und Rohstoffe,D30655 Hannover,Germany.
C.R. Ranero, R. yon Huene, and W. Weinrebe,
GEOMAR Research Center for Marine Geosciences,
Christian Albrechts Universita't, Wischhofstr. 1-3.
D-24148 Kiel, Germany.(rhuene@geomar.
de)
(Received June 25, 1999:
revised December 8, 1999;
acceptedDecember13, 1999.)