JOURNALOF GEOPHYSICAL RESEARCH, VOL. 102, NO. B6, PAGES 12,035-12,059, JUNE 10, 1997
The Chile ridge: A tectonic framework
$.F. Tebbens,1 S.C. Cande,2 L. Kovacs,3 J. C. Parra,4,5 J.L. LaBrecque,6 and H. Vergara7
Abstract.A newChile ridge tectonic framework isdeveloped based on satellite altimetry data, shipboardgeophysical data and, primarily, 38,500 km of magnetic data gathered on a joint U.S.- Chileanaeromagnetic survey. Eighteen active transforms with fossil fracture zones (FZs), includingtwo complex systems (the Chile FZ andValdivia FZ systems),have been mapped betweenthe northern end of theAntarctic-Nazca plate boundary (Chile ridge) at 35øS and the Chile margintriple junction at 47øS.Chile ridge spreading rates from 23 Ma to Presenthave been determinedand show slowdowns in spreading rates that correspond totimes of ridge-trench collisions.The ValdiviaFZ system,previously mapped as two FZs withan uncharted seismicallyactive region between them, is now recognized to bea multiple-offsetFZ system composedof six FZs separatedby shortridge segments 22 to 27 km long. At chron5A (-12 Ma),the Chile ridge propagated from the Valdivia FZ systemnorthward into the Nazca plate throughcrust formed 5 Myr earlierat thePacific-Nazca ridge. Evidence for this propagation event includesthe Friday and Crusoe troughs, located at discontinuitiesin the magnetic anomaly sequenceand interpreted as pseudofaults. This propagation event led to theformation of theFriday microplate,which resulted in thetransferal of crustfrom the Nazca plate to theAntarctic plate, and ina 500-kmnorthward stepwise migration of thePacific-Antarctic-Nazca triple junction. Rift propagation,microplate formation, microplate extinction, and stepwise triple junction migration arefound to occurduring large-scale plate motion changes and plate boundary changes in the southeastPacific.
Introduction improved due to the acquisition and analysis of additional geophysical traverses of the Chile ridge [Herron, 1972; The Chile ridge is a major branch of the global ridge Klitgord et aI., 1973; Handschumacher, 1976; Herron et al., system,extending from the junction of the Pacific, Antarctic, 1981] and Chile FZ [Anderson-Fontanaet al., 1987]. In the andNazca plates at the JuanFernandez microplate [Larson et immediate vicinity of the Chile margin triple junction, the al., 1992, and referencestherein] to the triple junctionof the chartingof ridge axis segments,FZs, and magneticisochrons Antarctic,Nazca, and SouthAmerican plates near 46øS along was furtherimproved by detailedshipboard surveys [Cande and the southernChile trench [Herron et al., 1981; Cande et aI., Leslie, 1986; Cande et al., 1987a]. 1987b](Figure 1). In January 1990, a team of scientistsfrom the U.S. Naval Theexistence of a mid-ocean'ridge between Rapa Nui (Easter Research Laboratory (NRL), Lamont-Doherty Earth Island)and southern South America was first predicted, on the Observatory(LDEO), the Servicio Nacional de Geologiay basisof seismicevidence, by Ewing and Heezen[1956]. The Mineria,and the Servicio Hidrografico y Oceanografico dola presenceof the Chile ridge was confirmedby bathymetricdata Armada de Chile conductedan aeromagneticsurvey of the collectedon the Scripps Institution of Oceanography Chile ridge. The aeromagneticsurvey and satellitealtimetry Downwindexpedition [Menard et al., 1964]. The spreading data [Sandwell,1993; Smith and Sandwell,1995] form the patternon the Chile ridge was first mappedby Herron and basis for an improvedcharting of the tectonicfeatures of the Hayes[1969] and Herron [1971], using shipboard data, and by Chile ridge and its flanks. Morganet aI. [1969],using aeromagnetic data. Knowledgeof This paper discussesthe tectonicfeatures observed along magneticanomalies and fracture zone (FZ) locationshas the Chile ridge and its flanks. The most significantnew feature is the Friday microplatewhich formedduring a chron 5A (-12 Ma) plate boundaryreorganization. A companion IDepartmentofMarine Science, University ofSouth Florida, St. paper[Tebbens and Cande,this issue]presents reconstructions Petersburg,Florida. 2ScrippsInstitution ofOceanography, LaJolla, California. of the southeastPacific sincethe late Oligocene,showing the 3NavalResearch Laboratory, Washington, D.C. evolution of the Chile ridge within the framework of the 4ServicioNacional deGeologia yMineria, Santiago, Chile. southern Pacific Ocean. 5NowatGEODATOS, Santiago, Chile. %amont-DohertyEarthObservatory ofColumbia University, Palisades,New York. Data 7ServicioHidrografico y Oceanografico dola Armada, Valparaiso, Chile. Tectonicinterpretations are presentedbased on all available topographic,magnetic, and satellite altimetry data in the Copyright1997by the American Geophysical Union. surveyregion, including the 1990 aeromagneticsurvey of the Papernumber 96JB0258 !. Chile ridge. Bathymetryand magneticdata from a 1988 RRS 0148-0227/97/96JB.02581$09.00 Charles Darwin survey are presented,including four ridge-
12,035 12,036 TEBBENSET AL.: CHILE RIDGE TECTONIC FRAMEWORK
120"W 115"W 11OøW 105"W 1O0"W 95øW 90"W 85øW 80"W 75øW 70"W 30øS
Nazca PlateFigures 7a & 7b Figures 4a & 4b 35øS Pacific Plate
40øS
ures 6a & 6 ,divi&..F.Z...•
45øS
-fa'•tao Figures 5a & 5b
'?. 5O S 50øS Antarctic to
.
120øW 115"W 110øW 105"W 100'W 95øW 90øW 85•W 80øW 75øW 70øW
Figure 1. Major tectonic featuresof the SouthPacific includingridge axes (heavy solid lines), FZs (solid and dashed lines are known and inferred locations, respectively), nontransformoffset traces (dash-dotted lines), fossil ridge axes (railroad track pattern), transferredlithosphere (striped areas),PAC-ANT-FAR triple junction trace (TJT) (dotted lines with heavy dots), and other tectonicdiscontinuities (heavy lines with dots). The 1990 aeromagnetic survey lines are shown (light dotted lines). Diagonal shading denotes transferred lithosphere. Location maps for Figures 4, 5, 6, and 7 are shown.
perpendicularprofiles south of the Guafo FZ. Fourpreviously The altitude at which aeromagneticdata are collectedaffects unpublishedaeromagnetic lines are included which were the resolutionof the recorded signal. When weather permitted, navigated with a combination of LORAN and inertial lines were flown at approximately 500 m elevation, whichis navigation: two lines cross the Chile ridge axis, one near less than 4.0 km over the seafloor at the ridge axis. At this 38.2øS and the other near 39.5øS; and two are east-west lines low altitude, the magneticsignal is comparableto a shipboard located between 35øS and 36øS, to the north of the Chile ridge survey. A few sectionswere flown at altitudesas high as3 krn, (D. W. Handschumacher,unpublished data, 1980). due to poor weatherand equipmentproblems, resulting in the In January1990, a P-3 Orion aircraftequipped with a proton loss of the short-wavelength information in these track precessionmagnetometer collected a total of 37,500 km of segments. aeromagneticdata (Figures 1 and 2). The survey lines were Magnetic anomalieswere identified by comparingthem to a oriented along seafloorspreading flow lines determinedprior model basedon the known sequenceof geomagneticpolarity to the survey by examining altimetry data for FZ orientations. reversals [Cande and Kent, 1992, 1995]. Representative Total intensity magnetic measurements were reduced to magneticprofiles are easily correlatableand identifiablewhen anomaly form by subtracting the 1985 International comparedto a modelprofile (e.g., Figure 3). We includedor, Geomagnetic Reference Field [International Association of when necessary, modified the magnetic anomaly Geomagnetismand Aeronomy (IAGA), 1987] correctedfor 5 identificationsand/or FZ locationsof Morgan et al. [1969], years of secular variation. The flights were almost entirely Herron and Hayes [1969], Herron [1971], Klitgordet al. navigatedwith Global PositioningSystem (GPS). During two [1973], Handschumacher [1976], Herron et al. [1981], short flight segmentswhen GPS was not available,OMEGA LaBrecque[ 1986], Candeand Leslie [ 1986], Anderson-Fontana navigationwas used. The OMEGA-navigatedsegments are the et al. [ 1986], Molnar et al. [ 1975], and Lonsdale [1994]. sections of the first and fourth aeromagnetic tracks north of Satelliteradar altimetry data were combined with shipboard the Guafo FZ and west of the ridge axis (181 and 355 km long, andaeromagnetic data to mapthe locationof ridgeaxes and respectively)(Figure 2). Small but significantdeviations (of FZs. Color imageswere createdusing the gravityanomaly the order of 4 kin) in the linearity of the magnetic isochrons gridof Smithand Sandwell[1995] whichincorporated satellite acrossthese track segmentsare believed to be artifactsdue to altimetryprofiles from ERS 1, Geosat,and Seasat. The grid inaccuracies in the OMEGA navigation relative to the GPS spacingis 1/20th of a degree in longitude. In this paper, navigation. In other regions with only GPS navigation, "gravityfield" refersto the Smithand Sandwell[1995] gravity offsetsof the order of 4 km are interpretedto be real. grid. TEBBENSET AL.: CHILERIDGE TECTON1C FRAMEWORK 12,037
100øW 95øW 90"W 85"W 80"W 75 øW _ ;-_-___ ß -...... --- ...... • ...... :...... •.•/(.n_,_e_.• ...... =•%:...... 57"ti,•e 5 .->-<:::'".:'.' .... / ...... d,! .._ ...... i -• -. --••;::.:_v. •: :'...... '...... I ...... '...... ' ...... ' ...... """"' :'•"1"'"•"• ...... 40 S I'...... ,-...... 40"S - ..--..--...... L•"2"'2'..... ;.L.'..._-.....-....-...----2-.•."•-'•"%5'•T-..-.'z'--•'•';";:•'...... Z/--- ._C':•.•j-__•_.-.• - - "---.-z.- - Valdivia FZ System
line..22...... ß line16 .
45øS 45"S GuanoFZ
GuamblinFZ
Ta:ttaøFZ .•..½.
50"S 50øS 1ooow 95øW 90øW 85øW 80øW 75øW
Figure 2. Location of aeromagneticprofiles shown in Figure 3. Conventionsas in Figure 1.
Locationof Ridge Axes and Fracture Zones In this work, FZs were located using magnetic and from Altimetry Data bathymetry data, where available, and using satellite altimetry data where traditional data coverage wa• sparseor unavailable. Chile ridge axis segmentshave previously been shown to The observedbathymetric expressionsof FZs differ from the havetopographic axial rift valleys [Herron and Hayes, 1969; theoretical models. The majority of FZs observed in Herron, 1972; Klitgord et al., 1973; Herron et al., 1981; bathymetry profiles in this survey region are topographic Cande and Leslie, 1986]. Small and Sandwell [1989] troughs, as also observed at FZs formed at the fast spreading demonstratedthat axial rift valleys are generally recordedas East Pacific Rise (EPR) [Fox and Gallo, 1989] and at the Kane lows in satellite altimetry data. Chile ridge axis segments FZ formed at the slow spreading Mid-Atlantic ridge [Miiller et (locatedwith magnetic data) are found to be associatedwith ai., 1990]. At the Kane FZ the mismatch between the gravityfield lows (Plate 1), which are interpreted to be axial topographiclow and geoid low for 37 profiles was--5 km, on rift valleys. In axial regionswhere the only geophysicaldata average [Miiller et al., 1990]. For a FZ associated with a arealtimetry measurements, ridge axis locationswere charted topographic trough the FZ location correspondsto the point alongaltimetry lows. at, or near, a geoid minimum [e.g., Mayes et al., 1990]. Theoreticalmodels vary in their predictionof the location Where no bathymetry data were available along the Chile of a FZ alonga geoidprofile. Sandwell[1984] considereda ridge, a topographictrough was presumed. FZs and transforms simple theoretical cooling model of the oceanic crust, were located at or near the geoid minimum, with the additional modelinga FZ as a stepin basementwith older,colder, denser constraint that on both flanks of each ridge axis segmentthe cruststepping up to younger,warmer, less densecrust. The distancebetween FZs was charted to be as equal as possibleat geoidmimicked that stepin basement,with the FZ locatedat conjugate distancesfrom the axis. themaximum gradient of thegeoid anomaly. Parmentier and Haxby [1986] consideredthermal bendingstresses which placedthe FZ slightlyoffset from the locationof the steepest Tectonic Maps and Interpretation gradient.Wessel and Haxby [1990] consideredadditional complexitiesand have a model which accountsfor slip and Detailed figures (index map, Figure 1.), including the includesflexure from a combination of thermal stressesand magnetic and bathymetric profiles on which the tectonic differentialsubsidence. Wessel and Haxby's[1990] model interpretation is based, were prepared for the regions north alsoplaces the FZ slightlyoffset toward the trough from the (Figures 4a and 4b) and south (Figures 5a and 5b) of the. locationof steepestgradient. Valdivia FZ system. A similar set of figures was compiledfor 12,038 TEBBENSET AL.: CHILERIDGE TECTONIC FRAMEWORK
line1 (dh016)
line2 (pro904) FZ 39 line3 (dh016) line4 (prn904) ine5 (prn908) FZ 40A / / / ! line6 (pro908) ValdiviaFZSystem line7 (pro909) I / / line8 (pro902) line9 (pro907) line10 (pro902) line11 (pro907) ChiloeFZ• • • line12 (pro901) line13 (pro901) line14 (pro903) line15 (pro901) line16 (pro903) line17 (prn901)
[600nT Guafo FZ I I t / • [ [ [ line 18 (chd36) line 19 (chd36)
6 5c 5A 5 4 3 2A AX 2A3 4 5 5A 5C
Model
Figure 3. Magneticprofiles used in the spreadingrate analysis, projected at N80øE. At bottomis a model magneticprofile created using the short interval spreading rates of Table2. Stipplingdenotes data not used for the spreadingrate analysis.
the regions southwest(Figures 6a and 6b) and northeast Anderson-Fontana et al. [1986], which is basedon an RN (Figures7a and7b) of the Chile ridge. Endeavorsurvey. At present,the Chile ridge extendsfrom the JuanFernandez Between the Chile and Valdivia transform fault systems microplate to the Chile margin triple junction and is there are six transform faults and associated FZs. With composedof 1380 km of ridge axis offset by 18 active improveddata, particularly altimetry data, we find the FZ transformfaults with fossilFZs, includingtwo complexfault orientationsin this regionare ~N88øE, about8 ø differentfrom systems. In addition, three nontransformoffsets and their off- those of Herron et al. [1981] and Mamrnerickx and Smith axis traceshave beenidentified. The lengthof eachridge [1980] (--N80øE). segment ranges from a maximum of 234 km between the FZ 37B (Figures4a and 4b) waspreviously uncharted. Two Chiloe and Guafo FZs to 39 km between FZ 37 and FZ 37B to gravityfield lows offsetby 94 km are interpretedto be ridge shorterlengths within the complexFZs (Table 1). The sense axis segmentsoffset by the transformassociated with FZ 37B. of shearis right-lateralalong all the transformfaults, except The locationof •e ridge axis segmentsouth of FZ 37B is FZ 40A immediatelynorth of the Valdiviasystem. furtherconstrained by a prominentcentral magnetic anomaly anda topographicallywell-defined rift valley. Thelocation of Chile Ridge North of the Valdivia FZ theridge axis segment north of FZ 37B is onlyconstrained by The tectonicmap of the northernChile ridge region,north a gravity field low. of the Valdivia FZ system,is basedon magneticprofiles Apparenttectonic complexities between FZ 37 andFZ 38 (Figures4a), bathymetryprofiles (Figure 4b) andthe gravity ßare observed in thegravity field. On the westflank there is a field (Plate 1). The 1300-km-longactive Chile transformfault linear low that extendssoutheast from FZ 37 at 98øWto the systemconnects the Juan Fernandezmicroplate to the ridgeaxis, just northof FZ 38B. Thereis a lessprominent northernmostsegment of the Chile ridge. The interpretationconjugate low on the east ridge flank. The trendsin the of the Chile transformfault and associatedFZs is after altimetrydata may record a pairof pseudofaultsthat formed by TEBBENSET AL.: CHILERIDGE TECTONIC FRAMEWOl• 12 039
C) 12,040 '-FEBBENSET AL.' CHILE RIDGE TECTONIC FRAMEWORK
! I I I TEBBENSETAL.: CHILE RIDGE TECTONIC FRAMEWORK 12,041
I I I
, .\
I I I I I I 1_ 12,042 TEBBENSET AL.: CHILE RIDGE TECTONIC FRAMEWORK
I I I
o '• (x) i
¸
¸
I I I I TEBBENSET AL.: CHILE RIDGE TECTONICFRAMEWORK 12 043
I I
C;
o•
o I
I I
I I I I 12,044 TEBBENSET AL.: CHILE RIDGE TECTONIC FRAMEWORK
N LI. o TEBBENSET AL.: CHILERIDGE TECTONIC FRAMEWORK 12 045
105"W 100øW 95øW 90 W
V !dwi•aFZ S . 5•
•r 5D
5D 5C
45"S 45"S 6 qE 5D 5C •'
., 2500mq 3500rn / ru,, . •,,• 4500m-•
,•, .
50"S 50"S
105øW 100øW 95"W 90øW
Figure 6b. Tectonic interpretationof the Antarctic plate at the boundaryof the southwestflank of the Chile ridge and southeastflank of the Pacific-Antarctic ridge and all available bathymetry data. The track lines represent3.5 km depth for the profiles. Contouredbasemap is satellite derived gravity field of Smith and Sandwell [1995]; contour interval is 5 mGal; labeling interval of 20 regal. Other symbols and conventions are the same as in Figure 6a.
a southwardpropagating ridge. The data are not definitive,and event which formed thesetroughs also transferredlithosphere thesefeatures are not includedin the tectonicinterpretation. from the Nazca plate to the Pacific plate, just west of the FZ 38B, previouslyuncharted, offsets magneticanomalies Friday trough, and formed a microplate, herein named the at the ridge axis by 102 km (Figures 4a). FZ 38B is a Friday microplate. prominentfeature in the gravity field (Plate 1) and is recorded Magnetic data. Between the Valdivia FZ systemand the asa bathymetriclow in availablebathymetry data (Figure4b). Chile FZ system, the flanks of the Chile ridge record BetweenFZ 40A and the Valdivia FZ system,there is a continuous magnetic anomaly sequences from the central small(-5 km), previouslyuncharted, off-axis isochronoffset anomaly to anomaly 5A at the Friday and Crusoe troughs namedFZ 40B. The shortoffset of anomalies2A and2 (Figure (Figure 4a). Across each of these troughs we observe a 4a) is constrainedby an aeromagneticline to the southof the discontinuity in the magnetic anomaly sequence. On the offset,and a satellite-navigatedcruise to the north (R/V Nazca plate, where more magneticdata are available, magnetic ThomasWashington, 1972, cruisesot02). A topographic profiles which cross the Crusoe trough (locations shown in profilealong the ridgeaxis shows no evidencefor a FZ at this Figure 8) are projectedalong flowlines and comparedto model latitude,suggesting no transformnow exists (Figure 4b). profiles (Figure 9). For the region between the central magnetic anomaly and anomaly 5A, model spreading rate Friday and Crusoe Troughs profiles were createdusing averageChile ridge spreadingrates The most prominentnew features chartedbetween the (Table 2). To the east of the Crusoe trough, from anomalies ValdiviaFZ and Chile FZ are a pair of topographictroughs 5D through 6B, model spreading rates were created using whichbound the limits of seafloorproduced by Chile ridge latitude-dependentspreading rates (Table 3). In this region the spre•,ding:the Friday and Crusoetroughs. The troughsare observed isochrons "fan" from north to south, with spreading visibleas continuouslows in the gravityfield (Plate 1). As rates increasingto the south(Figures 4a and 7a). Thus the use will be shownin the tollowingdiscussion, the propagation of a single (average) spreading rate for each interval for 12,046 TEBBENSET AL.: CHILE RIDGE TECTONIC FRAMEWORK
9-11HO TEBBENSET AL.: CHILERIDGE TECTONIC FRAMEWORK 12,047
E E E
%
!
i
,, %.
%
ß %
q
II I. 12,048 TEBBENSET AL.' CHILERIDGE TECTONIC FRA_MEWO•
Table 1. Propertiesof TransformFaults and Other Offsets of theChile Ridge
,,, ,,,
Name Latitude, Off set,* Offset, Type of Offset Distance, RidgeLength*, +øN km m.y. km km
,, , ,
Chile FZ -36.3 1124. 9 FZ 83. FZ 37 -37.0 167. 9 F-Z 31. FZ 37B -37.3 95. ? FZ? 79. FZ 38 -37.9 26. .93 F-Z 52. FZ38B -38.4 103. 3.19 F-Z 66. FZ39 -38.9 87. 2.57 FZ 142. FZ 40A -40.2 -49. 1.42 FZ 96.
Valdivia system Valdivia A -41.0 122. 27. Valdivia B -41.2 113. 24. Valdivia C -41.3 114. 24. Valdivia D -41.3 176. 25. Valdivia E -41.4 79. 22. Valdivia F -41.6 37. 28.* 153.
41 ø54' offset -41.9 12. nontransform 37._• 42 ø !2' offset -42.2 15. nontransform 88?
Chiloe FZs Chiloe (north) -42.9 31. 16. Chiloe (south) -43.0 30. 1.75 84.õ 229.
43ø48 ' offset 43.8 15. nontransform !45.õ
Guafo -44.6 292. 8.46 157. Guamb!in -45.7 82. 2.70 44. Darwin -45.9 53. 1.64 72.
, ,,
Total ridgelength is 1342kin, meanridge length is 74.55km, andridge length standard deviation is 58.38 km. Left stepis positive. Distancetonext FZ south. õDistance tonext nontransform offsetsouth Distance from nontransform offset to next FZ south.
profilesat differentlatitudes is notappropriateß The magnetic ridge. The observation of faster spreading rates is instead modelincludes a jump in the anomalysequence at the Crusoe consistent with the interpretation that crust between the trough,from anomaly5A to 5D. Crusoeand Selkirk troughsformed at the Pacific-Nazcaridge. Predictedmagnetic profiles for the regioneast of the Crusoe This interpretationis also supportedby plate reconstructions trough(anomalies 5D to 6A) requirespreading rates faster than [Tebbensand Cande,this issue]. those determined for the Chile ridge (Figure 9). The magneticanomaly sequence observed west of the Friday Handschumacher[1976] interpretedthat magneticanomalies trough(anomaly 5D and younger)continues east of the Crusoe 5E, 6, and 6A observedin this region formed at the Chile trough (anomaly 5E and older) (Figure 4a). This observation
95øW 90øW 85øW 80øW
.o
line ß ...... FZ 38
.o.ß line E (aho•)
40 øS 40øS
• ß i:!:!:i:i:i:i::.:.:.:.:.:.:------' ...... _ .,.,..-- __. -----"-' • • ..,.--- • • -.--..•...--. .----"- -..-- .....>¾•...... • ..----- .------•-•---..._....-.-- __... • ...... :i:i!i:i:!!!i • • '--' -'-' _.....----- ..---- ...-. --"- _ J _...... •.:'*"...•.------:.....•. ...--.- • • • • • _..... ------.----- •7 •,-;\ti2:lt,,•q•-=---'• .:,:-:,:::.:: ..--..•*,...'?"--'------..=...,.----- '"-' • / i--a.,-• • ---.- .,--.- ___ '.'--...a•'----- ,....•;.• ::::i:::::::: .----- .,**•....:.:.:.:.:::, • • _....- • ._-..- • ..___ ...-.,- • _..------'- .,,.:.:.a•-. • ,:
95øW 90øW 85øW 80øW Figure 8. Locationmap for magneticprofiles across the Crusoetrough which are modeledin Figure9. TEBBENSET AL.: CHILE RIDGE TECTONIC FRAMEWORK 12,049
[ 3 5 5A• 5C
LINE A
LINE B
6 6A
LINE C AX v
LINE D
LINE E
I 3 5 5A ' 5E 6
LINE F
t , ,, , ', ...... , , • , , , ,, Distance(km) 0 1O0 200 300 400 500 600 700 800 900 1000 11O0
Figure 9. Observedmagnetic profiles (heavy profiles) compared with modelprofiles created with the interpretationthat there is anage discordance at the Crusoe trough, with Chile ridge spreading rates to thewest of the troughand Pacific-Nazca spreading rates to the eastof the trough.The locationsof the ridgeaxis (verticalsolid line at left) andCrusoe trough (vertical dotted line) are as charted on the tectonic maps. 12,050 TEBBENSET AL.: CHILERIDGE TECTONIC FRAMEWORK
Table 2, NeogeneSpreading Rates of the Chileridge
Time Interval Time Interval, Half Spreading Half-SpreadingRate, (PolarityChron)* Ma Rate,mm/yr mm/yr
, , , ,
0-In(o) 0-0.780 26.5 2.6 25 ln(o)-2n(y) 0.780-1.770 30.3 3.0 25 2n(y)-2An.in(y) 1.770-2.581 32.8 4.9 26 31.3 2An. ln(y)-3n. ln(y) 2.581-4.180 29.6 3.7 27 3n.ln(y)-3An.ln(y) 4.180-5.894 35.0 3.8 31
3An. 1n(y )- 3B r.2n(y) 5.894-7.341 44.3 3.1 29 3Br.2n(y)-4An(y) 7.341-8.699 45.0 2.9 24 45.6 4An(y)-4Ar.2n(y) 8.699-9.580 48.9 4.3 22
4Ar.2n(y)-5 An. In(y) 9.580-!1.935 40.6 3.6 16 37.8 5An. ln(y)-SBn. ln(y) 11.935-14.800 35.5 2.8 10
5Bn.1 n(y)-5 Cn. 1 n(y) 14.800-16.014 46.2 4.9 7 5Cn.ln(y)-5 Dn(y) 16.014-17.277 55.5 2.8 8 51.4 5Dn(y)-5 En(y) 17.277-18.281 52.4 2.6 8
5En(y)-6n(y) 18.28!- 19.048 61.9 1.6 4 6n(y)-6An.in(y) 19.048-20.518 67.4 2.5 4 61.3 6An.1 n(y)-6B n. 1 n(y) 20.518-22.588 57.1 7.6 2 6Bn.1 n(y)-6Cn. in(y) 22.588-23.353 60.4 5.2 2
, ...... *Nomenclatureis afterHarland et al. [1990],as modified by Candeand Kent [1992, 1995] with an appended(o) {or (y)} to specifythe old {or young}age limit of eachchron. is consistentwith the interpretation that a ridge (now the known Chile ridge crust is anomalySAD(y), while just south northernChile ridge) propagatedthrough and separatedthis of the Chile FZ the oldestknown Chile ridge crust is younger, once continuoussequence of anomalies. Further, the oldest anomaly 5(0) (Figures 4a and 10). These magneticdata magneticanomalies observed west of the Friday trough, in a suggestpropagation initiated at the Valdivia FZ systemat regionof transferredlithosphere, become younger from south chron 5AD(y) and ended at the Chile FZ at chron 5(0). This (anomaly5D at 40øS) to north (anomaly5C at 39øS) (Figure interpretation indicates an average propagationrate of 136 4a). These observationsindicate the ridge propagatedthrough km/Myr (452 km north in 3.33 Myr). Although basedon progressivelyyounger crust from southto north. limited data, the northern Chile ridge propagationrate is Bathymetry data. The Friday and Crusoe troughsboth remarkablysimilar to the initial propagationrate foundfor the consist of rough topography, often two bathymetrichighs easternrift of the Eastermicroplate of -135 km/Myr [Naarand surrounding a central low (Figure 4b). Where satellite- Hey, 1991]. navigated bathymetric data are available, each trough is The jumps in the ages of magneticanomalies across both located at the central bathymetric low. This interpretationis the Friday and Crusoetroughs are each5 Myr. This agejump consistentwith previous models of rifting of old oceanic is expectedto correspondwith a stepin bathymetry,based on lithosphere [Mammerickx and Sandwell, 1986], with the the relationshipthat seafloordepth increases as a functionof modificationthat we find no evidencefor a precursorswell. the squareroot of its age [Parsonsand Sclater, 1977]. Such The troughs are roughly equidistantfrom the Chile ridge topographic steps have been observed across tectonic axis, with the distancebetween troughsdecreasing from south boundaries where there are age offsets of comparable to north (Figure 10), also suggesting northward ridge magnitude,such as FZs [e.g., Fox and Gallo, 1986], failedrifts propagation. Magnetic data also indicate northward [Mammerickx and Sandwell, 1986; Batiza, 1989] and propagation:just north of the Valdivia FZ systemthe oldest propagating rifts [Hey, !977]. Limited bathymetric data
Table 3. Pacific-FarallonHalf Spreadingrates Observed on the NazcaPlate
Rate Used in Model (Figure9), Anomaly Time Interval, Line A, Line B, Line C, Line D, Line E, Line F, Line G, Where Unknown, !nterval Ma mm/yr mm/yr mm/yr mm/yr mm/yr mm/yr mm/yr mm/yr
, ,,,
5C-5D 17.31-16.03 88.84 80. 5D-5E 18.32-17.31 92.03 80. 5E-6 19.08-18.32 59.11 73.56 73.12 87.17 97.41 80. 6-6A 20.55-19.08 56.92 80.68 97.15 88.92 89.05 94.69 80. 6A-6B 22.60-20.55 46.80 52.22 62.26 97.05 80. 6B-6C 23.36-22.60 116.49 107.17 120. 6C-7 24.72-23.36 126.55 120. 7-7A 25.48-24.72 85.36 83.26 120. 7A-8 25.80-25.48 155.25 88. 8-9 27.00-25.80 87.75 79. 9-10 28.22-27.00 79.02 81. TEBBENSET AL.: CHILE RIDGE TECTONIC FR.AMEWO• 12,051 260ø 265ø 270ø 275ø -35' -35'
...... Chile FZ 500 km km 90 km
495 km ......
530 546 km .40 ø .40 ø 1091 5?_7 5_64.k.•m - • •.• -_-_-•_-•_-..... -_-_ - -_-_ _ -_-..-_ 260ø 265ø 270ø 275ø Figure10. Observedpresent-day distances between theFriday and Crusoe troughs areshown inbold type to theeast of theaxis. Just north of the shaded lines, near each trough, are the trough to ridge axis distances. For a givensegment the trough to ridgeaxis distances are different for each flank, suggesting asymmetrical spreadingorridge jumps. The distance between thetroughs decreases fromsouth to north, which we interpret asevidence for a northwardpropagation of the ridge axis through the Nazca plate. Seetext for further evidence.The propagating ridge axis evolved tobecome the northern Chile ridge. reveala stepin the bathymetryacross the Friday and Crusoe 1993; Searleet al., 1993]. The lack of mirror symmetry troughs(Figure 4b) in agreementwith that predictedby betweenthe Fridayand Crusoe troughs may be explainedby Parsonsand Sclater's[1977] age versusdepth relationship their formingas pseudofaultsof a clockwise-rotatingFriday (Figure11). The magnitudeof the Fridaytrough step is greater microplate.The lack of symmetryis the primaryevidence that thanthat predictedby the model bathymetry(Figure 11), therewas a microplateand not an instantaneousrift jump. perhapsdue to the proximityof the profileto ValdiviaFZ A. In summary,two prominentbathymetric troughs in the The shapeof the Friday and Crusoetroughs are not mirror region north of the Valdivia FZ system,on the old side of imagesof each other. This lack of mirror symmetryof anomaly5A, boundthe oldestcrust formed at the Chileridge. pseudofaultsis also observedat the Easter and Juan Fernandez These troughs are interpreted to be the outer flanks of a microplateswhere the lack of symmetryhas been explained by propagatingrift. The ridge beganpropagating shortly before microplaterotation during propagation [e.g., Schoutenet al., chron5A, cuttingthrough crust formed5 Myr earlierat the
_ Chile RidgeAxis SelkirkTrough -lOOO- FridayTrough CrusoeTrough -2000 ~ -3000- ...... , .,..,•,./.,..'/"'"*'• ß,,x,•, .... -4000-
-5000-
-6000 -1500 -1000 -500 0 500 1000 1500 2000 Distancefrom ridge axis (km) Figure 11. Observedshipboard topography across the Friday(cruise c1803) and Crusoe(cruise sot02) troughs.Model bathymetry using the formula depth = 2500+ 350*(age1/2) (age in millionyears, depth in kilometers)after Parsons and Sclater [1977]. Onboth flanks, we model a gapin agefrom 12.6 Ma (chron5A- middle)to 17.6 Ma (5D-old), accordingto the magneticinterpretation (Figure 4a). The locationof the predictedgap, manifested as a stepin the modeledbathymetry, corresponds with the observedlocation of each trough.
, 12,052 TEBBENS ET AL.: CHILE RIDGE TECTONIC FRAMEWORK
Pacific-Nazca ridge. The plate boundary reorganizationthat withinthe Valdivia FZ formedat thesame spreading rate as the includedthis propagationformed the Friday microplate. magneticanomalies formed farther south, betweenthe ValdiviaFZ systemand the ChiloeFZ, the ageof theoldest V aldivia FZ System Valdivia FZ crust is ~chron 5C. The Valdivia FZ system was previously mapped as two transforms with their associated FZs (FZ 41 and the Valdivia Chile Ridge South of the Va!divia FZ System: Nontransform Offsets and Their Traces FZ) separatedby a seismicallyactive region with no identified ridges,transforms, or FZs [e.g., KIitgord et aI., 1973; Herron Betweenthe Valdivia FZ systemand the Chile margin triple et al., 1981; Mayes et al., 1990]. This region is now junction,the Chile ridge is offset by four transformfaults interpretedto be a FZ systemcomposed of six parallel FZs, (Figures5a and5b). The locationof the associatedFZs, the with seismicactivity in the vicinity of the ridge segmentsand Chiloe,Guafo, Guamblin, and Darwin FZs, have changed only transforms,separated by distancesranging from 22 to 27 km. slightly from earlier charts. FZ 41 is renamed FZ A of the Valdivia FZ system. The One modificationis that the ChiloeFZ, from magnetic Valdivia FZ regionis still sparselysurveyed; there are almost anomaly4A (8.7 Ma) to theridge axis, appears in bathymetric no magnetic data west of the ridge axis segments. The profilesto be composedof two parallelFZs 16 km apart southernfour ridge axis segmentsare well-recordedas lows in (Figure5b). The ChiloeFZ may be composedof two FZsin the gravity field, as are the trends of the transformsand FZs oldercrust as well,as suggestedby someof thebathymetry (Plate 1). profileson the eastside of the ridge axis (Figure5b). Evidencefor the five narrow corridorsof crustcomposing Alongthe Chileridge, south of the ValdiviaFZ system,the the Valdivia FZ system include offsets in the observed magneticdata are of sufficienttrack densityto indicatethe magneticanomaly sequences (Figures 4a and 5a): EachFZ has presenceof several nontransformoffsets with off-axis traces. been chartedbetween the observedoffsets in the anomaly Basedon the assumptionthat nontransformoffsets form at the sequencesand parallel to trendsobserved in the gravity field ridgeaxis andthus would be symmetricalon bothridge flanks, (Plate 1). Bathymetric profiles crossingthe Valdivia FZ nontransformoffset traces are chartedas pairsof featuresthat system record multiple troughs, consistent with the are symmetricalabout the ridge axis. interpretationof severalFZs, but the sparsedata do little to In the vicinity of the ridge axis, the nontransformoffsets further control transformfault and FZ locations(Figures 4b are observedas lows in the gravity field. Farther from the and 5b). ridge axis, in crust older than anomaly 4A (9 Ma), Within the narrow corridors of the Valdivia FZ system, nontransform offset traces that are charted on the basis of thereare remarkablywell-recorded magnetic anomalies (Figure magneticdata are usually not observedin the gravityfield. 5a). For example,a profile betweenValdivia FZs C and D on The dependenceof FZ gravitysignal on spreadingrate, with the east ridge flank recordsa continuousanomaly sequence lower amplitudesobserved at fasterspreading rates, has been from anomaly 2A (2.6 Ma) to 5AA (13.03 Ma), an interval of shownby Shaw [1988] and may explainthe lack of a signal over 10 Myr. Surprisingly,this profile collectedon the flank observedoff-axis, in crust formed during times of faster of a shortridge segmentis indistinguishablefrom profiles spreadingrates. collectedon the flanks of the longerridge segmentbetween Prior to the 1990 aeromagneticsurvey and the availability the Valdivia and ChiloeFZs in termsof observedspreading of Geosatand ERS 1 data, the only featuresidentified which rate and quality of magneticanomalies. Ridge axis locations offset the Chile ridge were simple transform faults. The within the Valdivia FZ system were determined by recognition of nontransform offset traces and FZ systems extrapolating from observed magnetic anomalies, using reveals an increasingly more complicated segmentation anomalyto ridge axis distancesobserved south of the Valdivia patternof the Chile ridge. FZ system,and, for the southernmostfour (of five) segments, correspondwith distinctaltimetry lows (Plate 1). Pacific-Antarctic-Nazca (Farallon) Triple Boundingthe V aldivia FZ systemcrust to the eastand west Junction Trace are WNW trendinglows in the altimetrydata (Plate 1) which are interpretedas tectonicboundaries. Evidence for the eastern The Pacific-Antarctic-Farallontriple junctiontrace, on the boundaryis a changein the orientationof the magnetic Antarcticplate between-•45ø30'S and 53øS, is after Candeet anomaliesand FZs acrossthe boundarywhich suggeststhat al. [1989]. At 50øS,this boundaryis well-constrained,as it is crustto the west formed at the Chile ridge while crustto the betweenmagnetic anomaly 16 formedon the westflank of the eastformed at thePacific-Nazca ridge. Thereis alsoa jumpin Chile ridge and magneticanomaly 16 formedon the eastflank the age of oceaniccrust across each boundary. For instance, of thePacific-Antarctic ridge (PAR) (Figure6a). The boundary eastof theeastern extension of ValdiviaFZ D thereis a jump is markedby a slighttrough in thebathymetry data (Figure 6b; from anomaly5C formedat the Valdivia FZ systemof the ~50øS, 93.5øW). Chileridge to anomaly6A formedat the Pacific-Nazcaridge West of the westernmostChiloe FZ and Valdivia FZ system, (Figure 5a). it appearsthat a trough(trough B) at 101øW,43øS (Figures 1 The oldest magnetic anomalies within the Valdivia FZ and 6b), separatescrust formed at the Chile ridge and crust systemare at least as old as anomaly 5B, as recordedon the formed at the PAR. At 43øS, the oldest identifiableisochron eastridge flank between FZs C andD andFZs E andF (Figure onthe west flank of theChile ridge is anomaly6A (Figure6a). 5a). To estimatethe oldestcrustal age within the ValdiviaFZ The oldest known isochron on the east flank of the PAR at system,we examinedthe distance,measured parallel to the 43øSis anomaly5A. Betweentrough B andanomaly 6A of the Valdivia FZs, between the oldest observed anomalies within Chileridge is a roughly360-km-wide region of unidentifiable eachsegment of theValdivia FZ systemand the gravity lows magnetic anomalies and rough topography. Plate boundingthe Valdivia FZ system. Assumingthat all crust reconstructionssuggest trough B formedat chron6(o), with a TEBBENSET AL.: CHILERIDGE TECTONIC FRAMEWORK 12,053 conjugatetrough A locatedon the westflank of thePacific- lOO Antarcticridge (Figure 1). With sparsedata, there are at least threepossible tectonic explanations. Trough B maybe (1) the Chile Ridge-TrenchCollisions pseudofaultof a propagatingridge, with trough A as the conjugatepseudofault; (2) one side of a separated -• 80t x x x-x-• (nonpropagating)fault, the conjugate to whichis troughA; or • 70 CHIE RIDGE (3)the rough-smooth boundary [i.e., Mamrnerickx, 1992] of a • 60 ' rapidslowdown in spreading,perhaps due to this ridge segmentbeing a former segmentof the EPR that was "captured"by the slowerspreading PAR. If the lastcase is correct,this region may have undergonean evolution similar • 40 • to that which formed the Friday and Crusoetroughs and the 30 ' Fridaymicroplate, just to the north,where the PAC-ANT-NAZ 20 triplejunction later (chron5A) jumpednorthward from the ValdiviaFZ systemto the Chile FZ. • 100 3A4A 5B5E 6C Discussion 0 5 10 15 20 25 Chile Ridge Spreading Rates Ag, (Ma) Chile ridge spreading rates were determined using 18 •igure 12n. Averageh•f spreadingrates of th• Chil• fidg• satellite navigated magnetic profiles aligned nearly for th• past25 Myr (Tabl• 2). PAR h•f-spr•ading ratesfor the perpendicularto the ridge axis (Figure 2). The magnetic sam• rim• intervals w•r, calculatedby int•olating b•tw•n profileswere projected along synthetic tracks striking N80øE, th• pol•s of Cande et al. [1995] and using rh•s• pol•s to d•t•n• gpr•adingrates for a potion of th• PAR adjacentto perpendicularto the ridge axis. Only well-formedmagnetic th• Chil• ridge, at 40øS, 112øW. •r• is a g•n•ral d•cmas•in anomalies, not crossing FZs or "higher-order offsets" spreading rate of th• Chil• ridge, p•rhaps related to the [Mt•cdonaldet al., 1991], were used (Figure 3). Average d•cr•as• in ag• of th• subducting slab of th• Nazca plate. spreadingrates were calculatedfor 17 short(.-.2 Myr) intervals Th•r• is a n•gativ• co,clarion b•tw•n spreadingrates of th• between chron 6C (23.3 Ma) and the Present (Table 2). Chil• fidg• •d PAR. Th• times of two ma•or slowdownsin Spreadingrates were also calculatedfor five longer(~5 Myr) Chil• fidg• spreadingrat•, 14 to 10 Ma •d 6 Ma, co•spond intervalswithin which the spreading rates were found to be with Chil• ridge/trenchcollisions. relativelyconstant or, in the case of chron 3A to the Present, variedin a uniformmanner (i.e., a slow decrease).All reported ratesare half-spreadingrates. As the Chile ridge is near the equatorof the Euler pole of pole). Similarly, the predictedspreading rate variationof the rotation, at any given time there is minimal variation in chron5A Nazca-Antarcticpole [Tebbensand Cande,this issue] spreadingrate along the entire ridge axis. The predicted was 0.8 mm/yr (37.8 mm/yr at the northernmostChile ridge along-axisspreading rate variation along the Chile ridge and37 mm/yrat the Chilemargin triple junction). As the usingthe chron 2A Nazca-Antarcticpole of Tebbensand Cande modeledvariations in spreadingrate along-axisare slight,and [thisissue] is 0.6 mm/yr (30.3 'mm/yr at the northernmost far lessthan one standarddeviation in spreadingrate (c•,Table segmentof the Chile ridge, 80ø from the pole, and 30.9 mrrdyr 2), the spreadingrates were not adjustedfor varyingdistance justnorth of the Chile margintriple junction,90 ø from the from the poles of rotation.
"-' WestFlank: East Flank
6C 5E 5B 4A 3A c?g . X 3A 4A 5B 5E 6C 0 •",'' • '"'' • .... '"'•'•'' •'•'' ...... •:'' "-•''' • •'' '-' •'' -25 -20 -15 -10 -5 0 5 10 15 20 25 Age (Ma) Figure 12b. Individualage versusdistance from the ridge•is points(circles) determined from magnetic profileson theeast and west fla•s of theChile ridge (Figure 3). AverageChile ridge spreading rates (Table 2 andFigure 12a) are annotated,and presented graphically (solid lines), with crossesmarking changes in averagespreading rate. Ch•ges in slopeindicate changes in spreadingrate. Seetext for discussion. 12,054 TEBBENSET AL.: CHILERItX3E TECTONIC FRAMEWORK
Our averagedspreading rates show that the Chile ridge was ridge at anomaly 5C. Evidence for this rotation is a 10o spreadingrapidly between chrons 6C and 5E (61.3 mm/yr), changein azimuth of magneticisochrons from chron5D to was slightly slower between chrons5E and 5B (51.4 mm/yr), chron5B at 43øS (Figure6a) and a changein the strikeof the slowed to intermediate rates between chrons 5B and 4A (37.8 Chiloe and Guafo FZs (Figure 6b). The counterclockwise mm/yr), increasedspeed between chrons4A and 3A (45.6 rotationof the Chile ridge placedthe Agassiztransform under mrn/yr), and sincethen has slowedto an averageof 31 mm/yr extensionand led to the initiationof the ValdiviaFZ system (Figure 12a). The trend toward slower spreadingrates in the spreadingwithin the separatedAgassiz transform. last few million years is supportedby the observationthat the The rotation of the Chile ridge from chrons5D to 5B also current spreadingrate, as measuredacross the central anomaly, placedthe othertransforms of the Chile ridge,all right-lateral is 26.5 mm/yr (Table 2). For variationsin spreadingrate on offsets, under extension, but only the Agassiz transform an anomaly by anomaly basis and for individual ridge developed into a multiple offset transform system. The segments,see Tebbens[ 1994]. relativelength of the Agassiztransform to otherChile ridge Anotherway to analyzespreading rates is to plot age versus transforms may explain why it was the only transformto distancefrom the ridge axis [e.g., Klitgord et al., 1975]. A separateduring the ridge rotation. The Agassiztransform was plot of distance versus age for all reversals used in the over 750 km long, more than 3 times longer than any other spreading rate calculations provides an independent Chile ridge transform. The Agassiztransform connected the determinationof the times of spreadingrate changes(Figure (then)northernmost Chile ridgeto the PAC-ANT-NAZtriple 12b). In particular, changesin slope indicate times when junction [Tebbensand Cande, this issue], analogousto the spreadingrates changed. There is minimal scatter in the present-dayChile transform. observed distance from the ridge axis at each age. The observedslopes are well fit on both flanks by the calculated Friday Microplate averagespreading rates (Table 2). Spreadingrates are foundto At chron 5A(o), the Chile ridge propagatednorthward from be generally symmetric,although the east flank is observedto the Valdivia FZ towardthe ChallengerFZ throughthen 5-Myr- have slightly s!ower than average rates for the intervals from old crust (approximatelyanomaly 5D) formed at the Pacific. 4 to 9 Ma and 16 to 19 Ma. Nazca ridge. The scars of this propagation event are the Interestingly, there is a correlation between the times of Crusoeand Friday troughs,the limits of the oldestcrust formed decreasesin spreading rates and the times of collisions of alongthis sectionof the Chile ridge. As a consequenceof this Chile ridge segmentswith the Chile trench. The middle propagationevent, crustwas transferredfrom the Nazcaplate Mioceneslowdown between 15 and 9 Ma roughlycorresponds to the Antarctic plate which comprisesthe Friday microplate to the time of col!isionof a major segmentof the Chile ridge core(Figures 1 and 13). with the Chi!e trenchbetween 14 and 10 Ma [Cande and Leslie, Figure 13 presentsa schematicrepresentation of the plate 1986]. The secondslowdown, at anomaly3A, is the time of boundariesjust before and after the chron 5A propagation the next major ridge-trench collision at 6 Ma [Cande and event and demonstrates the propagation-induced Leslie, 1986]. The most recent ridge-trench collision, reorganization of plate boundaries (upper) and the plate 100,000 years ago, occurred during an interval of normal boundary origins of the crust comprising each plate. The po!arity, which was the interval of slowest observed Chile chron5A propagationevent resultedin a major plateboundary spreadingrates, 0.78 Myr to Present(Table 2). Spreadingrate reorganization,discussed below. variationswithin the past 0.78 m.y. are not constrainedby Initial propagation. The propagation event appearsto the average values presentedhere, so whether this colli'sion have initiated from the northernmost transform fault of the correlates with a distinct change in spreading rate is not newly formed Valdivia transform fault system (Figure 13, resolved. Whether there is a significant causal relationship before). Intratransformorigins of propagatingridges of betweenChile ridge spreadingrates and Chile ridge-trench microplateshave been studiedby Bird and Naar [1994],who collisionsis unclear: a correspondingincrease in spreading found the eastern rifts of the Easter and Juan Fernandez rate during intervals of transformfault subructionis clearly microplatesalso apparentlyinitiated along transformfaults. observedduring the 10 and 6 Ma transformsubruction, but not The location of crust which later would form the core duringthe transformfault subductionpreceding the 3 Ma (and (transferredlithosphere) of the Fridaymicroplate is indicated 0.1 Ma) ridge-trenchcollisions (Table 2). for reference. The distinctdecrease in Chile ridge spreadingrate at 6 Ma During propagation. The propagatingridge separated (chron3A) was notedby Cande and Kent [1992] to be nearly Nazcaplate crust formed at the EPR andtransferred lithosphere synchronouswith other circum-Pacific events, including locatedbetween the propagatingridge and the EPR fromthe distinctincreases in spreadingrates on the Pacific-Antarctic Nazcaplate to the Antarcticplate (Figure 13, after,"Friday and SoutheastIndian ridges, a clockwiserotation in spreading microplatecore"). The EPR waswest of theinitial location of direction on the SoutheastIndian ridge [e.g., Munschyet al., the propagatingridge. The southernmostEPR becamean 1992], and the initial rifting of the Lau Basin at 5.6 Ma [Leg "overlappedridge" in microplateterminology [Naar, 1992] 135 Scientific Party, 1992]. These events are synchronous andformed the western microplate boundary. The transferred with a Chile ridge-trench collision, discussedabove, and lithospherewas not rigidlyattached to any of the majorplates slightlyprecede the initial propagationof the eastridge of the (Pacific, Antarctic,or Nazca plates). The transferred JuanFernandez microplate at 5.1 Ma [Bird and Naar, 1994]. lithosphere(labeled Friday microplate core in Figures1 and 13) andmaterial accreted at theridge axes compose the Friday Formation of the Valdivia FZ System microplate.Assuming the oceanicplates behaved rigidly, the Southof the Valdivia FZ, magneticprofiles (Figure 6a) and overlappedridge slowedand possiblychanged spreading gravity field trends (Figure 6b) document a 10ø directionrelative to therest of theEPR [e.g., Naar, 1992].In counterclockwiserotation in spreadingdirection on the Chile the gravityfield, thereis a slight,but distinct,linear low at TEBBENS ET AL.: CHILE RIDGE FECTONIC FRAMEWORK 12 055
...... • :!':'-"-':?g•,• ::...... PACIFIC: :••..•'II • Challenger e. - ß ' • '•• .... :-PLATE...... •-: :•••
...... ß.• ß • ...... :::: :: x:::::: •Iv3•t•g'-;'" ::::::::::: •'•'•:::•:••:::- .... ::::•:o•h•,•::,:::::. ' ½ :::::::::: •3•I•'•••"•••:'••:x:. •- -ooze- -/ ......
BEFORE AFTER
Figure 13. Schematic plate reconstruction(left) before and (right) after the chron 5A plate boundary reorganization The top row depictswhich crustcomprises each plate. Note that a region north of the Valdivia transtorreis transferredfrom the NAZ plate to the ANT plate. The bottom row depictsat which plate boundary the crust formed. Note that the sectionof ridge southof the ChallengerFZ changesfrom southernmostEPR to northernmostPAR due to the reorganization. Note that the northernmostChile ridge extends 500 km north, from terminating at the Valdivia transform(before) to the Chile transform(after).
~103øW on chron --5A crust on the east flank of the EPR, near Mathematician microplate, which is defined as an abandoned thewestern boundary of the Friday microplate(Plate 1). This transform fault of the O'Gorrnan FZ [Mammerickx et al., low trends parallel to the nearby magnetic isochrons(e.g., 1988]. The ridge to the west of the extinct microplate, the anomaly5A) on the Antarctic plate and appearsto reflect ridge southernmostsection of the EPR prior to propagation,became axisadjustments. Again assumingthe plateswere rigid, active the northernmostsegment of the Pacific-Antarcticridge. transformfaulting occurredalong the northern and southern The extinct Friday microplate is now boundedto the north boundariesof the microplate: the Chile transformand the by a poorly located portion of a former transform(perhaps formerAgassiz' transform[Tebbens and Cande, this issue], part of the Challenger FZ, Figure 13), to the south by a respectively. portionof Valdivia FZ A, to the west by chron5A cruston the Propagation complete. Propagationceased when the Pacific-Antarcticridge east flank, and to the east by chron5A propagatingridge reachedthe ChallengerFZ. The Challenger crust on the Chile ridge west flank. FZ formed at an offset of the Pacific-Farallon EPR. The new The chron 5A plate boundary reorganizationresulted in the ridgeaxis apparentlyrapidly adjustedto accommodatethe most recent northward migration in the location of the Pacific- sameplate motions as the Chile ridge southof the Valdivia FZ Antarctic-Nazcatriple junction, from the Valdivia FZ system systemand is now the northern Chile ridge (between the to the Chile FZ (Figure 13). ValdiviaFZ systemand the Chile FZ). The ChallengerFZ Implications to General Models of Microplate becamea reactivatedtransform (the present-dayChile Evolution transform)connecting the newnorthern Chile ridge to the new locationof the Pacific-Antarctic-Nazcatriple junction. The The evolutionof the Friday microplatediffers from that of extinctFriday microplatebecame incorporatedinto the other known extinct microplates. Based on microplates Antarcticplate. The extinctFriday microplate includes the worldwide, Mammerickx et al. [1988] described the evolution transferredNazca plate crust and the crustwhich formed at the of a microplate to involve (1) initial rifting, offset from microplateboundaries (Figure 13, lowerrow). Thesouthern anotherridge, (2) an intervalwhen both ridgesare active (.3) boundaryof the microplate, a portion of ValdiviaFZ A, which gradual failure of one ridge, and (4) resumptionof full hadbeen an active transform fault during rift propagation,was spreading at the dominant ridge and preservationof a abandoned.This is similarto the southernboundary of the microplate(initially) near this ridge. While more detailed 12,056 TEBBENS ET AL.: CH•E RIDGE TECTONIC FRAMEWORK
•mo_
•ro.
o TEBBENS ET AL.: CH/LE RIDGE TECTONIC FRAMEWORK 12,057 modelsof microplateevolution have been developed The ValdiviaFZ systemis composedof six closelyspaced [Schoutenetal., 1993,and references therein], these basic (22-27km offset)FZs separatingshort ridge segments which stepshave remained. Atthe Friday microplate wesee evidence have beencontinuously active for the last 15 Myr. The forthe first two steps described above, but neither ridge ever ValdiviaFZ systemevolved from a singlelong left-lateral failed.Instead, after microplate extinction, both ridge axes of offsettransform connecting the Chile ridge to the Pacific- themicroplate continued spreading and were "captured" as Antarctic-Nazcatriple junction at the time of a 10ø portionsof majorplate boundaries. This significant counterclockwise rotation in spreadingat chron5C. differenceresults in a processwe call "stepwise"triple A plateboundary reorganization at chron5A involvedthe junctionmigration. northwardpropagation of a ridgeaxis from the ValdiviaFZ systemto the Chile FZ; two pseudofaults,the Crusoeand Effectsof Ridge Subruction Fridaytroughs, were formed as a result.This event transferred The southernend of the Chile ridge is actively being lithospherefrom the Nazcaplate to theAntarctic plate to form subducteddown the Chile trench. Herron et al. [1981]were the the Friday microplate. The Friday microplatediffers from previouslystudied extinct microplates in thatboth ridge axes firstto make the surprisingobservation that the actively of themicroplate continued spreading; they were "captured" by subductingridge axis has very little effecton the seafloor majorridge axes. The eastand westridge axes of the extinct spreadingprocess near the trench axis. This observation has beenfurther supported by the observationson multichannel microplateare now segmentsof the Chile and Pacific- Antarcticridges, respectively. The chron5A reorganization, seismic,Sea Beam and GLORIA surveys[Cande et al., 1988; which includedrift propagation,microplate formation, and Candeet al., 1990; Tebbenset al., 1990]. The new tectonic microplateextinction with no failed ridge,resulted in a 500 mapsin thispaper also reveal no evidencefor Nazcaplate km northwardstepwise migration in the location of the breakupor Chile ridge rotationassociated with ridge Pacific-Antarctic-Nazcatriple junction. subduction.This differs from the subductionof the Pacific- Farallonridge off Northand Central Americas, where Menard [1978]found evidence for "pivotingsubduction," where small Acknowledgments.We thankthe dedicatedpersonnel of theNaval ResearchLaboratory Flight Detachment and OperationalServices. The portionsseparate from the mainoceanic plate and rotate. workwas initiated as part of S.F.T.'sdissertation when S.F.T. and S.C.C. Iterron et al. [1981] suggestedthe lack of "pivoting were both at Lamont-Doherty Earth Observatory of Columbia subruction"at the Chile ridgeis probablydue to the geometry University. G. Westbrookprovided magneticand bathymetricdata of plateinteraction: (1) the anglebetween the spreadingfrom RRS CharlesDarwin cruise36-88. D. Handschumacherprovided centeron the Chile ridge and the Peru-Chiletrench is only ProjectMagnet aeromagnetic data collectedby the Office of Naval Researchprogram element ONR-61153N underthe directionof H.C. about15 ø andperhaps too smallto requirepivoting; and (2) EppertJr. W. Haxbyprovided Geosat data and assistance prior to the theoffsets along the Chileridge displace the ridge crest away 1990 aeromagneticsurvey. D. Sandwellprovided access to computer fromthe trenchin the oppositesense to the Farallonplate programsand early accessto the Sandwell and Smith [1992] gravity boundaryin the northeastPacific. In addition,"pivoting grid. This paperbenefited from discussionswith and/orreviews of T. Atwater, S. Carbotte, W. Haxby, K. Klitgord, P. Lonsdale, M. subruction"in the northeast Pacific is associatedwith the Kuykendal!,D. Naar, M. Tivey, and, especially,H. Schoutenwho subructionof relatively young, uniform age Farallon crust. In explained his microplate model [Schoutenet al., 1993] and its contrast,while near zero-ageNazca crustis subductingat the implicationsfor the Fridaymicroplate. B. Batchelderdrafted Figures 4 collidingsection of the Chileridge, older Nazca plate crust (up and 5. This researchwas supportedby NationalScience Foundation to~35 Myr old)is subductingfarther north [Cande and Haxby, grantOCE-89-18612. 1991]. Platesubduction appears to be the dominantdriving forceof platemotions with older, colder,denser subducting References crustproviding greater driving force [Forsythand Uyeda, Anderson-Fontana,S., J. F. Engeln, P. Lundgen, R. L. Larson, and S. 1975;Lithgow-Bertelloni and Richards, 1995]. Subructionof Stein, Tectonicsand evolution of the Juan Fernandezmicroplate at theolder sections of Nazcaand Antarctic plates beneath South the Pacific-Nazca-Antarctictriple junction, J. Geophys.Res., 91, Americamay dominateChile ridge plate motions,including 2005-2018, 1986. the subduetingridge segment,and may help explain why Anderson-Fontana,S., J. F. Engeln, P. Lundgren,R. L. Larson,and S. Stein, Tectonics of the Nazca-Antarctic plate boundary, Earth "pivotingsubruction" is not observed. Planet. Sci. Lett., 86, 46-56, 1987. Batiza, R., Failed rifts, in The Geologyof North America,Vol. N, The Conclusions EasternPacific Oceanand Hawaii, vol. N, editedby E.L. Wintereret al., pp. 177-186,Geol. Soc. of Am., Boulder,Colo., 1989. Datafrom a recentaeromagnetic survey have been combined Bird, R. T., and D. F. Naar, Intra•transformorigins of mid-ocean microplates,Geology, 22, 987-990, 1994. withsatellite altimetry data to constructa detailedtectonic Cande, S.C., and W. F. Haxby, Eocene propagatingrifts in the mapof the Chile ridge. The tectonicinterpretation of this southwestPacific andtheir conjugatefeatures on the Nazcaplate, J. workis synthesizedin Figure 14. The Chile ridge extends Geophys.Res., 96, 19609-19622,1991. fromthe Juan Fernandez microplate to theChile Margin triple Cande, S.C., and D. V. Kent, A new geomagneticpolarity time scale for the Late Cretaceousand Cenozoic,J. Geophys.Res., 97, 13917- junctionand includes a total of 1380 km of ridgeaxis offset by 13951, 1992. 18 transformfaults, includingtwo complexfault systems Cande, S.C., and D. V. Kent, Revised calibration of the geomagnetic (Table 1). Three nontransformoffsets (and their off-axis polarity timescale for the Late Cretaceousand Cenozoic,J. traces)have also been identified. Geophys.Res., 100, 6093-6095, 1995. Cande,S.C., and R. B, Leslie,Late Cenozoictectonics of the southern Chfieridge spreading rates were determined for 17 intervals Chiletrench, J. Geophys.Res., 91, 471-496, 1986. between23 Ma andthe Present (Table 2). Timesof decreasesCande, S.C., R. B. Leslie, J. C. Parra, and M. Hobart, Interaction inspreading rate were found to correlatewith times of major between the Chile ridge and Chile trench: Geophysicaland Chileridge-trench collisions. geothermalevidence, J. Geophys.Res., 92, 495-520, 1987a. 12,058 TEBBENS ET AL.: CHILE RIDGE TECTONIC FRAMEWO•
Cande, S.C., P. R. Vogt, and P. J. Fox, Detailed investigationof a thesouthern (Pacific-Antarctic) EastPacific Rise, J.Geophys. Res., flowline corridorin the SouthAtlantic, II, A surveyof 50 to 70 m.y. 99, 4683-4702, 1994. old crust on the west flank (abstract), Eos Trans. AGU, 68, 1491, Macdonald,K.C., D. S. Scheirer, andS. M. Carbotte, Mid-ocean ridges: 1987b. Discontinuities,segments andgiant cracks, Science, 253, 986-994, Cande, S.C., S. D. Lewis, N. Bangs, S. F. Tebbens,and R. Forman, 1991. interactionof the Chile ridge and Chile trench: Preliminaryresults Mammerickx,J., TheFoundation seam,•unts: Tectonic setting of a from a Sea Beam and MCS survey,Eos Trans.AGU, 69, 1407, 1988. newlydiscovered seamount chain in theSouth Pacific, Earth Planet. Cande, S.C., J. L. LaBrecque, R. L. Larson, W. (2. Pitman II!, X. Sci. Lett., 113, 293-306, 1992. Golovchenko,and W. F. Haxby, Magneticlineations of the world's Mammerickx,J.,and D. Sandwell,Rifting of oldoceanic lithosphere, j. ocean basins, AAPG Map Set., Am. Assoc. of Pet. Geol., Tulsa, Geophys.Res., 91, 1975-1988,1986. Okla., 1989. Mammeriekx,J.,and S. M. Smith,GEBCO map of the Pacific Ocean, Cande, S.C., S. D. Lewis, N. Bangs, S. F. Tebbens, and G. K. sheet5-11, Can.Hydrogr. Serv., Ottawa, Ont., 1980. Westbrook,The Chile ridge/Chiletrench collision zone: A rift valley Mammerickx,J., D. F. Naar, andR. L. Tyce,The Mathematician disappearsbeneath a continentalmargin, Eos TransAGU, 71, 1565, paleopiate,J. Geophys.Res., 93, 3025-3040, 1988. 1990. Mayes,C. L., L. A. Lawver,and D. T. Sandwell,Tectonic history and Cande,S.C., C. A. Raymond,J. Stockand, W. F. Haxby, Geophysicsof newisochron chart of the SouthPacific, J. Geophys.Res., 95, 8543- the Pitmanfracture zone and Pacific-Antarcticplate motionsduring 8567, 1990. the Cenozoic, 270, 947-953, 1995. Menard,H. W., Fragmentationof the Farallon plate by pivoting Ewing, M., and B.C. Heezen, Some probiemsof Antarcticsubmarine subduction,J. Geol.,86, 99- ! 10, ! 978. geology,in Antarcticin theInternational Geophysical Year, Geophys. Menard,H. W., T. E. Chase,and S. M. Smith,Galapagos Rise in the Monogr. Set., vol. 1, edited by A. Crary et al., pp. 75-81, AGU, southeasternPacific, Deep SeaRes., 11, 233, 1964. Washington,D.C., !956. Molnar,P., T. Atwater,J. Mammerickx,and S. Smith,Magnetic Forsyth,D.., and S. Uyeda, On the relative importanceof the driving anomalies,bathymetry and the tectonic evolution of theSouth Pacific forcesof plate motion, Geophys.J. R. Astron. Soc., 43, 163-200, sincethe Late Cretaceous, Geophys. J.R. Astron. Soc., 40, 383-420, 1975. 1975. Fox, P. J., and D. G. Gallo, The geologyof North Atlantictransform Morgan,W., P. R. Vogt, and D. F. Falls,Magnetic anomalies and sea plate boundariesand their aseismicextensions, in The Geologyof floorspreading on the Chile rise,Nature, 222, !37, 1969. NorthAmerica, vol. M, The WesternNorth Atlantic Region, edited by M•iller, R. D., D. T. Sandwell,B. E. Tucholke,J. G. Sclater,and P. R. P.R. Vogt, pp. 157-172,Geol. Soc.of Am., Boulder,Colo., 1986. Shaw,Depth to basementand geoid expression of the KaneFracture Fox, P. J., and D. G. Gallo, Transforms of the eastern central Pacific, in Zone: A comparison,Mar. Geophys.Res., 13, 105-129,1990. The Geologyof North America, vol. N, The EasternPacific Ocean Munschy,M., J. Dyment,M. O. Boulanger,D. Boulanger,J. D. Tissot, and Hawaii, editedby E.L. Wintereret al., pp. 111-124,Geol. Soc. R. Schlichand M. F. Coffin,Breakup and seafloor spreading of Am., Boulder, Colo., 1989. betweenthe Kerguelen Plateau-Labuan Basin and the Broken Ridge- Handschumacher,D. W., Post-Eoceneplate tectonics of the eastern DiamantinaZone, Proc. Ocean Drill. ProgramSci. Results, 120, 931- Pacific, in The Geophysicsof the Pacific Ocean Basin and Its 944, 1992. Margin,Geophys. Monogr. Set., vol. 19,edited by G.H. Suttonet al., Naar,D. F., Microplates,in Encyclopediaof EarthSystem Science, vol. pp. 177-202,AGU, Washington,D.C., 1976. 3, pp.231-237, Academic, San Diego, Calif., 1992. Harland,W.B., R. Armstrong,A. Cox,L. Craig,A. Smith,and D. Smith, Naar,D. F., andR. N. Hey,Tectonic evolution of theEaster microplate, A geologicTime Scale 1989, 263 pp., CambridgeUniversity Press, J. Geophys.Res., 96, 7961-7993, 1991. New York, 1990. Parmentier,E. M., andW. F. Haxby,Thermal stresses in the oceanic Herron,E. M., Crustalplates and sea-floor spreading in thesoutheastern lithosphere:Evidence from geoid anomalies at fracturezones, J. Pacific,in AntarcticOceanology I, Antarct.Res. Set., vol. 15, edited Geophys.Res., 91, 7193-7204, 1986. by J.L.Reid, pp. 229-237, AGU, Washington,D.C., 1971. Parsons,B., andJ. G. Sclater,An analysisof thevariation of oceanfloor Herron,E. M., Sea-floorspreading and the Cenozoichistory of theeast- bathymetryand heat flow withage, J. Geophys.Res., 82, 803-827, 1977. centralPacific, Geol. Soc.Am. Bull., 83, 1671-1692, 1972. Herron,E. M., andD. E. Hayes,A geophysicalstudy of the Chile Ridge, Sandwell, D. T., Thermomechanicalevolution of oceanic fracture Earth Planet. Sci. Lett., 6, 77-83, 1969. zones,J. Geophys.Res., 89, 11401-11413,1984. Herron,E. M., S.C. Cande,and B. R. Hall,An activespreading center Sandwell,D., Globalmarine gravity grid and posterdeveloped collideswith a subductionzone: A geophysicalsurvey of theChile (abstract),Eos Trans.AGU, 74, 35, 1993. margin triple junction, in Nazca Plate: Crustal Formation and Sandwell,D., andW. H. F. Smith,Global marine gravity from ERS-1, AndeanConvergence, edited by L.D. Kulmet al., Mem. Geol.Soc. Geosatand Seasat reveals new tectonic fabric (abstract), Eos Trans. of Am., 154, 683-701, 1981. AGU,73 (43),Fall Meet. Suppl., 133, 1992. Hey,R., A new classof "pseudofaults"and their beating on plate Schouten,H., K. D. Klitgord,and D. G. Gallo,Edge-driven microplate tectonics:A propagatingrift model,Earth Planet. Sci. Lett., 37, 321- kinematics,J. Geophys.Res., 98, 6689-6701,1993. 325, 1977. Searle,R. C., R. T. Bird,R. I. Rusby,and D. F. Naar,The development InternationalAssociation of Geomagnetismand Aeronomy,Data of twooceanic microplates: Easter and Juan Fernandez microplates, ImagingWorking Group (IAGA, DIWG), International geomagnetic EastPacific Rise, J. Geol.Society London, 150, 965-976, 1993. referencefield revision 1987, !AGA News, 26, 87-92,1987. Shaw, P., An investigationof small-offsetfracture zone geoid waveforms,Geophys. Res. Lett., 15, 192-195,1988. Klitgord,K. D., S.P. Heustis, J.D. Mudieand R. L. Parker,An analysis Small,C., andD. T. Sandwell,An abruptchange in ridgeaxis gravity of near-bottommagnetic anomalies: Sea-floor spreading and the withspreading rate, J. Geophys.Res., 94, 17383-17392,1989. magnetizedlayer, Geophys. J. R. Astron.Soc., 43, 387-424,1975. Smith,W. H. F., and D. T. Sandwell,Marine gravity field from Klitgord,K. D., J'.D. Mudie,P. A. Larsonand J. A. Grow,Fast sea floor aleclassifiedGeosat and ERS- 1 altimetry(abstract), Eos Trans. AGU, spreadingon theChile ridge, Earth Planet Sci. Lett., 20, 93-99,1973. 76(46), Fall Meet. Suppl., F156, 1995. LaBreeque,J. L., SouthAtlantic Oceanand adjacentAntarctic Tebbens, S. F., Tectonic evolution of the southeastPacific from continentalmargin, Reg. Atlas Ser., Atlas 13, 21 sheets,Ocean Drill. Oligoceneto Present,Ph.D. thesis, Columbia Univ., New York, 1994. Program,Mar. Sci. Int., WoodsHole, Mass., 1986. Tebbens,S. F., andS.C. Cande,South Pacific tectonic evolution from Larson,R. L., R. C. Searle,M. C. Kleinrock,H. Schouten,R. T. Bird,D. earlyOligocene to Present:Pacific-Antarctic-Nazca triple junction F. Naar,R. i. Rusby,E. E. Hooftand H. Lasthiotakis,Roller-bearing migrations,J. Geophys.Res., this issue. tectonicevolution of the JuanFernandez microplate, Nature, 356, 571-576, 1992. Tebbens,S. F., G. K. Westbrook,S.C. Cande,N. Bangs,and S. D. Lewis, Preliminaryinterpretation of combinedGLORIA and Leg135 Scientific Party, Evolution of backarcbasins: ODP Leg 135, SeaBeamimagery in thevicinity of theChile margin triple junction LauBasin, Eos, Trans. AGU, 73, 241,243,246,1992. (abstract),Eos Trans. AGU, 70(43),1353, 1990. Lithgow-Bertelloni,C., and M. A. Richards,Cenozoic plate driving Wessel,P., andW. F. Haxby,Thermal stresses, differential subsidence, forces,Geophys. Res. Lett., 22, !317-1320,1995. andflexure at oceanicfracture zones, J. Geophys.Res., 95, 375- Lonsdale,P., Geomorphologyand structural segmentation of the crest of 391, 1990. TEBBENSETAL.: CHILE RIDGE TEL'TONIC FRAIVIEWO• 12,059
Wessel,P., andW. H. F. Smith,Free software helps map and display S.F. Tebbens,Department of MaxineSciences, University of South data,EOS Trans. AGU, 72, 441, 1991. Florida, St. PetersburgCampus, 140 Seventh Avenue South, St. Petersburg,FL 33701-5016.(e-mail: [email protected]) S.C.Can'"-"-•-'"de, ScrippsInstitution ofOceanography, MS0215, La Jolla, H. Vergara,Servicio Hidrografico y Oceanograficodo la Armada, CA 92093. (e-mail: [email protected]) Valparaiso,Chile. L. Kovacs,Naval ResearchLaboratory, Washington, DC 20375- 5350.(email: skip @hpSc.nrl.navy.mil) J.L. LaBrecque,Lamont-Doherty Earth Observatoryof Columbia University,PaliSades, NY 10964. (ReceivedDecember 5, 1994;revised June 24, 1996; J.C.Parra• GEODATOS, Santiago,Chile. acceptedAugust 20, 1996.)