Transcript
Geology doi: 10.1130/0091-7613(1999)027<1103:IMOTGP>2.3.CO;2 1999;27;1103-1106 Geology Dennis Geist, William White, Terry Naumann and Robert Reynolds transportIllegitimate magmas of the Galápagos: Insights into mantle mixing and magma Email alerting services articles cite this article to receive free e-mail alerts when newwww.gsapubs.org/cgi/alertsclick Subscribe to subscribe to Geologywww.gsapubs.org/subscriptions/ click Permission request to contact GSAhttp://www.geosociety.org/pubs/copyrt.htm#gsaclick official positions of the Society.citizenship, gender, religion, or political viewpoint. Opinions presented in this publication do not reflectpresentation of diverse opinions and positions by scientists worldwide, regardless of their race, includes a reference to the article's full citation. GSA provides this and other forums for thethe abstracts only of their articles on their own or their organization's Web site providing the posting to further education and science. This file may not be posted to any Web site, but authors may postworks and to make unlimited copies of items in GSA's journals for noncommercial use in classrooms requests to GSA, to use a single figure, a single table, and/or a brief paragraph of text in subsequenttheir employment. Individual scientists are hereby granted permission, without fees or further Copyright not claimed on content prepared wholly by U.S. government employees within scope of Notes Geological Society of America on July 17, 2011geology.gsapubs.orgDownloaded from INTRODUCTION In most volcanic provinces,the compositionsof lavas from individual volcanic centers are dis-tinctive; i.e.,the compositional variation of lavasfrom single volcanoes is far less than the compo-sitional variation of the entire volcanic province.This simple observation has been used to advan-tage in deciphering the transport of magma in dif-ferent tectonic environments. An example fromthe mid-ocean ridge system (Langmuir et al.,1986) shows that tectonic segmentation of theEast Pacific Rise is accompanied by petrologicsegmentation,whereby geochemical fingerprint-ing shows that magmas do not intrude and mixacross discontinuities in the ridge axis. Likewise,a study of arc magmatism in the area around No-varupta (Hildreth,1987) showed that magmamoved laterally between three separate volcanoes(Novarupta,Katmai,and Trident) immediatelybefore the A.D. 1912 eruption. The best-studiedexample of this sort of magmatic fingerprintingof ocean-island volcanoes is the big island of Hawaii,where lavas from each of the shieldshave unique and identifiable major and trace ele-ment and isotopic compositions (e.g.,Frey andRhodes,1993). On this basis,Rhodes etal.(1989) reasoned that between 2100 and 400 yrB.P. magmas from Mauna Loa penetrated Ki-lauea’s magmatic plumbing system.In the western Galápagos Islands,lavas fromeach of seven shield volcanoes are also composi-tionally distinct from their neighbors (McBirneyand Williams,1969; White et al.,1993),with sev-eral notable exceptions that are the subject of thispaper. We have found that four out of the fiveshields that have been studied in detail haveerupted a single flow of lava that is unlike everyother lava sampled from that volcano. The sam-ples were collected and analyzed during separatefield campaigns,so we are confident they werenot simply switched. We coin these illegitimatemagmas and believe that they provide evidencefor unusual circumstances of mantle mixing andmagma transport within the lithosphere. GEOLOGIC BACKGROUND The Galápagos are a hotspot-related group of islands that are adjacent to the Galápagosspreading center. Owing to the proximity of the Geology; December 1999; v. 27; no. 12; p. 1103–1106; 4 figures.1103 Illegitimate magmas of the Galápagos: Insights into mantle mixingand magma transport Dennis Geist* Department of Geology and Geological Engineering, University of Idaho, Moscow, Idaho 83844, USA William White Department of Geological Sciences, Cornell University, Ithaca, New York 14850, USA Terry Naumann Geology Department, University of Alaska, Anchorage, Alaska 99508, USA Robert Reynolds Department of Science, Central Oregon Community College, Bend, Oregon 97701, USA ABSTACTRoughly 1–2% of the flows erupted from flank vents of the western Galápagos shield vol-canoes have anomalous compositions. We call these illegitimate magmas because of their un-certain parentage. Because some illegitimate magmas are compositionally indistinguishablefrom lavas of an adjacent volcano and erupt from the flank facing the adjacent volcano,suchmagmas apparently result from lateral intrusion of magma from the adjacent volcano. Other il-legitimate magmas come from parts of the Galápagos plume that have incompletely mixed orresult from unusually advanced melting of part of the mantle. *E-mail:
[email protected]. Figure 1.Detailed map ofIsabela and Fernandina Is-lands shows distributionof vents (taken directlyfrom Chadwick and How-ard,1991) and locations offlows of illegitimate mag-mas (large filled circles).Base of each arrow is lo-cated on volcano that hascomposition closest tothat of illegitimate magma. on July 17, 2011geology.gsapubs.orgDownloaded from spreading center and the young,thin lithosphere,the islands are not regularly distributed in the di-rection of plate motion (Fig.1). Nevertheless,the islands are generally older to the east,andmost historical activity has occurred in the west(White et al.,1993). This study is restricted toyoung lavas erupted from the western volcanoesfor which we have detailed petrologic and geo-logic data. Five of the western volcanoes havenow been mapped and sampled in detail. Otherthan Sierra Negra,each has erupted a single ille-gitimate lava,one whose composition resemblesthat of another volcano.The compositional variation of Galápagoslavas has been attributed to three independentprocesses:plume-asthenosphere mixing,partialmelting,and crystal fractionation at differentdepths. First,isotopic compositions of Sr,Nd,and Pb indicate that the Galápagos mantle plumemixes dynamically with the shallow astheno-sphere (Geist et al.,1988; White et al.,1993;Kurz and Geist,1999). Second,concentrations of incompatible trace elements are controlledmostly by both plume-asthenosphere mixing andvarying extents of partial melting. Third,all sam-pled lavas from the western Galápagos haveevolved by cooling and crystallization in the lith-osphere,but at systematically different depths(Geist et al.,1998). Volcanoes with deep calderas(Cerro Azul,Wolf,Ecuador,and Fernandina)have high-level magma reservoirs,and the laststage of fractionation takes place at shallow 1104GEOLOGY,December 1999 Figure 2.Each illegitimatelava can be identified byratios of incompatible traceelements.Shaded fields indi-cate normal lavas from eachvolcano,and large open cir-cle marks illegitimate lava.Arrows connect illegitimatelavas (base of arrow) to fieldof normal lavas from thatvolcano (head of arrow).Figure 3.Isotopic ratios ofillegitimate lavas.Arrowsconnect illegitimate lavas(tail of arrow) to normallavas of their host volcano(head of arrow).Alsoshown are data for vol-cano that most closely re-semble illegitimate lavas. on July 17, 2011geology.gsapubs.orgDownloaded from depths. Magmas erupted from the volcanoes withshallower calderas (Sierra Negra,Alcedo,andDarwin) have evolved in the deeper crust andmantle lithosphere. EXAMPLES OF ILLEGITIMATEMAGMAVolcán Ecuador Volcán Ecuador is made up of two parts,anolder shield that has been dissected by sector col-lapse and a young rift zone to the east of the vol-cano (Rowland et al.,1994; Geist et al.,1997;White and Geist,1998). The oldest lavas areabout 90ka,and although no eruption has beenwitnessed from this remote volcano,the youngestflows are no more than a couple centuries old(Kurz and Geist,1999). Lavas from VolcánEcuador have a wide range of MgO concentra-tions,from >13% to <5%. Ratios of incompatibletrace elements are quite limited,however. For ex-ample,Nb/Zr ratios average 0.128 ± 0.010(1standard deviation; n = 149). A single sample,E95-13,is highly anomalous,with Nb/Zr of 0.074 (Fig.2A). This Nb/Zr ratio is almost pre-cisely that of Volcán Wolf’s lavas (0.079 ±0.007),and E95-13 has Sr and Nd isotopic ratiosthat are indistinguishable from those of Wolf andunlike any other Volcán Ecuador lava (Fig. 3B).E95-13 erupted from an isolated cone on thenorth flank of the east rift. The cone is one of theyoungest vents of the east rift,not a kipuka. Thiscone is surrounded by lavas of Volcán Ecuadorand is 2km from the nearest Wolf lava and 3kmfrom the nearest satellite vent of Wolf. Other thanthis one flow,all of the other lavas from the eastrift are indistinguishable from lavas erupted fromthe main shield of Volcán Ecuador. Volcán Wolf Volcán Wolf is the highest of the Galápagosvolcanoes and has the second-deepest caldera.The oldest exposed lava in the caldera walls is173 ± 20 ka ( 40 Ar/ 39 Ar),and the volcano hasbeen historically active. In contrast to neighbor-ing Ecuador,Volcán Wolf’s lavas are the most de-pleted,in Sr,Nd,and Pb isotopes and incompati-ble trace elements of all the western shields(White et al.,1993),and they have a low Nb/Zrratio,averaging 0.079 ± 0.007 (n=89; Fig.2A).Volcán Wolf’s lavas have a restricted range of MgO concentrations (Fig.2A).Sample W95-61 has a highly anomalous com-position,with Nb/Zr = 0.123,essentially identi-cal to the average value from Volcán Ecuador(Fig.2A). This specimen is also isotopically un-like all other Wolf lavas (Fig.3A). It plots closeto the field for Ecuador lavas,although justslightly displaced toward the Alcedo field. W95-61 was collected from a vent low on the south-west flank of the volcano,where it is intercalatedwith flows that are compositionally indistin-guishable from 54 other sampled lavas fromWolf’s lower flanks. Volcán Alcedo Volcán Alcedo is unique in the western Galá-pagos for the rhyolites that it has erupted (Geistet al.,1994,1995),but here we are concernedonly with the basalts that make up >99% of thevolume of the volcano. Alcedo has had only onehistorical eruption,and lavas as old as 150 ±50ka are exposed in the walls of the shallowcaldera,suggesting that Alcedo’s activity iswaning. Sample E-132 is from low on the southflank of Alcedo; although its Sr and Nd isotopicratios are unlike any other Alcedo lava,they aresimilar to those of Cerro Azul (Fig.3C). Incom-patible element ratios of E-132 are intermediatebetween those of other lavas of both Cerro Azuland Alcedo (Fig.2B). Cerro Azul Cerro Azul is one of the youngest and most ac-tive of the western shields,having erupted as re-cently as 1998. The oldest lavas at the base of thecaldera walls are 82 ± 17ka (Naumann andGeist,1999). Cerro Azul’s lavas are among themost enriched in the western part of the archipel-ago,indicating that they are derived from rela-tively pristine plume material (White et al.,1993). Sample CA36 has a Nb/Zr ratio of 0.116,which is much lower than the average of 0.158 ±0.013 from the other 80 analyzed samples(Fig.2B). Likewise,the La/Sm ratio of this sam-ple is 1.92,much lower than the volcano’s aver-age (2.92 ± 0.34). The sample is not isotopicallydifferent from the other Cerro Azul magmas(Fig.3D),although it is at the more depleted endof the Cerro Azul array. TYPE 1 ILLEGITIMACY:LATERAL INTRUSION The most obvious mechanism for the eruptionof an illegitimate magma is the intrusion of magma from one volcano into an adjacent vol-cano by lateral dike propagation (Fig.4). Thisscenario is likely if (1) the illegitimate lava iscompositionally indistinguishable from the lavasof an adjacent volcano and (2) the illegitimatelava erupted from the flank nearest the volcanowith its geochemical fingerprint.Two sites of lateral intrusion have been identi-fied by these criteria:intrusion of Wolf magmainto the east rift of Volcán Ecuador and intrusionof Ecuador magma into the southwest flank of Volcán Wolf. Volcán Ecuador’s illegitimatemagma has a Wolf isotopic signature (Fig.3B),and also has trace element ratios that are indistin-guishable from Wolf magmas (Fig.2A). Thesedata indicate that there was no mixing betweenthis magma and any Ecuador magma that mightbe stored beneath the east rift. Wolf’s illegitimate magma is compositionallysimilar to lavas from Volcán Ecuador. As notedby Rowland etal. (1994),the western end of theEcuador’s east rift strikes about 75°,but within2km of its contact with Wolf,it curves to about95° (Fig.1; the easternmost lavas mapped asEcuador by Rowland etal. [1994,Fig.5] were in-correctly assigned; they are Wolf lavas). Dikesthat feed the east rift would have to curve aroundto strike 180° and intrude another 2km to feedthe vent that erupted the illegitimate magma,which we consider unlikely. In addition,this lavais not a perfect match for Volcán Ecuador(Fig.3A),because its isotopic composition is in-termediate between Ecuador and Alcedo. Thus,the illegitimate magma at Wolf may well be of typeII. Wolf’s illegitimate lava is more magne-sian than all but two of its other lavas. These havea restricted range of MgO from 4.6% to 7.0%,which may be attributed to steady-state composi-tional buffering in the subcaldera magma cham-ber (e.g.,O’Hara,1977). A more magnesiancomposition might arise if the magma did notpass through the central magma chamber beforeintruding laterally. Thus,if this batch of magmaintruded from the Ecuador system,it did not en-counter Wolf’s subcaldera magma chamber.In an important study of illegitimacy in Hawaii,Rhodes etal. (1989) found that Kilauea’s magmastook on a Mauna Loa–like trace element signatureduring the interval 2000–400yrB.P. The anom-alous compositions are attributed to the lateral GEOLOGY,December 19991105 Figure 4.Illustration ofthree types of illegitimacy.(1) Magmas intrude laterallyfrom one volcano’s magmasupply system into adja-cent volcano.(2) Mantle do-mains have not completelymixed,preserving hetero-geneity.(3) Mantle melts togreater extents than is typi-cal,imparting odd trace ele-ment compositions on par-tial melts.Note that in allthree types,magma mustnot be incorporated intomain magma supply,whereit would lose its composi-tional uniqueness. on July 17, 2011geology.gsapubs.orgDownloaded from penetration of magma from Mauna Loa into Ki-lauea’s “high-level”plumbing system. The Galá-pagos cases of type 1 illegitimacy are different,however,because little mixing has occurred;hence penetration into the main magma bodies atthese volcanoes is rare. Instead,intrusion is intothe flanks of the volcanoes and does not involvemixing with the magma of the host volcano. TYPE 2 ILLEGITIMACY:INCOMPLETEMIXING The isotopic variation and much of the varia-tion in the ratios of incompatible trace elements of Galápagos lavas are due to a combination of het-erogeneity within the plume and mixing betweenthe mantle plume and the shallow asthenosphere(Geist et al.,1988; White et al.,1993; Harpp,1995). Any local variations in the ratio of plumeto asthenosphere would lead to local isotopic het-erogeneity in melts produced from that region. Inparticular,we would expect that melts derivedfrom volumes that have been less polluted by en-trained asthenosphere should have compositionsclose to that of the volcano immediately up-stream. This is exactly the characteristic of the il-legitimate lava from Wolf volcano,which is com-positionally similar to (but not identical to)Ecuador and may not be due to lateral intrusion.The existence of illegitimate magmas such asW95-61 thus suggests that the remarkable homo-geneity of lavas erupted from each of the westernvolcanoes reflects very efficient homogenizationwithin central magma chambers rather than theabsence of local mantle heterogeneity.Alcedo sample E-132 has isotopic composi-tions identical to those of Cerro Azul (Fig. 3). Itis highly unlikely that it erupted from a laterallyintruded dike,because such a dike would haveto pass beneath the Sierra Negra edifice (Fig. 1).Its srcin may be similar to that of Wolf lavaW95-61,i.e.,a product of melting of a region of mantle with a higher proportion of plume to as-thenosphere than that generally prevailing be-neath Alcedo. TYPE 3 ILLEGITIMACY:ADVANCEDMELTING The Cerro Azul illegitimate magma is isoto-pically similar to other Cerro Azul lavas (Fig.3D) but has unusually low incompatible ele-ment abundances (Fig. 2B). Modeling of therare earth element concentrations of normalCerro Azul magmas shows that they are due toaggregation of fractional melts,where the inte-grated extent of melting ranges from 2% to 5%(Naumann,1998). However,the illegitimatemagma from Cerro Azul is modeled to resultfrom about 8% total melting. This magma mustcome from a Cerro Azul–like source that was ei-ther hotter or ascended to shallower levels thannormal Cerro Azul mantle. It also bypassed thecentral magma reservoir. IMPLICATIONS FOR MAGMA TRANSPORT The identification of illegitimate magmas inthe Galápagos has several important implicationsto magma genesis and transport in this hotspotprovince. Foremost,each illegitimate magmaerupted from vents low on the flanks of the vol-cano:none is known to have erupted from a ventnear the summit,despite hundreds of analyses of summit lavas from these volcanoes. The westernGalápagos volcanoes have some of the most im-pressive calderas on Earth (Wood,1984),indicat-ing shallow magma chambers where the magmasevolve by cooling and crystal fractionation (Geistet al.,1998). In addition,the magma reservoirsmust act as mixing chambers,effectively dilutingeven small compositional variations imparted bydifferent batches of parental magmas. Type2 andtype3 illegitimate magmas must ascend throughthe lithosphere outside the main plumbing sys-tems and never interact with the normal magmas.Type 1 magmas may srcinate from the sub-caldera reservoirs of an adjacent volcano,butthey rarely,if ever,penetrate the main magmachamber of the volcano,as has happened at Ki-lauea (Rhodes et al.,1989).The illegitimate magmas also lend insight intothe mantle mixing and melting process. Specifi-cally,they support the notion that the isotopicvariation in island chains like the Galápagos isdue to mixing of different mantle reservoirs. Thetime scale of mixing in the mantle must be suchthat it is almost,but not totally,complete in thetime it takes for a volcano to drift about 30km(the typical spacing between Galápagos volca-noes),roughly 800k.y. for the Galápagos. ACKNOWLEDGMENTS This work is supported by National Science Foun-dation grants EAR-9405462 and EAR-9612110 toGeist and EAR-9628281 to White. Thanks to Paul Wal-lace and Marc Hirschmann for insightful reviews,andto the pioneer of illegitimacy,Mike Rhodes,for steer-ing us away from distasteful puns. REFERENCES CITED Chadwick,W. W.,and Howard,K. A.,1991,The pat-tern of circumferential and radial eruptive fis-sures on the volcanoes of Fernandina and Isabelaislands,Galápagos:Bulletin of Volcanology,v.53,p.259–275.Frey,F. A.,and Rhodes,J. M.,1993,Intershield geo-chemical differences among Hawaiian volcanoes:Implications for source compositions,meltingprocess and magma ascent paths:Royal Societyof London Philosophical Transactions,ser.A,v.342,p.121–136.Geist,D. J.,White,W. M.,and McBirney,A.R.,1988,Plume-asthenosphere mixing beneath the Galá-pagos Archipelago:Nature,v.333,p.657–660.Geist,D.,Howard,K. A.,Jellinek,A. M.,and Ray-der,S.,1994,The volcanic history of Volcán Al-cedo,Galápagos Archipelago:A case study of rhyolitic oceanic volcanism:Bulletin of Vol-canology,v.86,p.243–260.Geist,D.,Howard,K. A.,and Larson,P. B.,1995,Thegeneration of oceanic rhyolites by crystal frac-tionation:The basalt-rhyolite association at Vol-cán Alcedo,Galápagos Archipelago:Journal of Petrology,v.36,p.965–982.Geist,D.,White,W.,Harpp,K.,Albarede,F.,andReynolds,R.,1997,Volcanic development andpetrologic evolution of a collapsed Galápagosshield:Volcan Ecuador:Eos (Transactions,American Geophysical Union),v.78,p.780.Geist,D.,Naumann,T.,and Larson,P.B.,1998,Evo-lution of Galápagos magmas:Mantle and crustallevel fractionation without assimilation:Journalof Petrology,v.39,p.953–971.Harpp,K.,1995,Magmatic evolution at hotspots andmid-ocean ridges:Isotopic and trace elementstudies from the Galápagos Islands,the East Pa-cific Rise and the Mid-Atlantic Ridge. [Ph.D. the-sis]:Ithaca,New York,Cornell University,442p.Hildreth,W.,1987,New perspectives on the eruption of 1912 in the Valley of Ten Thousand Smokes,Kat-mai National Park,Alaska:Bulletin of Volcanol-ogy,v.49,p.680–693.Kurz,M.,and Geist,D.,1999,Dynamics and evolutionof the Galápagos hotspot from helium isotope geo-chemistry:Geochimica et Cosmochimica Acta.Langmuir,C. H.,Bender,J. F.,and Batiza,R.,1986,Petrological and tectonic segmentation of theEast Pacific Rise,5°30 ′ –14°30 ′ N:Nature,v.322,p.422–429.McBirney,A. R.,and Williams H.,1969,Geology andpetrology of the Galápagos Islands:GeologicalSociety of America Memoir118,197p.Naumann,T.,1998,Geology and petrology of CerroAzul volcano,Isabela Island,Galápagos Archi-pelago [Ph.D. thesis]:Moscow,University of Idaho,153 p.Naumann,T.,and Geist,D.,1999,Physical volcanol-ogy and structural development of Cerro Azulvolcano,Isabela island,Galápagos:Implicationsfor the development of Galápagos-type shieldvolcanoes:Bulletin of Volcanology.O’Hara,M. J.,1977,Geochemical evolution duringfractional crystalization of a periodically refilledmagma chamber:Nature,v.266,p.503–507.Rhodes,J. M.,Wenz,K. P.,Neal,D. A.,Sparks,J.W.,and Lockwood,J.P.,1989,Geochemical evidencefor invasion of Kilauea’s plumbing system byMauna Loa magma:Nature,v.337,p.257–260.Rowland,S. K.,Munro,D. C.,and Perez-Oviedo,V.,1994,Volcán Ecuador,Galápagos Islands:Ero-sion as a possible mechanism for the generationof steep-sided basaltic volcanoes:Bulletin of Vol-canology,v.56,p.271–283.White,W. M.,and Geist,D.,1998,Effects of sector col-lapse and caldera disruption on magma transportin Volcán Ecuador,Galápagos Islands:Eos(Transactions,American Geophysical Union),v.79,p.962.White,W. M.,McBirney,A.R.,and Duncan,R.A.,1993,Petrology and geochemistry of the Galápa-gos Islands:Portrait of a pathological mantleplume:Journal of Geophysical Research,v.98,p.19,533–19,564.Wood,C.,1984,Calderas:A planetary perspective:Jour-nal of Geophysical Research,v.89,p.8391–8406.Manuscript received May 13,1999Revised manuscript received August 23,1999Manuscript accepted September 14,19991106 Printed in U.S.A. GEOLOGY,December 1999 on July 17, 2011geology.gsapubs.orgDownloaded from
