Petrogenesis of mid-Proterozoic A-type granites: Casestudies from Fennoscandia (Finland) and Laurentia

(New Mexico)

by

Paula J. KosunenDivision of Geology and Mineralogy

Department of GeologyUniversity of Helsinki

P. O. Box 64FIN-00014 Helsinki, Finland

ACADEMIC DISSERTATION

To be presented, with the permission of the Faculty of Science of theUniversity of Helsinki, for public criticism in the small auditorium E204 of

Physicum, Kumpula, on February 6th, 2004, at 14 o’clock afternoon.

Helsinki 2004

2

Cover: Synplutonic minette pipe within the Jack Creek granite on the southeastern flank of theGila Middle Box, Burro Mountains, southwestern New Mexico. Photo: Tapani Rämö.

ISBN 952-91-6826-8 (paperback)ISBN 952-10-1648-5 (PDF)

YLIOPISTOPAINOHELSINKI 2004

3

Kosunen, P., 2004. Petrogenesis of mid-Proterozoic A-type granites: Case studies from Fennoscandia(Finland) and Laurentia (New Mexico). Academic dissertation, University of Helsinki, Finland.

Detailed studies of four granite plutons, two in southern Finland and two in southwestern New Mexico,indicate considerable petrologic variation for and bring new insights into the petrogenesis of the mid-Proterozoic A-type granites. The rapakivi plutons of Bodom and Obbnäs are the westernmost intrusions ofthe classic rapakivi area of southeastern Finland and show, despite their close proximity and temporalassociation, geochemical (including isotopic) differences indicative of different sources. The Bodom plutoncomprises a series of porphyritic granites (hornblende-, hornblende-biotite, and biotite-granite) and aneven-grained biotite-hornblende granite; the Obbnäs pluton is composed of porphyritic hornblende-biotitegranite intermingled with minor even-grained granodiorite. The granites of both plutons are aluminous A-type, but the Obbnäs granite exhibits higher average TiO2, CaO, MgO, P2O5, Ba, and Sr and lower FeO*/(FeO*+MgO), K2O, Rb, and Nb than the Bodom granites. Mineral chemistry implies that the plutons werecrystallized at ~ 900 to 800 ºC and ~ 2.7 to 5.0 kbar under low fO2. The two plutons have slightly differentinitial Nd isotope compositions (and possibly also different initial Sr isotopes) indicating a slightly olderoverall source for the Obbnäs pluton. Two sets of diabase dikes (Kopparnäs and Vihti) are associated withthe plutons; the dikes at Kopparnäs may be part of a different (high-Nb, high-Ti) magma type thatcompositionally approaches ocean island basalt, and are thus different from the typically tholeiitic rapakivi-associated mafic dikes. Quantitative modeling indicates that the Obbnäs pluton probably originated bypartial melting of a ferrodioritic/ferromonzo-dioritic source, with contribution from a mafic melt similar incomposition to the associated diabases; partial melt from a tonalitic source possibly influenced the evolutionof the granite. A more alkali feldspar –rich (granodioritic) source is required for the granites of the Bodompluton. New U-Pb data on the Obbnäs granite suggest an emplacement age of ~1640 Ma, whereas in theBodom pluton, two generations of zircon with different U-Pb ages are found, one defining an age of 1650Ma and another yielding an age of ~1638 Ma. This age difference of ~10 Ma may give a rough estimate ofthe time span required for the formation of a rapakivi melt in the lower crust and its emplacement at highercrustal levels.The northern Burro Mountains in southwestern New Mexico reveal two distinct, intimately juxtaposedMesoproterozoic magmatic suites represented by the granite plutons of Jack Creek and Redrock and theassociated mafic rocks. At ~1630 Ma, the newly formed Mazatzal crust was intruded by tholeiitic diabasewith a depleted-mantle –type Nd isotope composition but with enriched incompatible trace elementabundances. The potassic granite-minette suite of Jack Creek was emplaced ca. 1460 Ma, and was followedat ~1225-1220 Ma by the tholeiitic A-type granite-anorthosite suite of Redrock. The Jack Creek granitefrequently displays a nice rapakivi texture, but is only marginally A-type with lower Fe/Mg, Ga/Al, Zr, andREE, for instance, than in actual A-type granites. Minette is found as isolated enclaves, swarms of enclaves,and synplutonic dikes within the granite, and evidence for intensive commingling with the granite host isoften present. The Jack Creek granite-minette suite bears similarities to the 1.8 Ga post-collisional, bimodalintrusions with shoshonitic affinities found in southern Finland. The Redrock pluton is composed ofporphyritic hornblende- and hornblende-biotite granites and a later miarolitic biotite granite phase, whichshow the chemical and mineral composition typical of A-type granites. About 50 anorthosite/leucogabbroxenoliths are found scattered in the miarolitic biotite granite; they are excessively altered, but have chemicalcomposition similar to massif-type anorthosites. The diabase-minette-anorthosite sequence and the associatedsilicic rocks record dominantly juvenile additions to the cratonic margin and imply subcontinental enrichmentevents at ~1650 Ma (accretion), prior to 1460 Ma (potassic metasomatism), and at ~1220 Ma (magmaticunderplating); the latter two may have been controlled by a major transcurrent structure along the southernmargin of Laurentia.

Keywords: anorogenic, A-type granite, bimodal magmatism, geochemistry, Fennoscandia, Finland,Laurentia, minette, New Mexico, partial melting, petrogenesis, Proterozoic, radiogenic isotopes

Paula Kosunen, Department of Geology, P.O. Box 64, University of Helsinki, FIN-00014 Helsinki,Finland

4

INTRODUCTION

Mid-Proterozoic (1.8 – 1.0 Ga) A-type granitemagmatism of continental interiors has been atthe focus of active worldwide research for thelast two decades. These rocks, which presum-ably represent the most voluminous intraplatesilicic magmatism on Earth, are often aligned in alinear or semi-linear manner across Precambriancratons and are known from all continents (e.g.,Bridgwater & Windley, 1973; Anderson, 1983;Anderson & Morrison, 1992; Rämö & Haapala,1995).

These mid-Proterozoic granites are charac-teristically found as discordant, multiphase plu-tons that show subaluminous (or aluminous) A-type geochemical composition, are temporallyand spatially related to mafic rocks (bimodalmagmatic association), and often include rocktypes with rapakivi texture (e.g., Anderson, 1983;Haapala & Rämö, 1990; Emslie, 1991; Dall’Agnolet al., 1994; King et al., 1997; Frost et al., 2001).These plutons usually clearly postdate the lat-est preceding orogenic episode in any one re-gion and can thus be considered anorogenic incharacter (cf. Condie, 1991; Rämö & Haapala,1995; Rämö et al., 2002). The classic rapakivi gran-ites of Finland, for instance, were emplaced ~250to 400 m.y. after the formation of the ~1.9 GaSvecofennian crust (Haapala & Rämö, 1992).Many of these plutons were emplaced in an ex-tensional tectonic environment; this has beenclearly demonstrated for the Fennoscandianrapakivi granites (e.g., Haapala & Rämö, 1992;Korja & Heikkinen, 1995).

The petrogenesis of the mid-Proterozoic A-type granites, one of the key issues in evolutionand internal fractionation of Precambrian ter-ranes, has been ascribed to thermal perturba-tions in the subcontinental mantle resulting inmagmatic under- and intraplating and subse-quent anatexis of deep crust (e.g., Anderson,1983; Haapala, 1988; Elo & Korja, 1993; Rämö &Haapala, 1995). In this scenario, the mafic rocks(gabbroids, anorthosite, basaltic dikes, and rarebasalt lavas) are believed to be derivatives ofthe mantle melts that caused the anatexis, andformation of the silicic magmas has been as-cribed to dehydration melting of residual (granu-

litic) or fertile (yet relatively dry) lower crustaldomains (e.g., Collins et al., 1982; Christiansenet al., 1983; Clemens et al., 1986; Creaser et al.,1991) that consisted of metaigneous (dioritic togranodioritic) sources (e.g., Anderson & Cull-ers, 1978; Rämö, 1991; Andersson, 1997; PatiñoDouce, 1997; Dall’Agnol et al., 1999; Smith et al.,1999). Some workers have also suggested mantleinput to be essential for the formation of thesegranites, either by fractionation of mantle-de-rived magmas or through mixing of mantle-de-rived magmas with the anatectic melts (e.g., Barkeret al., 1975; Eklund et al., 1994; Shannon et al.,1997; Smith et al., 1999).

A currently debated issue is the lithologiccharacter of the source(s) of the mid-Protero-zoic anorogenic granites. The redox state of theserocks ranges from reduced (Haapala & Rämö,1990; Frost & Frost, 1997) to oxidized (Ander-son & Bender, 1989; Dall’Agnol et al., 1999) sug-gesting that their sources may have varied frommafic, shortly (or directly) mantle derived (re-duced) to more felsic (oxidized). Frost & Frost(1997) noted that many of these plutons (includ-ing the classic Finnish rapakivi granites) are re-duced (low fO2 during crystallization) implyingthat a major sedimentary component in them isnot feasible and that they rather reflect relativelymafic (~basaltic) source compositions. In linewith this, Frost et al. (1999) concluded the 1.43Ga Sherman batholith of Wyoming to have beengenerated by anatexis of mafic (tholeiitic) lowercrust. More silicic sources have, however, beensuggested by many previous, or parallel, stud-ies (e.g., Anderson & Cullers, 1978; Rämö, 1991;Andersson, 1997; Smith et al., 1999) to explainthe high silica and potassium contents of theseplutons.

The papers included in this thesis partici-pate in this discussion by presenting case stud-ies from two prominent Precambrian crustalunits, Fennoscandia and Laurentia. The studyareas comprise two rapakivi granite plutons(Bodom and Obbnäs) in southernmost Finland(southestern Fennoscandia) and two granite plu-tons (Redrock and Jack Creek) in the Redrockarea of northern Burro Mountains in southwest-ern New Mexico (southern Laurentia); the gran-ites of these four plutons are temporally and

5

THE PAPERS

The thesis is based on the following five pa-pers, which are referred to by their Roman nu-meral notation:

I Kosunen P. (1999). The rapakivigranite plutons of Bodom andObbnäs, southern Finland: petrogra-phy and geochemistry. Bulletin of theGeological Society of Finland 71:275-304.

II Kosunen, P.J., Rämö, O.T. &Vaasjoki M. Petrogenesis of theBodom and Obbnäs rapakivi plutons,southern Finland: Reduced, alumi-nous A-type granites from variablesources. Journal of Petrology, inreview.

III Mclemore, V.T., Rämö, O.T.,Kosunen, P.J., Haapala, I., Heizler,M. & McKee, C. (2000). Geology andgeochemistry of Proterozoic graniticand mafic rocks in the Redrock area,northern Burro Mountains, Grant

Paper I

Paper I focuses on the lithology, petrographyand overall whole-rock geochemistry of theBodom and Obbnäs plutons, which intrude the~1850 Ma Svecofennian basement of southern-most Finland. These relatively small (~5 by 15km), ~1645 Ma granite plutons show the typicalpetrographic and geochemical features of theProterozoic rapakivi granites in Finland andelsewhere, and are the westernmost intrusionsof the classic Wiborg rapakivi granite area ofsoutheastern Finland.

The Bodom pluton, located about 15 kmnorthwest of Helsinki, is lithologically more vari-able of the two comprising three varieties of por-phyritic granites (hornblende-, hornblende-bi-otite, and biotite-bearing) and an even-grainedgranite with slight textural variation. The con-tacts between the porphyritic granites, whichoccupy the northeastern and central parts ofthe pluton, are mostly gradational and difficult

spatially associated with assorted mafic rocksforming a bimodal magmatic association. TheBodom and Obbnäs plutons are the westernmostintrusions of the classic Wiborg rapakivi area.They are coeval, spatially closely related, andwere emplaced into the ~1.87 Ga Paleoproterozoicbedrock of southern Finland ~1.64 Ga ago alonga major shear zone. They display important litho-logic and compositional differences that allowelaboration on source composition and, in gen-eral, on issues relevant to the redox budget ofthe mid-Proterozoic A-type granites. In theRedrock area, the felsic and mafic rocks of theRedrock and Jack Creek plutons record repeatedintrusions into the cratonic rocks of the centralMazatzal province during the Mesoproterozoic.In addition to the questions related to the petro-genesis of A-type granites, they provide infor-mation about the ongoing processes at the pre-sumed craton margin and tectonic activity insouthern Laurentia during the Mesoproterozoictime.

County, New Mexico; A progressreport. In: Lawton, T.F., McMillan,N.J., McLemore, V.T., Austin, G., andBarker, J.M. (Eds.) SouthwestPassage, A Trip through the Phanerozoic. New Mexico GeologicalSociety Guidebook 51: 117-126.

IV McLemore, V. T., Dunbar, N.,Kosunen P. J., Rämö, O. T., HeizlerM., & Haapala, I. (2003). Geology andgeochemistry of the Redrock graniteand anorthosite xenoliths (Protero-zoic) in the northern Burro Moun-tains, Grant County, New Mexico.Bulletin of the Geological Societyof Finland 74: 7-52.

V Rämö, O.T., McLemore, V.T.,Hamilton, M.A., Kosunen, P.J.,Heizler, M. & Haapala, I. (2003).Intermittent 1630-1220 Ma magmatismin central Mazatzal province: Newgeochronologic piercing points andsome tectonic implications. Geology31: 335-338.

6

to pinpoint. The southwestern part of the plu-ton is formed by the even-grained granite thathas a sharp, well-defined contact against theporphyritic hornblende-biotite granite and prob-ably represents a separate intrusive phase. Theporphyritic granites are red or reddish brownand medium- to coarse-grained syenogranites.They contain alkali-feldspar megacrysts, whichare up to 4 cm in diameter and occasionallymantled by plagioclase. Ovoidal megacrysts anddrop quartz are common in the porphyritic horn-blende granite. The even-grained granite is red,medium-grained, and homogeneous, with grainsize of 2 to 5 mm. A weakly porphyritic texture infound in places. Common accessory mineralsinclude fluorite, allanite, zircon, apatite, and iron-titanium oxides.

The Obbnäs pluton is almost entirely com-posed of porphyritic hornblende-biotite granitethat gradually becomes more mafic to the south-west; the granite ranges from syeno- tomonzogranite. Magmatic microgranular enclavesof intermediate composition are found in thesouthern part of the pluton, and the southwest-ern tip of the Obbnäs Peninsula –Obbnäsudden–also contains hybrid granitoids. The porphy-ritic hornblende-biotite granite is red to reddish-brown and relatively coarse-grained with alkali-feldspar megacrysts up to 5 cm in diameter; ovoi-dal megacrysts and plagioclase mantles are morefrequent in the southern part of the pluton. Thehybrid rocks of Obbnäsudden comprise even-grained granodiorite and two types of porphy-ritic rocks, a sparsely porphyritic (granodiorite)and a densely porphyritic (granite), which dis-play well defined, but not cutting contactsagainst each other. The densely porphyritic hy-brid seems to grade into the actual Obbnäs gran-ite. The porphyritic hybrid rocks contain, in ad-dition to mostly ovoidal alkali feldsparmegacrysts, euhedral plagioclase crystals thatare coarser than the groundmass. Accessorytitanite, fluorite, allanite, zircon, apatite, and iron-titanium oxides are present in the actual Obbnäsgranite and the densely porphyritic hybrid; inthe sparsely porphyritic and the even-grainedhybrids, fluorite is missing.

Sporadic magmatic foliation, visible as par-allel orientation of the sub- to euhedral alkali

Paper II

Paper II deals with the petrogenesis of theBodom and Obbnäs plutons. It includes new U-Pb age determinations, a full set of whole-rockchemical data, mineral chemical data, and Nd-Sr-Pb isotope data that were utilized to determineconditions of crystallization for the two plutons,constrain their possible sources, and present apetrogenetic model able to explain the observed

feldspar megacrysts, is found in the marginalareas of the Bodom pluton. Magmatic foliationis more frequent in the Obbnäs pluton and nearthe northwestern rim, adjacent to the Porkkala-Mäntsälä shear zone, more intense foliation aswell as signs of brittle deformation are present.

The Bodom and Obbnäs granites show thetypical geochemical features of Proterozoic A-type (rapakivi) granites in Finland and elsewhere.They are metaluminous to weakly peraluminousand with average A/CNK of 1.00 and 1.05, re-spectively and have high FeOtot/(FeOtot+MgO)that averages at 0.94 for the Bodom and 0.87 forthe Obbnäs granites. The granites of both plu-tons have high Ga/Al (3.78 to 5.22 in Bodom and2.46 to 4.18 in Obbnäs) as well as HFS element(especially Zr, Nb, Ce, and Y) contents that clearlydefine them as A-type. The REE contents arehigh with LREE-enriched chondrite-normalizedpatterns and moderate (Obbnäs) to relativelystrong (Bodom) negative Eu-anomalies. Thecompositions of the porphyritic granites ofBodom overlap, but a trend of increasing aver-age SiO2 and decreasing average FeOtot, TiO2,Al2O3, MgO, and P2O5 is observed in the se-quence hornblende granite – hornblende-biotitegranite – biotite granite. The Obbnäs granite isenriched in CaO, TiO2, MgO, and FeO, and de-pleted in SiO2 and K2O compared to the Bodomgranites. Also, there are differences in the traceelement contents of the two plutons, most nota-bly Ba, Rb, and Sr: the average Rb/Ba and Rb/Srof the Bodom granites are 0.44 and 3.14, whilethe respective values for the Obbnäs granite areconsiderably lower, 0.14 and 0.89. Given the highamounts of Ba and Sr in the hybrid granitoids ofObbnäs, mixing might provide an explanation tothe composition of the Obbnäs granite.

7

differences between them. Two sets of diabasedikes associated with the plutons are includedin the discussion.

The ~1.65 Ga, aluminous A-type rapakiviplutons of Bodom and Obbnäs (southern Fin-land) show geochemical (including isotopic) dif-ferences indicative of different sources. TheBodom pluton consists of a series of porphy-ritic granites (hornblende, hornblende-biotite,and biotite granite) and an even-grained biotite-hornblende granite. The Obbnäs pluton is litho-logically more monotonous with a dominatingporphyritic hornblende-biotite granite inter-mingled with a minor even-grained granodiorite.Two sets of diabase dikes crop out west of theplutons in Kopparnäs (close to Obbnäs) and inVihti (close to Bodom). They strike WNW, areup to 15 m wide, and correlate with the tholeiiticdiabase dikes associated with the southeasternFinnish rapakivi intrusions.

The new U-Pb data on the Obbnäs granitesuggest an emplacement age of ~1640 Ma forthe pluton and support the previous results. Twogenerations of zircon with different U-Pb agesare found in the Bodom pluton: a dark brownvariety defining an age of 1650 Ma and a lightbrown variety yielding an age of ~1638 Ma. Thisage difference of ~10 Ma may give a rough esti-mate of the time span required for the formationof a rapakivi melt in the lower crust and its intru-sion to higher crustal levels.

Mineral chemical data include analyses offeldspars, mica, amphibole, and Fe-Ti oxides. Therocks of both plutons contain iron-rich biotiteand amphibole typical of the rapakivi granites.The Fe/(Fe+Mg) of the biotites from Bodom var-ies from 0.86 to 0.98 while those from Obbnäshave values ranging from 0.73 to 0.80; amphib-oles, which are classified as hastingsite, showvalues from 0.96 to 0.99 and 0.83 to 0.89, respec-tively. Al-in-hornblende barometry suggestspressures of ~ 2.7 to 5.0 kbar for the two plu-tons. Even though these values are similar tothe ones reported recently for the Wiborgbatholith, they are probably over-estimates dueto the high Fe/(Fe+Mg) of the amphiboles. Zir-con saturation temperatures range from ~ 950 to830 ºC for the two plutons; amphibole-plagio-clase equilibria also records reasonable mag-

matic temperatures yielding values about 50-140ºC lower than zircon saturation. Oxygen fugac-ity estimates based on the compositions of themafic silicates (especially amphibole) suggestthe two plutons to have crystallized under lowfO2, Obbnäs under fO2 similar to that implied forthe Wiborg batholith and Bodom under evenmore reducing conditions.

The two plutons have slightly different ini-tial Nd isotope compositions indicating a slightlyolder overall source for the Obbnäs pluton: inthe Bodom pluton, the average εNd (at 1640 Ma)and TDM values are –0.8 and 2.02 Ga, respec-tively, while the corresponding values for theObbnäs pluton are –1.7 and 2.07 Ga. The εNdvalues of the associated diabase dikes are, onaverage, slightly higher than those of the Bodomand Obbnäs granites, and their TDM model agesrange from 2.02 to 2.10 Ga. The initial 87Sr/86Srratios scatter widely, but it is possible that asmall difference exists between the plutons, withSri of 0.7040 for Bodom and 0.7047 for Obbnäs.The Pb isotopic data also show considerablescatter and indicate long-term Th/U of ~4 for thegranites of both plutons.

Quantitative modeling suggests that partialmelts from a ferro(monzo)dioritic and a tonaliticsource contributed to the Obbnäs pluton, to-gether with a mafic melt similar in compositionto the Vihti diabases. The partial melts mightseparately have been mixing with the mafic ma-terial producing the even-grained granodioriteand the porphyritic granite, which then inter-mingled during intrusion, or the mixture of theferrodiorite partial melt and mafic melt could haveinteracted with the tonalite partial melt to pro-duce the granite. The geochemical features ofthe porphyritic granites of Bodom are best re-produced by partial melting of a more alkali feld-spar –rich, possibly granodioritic, source.

Paper III

Paper III introduces the lithology and petrologyof the Proterozoic igneous rocks in the Redrockarea, northern Burro Mountains, southwesternNew Mexico. It also reports 40Ar/39Ar geo-chronological data as well as preliminarygeochemical and Nd isotopic results.

8

The Redrock area comprises a complex Pro-terozoic terrain consisting of metamorphic rocks(Bullard Peak and Ash Creek series, 1550-1570Ma) intruded by assorted granitic and maficrocks. These are:

(1) ~1445 Ma (U/Pb; Hedlund, 1980) BurroMountains granite, which is exposed south ofthe actual study area, is a pinkish-gray, medium-grained, in places foliated biotite granite. It iscalc-alkaline, metaluminous to peraluminous, andprobably consists of more than one pluton asindicated by its variable geochemical composi-tion.

(2) ~1380 Ma (U/Pb; Hedlund, 1980)gneissic granite/granodiorite is gray, medium-to coarse-grained, in places porphyritic biotite-hornblende granite/granodiorite that is typicallyfoliated. It is exposed in the eastern part of thestudy area, where it is occasionally found incontact with Jack Creek rapakivi granite. Thegneissic granite is calc-alkaline and metalumi-nous, and has lower SiO2 and higher TiO2, CaO,MnO, Ba, and V than the other three granites.

(3) Jack Creek rapakivi granite is pink-gray to red-orange, medium- to coarse-grainedand characterized by large alkali feldspar phe-nocrysts occasionally mantled by plagioclase(rapakivi texture). This granite is the major rocktype in the eastern part of the study area and isexposed from Lydian Peak eastward to theSchoolhouse Mountain fault. The Jack Creekrapakivi granite is calc-alkaline, weakly peralu-minous, high-K granite with marginal A-typegranite geochemical characteristics. It differsfrom the classic Proterozoic rapakivi granites ofFinland in having lower Fe/Mg, Ga/Al, Zr, andHREE, for example. 40Ar/39Ar gives ages of ~1220Ma and 1200 Ma for the Jack Creek rapakivi gran-ite, but field relations suggest it to be older thanthe Redrock granite.

(4) Minette is found in isolated enclaves,swarms of enclaves, and synplutonic dikeswithin the Jack creek rapakivi granite. The ma-jority of the minette swarms and synplutonicdikes are found along Gila Middle Box. Theminette is dark gray, fine- to medium-grained,and porphyritic with dark mica and, probably,altered pyroxene phenocrysts. The minette of-ten shows evidence for intensive commingling

with the rapakivi granite, and pillows of hybridmaterial are common on the edges of synplutonicdikes and within swarms of enclaves. 40Ar/39Arage for the minette is ~1420 Ma, Sm/Nd isoch-ron yields an age of ~1135 Ma.

(5) Redrock granite forms the western partof the study area extending from Smith Canyonsoutheastward to Ash Creek. It consists of fourphases, which are texturally and mineralogicallydistinct, but have similar geochemical composi-tions and are interpreted to be a single, zonedpluton. Hornblende granite is orange to red-brown, coarse-grained and often slightly foli-ated. East of Ash Creek, the hornblende graniteis in contact with the Jack Creek rapakivi graniteand appears to be chilled against it. Biotite-horn-blende granite, which is the dominant phase ofthe Redrock granite, is orange to red-brown,medium- to coarse-grained, and grades into thehornblende granite. Miarolitic biotite graniteis orange-pink, fine- to medium-grained, and char-acterized by nearly circular miarolitic cavities upto 15 cm in diameter. This phase has sharp or, inplaces, brecciated contacts against the horn-blende granite and it is interpreted to be a later(younger) phase in the upper part of the pluton.Alkali-feldspar granite is found as small lensesand dikes within the miarolitic biotite granite,from which it differs by a lighter color (white tolight pinkish-gray), fewer miarolitic cavities, andabsence of biotite. The contacts are gradational,and this phase is most likely a differentiate ofthe miarolitic biotite granite. All phases of theRedrock granite are calc-alkaline, metaluminousto peraluminous, and show geochemical char-acteristics of A-type granites. Sm/Nd isochronage for the Redrock granite is ~1328 Ma; 40Ar/39Ar gives ages of ~1210 Ma and 1205 Ma forthe hornblende granite phase.

(6) About 50 anorthosite/leucogabbroxenoliths are found in a northeast-trending zonewithin the miarolitic biotite granite phase of theRedrock granite. These bodies are up to 270 mlong and 30 m wide and commonly brecciatedby diabase/gabbro and the miarolitic granite. Theanorthosites are heterogeneous, tan to white-gray and black, fine to coarse-grained, and con-tain white patches of leucogabbro. Though of-ten heavily altered, the Redrock anorthosites are

9

Paper IV

Paper IV concentrates on the Redrock graniteand spatially associated anorthosites in thenorthern Burro Mountains, southwestern NewMexico, and discusses in detail their lithologyand petrography as well as mineral, whole-rock,and isotope geochemistry.

A rare spatial association of A-type granitesand anorthosites is found in the Redrock area inthe northwestern Burro Mountains. The oldestrocks in the area are the metamorphic Bullard

similar in chemistry to other anorthosites world-wide. The ages for the anorthosite xenoliths are~1220 Ma and 1230 Ma (40Ar/39Ar) and ~1326Ma (Sm-Nd isochron).

(7) Diabase/gabbro/diorite dikes andsmall stocks are black to dark gray and fine- tomedium-grained. These mafic rocks show localfoliation, and they have intruded all of the gra-nitic and metamorphic rocks. It is likely that therehas been two or more periods of mafic magmatismin the area.

(8) Several kinds of felsic dikes intrude themain granites: the Redrock, Jack Creek rapakivi,and Burro Mountains granites are intruded byfine-grained biotite and K-feldspar granitedikes, and rhyodacite/dacite porphyry dikes arefound intruding the Redrock granite. Dikes ofthe former group occasionally cut the latter. Thecontacts of these dikes are sharp and locallychilled along the margins, indicating them to beyounger than the main granites of the area. An40Ar/39Ar age of ~1210 Ma exists from one of thefine-grained biotite granite dikes.

The available age data (40Ar/39Ar, Sm/Nd iso-chron) may reflect later thermal events ratherthan actual crystallization. The clustering of 40Ar/39Ar age dates around 1200 Ma suggests that aheating event or uplift took place at that time.Geochemical and isotopic data indicate that theRedrock granite, Jack Creek rapakivi granite, andBurro Mountains granite were derived from (atleast slightly) different sources. These wereclearly more juvenile than the metamorphic rocksof the region and may have included a signifi-cant Middle Proterozoic mafic lower crust/sub-continental lithospheric mantle component.

Peak and Ash Creek metamorphic rocks that com-prise a variety of quartzo-feldspathic gneissesand schists, biotite and hornblende schists,amphibolite, phyllite, quartzite, and serpentine-carbonate rocks. The interlayered units of thesemetaigneous and metasedimentary rocks are dis-rupted by the intrusion of the Redrock granite,and they locally show signs of contact meta-morphism adjacent to the granite. The Redrockgranite is divided into four phases that are tex-turally and mineralogically distinct, but havechemical compositions that imply them to begenetically related. The bulk of the granite isformed by hornblende and hornblende biotitegranites that are rather similar in appearance andgrade into each other. These granites are orangeto reddish brown and medium- to coarse-grained,and consist of plagioclase, alkali feldspar, quartz,hornblende, and biotite (none to trace amountsin the hornblende granite); accessory mineralsinclude zircon, titanite, ilmenite, and apatite. Thethird phase, miarolitic biotite granite, is found inthe southeastern part of the pluton, where it isin contact with the hornblende granite. The con-tacts are sharp and, on occasion, brecciated in-dicating the miarolitic biotite granite to be ayounger phase. This orange to pink, fine- tomedium-grained granite is characterized by cir-cular to elliptical miarolitic cavities up to 15 cmin diameter, and consists of plagioclase, alkalifeldspar, quartz, biotite, muscovite, and traceamounts of hornblende, zircon, titanite, magne-tite, ilmenite, monazite, andradite garnet, fluo-rite, and apatite. Locally, the miarolitic biotitegranite is pervasively altered to epidote, quartz,and chlorite; aggregates of epidote, quartz, he-matite, and chlorite are also found withinmiarolitic cavities. This magmatic-hydrothermalalteration is interpreted to result from degassingof the underlying pluton. Within the miaroliticbiotite granite, small lenses of white to light pink-ish-gray, fine-grained alkali feldspar granite arefound. This phase contains alkali-feldspar,quartz, and varying amounts of plagioclase. Thecontacts with the miarolitic biotite granite aregradational, and the alkali-feldspar granite ismost likely a differentiate of the former. About50 anorthosite/leucogabbro xenoliths are scat-tered throughout a northeast-trending zone

10

within the miarolitic biotite granite phase of theRedrock granite. The largest of these bodies is270 m long and 30 m wide. The anorthosites areheterogeneous, altered, and commonly brecci-ated by diabase/gabbro and the miarolitic gran-ite. They are tan to white-gray and black, fine tocoarse-grained, and composed of unzoned,highly altered plagioclase (up to 50 cm long)with hornblende or pyroxene, chlorite, biotite,relict magnetite-ilmenite, and rare titanite, apa-tite, pyrite, olivine, and zircon. The Redrock an-orthosites are best classified as massif-type,though it is made difficult by extensive alter-ation due to hydrothermal influence of theRedrock granite.

Mineral chemical data were acquired on feld-spars, amphiboles, and micas from the horn-blende, biotite-hornblende, and miarolitic biotitegranite phases, and on feldspars and amphib-oles from the anorthosites. The Redrock granitephases all contain potassic alkali-feldspar (or-thoclase) and sodic plagioclase (albite to oligo-clase). The feldspars in the anorthosite samplesappear to preserve no igneous textures, and theircompositions do not, thus, represent originaligneous ones. The least altered plagioclase islabradorite. Amphiboles from the hornblendeand biotite-hornblende granites (ferro-actinoliteto hastingsitic hornblende) are rich in iron, withFe/(Fe+Mg) ranging from 0.7 to 1.0. Amphibolesfrom one of the anorthosites define a wide com-positional field, especially with respect to Fe andMg, which is interpreted as a secondary, alter-ation-related effect. Biotites of the hornblendeand biotite-hornblende granites are iron-richannites with Fe/(Fe+Mg) of 0.7 to 1.0. The com-position of micas in the miarolitic biotite granitevaries widely. Some of the muscovites may benear primary, whereas others clearly result fromsecondary alteration. The magmatic muscovitescontain significant TiO2 (near 0.5 wt.%), whilethe hydrothermal ones contain nearly none.

The Redrock granite is metaluminous toperaluminous, marginally alkaline to subalkaline,and exhibits the geochemical features of A-typegranites. The four phases are similar in bulkgeochemical composition, which implies acomagmatic relationship. The granites increasein SiO2, U, Th, and Rb, and decrease in TiO2,

total Fe, MnO, CaO, P2O5, Zr, Y, and Ba in thesequence hornblende granite – biotite-horn-blende granite – miarolitic biotite granite. Thehornblende and biotite-hornblende phases showthe high Fe/Mg of A-type granites; in themiarolitic granite the ratio is lower and more vari-able. The Redrock granite has average Rb/Ba of1.6, Rb/Sr of 7.6, F of 871 ppm, and shows dis-tinctive A-type characteristics in the Ga/Al-dis-crimination diagrams. Chondrite-normalized REEpatterns are relatively flat, slightly LREE-en-riched, save for the alkali-feldspar granite, whichis slightly enriched in HREE. The negative Eu-anomaly becomes more pronounced from thehornblende granite to the alkali-feldspar gran-ite, consistent with a typical magmatic evolu-tionary sequence.

The four Redrock granite phases fall on aSm-Nd isochron with an age of ~1330 Ma andinitial εNd of +2.7. This isochron age is some-what higher than that indicated by 40Ar/39Ar andpreliminary U/Pb; recalculated at 1220 Ma, theinitial εNd range from +1.3 to 2.2; the TDM are be-tween 1494 and 1666 Ma. In the Sm-Nd isochrondiagram, the anorthosite samples plot roughlyalong the Redrock granite isochron, and theirinitial εNd (at 1220 Ma) and TDM range from +1.5to +2.2 and 1543 to 1636 Ma, suggesting that theanorthosites could represent a magmatic sys-tem that was roughly coeval with the Redrockgranite. The initial 87Sr/86Sr ratios of the indi-vidual granite samples are less than 0.7 at 1220Ma, and suggest that the Rb-Sr systematics havebeen disturbed since the crystallization of thegranite. The initial 87Sr/86Sr of the least radio-genic anorthosite sample is 0.70449 at 1220 Maand may represent the initial ratio of the an-orthosites. The Nd isotopic composition of themetamorphic rocks is very different from that ofthe granites and anorthosites (εNd of –3 to-4 at1220 Ma) showing that the granites could nothave been derived from a crustal source withthis Nd isotopic composition, but required a morejuvenile protolith. This could have been a lowercrust domain formed from the ~1650 Ma crustby hybridisation caused by mafic underplatingslightly prior to the generation of the Redrockgranite and anorthosites.

11

Paper V

Paper V presents U-Pb age data on the majormafic and felsic rocks in the Redrock area,northern Burro Mountains, southwestern NewMexico, and discusses their implications on thetectonic evolution of southern Laurentia.

The Precambrian bedrock of the Redrockarea, west and east of Gila Middle Box in thenorthern Burro Mountains, comprises threeclosely spaced igneous suites: a large diabasedike/sill, the Jack Creek granite - minette suite,and the Redrock granite - anorthosite suite. Thesubcircular ~40 km2 Redrock pluton in the west-ern part of the area is composed of coarse-grained hornblende granite, medium- to coarse-grained biotite-hornblende granite, and fine- tomedium-grained miarolitic biotite granite. About50 xenoliths of coarse anorthosite/leucogabbroup to 30 m wide by 270 long are found within themiarolitic granite. The granites are A-type, andthe anorthosite xenoliths are similar to Protero-zoic massif-type anorthosites. The eastern partof the area exposes the Jack Creek pluton, whichis composed of homogeneous, medium to coarse-grained biotite granite with alkali-feldsparmegacrysts occasionally mantled by plagioclase(rapakivi texture). This granite is transitional be-tween A- and I-type and contains numeroussynplutonic dikes and enclaves of minette. Thedikes can be up to 30 m wide and several hun-dred meters long, and are typically mingled withthe granite forming swarms of enclaves. In thecentral part of the area, intruded into metamor-phic rocks of presumed Mazatzal age that nearlyseparate the two granite plutons, a relativelylarge, medium-grained, tholeiitic diabase dike/sill is found.

U-Pb age determinations on zircon yield 1633Ma as the crystallization age of the diabase. Zir-cons from the Jack Creek granite and minettegive upper-intercept ages of 1461 and 1465 Ma,respectively, and the age of the pluton is con-sidered to be ~1460 Ma. The Redrock pluton isabout 200 m.y. younger: the anorthosite xeno-liths are dated at 1225 Ma, and the crystalliza-tion age of the granite is 1220 Ma. Thus, thenorthern Burro Mountains show three principalages of Proterozoic magmatism, each with a dis-

tinct magmatic assemblage. The Nd isotope composition yields TDM

ages of 1930-1760 Ma to the metamorphic rocksof the area; these ages are older than those pre-viously reported from central Mazatzal province,and suggest the presence or recycling of sig-nificantly older Paleoproterozoic crust. Shortlyafter accretion, the central Mazatzal provinceunderwent extension that led to the emplace-ment of the diabase dike/sill at 1633 Ma. Thediabase may have been derived from subconti-nental lithosphere stabilized in the accretion pro-cess. The sub-Mazatzal mantle underwent po-tassic metasomatism prior to 1460 Ma, as indi-cated by the enrichment of K, LREE, and otherincompatible elements in the minettes; the gran-ites of the bimodal Jack Creek suite probablyrepresent partial melting of the newly formedlower crust. A subsequent episode of bimodalmagmatism took place at 1225-1220 Ma, as evi-denced by the A-type granite – anorthosite suiteof Redrock. Compared to the Jack Creek suite,Nd isotopes point to a slightly younger ances-try for the anorthosites/ leucogabbros ofRedrock, which display an island-arc tholeiiteaffinity. It is therefore likely that an astheno-spheric input to the Mazatzal lithosphere tookplace between 1460 and 1225 Ma, leading, even-tually, to fusion of juvenile lower crust and em-placement of the Redrock pluton.

The minettes of the Jack Creek pluton wereprobably formed in a continental rift zone asso-ciated with metasomatized mantle and resultantalkali- and carbonate-rich magmatism, and an ex-tensional regime is favored also for the juvenile,bimodal Redrock pluton. The two plutons areon a proposed transcurrent zone, thought tohave extended along the southern margin ofLaurentia from western Texas to southeasternCalifornia ca. 1250-1080 Ma (Bickford et al., 2000),and could have been formed by extension andmantle upwelling associated with such a zone;the 1460 Ma Jack Creek pluton could indicatethe zone to have already been active in the earlyMesoproterozoic.

12

DISCUSSION

Comparisons

Several similarities between the two study areasare evident. Both comprise two relatively small(~40 to 60 km2) granite plutons that postdate thesurrounding Paleoproterozoic bedrock by atleast 100 m.y. On both locations, one of theplutons shows clear A-type geochemicalcharacteristics similar, for example, to those ofthe classic rapakivi granites of southwesternFinland, whereas the other more or less differsfrom these. The ‘clearly A-type’ plutons – Bodomin Finland and Redrock in New Mexico – aremulti-phase intrusions that contain a series ofgranites reflecting fractionation of a parent melt,and a more evolved phase that intrudedsomewhat later. Although these plutons are partof a bimodal magmatic association implied bythe (presumably) coeval diabases and an-orthosites, they do not show direct evidence ofmagma mingling or mixing. The other twoplutons, Obbnäs and Jack Creek, are composedof only one type of granite, and are mingledlocally with more mafic material.

In aluminum saturation diagrams, the gran-ites of the Bodom pluton straddle the metalumi-nous/peraluminous –boundary, as do the Red-rock granites. The latter show, however, a widerrange in their A/CNK –values (Fig. 1a). Both theObbnäs and the Jack Creek granites are slightlyperaluminous. The granites from New Mexicoare, on average, higher in silica than the Finnishones, even though the SiO2 contents vary quitea bit within each pluton; the Redrock granitesshow the widest range of variation, also in othermajor elements (Fig. 1). Although different fromeach other, the Bodom and Obbnäs granites havethe high Fe/Mg typical of A-type granites (aver-age FeOtot/(FeOtot+MgO) of 0.94 and 0.87, respec-tively) and show ‘anorogenic’ affinity (Fig. 1b).The iron content and FeOtot/(FeOtot+MgO) of theRedrock granites vary strongly, the latter from0.35 to 0.98 (average 0.84). The Jack Creek plu-ton is clearly different from the other three: theaverage FeOtot/(FeOtot+MgO) is as low as 0.63,mostly because of the high MgO content of thegranite, and five out of six samples are clearly

Obbnäs granite

Bodom granites

Redrock granites

Jack Creek granite

Post-orogenic granites withshoshonitic affinities

60 70 80 900

1

2

3

SiO2

K O/Na O2 2

c)

60 70 80 900.5

1.0

1.5

SiO2

A/CNK

Peraluminous

Metaluminous

a)

RRG+CEUG

60 65 70 75 800.5

0.6

0.7

0.8

0.9

1.0

IAG+CAG+CCG

POG

SiO2

FeO*/(FeO*+MgO) b)

‘Anorogenic’

‘Orogenic’

Fig. 1. The Bodom, Obbnäs, Redrock, and Jack Creekgranites plotted in (a) A/CNK vs. SiO2, (b) FeOtot/(FeOtot+MgO) vs. SiO2, and (c) K2O/Na2O vs. SiO2 varia-tion diagrams. The fields in (b) are from Maniar andPiccoli (1989): RR - rift related, GEU - continentalepeirogenic uplift, PO - post-orogenic, IA - island arc,CA - continental arc, and CCG - continental collisiongranitoids. Data for post-collisional granites from Eklundet al. (1998).

13

‘orogenic’ in character (Fig. 1b; Table 6 in PaperIV; unpublished data on Jack Creek granite).Bodom and Obbnäs granites are also more po-tassic, on average, than the granites of Redrockand Jack Creek (Fig 1c).

In addition to the Fe/Mg, a clear differenceexists in the Rb, Sr, and Ba contents of the twoFinnish plutons, with the Obbnäs granite beingricher in Ba and Sr, and poorer in Rb than thegranites of Bodom, on average (Fig. 2a). TheRedrock granites resemble those of Bodom inthat they are relatively low in Sr and high in Rb;the Rb content of some samples is even higherthan that of the Bodom granites. The Jack Creekgranite has similar average content of Ba thanthe Obbnäs granite, but differs from all the otherones by its higher Sr (Fig. 2a). The characteris-tics of the four plutons are nicely shown in theGa/Al- and major vs. trace element –discrimina-tion diagrams (Figs. 2b and c). Bodom, with itshigh Ga/Al and high contents of HFS elements,is clearly A-type. So is Obbnäs, although not sostrongly: in the Ga/Al –diagrams especially, ittends towards the field of S-, M-, and I-typegranites (Fig. 2b). Redrock shows quite a lot ofvariation, but can be clearly labelled as A-typedespite the samples that cluster rather close tothe field of ‘other’ granites. Jack Creek is onlymarginally A-type plotting partially into the fieldof S-, M-, and I-type granites or their fraction-ates (Figs. 2b and c).

Spidergrams showing average contents ofselected LIL and HFS elements in the differentgranite types of the four plutons are shown inFig. 3. Respective rock types of the similar plu-tons (Bodom/Redrock, Obbnäs/Jack Creek) areplotted as pairs to better display the possibledifferences and similarities. The granites ofBodom and Redrock are, on average, enrichedin LIL and HFS elements compared to the aver-age continental crust, Bodom more strongly thanRed-rock. Exceptions to this are Sr, P, and Ti,which are depleted in all granite types of thesetwo plutons, as are the transition metals Ni andCr. In addition to these, the hornblende and bi-otite –bearing granites of both plutons showsmall negative anomalies in Ba and Nb (Fig. 3b);the Ba-depletion increases in the more evolvedgranites (Figs. 3c and d). Overall, the spidergrams

1 2 3 4 5 60

10

20

30

40

50

60

70

80

90

100

10000*Ga/Al

Nb

A-typeS-, M-,and I-type

b)

Rb Ba

Sr

Rb Ba

Sr a)

A-type

S-, M-,and I-type

Fractionatedgranites

c)

100 1000 20002

10

30

Zr+Nb+Ce+Y

(K2O+Na2O)/CaO

Obbnäs granite

Bodom granites

Redrock granites

Jack Creek granite

Post-orogenic granites withshoshonitic affinities

Fig. 2. The Bodom, Obbnäs, Redrock, and Jack Creekgranites plotted in (a) Rb-Sr-Ba, (b) Nb vs. 10000xGa/Al, and (c) (K2O+Na2O)/CaO vs. Zr+Nb+Ce+Y varia-tion diagrams. The fields for the S-, M-, and I-typegranites in diagrams (b) and (c) are from Whalen et al.(1987). Data for post-collisional granites from Eklundet al. (1998).

14

for the different granite types of Bodom resembleeach other and those of the Redrock granites.The alkali-feldspar granite of Redrock showsLREE and Nb deviating from the others (lowerLREE, high Nb; Fig. 3d), probably due to its highly

0.02

0.1

1

10

30

Rb

Ba

K

Th

U

Sr

La

Ce

Nb

Nd

P

Zr

Eu

Ti

Tb

Y

Yb

Ni

Cr

Hornblende granites

0.02

0.1

1

10

30

Rb

Ba

K

Th

U

Sr

La

Ce

Nb

Nd

P

Zr

Eu

Ti

Tb

Y

Yb

Ni

Cr

Hornblende- and biotite-bearing granites

0.02

0.1

1

10

30

Rb

Ba

K

Th

U

Sr

La

Ce

Nb

Nd

P

Zr

Eu

Ti

Tb

Y

Yb

Ni

Cr

Biotite granites

0.02

0.1

1

10

30

Rb

Ba

K

Th

U

Sr

La

Ce

Nb

Nd

P

Zr

Eu

Ti

Tb

Y

Yb

Ni

Cr

‘Late phase’ granites

0.02

0.1

1

10

30

Rb

Ba

K

Th

U

Sr

La

Ce

Nb

Nd

P

Zr

Eu

Ti

Tb

Y

Yb

Ni

Cr

Obbnäs and Jack Creek granites

Obbnäs granite

Bodom granites

Redrock granites

Jack Creek granite

Fig. 3. Average contents of selected LIL and HFS elements and transition metals in the different granite types ofthe Bodom, Obbnäs, Redrock, and Jack Creek plutons. Values normalized to average continental crust (Weaver andTarney, 1984)

evolved nature. The trace element contents ofthe Obbnäs and, especially, Jack Creek graniteare closer to those of the average continentalcrust (Fig. 3e). The overall shape of the Obbnässpidergram is similar to those of Bodom and

15

Redrock, the depletions in Sr, P, and Ti are not,however, as pronounced. In the Jack Creek gran-ite, the HFS element contents are relatively closeto the average continental crust, enriched ordepleted only by a factor of two. Ni and Cr arealso less depleted than in the granites of theother plutons.

Trace element concentration of an igneousrock can be thought of as a reflection of its sourceand the processes that affected its formation,and spidergrams have been successfully utilizedin depicting basalt chemistry and constrainingpossible sources for them. Even though felsicplutonic rocks are far more complex, it can bepresumed that granites with similar spidergramscan have similar genetic histories. Quantitativemodeling suggests that the Bodom granites wereformed by partial melting of a relatively felsic(granodioritic) crustal source, whereas thesource for the Obbnäs pluton was more mafic(ferrodioritic to tonalitic) and further modifiedby interaction with mafic melt (Paper II Discus-sion). The similarity of the Bodom and Redrockspidergrams implies that the source of the gran-ites of the latter pluton might also have beenrelatively felsic, and that mafic melts did not haveany significant (direct) influence in their forma-tion. The composition of the Jack Creek granite,on the other hand, was probably somewhat modi-fied by mafic (minette) melt, and/or the sourcefor it was quite mafic.

A post-collisional connection?

Eklund et al. (1998) described a series of small1.8 Ga post-collisional, bimodal intrusions insouthern Finland and Russian Karelia (Fig. 4).The rocks of these intrusions range fromultramafic, calc-alkaline potassium lamprophyresto peraluminous high-BaSr granites, and show ashoshonitic affinity with K2O + Na2O > 5%, K2O/Na2O > 0.5, and Al2O3 > 9% over a wide spectrumof SiO2. The mafic rocks of the series are enrichedin P, F, Ba, Sr, and LREE, with depletions notedfor HFS elements Ti, Nb, and Ta (also in thefelsic rocks). Eklund et al. (1998) proposedenriched (metasomatized) lithospheric mantlewith little or no crustal input as the source forthe post-collisional intrusions, and further

suggested that the influx of enriched lithosphericmantle material into the crust continued as lateas ~1615 Ma ago (Eklund and Shebanov, 2003).The Obbnäs granite resembles these post-collisional, ‘shoshonitic’ granites, especially inhaving the high Ba and Sr (see Figs. 1 and 2),and it is not completely impossible that thepluton might have tapped a source enriched inthese elements by the shoshonitic event. Themafic rocks associated with the Obbnäs graniteare, however, tholeiitic instead of calc-alkaline.The features seen in the post-collisionalintrusions are also observed in the minette-granite association of Jack Creek. The minettesdo not show the extreme trace elementenrichment reported by Eklund et al. (1998) forthe mafic rocks of the association in RussianKarelia. They are, nevertheless, clearly moreenriched (up to 1.7 wt.% P2O5, 7600 ppm F, 3720ppm Ba, 540 ppm Sr, 283 ppm La, and 630 ppmCe) than the other mafic rocks in the Redrockarea, and the Jack Creek pluton can, perhaps, belikened to the less enriched post-collisionalintrusions of southwestern Finland, where themetasomatizing fluid was poorer in CO2 and F.

Colliding continents and ripping crust:On tectonics

The rapakivi granites of southern Finland andadjacent Russia are found along a roughly E-Wtrending belt extending from the eastern side ofLake Ladoga, Russia, to the Åland archipelagoin southwestern Finland (Fig. 4a). Tectonicevents that preceded the rapakivi event include(1) accretion of the early Proterozoic arc complexof western Finland (AWF) to the Archaean cratonin the east (syn-orogenic stage at 1.89-1.86 Ga);(2) a second collision from the south (the arccomplex of southern Finland, ASF) against thepreviously accreted complex leading to tectonicthickening of the crust, remelting, and formationof the S-type granites of southern Finland (late-orogenic stage at 1.84-1.80 Ga); (3) intrusion ofsmall, post-collisional, 1.8 Ga granite-monzonite-lamprophyre plutons of southern Finland,probably preceded by mantle metasomatism andenrichment in the region some 40-50 m.y. afterthe peak of the regional metamorphism (post-

16

orogenic stage at 1.81-1.77 Ga) (e.g., Eklund etal., 1998; Haapala & Rämö, 1992; Nironen, 1997,2003). The rapakivi granites and associated maficrocks intruded 120-230 m.y. later in extensionaltectonic regime. This is rather convincinglydemonstrated by the extensive, subparallelswarms of tholeiitic diabase dikes associatedwith the rapakivi granites (e.g., Haapala & Rämö,1992), as well as by recent deep seismicsoundings that have shown the rapakivi granitesand the dike swarms to inhabit a relatively thincrust (Luosto, 1991; Korja, 1991). Seismic studiesalso indicate the existence of a mafic underplateand a body of anorthosite and gabbro below the~10 km thick Wiborg batholith (Korja, 1991).

Similar ‘stages’ are found in southernLaurentia, and are recorded in the bedrock ofthe Redrock area. Three Proterozoic crustal prov-inces, Mojave, Yavapai, and Mazatzal (Fig. 4b)amalgamated onto the Archaean craton in thenorth at 1.8-1.65 Ga (e.g., Condie, 1992). The lowercrust/lithospheric mantle was subsequentlymetasomatized and, at 1.46 Ga, remelted in a pre-sumably extensional tectonic environment, lead-ing to intrusion of a bimodal granite-minette as-sociation (Jack Creek). This high-K (‘shosho-nitic’) intrusion was then followed by the A-type Redrock pluton about 220 m.y. later. Com-pared to southern Finland, the enrichment ofthe lower crust/lithospheric mantle appears,however, to have been of relatively local nature,as other ~1.46 Ga granitic intrusions of south-western United States are typically A-type (e.g.,Van Schmus et al., 1996); in addition to Jack Creek,only the 1.41 Ga Mountain Pass carbonate-sy-enite-shonkinite intrusion in southeastern Cali-fornia (cf. Calzia & Rämö, 2000) shows potassicaffinity.

It is generally accepted that continental col-lisions lead to over-thickening and destabiliza-tion of the lithosphere, which will, eventually,try to return to normal thickness by uplift anderosion (powered by isostatic rebound) and/orby delamination of the lithospheric mantle andorogenic collapse (e.g., England, 1993; Dewey,1998; Marotta et al., 1998) These processes caninitiate lithospheric extension and crustal melt-ing, which, in turn, can lead to mafic underplatingand extension-related magmatism. But is this the

process generating the A-type granites, andwhat is the role of the high-K (‘shoshonitic’)magmatism? Is it just a coincidence that theserocks are found in similar ‘positions’ precedingthe A-type granites both in southern Finlandand in the Redrock area?

Marotta et al. (1998) modeled the thermo-mechanical evolution of the lithosphere under aconvergence regime and reported four majorstages that characterize mantle unrooting(delamination sensu lato): orogenic growth,gravitational instability and mantle breakup,sinking of the detached root, and relaxation ofthe system. The gravitational instability startswhen the lithosphere has reached a thicknessabout twice its initial value (about 50 Ma fromthe onset of the collision) and culminates in thebreakup of the lithospheric root (around 80 Ma).During the sinking of the detached root (from 80to 100 Ma), isostatic rebound produces uplift ofthe crust associated with the replacement oflithospheric mantle by hotter asthenosphere.This, together with cessation of convergence(after 80 Ma), produces a change in the stressregime from compression to extension (after 110Ma). The time span between commencement ofthe orogeny and mantle unrooting dependsmostly on the rheology and, to lesser extent, onconvergent velocity and width of the deforma-tion zone (Marotta et al. assumed viscous rheol-ogy and low convergence rates to derive thetime frame given above).

In southern Finland and adjacent Russia, the‘shoshonitic’ post-collisional plutons form a NE-SW zone relatively close to the boundary be-tween two major Proterozoic crustal provinces(the arc complexes of western and southern Fin-land) and, in the east, the Archaean craton; theyalso seem to lie close to the northern edge of thebelt of the lateorogenic S-type granites of south-ern Finland (Fig. 4a). The lithospheric mantle inthis area was, perhaps, rather extensively af-fected by subduction-related metasomatism dur-ing the collision of these crustal blocks. If theassumption is made that lithospheric unrootingtook place, the 1.83 Ga lateorogenic S-type gran-ites could actually represent crustal melting thattook place during the sinking of the detachedroot and upwelling of the asthenosphere; the

17

Archean crust ( 2.8 Ga)~

Paleoproterozoic crust ( )~1.9 Ga

1.8 Ga post-collisional shoshonitic intrusions

Phanerozoic sedimentary rocks

Gabbro, anorthosite

Rapakivi granite Diabase dikes

Rapakivi complexes (1.65-1.53 Ga)

Russia

Gulf of Finland

LakeLadoga

Åland1576-1568 Ma

Finland

Wiborg1646-1615 Ma

Ahvenisto1643-1632 Ma

60o

24o

BothnianSea

30o

Salmi1547-1530 Ma

Vehmaa1570 Ma

Eurajoki

Reposaari

Siipyy

Kökars-fjärden

1574 Ma

Fjälskär

Obbnäs

Bodom Onas

Suursaari

100 km

Helsinki

VIHTI

The HÄMESWARM

KOPPARNÄS

Porkkala-Mäntsäläshear zone

B

A

200 km

Llano FrontMAZATZAL

YAVAPAI

MOJAVE

ARCHEAN

40O

32O

110O

BurroMountains

MountainPass

102 WO

Peipohja

Suomenniemi1640-1635 Ma

Laitila1570-1540 Ma

ASF

AWF

Exposed Precambrian bedrock

Northern edge of the potassium S-typegranites of southern Finland

Fig. 4. (a) Simplified lithological map of southern Finland and adjacent Russia showing the location of rapakivigranites and the associated mafic rocks (Bodom and Obbnäs plutons are framed) and the 1.8 Ga post-collisional‘shoshonitic’ intrusions. AWF - arc complex of western Finland, ASF - arc complex of southern Finland. (b) A mapshowing the three Proterozoic crustal provinces of Laurentia and the exposed Precambrian bedrock of southwest-ern USA. Frames show the locations of the Burro Mountains and the Mountain Pass carbonate-syenite-shonkiniteintrusion. The dash line marks the location of the transcurrent zone of Bickford et al. (2000). (a) modified afterRämö (1991), Eklund et al. (1998), and Nironen (2003); (b) after Fig. 1 in Paper V.

post-collisional ‘shoshonitic’ intrusions werethen intruded 30 m.y. later, after cessation of thecollision in an (incipient) extensional stress re-gime, and tapped the remnants of the enrichedlithospheric mantle. How long did the extensioncontinue? Was it still going on at 1.65 Ga, or wasthere a plume that reactivated it at that time? Inthe plume scenario, the asthenospheric materialreplacing the detached root could have formeda new lithospheric mantle/lower crust, whichacted as a source for the rapakivi melts causedby plume induced extension and melting (this isin agreement with the ferrodioritic source sug-gested for the Obbnäs pluton, see Paper II).However, direct evidence of a mantle plume un-

der Fennoscandia around the time of rapakivimagmatism is sparse: a possible plume signa-ture is detected in the rapakivi related diabasedikes of Kopparnäs (cf. Paper II).

In southern Laurentia, high-K (‘shosho-nitic’) magmatism preceding A-type granitesseems to be much more rare than in southernFinland. The two known occurrences, the JackCreek granite-minette association in New Mexicoand the Mountain Pass carbonate-syenite-shonkinite intrusion in southeastern California,intruded about 190 m.y. after the end of the col-lision. Both of these intrusions are located closeto a major transcurrent zone proposed byBickford et al. (2000) to have extended along the

18

southern margin of Laurentia from western Texasto southestern California (Fig. 4b). It may be thatthis major crustal structure controlled the localpotassic metasomatism that affected the lithos-pheric mantle sometime between ~1.63 and 1.46Ga (at least in the Redrock area). It may alsohave provided the mechanism for repeated ex-tension, mantle upwelling, and melting, regis-tered by the two granite plutons and the associ-ated mafic rocks of Redrock.

Concluding remarks

- Compositional (including isotopic) dif-ferences between the ~1640 Ma aluminous A-type granite plutons of Bodom and Obbnäs,southern Finland, imply derivation from differ-ent sources: a relatively mafic (ferrodioritic totonalitic) crustal source is needed for the Obbnäspluton, whereas the Bodom pluton requires amore felsic (granodioritic) source. In addition,the composition of the Obbnäs pluton was prob-ably influenced by interaction with a mafic melt.

- Two distinct, intimately juxtaposedmagmatic suites are found in the northern BurroMountains, southwestern New Mexico: the~1460 Ma potassic granite-minette suite of JackCreek and the ~1220 Ma tholeiitic A-type gran-ite-anorthosite suite of Redrock. These, togetherwith a ~1630 Ma tholeiitic diabase, record domi-nantly juvenile additions to the cratonic marginof southern Laurentia and imply subcontinentalenrichment events and repeated mantle meltingin an extensional (or possibly transtensional)environment.

- A set of diabase dikes associated withthe Bodom and Obbnäs granites (the dikes atKopparnäs) may represent a magma type thatcompositionally approaches ocean islandbasalts. This would support a mantle plume as adriving mechanism for the extension and crustalmelting associated with the rapakivi magmatismin southeastern Fennoscandia.

- Overall, the studied suites of southernFennoscandia and Laurentia show that the mid-Proterozoic anorogenic bimodal granitemagmatism so characteristic of this era can bepetrologically outstandingly variable.

ACKNOWLEDGMENTS

I want to express my gratitude to the followingpersons, without whom this work would not ex-ist, at least not in this form. Professor IlmariHaapala suggested the Bodom and Obbnäs gran-ites for the topics of my study and provided hisexpertise on rapakivi granites when needed. Pro-fessor Tapani Rämö took me into ‘the Redrockproject’, gave invaluable advice both at the of-fice and in the field, and remained patient andunderstanding during moments of frustration(there were a few…). Dr. Virginia McLemore (anawesome driver, organizer, and geologist) andDr. Matti Vaasjoki (a great story-teller and U-Pbspecialist) were important and inspiring co-work-ers. Professor Calvin Barnes and Dr. ArtoLuttinen reviewed the thesis manuscript, pro-vided constructive criticism and helpful com-ments, and untied some knots (mostly in mystomach). Dr. Seppo Lahti provided material andshared information about Bodom and Obbnäs,and performed the mineral chemical analyses.Commandant of the Upinniemi Naval Basegranted access to restricted military premises,provided transport to certain islands, and ar-ranged living quarters at the base during field-work on the Obbnäs pluton. Also, thanks to mycolleagues and friends for inspiring, refreshing,and -most of all- entertaining coffee break dis-cussions!

This study was funded by the Academy ofFinland from 1996-2001 (project 36002, led byTapani Rämö), Geological Graduate School, andthe University of Helsinki (grant for finishingdoctoral dissertations), and is a contribution toIGCP Project 426 (Granite Systems and Protero-zoic Lithospheric Processes).

REFERENCES

Anderson, J. L. (1983). Proterozoic anorogenicgranite plutonism of North America. In:Medaris, L. G. Jr., Byers, C. W., Mickelson,D. M. & Shanks, W. C. (Eds.) ProterozoicGeology: Selected papers from an interna-tional Proterozoic symposium. GeologicalSociety of America, Memoir 161, 133-154.

Anderson, J. L. & Cullers, R. L. (1978). Geochem-

19

istry and evolution of the Wolf Riverbatholith, a Late Precambrian rapakivi mas-sif in North Wisconsin, U.S.A. PrecambrianResearch 7, 287-324.

Anderson, J. L. & Morrison, J. (1992). The roleof anorogenic granites in the Proterozoiccrustal development of North America. In:Condie, K. C. (Ed.) Proterozoic crustal evo-lution. Developments in Precambrian ge-ology 10, 263-299.

Andersson, U. (1997). Petrogenesis of some Pro-terozoic granitoid suites and associatedbasic rocks in Sweden (geochemistry andisotope geology). Sveriges GeologiskaUndersökning, Rapporter och Med-delanden 91, 213 pp.

Barker, F., Wones, D. R., Sharp, W. N. &Desborough, G. A. (1975). The Pikes Peakbatholith, Colorado Front Range, and amodel for the origin of the gabbro-an-orthosite-syenite-potassic granite suite.Precambrian Research 2, 97-160.

Bickford, M.E., Soegaard, K., Nielsen, K.C., andMcLelland, J.M. (2000). Geology and geo-chronology of Grenville-age rocks in the VanHorn and Franklin Mountains area, westTexas: Implications for the tectonic evolu-tion of Laurentia during the Grenville. Geo-logical Society of America Bulletin 112,1134-1148.

Bridgwater, D. & Windley, B. F. (1973). An-orthosites, post-orogenic granites, acid vol-canic rocks and crustal development in theNorth Atlantic Shield during the Mid-Prot-erozoic. Geological Society of South Africa,Special Publication 3, 307–317.

Calzia, J. P. & Rämö, O. T. (2000). Late Cenozoiccrustal extension and magmatism, southernDeath Valley region, California. In: Lagasonet al. (Eds.) Great Basin and Sierra Nevada.Geological Society of America Field Guide2, 135-164.

Christiansen, E. H., Burt, D. M., Sheridan, M. F.& Wilson, R. T. (1983). The petrogenesis oftopaz rhyolites from the western UnitedStates. Contributions to Mineralogy andPetrology 83, 16–30.

Clemens, J. D., Holloway, J. R. & White, A. J. R.(1986). Origin of an A-type granite: experi-

mental constraints. American Mineralogist71, 317-324.

Collins, W. J., Beams, S. D., White, A. J. R. &Chappel, B. W. (1982). Nature and origin ofA-type granites with particular reference tosoutheastern Australia. Contributions toMineralogy and Petrology 80, 189-200.

Condie, K. C. (1991). Precambrian granulites andanorogenic granites, are they related? In:Haapala, I. & Condie, K. C. (Eds.) Precam-brian Granitoids - Petrogenesis, Geochem-istry and Metallogeny. Precambrian Re-search 51, 161-172.

Condie, K. C. (1992). Proterozoic terranes andcontinental accretion in southwestern NorthAmerica. In: Condie, K. C. (Ed.) Proterozoiccrustal evolution. Amsterdam, Elsevier, p.447-480.

Creaser, R. A., Price, R. C. & Wormald, R. J. (1991).A-type granites revisited: assessment of aresidual-source model. Geology 19, 163-166.

Dall’Agnol, R., Lafon, J.-M. & Macambira, M. J.B. (1994). Proterozoic anorogenic magmatismin the Central Amazonian Province, Amazo-nian craton: geochronological, petrologicaland geochemical aspects. In: Haapala, I. &Rämö, O. T. (Eds.) IGCP Project 315 Publi-cation No. 12. Mineralogy and Petrology50, 113–138.

Dall’Agnol, R., Rämö, O. T., Magalhães, M. S. &Macambira, M. J. B. (1999). Petrology of theanorogenic, oxidised Jamon and Musa gran-ites, Amazonian craton: implications for thegenesis of Proterozoic A-type granites.Lithos 46, 431–462.

Dewey, J. F. (1988) Extensional collapse oforogens. Tectonics 7, 1123-1139.

Eklund, O. & Shebanov, A. (2003). The prolongedSvecofennian post-collisional shoshoniticmagmatism in the southern part of theSvecofennian domain – a case study of theÅva granite-lamprophyre ring complex. In:Rämö, O.T., Kosunen, P.J., Lauri, L.S., &Karhu, J.A. (Eds.) Granitic Systems – Stateof the Art and Future Avenues. An Interna-tional Symposium in Honor of ProfessorIlmari Haapala, January 12-14, 2003, Ab-stract Volume. Helsinki University Press,41-43.

20

Eklund, O., Fröjdö, S. & Lindberg, B. (1994).Magma mixing: the petrogenetic link be-tween anorthositic suites and rapakivi gran-ites, Åland, SW Finland. Mineralogy andPetrology 50, 3-19.

Eklund O., Konopelko D., Rutanen H., Fröjdö S.,Shebanov A.D. (1998). 1.8 Ga Svecofennianpost-collisional shoshonitic magmatism inthe Fennoscandian shield. Lithos 45, 87-108

Elo, S. & Korja, A. (1993). Geophysical interpre-tation of the crustal and upper mantle struc-ture in the Wiborg rapakivi granite area,southeastern Finland. Precambrian Re-search 64, 273–288.

Emslie, R. F. (1991). Granitoids of rapakivi gran-ite - anorthosite and related associations.Precambrian Research 51, 173-192.

England, P. (1993). Convective removal of ther-mal boundary layer of thickened continen-tal lithosphere: a brief summary of causesand consequences with special reference tothe Cenozoic tectonics of the Tibetan Pla-teau and surrounding regions. Tectono-physics 223, 67-73.

Frost, B. R., Barnes, C. G., Collins, W. J., Argulus,R. J., Ellis, D. J. & Frost, C. D. (2001).geochemical classification for granitic rocks.Journal of Petrology 42, 2033-2048.

Frost, C. D. & Frost, B. R. (1997). Reducedrapakivi-type granites: the tholeiite connec-tion. Geology 25, 647-650.

Frost, C. D., Frost, B. R., Chamberlain, K. R. &Edwards, B. R. (1999). Petrogenesis of the1.43 Ga Sherman Batholith, SE Wyoming,USA: a reduced, rapakivi-type anorogenicgranite. Journal of Petrology 40, 1771-1802.

Haapala, I. (1988). Metallogeny of the Protero-zoic rapakivi granites of Finland. In: Taylor,R. P. & Strong, D. F. (Eds.) Recent advancesin the geology of granite-related mineraldeposits. Canadian Institute of Mining andMetallurgy, Special Volume 39, 124–132.

Haapala, I. & Rämö, O. T. (1990). Petrogenesis ofthe Proterozoic rapakivi granites of Finland.In: Stein, H. J. & Hannah, J. L. (Eds.) Ore-bearing granite systems: Petrogenesis andmineralizing processes. Geological Soci-ety of America, Special Paper 246, 275–

286.Haapala, I. & Rämö, O. T. (1992). Tectonic set-

ting and origin of the Proterozoic rapakivigranites of southeastern Finland. Transac-tions of the Royal Society of Edinburgh,Earth Sciences 83, 165-171.

Hedlund, D.C. (1980). Geologic map of theRedrock NE quadrangle, Grant County, NewMexico. U.S. Geological Survey, Miscella-neous Field Studies Map MF-1264, Scale1:24,000.

King P. L., White A. J. R., Chappell B. W. & AllenC. M. (1997). Characterization and origin ofaluminous A-type granites from the LachlanFold Belt, Southeastern Australia. Journalof Petrology 38, 371-391.

Korja, A. (1991). Crustal and upper mantle struc-ture of the Wiborg batholith, SE Finland.Institute of Seismology, University ofHelsinki, Report S-25, 97-101.

Korja, A. & Heikkinen, P. (1995). Proterozoic ex-tensional tectonics of the centralFennoscandian Shield: results from the Bal-tic and Bothnian Echoes from the Lithos-phere experiment. Tectonics 14, 504–517.

Luosto, U. (1991). Moho depth maps of theFennoscandian shield based on seismic re-fraction data. Institute of Seismology, Uni-versity of Helsinki, Report S-25, 43-49.

Maniar, P. D. & Piccoli, P. M. (1989). Tectonicdiscrimination of granitoids. GeologicalSociety of America, Bulletin 101, 635-643.

Marotta, A. M., Fernàndez, M.& Sabadini, R.(1998). Mantle unrooting in collisional set-tings. Tectonophysics 296, 31-46.

Nironen, M. (1997). The Svecofennian Orogen:a tectonic model. Precambrian Research 86,21-44.

Nironen, M. (2003). Renewed classification ofProterozoic orogenic granitoid rocks in Fin-land. In: Rämö, O.T., Kosunen, P.J., Lauri,L.S., & Karhu, J.A. (Eds.) Granitic Systems– State of the Art and Future Avenues. AnInternational Symposium in Honor of Pro-fessor Ilmari Haapala, January 12-14,2003, Abstract Volume. Helsinki Univer-sity Press, 77-78.

Patiño Douce, A. E. (1997). Generation ofmetaluminous A-type granites by low-pres-

21

sure melting of calc-alkaline granitoids. Ge-ology 25, 743-746.

Rämö, O. T. (1991). Petrogenesis of the Protero-zoic rapakivi granites and related basic rocksof southeastern Fennoscandia: Nd and Pbisotopic and general geochemical con-straints. Geological Survey of Finland,Bulletin 355, 161 p.

Rämö, O. T. & Haapala, I. (1995). One hundredyears of Rapakivi Granite. Mineralogy andPetrology 52, 129-185.

Rämö, O. T., Dall’Agnol, R., Macambira, M. J. B.,Leite, A. A. S. & de Oliveira D. C. (2002). 1.88Ga oxidized A-type granites of the Rio Mariaregion, eastern Amazonian craton, Brazil:positively anorogenic! Journal of Geology110, 603-610.

Shannon, W., Barnes, C. & Bickford, M. (1997).Grenville magmatism in west Texas: petrol-ogy and geochemistry of the Red Bluff gra-nitic suite. Journal of Petrology 38, 1279–1305.

Smith, D. R., Noblett, J., Wobus, R. A., Unruh,D., Douglass, J., Beane, R., Davis, C.,Goldman, S., Kay, G., Gustavson, B., Saltoun,B. & Stewart, J. (1999). Petrology andgeochemistry of late-stage intrusions of theA-type, Mid-Proterozoic Pikes Peakbatholith (Central Colorado, USA): implica-tions for petrogenetic models. PrecambrianResearch 98, 271-305.

Van Schmus, W. R., Bickford, M. E. & Turek, A.(1996). Proterozoic geology of the east-cen-tral Midcontinent basement. In: van derPluijm, P. A. & Catacosinos, P. A. (Eds.) Base-ment and basins of eastern North America.Geological Society of America SpecialPaper 308, 7-32.

Weaver, B. & Tarney, J. (1984). Empirical ap-proach to estimating the composition of thecontinental crust. Nature 310, 575-575

Whalen, J. B., Currie, K. C. & Chapell, B. W.(1987). A-type granites: geochemical char-acteristics, discrimination and petrogenesis.Contributions to Mineralogy and Petrol-ogy 95, 407-419.