sign in sign up
<<
Home , Eclogite, Granulite

J. metamorphic Geol., 2003, 21, 65–80

Deep in the Heart of Dixie: Pre-Alleghanian and HP in the Carolina , South Carolina, USA

J. W. SHERVAIS,1 A. J. DENNIS,2 J. J. MCGEE3 AND D. SECOR3 1Department of Geology, Utah State University, Logan UT, 84322, USA ([email protected]) 2Department of Geology and Biology, University of South Carolina, Aiken SC 29801, USA 3Department of Geological Sciences, University of South Carolina, Columbia SC, 29208, USA

ABSTRACT The central part of the Carolina terrane in western South Carolina comprises a 30 to 40 km wide zone of high grade that are distinct from facies metavolcanic rocks of the Carolina belt (to the SE) and facies metavolcanic and metaplutonic rocks of the Charlotte belt (to the NW). This region, termed the Silverstreet domain, is characterized by penetratively deformed gneisses, granitic gneisses, and . Mineral assemblages and textures suggest that these rocks formed under high-pressure metamorphic conditions, ranging from eclogite facies through high-P granulite to upper amphibolite facies. Mafic rocks occur as amphibolite dykes, as metre-scale blocks of coarse-grained -clinopyroxene amphibolite in felsic , and as residual boulders in deeply weathered felsic gneiss. Inferred has been replaced by a vermicular symplectite of sodic in , consistent with decompression at moderate to high temperatures and a change from eclogite to granulite facies con- ditions. All samples have been partially or wholly retrograded to amphibolite assemblages. We infer the following P-T-t history: (1) eclogite facies P-T conditions at ‡ 1.4 GPa, 650–730 °C (2) high-P granulite facies P-T conditions at 1.2–1.5 GPa, 700–800 °C (3) retrograde amphibolite facies P-T conditions at 0.9–1.2 GPa and 720–660 °C. This metamorphic evolution must predate intrusion of the 415 Ma Newberry and must postdate formation of the Charlotte belt and Slate belt arcs (620 to 550 Ma). Comparison with other medium temperature and high pressure suggests that these assemblages are most likely to form during collisional orogenesis. Eclogite and high-P granulite facies metamorphism in the Silverstreet domain may coincide with a 570–535 Ma event documented in the western Charlotte belt or to a late Ordovician-early Silurian event. The occurrence of these high-P assemblages within the Carolina terrane implies that, prior to this event, the western Carolina terrane (Charlotte belt) and the eastern Carolina terrane (Carolina Slate belt) formed separate . The collisional event represented by these high-pressure assemblages implies amalgamation of these formerly separate terranes into a single composite terrane prior to its accretion to Laurentia. Key words: amphibolite; Carolina terrane; southern Appalachians; eclogite; HP granulite.

during collision (e.g. Carswell, 1990; O’Brien & Ro¨ t- INTRODUCTION zler, 2003). In many areas, these rocks are commonly High-pressure granulites, characterized by the ortho- associated with retrogressed felsic gneisses that were -free assemblage Grt + Cpx + Pl ± Qtz, originally cofacial with the enclosed eclogites (e.g. comprise a newly recognized subfacies transitional Cuthbert & Carswell, 1990; Cuthbert et al., 2000; between plagioclase-free eclogites and orthopyroxene- O’Brien et al., 1990). bearing granulites (Pattison, 2003). O’Brien & Ro¨ tzler, The eastern margin of North America in the 2003) distinguished two varieties of high-P granulite: southern and central Appalachians comprises a tec- ultra-high temperature assemblages with melt reaction tonic collage of terranes that formed in exotic locations textures, and medium-T, high-P assemblages (700– during the late Neoproterozoic through early Palaeo- 850 °C, 1.0–1.4 GPa) that overprint former eclogite zoic, and were subsequently accreted to Laurentia facies assemblages. Like medium temperature (MT) during the mid- to late Palaeozoic (Williams & eclogites (Carswell, 1990), the medium-T, high-P Hatcher, 1983; Secor et al., 1983; Horton et al., 1989, granulite subfacies is typically associated with colli- 1991; van Staal et al., 1998). These exotic terranes sional orogens, which form in tectonically thickened evolved independently of Laurentia for much of their arc or continental , typically in response to the existence, and preserve evidence of orogenic and attempted of an arc or continental margin magmatic events that are not observed in Laurentia.

Ó Blackwell Science Inc., 0263-4929/03/$15.00 65 Journal of Metamorphic Geology, Volume 21, Number 1, 2003 66 J. W. SHERVAIS ET AL.

Fig. 1. Regional geology of the southern Appalachians, showing principal sub-divi- sions, including the Carolina terrane (pale grey), the Blue Ridge terrane (dark grey), the Inner Piedmont terrane, and the Atlantic Coastal Plain. Rocks of the Inner Piedmont terrane (including the Chauga belt) and the Carolina terrane (including the Charlotte belt, the Carolina Slate belt, and the Kings Mountain belt [KMB]) are all exotic to North America.

One of the most extensive of these exotic peri- The Carolina terrane has been divided into three belts with dif- Gondwana terranes is the Carolina terrane, which ferent metamorphic and petrological characteristics: (1) the Kings Mountain belt, which consists of greenschist facies mafic metavol- comprises a large portion of the southern Appalachian canic rocks and forms the north-western margin of the Carolina orogen east of the Blue Ridge province (Secor et al., terrane; (2) the Charlotte belt, which consists largely of lower to 1983; Fig. 1). The Carolina terrane is an exotic middle amphibolite facies, dominantly mafic metavolcanic and meta- Avalonian terrane that originally formed adjacent to plutonic rocks; and (3) the Carolina Slate belt, which is dominated by Gondwana in the late Neoproterozoic, and was not low-grade (greenschist to subgreenschist) felsic metavolcanic rocks with subordinate mafic lavas and mudstones (Fig. 1). accreted to Laurentia until the mid- to late Palaeozoic The Carolina terrane was metamorphosed and ductilely deformed (Secor et al., 1983; Williams & Hatcher, 1983). during the latest Neoproterozoic to early Cambrian (Dennis & We have recently re-examined a little known Wright, 1995, 1997; Hibbard & Samson, 1995; Barker et al., 1998). occurrence of high-P granulite and amphibolite, with Metamorphism and ductile deformation resulting from the Alle- ghanian (320 Ma) collision of Laurentia and Gondwana is an inferred MT eclogite precursor, within the central restricted to narrow shear zones which separate broad zones con- part of the Carolina terrane (Dennis et al., 2000). taining older fabric and mineral assemblages (e.g. Secor et al., 1986; These rocks, which were originally interpreted as Dallmeyer et al., 1986; Horton et al., 1989; Horton & Dicken, 2001). pyroxene-bearing garnet amphibolites, contain relict The Charlotte belt was intruded by a suite of undeformed Devonian garnet-pyroxene-plagioclase assemblages that record a and granitoids (400 Ma; McSween et al., 1991) that cross- cut regional foliation and mark the upper age limit of penetrative previously unrecognized episode of eclogite trans- deformation within most of the terrane. itional to medium temperature HP granulite facies The exotic nature of the Carolina terrane is shown clearly by the metamorphism within the Carolina arc terrane. This occurrence of a diverse Middle Cambrian peri-Gondwanan trilobite event has broad implications for the evolution of the fauna in the Carolina Slate belt (Samson et al., 1990). In addition, combined field-geochronological studies have shown that metamor- southern Appalachians, and for models of metamor- phic fabric in most of the Carolina terrane formed prior to 535 Ma, phism and exhumation in accreted arc terranes approximately coeval with the rift-drift transition on the Laurentian in general. We present here a first look at these margin (Dennis & Wright, 1995, 1997; Hibbard & Samson, 1995; newly discovered high pressure rocks, their inferred Barker et al., 1998). P–T–t history and some tectonic implications of their occurrence. Field Occurrence of High-Pressure Rocks The boundary between the Charlotte belt and the Slate belt in central ECLOGITES AND GRANULITES OF THE South Carolina comprises a 30-km wide zone of high grade gneisses CAROLINA TERRANE that are distinct from less highly deformed amphibolite facies (dominantly) mafic rocks of the Charlotte belt (to the northwest) and low-grade felsic metavolcanic rocks of the slate belt (to the south- Regional Setting east; Fig. 2). This region, termed the Silverstreet domain, consists of The Carolina terrane in the southern Appalachians is a calc-alkaline high-grade felsic biotite gneisses, granitic gneisses, and amphibolites island arc that is exotic to Laurentia and does not share a common that form the SE margin of the Charlotte belt (Secor et al., 1982, history with North America until the late Palaeozoic Alleghanian 1988; Halik, 1983; Hauck, 1984). The Silverstreet domain is intruded orogeny (Fig. 1). It is largely Neoproterozoic in age but includes by the undeformed early Devonian Newberry granite (415 ± 9 Ma; sections of early to middle Cambrian age (Secor et al., 1983; Samson Fullagar, 1981; Samson & Secor, 2000), which cross-cuts regional et al., 1990; Shervais et al., 1996; Dennis & Shervais, 1996; Wortman foliation and includes of sheared and foliated country . et al., 2000). Recent field and geochronological studies show The age of this pluton represents an uppermost age limit for that the Carolina terrane formed during two major episodes of arc formation and deformation of rocks in the Silverstreet domain. The Silverstreet domain is bounded to the north and south by magmatism at 620 Ma and 550 Ma (Dennis & Wright, 1997; Heatherington et al., 1996). shear zones. A variety of field studies have demonstrated that the DEEP IN THE HEART OF DIXIE 67

34¡30'N High grade Silverstreet domain: Clinton eclogite, high-P granulite and ont ne enclosing felsic gneisses Piedm shear zo er ek 81¡W nn E Cre c. 414 ± 8 Ma (U-Pb z) Newberry granite nt: I verE tra Bea E een + + c. 295 ± 4 Ma (Rb-Sr w.r.) e R + mir + + Winnsboro granite hit Clinton Joanna W ? Newberry NW Whitmire S Blair Salem X-roads Little Mountain metatonalite orthogneiss

E E known eclogite - high P granulite localities Wateree L. 7.5' quadrangles indicated and named Newberry granite Winnsboro granite in SE corner. This report focussed on 414 ± 8 Ma U-Pb z. 295 ± 2 Rb-Sr w.r. quadrangles shown in bold.

Cross Hill Bush River Newberry W Newberry E Pomaria Jenkinsville area of Secor and others (1982) L. Greenwood E investigation outlined in dash E E neiss orthog y Hill Mesozoic brittle faults Stone Orientation of some major structures Little Mtn in Carolina slate belt and Kiokee belt metatonalite 550 ± 4 Ma U-Pb z. Dyson Chappells Silverstreet Prosperity Little Mtn Chapin

t bel Lake Murray d Kiokee Carolina slate belt an

82¡W Columbia 34¡N Good Hope Saluda N Denny Delmar

Fig. 2. Geological map showing location of eclogite ⁄ high pressure granulite-bearing Silverstreet domain of the Charlotte belt relative to the Carolina Slate belt and the Whitmire reentrant of the Inner Piedmont. Capital ÔEÕ shows location of known eclogite ⁄ granulite blocks. boundary between the high-grade gneisses of the Silverstreet domain and the Carolina Slate belt is a fault over much of its length, but its geometry and kinematics are not known (e.g. Dennis et al., 2000; Offield, 1995; Offield & Sutphin, 2000; Secor et al., 1988, 1989). It is inferred to be a normal fault in part because it juxtaposes high-grade Charlotte belt rocks against low-grade Slate belt rocks. Locally, however, it can be demonstrated that most recent ductile motion along the Stony Hill orthogneiss was right lateral, based on com- posite planar fabric and asymmetric porphyroclasts (Dennis et al., 2000). The northern margin of the Silverstreet domain is a 10-km wide, E-W trending ductile shear zone (Beaver Creek shear zone) with dextral shear sense indicators that separates it from less deformed rocks of the Charlotte belt (West, 1998). Mafic rocks in less deformed parts of the Silverstreet domain form amphibolite dykes up to 20 cm thick that are oriented parallel to regional foliaton. In more deformed areas, isolated metre-scale blocks within felsic gneiss are interpreted to represent boudinaged mafic dykes (Fig. 3). In many areas, these blocks form residual boulders that have weathered out of the felsic gneisses; where these Fig. 3. Field photo of eclogite block in felsic gneiss of Beaver occur in flat upland terrain they are interpreted to be approximately Creek shear zone. Block is about 1.5 m across, with a trapezoidal in place. shape; foliation in the shear zone wraps around the block. This Retrogressed eclogite and high-P granulite assemblages commonly and other blocks were sampled using portable core drill. are preserved in the cores of these isolated blocks. Blocks with relict high-pressure assemblages are found within the Beaver Creek shear zone along the north side of the Silverstreet domain and as residual boulders near the centre of the terrane, south of the Newberry granite, and clearly outside of the Beaver Creek shear zone (Fig. 2). and minor . forms irregular veins and patches. Epi- Foliation in the shear zone wraps around the eclogite blocks and dote, , plagioclase, and oxides are also found as clearly postdates eclogite formation. inclusions in garnet, with epidote being the dominant inclusion phase. Representative electron microprobe analyses from one sample are presented in Table 1; these data are presented graphically in Mineralogy and Mineral Chemistry Fig. 5. Analytical methods are presented in Appendix A: Methods. Where it has been well preserved, diopside is characterized by a Mafic rocks that preserve high-pressure assemblages are modally vermicular symplectite of sodic plagioclase (An15)22) that we infer diverse with 20–40% pink garnet, 20–60% green, diopsidic clino- represents the breakdown of omphacite; this is clearly shown by both pyroxene, 15–45% hornblende, up to 10% plagioclase, and 3–5% BSE images and high-resolution X-ray composition maps of the ilmenite, with accessory , epidote, apatite, , and symplectites (Fig. 6). The diopside contains about 15% jadeite calcite (e.g. Libby & Carpenter, 1969). Relict garnet and pyroxene component, but modal reconstruction (see Appendix A: Methods) grains are up to 1 cm diameter, but typical grain size for relict phases suggests that primary omphacite contained 30% jadeite. The is 1–3 mm (Fig. 4). In , hornblende, plagioclase and reconstructed omphacite is presented in Table 1. ilmenite replace clinopyroxene, while garnet is replaced along its The breakdown of omphacite to diopside + plagioclase sym- margins by kelyphitic intergrowths of plagioclase with hornblende plectite is commonly observed in high-P granulites after a medium 68 J. W. SHERVAIS ET AL.

Fig. 4. Probe mount (2.5 cm diameter) of eclogite ⁄ granulite from central Carolina terrane, sample NEW-1–3. Pink ¼ garnet, pale green ¼ pyroxene and pyroxene-plagio- clase symplectite, brown ⁄ dark green ¼ hornblende, clear ¼ plagioclase or calcite, black ¼ ilmentite or Fe-oxides. Note plagioclase-rich kelyphite rims on garnet. temperature eclogite assemblage, and is consistent with decompres- inner zone (0 to 55% of radius) that is low in Grs and Prp, and high sion at moderate to high temperatures (e.g. Elvevold & Gilotti, 2000; in Alm and Sps, and (b) an outer zone (60 to 100%) that is higher in O’Brien & Ro¨ tzler, 2003). The inferred primary assemblage Grs and Prp, and lower in Alm and Sps (Fig. 9). All grains exhibit a omphacite + garnet ± rutile is consistent with formation under sharp increase in Grs and decrease in Alm at the transition (c. 60% of eclogite facies conditions, whereas the observed assemblage diopside grain radius) that implies an abrupt change in growth history. + plagioclase + garnet represents high pressure granulite facies conditions (Galan & Marcos, 2000; Cooke et al., 2000; Pattison, 2003; O’Brien & Ro¨ tzler, 2003). The breakdown of diopside and RESULTS garnet to form + plagioclase + epidote + ilmenite represents final equilibration under amphibolite facies conditions. Geothermobarometry Formation of amphibolite occurred in two stages. The first is rep- resented by aluminous pargasite and relatively calcic plagioclase We infer from the data presented above that the garnet (An26)53), which replace diopside + sodic plagioclase symplectites; cores formed during prograde metamorphism at the second is represented by magnesian pargasite and more sodic greenschist or amphibolite facies conditions, followed plagioclase (An17)22) which replace both diopside and garnet. Cal- culated hornblende-plagioclase temperatures (next section) suggest by growth of the garnet mantles at eclogite (Grt- that the aluminous pargasite-calcic plagioclase pairs formed at higher Omp-Rt) and then high-P granulite facies conditions temperatures than the magnesian pargasite-sodic plagioclase pairs, (Grt-Di-Hbl-Pl-Ilm); retrograde metamorphism in the and that they are closely associated with the high-P granulite assemblages. amphibolite facies resulted in the breakdown of garnet High resolution X-ray composition maps of garnet show two and formation of the late Hbl-Pl-Ilm assemblage. distinct growth zones (Fig. 7). The inner zone is enriched in Mn and Because these rocks experienced a range of metamor- Fe, the outer zone is enriched in Ca and Mg. The Ca and Fe X-ray phic conditions, a number of assumptions are made in maps show a sharp interface between the inner core and the outer determining which compositions to use for thermo- , while Mg and Mn show smooth, continuous zoning profiles (Fig. 7). Note that these garnet are generally not symmetrically barometry. zoned: the centre of growth typically lies close to one edge of the We assume that the reconstructed omphacite was in grain. In the example shown here, several small spessartine-rich equilibrium with the more Mg-Ca-rich mantles of the garnet cores (seen as high Mn spots in the X-ray maps) have been garnet, and that the diopside-plagioclase symplectites subsumed by the garnet mantle as it grew. Compositional profiles selected to traverse from the true core to were in equilibrium with the more Mg-Ca-rich, rim confirm these trends. A 1500-lm traverse of the grain mapped in Fe-poor outermost rims of the garnet (e.g. Fig. 9). For Fig. 7 shows smooth profiles for (Prp) and spessartine (Sps), purposes of calculation, three garnet mantle composi- and sharp steps in profiles for grossular (Grs) and almandine (Alm; tions were used: (a) an average of all garnet mantles Fig. 8). Profiles for three additional garnet are shown in Fig. 9, from the profile shown in Table 1a (b) an average of all scaled to percentage of total grain radius. Although these three grains vary somewhat in their innermost core compositions (Grs, garnet mantles in Fig. 9, from 65 to 99% of grain Sps), they display consistent profiles for all elements, with (a) an radius, and (c) the garnet mantle farthest from the Table 1a. Garnet analyses, profile of single large garnet in sample NEW-1. Garnet formulae per 12 oxygen.

Distance from Core Core Core Core Core Core Core Core Core Core Core Core Core Mantle Mantle Mantle Mantle Mantle Centre lm 1 51 204 255 306 508 559 610 762 813 863 914 1016 1117 1168 1270 1371 1472

SiO2 37.54 37.48 37.64 37.67 37.35 37.42 36.94 36.89 37.28 37.27 37.10 37.54 37.39 37.21 37.24 37.81 37.81 37.86 TiO2 0.11 0.10 0.09 0.12 0.12 0.10 0.10 0.07 0.13 0.10 0.12 0.07 0.11 0.07 0.17 0.08 0.09 0.17 Al2O3 21.37 21.40 21.59 21.37 21.54 21.71 21.46 21.44 21.32 21.25 21.30 21.30 21.27 21.46 21.27 21.49 21.46 21.31 FeO 29.37 28.82 28.86 29.07 29.00 28.93 28.88 28.82 28.97 28.93 29.05 29.17 28.46 27.69 27.27 27.20 26.68 26.53 MnO 1.60 1.49 1.27 1.14 1.07 0.81 0.75 0.68 0.47 0.37 0.38 0.32 0.20 0.12 0.12 0.13 0.10 0.12 MgO 2.41 2.33 2.36 2.39 2.40 2.38 2.39 2.48 2.62 2.68 2.76 2.93 3.09 3.34 3.25 3.33 3.39 3.47 CaO 8.62 9.28 9.34 9.41 9.44 9.73 9.87 9.92 9.88 9.84 9.79 9.64 9.90 10.38 10.73 10.71 11.10 11.10 Na2O 0.03 0.03 0.03 0.03 0.04 0.03 0.03 0.00 0.01 0.02 0.04 0.00 0.02 0.05 0.06 0.04 0.02 0.04 K2O 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Cr2O3 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.03 0.01 0.00 Sum 101.03 100.94 101.19 101.20 100.94 101.11 100.42 100.30 100.68 100.46 100.54 100.98 100.43 100.31 100.12 100.81 100.65 100.59 Si 2.968 2.963 2.965 2.969 2.952 2.949 2.937 2.936 2.951 2.956 2.943 2.959 2.957 2.941 2.947 2.964 2.964 2.969 Ti 0.006 0.006 0.005 0.007 0.007 0.006 0.006 0.004 0.007 0.006 0.007 0.004 0.007 0.004 0.010 0.005 0.005 0.010 Al 1.991 1.995 2.004 1.985 2.006 2.017 2.011 2.011 1.989 1.986 1.991 1.979 1.982 1.998 1.984 1.985 1.983 1.970 Fe2+ 1.942 1.906 1.901 1.916 1.917 1.907 1.920 1.918 1.918 1.918 1.927 1.923 1.882 1.830 1.805 1.783 1.750 1.740 Mn 0.107 0.100 0.085 0.076 0.071 0.054 0.051 0.046 0.031 0.025 0.025 0.021 0.013 0.008 0.008 0.009 0.006 0.008 Mg 0.284 0.275 0.277 0.281 0.282 0.280 0.283 0.294 0.309 0.317 0.326 0.345 0.365 0.394 0.383 0.389 0.396 0.405 Ca 0.730 0.786 0.789 0.795 0.799 0.821 0.841 0.845 0.838 0.836 0.832 0.814 0.839 0.878 0.910 0.899 0.932 0.933 Na 0.004 0.005 0.005 0.004 0.006 0.004 0.005 0.000 0.002 0.003 0.005 0.000 0.003 0.007 0.010 0.006 0.003 0.005 K 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Cr 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.002 0.001 0.000 Pyrope 9.3 9.0 9.1 9.2 9.2 9.1 9.1 9.5 10.0 10.2 10.5 11.1 11.8 12.7 12.3 12.6 12.8 13.1 Almandine 63.4 62.1 62.3 62.5 62.5 62.3 62.0 61.8 62.0 62.0 62.0 62.0 60.7 58.8 58.1 57.9 56.7 56.4 Spessartine 3.49 3.26 2.79 2.48 2.31 1.76 1.65 1.48 1.00 0.81 0.80 0.68 0.42 0.26 0.26 0.29 0.19 0.26 Grossular 23.8 25.6 25.9 25.9 26.0 26.8 27.2 27.2 27.1 27.0 26.8 26.2 27.1 28.2 29.3 29.2 30.2 30.2

Table 1b. Pyroxene (6 oxygen), hornblende (23 oxygen), and (8 oxygen) analyses from eclogite ⁄ granulite sample NEW-1.

Di Di Omp Hbl adj Hbl adj actinolite Pl in Di Pl in Di Pl in Di Pl in Di Pl in Di Pl in Di Pl in Di Pl in Di Pl in Di Interstitial Pl Interstitial Pl Interstitial Pl Interstitial Pl Interstitial Pl Interstitial Pl

SiO2 50.64 51.66 51.50 42.17 41.52 48.42 62.16 63.60 64.75 63.57 63.85 63.78 64.11 64.65 63.81 59.76 58.52 60.83 60.06 58.82 54.44 TiO2 0.32 0.18 0.26 1.11 1.11 0.73 0.51 0.02 0.00 0.02 0.00 0.02 0.01 0.03 0.08 0.04 0.06 0.00 0.00 0.00 0.00 Al2O3 4.2 3.47 9.36 11.95 12.72 6.20 22.55 22.36 22.41 23.09 22.50 23.06 22.68 22.41 22.63 25.27 26.32 25.55 25.19 26.31 28.75 FeO 11.85 11.03 9.61 18.90 18.87 15.63 0.33 0.23 0.25 0.38 0.25 0.26 0.30 0.24 0.28 0.18 0.21 0.26 0.09 0.16 0.28 MnO 0.1 0.07 0.08 0.07 0.08 0.06 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 MgO 11.45 10.69 9.21 9.46 9.34 13.01 0.01 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

CaO 19.45 20.27 16.51 11.53 11.26 11.89 4.25 3.19 3.14 3.63 3.23 3.77 3.29 2.94 3.43 6.46 7.51 5.81 5.89 7.35 10.34 DIXIE OF HEART THE IN DEEP Na2O 1.42 1.74 3.47 1.93 2.08 1.01 8.24 9.10 9.00 8.87 9.06 8.83 8.89 9.29 8.91 7.39 6.87 6.92 6.75 6.42 5.22 K2O 0.05 0.01 0.07 0.36 0.43 0.19 0.08 0.09 0.12 0.11 0.07 0.09 0.13 0.10 0.10 0.03 0.05 0.07 0.04 0.04 0.01 Cr2O3 0.01 0.00 0.01 0.00 0.01 0.02 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Sum 99.43 99.10 100.00 97.48 97.41 97.14 98.14 98.61 99.66 99.67 98.97 99.80 99.43 99.65 99.24 99.13 99.54 99.43 98.02 99.09 99.04 Si 1.917 1.958 1.937 6.400 6.311 7.172 2.796 2.838 2.854 2.812 2.838 2.815 2.836 2.851 2.830 2.678 2.621 2.703 2.705 2.638 2.473 Ti 0.009 0.005 0.007 0.127 0.127 0.081 0.017 0.001 0.000 0.001 0.000 0.001 0.001 0.001 0.003 0.001 0.002 0.000 0.000 0.000 0.000 Al 0.187 0.155 0.415 2.137 2.278 1.083 1.196 1.176 1.165 1.203 1.179 1.200 1.183 1.165 1.183 1.335 1.390 1.338 1.337 1.390 1.539 Fe2+ 0.375 0.350 0.302 2.399 2.398 1.936 0.013 0.009 0.009 0.014 0.009 0.009 0.011 0.009 0.010 0.007 0.008 0.010 0.004 0.006 0.011 Mn 0.003 0.002 0.003 0.009 0.010 0.008 0.000 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Mg 0.646 0.604 0.516 2.140 2.116 2.871 0.001 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Ca 0.789 0.823 0.665 1.874 1.834 1.887 0.205 0.153 0.148 0.172 0.154 0.179 0.156 0.139 0.163 0.311 0.361 0.277 0.284 0.353 0.503 Na 0.1042 0.128 0.253 0.568 0.613 0.290 0.719 0.787 0.769 0.761 0.781 0.755 0.763 0.794 0.766 0.642 0.596 0.597 0.590 0.559 0.460 K 0.002 0.001 0.003 0.070 0.083 0.035 0.005 0.005 0.007 0.006 0.004 0.005 0.007 0.005 0.006 0.002 0.003 0.004 0.002 0.002 0.001 Cr 0.000 0.000 0.000 0.000 0.001 0.003 0.000 0.000 0.000 0.000 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 69 70 J. W. SHERVAIS ET AL.

Table 1c. Hornblende-plagioclase pairs from retrograded eclogite ⁄ granulite sample NEW-1.

Sample# Hb-1a Hb-2a Hb-2b Hb-3a Hb-4a Hb-5b Hb-6a Hb-6c Hb-8 Hb-8a Hb-9 Hb-10 Hb-11 Hb-12

SiO2 39.97 37.91 39.84 42.00 40.88 40.82 40.65 42.44 42.57 43.46 39.38 40.76 42.99 40.97 TiO2 1.28 1.28 1.45 0.87 1.31 0.54 0.88 1.32 1.41 1.33 0.88 0.96 1.11 1.00

Al2O3 13.58 16.39 14.33 11.69 12.54 13.80 13.16 10.33 11.03 10.22 14.98 13.43 11.04 12.97 Cr2O3 0.02 0.01 0.01 0.01 0.02 0.01 0.02 0.02 0.00 0.00 0.00 0.00 0.00 0.00

Fe2O3 6.01 6.19 5.56 6.26 6.30 6.72 6.84 5.77 5.13 5.05 7.02 6.75 5.27 6.52 FeO 12.56 11.83 11.86 12.34 12.97 12.95 13.96 14.21 11.81 11.42 12.28 11.44 12.31 11.93 MnO 0.04 0.05 0.05 0.04 0.03 0.08 0.08 0.06 0.03 0.01 0.05 0.00 0.04 0.02 MgO 9.27 8.64 9.65 10.01 9.42 8.88 8.48 9.50 11.03 11.52 8.74 10.01 10.68 9.87 CaO 11.61 11.72 11.65 11.72 11.55 11.63 11.67 11.42 11.44 11.83 11.54 11.97 11.79 11.85

Na2O 2.09 2.22 2.17 1.85 2.04 2.00 2.03 1.89 1.93 1.75 2.15 1.99 1.80 1.99

K2O 0.55 0.66 0.54 0.38 0.45 0.49 0.46 0.37 0.55 0.36 0.51 0.48 0.42 0.46 H2O 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 2.00 Total 98.98 98.90 99.11 99.17 99.51 99.92 100.23 99.33 98.93 98.95 99.53 99.79 99.45 99.58 Si 6.053 5.749 5.998 6.316 6.162 6.124 6.120 6.419 6.384 6.490 5.935 6.092 6.419 6.145 Ti 0.146 0.146 0.164 0.098 0.149 0.061 0.100 0.151 0.159 0.149 0.100 0.108 0.125 0.112 Al 2.425 2.929 2.544 2.074 2.228 2.441 2.335 1.842 1.951 1.800 2.661 2.367 1.944 2.293 Cr 0.002 0.001 0.002 0.001 0.002 0.001 0.002 0.002 0.000 0.000 0.000 0.000 0.000 0.000 Fe3+ 0.685 0.707 0.630 0.709 0.715 0.759 0.775 0.656 0.579 0.567 0.796 0.760 0.592 0.736 Fe2+ 1.592 1.501 1.493 1.552 1.635 1.625 1.757 1.798 1.481 1.426 1.548 1.430 1.537 1.497 Mn 0.006 0.006 0.007 0.004 0.004 0.010 0.010 0.008 0.004 0.001 0.006 0.000 0.005 0.003 Mg 2.092 1.952 2.165 2.243 2.115 1.987 1.902 2.141 2.465 2.563 1.964 2.229 2.377 2.206 Ca 1.885 1.904 1.879 1.889 1.866 1.869 1.883 1.851 1.839 1.893 1.863 1.917 1.887 1.905 Na 0.614 0.654 0.633 0.540 0.597 0.581 0.592 0.555 0.560 0.507 0.627 0.577 0.522 0.577 K 0.106 0.128 0.103 0.073 0.087 0.093 0.088 0.072 0.105 0.069 0.097 0.092 0.080 0.089 Plagioclase Pl-1a Pl-2a Pl-2b Pl-3a Pl-4a Pl-5b Pl-6a Pl-6c Pl-8 Pl-8a Pl-9 Pl-10 Pl-11 Pl-12

SiO2 64.89 54.44 60.06 64.71 58.82 57.81 58.09 64.30 64.70 65.71 62.40 63.95 65.38 60.83 Al2O3 23.30 28.75 25.19 23.60 26.31 26.85 27.44 23.35 23.46 23.49 24.69 23.88 24.52 25.55 CaO 3.03 10.34 5.89 3.23 7.35 7.99 8.50 3.29 2.75 2.70 4.84 3.53 2.98 5.81

Fe2O3 0.21 0.28 0.09 0.13 0.16 0.22 0.18 0.25 0.13 0.15 0.18 0.22 0.20 0.26

Na2O 7.61 5.22 6.75 7.46 6.42 6.41 6.24 8.68 7.46 7.42 7.14 7.40 7.26 6.92

K2O 0.06 0.01 0.04 0.07 0.04 0.04 0.03 0.09 0.09 0.10 0.06 0.07 0.10 0.07 Total 99.10 99.04 98.02 99.19 99.09 99.31 100.47 99.95 98.56 99.56 99.30 99.04 100.46 99.43 Si 2.855 2.473 2.705 2.845 2.638 2.597 2.582 2.825 2.856 2.869 2.7620 2.822 2.834 2.703 Al 1.209 1.540 1.337 1.223 1.391 1.422 1.437 1.209 1.221 1.209 1.288 1.242 1.253 1.338 Ca 0.143 0.503 0.284 0.152 0.353 0.385 0.405 0.155 0.130 0.126 0.230 0.167 0.139 0.277 Fe3+ 0.008 0.011 0.004 0.005 0.006 0.008 0.007 0.009 0.005 0.005 0.007 0.008 0.007 0.010 Na 0.649 0.460 0.590 0.636 0.559 0.558 0.538 0.739 0.639 0.628 0.613 0.633 0.610 0.597 K 0.004 0.001 0.002 0.004 0.002 0.002 0.001 0.005 0.005 0.005 0.003 0.004 0.006 0.004 Xalbite 0.82 0.48 0.67 0.80 0.61 0.59 0.57 0.82 0.83 0.83 0.73 0.79 0.81 0.68

Wo 50 Diopside centre of the grain in Table 1a (but not the rim), cor- Jadeite responding to the analysis at 1371 lm. For the garnet 600 °C 800 °C rim compositions, we used the garnet rim from 1000 °C Table 1a (at 1472 lm) and the average rims of the three garnet shown in Fig. 9. Garnet-pyroxene tem- B peratures were calculated using the calibrations of Ellis & Green (1979) and Powell (1985). ° Because there is no indication of primary plagioclase 1200 C ° 1000 C in equilibrium with omphacite, only minimum pres- sures are estimated for the inferred eclogite assemblage Reconstructed using the plagioclase in diopside symplectite as a Omphacite proxy, using the diopside-plagioclase-garnet-quartz (Newton & Perkins, 1982; Powell & Holland, 1988; A Interstitial Plagioclase Moecher et al., 1988), and -jadeite-quartz geo- Exsolved Diopside barometers (Holland, 1980). For the high-P granulite Plagioclase Symplectite in Pyroxene assemblage, we used diopside-plagioclase-garnet- DiHd to EnFs quartz (Newton & Perkins, 1982; Powell & Holland, 1988; Moecher et al., 1988) to estimate pressure, and Fig. 5. Mineral data from eclogite ⁄ granulite sample NEW-2: the jadeite content of clinopyroxene geobarometer of (a) Jadeite-high Ca Px-Low Ca Px ternary plot, showing analyzed Carswell & Harley (1990), to give a minimum pressure pyroxene (squares) and reconstructed omphacite (X); (b) pyroxene for the eclogite assemblage. quadrilateral plot showing analyzed pyroxene (squares), along For hornblende-bearing assemblages, the garnet- with temperature contours of Lindsley & Anderson (1983); (c) albite corner of the feldspar ternary, showing compositions hornblende (Graham & Powell, 1984) thermometer and of plagioclase symplectite in diopside (open) and interstitial garnet-plagioclase-hornblende-quartz geobarometer plagioclase associated with hornblende (closed). (Kohn & Spear, 1989, 1990), were used, taking only the Fig. 6. X-ray composition maps of diopside-sodic plagioclase symplectites (¼ former omphacite) surrounded by hornblende, plagioclase, garnet, and calcite. (A) Mg map, New-1; (B) Ca map, New-1; (C) Al map, 3080E (D) Al map, 3080E. (A, B) Field of view ¼ 5 mm, hotter colours equal higher concentrations. Note shapes of the diopside- plagioclase-hornblende aggregates, which seem to pseudomorph the primary omphacite. (C, D) Field of view 2.5 mm. Lighter shades ¼ higher concentrations.

A B C D EPI H ER FDIXIE OF HEART THE IN DEEP

Fig. 7. X-ray maps of zoned garnet surrounded by hornblende with minor plagioclase and ilmenite. A ¼ Fe, B ¼ Mn, C ¼ Mg, D ¼ Ca. Garnet has high Mn and Fe in core, with higher Mg and Ca in mantle. Note the sharp contact between the inner garnet core and the outer garnet mantle seen clearly in the Fe and Ca X-ray maps. Note also the small garnet cores (high Mn and Fe, low Ca and Mg) that have been subsumed by the garnet mantle. Hotter colours equal higher concentration in Fe, Mn, and Mg; darker blue equals higher Ca. Black line in A is approximate location of line profile (Table 1a). Field of view is 5 mm in all maps. 71 72 J. W. SHERVAIS ET AL.

(Fig. 5). The results from this graphical solvus thermo- meter 700–800 °C, or up to 50 °C higher than results from the garnet rim-diopside thermometers (700–750 °C). Garnet-clinopyroxene temperatures were calculated using stoichiometry to partition total Fe between Fe2+ and Fe3+; calculation of Fe3+ stoi- chiometrically is strongly dependent on analytical precision and generally overestimates Fe3+ in pyroxene due to vacancies in the pyroxene lattice (e.g. Robinson, 1980). Calculation of garnet-diopside temperatures assuming total Fe as Fe2+ results in temperatures 50–80 °C higher than those using calculated Fe3+ (Table 2). The c. 50 °C difference between the pyrox- ene solvus temperatures and garnet-diopside tempera- tures calculated here probably results from the overestimation of Fe3+ using stoichiometry. Variations in hornblende-plagioclase temperatures correlate with texture and composition, as noted ear- lier. Aluminous pargasite and calcic plagioclase that replace diopside-plagioclase symplectites formed at higher temperatures than the magnesian pargasite and sodic plagioclase that replace garnet and form the groundmass in highly amphibolitized samples. Fig- ure 10 shows hornblende-plagioclase temperatures as a function of plagioclase composition at 1.2 GPa, which is the mean pressure calculated for these assemblages. The high temperature hornblende-plagioclase assem- blage clearly replaces pre-existing diopside-plagioclase symplectites, but formed at similar temperatures and P pressures, probably in response to increased H2O during the thermal peak; we speculate that this water must have come from dehydration reactions in the enclosing felsic gneisses. Using the data and methods described above, the following equilibration conditions are suggested for each stage of metamorphism, with each diagnostic Fig. 8. Profile of single large garnet grain, from centre to rim. Distance from rim in lm. Note sharp increase in Grs (CaO) and assemblage shown in parentheses: decrease in Alm (FeO) at around 1150 lm. Pyr (MgO) shows (1) Eclogite facies metamorphism of a mafic protolith less precipitous increase, Sps (MnO) decays exponentially (garnet mantle + omphacite + rutile) at 650–730 °C toward rim. and ‡ 1.4 GPa. (2) HP granulite facies conditions during decompres- outermost rim analysis of the garnet in Table 1a and sion (garnet rim + diopside + pargasite I + plagio- selected hornblende-plagioclase pairs from Table 1c. clase) at 700–800 °C and 1.2–1.5 GPa. Hornblende-plagioclase temperatures (Holland & (3) Amphibolite facies conditions (pargasite II + Blundy, 1994) were also calculated using the data from plagioclase + ilmenite ± epidote) at 660 °C to 720 °C Table 1c, which represent adjacent hornblende-plagio- and 0.9–1.2 GPa. clase pairs from a range of textural associations. We These data are summarized in Table 2 and Fig. 11, also used GRIPS (garnet-rutile-ilmenite-plagioclase- which depicts the calculated equilibria for various silica, Bohlen & Liotta, 1986) to estimate pressure, since mineral associations and facies with ellipses that ilmenite is associated with the hornblende-forming overlie the intersection of geothermometer and geo- reactions. barometer equilibria for each stage of metamorphism. Calculation of equilibrium T-P conditions was car- In the absence of primary plagioclase, the pressure ried out using the program GTB of Spear & Kohn estimated for the eclogite assemblage (P ¼ 1.4 GPa) (2001), and the hornblende-plagioclase program of is a minimum equilibration pressure; this may be Holland & Blundy (1994). The pyroxene quadrilateral lowered somewhat based on the high Fe content of the thermometer of Lindsley & Anderson (1983) is used sample, but may also be considerably higher (Carswell for diopside (but not reconstructed omphacite) because & Harley, 1990). The equilibration pressures of the its nonquadrilateral components are less than 20% HP granulite assemblage (1.2–1.5 GPa) are well DEEP IN THE HEART OF DIXIE 73

3.5

2.5

1.5

0.5

Fig. 9. Combined profiles of three garnet grains in sample NEW-1, scaled to percent radius of grain. Note sharp increase in Grs (CaO) and decrease in Alm (FeO) at around 60% of total radius. Pyr (MgO) shows less precipitous increase, Sps (MnO) decays exponentially toward rim.

Table 2. Summary of T-P data for high-P mafic lithology from Silverstreet domain. Numbers in brackets refer to references listed below.

Assemblage Facies Themometer Temperature Barometer Pressure

1 Eclogite Grt-Cpx 650–730 °C @1.4 GPa [1,2] Grt-Cpx-Pl-Qtz > 1.3 GPa @700 °C[6] Jd-Ab-Qtz > 1.5 GPa @700 °C [7,8] > 1.4 Gpa @700 °C[9] 2 High P Granulite Grt-Cpx 700–750 °C @1.2–1.5 GPa [1,2] Grt-Cpx-Pl-Qtz 1.2 GPa @800 °C [6,7] Grt-Cpx 760–780 °C @1.2–1.5 GPa [1,2]* 1.5 GPa @800 °C [7,8] Cpx Solvus 700–800 °C[3] Grt-Hbl 770–820 °C @1.2 GPa [4] Grt-Hbl-Pl-Qtz 1.2–1.5 GPa @800 °C [10] Hbl-Pl 760–830 °C @1.2 GPa [5] GRIPS 1.2–1.5 GPa @800 °C [11] 3 Amphibolite Grt-Hbl 660–775 °C @1.2 GPa [4] Grt-Cpx-Pl-Qtz 1.0–1.2 GPa @700 °C [10] Hbl-Pl 690–740 °C @1.2 GPa [5] GRIPS 1.0–1.2 GPa @675 °C [11]

References [1] Ellis & Green (1979); [2] Powell (1985); [3] Lindsley & Anderson (1983); [4] Graham & Powell (1984); [5] Holland & Blundy (1994); [6] Newton & Perkins (1982); [7] Powell & Holland (1988); [8] Moecher et al. (1988); [9] Holland (1980); [10] Kohn & Spear (1989); [11] Bohlen & Liotta (1986). All temperatures with Fe3+ correction except *. constrained since plagioclase and quartz are present in both assemblages. Protolith of mafic boudins and layers The conditions calculated for these assemblages Field relations suggest that protoliths of the mafic imply a clockwise pressure-temperature-time (P-T-t) boudins and layers were originally mafic dykes intru- path (Fig. 11). The clockwise P-T-t configuration is ded into the more abundant felsic gneisses that com- consistent with models that involve collision of large pose the country rock of the Silverstreet domain. continental or arc blocks, where one block is thrust Whole rock analyses presented here (Table 3) and beneath another (high pressure at relative low tem- elsewhere (Dennis & Shervais, 1991, 1996) show that peratures) and then rebounds to an equilibrium the felsic gneisses were derived from arc-related felsic when the block is exhumed during to intermediate composition metavolcanic and meta- uplift (England & Thompson, 1984). Based on the plutonic rocks of Charlotte belt affinity. preservation of primary zoning profiles in the garnet at Most of the mafic boudins and dykes are basaltic in temperatures up to 800 °C, we suggest that uplift and composition, with SiO2 50%, MgO 6–8%, cooling must have been relatively rapid after peak FeO* 10% and TiO2 1–3% (Fig. 12). These metamorphic conditions were reached (O’Brien, 1997; compositions are typical of oceanic , but are too Cooke et al., 2000). high in TiO2 to represent arc-related high-alumina 74 J. W. SHERVAIS ET AL.

840 basalts. Two samples are somewhat unusual ferro- basalts, with SiO2 41%, MgO 6%, FeO* 820 [email protected] GPa 15–18%, and TiO2 2.5–3.5% (Fig. 12). They are chemically equivalent to evolved tholeiitic basalts, with 800 very low Cr and Ni (<50 p.p.m). Similar ferrobasaltic 780 compositions were reported by Dal Piaz & Lombardo (1986), Galan & Marcos (2000), and Will & Schma¨ - 760 dicke (2001). All of the basalts and ferrobasalts studied here can be classified as mid-ocean ridge basalts or 740 ocean island basalts using geochemical discrimination 720 plots such as Ti-V and Ti-Zr (Fig. 13; Shervais, 1982;

Hbl-Pl Temperatures Pearce & Cann, 1973). This oceanic affinity contrasts 700 with the characteristic arc-related compositions of the surrounding felsic to intermediate gneisses and other 680 rocks of the Charlotte belt (Dennis & Shervais, 1991, 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1996). Similar relationships are observed in the Pro- Albite terozoic Nagssugtoqidan mobile belt of east (Messiga et al., 1990). Fig. 10. Plot of hornblende-plagioclase temperatures at 1.0 GPa pressure, using Holland & Blundy (1994), as a function of plagioclase composition. The strong correlation of calculated DISCUSSION temperature with composition is consistent with observed changes in assemblage and texture. Tectonic setting of high pressure metamorphism The equilibration conditions calculated here imply a clockwise P-T-t path for mafic rocks of the Silverstreet domain, consistent with collision of two crustal blocks 2 during attempted subduction of one of these blocks. Combined P-T path for These conditions are similar to those inferred for other Silverstreet domain medium temperature eclogites and related high-pres- sure granulites (e.g. Carswell & O’Brien, 1993; Galan & Marcos, 2000; Cooke et al., 2000; Will & Schma¨ - 1.6 Eclogite dicke, 2001; O’Brien & Ro¨ tzler, 2003). In many cases, 1 medium temperature eclogites are associated with High-P continent-continent collisions, as seen in the Norwe- 2 Granulite gian and Greenland Caledonides (Krogh, 1982; Cuth- bert & Carswell, 1990; Cuthbert et al., 2000; Elvevold 1.2 & Gilotti, 2000) and the Variscan foldbelts of central Europe (O’Brien et al., 1990; Galan & Marcos, 2000; 3 Cooke et al., 2000; Will & Schma¨ dicke, 2001; O’Brien &Ro¨ tzler, 2003). P GPa Amphibolite 0.8 The Kohistan arc terrane of northern Pakistan may provide the closest analogue to the rocks studied here, because it involves collision of an island arc with a continental margin (Jan & Howie, 1981; Coward et al., 1 = Eclogite Gt-Omp 1987; Khan et al., 1989). High pressure granulites of 2 = Granulite Gt-Cpx-Hb-Pl 0.4 the Jijal complex formed at the base of the Kohistan 3 = Amphibolite Gt-Hb-Pl-Ilm arc, possibly during its amalgamation with Asia (e.g. Coward et al., 1987; Khan et al., 1989). It was exposed during the subsequent collision of India with the com- bined Kohistan-Asia block, during which Kohistan was 0 400 500 600 700 800 900 in the upper plate of the collision (Coward et al., 1987). The c. 450 Ma eclogites of the Eastern Blue Ridge T ˚C province of the southern Appalachians are not related to the MT eclogite ⁄ HP granulites from the Carolina Fig. 11. P-T-t plot showing clockwise path, with isobaric terrane described here. The Blue Ridge eclogites are heating from granulite to hornblende granulite assemblages. interpreted as low temperature (LT) eclogites formed 1 ¼ inferred eclogite (reconstructed omphacite-plagio- clase-garnet mantles; minimum pressures only); 2 ¼ high-pressure during the subduction of Ordovician granulite (diopside-pargasite-plagioclase-garnet rims-ilmenite); beneath the Inner Piedmont terrane (Willard & Adams, 3 ¼ amphibolite (pargasite II-plagioclase-ilmenite-epidote). 1994; Adams et al., 1995; Adams & Trupe, 1997). These DEEP IN THE HEART OF DIXIE 75

Table 3. Whole rock major and trace element analyses by X-ray fluorescence.

Sample# 3080 NEW-2 3071UP 3071 A 3071B 3071C 3071–4 3071–5 3071 FG 3025 A 3025B Rock Ferro- Ferro- 3070–1 3070–2 3070–3 3070–4 Amphib- Mafic Mafic Mafic Amphib- Amphib- Felsic Felsic Felsic type basalt Basalt Basalt Basalt Basalt olite dyke boudin boudin boudin olite dyke olite dyke Gneiss Gneiss Gneiss

SiO2 41.7 41.5 50.1 49.8 49.1 50.4 51.0 49.3 50.4 49.3 50.3 48.0 54.9 60.6 60.0 TiO2 2.77 3.60 0.54 1.20 1.31 1.32 1.14 2.56 3.15 3.12 2.05 1.24 1.16 0.89 0.88

Al2O3 16.24 13.01 16.86 13.50 13.70 13.32 15.10 16.20 15.36 15.67 12.69 15.01 19.88 16.53 16.79

Fe2O3 16.70 20.36 7.38 12.11 11.71 13.12 10.69 11.66 12.70 12.37 14.23 11.66 9.93 8.37 8.54 MnO 0.13 0.18 0.13 0.20 0.22 0.20 0.17 0.17 0.16 0.20 0.25 0.20 0.08 0.33 0.33 MgO 5.61 6.48 8.65 8.00 7.83 5.83 6.67 5.81 5.84 5.82 5.83 6.75 2.49 1.17 1.24 CaO 13.51 11.30 11.48 12.31 12.87 13.18 10.92 10.43 9.46 10.10 10.36 13.06 2.96 3.69 3.87

Na2O 2.09 2.10 2.72 2.57 2.20 2.03 2.91 1.70 1.78 1.57 2.21 1.99 4.98 4.16 4.29

K2O 0.176 0.209 0.606 0.39 0.398 0.132 1.091 0.956 0.495 0.697 1.228 0.707 2.948 3.006 2.903 P2O5 0.100 0.076 0.045 0.106 0.094 0.163 0.090 0.498 0.450 0.481 0.224 0.112 0.031 0.123 0.145 Total 99.03 98.84 98.47 100.18 99.43 99.71 99.79 99.25 99.82 99.36 99.40 98.76 99.31 98.84 98.94 p.p.m. Nb 2.4 2.5 1.4 2.8 3.3 3.2 4.7 21.1 25.9 25.2 9.5 3.3 23 12.1 12.3 Zr 66 32 33 60 64 61 65 230 273 244 122 69 177 259 273 Y 171414232320223742373728135963 Sr 937 100 180 59 116 85 210 548 367 436 133 274 255 422 418 Rb 5 4 19 7 6 5 15 47 13 33 20 15 160 55 53 Sc 50.1 52.6 40.3 38.3 44.2 41.7 36.1 35.1 39.2 37.6 39.2 51.2 10.2 8 9.4 V 666 734 171 309 324 341 293 363 438 397 406 325 172 70 82 Cr 30 43 259 288 283 76 125 160 185 186 65 268 123 10 15 Ni 8 18 158 78 63 69 65 67 49 55 48 112 79 8 10 Cu 46 107 107 96 90 44 62 31 9 30 138 148 121 41 29 Zn 117 124 47 88 87 93 86 90 98 89 112 99 116 81 85 Ba 34 19 13 85 117 12 77 bdl bdll bdl 26 168 365 659 586

Note: Major elements in weight percentage oxide, trace elements in p.p.m.

(a) (b) (c)

(d) (e) (f)

Fig. 12. Harker diagrams for eclogite ⁄ high-P granulites of the Silverstreet domain, along with three felsic gneiss host rocks. Plots show SiO2 vs. (a) TiO2 (b) FeO* (c) MgO and CaO (d) Na2O and K2O (e) Cr and Ni p.p.m., and (e) Zr p.p.m. rocks were preserved in thrust sheets during the Alle- from Laurentia, and prior to accretion of the Caro- ghanian collision of North America with Gondwana in lina terrane to the Laurentian margin (Late Palaeo- the late Palaeozoic (Stewart et al., 1997). zoic). Since the Silverstreet domain appears to be part of the Charlotte belt and may represent in part the infrastructure of the Charlotte belt arc, subduction Regional Implications polarity during amalgamation of the Charlotte belt One conclusion that seems inescapable at this time is arc with the Carolina slate belt arc must have been to that the Carolina terrane, as it is now understood, is a the SE (present day co-ordinates; Fig. 14). That is, composite terrane composed of two previously unre- the Charlotte belt arc formed part of the lower plate lated arcs: the Charlotte belt arc and the Carolina assemblage that was over-ridden by the Carolina slate slate belt arc. Amalgamation of these two arcs to belt arc (which formed part of the upper plate of the form the Carolina terrane must have occurred far subduction zone). 76 J. W. SHERVAIS ET AL.

800 assemblage that was active at the time of collision; this 10 is not observed in the Goochland terrane. 700 V ppm The high pressure metamorphic event documented 20 Ferrobasalts here must predate the c. 415 Ma Newberry granite 600 ARC (which is unaffected by the amphibolite facies over- MORB print, and contains xenoliths of the sheared felsic 500 gneisses which host the eclogites), and it must postdate 50 400 formation of the Carolina arc (c. 620 to 550 Ma; Basalts Dennis et al. 1997; Dennis & Wright, 1997; Samson 300 et al., 1995; Wortman et al., 1996, 2000; Barker et al., 1998). High-P metamorphism may coincide with the c. 200 535 Ma collisional event documented in the western Charlotte belt by Dennis & Wright (1997), or it may 100 Felsic Gneiss represent an older, as yet unrecognized collisional 0 event that occurred within the Carolina terrane prior 0 5000 10000 15000 20000 25000 to its amalgamation with Laurentia. The occurrence of MORB-chemistry dykes in the Ti ppm 25000 roots of the arc implies that prior to its collision and Ti ppm partial subduction beneath the Slate belt arc, the Charlotte belt-Silverstreet arc may have collided with 20000 ferro-basalts and over-ridden an active spreading centre. This is consistent with the collision of two arcs that were 15000 originally separated by a spreading centre, which would have to be consumed before collision of the two arcs could occur (Fig. 14). It may also explain the high 10000 MORB temperatures and isobaric heating observed in the felsic gneiss Silverstreet domain. 5000 Calc-alkaline IAT Uplift and Exhumation 0 0 50 100 150 200 250 300 High-pressure, MT granulites ⁄ eclogites of the Silver- Zr ppm street domain were exhumed and cooled extremely rapidly, as shown by the preservation of prograde Fig. 13. Ti-V and Ti-Zr plots for eclogite ⁄ granulite blocks zoning profiles in garnet that formed at 660–820 °C from Silverstreet domain, showing MORB ⁄ OIB affinities of (e.g. O’Brien, 1997). These profiles could not persist if the mafic rocks. the boudins were held for long times at such high temperatures. Exhumation of high-P metamorphic Alternatively, Hibbard & Samson (1995) have sug- rocks involves two related problems: (1) a driving force gested that collision between Carolina and the Gren- for uplift of the crust, and (2) the dominant mechanism ville-aged Goochland terrane of eastern Virginia for exhumation – erosion vs. low-angle normal faulting (Farrar, 1984) might be responsible for the meta- (Jamieson & Beaumont, 1989). In this context, uplift morphic fabric of the western Carolina terrane. refers to upward movement with respect to a fixed Mueller et al. (1996) and Heatherington et al. (1996) datum, whereas exhumation refers to unroofing and have presented results based on their work in North movement to lower lithostatic pressures (Jamieson & Carolina that suggest a swing in Nd isotopic compo- Beaumont, 1989; Ring et al., 1999). sitions of Carolina lavas from strongly positive eNd to There are two possible models that may drive the near 0 or even negative at approximately the Cambrian uplift of deeply buried rocks in subduction zones or boundary. A complex collision between a c. 1.1 Ga collision zones, both related to the buoyancy of crustal continental fragment and some portion of the Carolina rocks at depth: (1) the buoyancy of tectonically thicken composite terrane could explain this observed swing in crust when material is removed from the upper crust isotopic compositions and some of the xenocrystic by erosion or faulting, or (2) break-off of the sub- zircon in Albemarle Group lavas from the North ducting slab, thus removing slab pull and allowing the Carolina slate belt. Subsequent terrane dispersal subducting crust to return buoyantly to the surface during postcollisional strike-slip faulting could be (von Blanckenburg & Davies, 1995; Ernst et al., 1997). responsible for the present disposition of high pressure We favour the slab breakoff model as the release rocks far from the present outcrop belt of the mechanism for uplift, because it is a logical conse- Goochland terrane. However, since the Charlotte belt- quence of the partial subduction of a buoyant block Silverstreet arc must have been in the lower plate of the attached to oceanic lithosphere, and because the two collision, the upper plate would have to include an arc arcs involved in this collision (Charlotte belt, Slate DEEP IN THE HEART OF DIXIE 77

Fig. 14. Model for collision and amalgamation of the Charlotte belt ⁄ Slate belt arcs. Stage 1: convergence of Charlotte belt and Slate belt arcs, separated by mid-ocean ridge spreading centre; Stage 2: collision of Charlotte belt with MOR, emplacement of MORB composition dykes into infrastructure of the arc; Stage 3: continued convergence of Charlotte belt and Slate belt arcs, with detachment and sinking of subducted lithosphere; Stage 4: collision of Charlotte belt and Slate belt arcs, with eclogite facies metamorphism of Charlotte belt arc infrastructure; Stage 5: and sinking of oceanic lithosphere at leading edge of Charlotte belt arc, followed by rapid uplift and exhumation of Charlotte belt infrastructure; granulite and then amphibolite facies overprint during exhumation; Stage 6: postcollisional ÔCarolina terraneÕ compositer arc; later arc volcanism and sedimentation may represent overlap assemblages that postdate suture; later reactivation of suture during Alleghanian sinistral shear? belt) were likely too thin to generate great thicknesses progressive ⁄ retrogressive metamorphism under eclog- of crust. In addition, it would allow isobaric heating ite, high pressure granulite, and amphibolite facies followed by sudden and rapid pressure release and conditions. These rocks define a clockwise P-T-t path, cooling. consistent with collision and partial subduction. We Exhumation, defined as the return of deeply buried suggest that this collision may have occurred during rocks to the surface, is generally driven either by ero- amalgamation of the Charlotte belt to the Carolina sion, low-angle normal faulting, or both; ductile Slate belt to form the composite Carolina terrane that extension of the crust or lithosphere is slow and cannot was later accreted to Laurentia. account for more than a small fraction of exhumation Our conceptual model for this collision, based on the (Ring et al., 1999). Erosion and low-angle normal data discussed above, is outlined in Fig. 14. In stage 1, faulting both operate at similar rates (5–13 km Myr) 1 the Charlotte belt arc and the Slate belt arc face one for erosion, 5–10 km Myr)1 for normal faulting) and another across an active spreading centre. The Char- either can account for the rapid cooling required to lotte belt arc over-rides this spreading centre in stage 2, preserve the growth zoning in garnet. In the case of the leading to the emplacement of MORB composition Silverstreet domain, we suggest that low angle faulting dykes in the dominantly felsic arc basement. By stage was the dominant process, because the fault contact 3, the Charlotte belt arc became extinct and attached between the Charlotte belt and Slate belt removes a to oceanic crust still subducting beneath the Slate belt significant thickness of metamorphic section (c. 14 km) arc. Collision occurred during stage 4, with the and juxtaposes terranes with significantly different Charlotte belt arc in the lower plate being partially . Clearly, this process must have over-ridden by the Slate belt arc, leading to the high- been aided by rapid erosion to remove material from pressure metamorphism at eclogite and high-P granu- the footwall, but erosion alone cannot account for the lite facies conditions. During stage 5 the subducting juxtaposition of terranes with distinctly different oceanic slab broke off, allowing rapid exhumation metamorphic grades and palaeodepths (Platt, 1986, of the Charlotte belt arc basement, possibly along a 1993). low-angle normal fault. Finally (stage 6) subduction was re-established beneath the combined Charlotte belt-Slate belt arc with renewed volcanism and plut- CONCLUSIONS onism, and possible reactivation of the suture as a Mafic meta-igneous rocks found along the boundary high-angle transcurrent structure. Further work is between the Charlotte belt and the Carolina Slate belt needed to refine and test this model in other parts of preserve metamorphic phase assemblages that imply the Carolina terrane. 78 J. W. SHERVAIS ET AL.

ACKNOWLEDGEMENTS Appalachians. Part II. geochronological constraints on the tectonothermal evolution of the eastern Piedmont in South The authors thank R. Jamieson and J. Gilloti for their Carolina. Geological Society of America Bulletin, 97, 1329– thoughtful and detailed reviews, A. Indares for edi- 1344. torial handling, and especially M.J. Kohn for discus- Dennis, A. J. & Shervais, J. W., 1991. Evidence for arc rifting along the Carolina terrane boundary in northwestern South sions of thermobarometry issues. We also thank Carolina. Geology, 19, 226–229. A. Bell and M. A. Jones (deceased) for assistance in the Dennis, A. J. & Shervais, J. W., 1996. The Carolina Terrane in field. This work was supported by NSF grant northwestern South Carolina: Insights into the development EAR9805159 to Dennis and Shervais. of an evolving island arc. In: Avalonian and Related Peri- Gondwanan Terranes of the Circum North Atlantic. Special Paper 304, (eds Nance, R. D. & M. D. Thompson), p. 237– REFERENCES 256. Geological Society of America, Boulder. Dennis, A. J., Shervais, J. W. & Secor, D. T., 2000. Newberry, Adams, M. G., Stewart, K. G., Trupe, C. H. & Willard, R. A., South Carolina Eclogite. Structural Setting and Style of 1995. Tectonic Significance of High-Pressure Metamorphic Occurrence. In: A Compendium of Field Trips of South Rocks and Dextral Strike Slip Faulting in the Southern Carolina Geology, (eds C. Abate & A. B. Maybin), 29–38. Appalachians. In: New Perspectives in the Appalachian- South Carolina Geological Survey, Columbia Caledonian Orogen, Geological Association of Canada Special Dennis, A. J. & Wright, J. E., 1995. Mississippian (ca. 326–323 Paper (eds. Hibbard, J., van Staal. C. R. Cawood, P. & Ma) U-Pb crystallization ages for two granitoids in Spartan- Colman-Sid, S.), 41, 21–42. burg and Union Counties, South Carolina. In: Geology of the Adams, M. G. & Trupe, C. H., 1997. Conditions and timing of Western Part of the Carolina Terrane, (ed. Dennis, A. J.) South metamorphism in the Blue Ridge thrust complex, northwest- Carolina Geology 38, 23–28. ern North Carolina and eastern Tennesee. In: Paleozoic Dennis, A. J. & Wright, J. E., 1997. The Carolina terrane in Structure, Metamorphism, and Tectonics of the Blue Ridge of northwestern South Carolina, USA: Late precambrian-Cam- Western North Carolina, Field Trip Guide, (eds Stewart, K. G. brian deformation and metamorphism in a peri-Gondwana Adams, M. G. & Trupe, C. H.), pp. 33–48. Carolina oceanic arc. Tectonics, 16, 460–473. Geological Society, Banner Elk. Ellis, D. J. & Green, D. H., 1979. An experimental study of the Barker, C. A., Secor, D. T., Pray, J. R. & Wright, J. E., 1998. effect of Ca upon garnet-clinopyroxene Fe-Mg exchange Age and deformation of the Longtown metagranite, South equilibria. Contributions to Mineralogy and , 71, Carolina Piedmont: a possible constraint on the origin of the 13–22. Carolina terrane. Journal of Geology, 106, 711–723. Elvevold, S. & Gilotti, J. A., 2000. Pressure-temperature evolution von Blanckenburg, F. & Davies, J. H., 1995. Slab breakoff: a of retrogressed eclogites, Weinschenk Island, North- model for syncollisional magmatism and tectonics in the Alps. East Greenland Caledonides. Lithos, 53, 127–147. Tectonics, 14, 120–131. England, P. C. & Thompson, A. B., 1984. Pressure-temperature Bohlen, S. R. & Liotta, J. J., 1986. A barometer for garnet time paths of regional metamorphism; I, Heat transfer during amphibolites and garnet granulites. Journal of Petrology, 27, the evolution of regions of thickened . Journal 1025–1034. of Petrology, 25, 894–928. Carswell, D. A., 1990. Eclogites and the eclogite facies: defini- Ernst, W. G., Maruyama, S. & Wallis, S. R., 1997. Bouyancy tions and classification. In: Eclogite Facies Rocks. (ed. D.A. driven rapid exhumation of ultra-high pressure metamor- Carswell), p. 1–13. Blackie, London. phosed continental crust. Proceedings of the National Academy Carswell, D. A. & Harley, S. L., 1990. Mineral barometry and of Science, 94, 9532–9537. thermometry. In: Eclogite Facies Rocks. (ed. D.A. Carswell), Farrar, S. S., 1984. The Goochland granulite terrane: Remobi- p. 83–110. Blackie, London. lized Grenville basement in the eastern Virginia Piedmont. Carswell, D. A. & O’Brien, P. J., 1993. Thermobarometry and In: The Grenville Event in the Appalachians and Related Topics. geotectonic significance of High-Pressure Granulites: Exam- Special Paper 194. (ed. Bartholomew, M. J.), pp. 215–227. ples from the of the Bohemian Massif in Geological Society of America, Boulder. lower Austria. Journal of Petrology, 22, 85–126. Fullagar, P. D., 1981. Summary of Rb-Sr whole rock ages for Cooke, R. A., O’Brien, P. J. & Carswell, D. A., 2000. Garnet South Carolina. South Carolina Geology, 25, 29–32. zoning and the identification of equilibrium mineral composi- Galan, G. & Marcos, A., 2000. The metamorphic evolution of tion in high-pressure-temperature granulites from the Molda- the pressure mafic granulites of the Bacariza formation nubian zone, Austria. Journal of Metamorphic Geology, 18, (Cabo Ortegal complex), Hercynian belt, NW Spain. Lithos, 551–569. 54, 139–171. Coward, M. P., Butler, R. W. H., Khan, M. Asif & Knipe, R. J., Graham, C. M. & Powell, R., 1984. A garnet-hornblende 1987. The tectonic history of Kohistan and its implications for geothermometer; calibration, testing, and application to the Himalayan structure. Journal of the Geological Society of Pelona , Southern California. Journal of Metamorphic London, 144, 377–391. Geology, 2, 13–31. Cuthbert, S. J. & Carswell, D. A., 1990. Formation and Halik, R. S., 1983. Characterization of the Charlotte belt– exhumation of medium temperature eclogites in the Scandi- Carolina slate belt boundary in the Silverstreet and Denny navian Caledonides. In: Eclogite Facies Rocks. (ed. D.A. quadrangles, South Carolina: A stratigraphic, structural and Carswell), p. 180–203. Blackie, London. metamorphic study. MSc Thesis University of South Carolina, Cuthbert, S. J., Carswell, D. A., Krough-Ravna, E. J. & Wain, Columbia. A., 2000. Eclogites and eclogites in the Western Gneiss Hauck, M. L., 1984. Geologic characterization of the Charlotte Region, Norwegian Caledonides. Lithos, 52, 165–195. belt in South Carolina. MSc Thesis University of South Dal Piaz, G. V. & Lombardo, B., 1986. Early Alpine eclogite Carolina, Columbia. metamorphism in the Monte Rosa-Gran Paradiso Heatherington, A. L., Mueller, P. A. & Nutman, A. P., 1996. basement nappes of the northwestern Alps. In: and Neoproterozoic magmatism in the Suwannee terrane: Eclogites, Memoir 164, (eds B. W. Evans & G. H. Brown), Implications for terrane correlations. In: Avalonian and pp. 249–265. Geological Society of America, Boulder. Related Peri-Gondwanan Terranes of the Circum-North Atlan- Dallmeyer, R. D., Wright, J. E., Secor, D. T. & Snoke, A. W., tic, Special Paper 304 (eds Nance, R. D. & Thompson, M.), 1986. Character of the Alleghanian orogeney in the southern pp. 219–236. Geological Society of America, Boulder. DEEP IN THE HEART OF DIXIE 79

Hibbard, J. P. & Samson, S. D., 1995. Orogenesis Exotic to the Atlantic, Special Paper 304 (eds R. D. Nance & M. D. Iapetan Cycle in the Southern Appalachians. In: Current Thompson), pp. 207–217. Geological Society of America, Perspectives in the Appalachian Caledonian Orogen, Special Boulder. Paper 41, (eds Hibbard, J.P., van Staal.C., Cawood, P.), Newton, R. C. & Perkins, D., 1982. Thermodynamic calibration pp. 191–205. Geological Association of Canada, Ottawa. of geobarometers based on the assemblages garnet – plagio- Holland, T. J. B., 1980. The reaction albite ¼ jadeite + quartz clase – orthopyroxene (clinopyroxene) – quartz. American determined experimentally in the range 600–1200 °C. Ameri- Mineralogist, 67, 203–222. can Mineralogist, 65, 129–134. O’Brien, P. J., 1997. Garnet zoning and reaction textures in Holland, T. & Blundy, J., 1994. Non–ideal interactions in calcic overprinted eclogites, Bohemian Massif, European Variscides: and their bearing on amphibole-plagioclase ther- a record of their thermal history during exhumation. Lithos, mometry. Contributions to Mineralogy and Petrology, 116, 41, 119–133. 433–447. O’Brien, P. J., Carswell, D. A. & Gebauer, D., 1990. Eclogite Horton, J. W. & Dicken, C. L., 2001. Preliminary Digital formation and distribution in the European Variscides. Geologic Map of the Appalachian Piedmont and Blue Ridge, In: Eclogite Facies Rocks. (ed. D.A. Carswell), pp. 204–224. South Carolina Segment. US Geological Survey Open-File Blackie, London. Report 01–298, Scale 1: 500,000, CD ROM. O’Brien, P. J. & Rotzler, J., 2003. High-Pressure granulites: Horton, J. W., Drake, A. A. & Rankin, D. W., 1989. Formation, Recovery of peak conditions, and implications for Tectonostratigraphic terranes and their Paleozoic boundaries tectonics. Journal of Metamorphic Geology, this volume. in the central and southern Appalachians. In: Terranes in the Offield, T. W., 1995. Structural contrasts of the Carolina slate Circum-Atlantic Paleozoic Orogens, Special Paper 230. (ed. belt and Charlotte belt . In: South Carolina, Dennis, A. J., ed. Dallmeyer, R. D.), pp. 213–245. Geological Society of Geology of the Western Part of the Carolina Terrane, South America. Boulder. Carolina Geology 38, 61–70. Horton, J. W., Drake, A. A., Rankin, D. W. & Dallmeyer, R. Offield, T. W. & Sutphin, D., 2000. Revised Stratigraphic and D., 1991. Preliminary tectonostratigraphic terrane map of the Tectonic Framework of the Carolina Slate Belt from Southern central and southern Appalachians. U.S. Geological Survey Virginia to the South Carolina-Georgia Border (with Observa- Miscellaneous Investigations Series, Map I-2163, scale 1: tions Bearing on the Origin of Slate-Belt Gold Deposits). U.S. 2,000,000, 1 sheet. Geological Survey Open File Report 99–2, CD-ROM con- Jamieson, R. A. & Beaumont, C., 1989. Deformation and taining text and digital maps, scales 1: 100,000 and 1: 500,000. metamorphism in convergent orogens: a model for uplift and Pattison, D. R. M., 2003. Petrogenetic significance of ortho- exhumation of metamorphic terranes. In: Evolution of pyroxene-free garnet + clinopyroxene + plagioclase-bearing Metamorphic Belts. Special Publication 43, (eds. Daly, J. S., metabasites with respect to the amphibolite and granulite Cliff, R. A. & Yardley, B. W. D.), pp. 117–129. Geological facies. Journal of Metamorphic Geology, this volume. Society, London. Pearce, J. A. & Cann, J. R., 1973. Tectonic setting of basic Jan, M. Q. & Howie, R. A., 1981. The mineralogy and volcanic rocks determined using trace element analyses. geochemistry of the metamorphosed basic and ultrabasic and Planetary Science Letters, 19, 290–300. rocks of the Jijal complex, Kohistan, NW Pakistan. Journal of Pin, C. & Vielzeuf, D., 1988. Les granulites de haute pression Petrology, 22, 85–126. d’Europe moyenne terrains d’une subduction eo-hercynienne: Khan, M. A., Jan, M. Q., Windley, B. F., Tarney, J. & Thirwall, Implications sur l’origine des groupes leptyno-amphiboliques. M. F., 1989. The Chilas mafic-ultramafic complex; the root of Bulletin Des Societie Geologique , 8, 13–200. the Kohistan island arc in the Himalaya of northern Pakistan. Platt, J. P., 1986. Dynamics of orogenic wedges and the uplift of Geological Society of America Special Paper, 232, 75–94. high-pressure metamorphic rocks. Geological Society of Amer- Kohn, M. J. & Spear, F. S., 1989. Empirical calibration of ica Bulletin, 97, 1037–1053. geobarometers for the assemblage garnet-hornblende-plagio- Platt, J. P., 1993. Exhumation of high pressure rocks: a review of clase-quartz. American Mineralogist, 74, 77–84. concepts and processes. Terra Nova, 5, 119–133. Kohn, M. J. & Spear, F. S., 1990. Two new barometers for Pouchou, J.-L. & Pichoir, F., 1991. Quantitative analysis of garnet amphibolites with applications to SE Vermont. homogeneous or stratified microvolumes applying the model American Mineralogist, 75, 89–96. ÔPAPÕ. In: Electron Probe Quantitation, (eds K. F. J. Heinrich Krogh, E. J., 1982. Metamorphic evolution of Norwegian & D. E. Newbury), pp. 31–76. Plenum Press, New York. country rock eclogites as deduced from mineral inclusions Powell, R., 1985. Regression diagnostics and robust regression in and compositional zoning of . Lithos, 15, 305–321. geothermometer ⁄ geobarometer calibration: the garnet-clino- Libby, W. G. & Carpenter, J. R., 1969. Petrography and pyroxene geothermometer revisited. Journal of Metamorphic geochemsitry of a mafic granofels in Newberry County, South Geology, 3, 231–243. Carolina. Southeastern Geology, 10, 55–62. Powell, R. & Holland, T. J. B., 1988. An internally consistent Lindsley, D. H. & Anderson, D., 1983. A two-pyroxene thermodynamic dataset with uncertainties and correlations: 3. thermometer. Journal of Geophysical Research, 88, A887–A906. Applications to geobarometry, worked examples and a com- McSween, H. Y., Speer, J. A. & Fullagar, P. D., 1991. Plutonic puter program. Journal of Metamorphic Geology, 6, 173–204. rocks. In: The Geology of the Carolinas, Carolina Geological Ring, U., Brandon, M. T., Willett, S. D. & Lister, G. S., 1999. Society Fiftieth Anniversary Volume. (eds Horton, J W. Exhumation processes. In: Ring, U., Brandon, M. T., Lister, & Zullo, V.A.), pp. 109–126. University of Tennessee Press, G. S. & Willett, S. D., eds. Exhumation Processes: Normal Knoxville. Faulting, Ductile Flow and Erosion. Special Publication 154, Messiga, B., Tribuzio, R. & Vannucci, R., 1990. Mafic and 1–27. Geological Society, London. ultramafic pods with eclogitic relics from the Robinson, P., 1980. The Composition Space of Terrestrial Nagssugtoqidian mobile belt of East Greenland. Lithos, 25, . Reviews in Mineralogy, 7, 419–494. 101–118. Rubie, D. C., 1990. Role of kinetics in the formation and Moecher, D. P., Essene, E. & Anovitz, L. M., 1988. Calculation preservation of eclogites. In: Eclogite Facies Rocks. (ed. D.A. and application of cpx-gar-plag-qtz geobarometer. Contribu- Carswell), pp. 111–140. Blackie, London. tions to Mineralogy and Petrology, 100, 92–106. Samson, S. D., 1995. Is the Carolina Terrane Part of Avalon?. Mueller, P. A., Kozuch, M., Heatherington, A. L., Wooden, J. In: Current Perspectives in the Appalachian Caledonian Orogen, L., Offield, T. W., Koeppen, R. P. & Klein, T. L., 1996. Special Paper 41, (eds Hibbard, J.P., van Staal.C., Cawood, Evidence for Mesoproterozoic basement in the Carolina P.), pp. 253–264. Geological Association of Canada, Ottawa. terrane and speculations on its origin. In: Avalonian Samson, S. D., Hibbard, J. P. & Wortman, G. L., 1995. Nd and Related Per-Gondwanan Terranes of the Circum-North isotopic evidence for juvenile crust in the Carolina terrane, 80 J. W. SHERVAIS ET AL.

southern Appalachians. Contributions to Mineralogy and Spear, F. S. & Kohn, M. A., 2001. GTB: Program Geothermo- Petrology, 121, 171–184. barometry, Version 2.1. February 2001. computer program. Samson, S., Palmer, A. R., Robison, R. A. & Secor, D. T, 1990. http://ees2.geo.rpi.edu/MetaPetaRen/Frame_software.html. Biogeographical significance of Cambrian trilobites from the van Staal, C. R., Dewey, J. F., MacNiocaill, C. & McKerrow, Carolina slate belt. Geological Society of America Bulletin, W. S., 1998. The Cambrian-Silurian tectonic evolution of the 102, 1459–1470. northern Appalachians and British Caledonides: history of a Samson, S. D. & Secor, D. T., 2000. New U-Pb geochronological complex, west and southwest Pacific-type segment of Iapetus. evidence for a Silurian magmatic event in central South In: Lyell: the Past Is the Key to the Present. Special Carolina. Geological Society of America Abstracts with Publication, 143, (eds Blundell, D. J. & Scott, A. C.), Programs, 32/4. pp. 199–242. Geological Society, London. Secor, D. T., 1988. Geology of the eastern Piedmont in central Stewart, K. G., Adams, M. G. & Trupe, C. H., 1997. South Carolina. In: D. T. Secor, ed. Southeastern Geological Paleozoic structural evolution of the Blue Ridge thrust Excursions Geological Society of America, Southeastern complex, western North Carolina. In: Paleozoic Structure, Section, Field Trip Guidebook for 1988. 204–225. Metamorphism, and Tectonics of the Blue Ridge of Western Secor, D. T., Dennis, A. J., Maher, H. D. & Sacks, P. E., 1989. North Carolina, Field Trip Guide, (eds Stewart, K. G., Adams, Geology and terrane analysis of the eastern Piedmont, South M. G. & Trupe, C. H.), pp. 21–31. Carolina Geological Carolina and Georgia. In: Dallmeyer, R. D., ed. Tectonos- Society, Banner Elk. tratigraphic Expression of Terrane Accretion in the Southern West, T. E., 1998. Structural analysis of the Carolina–Inner Appalachian Orogen: a Geotraverse Excursion, the University Piedmont terrane boundary: Implication for the age and of Georgia, I G C, p. 233. Guidebook Series, 8 ⁄ 25–8 ⁄ 91 and kinematics of the central Piedmont suture, a terrane boundary 9 ⁄ 1–9 ⁄ 17. that records Paleozoic Laurentia–Gondwana interactions. Secor, D. T., Peck, L. S., Pitcher, D. M. et al., 1982. Geology of Tectonics, 17, 379–394. the area of induced seismic activity at Monticello Reservoir, Will, T. M. & Schma¨ dicke, E., 2001. A first find of retrogressed South Carolina. Journal of Geophysical Research, 87, 6945– eclogites in the Odenwald crystalline complex, mid-German 6957. Crystalline Rise, Germany: evidence for a so far unrecognized Secor, D. T., Samson, S. L., Snoke, A. W. & Palmer, A. R., high-pressure metamorphism in the central Variscides. Lithos, 1983. Confirmation of the Carolina slate belt as an exotic 59, 109–125. terrane. Science, 221, 649–650. Willard, R. A. & Adams, M. G., 1994. Newly discovered Secor, D. T., Snoke, A. W., Bramlett, K. W., Costello, O. P. & eclogite in the southern Appalachian orogen, northwestern Kimbrell, O. P., 1986. Character of the Alleghanian orogeney North Carolina. Earth and Planetary Science Letters, 123, in the southern Appalachians. Part I. Alleghanian deforma- 61–70. tion in the eastern Piedmont of South Carolina. Geological Williams, H. & Hatcher, R. D. 1983. Appalachian suspect Society of America Bulletin, 97, 1314–1328. terranes. In: Contributions to the Tectonics and Geophysics of Shervais, J. W., 1982. Ti-V plots and the petrogenesis of modern Mountain Chains. Memoir 158, (eds Hatcher, R. D., Williams, and ophiolitic lavas. Earth and Planetary Science Letters, 59, H. & Zietz, I.), pp. 33–53. Geological Society of America, 101–118. Boulder. Shervais, J. W., Shelley, S. A. & Secor, D. T., 1996. Wortman, G. L., Samson, S. D. & Hibbard, J. P., 1996. Geochemistry of volcanic rocks of the Carolina and Augusta Discrimination of the Milton belt and Carolina terrane in the terranes in central South Carolina: An exotic rifted volcanic southern Appalachians: a Nd isotopic approach. Journal of arc?. In: Avalonian and Related Per-Gondwanan Terranes of the Geology, 104, 239–247. Circum-North Atlantic, Special Paper 304. (eds Nance, R. D. Wortman, G. L., Samson, S. D. & Hibbard, J. P., 2000. Precise & M. D. Thompson), pp. 237–256. Geological Society of U-Pb zircon constraints on the earliest magmatic history of America, Boulder. the Carolina terrane. Journal of Geology, 108. Spear, F. S., 1993. Metamorphic Phase Equilibria and Pressure- Temperature Time Paths. Monograph 1, Mineralogical Society of America, Washington DC. Received 26 October 2001; revision accepted 8 July 2002.

APPENDIX: METHODS Omphacite was reconstructed from diopside-plagioclase symplec- tite by acquiring a series of BSE images on different diopside grains, All minerals were analyzed on the Cameca SX-50 electron micro- which were then segmented in the Cameca image analysis software to probe at the University of South Carolina using natural and syn- recover the proportion of plagioclase in diopside. Results for all thetic mineral standards from the Smithsonian Institution; grains imaged were around 20% modal plagioclase. The inferred operating conditions were typically 20 KV at 25 nA. Data reduction omphacite composition was reconstructed by converting the average was carried out using the Cameca implementation of the phi-rho-z composition of plagioclase in symplectite to a mineral formula algorithm (Pouchou & Pichoir, 1991). Large area X-ray composi- and subtracting one mole of SiO2 to create jadeite. The jadeite tion mapping was carried out in stage mode using four fixed formula was converted back to weight% oxide and mixed with the wavelength spectrometers in conjunction with the backscattered average composition of diopside in 20 : 80 proportions to create electron (BSE) diodes. Typical maps are 512 · 512 pixels with a step omphacite. size 10 lm and dwell time of 200 milliseconds, for a total area scanned Calculation of Fe3+/Fe2+ by stoichiometry for thermobarometry of 5.0 · 5.0 mm and an analysis time of about 14 h per map. followed method of Spear (1993).