Upper and lower crust recycling in the source of CAMP basaltic dykes from southeastern North America

Earth and Planetary Science Letters 376 (2013) 186–199

Contents lists available at SciVerse ScienceDirect

Earth and Planetary Science Letters www.elsevier.com/locate/epsl

Upper and lower crust recycling in the source of CAMP basaltic dykes from southeastern North America Sara Callegaro a,∗ , Andrea Marzoli a , Hervé Bertrand b , Massimo Chiaradia c , Laurie Reisberg d , Christine Meyzen a , Giuliano Bellieni a , Robert E. Weems e , Renaud Merle f a

Dipartimento di Geoscienze, Università di Padova, via Gradenigo 6, 35100 Padova, Italy Laboratoire de Géologie de Lyon – UMR CNRS 5276, Université Lyon et Ecole Normale Supérieure de Lyon, 46 Allée d’Italie, 69364 Lyon Cedex 7, France c Section des Sciences de la Terre, Université de Genève, 13 rue des Maraîchers, 12011 Genève, Switzerland d Centre de Recherches Pétrographiques et Géochimiques (CRPG)–CNRS UMR 7358, Université de Lorraine, BP 20, 54501 Vandoeuvre-les-Nancy Cedex, France e Paleo Quest, 14243 Murphy Terrace, Gainesville, VA 20155, USA f Department of Applied Geology, Curtin University, GPO Box 25 U1987, Perth, WA 6845, Australia b

a r t i c l e

i n f o

Article history: Received 31 December 2012 Received in revised form 13 June 2013 Accepted 16 June 2013 Available online 17 July 2013 Editor: T.M. Harrison Keywords: dykes CAMP Appalachians Os–Sr–Nd–Pb isotopes mantle source crustal recycling

a b s t r a c t The densest dykes swarm of the Central Atlantic magmatic province (CAMP) occur in southeastern North America (SENA) and were intruded between 202 and 195 Ma during Pangea break-up. New combined geochemical data (major and trace elements, Sr–Nd–Pb–Os isotopes) constrain the mantle source of these magmatic bodies and their evolution path. While Sr–Nd isotopic compositions for SENA rocks (87 Sr/86 Sr200Ma 0.70438–0.70880 and 143 Nd/144 Nd200Ma 0.51251–0.51204) fall within the low-Ti CAMP field, Pb–Pb isotopes (206 Pb/204 Pb200Ma 17.46–18.85, 207 Pb/204 Pb200Ma 15.54–15.65, 208 Pb/204 Pb200Ma 37.47–38.76) are peculiar to this area of the CAMP and cover a considerable span of compositions, especially in 206 Pb/204 Pb200Ma . Given the generally unradiogenic Os isotopic compositions (187 Os/188 Os200Ma 0.127–0.144) observed and the lack of correlation between these and other geochemical markers, crustal contamination during the evolution of SENA dykes must have been limited (less than 10%). Thus the isotopic variation is interpreted to reside primarily within the mantle source. These observations, coupled with typical continental signatures in trace elements (positive anomaly in Pb and negative anomalies in Ti and Nb), require another means of conveying a continental flavor to these magmas, which is here hypothesized to be the shallow recycling within the upper mantle of subducted lower and upper crustal materials. Pseudo-ternary mixing models show that a maximum of 10% recycled crust is enough to explain their trace element patterns as well as their isotopic heterogeneity. Looking at the larger picture of the origin of the CAMP, the thermal contribution of a mantle plume cannot be ruled out due to the relatively high mantle potential temperatures (1430–1480 ◦ C) calculated for highFo SENA olivines. Nevertheless, our results suggest that the chemical involvement of a mantle plume is negligible (less than 5%) if either a C- or an EM-flavored plume is considered. Rather, the possibility of a PREMA-flavored mantle plume, enriched by 5–20% recycled crustal material, remains a possible, though less plausible, source for these tholeiites. © 2013 Elsevier B.V. All rights reserved.

1. Introduction

Large Igneous Provinces (LIPs) have drawn growing attention from the scientific community due to their synchrony with major mass extinctions and continental break-up events and their potential role in triggering these events. For this reason, among others, it is interesting to inspect the mantle source of the Central At-

*

Corresponding author. E-mail address: [email protected] (S. Callegaro).

0012-821X/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.epsl.2013.06.023

lantic magmatic province (CAMP), as well as the modalities of its emplacement. Rapidly outpoured or intruded around the Triassic–Jurassic boundary (peak activity centered at ca. 201 Ma; Marzoli et al., 2011) over a surface exceeding 107 km2 , CAMP tholeiitic dykes, sills, intrusions and flows are spread over four continental masses surrounding the central Atlantic Ocean (Marzoli et al., 1999). As is true for many other continental flood basalt (CFB) provinces, the origin of the CAMP is still a matter of debate. Investigators ascribe it either to the impingement of a mantle plume (Cebrià et al., 2003; Hill, 1991; Wilson, 1997) or to lithospheric partial melting after thermal incubation under the

S. Callegaro et al. / Earth and Planetary Science Letters 376 (2013) 186–199

187

Pangaea supercontinent (Coltice et al., 2007; McHone, 2000). This argument concerns the entire magmatic province, including its eastern North American (ENA) part (i.e. Beutel et al., 2005; De Boer and Snider, 1979; Heatherington and Mueller, 1999; Pegram, 1990). In this framework, our study is centered on the dyke swarms from southeastern North America (SENA). This region of the CAMP displays the highest density of dykes (along with the Taoudenni dyke swarm from Mali; Verati et al., 2005), and is considered by several authors to be the site of the first mantleplume impingement (e.g. De Boer and Snider, 1979; Hill, 1991; May, 1971; Wilson, 1997) and of first continental break-up in Pangea (Schlische et al., 2003). A previous study of this region of the CAMP (Pegram, 1990) suggested that magmatism resulted from partial melting of the subcontinental lithospheric mantle (SCLM), variously enriched during an early Proterozoic subduction event (pre-Grenville, ca. 1.6 Ga), and that CAMP magmas were not contaminated by the continental crust. By providing a large set of major and trace element data, combined with Sr–Nd–Pb–Os isotopic data (recalculated to the ca. 200 Ma crystallization age) on near-primitive basaltic dykes, we provide further constraints on the petrogenesis of SENA CAMP magmas. Furthermore, we propose a model for the SENA CAMP mantle source, in which the shallow mantle has been enriched by recycled upper and lower crustal material subducted during the Paleozoic. Such an interpretation highlights the influence of shallow mantle processes in the genesis of the CAMP and is consistent with early ideas of Puffer (2001) and with recent findings for the Paraná CFB (Rocha-Junior et al., 2012). 2. Geologic context 2.1. Tectonic setting The geologic history and evolution of eastern North America (ENA) are traced through the development of more than one entire Wilson cycle, from the assemblage of the Rodinia supercontinent (Mesoproterozoic, 1.3–1.1 Ga; Li et al., 2008) to the breakup of Pangaea (Early Jurassic). The two main convergent phases are represented by the Proterozoic Grenville (e.g. McLelland et al., 2010) and the late Paleozoic Appalachian orogenies. The latter is related to the Early Paleozoic closure of the Iapetus and Rheic oceans during a succession of orogenic events (Taconian, Acadian and Alleghanian orogenies), leading to the accretion of microcontinents (superterranes) of Gondwanan affinity to Laurentia, i.e. Carolina, Avalon, Ganderia and Meguma allochtonous terranes (e.g., Van Staal et al., 2009) and eventually resulting in the continental collision that assembled the supercontinent Pangaea (Bradley, 1982). Limited subduction of oceanic crust in the South supposedly led to a) sparse intrusion of mafic suprasubduction-zone plutons (Esawi, 2004) and b) subduction of continental crust that totally shut off mafic pluton emplacement (Hatcher, 2010; Hibbard et al., 2010). The extensional phases were characterized instead by widespread magmatism. The rifting marking Rodinia breakup (Neoproterozoic, 750–700 Ma) and opening of the Iapetus ocean (broadly along the present North American margin; Scotese et al., 1979) was associated with bimodal, rift-related, volcanism (Bakersville dykes, Catoctin volcanic province; Badger and Sinha, 1988; Goldberg et al., 1986; Hibbard et al., 2002). The breakup of Pangaea was heralded by a Middle and Late Triassic extensional phase and culminated with the early Jurassic opening of the central Atlantic Ocean. This event occurred as a rift-to-drift transition, proceeding from South to North (e.g., Schlische et al., 2003). In this extensional geodynamic framework, close to the Triassic–Jurassic boundary, the SENA CAMP magmas were intruded in the southeastern U.S.A.

Fig. 1. Sampling sites (stars) are plotted on an overview map of the southeastern North American diabase dyke swarm as it appears from the compilation of Ragland et al. (1983), based in turn on aeromagnetic mapping of the area. Triassic rift basins of the Newark Supergroup are outlined as grey areas. The thick gray line represents the Fall Line, marking the limit between the Appalachian Piedmont area (to the NW) and the Coastal Plain (SE). U.S.A. state acronyms are as follows: GA Georgia, SC South Carolina, NC North Carolina, VA Virginia, DE Delaware, NJ New Jersey and PA Pennsylvania. Inset: CAMP outcrops are represented on a Late Triassic geographic reconstruction (modified after Marzoli et al., 2011).

2.2. CAMP in eastern North America CAMP relics in eastern North America (yielding an 40 Ar/39 Ar dated peak activity at 201.5 ± 0.9 Ma; Marzoli et al., 2011) are represented by: a) the here studied SENA diabase dyke swarms (and the Durham sill) occurring from Alabama to Virginia (Fig. 1) (Beutel et al., 2005; Hames et al., 2000; Nomade et al., 2007; Pegram, 1990; Ragland et al., 1983). Most studied (7 samples) SENA dykes yielded robust 40 Ar/39 Ar plateau ages on plagioclase ranging between 202.1 ± 1.5 and 199.4 ± 1.7 Ma, while two samples gave a significantly younger age at 196.1 ± 2.2 Ma and 195.2 ± 2.0 Ma (Beutel et al., 2005; Hames et al., 2000; Nomade et al., 2007; recalculated after Renne et al., 2010); b) the basaltic lava flows (e.g. Puffer, 1992), Early Jurassic Newark Supergroup basins, from Virginia (Culpeper basin) to Nova Scotia (Fundy basin; mean 40 Ar/39 Ar 201.7 ± 0.7 Ma; Jourdan et al., 2009; Marzoli et al., 2011); c) sub-surface diabase dykes and possibly basaltic flows interbedded within sedimentary basins of Florida, Georgia and Alabama (Heatherington and Mueller, 1999) and d) wedge-shaped basaltic bodies offshore of eastern North America (Heffner et al., 2012; Holbrook and Kelemen, 1993). This study is centered on CAMP diabase dykes (dolerites) cropping out in the southern portion of the Appalachians. We sampled and analyzed 73 dykes from Georgia to Virginia and the Durham sill (3 samples) in North Carolina (Fig. 1). SENA CAMP dykes intruded the Piedmont area and the Coastal Plain of the Appalachians (eastern U.S.A.), cutting through either the crystalline base-

188

S. Callegaro et al. / Earth and Planetary Science Letters 376 (2013) 186–199

Cr-spinel (20–36 wt% Cr2 O3 ), Ti-magnetite (11–22 wt% TiO2 ), ilmenite (46–48 wt% TiO2 ), as well as infrequent orthopyroxene. Samples devoid of olivine phenocrysts (Q-normative rocks, mainly) contain pigeonite and rare fine-grained granophyre in the groundmass. Primary sulphides are also observed in some samples. Alteration phases such as serpentine and iddingsite replace olivine crystals in some rocks, while plagioclase may display sericite coatings. Devitrified glass, hydrous minerals (amphibole, chlorite, biotite, zeolites) and carbonates are present in the groundmass, as hydrothermal or alteration phases. 3.2. Mineral chemistry and thermobarometry

Fig. 2. Total Alkali versus Silica (TAS) classification diagram (LeMaitre et al., 2002) of the 76 sampled ENA dykes and sills. Black circles: LFO samples; white circles: HFO samples; grey triangles: LFQ samples; white triangles: HFQ samples. The black dashed line separates alkaline from subalkaline compositions.

ment or the Triassic sediments of the Newark Supergroup basins (Schlische et al., 2003). The sampled dykes are near vertical and vary in thickness from a few to 150 m (e.g. the Pageland dyke), showing a maximum elongation of 250 km (Ragland et al., 1983). They vary in trend from NW (N 10◦ –30◦ W) in the south, to N-S, to mostly NE (N 30◦ E) in the north. Minor occurrences of E-W dykes were also sampled. Though previously considered to result from the upwarping effect of a mantle plume emplaced under the Carolinas (De Boer and Snider, 1979), recent geochronological and paleomagnetic studies convincingly link these different trends to a rapidly changing stress field (Beutel et al., 2005). Cross-cutting relationships, palaeomagnetic and 40 Ar/39 Ar data suggest that the N-S swarm is slightly younger than NW-trending dykes (Beutel et al., 2005; Ragland et al., 1983). 3. Results Major elements were analyzed by X-ray fluorescence (XRF; Phillips PW1404) at the University of Padova. Trace elements were measured with a VG Element plasma quadrupole II ICP-MS at the Washington State University. Sr–Nd–Pb isotopes were analyzed by Thermal Ionization Mass Spectrometry (TIMS), with a Thermo TRITON or a Finnigan MAT262 mass spectrometer at the University of Geneva. Re–Os isotopic analyses were performed at the Université de Lorraine (CRPG Vandoeuvre-les-Nancy) by negative thermal ionization mass spectrometry (Finnigan MAT262). Mineral chemistry (pyroxenes, olivine, plagioclase, oxides) was investigated with a CAMECA SX50 electron microprobe (EMP) at the IGG-CNR of Padova. Further details on the analytical are reported in the online Supplementary Material. 3.1. Petrography Based on major element compositions (XRF; see Supplementary Material), the investigated rocks (maximum loss on ignition, LOI = 2.6 wt%) are classified as basalts or basaltic andesites (Fig. 2) and range in CIPW normative composition from quartz (Q-) to olivine-hypersthene (Ol/Hy) normative. SENA dykes display simple mineralogical assemblages, typical of tholeiitic dolerites and appear either olivine- or pigeonitephyric, with a broad grain size spectrum, varying from fine-grained (smaller dykes and margins of the thicker dykes) to sub-pegmatitic (Durham sill). Textures range from aphyric to strongly porphyritic, and can be intersertal, intergranular or ophitic. Most dykes have phenocrysts of olivine (Fo89–48 ), augite (En56–33 Fs9-30 Wo42–36 ), plagioclase (An84–47 ), and oxides such as

Porphyritic rocks contain olivine with high-Fo cores (up to Fo89 ) which are close to equilibrium with the host whole-rock compositions, for an olivine/melt KD (Fe/Mg) of 0.30 ± 0.03 (Roeder and Emslie, 1970), whereas their clinopyroxenes have low Mg# (81–67) suggesting later crystallization. In contrast, in ophitic rocks, olivines have relatively low Fo contents (72–61), whereas clinopyroxenes are relatively Mg-rich (Mg# 90–80). Olivines show a good positive correlation between Ni (473–3344 ppm) and Fo (58–89), and a negative correlation between Mn (708–5275 ppm) and Fo. Pressures and temperatures were estimated for 6 samples, basing on the thermobarometers of Putirka (2008) for major element compositions of minerals, which are in chemical equilibrium with their host whole-rock. Olivine crystallization temperatures range from 1325 to 1350 ± 15 ◦ C for porphyritic rocks, in which olivine appears to be the first liquidus phase. Clinopyroxene cores yield maximum crystallization pressures ranging between 0.85 and 0.5 GPa and temperatures of 1237–1222 ◦ C. Mantle potential temperature (T p ; McKenzie and Bickle, 1988), calculated for dry conditions for high-Fo (> 87) olivine cores of high-MgO (> 12 wt%) rocks, yields values of 1487–1466 ◦ C if calculated after Putirka (2008), while the same rocks (CS28 and CS57) yield T p around 1440–1432 ◦ C if calculated with PRIMELT2 (Herzberg and Asimow, 2008). Despite some uncertainty on the H2 O content and redox conditions of the magmas, which we assume to be near-anhydrous and poorly oxidizing (QFM), based on the mineralogical compositions, we note that the calculated T p values are consistent with those published for the CAMP by Herzberg and Gazel (2009), who reported mantle T p reaching a maximum value of 1480 ◦ C. Such T p are significantly above the range of the average ambient mantle (1300–1400 ◦ C), but well below maximum T p values for the plume-related Deccan and Siberian Trap CFBs (Sobolev et al., 2011). 3.3. Major and trace elements A wide compositional range is observed, varying from nearprimary to quite evolved (Mg# = 74–43; MgO = 14.8–4.7 wt%; Fig. 3), though all the rocks are low in TiO2 (0.40–1.35 wt%). Four groups can be defined on the basis of major element geochemistry and CIPW normative compositions. Ol/Hy-normative rocks can be grouped as low-Fe samples (hereafter, LFO; Fe2 O3 tot < 13.0 wt%) and high-Fe samples (HFO; Fe2 O3 tot > 13.0 wt%). Similarly, Qnormative samples with less than 12.0 wt% Fe2 O3 tot are classified as low-Fe (LFQ), while those with more than 12.0 wt% Fe2 O3 tot as high-Fe (HFQ). The four normative groups can also be clearly identified in terms of other major element contents and display different evolutionary trends (Fig. 3). In particular, HFO rocks are enriched in alkali elements (Na2 O and K2 O 2.4–3.7 vs 1.6–2.9 wt%), TiO2 and P2 O5 and are low in SiO2 (46.3–48.4 vs 46.7–50.2 wt%), CaO and Al2 O3 , relative to LFO basalts at similar MgO. Trace elements were measured on 45 SENA dykes (whole rocks with LOI < 2.0 wt% (Supplementary Material). All are moderately

S. Callegaro et al. / Earth and Planetary Science Letters 376 (2013) 186–199

189

Fig. 3. Major elements variation diagrams for 76 ENA diabase samples. Dashed and dotted lines represent liquid lines of descent calculated with MELTS (Ghiorso and Sack, 1995) starting from a LFO (CS28) and a HFO sample (CS9), respectively. Fractional crystallization was modeled at 0.3 GPa for near-anhydrous compositions (0.3 wt% H2 O in starting magma), with oxygen fugacity at the Quartz–Fayalite–Magnetite buffer. Low pressure fractional crystallization of olivine + spinel and then plagioclase + augite assemblages yields liquid lines of descent that are consistent with the major element variations displayed by the studied rocks. The quite evolved LFQ and HFQ SENA rocks can be derived after ca. 40–60% fractional crystallization from, respectively, LFO magmas and both HFO or LFO magmas. Other symbols represent the compositions of experimental partial melts from peridotites, plotted for comparison. Grey squares mark experimental melts of Hirose and Kushiro (1993) obtained from a fertile spinel lherzolite (HK66, from Hawaii Islands) at 2.5 (large squares) and 3.0 GPa (small squares). Dark grey is for lower (ca. 10%) and light grey for higher (ca. 20%) melting degrees. Grey diamonds represent experimental melts obtained from Longhi (2002) from a lherzolite with a chemical composition representative of the primitive upper mantle, at 3 GPa. Dark grey indicates low (1%) melting degree, light grey indicates higher (10%) extent of melting.

enriched in incompatible elements (IE), with their Rare Earth elements (REE) patterns broadly resembling those of enriched (E) MORBs (McDonough and Sun, 1995). However, unlike E-MORBs and within-plate oceanic basalts in general, SENA dykes are depleted in Nb, Ta and Ti and enriched in Rb, Ba and Pb, therefore displaying a clear continental signature as observed for most CAMP basalts (e.g., Merle et al., 2011a). REE (Fig. 4(b)) are moderately enriched compared to chondritic values (McDonough and Sun, 1995), with mildly fractionated patterns (i.e. LaC /YbC : 0.54–2.60; LaC /SmC : 1.18–2.53; SmC /YbC : 0.42–1.40; subscript C: chondrite normalized). Broad overlap exists between REE patterns of Ol/Hyand Q-normative SENA dykes, but the latter yield generally both slightly more enriched compositions and more fractionated patterns. 3.4. Sr–Nd–Pb isotopes Isotopic analyses for Sr, Nd and Pb were carried out on 28 whole rocks (LOI < 1.8 wt%). All isotopic ratios are age-

corrected to the ca. 200 Ma crystallization age (Table 1). They exhibit a wide range in 87 Sr/86 Sr200Ma (0.70438–0.70883) and 143 Nd/144 Nd200Ma (0.51251–0.51204; ε Nd = −6.67 to +2.42), overlapping the broad negative Sr vs Nd isotopic correlation displayed by low-Ti basalts from the entire CAMP, including those of ENA lava flows (Merle et al., 2011b; Pegram, 1990; Tollo and Gottfried, 1992; Fig. 5). Rocks from the four groups entirely overlap in the Sr–Nd isotopic space, with LFO samples sweeping the entire range of values. The large Sr–Nd isotopic variability is also mirrored by Pb isotopes (206 Pb/204 Pb200Ma 17.41–18.61, 207 Pb/204 Pb200Ma 15.54–15.63, 208 Pb/204 Pb200Ma 37.32–38.38), with the sampled SENA dykes extending to far lower 206 Pb/204 Pb and 208 Pb/204 Pb values with respect to the other low-Ti CAMP rocks (Fig. 5). In particular, in Pb–Pb spaces, samples define a positive linear trend, which is displaced above (but not exactly parallel to) the Northern Hemisphere Reference Line (NHRL; Hart, 1984). A measure of this displacement is given by the parameters 7/4 (10–18 and 8/4 (19–73;  values defined after Hart et al., 1986). In addition, they depict a broadly positive correlation in Nd–Pb iso-

190

Sample

Group

Rb

Sr

Sm

Nd

U

Pb

Th

CS2

HFQ

15.1

117.7

3.7

11.6

0.52

3.3

1.92

CS46

HFQ

17.7

209.8

3.5

13.1

0.63

3.9

Re

Os

87

Rb

86 Sr

0.372

2.04

0.245

147

Sm

144 Nd

0.191 0.164

238

U

204 Pb

25.592 25.513

235

U

204 Pb

0.188 0.187

232

Th

204 Pb

187

Re

188 O s

87

Srm 86 Sr

143

Ndm 144 Nd

206

Pbm 204 Pb

207

Pbm 204 Pb

208

Pbm 204 Pb

187

Osm 188 O s

87

Sri 86 Sr

143

Ndi 144 Nd

206

Pbi 204 Pb

207

Pbi 204 Pb

208

Pbi 204 Pb

31.902

0.706511

0.51271

18.842

15.649

38.764

0.705474

0.51246

18.522

15.633

38.381

31.804

0.705657

0.51264

18.847

15.626

38.554

0.704977

0.51243

18.516

15.609

38.207

GDFT2

HFQ

21.5

103.5

3.0

11.6

0.61

3.7

2.26

0.601

0.156

43.545

0.320

47.900

0.707005

0.51261

18.899

15.645

38.762

0.705332

0.51239

18.563

15.628

38.358

CS23

LFQ

13.8

138.9

2.3

8.7

0.43

3.1

1.62

0.287

0.159

9.718

0.071

43.004

0.705738

0.51263

18.816

15.621

38.606

0.704940

0.51241

18.539

15.607

38.266

CS44

LFQ

11.8

0.10

2.6

0.65

0.151

0.148

8.474

0.062

29.414

0.707597

0.51229

17.558

15.561

37.656

0.707177

0.51209

17.484

CS77

LFQ

19.7

108.0

2.1

8.9

0.30

3.3

2.01

0.528

0.140

12.827

0.094

40.340

0.709192

0.51238

18.568

15.625

38.561

0.707722

0.51219

18.383

15.616

38.161

CS7

HFO

11.8

108.2

3.3

10.8

0.15

2.6

0.92

0.317

0.182

9.386

0.069

26.251

0.707041

0.51251

17.877

15.556

37.798

0.706159

0.51228

17.760

15.550

37.571

CS9

HFO

20.4

119.1

2.8

9.3

0.31

2.6

1.46

0.496

0.186

7.264

0.706818

0.51251

18.427

15.601

38.472

18.187

15.589

38.107

CS17

HFO

9.9

194.6

3.6

13.0

0.16

1.9

0.95

0.148

0.170

20.891

0.153

76.177

0.707215

0.51251

18.482

15.621

38.571

0.706804

0.51229

18.316

15.613

38.239

CS22

HFO

6.8

174.0

3.4

11.8

0.14

1.7

0.93

0.113

0.176

9.923

0.073

37.608

0.705905

0.51253

18.455

15.619

38.603

0.705592

0.51230

18.291

15.610

38.244

CS26

HFO

6.8

109.5

2.3

8.1

0.04

1.5

0.22

0.706594

0.51209

17.413

15.555

37.399

CS32

HFO

12.8

265.1

4.9

19.4

0.21

3.4

1.17

0.705823

0.51225

18.048

15.591

37.920

CS14

LFO

13.8

86.9

2.5

9.1

0.15

2.3

0.91

0.881

CS28

LFO

7.5

128.3

1.8

6.4

0.11

1.7

0.43

0.626

CS31

LFO

8.5

108.1

2.1

7.0

0.25

1.9

0.87

CS41

LFO

4.8

97.0

1.4

4.5

0.03

0.8

0.18

10.2

196.0

182.3

2.9

1.9

1.368

1.167

0.905

0.436

36.728

3.248

9.976

0.073

33.183

10.280

0.076

39.241

0.924

0.458

0.165

8.510

0.063

33.015

4.600

0.708040

0.51240

18.012

15.581

38.015

0.144

0.706766

0.51219

17.884

15.574

37.757

0.415

0.168

0.170

2.357

0.017

15.972

7.300

0.705913

0.51251

18.028

15.583

37.795

0.161

0.705445

0.51229

17.903

15.577

37.632

0.226

0.178

5.763

0.042

39.595

0.705131

0.51273

18.756

15.621

38.502

0.704501

0.51250

18.490

15.608

38.200

0.188

3.622

0.027

22.277

1.420

0.705935

0.51250

17.489

15.539

37.472

0.162

0.705533

0.51226

17.420

15.535

37.328

0.705926

0.51248

17.911

15.583

37.748

0.705575

0.51227

17.779

15.577

37.570

8.806

0.704719

0.51274

18.767

15.617

38.568

0.165

0.704388

0.51250

18.610

15.609

38.330

0.705827

0.51258

17.813

15.560

37.604

0.705497

0.51234

17.705

15.555

37.464

0.705528

0.51274

18.631

15.616

38.443

0.705280

0.51249

18.540

15.611

38.264

0.139 0.138

0.126

0.166

7.348

0.054

35.530

0.503

0.276

0.119

0.179

5.210

0.038

33.021

0.119

0.187

5.005

0.037

34.718

0.089

0.185

1.576

0.012

9.301

LFO

7.9

7.0

0.11

1.6

0.45

CS49

LFO

6.1

148.0

1.9

6.5

0.11

1.4

0.52

CS51

LFO

5.7

138.7

2.2

7.1

0.08

1.5

0.32

CS55

LFO

4.5

147.6

2.3

7.6

0.09

2.0

0.54

1.024

0.132

CS57

LFO

4.4

139.7

1.7

6.1

0.07

0.23

0.911

0.258

CS60

LFO

8.4

87.6

1.6

5.0

0.12

1.3

0.47

1.011

0.310

CS62

LFO

6.7

94.8

1.8

5.6

0.13

1.5

0.51

CS64

LFO

7.6

127.8

2.0

6.0

0.07

1.5

0.43

0.983

CS73

LFO

13.9

3.0

0.69

0.854

37.490

15.597

38.141

0.183

0.266

0.140 0.128 0.136 0.131 0.135

0.090

0.170

21.691

0.882

0.705141

0.51248

17.931

15.567

37.588

0.195

0.704890

0.51226

17.825

15.561

37.464

0.276

0.196

4.016

0.030

25.822

2.350

0.705933

0.51260

18.392

15.626

38.269

0.180

0.705164

0.51234

18.208

15.617

38.037

0.192

3.944

0.029

16.295

0.705601

0.51260

18.357

15.616

38.201

0.705036

0.51234

18.183

15.607

37.973

0.197

8.267

0.061

29.787

29.274

0.705603

0.51276

18.633

15.608

38.438

0.229

0.705124

0.51250

18.535

15.603

38.250

0.132

14.275

19.551

37.469

0.209

0.51204

17.463

0.144

0.172

0.213

0.028

15.557

18.172

0.203 0.164

3.819

37.960

17.462

0.51245

0.130

0.175

0.145

0.51232

0.706212

0.51226

0.153

0.452

0.707097

0.705438

37.496

0.181

0.880

12.990

0.155

Osi 188 O s

0.140

CS48

1.2

0.053

15.557

187

87.7

2.1

8.9

0.10

0.458

0.140

2.157

15.542

37.322

FTNC4

LFO

8.3

111.2

1.6

5.9

0.15

1.7

0.66

0.216

0.165

4.139

0.030

17.873

0.706808

0.51241

18.165

15.630

38.205

0.706208

0.51220

17.983

15.621

37.950

FTNC13

LFO

13.6

170.3

2.3

10.3

0.18

2.8

0.95

0.231

0.137

4.833

0.036

0.016

23.228

0.708215

0.710105

0.51227

0.51222

18.174

17.531

15.609

15.545

38.280

0.707571

0.708830

0.51209

18.047

15.602

38.062

NC2B

LFO

1.8

96.5

1.4

4.5

0.04

0.7

0.20

0.053

0.189

3.379

0.025

13.930

0.705051

0.51250

17.842

15.587

37.772

0.704905

0.51225

17.728

15.582

37.602

0.128

S. Callegaro et al. / Earth and Planetary Science Letters 376 (2013) 186–199

Table 1 Measured and initial (200 Ma) isotopic ratios and parent/daughter trace element data for the analyzed SENA CAMP rocks. Data are reported in ppm for Sr, Nd, U, Th and Pb, in ppb for Re and Os. Details on the analytical methods and on the uncertainties of measured and initial isotopic data are given in the Supplementary Material Table A.

S. Callegaro et al. / Earth and Planetary Science Letters 376 (2013) 186–199

191

petrogenetic histories spread over a very wide geographic area. Therefore they may have had non-uniform (though mantle-like) initial isotopic ratios and slightly variable ages, and so would not meet the strict conditions required for development of a true isochron. Nevertheless, the Re–Os age agrees well with the 40 Ar/39 Ar ages (201.4–195.2 Ma; Beutel et al., 2005; Hames et al., 2000; Nomade et al., 2007; all data recalculated after Renne et al., 2010), suggesting that the Re–Os system was not significantly disturbed after magmatic crystallization. SENA diabases display a narrow range in 187 Os/188 Os200Ma (0.1278 ± 0.0009–0.1442 ± 0.0023; 2σ , including all identified sources of uncertainty; cf. Supplementary Material), similar to those of the few previously analyzed CAMP basalts (from northern Brazil; Merle et al., 2011a). HFO and LFO samples are indistinguishable in terms of Os isotopic composition and overlap with the Primitive Upper Mantle (PUM) value at 200 Ma (0.1281 ± 0.0008, derived from the present-day value of PUM of 0.1296; Meisel et al., 2001). [Re] is not correlated with 187 Os/188 Os200Ma values, and neither is [Os], even if two samples yielding distinctly higher [Os] (CS9, HFO: 0.90 ppb and CS14, LFO: 0.92 ppb) are among those with the lowest isotopic initial values (ca. 0.128–0.130). 187 Os/188 Os200Ma values do not correlate with Sr, Nd or Pb isotopes, and do not show any systematic variation with geographic position (or orientation) of the dykes. 4. Discussion

Fig. 4. (a) Primitive mantle-normalized (McDonough and Sun, 1995) trace element and (b) chondrite-normalized REE contents of analyzed SENA tholeiites (gray field). The compositions of the most primitive (MgO-rich) sample per normative group are shown in detail in both diagrams. For comparison, E-MORB average composition (McDonough and Sun, 1995) is shown (black dashed line). Additional points plotted in (b), represent the REE contents of 5% (grey diamonds) and 10% (white diamonds) partial melts of a hypothetical upper mantle. The composition of the mantle source rock is obtained by adding 8% local upper crust (Pe-Piper and Jansa, 1999) and 2% lower crust (Rudnick and Fountain, 1995) to 90% average depleted MORB mantle (DMM; Workman and Hart, 2005). The modal composition of a spinel lherzolite was considered (55% olivine, 25% orthopyroxene, 18% clinopyroxene, 2% spinel; Hirose and Kushiro, 1993), and partial melting was modeled within the spinel stability field. Distribution coefficients between solid and melt are as in Johnston and Schwab (2004) for clinopyroxene and spinel, as in Salters and Longhi (1999) for orthopyroxene and as in McKenzie and O’Nions (1991) for olivine.

topic space. There is no correlation between isotopic compositions of SENA dykes and either their geographic position/orientation or the lithology of the intruded country rocks. Also, systematic isotopic variations are never visible as a function of MgO or other differentiation indexes. 3.5. Os isotopes Re–Os isotopic analyses were performed on 11 Ol/Hy-normative samples (Table 1 and Supplementary Material Table A). Q-normative samples were not analyzed because highly differentiated, i.e. likely to yield very low Os contents. Measured Os concentrations range from 0.13 to 0.92 ppb, while [Re] varies between 0.5 and 1.4 ppb, with relative uncertainties of 0.3–2.1% and 2.2–6.7%, respectively. A rough positive correlation (not shown) exists between [Os] and MgO contents, consistent with the compatible behavior of Os in magmatic processes. No correlation is observed between [Re] and MgO. Measured 187 Os/188 Os ratios range between 0.1435 and 0.2656, and plotted against 187 Re/188 Os result in an errorchron yielding an age of 211 ± 22 Ma (MSWD = 26), with initial 187 Os/188 Os of 0.1318 ± 0.0067 (Fig. 6(a)). The imprecision of this age is unsurprising since the samples were collected from dykes with varying

High-MgO and low 187 Os/188 Os SENA dykes present geochemical attributes that are typical of slightly differentiated mantle derived magmas. Nevertheless, their incompatible element patterns show continental characteristics, i.e. negative Nb, Ta and Ti and positive Pb anomalies (e.g. Kelemen et al., 1993; Plank, 2005), as well as “enriched” Sr–Nd–Pb isotopic compositions, that is, isotopic signatures indicative of long-term incompatible element enrichment of the source. These geochemical characteristics are ubiquitous in low-Ti CAMP basalts and other CFBs (e.g., Rocha-Junior et al., 2012). However, SENA CAMP dykes display also a large isotopic variability for each of the normative groups, decoupled from distinct major element and IE contents for the four groups (in particular for high MgO HFO vs LFO near-primitive magmas, the former being slightly more enriched in IE). These geochemical features will be discussed in terms of open and closed system magmatic differentiation in the following sections. 4.1. Open system magmatic evolution Open system magmatic evolution for SENA tholeiitic magmas was modeled for Sr–Nd–Pb–Os isotopes assuming both simple AFC (Assimilation Fractional Crystallization; De Paolo, 1981), and EC- (Energy Constrained) AFC (Spera and Bohrson, 2001; Fig. 6(b) and Fig. 7) involving upper and lower crustal contamination. The trace element composition chosen for the upper crustal contaminant corresponds to that of the average upper crust of Taylor and McLennan (1995), with isotopic characteristics based on those provided by several authors for the Southern (e.g. Dennis et al., 2004; Harper and Fullagar, 1981; Samson et al. 1995a, 1995b) and Northern (Ayuso, 1986; Dorais et al., 2008) Appalachians. Despite the significant internal variability, the accreted terranes (i.e. Carolina terrane) of the Southern Appalachians are in general less isotopically “enriched” than those of the Northern Appalachians, probably reflecting the juvenile nature and relatively young age (ca. 600 Ma; Samson et al., 1995b) of the southern terranes. We therefore modeled the assimilation of the upper crust using compositions encompassing the wide range of available crustal data, from very “crustal” (e.g., 87 Sr/86 Sr 0.730) to mantle-like values, but consider a very modestly “enriched” isotopic signature to be most

192

S. Callegaro et al. / Earth and Planetary Science Letters 376 (2013) 186–199

Fig. 5. Isotopic diagrams showing combined initial (200 Ma) Sr–Nd–Pb data for the 28 SENA CAMP dykes. Low-Ti and high-Ti CAMP rocks data are taken from Merle et al. (2011a) and references therein, isotopic fields for ENA flows are from Merle et al. (2011b) and isotopic data for Atlantic MORBs older than 120 Ma located off shore of North and South Carolina are from Janney and Castillo (2001). Isotopic data for the Northern Hemisphere Reference Line (NHRL) are back corrected to 200 Ma using DMM trace element compositions (Workman and Hart, 2005). The DMM reservoir is plotted following Douglass and Schilling (2000) for the isotopic signatures and recalculated to 200 Ma using trace element contents as in Workman and Hart (2005).

Fig. 6. (a) Re–Os isochron plot of the 11 analyzed ENA CAMP tholeiites. Age and uncertainties were calculated using the Isoplot software (Ludwig, 2003). (b) 187 Os/188 Os200Ma ratios plotted as a function of [Os]. CAMP field is marked using data from Merle et al. (2011a, 2011b). Contamination paths for the upper (dotted line) and the lower (dashed line) crust are calculated with EC-AFC modeling (Spera and Bohrson, 2001). Indicated percentages of assimilation refer to Ma parameter of Spera and Bohrson (2001). Uncertainties on each data point are 2σ and include in-run errors as well as blank and weighing uncertainties. Uncertainties on 187 Os/188 Os200Ma values range between 0.8 and 3.1%.

S. Callegaro et al. / Earth and Planetary Science Letters 376 (2013) 186–199

Fig. 7. Contamination paths for assimilation of upper (black line) and lower crust (grey line) are calculated with EC-AFC modeling (Spera and Bohrson, 2001) and plotted on Sr–Nd (a), Pb–Os (b) and 206 Pb/204 Pb vs 207 Pb/204 Pb (c) isotopic spaces, along with values of analyzed SENA CAMP rocks. Marked percentages of assimilation refer to Ma parameter of Spera and Bohrson (2001).

representative of average Southern Appalachian upper crust (Average UC, Supplementary Material Table A; 87 Sr/86 Sr200Ma 0.708; 143 Nd/144 Nd200Ma 0.5125; 206 Pb/204 Pb200Ma 18.5; 207 Pb/204 Pb200Ma 15.53; 208 Pb/204 Pb200Ma 38.1). Given the lack of data for the local lower crust (no xenoliths have been documented so far in the SENA region), a plausible composition was chosen to represent a deep crustal contaminant, with trace element compositions as in Rudnick and Fountain (1995) and isotopic compositions

193

(87 Sr/86 Sr200Ma 0.7054; 143 Nd/144 Nd200Ma 0.5113; 206 Pb/204 Pb200Ma 17; 207 Pb/204 Pb200Ma 15.52; 208 Pb/204 Pb200Ma 37.10) in agreement with those of Grenvillian or Paleozoic lower crusts (e.g. Meyzen et al., 2005; Zartman et al., 2013). Grenvillian crust of Laurentian (continental) affinity (e.g. Samson et al., 1995b) can be envisaged to underlie the accreted, mostly juvenile, upper crust of the Carolina terrane. The 187 Os/188 Os200Ma signatures estimated for average upper and lower crust (0.46 and 0.26, respectively), were calculated for continental crusts of Upper Neoproterozoic age (ca. 600 Ma), supposing a juvenile origin (as reported for the Carolina terrane in Samson et al., 1995b) and using 187 Re/188 Os ratios of 50 and 20, respectively (Shirey and Walker, 1998; Saal et al., 1998). In the example shown in Fig. 7 (parameters detailed in the Supplementary Material), sample CS49 (LFO; MgO 9.12 wt%) was chosen as the starting composition, since it has both the highest 144 Nd/143 Nd200Ma (0.51251) and the lowest 87 Sr/86 Sr200Ma (0.70438), along with high Pb isotopic composition (e.g., 206 Pb/ 204 Pb200Ma 18.61). We note however that similar results would apply for other low 187 Os/188 Os200Ma rocks. Starting from the relatively low 187 Os/188 Os200Ma of CS49 (0.1354; [Os] 276 ppt) the highest observed 187 Os/188 Os200Ma values (e.g., 0.144, [Os] 213 ppt, for CS73; Fig. 6(b)) are reached with a very limited amount of assimilation of either upper (ca. 5%) or lower (ca. 7–9%) crust, following AFC or ECAFC calculations. Even lower amounts of assimilation would be permitted if higher 187 Os/188 Os values, more typical of average continental crust, were assumed for the contaminant. These limited maximum amounts of crustal assimilation would allow for only very minor shifts of Sr–Nd–Pb isotopic ratios, unless the most extreme isotopic compositions of ENA crustal rocks are considered (e.g. 87 Sr/86 Sr 0.750, 143 Nd/144 Nd 0.5114, for acidic plutons cropping out in the Meguma terrane of the Northern Appalachians; Clarke et al., 1997). However, such extreme compositions are not representative of average ENA crust and, more importantly are inherited from the Grenvillian crust, which because of its age would have high 187 Os/188 Os ratios, thus further limiting the amount of allowable contamination. If other isotopic compositions (cf. UC avg, UC min and LC in the Supplementary Material) are taken into account to model the upper or lower crustal contamination, 40% assimilation would be required to reach our most enriched Sr–Nd isotopic compositions starting from CS49, but the modeled contamination paths would not overlap the observed distribution of the samples in Sr–Nd. To reproduce the entire Pb isotopic spread of SENA dykes starting from sample CS49 (high 206 Pb/204 Pb200Ma ), the calculated amount (> 56% following ECAFC calculations) of lower crust assimilation would not be consistent with Os isotopic variations and in particular with the observation that low 187 Os/188 Os200Ma samples (< 0.135) display 206 Pb/204 Pb200Ma from 17.4 to 18.5. Mantle-derived basalt also can be contaminated by a third process, described by e.g. Kerr et al. (1995) and referred to as crustal assimilation during turbulent magma ascent (ATA). For such two-component mixing model, assimilation of ca. 20% upper (Os = 31 ppt; Peucker-Ehrenbrink and Jahn, 2001) and 35% lower crust (Os = 41 ppt; Saal et al., 1998) would be required to drive 187 Os/188 Os from 0.134 to 0.144. With these amounts of assimilated crust, ATA modeling would reproduce also the highest Sr and lowest Nd isotopic compositions of SENA dykes, but it fails to reproduce the Pb isotopic variability and ATA trends would not reproduce the general Sr–Nd–Pb isotopic trends observed for all SENA dykes (see graphs in the Supplementary Material Table A). Moreover, the large amounts of assimilated crust required by ATA are clearly at odds with the generally little evolved compositions (e.g. high MgO) and low IE contents of most samples. Furthermore, in contamination through ATA the hottest and thus most turbulent and less evolved samples bear the most contaminated signatures (Kerr et al., 1995), a feature that is not observed in SENA dykes.

194

S. Callegaro et al. / Earth and Planetary Science Letters 376 (2013) 186–199

In summary, taking into account Sr–Nd–Pb and Os isotopic data and the absence of expected correlations, we can a) constrain the amount of crustal contamination to a maximum of ca. 10%, and b) interpret the isotopic heterogeneity as a heritage of the mantle source. 4.2. Parental magmas and mantle melting Before trying to evaluate potential magma source variations, we must examine the effects of closed system magmatic differentiation. This has been done by calculating liquid lines of descent (Fig. 3) with MELTS (Ghiorso and Sack, 1995), assuming mid-crustal pressure (as constrained by clinopyroxene barometry), moderately oxidizing conditions and low water contents (consistent with the composition and mineralogy of SENA tholeiites). Notably, HFO and LFO magmas appear to have evolved along distinct liquid lines of descent, suggesting the existence of (at least) two distinct parental magma types, issued from either distinct mantle sources or distinct melting regimes. A first attempt to define the SENA CAMP mantle source hinges on comparing our most primitive rocks (Ol/Hy-normative) with experimental melts derived from different fertile mantle lithologies (e.g. lherzolites or pyroxenites; Hirose and Kushiro, 1993; Keshav et al., 2004; Longhi, 2002). According to the approach of Herzberg and Asimow (2008) or Sobolev et al. (2007), variations in major elements between different normative groups might be related to different mantle source lithologies, where LFO and HFO samples (the latter being systematically lower in Ca and Si, higher in Ti, Fe and Na than LFO samples) would be produced from a dominantly lherzolitic and a pyroxenitic mantle source, respectively. We use the definition of pyroxenite as in Sobolev et al. (2007), or Herzberg and Asimow (2008), that is, a near olivine-free mantle lithology produced by reaction of melts issued from eclogites (considered as the high-pressure expressions of subducted oceanic crust) with the ambient peridotite. At first glance, such a pyroxenitic source component could reconcile both the slight enrichment observed in incompatible elements for HFO samples (Fig. 4) and the relatively high (2188–3340 ppm) Ni content observed in Mg-rich (Fo87–89 ) olivines, which is considered as a pyroxenite fingerprint (cf. Sobolev et al., 2007; yet cf. also Putirka et al., 2011). We note that ancient recycled oceanic crust and ambient lherzolite would be expected to produce melts having different IE and isotopic attributes (i.e. possibly enriched flavors being conveyed by the pyroxenite), whereas HFO and LFO samples show overlapping isotopic signatures (Fig. 5). In addition, trace elements patterns observed in SENA dykes do not display the typical trace element traits (enrichment in Nb and Ta, high Ce/Pb) of OIBs derived from a recycled oceanic crust component (Willbold and Stracke, 2006). Though not ruling out a minor contribution from a pyroxenitic source rock for SENA basalts in general, we attribute the systematic differences in major element geochemistry between normative groups to differences in melting pressure and extent of the same dominantly peridotitic mantle source (e.g. the pyroxene-rich peridotite HK66 of Hirose and Kushiro, 1993; Fig. 3). The slightly lower SiO2 and higher Fe2 O3 tot of HFO with respect to LFO basalts would thus suggest slightly higher average melting depth of the former, while the higher Na2 O and K2 O of the HFO (Fig. 2) would argue for lower degrees of melting. Furthermore, experimental (Hirose and Kushiro, 1993) and calculated partial melts from spinel lherzolites (Longhi, 2002) show how CaO content in the melt increases as a function of melting degree, as long as clinopyroxene is a residual phase, which is consistent with the major element evolution displayed by both HFO (low CaO) and LFO (high CaO) samples (Fig. 3). The generation of HFO and LFO at slightly different melting degrees would also be consistent with the incompatible trace element contents (e.g. La, Ce, Zr and Re), which are generally higher

in HFO than in LFO rocks (Fig. 4), as are ratios of strongly to moderately incompatible elements (e.g., La/Yb 2.08–3.83 in HFO vs 1.24–3.52 in LFO). A change in melting pressure and degree from HFO to LFO magmas is also compatible with the extensional geodynamic context of CAMP magmatism preceding the break-up of Pangea and with the possibly slightly older age of N-W trending dykes (all LFO) compared to N-S and E-W dykes (Beutel et al., 2005; Ragland et al., 1983). Potentially, in both active and passive rifting scenarios, the degree of melting and the importance of shallow melting are expected to increase with time. 4.3. Isotopic constraints on mantle sources of the ENA dykes In Section 4.1 we showed that the extent of crustal contamination of the SENA dykes was probably quite limited. Therefore, the continental characteristics of SENA rocks (e.g., high field strength element anomalies and Sr–Nd–Pb isotopes clearly different from those of typical OIBs or MORBs) might be interpreted, following for example Heinonen et al. (2010), as being inherited from a dominant sub-continental lithospheric mantle (SCLM) source, or from extensive contamination of an asthenospheric parental magma with SCLM-derived melts. However, unlike in the Karoo province (e.g., Jourdan et al., 2007), there is no ultra-alkaline (kimberlitic– lamproitic) Triassic–Jurassic magmatism along the ENA margin testifying to the existence of enriched zones in the SCLM. Pegram (1990) proposed that the SENA magmas issued from an SCLM variously enriched in IE during early Proterozoic times. While such a hypothesis may explain the large range in observed isotopic compositions and the generally high 7/4 and 8/4, it is however not consistent with the lack of correlation between SENA isotopic and IE compositions. More importantly, SENA tholeiites show initial 187 Os/188 Os200Ma (> 0.127) too radiogenic to be derived from a ca. 2 Ga SCLM source dominated by depleted mantle, whose 187 Os/188 Os is likely to be lower than 0.125 (Carlson, 2005). Considering that peridotites have high Os concentrations relative to those of potential metasomatic fluids and melts, even enriched zones of the SCLM should have unradiogenic Os isotopic compositions. In contrast, the initial Os isotopic signatures of SENA dykes are similar to or slightly more radiogenic than those ascribed to the primitive mantle. To a first approximation, SENA rocks point towards enriched mantle domains in Pb–Pb isotopic spaces (Fig. 8), but they evidently require an additional end-member distinct from these enriched poles. Modeling of the involvement of various amounts of EMI and EMII components (Zindler and Hart, 1986) along with a main contribution from the upper asthenosphere (depleted MORB mantle, DMM) or from a C-type (Hanan and Graham, 1996) plume component fails to reproduce the data. Thus, we explore the possibility that the enriched end-members sampled by SENA magmas may be located in the shallow mantle, more specifically in the metasomatized suprasubduction environment (cf. Puffer, 2001). Here we present a model in which the observed enriched isotopic signatures are conveyed by the introduction of small portions of recycled continental crust into the depleted suprasubductive mantle. Recycled crust dominates the trace element budget (Fig. 4(b)), due to its extremely enriched character with respect to the ambient upper mantle (e.g. depleted MORB mantle, DMM; Salters and Stracke, 2004; Workman and Hart, 2005), and therefore exerts a strong control both on the IE patterns and on Sr–Nd–Pb isotopic systematics. In contrast, 187 Os/188 Os signatures are buffered by the ambient peridotite, given its 1–2 orders of magnitude higher [Os] (1–4 ppb; Harvey et al., 2011) with respect to that of average continental crust (< 0.05 ppb; Shirey and Walker, 1998). A ternary mixing model was thus developed among DMM, recycled upper crust and recycled lower crust (Fig. 9; further details of the model in the caption). The upper and lower

S. Callegaro et al. / Earth and Planetary Science Letters 376 (2013) 186–199

195

(Douglass and Schilling, 2000) as representative of the DMM composition, due to geographic proximity. Trace element data for the average DMM were based on Workman and Hart (2005). Isotopic signatures are all back-corrected to 200 Ma. Given these end-member compositions, mass balance calculations (in a pseudo-binary mixing model; cf. Douglass and Schilling, 2000) constrain the contribution of the recycled crustal material to be small (2–10%) compared to the largely dominant depleted upper mantle (DMM) component (> 90%). The large range of isotopic compositions, in particular for Pb isotopes of SENA dykes, can be explained by variable relative amounts of lower vs upper crustal components. This interpretation seems to be consistent also with the rough negative correlation observed between 206 Pb/204 Pb200Ma and Eu anomalies (calculated as EuC /[(SmC × GdC )1/2 ]; Supplementary Material Fig. C). As observed by Willbold and Stracke (2010), upper and lower crustal recycling in the mantle should be accompanied by, respectively, negative and positive Eu anomalies, which characterize the two different types of crust (0.66 for the UC and 1.12 for the LC; cf. Rudnick and Fountain, 1995 and Taylor and McLennan, 1995). Finally, it should be noted that adding crustal components directly into the low-IE mantle source has a much larger effect on Sr–Nd–Pb isotopic signatures than crustal assimilation en route to the surface of basaltic magmas with relatively high IE contents (see Section 4.1), because of the difference in incompatible element concentrations between peridotites and basalts. 4.4. Plume component?

Fig. 8. Sr–Nd–Pb isotopes for SENA dykes are plotted along with several OIBs representative of C (Canary, Cape Verde), EMI (Tristan da Cunha, Pitcairn) or EMII-type (Society, Samoa) mantle components (Willbold and Stracke, 2010; Gibson et al., 2005; Hofmann, 2003; Zindler and Hart, 1986).

continental crust end-members have compositions of early Paleozoic acidic and early Proterozoic basic rocks, respectively, consistent with circum-Atlantic terranes (e.g., Meguma granites from North America or old granulites from cratonic or circum-cratonic areas of Gondwana or North America). The crustal end-members are characterized by more and less radiogenic Pb isotopic compositions, respectively, both at high 7/4. To represent the upper crustal signature we selected the composition of a local crust from a dataset from the Meguma terrane (Pe-Piper and Jansa, 1999), taking into account that a) like the Carolina, Avalon and Ganderia terranes, Meguma has been recognized to be an allochtonous terrane of peri-Gondwananan origin (Hatcher, 2010; Pe-Piper and Jansa, 1999) and b) being chiefly composed of metasediments, the Meguma terrane potentially represents the average product of a vast area of exposed crust (Pe-Piper and Jansa, 1999). Moreover, this composition is consistent with that of sediments from the Atlantic Ocean (Hoernle, 1998). Due to the lack of chemical and isotopic constraints on the circum-Atlantic lower crust, mean trace element (Rudnick and Fountain, 1995) and isotopic compositions from a global compilation of early Proterozoic lower continental crustal xenoliths (e.g. Meyzen et al., 2005; Zartman et al., 2013) were used, as in previously discussed AFC calculations. We chose the isotopic signature of a Mid Atlantic Ridge MORB compilation

We explore the possibility that SENA basalts were generated from a mantle-plume source, which may be similar either to the present-day central Atlantic OIBs, yielding a dominant Ccomponent (Hanan and Graham, 1996), or to near-primitive deep mantle material which contributed to the generation of some LIPs (Jackson and Carlson, 2011). Trace element characteristics of the SENA samples argue against a plume source since, with the partial exception of Samoan OIBs, plume-related basalts show different trace element patterns, being chiefly enriched in Nb and Ta and in Ce/Pb ratios, contrary to SENA dykes. Also, isotopic data do not support a C-flavored plume component for the SENA basalts, since plume-related OIBs from the central Atlantic (e.g., Hofmann, 2003) yield quite radiogenic Pb isotopic ratios (206 Pb/204 Pb ∼ 19.3). Such ratios are higher in 206 Pb/204 Pb than those of all SENA (and CAMP) basalts and would fail to explain the high 7/4 of CAMP basalts in general (C mantle pole lies on the NHRL). Thus any chemical contribution from a C component should be limited (< 5%; Fig. 9(c)). Similarly, OIBs from the southern Atlantic which yield an EM-1 signature (e.g. Tristan da Cunha; Gibson et al., 2005) overlap partially the Sr–Nd isotopic composition of SENA dykes, but are distinct in terms of Pb isotopes (Fig. 8). Jackson and Carlson (2011) suggested a non-chondritic deep mantle reservoir as a source for many < 250 Ma LIPs (though not for the CAMP), similar to the prevalent mantle (PREMA component) of Zindler and Hart (1986) in terms of Sr–Nd–Pb isotopic compositions. Notably, the PREMA reservoir (back-corrected to 200 Ma) is isotopically fairly similar to the DMM mantle pole, though more enriched in trace elements. Ternary-mixing models of the PREMA component along with local upper and lower crust reproduce the isotopic compositions of SENA basalts if a 5 to 20% contribution of recycled crust is considered (mixing graphs in the Supplementary Material). However, though we cannot exclude that a PREMA-flavored mantle plume was a source component for the SENA CAMP magmatism, we still tend to favor the DMM as the principal source for several reasons; i.e. a) a rising mantle plume should melt significantly at high pressure, within the garnet stability zone, producing magmas with trace element signature (e.g., high Sm/Yb) different from those

196

S. Callegaro et al. / Earth and Planetary Science Letters 376 (2013) 186–199

Fig. 9. Plots of pseudo-binary mixing calculations (e.g. Douglass and Schilling, 2000) between DMM, lower crust and upper crust to describe Sr–Nd–Pb isotopic variability of ENA dykes. Grey lines mark mixing paths, with tick marks every 10% step. Additional lines marking the 90% and 80% limits of DMM component were drawn. Input data are reported in the Supplementary Material Table C. The DMM end-member is calculated from a Mid Atlantic Ridge MORB compilation (Douglass and Schilling, 2000), due to geographic proximity. Isotopic and trace element data for the upper crust reflect composition of local Meguma crust reported in Pe-Piper and Jansa (1999). Sr elemental data for the upper crust are taken from Taylor and McLennan (1995). Lower crust isotopic values are selected to be within the range of global lower crustal xenolith data (compiled by Meyzen et al., 2005). Trace element composition of the lower crust component is also selected to be within the range of xenolith data and of values reported by Taylor and McLennan (1995) and Rudnick and Fountain (1995). Mixing lines including a C-component signature (Hanan and Graham, 1996) were drawn in the 206 Pb/204 Pb vs 207 Pb/204 Pb. The C reservoir was mixed with three different end members, represented in turn by a) 8% lower crust, 2% upper crust, 90% DMM, b) 7% lower crust, 3% upper crust, 90% DMM and c) 5% lower crust, 5% upper crust, 90% DMM. A maximum chemical contribution of a C-flavored mantle plume component is thus isotopically constrained to 5%. Isotopic composition of the PREMA reservoir (Zindler and Hart, 1986) is also reported (see Supplementary Material for the ternary mixing models involving the PREMA reservoir). All mantle poles and NHRL are recalculated to 200 Ma.

observed for SENA dykes (Fig. 4(b)); b) trace element patterns of CFBs derived from the non-chondritic deep-mantle reservoir (cf. Jackson and Carlson, 2011) are different from those observed for SENA tholeiites, and c) the PREMA would require enrichment by up to 20% recycled crust, twice the amount required to enrich a DMM source, and smaller amounts of recycled material seem more plausible. Finally, calculated mantle potential temperatures for SENA rocks, inferred from olivine or from Mg-rich whole-rock compositions, are around 1430–1480 ◦ C (this study and Herzberg and Gazel, 2009). Even if slightly hotter than the average ambient upper mantle (1300–1400 ◦ C), such Tp values are clearly lower than those of mantle-plume related CFBs (T p up to 1650 ◦ C for Deccan and Siberian traps; Herzberg and Gazel, 2009; Sobolev et al., 2011). 4.5. Constraints on the geodynamic evolution of the ENA margin Compared to Late Neoproterozoic mafic rocks from the SENA area related to the early stages of Rodinia disruption (Bakersville dykes, ca. 730 Ma, Tennessee; and Catoctin volcanic province, 710–560 Ma, Virginia; 87 Sr/86 Sr200Ma ∼ 0.7050 and 0.7040; Badger

and Sinha, 1988; Goldberg et al., 1986), SENA CAMP diabases display an enriched Sr isotopic composition. Assuming that they issued from the same upper mantle, an enrichment of the mantle source must have taken place between the times of the Rodinia and Pangea breakup events (ca. 600 Ma and 200 Ma, respectively), i.e. possibly during the Carboniferous Acadian–Neoacadian subduction (ca. 410–330 Ma; Badger and Sinha, 1988; Hatcher, 2010; Sinha and Zietz, 1982). Transport of continental upper crustal material as oceanic sediments into the mantle wedge and overlying lithosphere can occur at convergent margins, while lower continental crust can be eroded from the obducted plate during subduction erosion or crustal delamination (Levin et al., 2000; Nelson, 1992; Stern, 2011; Willbold and Stracke, 2010). Therefore, subducted sediments and delaminated portions of obducted lower continental crust may have reacted with the ambient shallow lherzolitic mantle possibly forming enriched (pyroxenitic) zones. Since this suprasubductive mantle did not produce (significant amounts of) basaltic magmatism during the Paleozoic (Hatcher, 2010), the recycled crustal portions may have been secured in the shallow mantle at the base of the lithosphere, thus largely escaping con-

S. Callegaro et al. / Earth and Planetary Science Letters 376 (2013) 186–199

vective homogenization. Small scale heterogeneities were thus preserved and reflected in different isotopic signatures of SENA-CAMP dykes separated by only small distances (up to a few kilometers). 5. Conclusion The Southern ENA margin is intruded by swarms of closely spaced dykes and shallow sills belonging to the CAMP but presenting peculiar isotopic characteristics (e.g., highly variable 206 Pb/204 Pb200Ma ), unobserved so far in other CAMP basalts. Different depths and extents of melting of an enriched upper mantle, in an extensional geodynamic context, may have produced the two distinct high-MgO magma types (HFO and LFO) recognized in the SENA area. Os isotopes constrain the amount of assimilated crust to be lower than ca. 10% and argue in favor of an upper mantle source with an undepleted Os composition. However, uniform and typically continental incompatible element patterns (Pb enrichment, Nb–Ta–Ti depletion) along with enriched and heterogeneous Sr–Nd–Pb isotopic signatures require the presence of enriched (and heterogeneous) domains of crustal provenance directly in the mantle source. In addition to ambient depleted mantle or (less likely) to plume-related deep non-chondritic mantle, at least two components are necessary to thoroughly describe the geochemical features of ENA CAMP tholeiites. In particular, recycling of lower continental crust in the shallow mantle is envisaged as the conveyor of the extremely low 206 Pb/204 Pb (at relatively high 207 Pb/204 Pb) end-member signature. Conceivably, since these signatures are not witnessed in any other CAMP region (not even in very proximal northern ENA CAMP volcanics), the process responsible for lower crust recycling may have been geographically restricted to the SENA area, whose underlying mantle thus represents an exception rather than the rule for the CAMP in general. Lower and upper continental crustal domains were plausibly incorporated within the mantle underlying ENA during Paleozoic subduction events. Acknowledgements We are warmly grateful to S. Howard, E.K. Beutel, M. Higgins, R. Crawford, P. Bradley and H. Hannah for field planning and assistance. We thank R. Carampin, D. Pasqual (IGG-CNR and University of Padova) and C. Zimmermann (CRPG, Vandoeuvre-lès-Nancy) for assistance during EMP, XRF and Re–Os analyses, S. Nomade for providing four samples, and G. Cavazzini (CNR-Padova) for discussion of the isotopic data. Financial support was from CA.RI.PA.RO. (Eccellenza 2008 to A.M.) and PRIN (2008 to A.M.). Detailed and very constructive reviews of F. Jourdan, O. Nebel and an anonymous reviewer substantially helped us to improve this contribution. Appendix A. Supplementary material Supplementary material related to this article can be found online at http://dx.doi.org/10.1016/j.epsl.2013.06.023. References Ayuso, R.A., 1986. Lead-isotopic evidence for distinct sources of granite and for distinct basement in the Northern Appalachians, Maine. Geology 14, 322–325. Badger, R.L., Sinha, A.K., 1988. Age and Sr isotopic signature of the Catoctin volcanic province: Implications for subcrustal mantle evolution. Geology 16, 692–695. Beutel, E.K., Nomade, S., Fronabarger, A.K., Renne, P.R., 2005. Pangea’s complex breakup: A new rapidly changing stress field model. Earth Planet. Sci. Lett. 236, 471–485. Bradley, D.C., 1982. Subsidence in late Paleozoic basins in the northern Appalachians. Tectonics 1, 107–123. Carlson, R.W., 2005. Application of the Pt–Re–Os isotopic systems to mantle geochemistry and geochronology. Lithos 82, 249–272.

197

Cebrià, J.M., Lopez-Ruiz, J., Doblas, M., Martins, L.T., Munha, J., 2003. Geochemistry of the Early Jurassic Messejana–Plasencia dyke (Portugal–Spain); implications on the origin of the Central Atlantic Magmatic Province. J. Petrol. 44, 547–568. Clarke, D.B., MacDonald, M.A., Tate, M.C., 1997. Late Devonian mafic–felsic magmatism in the Meguma Zone, Nova Scotia. Geol. Soc. Am. Mem. 191, 107–127. Coltice, N., Phillips, B.R., Bertrand, H., Ricard, Y., Rey, P., 2007. Global warming of the mantle at the origin of flood basalts over supercontinents. Geology 35, 391–394. De Boer, J., Snider, F.G., 1979. Magnetic and chemical variations of Mesozoic diabase dikes from eastern North America: Evidence for a hotspot in the Carolinas. Geol. Soc. Am. Bull. 90, 185–198. Dennis, A.J., Shervais, J.W., Mauldin, J., Maher Jr., H.D., Wright, J.E., 2004. Petrology and geochemistry of Neoproterozoic volcanic arc terranes beneath the Atlantic Coastal Plain, Savannah River Site, South Carolina. Geol. Soc. Am. Bull. 116, 572–593. De Paolo, D.J., 1981. Trace element and isotopic effects on combined wallrock assimilation and fractional crystallization. Earth Planet. Sci. Lett. 53, 189–202. Dorais, M.J., Workman, J., Aggarwal, J., 2008. The petrogenesis of the Highlandcroft and Oliverian plutonic suites, New Hampshire: Implications for the structure of the Bronson Hill terrane. Am. J. Sci. 308, 73–99. Douglass, J., Schilling, J.-G., 2000. Systematics of three-component, pseudo-binary mixing lines in 2D isotope ratio space representations and implications for mantle plume–ridge interaction. Chem. Geol. 163, 1–23. Esawi, E.K., 2004. Evidence from the Farmington pluton for Early Devonian subduction-related magmatism in the Carolina zone of central North Carolina. J. Geodyn. 37, 531–548. Ghiorso, M.S., Sack, R.O., 1995. Chemical mass transfer in magmatic processes, IV. A revised and internally consistent thermodynamic model for the interpolation and extrapolation of liquid–solid equilibria in magmatic systems at elevated temperatures and pressures. Contrib. Mineral. Petrol. 119, 197–212. Gibson, S.A., Thompson, R.N., Day, J.A., Humphris, S.E., Dickin, A.P., 2005. Meltgeneration processes associated with the Tristan mantle plume: Constraints on the origin of EM-1. Earth Planet. Sci. Lett. 237, 744–767. Goldberg, S.A., Butler, J.R., Fullagar, P.D., 1986. The Bakersville dike swarm; geochronology and petrogenesis of late Proterozoic basaltic magmatism in the southern Appalachian Blue Ridge. Am. J. Sci. 286, 403–430. Hames, W.E., Renne, P.R., Ruppel, C., 2000. New evidence for geologically instantaneous emplacement of earliest Jurassic Central Atlantic magmatic province basalts on the North American margin. Geology 28, 859–862. Hanan, B.B., Graham, D.W., 1996. Lead and helium isotope evidence from oceanic basalts for a common deep source of mantle plumes. Science 272, 991–995. Harper, S.B., Fullagar, P.D., 1981. Rb–Sr ages of granitic gneisses of the Inner Piedmont belt of northwestern North Carolina and southwestern South Carolina. Geol. Soc. Am. Bull. 92, 864–872. Hart, S.R., 1984. A large-scale isotope anomaly in the Southern Hemisphere mantle. Nature 309, 753–757. Hart, S.R., Gerlach, D.C., White, W.M., 1986. A possible new Sr–Nd–Pb mantle array and consequences for mantle mixing. Geochim. Cosmochim. Acta 50, 1551–1557. Harvey, J., Dale, C.W., Gannoun, A., Burton, K.W., 2011. Osmium mass balance in peridotite and the effects of mantle-derived sulphides on basalt petrogenesis. Geochim. Cosmochim. Acta 75, 5574–5596. Hatcher, R.D., 2010. The Appalachian orogen: A brief summary. Geol. Soc. Am. Mem. 206, 1–19. Heatherington, A.L., Mueller, P.A., 1999. Lithospheric sources of North Florida, USA tholeiites and implications for the origin of the Suwannee terrain. Lithos 46, 215–233. Heffner, D.M., Knapp, J.H., Akintunde, O.M., Knapp, C.C., 2012. Preserved extent of Jurassic flood basalt in the South Georgia Rift: A new interpretation of the J horizon. Geology 40, 167–170. Heinonen, J.S., Carlson, R.W., Luttinen, A.V., 2010. Isotopic (Sr, Nd, Pb, and Os) composition of highly magnesian dikes of Vestfjella, western Dronning Maud Land, Antarctica: A key to the origins of the Jurassic Karoo large igneous province. Chem. Geol. 277, 227–244. Herzberg, C., Asimow, P.D., 2008. Petrology of some oceanic island basalts: PRIMELT2.XLS software for primary magma calculation. Geochem. Geophys. Geosyst. 9, http://dx.doi.org/10.1029/2008GCQ09001. Herzberg, C., Gazel, E., 2009. Petrological evidence for secular cooling in mantle plumes. Nature 458, 619–623. Hibbard, J.P., Stoddard, E.F., Secor, D.T., Dennis, A.J., 2002. The Carolina Zone: Overview of Neoproterozoic to Early Paleozoic peri-Gondwanan terranes along the eastern flank of the southern Appalachians. Earth-Sci. Rev. 57, 299–339. Hibbard, J.P., van Staal, C.R., Rankin, D.W., 2010. Comparative analysis of the geological evolution of the northern and southern Appalachian orogen: Late Ordovician–Permian. Geol. Soc. Am. Mem. 206, 51–69. Hill, R.I., 1991. Starting plumes and continental break-up. Earth Planet. Sci. Lett. 104, 398–416. Hirose, K., Kushiro, I., 1993. Partial melting of dry peridotites at high pressures: Determination of compositions of melts segregated from peridotite using aggregates of diamond. Earth Planet. Sci. Lett. 114, 477–489.

198

S. Callegaro et al. / Earth and Planetary Science Letters 376 (2013) 186–199

Hoernle, K., 1998. Geochemistry of Jurassic Oceanic Crust beneath Gran Canaria (Canary Islands): Implications for crustal recycling and assimilation. J. Petrol. 39, 859–880. Hofmann, A.W., 2003. Sampling mantle heterogeneity through oceanic basalts: Isotopes and trace elements. In: Holland, H.D., Turekian, K.K. (Eds.), Treatise on Geochemistry, the Mantle and Core. Elsevier, Oxford, UK, pp. 61–101. Holbrook, W.S., Kelemen, P.B., 1993. Large igneous province on the U.S. Atlantic margin and implications for magmatism during continental breakup. Nature 364, 433–436. Jackson, M.G., Carlson, R.W., 2011. An ancient recipe for flood-basalt genesis. Nature 476, 316–319. Janney, P.E., Castillo, P.R., 2001. Geochemistry of the oldest Atlantic crust suggests mantle plume involvement in the early history of the Central Atlantic Ocean. Earth Planet. Sci. Lett. 192, 291–302. Johnston, A.D., Schwab, B.E., 2004. Constraints on clinopyroxene/melt partitioning of REE, Rb, Sr, Ti, Cr, Zr, and Nb during mantle melting: First insights from direct peridotite melting experiments at 1.0 GPa. Geochim. Cosmochim. Acta 68, 4949–4962. Jourdan, F., Bertrand, H., Schäerer, U., Blichert-Toft, J., Feraud, G., Kampunzu, A.B., 2007. Major and trace element and Sr, Nd, Hf, and Pb isotope compositions of the Karoo large igneous province, Botswana–Zimbabwe: lithosphere vs mantle plume contribution. J. Petrol. 48, 1043–1077. Jourdan, F., Marzoli, A., Bertrand, H., Cirilli, S., Tanner, L.H., Kontak, D.J., McHone, G., Renne, P.R., Bellieni, G., 2009. 40 Ar/39 Ar ages of CAMP in North America: Implications for the Triassic–Jurassic boundary and the 40 K decay constant bias. Lithos 110, 167–180. Kelemen, P.B., Shimizu, N., Dunn, T., 1993. Relative depletion of niobium in some arc magmas and the continental crust: Partitioning of K, Nb, La and Ce during melt/rock reaction in the upper mantle. Earth Planet. Sci. Lett. 120, 111–134. Kerr, A.C., Kempton, P.D., Thompson, R.N., 1995. Crustal assimilation during turbulent magma ascent (ATA); new isotopic evidence from the Mull Tertiary lava succession, N. W. Scotland. Contrib. Mineral. Petrol. 119, 142–154. Keshav, S., Gudfinnsson, G.H., Sen, G., Fei, Y., 2004. High-pressure melting experiments on garnet clinopyroxenite and the alkalic to tholeiitic transition in oceanisland basalts. Earth Planet. Sci. Lett. 223, 365–379. LeMaitre, R.W., Streckeisen, A., Zanettin, B., Le Bas, M.J., Bonin, B., Bateman, P., Bellieni, G., Dudek, A., Efremova, S., Keller, J., Lamere, J., Sabine, P.A., Schmid, R., Sorensen, H., Woolley, A.R. (Eds.), 2002. Igneous Rocks: A Classification and Glossary of Terms, Recommendations of the International Union of Geological Sciences Subcommission of the Systematics of Igneous Rocks. Cambridge University Press, UK. Levin, V., Park, J., Brandon, M.T., Menke, W., 2000. Thinning of the upper mantle during late Paleozoic Appalachian orogenesis. Geology 28, 239–242. Li, Z.X., Bogdanova, S.V., Collins, A.S., Davidson, A., De Waele, B., Ernst, R.E., Fitzsimons, I.C.W., Fuck, R.A., Gladkochub, D.P., Jacobs, J., Karlstrom, K.E., Lu, S., Natapov, L.M., Pease, V., Pisarevsky, S.A., Thrane, K., Vernikovsky, V., 2008. Assembly, configuration, and break-up history of Rodinia: A synthesis. Precambrian Res. 160, 179–210. Longhi, J., 2002. Some phase equilibrium systematics of lherzolite melting: I. Geochem. Geophys. Geosyst. 3, Q11K11, http://dx.doi.org/10.1029/ 2001GC000204. Ludwig, K.R., 2003. User’s Manual for Isoplot 3.00. Berkeley Geochronology Center Spec. Publ. 74 pp. Marzoli, A., Renne, P.R., Piccirillo, E.M., Ernesto, M., Bellieni, G., De Min, A., 1999. Extensive 200-million-year-old continental flood basalts of the Central Atlantic magmatic province. Science 284, 616–618. Marzoli, A., Jourdan, F., Puffer, J.H., Cuppone, T., Tanner, L.H., Weems, R.E., Bertrand, H., Cirilli, S., Bellieni, G., De Min, A., 2011. Timing and duration of the Central Atlantic magmatic province in the Newark and Culpeper basins, eastern U.S.A. Lithos 122, 175–188. May, P.R., 1971. Pattern of Triassic diabase dikes around the North Atlantic in the context of predrift position of the continents. Geol. Soc. Am. Bull. 82, 1285–1292. McDonough, W.F., Sun, S.-S., 1995. The composition of the Earth. Chem. Geol. 120, 223–254. McHone, J.G., 2000. Non-plume magmatism and rifting during the opening of the central Atlantic Ocean. Tectonophysics 316, 287–296. McKenzie, D., Bickle, M.J., 1988. The volume and composition of melt generated by extension of the lithosphere. J. Petrol. 29, 625–679. McKenzie, D., O’Nions, R.K., 1991. Partial melt distributions from inversion of rare Earth element concentrations. J. Petrol. 32, 1021–1091. McLelland, J.M., Selleck, B.W., Bickford, M.E., 2010. Review of the Proterozoic evolution of the Grenville Province, its Adirondack outlier, and the Mesoproterozoic inliers of the Appalachians. In: Tollo, R.P., Bartholomew, M.J., Hibbard, J.P., Karabinas, P.M. (Eds.), From Rodinia to Pangea: The Lithotectonic Record of the Appalachian Region. Geol. Soc. Am. Mem. 206, 1–29. Meisel, T., Walker, R.J., Irving, A.J., Lorand, J.-P., 2001. Osmium isotopic compositions of mantle xenoliths: A global perspective. Geochim. Cosmochim. Acta 65, 1311–1323. Merle, R., Marzoli, A., Bertrand, H., Reisberg, L., Verati, C., Zimmermann, C., Chiaradia, M., Bellieni, G., Ernesto, M., 2011a. 40 Ar/39 Ar ages and Sr–Nd–Pb–Os

geochemistry of CAMP tholeiites from Western Maranhão basin (NE Brazil). Lithos 122, 137–151. Merle, R., Marzoli, A., Bertrand, H., Reisberg, L., Chiaradia, M., Bellieni, G., 2011b. Sr–Nd–Pb–Os isotopes of CAMP tholeiites from northeast America. Goldschmidt Conference Abstracts 2011, 1457. Meyzen, C.M., Ludden, J.N., Humler, E., Luais, B., Toplis, M.J., Mével, C., Storey, M., 2005. New insights into the origin and distribution of the DUPAL isotope anomaly in the Indian Ocean mantle from MORB of the Southwest Indian Ridge. Geochem. Geophys. Geosyst. 6, Q11K11, http://dx.doi.org/10.1029/2005GC000979. Nelson, K.D., 1992. Are crustal thickness variations in old mountain belts like the Appalachians a consequence of lithospheric delamination? Geology 20, 498–502. Nomade, S., Knight, K.B., Beutel, E., Renne, P.R., Verati, C., Féraud, G., Marzoli, A., Youbi, N., Bertrand, H., 2007. Chronology of the Central Atlantic Magmatic Province: Implications for the Central Atlantic rifting processes and the Triassic– Jurassic biotic crisis. Palaeogeogr. Palaeoclimatol. Palaeoecol. 244, 326–344. Pegram, W.J., 1990. Development of continental lithospheric mantle as reflected in the chemistry of the Mesozoic Appalachian Tholeiites, U.S.A. Earth Planet. Sci. Lett. 97, 316–331. Pe-Piper, G., Jansa, L.F., 1999. Pre-Mesozoic basement rocks offshore Nova Scotia, Canada: New constraints on the accretion history of the Meguma terrane. Geol. Soc. Am. Bull. 111, 1773–1791. Peucker-Ehrenbrink, B., Jahn, B., 2001. Rhenium–osmium isotope systematics and platinum group element concentrations: Loess and the upper continental crust. Geochem. Geophys. Geosyst. 2. Plank, T., 2005. Constraints from thorium/lanthanum on sediment recycling at subduction zones and the evolution of the continents. J. Petrol. 46, 921–944. Puffer, J.H., 1992. Eastern North American flood basalts in the context of the incipient breakup of Pangea. In: Puffer, J.H., Ragland, P.C. (Eds.), Eastern North American Mesozoic Magmatism. Geol. Soc. Am. Sp. Paper 268, 95–119. Puffer, J.H., 2001. Contrasting high field strength element contents of continental flood basalts from plume versus reactivated-arc sources. Geology 29, 675–678. Putirka, K., 2008. Thermometers and barometers for volcanic systems. In: Putirka, K., Tepley, F. (Eds.), Minerals, Inclusions and Volcanic Processes. Rev. Mineral. Geochem., Mineral. Soc. Am. 69, 61–120. Putirka, K., Ryerson, F.J., Perfit, M., Ridley, W.I., 2011. Mineralogy and composition of the oceanic mantle. J. Petrol. 52, 279–313. Ragland, P.C., Hatcher Jr., R.D., Whittington, D., 1983. Juxtaposed Mesozoic diabase dike sets from the Carolinas: A preliminary assessment. Geology 11, 394–399. Renne, P.R., Mundil, R., Balco, G., Min, K., Ludwig, K.R., 2010. Joint determination of 40 K decay constants and 40 Ar∗ /40 K for the Fish Canyon sanidine standard, and improved accuracy for 40 Ar/39 Ar geochronology. Geochim. Cosmochim. Acta 74, 5349–5367. Rocha-Junior, E.R.V., Puchtel, I.S., Marques, L.S., Walker, R.J., Machado, F.B., Nardy, A.J.R., Babinski, M., Figueiredo, A.M.G., 2012. Re–Os isotope and highly siderophile element systematics of the Paranà Continental flood basalts (Brazil). Earth Planet. Sci. Lett. 337–338, 164–173. Roeder, P.L., Emslie, R.F., 1970. Olivine–liquid equilibrium. Contrib. Mineral. Petrol. 29, 275–289. Rudnick, R.L., Fountain, D.M., 1995. Nature and composition of the continental crust—a lower crustal perspective. Rev. Geophys. 33, 267–309. Saal, A.E., Runick, R.L., Ravizza, G.E., Hart, S.R., 1998. Re–Os isotope evidence for the composition, formation and age of the lower continental crust. Nature 393, 58–61. Salters, V.J.M., Longhi, J., 1999. Trace element partitioning during the initial stages of melting beneath mid-ocean ridges. Earth Planet. Sci. Lett. 166, 15–30. Salters, V.J.M., Stracke, A., 2004. The composition of the depleted mantle. Geochem. Geophys. Geosyst. 5, http://dx.doi.org/10.1029/2003GC000597. Samson, S.D., Coler, D.G., Speer, J.A., 1995a. Geochemical and Nd–Sr–Pb isotopic composition of Alleghanian granites of the southern Appalachians: Origin, tectonic setting, and source characterization. Earth Planet. Sci. Lett. 134, 359–376. Samson, S.D., Hibbard, J.P., Wortman, G.L., 1995b. Nd isotopic evidence for juvenile crust in the Carolina terrane, southern Appalachians. Contrib. Mineral. Petrol. 121, 171–184. Schlische, R.W., Withjack, M.O., Olsen, P.E., 2003. Relative timing of CAMP, rifting, continental breakup, and basin inversion: Tectonic significance. In: Hames, W.E., McHone, J.G., Renne, P.R., Ruppel, C. (Eds.), The Central Atlantic Magmatic Province: Insights from Fragments of Pangea. In: AGU Geophys. Monogr., vol. 136, pp. 1–15. Scotese, C.R., Bambach, R.K., Barton, C., Van der Voo, R., Ziegler, F.A., 1979. Paleozoic base maps. J. Geol. 87, 217–277. Shirey, S.B., Walker, R.J., 1998. The Re–Os isotope system in cosmochemistry and high-temperature geochemistry. Ann. Rev. Earth Planet. Sci. Lett. 26, 423–500. Sinha, A.K., Zietz, I., 1982. Geophysical and geochemical evidence for a Hercynian magmatic arc, Maryland to Georgia. Geology 10, 593–596. Sobolev, A.V., Hofmann, A.W., Kuzmin, D.V., Yaxley, G.M., Arndt, N.T., Chung, S.-L., Danyushevsky, L.V., Elliott, T., Frey, F.A., Garcia, M.O., Gurenko, A.A., Kamenetsky, V.S., Kerr, A.C., Krivolutskaya, N.A., Matvienkov, V.V. , Nikogosian, I.K., Rocholl, A.,

S. Callegaro et al. / Earth and Planetary Science Letters 376 (2013) 186–199

Sigurdsson, I.A., Sushchevskaya, N.M., Teklay, M., 2007. The amount of recycled crust in sources of mantle-derived melts. Science 316, 412–417. Sobolev, S.V., Sobolev, A.V., Kuzmin, D.V., Krivolutskaya, N.A., Petrunin, A.G., Arndt, N.T., Radko, V.A., Vasiliev, Y.R., 2011. Linking mantle plumes, large igneous provinces and environmental catastrophes. Nature 477, 312–316. Spera, F.J., Bohrson, W.A., 2001. Energy-constrained open-system magmatic processes I: General model and energy-constrained assimilation and fractional crystallization (EC-AFC) formulation. J. Petrol. 42, 999–1018. Stern, C.R., 2011. Subduction erosion: Rates, mechanisms, and its role in arc magmatism and the evolution of the continental crust and mantle. Gondwana Res. 20, 284–308. Taylor, S.R., McLennan, S.M., 1995. The geochemical evolution of the continental crust. Rev. Geophys. 33, 241–265. Tollo, R.P., Gottfried, D., 1992. Petrochemistry of Jurassic basalt from eight cores, Newark basin, New Jersey. In: Puffer, J.H., Ragland, P.C. (Eds.), Eastern North American Mesozoic Magmatism. Geol. Soc. Am. Sp. Paper 268, 233–260. Van Staal, C.R., Whalen, J.B., Valverde-Vaquero, P., Zagorevski, A., Rogers, N., 2009. Pre-carboniferous, episodic accretion-related, orogenesis along the Laurentian margin of the northern Appalachians. In: Murphy, J.B., Keppie, J.D., Hynes, A.J. (Eds.), Ancient Orogens and Modern Analogues. In: Geol. Soc. London, Special Publications, vol. 327, pp. 271–316.

199

Verati, C., Bertrand, H., Féraud, G., 2005. The farthest record of the Central Atlantic Magmatic Province into West Africa craton: Precise 40 Ar/39 Ar dating and geochemistry of Taoudenni basin intrusives (northern Mali). Earth Planet. Sci. Lett. 235, 391–407. Willbold, M., Stracke, A., 2006. Trace element composition of mantle end-members: Implications for recycling of oceanic and upper and lower continental crust. Geochem. Geophys. Geosyst. 7, Q04004, http://dx.doi.org/10.1029/2005GC001005. Willbold, M., Stracke, A., 2010. Formation of enriched mantle components by recycling of upper and lower continental crust. Chem. Geol. 276, 188–197. Wilson, M., 1997. Thermal evolution of the Central Atlantic passive margins: continental break-up above a Mesozoic superplume. J. Geol. Soc. London 154, 491–495. Workman, R.K., Hart, S.R., 2005. Major and trace element composition of the depleted MORB mantle (DMM). Earth Planet. Sci. Lett. 231, 53–72. Zartman, R.E., Kempton, P.D., Paces, J.B., Downes, H., Williams, I.S., Dobosi, G., Futal, K., 2013. Lower-crustal xenoliths from Jurassic kimberlite diatremes, Upper Michigan (USA): Evidence for Proterozoic Orogenesis and plume magmatism in the lower crust of the Southern Superior Province. J. Petrol. 54, 575–608. Zindler, A., Hart, S.R., 1986. Chemical geodynamics. Ann. Rev. Earth Planet. Sci. 14, 493–571.