International Journal of Earth Sciences

, Volume 99, Issue 8, pp 1773–1790

SHRIMP U–Pb zircon geochronology and Sr–Nd isotopic systematic of the Neoproterozoic Ghimbi-Nedjo mafic to intermediate intrusions of Western Ethiopia: a record of passive margin magmatism at 855 Ma?

Authors

    • Department of GeoscienceShimane University
    • Chesapeake College
  • Jun-Ichi Kimura
    • Institute for Research on Earth Evolution (IFREE)Japan Agency for Marine-Earth Science and Technology (JAMSTEC)
  • Daniel J. Dunkley
    • National Institute of Polar Research (NIPR)
  • Kenichiro Tani
    • Institute for Research on Earth Evolution (IFREE)Japan Agency for Marine-Earth Science and Technology (JAMSTEC)
  • Hiroto Ohira
    • Department of GeoscienceShimane University
Original Paper

DOI: 10.1007/s00531-009-0481-x

Cite this article as:
Woldemichael, B.W., Kimura, J., Dunkley, D.J. et al. Int J Earth Sci (Geol Rundsch) (2010) 99: 1773. doi:10.1007/s00531-009-0481-x

Abstract

The reworked Pre-Neoproterozoic and juvenile Neoproterozoic terrane of the Western Ethiopian Shield (WES) consists of three N–S trending terranes. These are the western migmatitic gneissic terrane, the central metavolcano sedimentary terrane (CVST) and the eastern migmatitic gneissic terrane. The eastern part of the CVST mostly consists of suture-related ultramafic-metasedimentary complexes, whereas metavolcanics predominate in the western part. Gabbroic to granitic intrusions frequently occur in the CVST and in adjacent areas. New zircon SHRIMP U–Pb ages for two gabbros and three diorites in the Ghimbi-Nedjo region of the WES indicate magmatic crystallization ages. Two pulses of magmatism, at 860–850 and 795–785 Ma, are documented with the former for the first time. The tholeiitic Kemashi diorite and Bikilal-Ghimbi gabbros have oceanic affinities and yield U/Pb zircon ages of 856.3 ± 9.8 and 846.0 ± 7.6 Ma, respectively. The calc-alkaline Gebeya Kemisa pyroxene diorite, and the Senbet Dura hornblende diorite plus the tholeiitic Wayu Meni gabbro, which collectively have arc-back arc characteristics are indistinguishable at ages of 794.3 ± 9.4, 787.7 ± 8.8 and 778.1 ± 6.3 Ma, respectively. Positive εNd (4.5–7.0) and low initial 87Sr/86Sr (0.7029 ± 0.0002) and a mean TDM model age of 0.95 Ga for the Ghimbi-Nedjo region (mean TDM model age of 0.95 Ga for the WES overall) indicate that the magmas were generated from juvenile Neoproterozoic depleted mantle sources, with no discernable involvement of pre-Neoproterozoic continental crust. The occurrence of gabbros and diorites with oceanic tholeiite affinities combined with the new ages suggests that the intrusions were emplaced in the earliest stages of the rifting of Rodinia. This event in the WES led to the development of a passive margin and associated plume-type magmatism at ~855 Ma. The two intrusive groups with differing magma chemistry and ages suggest that the earliest magmatism was tholeiitic and associated with the passive margin system followed by continental breakup to form the Mozambique Ocean. The combination of tholeiitic and calc-alkaline magmatism was related to arc and back-arc basin formation and later terrane accretion (~830–690 Ma).

Keywords

SHRIMP U–Pb datingIntrusionsNeoproterozoicPassive marginWestern Ethiopian ShieldEAAO

Introduction

The Neoproterozoic East African–Antarctic Orogen (EAAO) in Western Ethiopia occupies a position of particular significance, as it lies in the transition between the Arabian–Nubian Shield (ANS: low-grade metavolcanics and metasediments) in the north, the Mozambique Belt (MB: predominantly high-grade gneiss) in the south, and the Saharan metacraton in the west (Fig. 1a). The terrane is also associated with mafic–ultramafic belts of suture zones and syn- to post-tectonic gabbroic to granitic intrusions. The Ghimbi-Nedjo region is part of the Western Ethiopian Shield (WES) where an N–S trending suture zone consisting of metavolcanic, metasedimentary and ultramafic complexes is bounded by gneissic terranes to the east and to the west (e.g., Abraham 1989; Abdelsalam and Stern 1996; Allen and Tadesse 2003). The field relations, petrography, major and trace element geochemistry and genesis of the mafic to intermediate intrusions of the Ghimb-Nedjo region were recently studied by Woldemichael and Kimura (2008a, b), who interpreted that the Bikilal-Ghimbi gabbro body had intraplate-type tholeiitic parental magma, suggesting plume-type magmatism. However, the ages and isotopic characteristics of the Ghimbi-Nedjo region plutons, especially the mafic to intermediate intrusions, have not been studied to date. Comprehensive and systematic studies of the WES are lacking and those available mostly deal with felsic intrusions or granitic rocks (Ayalew et al. 1990; Kebede et al. 2001a; Grenne et al. 2003); these studies have recognized three generations of magmatism (~830–810 Ma; ~780–700 and ~620–550 Ma) in a subduction-related island arc and intraplate post-collisional environments.
https://static-content.springer.com/image/art%3A10.1007%2Fs00531-009-0481-x/MediaObjects/531_2009_481_Fig1_HTML.gif
Fig. 1

a Map of the Arabian–Nubian Shield (ANS; dashed ellipsoid, modified from Stern et al. 2006) showing the location of the study areas and regions where pre-Neoproterozoic crust is found. Ages for pre-Neoproterozoic crustal tracts are from Agar et al. (1992), Sultan et al. (1994), Stern (1994), Kröner and Sassi (1996), Whitehouse et al. (1998). Part of the Mozambique Belt (MB) is indicated. b Simplified geological map of the Western Ethiopian Shield (WES; modified after Tefera and Berhe 1987; Tefera 1991; Tefera et al. 1996; Tadesse and Allen 2004; Woldemichael and Kimura 2008a). Ages are from Ghimbi-Nedjo (this work); Kilaj-Wombera (Grenne et al. 2003); Ghimbi (Kebede et al. 2001a, 2007) and Gore-Gambella (Ayalew et al. 1990). EMGT eastern migmatitic high-grade gneissic terrane, WMGT western migmatitic high-grade gneissic terrane, CVST central metavolcano-sedimentary terrane

This study investigates the ages and Sr–Nd isotopic compositions of gabbroic to dioritic plutons of the Ghimbi-Nedjo region, the central part of the WES. For this, we have carried out Sr–Nd isotopic analyses and SHRIMP dating of zircons from five WES plutons: the Wayu Meni olivine gabbro, Senbet Dura hornblende diorite, Gebeya Kemisa pyroxene diorite, Kemashi hornblende diorite and the Bikilal-Ghimbi hornblende gabbro (Fig. 2). Combined with local and regional geological, geochemical and geochronological data, the results shed light on the Neoproterozoic geological evolution of the WES and the EAAO.
https://static-content.springer.com/image/art%3A10.1007%2Fs00531-009-0481-x/MediaObjects/531_2009_481_Fig2_HTML.gif
Fig. 2

Geological map of part of the WES in the Ghimbi region with all five plutons indicated (modified after Alemu and Abebe 2000; Allen and Tadesse 2003)

Regional geology and tectonic setting

The term Pan-African originally referred to a sequence of tectonothermal events at 500 ± 100 Ma within Africa and Gondwana (Kennedy 1964). Kröner (1984) later included orogenic events in the time range 950–450 Ma, and identified the Arabian–Nubian Shield (ANS) as one of the major orogenic belts formed during Neoproterozoic time. A very early Pan-African orogenic event called Bayudian event (920–900 Ma; Küster et al. 2008) is recorded at the western boundary of the ANS, i.e., at the eastern boundary of the Saharan Metacraton. The locus of collision between east and west Gondwana is marked by the East African Orogen (EAO: Stern 1994; Meert 2003). The southern EAO continues through the largely metamorphosed and reworked Neoproterozoic crust south through a reconstructed Gondwana into Antarctica, such that (Jacobs and Thomas 2004; Stern 2008) renamed the entire ~8,000 km long belt as the East Africa–Antarctic Orogen (EAAO). Pan-African rocks are globally widespread, but are especially common in Africa (Stern 2008), and are particularly important in the EAAO. The EAAO formed during the ‘Pan-African’ assembly of Gondwana, and includes complex intra-oceanic and continental margin igneous rocks that were deformed after a protracted sequence of tectonothermal events (Stern 1994). The EAAO stretches south from Israel and Jordan through to Tanzania and into Antarctica (Stern 1994, 2008; Jacobs and Thomas 2004). The transition between the juvenile ANS and MB is located within Ethiopia, Eritrea, Sudan and Somalia.

Two major crustal types are identified in the EAAO. The juvenile (i.e., mantle derived) ANS in the north is dominated by low-grade volcano-sedimentary rocks in association with plutons and ophiolitic remnants (Fig. 1a). The second crustal type is a tract of older remobilized crust to the south of the ANS, known as the Mozambique Belt (MB; Fig. 1a). The transition between the juvenile ANS and MB is located within Ethiopia, Eritrea, Sudan and Somalia. The WES (Fig. 1b) lies near the transition between ANS and MB, and also adjacent to the enigmatic “East Saharan Metacraton”, which consists of older crust that was extensively remobilized during the Neoproterozoic time (Abdelsalam et al. 2002).

The tectonic evolution of the WES (Fig. 1b) overall has been interpreted in terms of early rifting and associated sedimentation, followed by subduction and island-arc formation, arc-accretion and finally protracted continent–continent (East and West Gondwana) collision (e.g., Kazmin et al. 1978; Stern 1994). Linear belts of highly deformed mafic–ultramafic bodies within low-grade (ANS-like) terranes have been interpreted as suture zones containing dismembered ophiolitic rocks (e.g., Kazmin et al. 1978; Berhe 1990; Ayalew et al. 1990; Abdelsalam and Stern 1996; Allen and Tadesse 2003; Tadesse and Allen 2005). Tadesse and Allen (2005) named this dismembered ophiolites as the Tuludimtu Ophiolite. However, unequivocal evidences for dismembered ophiolites are often lacking in the WES. In contrast, the ultramafic bodies have been regarded as an Alaskan-type (Mogessie et al. 2000) concentrically zoned intrusions emplaced into an extensional arc or back-arc environment (Braathen et al. 2001; Grenne et al. 2003). The southern part of the WES grades into the lithologically similar Mozambique Belt rocks of NW Kenya. However, the general lack of pre-Neoproterozoic ages and the scarcity of granulite-facies metamorphic relicts in the WES suggest that the relationship between the WES gneissic terranes and those of the Mozambique Belt is complex (Johnson et al. 2004).

Although several lithologically and structurally distinct domains are recognized within the WES (Table 1), the division by Woldemichael and Kimura (2008b) is adopted here. They broadly subdivided the region into three terranes: the eastern migmatitic high-grade gneissic terrane (EMGT), the western migmatitic high-grade gneissic terrane (WMGT), and the central metavolcano-sedimentary terrane (CVST; Fig. 1b). EMGT basically corresponds to the Geba domain of Tefera and Berhe (1987), the Aba Sina domain of Alemu and Abebe (2000) and the Didesa domain of Allen and Tadesse (2003). The CVST mostly correlates with the Birbir domain of Tefera and Berhe (1987), the Katta domain of Alemu and Abebe (2000) and the Kemashi and Dengi domains of Allen and Tadesse (2003). The WMGT corresponds to the gneissic parts of the Birbir domain of Tefera and Berhe (1987), the Katta domain of Alemu and Abebe (2000) and the Jamoa-Ganti block of the Dengi domain of Allen and Tadesse (2003). Table 1 (modified after Alemu 2005) summarizes the various names and inconsistencies of the domains and terranes proposed by several authors for the WES. The CVST is further divided into an eastern zone dominated by metasediment-ultramafic suture-related rocks (the Kemashi domain of Allen and Tadesse 2003) and a western zone dominated by volcanic arc-related metavolcanics (Dengi domain of Allen and Tadesse 2003). The eastern CVST is a relatively narrow N–S trending strip of ultramafic and mafic volcanic rocks and related plutons. These are associated with marine sediments and thus are considered to be dismembered ophiolite (Tadesse and Allen 2005). The eastern CVST has been interpreted as a suture zone by several authors (e.g., Kazmin et al. 1978; Berhe 1990; Tadesse and Allen 2005). EMGT consists of migmatitic biotite to biotite–hornblende granitoid orthogneiss, whereas WMGT is dominated by biotite to biotite–hornblende gneiss (Alemu and Abebe 2000). Early-, syn-, late-, post-tectonic gabbroic to dioritic intrusions occur in all of the three terranes (Figs. 1b, 2).
Table 1

Litho-tectonic divisions of the Precambrian rocks of Western Ethiopia, modified after Alemu 2005

Litho-tectonic units

UNDP (1972)

Kazmin et al. (1979)

Davidson (1983)

Tefera and Berhe (1987)

Tefera (1991)

Tefera et al.(1996)

Alemu and Abebe (2000)

Seyid (2002)

Allen and Tadesse (2003)

This study and Woldemichael and Kimura (2008b)

Low-grade volcanic sediments and associated intrusive rocks

Birbir Group: low-grade metavolcanic sediments

Zone IV: metavolcanic metasedimentary belt

Akobo Domain: low-grade metavolcanic sediments and associated mafic and ultramafic rocks

Birbir Domain: low-grade metavolcanic sediments and associated mafic to felsic rocks

Central domain: metavolcanic sediments and associated mafic to intermediate intrusives

Tulu Dimtu and birbir Group: low-grade metavolcano-sedimentary rocks and associated mafic-ultramafics

Katta Domain: low-grade metavolcano-sedimentary rocks and associated intrusives

Aja Domain: metavolcano-sedimentary schist

Kemashi and Dengi Domains: ophiolitic–ultramafic–mafic volcanic-intrusive rocks with sedimentary rocks

CVST: Central Volcano Sedimentary Terrane. further subdivided as eastern CVST and western CVST

Metavolcanic group: metabasalt and amphibolites with minor gabbro and ultramafics

Zone III: dioritic granodioritic batholiths and associated metavolcanics

  

Gesengesa Domain: greenschist to amphibolite grade metasedimentary gneiss and schist

 

Afa and Chochi Domain: greenschist to amphibolites grade metasedimentary gneiss and schist

Beles Domain: metavolcano-sedimentary rocks with slivers of gneiss

Sirkole Domain: alternating gneiss and metavolcano-sedimentary sequences, and plutons

 
 

Zone II: ophiolitic belt

        

High-grade gneiss and migmatites

High-grade gneiss and schist

Zone V: western block of high-grade pre-Pan-African basement

Surma Domain: foliated and sheared granitoid gneiss

Baro Domain: high-grade dominantly supercrustal gneiss and schist

Kurmuk Domain: ortho and para gneiss

Baro Group: ortho and para gneiss

 

Guba Domain: ortho and para gneiss

Didessa Domain: moderate-grade gneisses and gabbros to granitoids plutons

EMGT: eastern migmatitic high-grade gneissic terrane

Zone I: eastern block of high-grade pre-Pan-African basement

Hamar Domain: gneiss and migmatites with minor granulites

Geba Domain: orthogneiss and migmatites

 

Alghe Group: orthogneiss and migmatites

Aba Sina Domain: high-grade gneiss and migmatites

 

Daka Domain: moderate-high-grade gneisses and granitoids

WMGT: western migmatitic high-grade gneissic terrane

Studies of the geochronology and the geochemistry of WES plutonic rocks are limited and are based mainly on granitic rocks. Those available give Neoproterozoic ages, consistent with those obtained for other parts of the ANS. Three generations of plutonism are recognized in the WES (Ayalew et al. 1990; Ayalew and Peccerillo 1998; Kebede et al. 1999; 2001a, b; Grenne et al. 2003). According to these authors, “pre-kinematic” (~830–810 Ma) and “syn-kinematic” plutons (~780–700 Ma) were subduction-related, emplaced in intraoceanic island-arc environments. The late- and post-kinematic plutons (~620–550 Ma) are considered to represent both subduction-related and intra-plate post-collision intrusions. However, we prefer not to use the words “pre-kinematic” or “pre-tectonic”, preferring the term “early-tectonic” for magmatism, which occurred at the earliest stage. We adopt this terminology because the terms “pre-kinematic” or “pre-tectonic” should refer to bodies emplaced before any EAAO tectonic activity. There is no evidence that WES igneous activity occurred prior to any EAAO tectonic event, although the age and affinity of polydeformed gneisses and most WES intrusions were not well studied. Johnson et al. (2004) found that igneous protoliths of EMGT gneisses yielded zircon U–Pb ages of 830–785 Ma, whereas granitoid orthogneiss yielded a zircon SHRIMP age of 776 ± 12 Ma (Kebede et al. 2007).

The field relations, petrography, geochemistry and petrogenesis of the mafic to intermediate intrusions of the study area were described by Woldemichael and Kimura (2008a, b), showing both oceanic island tholeiite (OIT) and arc-back arc basin (BAB) geochemical signatures. The age relationship of these intrusions will be shown in later sections.

Samples and analytical methods

Isotope analysis

Nine bulk rock samples (three olivine gabbros, one hornblende gabbro and five diorites) were analyzed for Sr and Nd isotopes. Sr and Nd element separation procedures followed Izumi et al. (1994, 1995). The reagents used were ultrapure hydrofluoric and nitric acids and precise measurement grade hydrochloric acid. Samples were analyzed using a multiple collector–ICP-MS (Thermo Elemental VG Plasma 54) at Shimane University. SRM987 (for Sr), La Jolla and JMC (for Nd) standards and JB-2 igneous rock standard (for Nd and Sr) were analyzed together with every six samples. Mass fractionation during analyses was corrected for first by internal normalization to canonical stable isotope ratios of 86Sr/88Sr = 0.1194 and 146Nd/144Nd = 0.7219. Further mass fractionation that cannot be corrected for by the conventional method for MC-ICP-MS (e.g., Vance and Thirlwall 2002) was corrected by normalization against SRM987 (87Sr/86Sr = 0.71024) and JMC (143Nd/144Nd = 0.512241) using exponential fractionation correction method by Patchett et al. (1981). Results for La Jolla [143Nd/144Nd = 0.511874 ± 15 (n = 5)] and JB-2 [87Sr/86Sr = 0.703772 ± 19 (n = 2), 143Nd/144Nd = 0.513080 ± 15 (n = 2)] agreed well with the conventional thermal ionization mass spectrometry (TIMS) results of Izumi et al. (1994, 1995: 143Nd/144Nd = 0.511850 ± 5 for La Jolla and 87Sr/86Sr = 0.703750 ± 14 and 143Nd/144Nd = 0.513090 ± 6 for JB-2).

SHRIMP U–Pb analytical method

Samples and location

The Ghimbi-Nedjo area in Western Ethiopia is represented by the Ethiopian Mapping Authority 1: 250,000 topographic map sheets number NC36-12. It is located in the UTM zone of 36P. The sampling location is listed as easting and northing coordinate pair in the metric system. Rock samples weighing ~5–7 kg were collected from five plutons: (1) Wayu Meni olivine gabbro (764215 mE, 1013750 mN), which outcrops within the CVST metavolcanics; (2) Senbet Dura hornblende diorite (774953 mE, 1067627 mN) and (3) Gebeya Kemisa pyroxene diorite (782366 mE, 1956912 mN), both from the boundary between the CVST and WMGT; (4) Kemashi hornblende diorite (808404 mE, 1045928 mN) of the CVST metasediment-ultramafic suture zone; and (5) Bikilal-Ghimbi hornblende gabbro (774953 mE, 1067627 mN) of the EMGT. Locations are shown in Fig. 2. The Bikilal-Ghimbi gabbro, Wayu Meni olivine gabbro and Kemashi hornblende diorite have tholeiitic affinities, whereas Senbet Dura diorite and Gebeya Kemisa pyroxene diorite are calc-alkaline (Woldemichael and Kimura 2008a, b). The samples were crushed and panned to separate the coarser mineral fraction, followed by sieving with 300-µm mesh using running water. Magnetic minerals were separated using a hand magnet and a Frantz magnetic separator. Zircon was further concentrated by using heavy liquids (bromoform and diiodomethane) and handpicking using a binocular microscope. Samples from the Bikilal-Ghimbi hornblende gabbro and Senbet Dura hornblende diorite were treated with HF and HNO3 prior to handpicking to digest abundant apatite and sulfides.

Analytical conditions and standards for geochronology

Ages were determined using SHRIMP U–Pb zircon dating techniques. Zircons were selected for dating following cathodoluminescence (CL) imaging (Fig. 3), carried out at the National Institute of Polar Research (NIPR), Tokyo, Japan. U–Th–Pb analyses of zircons were conducted using the SHRIMP II ion microprobe at the NIPR following analytical procedures described by Williams (1998). The spot size of the ion beam was ~25 × 20 µm. Abundance of U was calibrated against zircon standard SL13 (U = 238 ppm) provided by the Australian National University. Sample U–Pb measurements were calibrated against zircon standard FC1 (1,099 Ma; Paces and Miller 1993) using a calibration exponent of 2. Common lead correction estimated from 204Pb using the Stacey and Kramers (1975) model for the approximate age of each analysis was negligible for most analyses. The 204Pb corrected ratios have more spread results than the uncorrected ratios. Therefore, ratios and concordia ages are presented uncorrected for common Pb. Data processing was done with SQUID–ISOPLOT (Ludwig 2003). Detailed analytical data are given in Table 2, and concordia plots are presented in Fig. 4, in which individual analyses are shown with 68.3% confidence levels (1σ uncertainties), whereas both the concordia intercept ages and the weighted mean ages are quoted at the 95% confidence level (2σ uncertainties).
https://static-content.springer.com/image/art%3A10.1007%2Fs00531-009-0481-x/MediaObjects/531_2009_481_Fig3_HTML.jpg
Fig. 3

Cathodoluminescence images of selected zircon grains analyzed by SHRIMP. CL Cathodoluminescence

Table 2

SHRIMP U–Pb–Th analytical data for zircons and determined ages for the intermediate to mafic intrusions of Western Ethiopia

Spot

%206Pbc

U (ppm)

Th (ppm)

232Th/238U

206Pb* (ppm)

206Pb/238UAge

207Pb/206PbAge

% Discordant

238U/206Pb

±%

207Pb/206Pb

±%

207Pb/235U

±%

Err corr

(1)

1σ err

(2)

1σ err

(1)

1σ err

K-1.1

167

79

0.49

20.1

847.9

10.4

847.9

10.8

848

29

0

7.114

1.3

0.06732

1.4

1.305

1.9

0.688

K-2.1

0.11

183

103

0.58

22.3

856.3

10.4

855.4

10.7

883

27

3

7.039

1.3

0.06848

1.3

1.341

1.8

0.702

K-3.1

164

65

0.41

20.2

865.1

10.7

865.4

11.1

855

29

−1

6.963

1.3

0.06757

1.4

1.338

1.9

0.683

K-4.1

0.10

125

62

0.52

15.0

844.5

10.9

843.7

11.3

869

33

3

7.144

1.4

0.06303

1.6

1.313

2.1

0.652

K-5.1

269

153

0.59

33.2

867.3

10.1

867.8

10.4

853

23

−2

6.944

1.2

0.06750

1.1

1.340

1.7

0.746

K-6.1

0.21

122

58

0.49

14.7

843.1

11.0

841.5

11.4

894

39

6

7.157

1.4

0.06883

1.9

1.326

2.3

0.595

BG-1.1

166

54

0.33

20.3

856.6

7.8

857.0

8.1

844

26

−1

7.037

0.97

0.06720

1.2

1.317

1.6

0.616

BG-2.1

115

34

0.30

14.0

852.9

8.9

853.3

9.2

843

30

−1

7.069

1.1

0.06716

1.5

1.310

1.8

0.605

BG-3.1

0.06

92

21

0.23

11.1

841.5

9.0

841.1

9.3

855

36

2

7.171

1.1

0.06757

1.7

1.299

2.1

0.547

BG-4.1

0.38

107

60

0.58

12.0

794.0

8.1

791.2

8.3

888

32

12

7.629

1.1

0.06863

1.6

1.240

1.9

0.569

BG-5.1

120

38

0.33

14.3

842.5

8.2

843.2

8.5

819

30

−3

7.162

1.0

0.06640

1.4

1.278

1.8

0.586

BG-6.1

121

61

0.52

14.7

852.5

8.4

853.3

8.7

828

31

−3

7.073

1.1

0.06667

1.5

1.300

1.8

0.581

BG-7.1

0.25

118

57

0.50

14.2

843.1

8.3

841.1

8.6

904

29

7

7.157

1.1

0.06918

1.4

1.333

1.8

0.593

BG-8.1

58

13

0.24

6.84

829.5

10.2

831.7

10.7

757

45

−9

7.282

1.3

0.06447

2.2

1.221

2.5

0.521

BG-9.1

0.17

59

20

0.34

7.15

843.7

10.4

842.4

10.9

885

44

5

7.151

1.3

0.06855

2.1

1.322

2.5

0.525

BG-10.1

70

24

0.36

8.47

844.2

13.7

844.4

14.1

838

40

−1

7.147

1.7

0.06701

1.9

1.293

2.6

0.666

GK-1.1

0.04

201

151

0.78

22.4

787.5

9.6

787.1

9.9

798

28

1

7.697

1.3

0.06574

1.3

1.178

1.9

0.692

GK-2.1

264

139

0.54

29.8

794.5

9.6

795.9

10.0

746

27

−6

7.624

1.3

0.06412

1.3

1.160

1.8

0.704

GK-3.1

0.05

563

287

0.53

64.3

804.9

9.0

804.6

9.3

817

18

1

7.519

1.2

0.06632

0.86

1.216

1.5

0.810

GK-4.1

450

371

0.85

50.9

797.8

11.7

798.0

12.0

791

21

−1

7.590

1.6

0.06552

1.0

1.190

1.9

0.837

GK-5.1

0.95

354

153

0.45

39.2

782.0

11.0

775.1

11.2

1,009

22

29

7.753

1.5

0.07283

1.1

1.295

1.9

0.802

GK-6.1

0.09

255

184

0.74

28.3

783.1

9.8

782.4

10.1

806

29

3

7.742

1.3

0.06600

1.4

1.175

1.9

0.692

SD-2.1

0.04

192

95

0.51

21.5

788.9

9.5

788.6

9.8

798

28

1

7.682

1.3

0.06575

1.3

1.180

1.8

0.692

SD-3.1

0.00

418

165

0.41

47.7

803.3

8.9

803.3

9.2

803

19

0

7.535

1.2

0.06591

0.89

1.206

1.5

0.796

SD-4.1

0.05

246

143

0.60

26.9

773.8

9.0

773.5

9.3

787

25

2

7.840

1.2

0.06539

1.2

1.150

1.7

0.725

SD-5.1

219

131

0.62

25.1

807.7

9.6

809.3

9.9

751

27

−7

7.492

1.3

0.06428

1.3

1.183

1.8

0.703

SD-6.1

0.08

152

72

0.49

17.1

790.2

9.8

789.6

10.2

811

32

3

7.668

1.3

0.06613

1.5

1.189

2.0

0.658

SD-7.1

0.12

109

62

0.59

11.9

772.7

10.3

771.8

10.6

803

38

4

7.853

1.4

0.06589

1.8

1.157

2.3

0.614

SD-8.1

0.25

223

142

0.66

23.8

755.4

14.8

753.5

15.2

821

27

9

8.044

2.1

0.06645

1.3

1.139

2.4

0.853

SD-9.1

0.24

212

130

0.63

23.2

770.2

9.2

768.4

9.4

831

26

8

7.880

1.3

0.06678

1.3

1.169

1.8

0.709

WM-1.1

0.14

205

113

0.57

22.5

773.5

6.9

772.4

7.1

811

26

5

7.844

0.94

0.06613

1.2

1.162

1.5

0.610

WM-2.1

0.08

394

319

0.84

43.8

784.3

6.2

783.7

6.4

804

18

3

7.730

0.84

0.06592

0.86

1.176

1.2

0.700

WM-3.1

227

103

0.47

24.7

767.3

7.0

768.1

7.2

736

27

−4

7.912

0.97

0.06382

1.3

1.112

1.6

0.606

WM-4.1

187

124

0.69

20.9

787.7

7.1

788.1

7.4

771

26

−2

7.694

0.96

0.06490

1.2

1.163

1.6

0.611

WM-5.1

331

236

0.74

36.4

777.2

6.4

777.5

6.6

764

20

−2

7.805

0.87

0.06470

0.95

1.143

1.3

0.676

WM-6.1

314

167

0.55

34.9

785.3

9.0

786.4

9.2

749

20

−5

7.719

1.2

0.06424

0.97

1.147

1.6

0.781

WM-7.1

621

613

1.02

68.8

781.8

5.9

782.4

6.1

763

15

−2

7.755

0.8

0.06465

0.69

1.149

1.1

0.758

WM-8.1

521

331

0.66

58.1

786.3

14.1

786.5

14.5

777

16

−1

7.709

1.9

0.06509

0.75

1.164

2.0

0.931

Errors are 1σ, Pbc and Pb* indicate the common and radiogenic portions, respectively, as estimated by assuming age concordance

Error in standard calibration was 0.42% for samples K, GK and SD; 0.36% for samples BG and WM (not included in above errors, but required when comparing data from different mounts)

(1) No common Pb correction (2) Common Pb corrected by assuming 206Pb/238U–207Pb/235U age concordance

K Kemashi, BG Bikilal-Ghimbi, GK Gombo-Kora, SD Senbet Dura, WM Wayu Meni

https://static-content.springer.com/image/art%3A10.1007%2Fs00531-009-0481-x/MediaObjects/531_2009_481_Fig4_HTML.gif
Fig. 4

Concordia plots for zircons from Kemashi diorite, Bikilal-Ghimbi hornblende gabbro Gebeya Kemisa pyroxene diorite, Senbet Dura hornblende gabbro and Wayu Meni olivine gabbro. 2σ error ellipses

Results

External morphology and internal structure

Zircons from all intrusions occur as fragments and whole grains. The zircons were fragmented by crushing. The zircons recovered from the Bikilal-Ghimbi hornblende gabbro, Gebeya Kemisa pyroxene diorite and Senbet Dura hornblende diorite are euhedral, whereas those from the Kemashi hornblende diorite are subhedral to euhedral. Euhedral to rounded zircons were extracted from the Wayu Meni olivine gabbro. All zircons have high to moderate cathodoluminescence, show simply graded, sector and oscillatory internal zoning, with little evidence of recrystallization or overgrowth. Several grains from the Wayu Meni gabbro have high cathodoluminescence margins of relatively unzoned zircon that truncate euhedral zoning in grain cores. Analysis points were chosen in areas of moderate cathodoluminescence, in oscillatory-zoned zircon grains or fragments (Fig. 3).

Th and U contents in zircon

Th/U ratios of zircons have been suggested by some authors to be useful as indicators of either magmatic or metamorphic growth (e.g., Williams and Claesson 1987; Maas et al. 1992; Williams et al. 1996; Carson et al. 2002). In general, zircons grown under metamorphic conditions frequently have low Th/U, typically 0.05–0.1 (Williams and Claesson 1987; Maas et al. 1992). The zircons analyzed here have Th/U > 0.23, mostly >0.40, and are clearly igneous in origin. The uniformity of zircon zoning textures, U–Th composition and spot ages from all analyzed grains are all features consistent with crystallization from magma (Table 2).

SHRIMP ages

SHRIMP ages for Kemashi hornblende diorite (tholeiitic)

Six grains were analyzed (Table 2; K-1.1 to K-6.1) and had U contents of 122–269 ppm and Th contents of 58–153 ppm. Th/U varied between 0.4 and 0.6. The weighted mean of all six 204Pb-corrected 206Pb/238U ages was 852.4 ± 11.2 Ma (95% confidence, MSWD = 0.92). The six analyses (common Pb uncorrected) define a SQUID-calculated concordia age of 856.3 ± 9.8 Ma (95% confidence, MSWD = 0.67, probability of concordance = 0.49) (Fig 4a).

SHRIMP ages for Bikilal-Ghimbi hornblende gabbro (tholeiitic)

Ten grains were analyzed (Table 2; BG-1.1 to 10.1). Analyses yielded U contents of 58–166 ppm and Th contents of 13–61 ppm; Th/U varied between 0.2 and 0.6. Excluding one disconcordant analysis (4.1), the weighted mean of nine 204Pb-corrected 206Pb/238U ages was 844.0 ± 12.1 Ma (95% confidence, MSWD = 0.40). Excluding analysis 4.1, the remaining nine analyses (common Pb uncorrected) define a SQUID-calculated concordia age of 846.0 ± 7.6 Ma (95% confidence, MSWD = 1.0, probability of concordance = 0.82) (Fig 4b).

SHRIMP ages for Gebeya Kemisa pyroxene diorite (calc-alkaline)

Six grains were measured (Table 2; GK-1.1 to 6.1). Analyses indicated U contents of 201–563 ppm, Th contents of 151–371 ppm, and Th/U between 0.5 and 0.9. Excluding one discordant analysis 5.1, the weighted mean of all five 204Pb-corrected 206Pb/238U ages was 790.6 ± 10.5 Ma (95% confidence, MSWD = 1.15). The same five analyses defined (common Pb uncorrected) a SQUID-calculated concordia age of 794.3 ± 9.4 Ma (95% confidence, MSWD = 1.0, probability of concordance = 0.98) (Fig 4c).

SHRIMP ages for Senbet Dura hornblende diorite (calc-alkaline)

Eight analyses were made (Table 2; SD-2.1 to 9.1), giving U contents of 109–418 ppm, Th contents of 62–165 ppm, and Th/U between 0.4 and 0.7. Excluding one slightly younger, concordant analysis 8.1 and reverse-discordant analysis 5.1, the weighted mean of 6 204Pb-corrected 206Pb/238U ages was 776.7 ± 10.5 Ma (95% confidence, MSWD = 1.28). The same six analyses (common Pb uncorrected) defined a SQUID-calculated concordia age of 787.7 ± 8.8 Ma (95% confidence, MSWD = 1.2, probability of concordance = 0.08) (Fig 4d).

SHRIMP ages for Wayu Meni olivine gabbro (tholeiitic)

Eight grains were analyzed (Table 2; WM-1.1 to 8.1). The analyses had 187–621 ppm U, 113–613 ppm Th, and Th/U between 0.5 and 1.0. All analyses were concordant within 95% confidence. The weighted mean of all eight 204Pb-corrected 206Pb/238U ages was 779.5 ± 11.1 Ma (95% confidence, MSWD = 0.40). All eight analyses (common Pb uncorrected) defined a SQUID-calculated concordia age of 778.1 ± 6.3 Ma (95% confidence, MSWD = 1.1, probability of concordance = 0.33) (Fig 4e).

Sr–Nd Isotope systematics

The Nd and Sr isotope compositions of both tholeiitic and calc-alkaline intrusions were determined. The initial Nd and Sr isotope ratios and epsilon values were calculated using the SHRIMP U–Pb ages, and are summarized in Table 3 and illustrated in Fig. 5. The initial 143Nd/144Nd ranged from 0.51178 ± 10 to 0.51191 ± 10 and εNd ranged from +4.2 to +7.0 (Table 3, Fig. 5). Initial 87Sr/86Sr for the samples ranged from 0.70277 ± 10 to 0.70319 ± 10 and εSr −5.5 to −10.4 (Table 3). The calc-alkaline samples showed systematically slightly higher εSr (−5.5 to −8.1) than the tholeiitic samples. The εNd–εSr plot shows that all gabbros fall in the fields of ANS plutonic rocks (Fig. 5), indicating derivation from depleted mantle with no identifiable contribution from pre-Neoproterozoic crust or ancient lithospheric mantle.
Table 3

Sr–Nd isotopic composition of the intrusions of Western Ethiopia

 

Sample

87Sr/86Sr

143Nd/144Nd

Rb (ppm)

Sr (ppm)

Sm (ppm)

Nd (ppm)

87Rb/86Sr

Sr (i)

147Sm/144Nd

Nd (i)

t (Ma)

εSr

εNd

TDM (Ga)

BG OL GB

BG12

0.702811

0.51306

0.33

312

1.1

3.1

0.003

0.70277

0.21538

0.511871

840

−10.4

6.2

 

BG OL GB

BG28

0.702912

0.51288

0.36

515

0.7

2.2

0.002

0.70289

0.20080

0.511778

840

−8.8

4.4

 

BG HB GB

BG57

0.702913

0.51296

1.26

413

2.8

9.0

0.009

0.70281

0.18979

0.511911

840

−10.0

7.0

 

WM OL GB

NK37B

0.703307

0.51268

5.00

365

5.6

21.6

0.040

0.70287

0.15601

0.511888

774

−10.2

4.9

0.97

K HB DT

NK24

0.705363

0.51252

23.7

327

3.9

18.0

0.210

0.70282

0.12889

0.511802

847

−9.6

5.0

0.95

GK PYX DT

NK02

0.704057

0.51248

31.4

993

3.6

17.7

0.092

0.70303

0.12243

0.511843

788

−7.7

4.3

0.95

SD DT

NK08

0.703721

0.51245

20.1

948

3.8

19.5

0.061

0.70304

0.11669

0.511850

779

−7.7

4.2

0.94

SD DT

NK13

0.703900

0.51244

25.3

1,152

4.3

22.4

0.064

0.70319

0.11524

0.511855

779

−5.5

4.3

0.94

SD DT

NK17

0.703196

0.51272

4.10

718

1.0

3.5

0.017

0.70301

0.16932

0.511861

779

−8.1

4.4

 

Ol Olivine, Hb Hornblende, Pyx Pyroxene, Gb Gabbro, DT Diorite, BG Bikilal-Ghimbi, WM Wayu Meni, K Kemashi, GK Gebeya Kemisa, SD Senbet Dura

TDM model ages are after Stern (2002) using the depleted mantle evolution model of Nelson and DePaolo (1984) and are calculated for samples with 147Sm/144Nd < 0.165

https://static-content.springer.com/image/art%3A10.1007%2Fs00531-009-0481-x/MediaObjects/531_2009_481_Fig5_HTML.gif
Fig. 5

Data points are shown as εNd versus εSr diagram for the Bikilal-Ghimbi gabbro. The field of the Arabian Nubian shield (ANS) rocks is after Stern and Kröner (1993); all epsilon values are adopted from GEOROC database metafile (GEOROC 2005)

Nd model ages (crust formation ages) were calculated assuming derivation of juvenile crust from depleted, asthenospheric mantle following Stern (2002), using the DePaolo depleted mantle model (DePaolo 1981; Nelson and DePaolo 1984). Samples with 147Sm/144Nd > 0.165 were excluded as they do not give reliable results (Stern 2002). Five of the samples had ratios <0.165 and yielded Nd model ages ranging from 0.93 to 0.97 Ga (Table 3; Fig. 6; Mean TDM = 0.95 Ga). The model ages were about 100 Ma older than the crystallization ages reported, but were younger than the Mesoproterzoic and Archean ages. This is consistent with the derivation of this part of the crust from a depleted mantle source, perhaps at 0.93–0.97 Ga, confirming the juvenile nature (mantle-derived) of ANS magmatism in the WES during the Neoproterozoic (Stern 2002).
https://static-content.springer.com/image/art%3A10.1007%2Fs00531-009-0481-x/MediaObjects/531_2009_481_Fig6_HTML.gif
Fig. 6

Nd isotopic evolution of the Ghimbi-Nedjo mafic to intermediate intrusions of Western Ethiopian Shield. Model for the evolution of the depleted mantle is from DePaolo (1981) and Nelson and DePaolo (1984); data calculated after Stern (2002) with 147Sm/144Nd < 0.165. Emplacement ages and εNd (t) of the intrusions is denoted by the symbols. TH tholeiitic, CA calc-alkaline, IA island arc, OIT oceanic island tholeiite, BAB back-arc basin, HB hornblende, OL olivine, PYX pyroxene, GB gabbro, DT diorite

Discussion

Early tectonic ages (~870–830 Ma ages)

The oldest dated intrusions are the Kemashi hornblende diorite (856.3 ± 9.8 Ma) and the Bikilal-Ghimbi hornblende gabbro with age of 846.0 ± 7.6 Ma that are interpreted as intra-plate tholeiitic intrusions with plume-type oceanic island tholeiite (OIT) geochemical signatures (Fig. 7, by Woldemichael and Kimura 2008a, b). These ages are older than the previously dated intrusions in WES (Table 4; Figs. 8, 9) and include the U–Pb zircon ages of Goma granodiorite (814 ± 2 Ma) and the Birbir quartz diorite 828 +9/−2 Ma from Gore-Gambella area (Ayalew et al. 1990), along with the Pb/Pb zircon evaporation ages of Ujjukka granite and granodiorite (815 ± 5 Ma) from the Ghimbi area (Kebede et al. 2001a). In contrast to the OIT Kemashi hornblende diorite and Bikilal-Ghimbi hornblende gabbro, the other plutons (Ayalew et al. 1990 and Kebede et al. 2001a) are subduction related, emplaced in intra-oceanic island-arc environments.
https://static-content.springer.com/image/art%3A10.1007%2Fs00531-009-0481-x/MediaObjects/531_2009_481_Fig7_HTML.gif
Fig. 7

Summarized primitive mantle (Sun and McDonough 1989) normalized plot of the dated intrusions of the Western Ethiopian Shield, modified after Woldemichael and Kimura (2008a, b). Hawaiian (Kilauea magma–Hilina lava) primitive tholeiite (Kimura et al. 2006), Northeast Japan island arc (Kimura and Yoshida 2006) and Mariana Back-arc basin (Pearce et al. 2005) data are used for comparison. GB gabbro, DT diorite, TH tholeiitic, CA calc-alkaline, OIT oceanic island tholeiite, IA island arc, BAB back-arc basin

Table 4

Isotopic age and tectonic setting summary of Neoproterozoic rocks in Western Ethiopia Shield (WES)

Subdivision

Locality

Rock type

Method

Age (Ma)

Tectonic setting

WMGT

Barol (S or C)

Orthogneiss

U–Pba

798 ± 12

CVST/WMGT

Baro (S)

Granite

U–Pbb

783 ± 15

Syn-CA-IAf

Senbet Dura (C)

Hornblende diorite

SHRIMP (this study)

788 ± 9

Syn-CA-IAg

Gebeya Kemisa (C)

Pyroxene diorite

SHRIMP (this study)

794 ± 9

Syn-CA-IAg

CVST

Kemashi (C)

Hornblende diorite

SHRIMP (this study)

856 ± 10

Early-TH/OITg

Goma/Birbir (S)

Granodiorite/quartz diorite

U–Pbb

814 ± 2/828 ± 5

Early-CA-IAf

Kilaj (N)

Quartz diorite

U–Pb(m,c)

866 ± 20

Early-CA- BAB/IAc

Wayu Meni (C)

Olivine gabbro

SHRIMP (this study)

778 ± 6

Syn-TH-OIT/BABg

Dogi/Duksi (N)

Granodiorite

U–Pb(n,c)/Pb–Pbc

651 ± 5/699 ± 2

Post-CA/BAB-IAc

Ganjii (C)

Monzogranite

Pb–Pbd

622 ± 7

Post-CA/WPGh

Mao/Bonga (S)

Granite/monzonite

U–Pbb

541 ± 15/571 ± 10

Post-CA-IAf

EMGT

Suqii-Wagga/Guttin (C)

Granite

Pb–Pbd/U–Pbd

698 ± 27/730 ± 2

Post-CA-WPGi

Bikilal-Ghimbi (C)

Hornblende gabbro

SHRIMP (this study)

846 ± 8

Early-TH/OIT(g,j)

Ujjukka (C)

Granite/granodiorite

Pb–Pbd

815 ± 5

Early-CA-VAG(d,k)

Gebal (S or C)

Orthogneiss

U–Pba

811 ± 25

Homa (C)

Granitoid orthogneiss

SHRIMPe

776 ± 12

S southern-WES, N northern-WES, C central-WES, WMGT western migmatitic gnessic terrane, EMGT eastern migmatitic gnessic terrane, CVST central metavolcano-sedimentary terrane, CA Calc-alkaline, TH tholeiitic, IA island arc, OIT oceanic island tholeiite, BAB back-arc basin, WPG within in plate granite, VAG volcanic arc granite

aJohnson et al. (2004); b Ayalew et al. (1990); c Grenne et al. (2003); d Kebede et.al. (2001a); e Kebede et al. (2007); f Ayalew and Peccerillo (1998); g Woldemichael and Kimura (2008b); h Kebede and Koeberl (2003); i Kebede et al. (2001b); j Woldemichael and Kimura (2008b); k Kebede et al. (1999)

lDomain-name; m Imprecise age; n Titanite

https://static-content.springer.com/image/art%3A10.1007%2Fs00531-009-0481-x/MediaObjects/531_2009_481_Fig8_HTML.gif
Fig. 8

Simplified geological map with dated rocks of the Western Ethiopian Shield. The central part is represented by: a the Ghimbi-Nedjo area modified after Alemu and Abebe (2000), Allen and Tadesse (2003), and Woldemichael and Kimura (2008b). The northern part of WES is represented by: b the Kilaj–Wombera area, modified after Grenne et al. (2003). The southern is represented by: c the Gore-Gambella area, modified after Ayalew et al. (1990). Data summarized in Table 4. Inset shows the WES and the relative positions of the three areas as indicated by the boxes. SW Suqii-Wagga granite, UJ Ujjukka granite, GJ Ganjii granite

https://static-content.springer.com/image/art%3A10.1007%2Fs00531-009-0481-x/MediaObjects/531_2009_481_Fig9_HTML.gif
Fig. 9

a Histogram of robust U–Pb, Pb–Pb and Rb–Sr ages for the rocks of the Arabian–Nubian Shield showing the entire age range, including Archean and inherited materials (previous studies; modified after Johnson and Woldehaimanot 2003). b Neoproterozoic subset including ages from the WES (data from Table 4) and the Ghimbi-Nedjo area (SHRIMP ages, this study). Apart from this study, only one other SHRIMP age is available from the WES. TH tholeiitic, CA calc-alkaline, OIT oceanic island tholeiite, IA island arc, BAB back-arc basin

The oldest volcanic rocks in the ANS (Fig. 9) are oceanic tholeiitic basalts, erupted from about 900–870 Ma, interpreted to have formed at volcanic passive margins (Stern 1994) or associated with a mantle plume, perhaps in an oceanic plateau (Stein and Goldstein 1996; and references therein). The controversial issue of the mantle reservoir(s) that feed large igneous provinces is the subject of many papers (e.g., Anderson 1994; Courtillot et al. 2003; Coffin and Eldholm 1994). However, it has been suggested that in addition to plate tectonic forces, sub-vertical flows of hot and buoyant mantle are associated with plate breakup and the formation of volcanic passive margins (e.g., Geoffroy 2005; and references therein). Volcanic passive margins form when rifting is accompanied by significant mantle melting, with volcanism occurring after and/or during continental breakup, magmas are either extruded as basalt flows or intruded into or beneath the crust. Transitional crust of volcanic passive margins is composed of basaltic igneous rocks, including lava flows, sills, dykes and gabbro (e.g., White and McKenzie 1989; Coffin and Eldholm 1994; Courtillot et al. 1999; Boillot and Froitzheim 2001; Geoffroy 2005). Gabbros associated with rifted continental margins are supposed to represent either magmatic underplating of continental crust, passively exhumed during much later rifting or, alternatively, initiation of partial melting during late rifting and incipient sea floor spreading. Both types of occurrence have been described from present-day ocean–continent transitions (e.g., Schärer et al. 1995, 2000; Cornen et al. 1999; Boillot and Froitzheim 2001).

Direct evidence of early rifting and passive margin development in the EAAO may be preserved in sedimentary successions in Kenya and Sudan (e.g., Key et al. 1989; Kröner et al. 1987; Stern 1994). A westward-facing (present geometry) passive margin with respect to the Mozambique Ocean may have developed in Western Ethiopia during early Neoproterozoic time, upon which carbonate and clastic sediments were deposited (Braathen et al. 2001; Tadesse and Allen 2005). The Kemashi hornblende diorite (856.3 ± 9.8 Ma) and the Bikilal-Ghimbi hornblende gabbro (846.0 ± 7.6 Ma) may have intruded into a passive margin setting implying that the early oceanic magmatism in WES of the EAAO is related to the early stage of seafloor spreading that later led to the full opening of the Mozambique Ocean.

Even though no dating has been done, Tadesse and Allen (2005) interpreted the dismembered Tuludimtu ophiolite as representing oceanic crust, with no evidence for pre-Neoproterozoic continental crust. The associated metasediments are a mixture of deep marine quartzites and pelagic muds, tectonically interleaved with a turbiditic sequence containing lenses of carbonate. The presence of carbonate lenses suggests a shallow marine depositional environment, perhaps related to a continental margin setting (Tadesse and Allen 2005). The Kemashi hornblende diorite is part of the assemblage interpreted as dismembered Tuludimtu ophiolite. We hence consider these intrusions and associated magmatism as the first clear evidence of magmatism formed at the passive margin in the EAAO. The origin of the large Bikilal-Ghimbi gabbro body and the Kemashi diorite, which have a geochemical signature of intra-plate plume-related tholeiitic parental magma magmatism (Fig. 7; Woldemichael and Kimura, 2008a, b), is consistent with rifting to form a passive margin, early in the evolution of the EAAO.

Syn-tectonic (830–700 Ma ages)

The Wayu Meni olivine gabbro (778.1 ± 6.3 Ma) followed by Senbet Dura hornblende diorite (787.7 ± 8.8 Ma) and Gebeya Kemisa pyroxene diorite (794.3 ± 9.4 Ma) are almost contemporaneous. These ages are similar to “syn-tectonic” granitic plutons of Baro domain of Gore-Gambella area with U/Pb emplacement age of 783 +19/−14 Ma (Ayalew et al. 1990) and “syn-tectonic” granites from the Ghimbi area, which have ages of 698 and 730 Ma (Kebede et al. 2001b; Table 4; Figs. 8, 9). Woldemichael and Kimura (Fig. 7; 2008b) concluded that the intrusions at the boundary between the western CVST metavolcanics and WMGT are calc-alkaline arc-type (the Gebeya Kemisa and Senbet Dura diorites) and those within the CVST metavolcanics (Wayu Meni gabbro) are tholeiitic back-arc basin (BAB) in origin. Collectively, these plutons have the trace element signature of an arc-BAB system. This is consistent with previous interpretations that the “syn-kinematic” plutons (~780–700 Ma) were subduction-related and were emplaced in intra-oceanic island-arc environments (Ayalew and Peccerillo 1998; Kebede et al. 1999, 2001a, b; Grenne et al. 2003).

The distribution of the dated intrusions from the differing magma suites in the Ghimbi-Nedjo area shows spatial and temporal variation. Calc-alkaline arc-type and tholeiitic BAB-plutons occur in the western terrane (i.e., WMGT and western CVST), whereas the eastern terrane (EMGT and eastern CVST) is dominated by tholeiitic OIT-type magmas (Fig. 7; Woldemichael and Kimura, 2008a, b). This is supported by the age relationship that tholeiitic OIT-type magmas are early-tectonic (or syn-rifting; 846–856 Ma), whereas the calc-alkaline arc-type and tholeiitic BAB-type plutons are syn-tectonic (post-rifting and syn-sea floor spreading; subduction related; 778–794 Ma). All the other dated intrusions from previous studies of supposedly “pre-tectonic” and “syn-tectonic” are hence “early to syn-tectonic”.

Tadesse and Allen (2004) studied the geochemistry of Kemashi domain metavolcanics (eastern CVST of this study) and suggested that they represent oceanic crust that formed at a spreading ridge, whereas metavolcanics from the Dengi and Sirkole domains (western CVST of this study) represented an island arc. If the “imprecise age” of 866 ± 20 Ma (U–Pb zircon) from the quartz diorite of the Kilaj intrusive complex and the Rb–Sr whole-rock errorchron of 873 ± 82 Ma from the Debesa metavolcanics (Table 4; Grenne et al. 2003) north of this study area are accurate, these are indicative of the coeval nature of the intrusions and the metavolcanics. These metavolcanics and intrusions show a transition between typical calc-alkaline and E-MORB suites, similar to continental margin magmatism (Grenne et al. 2003). However, re-evaluating the other metavolcanics and intrusions with SHRIMP ages is required. This is beyond the scope of our present study.

Isotopic constraints

The εNd–εSr plot shows that all gabbros fall in the fields of ANS plutonic rocks (Fig. 5). The result suggests depleted sources with no discernable involvement of older continental crust. If significant older crust was involved, the isotopic compositions of the magmas would have extended to more negative epsilon Nd and positive epsilon Sr values. Therefore, there is no significant involvement of pre-Neoproterozoic crust. The calc-alkaline island arc-type plutons show slightly higher εSr ranging from −5.5 to −8.1, suggesting involvement of trace amounts of sediments in generating these magmas, or assimilation of minor radiogenic ANS crust. The difference is subtle, but is consistent with the inferred origin of the magma types.

The Nd model ages ranges from 0.93 to 0.97 (TDM mean = 0.95 Ga; Fig. 6) and show variation between the calc-alkaline arc-back arc-type and tholeiitic OIT-type magmas. The mean depleted mantle model age for the entire WES is 0.92 Ga (Fig. 10). This was also calculated for other parts of WES (Grenne et al. 2003; Kebede and Koeberl 2003; Johnson et al. 2004) following Stern (2002), excluding samples with 147Sm/144Nd > 0.165 and using the DePaolo depleted mantle model (DePaolo 1981; Nelson and DePaolo 1984). The mean value from the WES correspond well with those calculated by Stern (2002) based on existing Nd isotopic data from northern Ethiopia and Eritrea (mean value 0.87 Ga) and southern Ethiopia (1.13 Ga), respectively. The mean Nd model age for Western Ethiopia (0.92 Ga) perhaps indicates that the transition between the northern Ethiopia and southern Ethiopia lies in Western Ethiopia, reflecting the overall big transition of the northern ANS and southern MB of the EAAO. Stern (2002) suggested that the enigmatic nature of the crust in the transition between the juvenile crust in the northern EAO and the Archean–Paleoproterozoic crust (MB) in the southern EAO lies in southern Ethiopia and northern Kenya.
https://static-content.springer.com/image/art%3A10.1007%2Fs00531-009-0481-x/MediaObjects/531_2009_481_Fig10_HTML.gif
Fig. 10

Histogram of Nd model ages of the Western Ethiopian Shield, showing data of this study and data recalculated from previous works (Grenne et al. 2003; Kebede and Koeberl 2003; Johnson et al. 2004). Model for the evolution of the depleted mantle is from DePaolo (1981) and Nelson and DePaolo (1984); data calculated after Stern (2002) with 147Sm/144Nd < 0.165

Implications for the geodynamic evolution of the Western Ethiopia Shield

The WES has strong parallels to modern orogenic belts that evolved by the classic Wilson-style plate tectonic cycle (e.g. Tadesse and Allen 2005). The Ghimbi-Nedjo region yields important insights into the tectonic development of the WES in Neoproterozoic time. A geodynamic evolutionary model that emphasizes the evolution of a passive margin formation in that Ghimbi-Nedjo region is outlined in Fig. 11.
https://static-content.springer.com/image/art%3A10.1007%2Fs00531-009-0481-x/MediaObjects/531_2009_481_Fig11_HTML.gif
Fig. 11

Proposed geodynamic model of the Ghimbi-Nedjo area of the Western Ethiopian Shield. TH tholeiitic, CA calc-alkaline, OIT oceanic island tholeiite, IA island arc, BAB back-arc basin, WMGT western migmatitic gneissic terrane, EMGT eastern migmatitic gneissic terrane, CVST central volcano sedimentary terrane

The pre-rifting (>~900 Ma) stage is defined by a complex structural history and a metamorphic cycle predating Pan-African orogenesis, but evidence of these older rocks has proven elusive. The igneous protoliths of the gneissic terrane have ages of ~798 Ma (U–Pb; Johnson et al. 2004) and ~776 Ma (zircon SHRIMP age; Kebede et al. 2007) which is contradictory to the assumption that the gneissic terrane are older. The mean TDM Nd model age of 0.92 Ga is not conclusive enough (Fig. 10). However, older Nd model ages (>1.00 Ga) may suggest a contribution from reworked ancient crust or sediments in the protolith of Neoproterozoic rocks from Western Ethiopia. (Fig. 10; Kebede and Koeberl 2003; Grenne et al. 2003; Johnson et al. 2004).

Early rifting (~900–860 Ma, Fig. 11a) might have been initiated during the early Neoproterozoic break-up of the Rodinia supercontinent. Stern (1994) argued that rifting of the continent started at ~870 Ma, followed by passive margin formation (860–830 Ma, Fig. 11b), but offered scant evidence for this interpretation. The oceanic island tholeiitic signature together with U–Pb SHRIMP magmatic ages from the Kemashi hornblende diorite (856.3 ± 9.8 Ma) and the Bikilal-Ghimbi hornblende gabbro (846.0 ± 7.6 Ma) are clear indications of early oceanic magmatism that could have been associated with early Neoproterozoic rifting as part of EAAO development and which were emplaced in a passive margin environment. The reliable ages reported here, supported by chemical compositions of the plutons (Fig. 7; Woldemichael and Kimura 2008a, b) are consistent with magmatism associated with passive margin formation. The opening of the Mozambique Ocean eventually resulted in subduction, forming arcs and back-arc basins (830–750 Ma, Fig. 11c). The calc-alkaline arc-type Gebeya Kemisa and Senbet Dura diorites and the BAB-like tholeiitic Wayu Meni gabbro (ages of 794.3 ± 9.4, 787.7 ± 8.8, and 778.1 ± 6.3 Ma, respectively) were emplaced at this stage. Other subduction-related calc-alkaline granitoids were also emplaced in intra-oceanic arc environments (Ayalew et al. 1990; Ayalew and Peccerillo 1998; Kebede et al. 1999, 2001a, b; Grenne et al. 2003). Orogeny (750–650 Ma, Fig. 11d) is characterized by basin closure, arc assembly, and terrane accretion associated with the generation of calc-alkaline arc intrusions (Ayalew et al. 1990; Ayalew and Peccerillo 1998; Braathen et al. 2001). This collision orogeny continued until ~650 Ma. The postorogenic period (~650–550 Ma, Fig. 11e) terminated at around 550 Ma with emplacement of post-collisional plutons (Ayalew et al. 1990; Ayalew and Peccerillo 1998; Kebede et al. 1999, 2001a, b; Grenne et al. 2003).

Conclusions

The WES consists of metavolcanic, metasedimentary and ultramafic complexes, bounded by gneissic terrane to the east and west. The WES is associated with gabbroic to granitic intrusive bodies. WES dioritic to gabbroic intrusions are (1) early-tectonic syn-rift tholeiitic Kemashi diorite and Bikilal-Ghimbi gabbros, with ages of 856 and 846 Ma, respectively; and (2) syn-tectonic (post-rift and syn-sea floor spreading) BAB-type tholeiitic Wayu Meni gabbro (778 Ma) and calc-alkaline arc-type Senbet Dura hornblende diorite (787 Ma) and Gebeya Kemisa pyroxene diorite (794 Ma). Their positive εNd (4.5–7.0) and the negative εSr (−5.5 to −10.5) values lie in the middle of the fields of ANS plutonic rocks, suggesting depleted mantle sources with no significant involvement of a pre-Pan-African continental crust. Calculated mean Nd depleted mantle model ages of 0.95 Ga for the Ghimbi-Nedjo region and 0.92 Ga for the entire WES also suggest a depleted juvenile Neoproterozoic mantle source. The results are consistent with the ANS development model whereby the earliest stage of EAAO development began with rifting associated with the breakup of Rodinia, leading to passive margin formation at about ~860–830 Ma, associated with opening of the Mozambique Ocean. This was followed by seafloor spreading, arc and back-arc basin formation, between 830 and 690 Ma. Calc-alkaline WES plutons were emplaced at 790–770 Ma suggesting an oceanic arc existed in the WES.

Acknowledgments

This work was carried out at the Shimane University with the highly appreciated scholarship support from Monbukagakusho (the Japanese Ministry of Education, Culture, Sport, Science, and Technology) to B.W.W. We greatly appreciate the support of the Geological Survey of Ethiopia for logistical assistance in the field, and for granting extended leave to B.W.W. for this study. We also thank the National Institute of Polar Research for access to SHRIMP analytical facilities. Our thanks to Dr. R. J. Stern of the University of Texas at Dallas, USA, for thoughtful comments on an earlier version of the manuscript and to Dr. B. P. Roser of Shimane University, Japan, for valuable comments on the draft manuscript. Comments from Dr. M. G. Abdelsalam and an anonymous reviewer were helpful in improving the quality of the manuscript.

Copyright information

© Springer-Verlag 2009