Scholarly article on topic 'Petrogenesis and origin of modern Ethiopian rift basalts: Constraints from isotope and trace element geochemistry'

Petrogenesis and origin of modern Ethiopian rift basalts: Constraints from isotope and trace element geochemistry Academic research paper on "Earth and related environmental sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Lithos
Keywords
{"Ethiopian rift basalts" / "Mixed source" / Petrogenesis / "Veined lithosphere"}

Abstract of research paper on Earth and related environmental sciences, author of scientific article — D. Ayalew, S. Jung, R.L. Romer, F. Kersten, J.A. Pfänder, et al.

Abstract The source of continental rift-related basalts and their relation to rifting processes is a continuous matter of debate. We present major and trace element and Sr, Nd, Hf and Pb isotope data for axial rift basalts from eight volcanic centres (Ayelu, Hertali, Dofan, Fantale, Kone, Bosetti and Gedemsa, from NE to SW) in Afar and Main Ethiopian Rift (MER) to assess their source regions and their genetic relationships. These lavas have geochemical characteristics, i.e., a peak at Ba, Nb and troughs at K and Rb in primitive mantle-normalised multielement diagrams, which are consistent with predominant melting of an amphibole-bearing lithospheric mantle. However, the isotopic compositions for these lavas are heterogeneous (87Sr/86Sr=0.70354–0.70431, 143Nd/144Nd=0.51280–0.51294, 176Hf/177Hf=0.28301–0.28315, 206Pb/204Pb=18.48–19.31, 207Pb/204Pb=15.53–15.62, 208Pb/204Pb=38.61–39.06) and require various mantle reservoirs with distinctive isotopic signatures. The range of isotopic compositions requires the involvement of three distinct source components from the asthenospheric and veined lithospheric mantle. Progressive rifting leads to lithosperic thinning and upwelling of hot asthenospheric mantle, which induces melting of the veined lithospheric mantle. The trace element characteristics of the lavas are dominated by the vein material, which has a higher trace element content than the surrounding mantle. The isotopic composition of the vein material, however, is not very different from the ambient mantle, giving rise of apparent uncoupling of trace element and isotope constraints for the melt source. The uprising basaltic liquids in part inherit a lithospheric trace element signature, while their isotopic compositions are mostly unaffected due to short residence times within the lithosphere in context with progressive rifting and lithospheric thinning. Thus, the geochemical and isotope data are consistent with a multi-component source prevailing beneath the Afar and MER areas in which the basalts are generated during progressive rifting and, thus, passive upwelling of a mantle source.

Academic research paper on topic "Petrogenesis and origin of modern Ethiopian rift basalts: Constraints from isotope and trace element geochemistry"

Accepted Manuscript

Petrogenesis and origin of modern Ethiopian rift basalts: Constraints from isotope and trace element geochemistry

D. Ayalew, S. Jung, R.L. Romer, F. Kersten, J.A. Pfänder, D. Garbe-Schonberg

PII: S0024-4937(16)30028-7

DOI: doi: 10.1016/j.lithos.2016.04.001

Reference: LITHOS 3887

To appear in:

LITHOS

Received date: Accepted date:

23 September 2015 5 April 2016

Please cite this article as: Ayalew, D., Jung, S., Romer, R.L., Kersten, F., Pfander, J.A., Garbe-Schonberg, D., Petrogenesis and origin of modern Ethiopian rift basalts: Constraints from isotope and trace element geochemistry, LITHOS (2016), doi: 10.1016/j.lithos.2016.04.001

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Petrogenesis and origin of modern Ethiopian rift basalts: constraints from isotope and trace element geochemistry

Ayalew, D.1, Jung, S.2, Romer, R.L.3, Kersten, F.4, Pfänder, J.A.4, Garbe-Schönberg, D.5

1School of Earth Sciences, Addis Ababa University, P.O. Box 1176, Addis Ababa, Ethiopia

Mineralogisch-Petrographisches Institut, Universität Hamburg, Grindelallee 48, 20146 Hamburg, Germany

3Helmholtz-Zentrum Potsdam, Deutsches GeoForschungsZentrum GFZ, Telegrafenberg, 14473 Potsdam, Germany

4Geologisches Institut, Technische Universität-Bergakademie Freiberg, Bernhard-von-Cotta Str.2, 09599 Freiberg, Germany

5Institut für Geowissenschaften, Abteilung Geologie, Universität Kiel, Ludewig-Meyn-Strasse 10, 24118 Kiel, Germany

Abstract

The source of continental rift-related basalts and their relation to rifting processes is a continuous matter of debate. We present major and trace element and Sr, Nd, Hf and Pb isotope data for axial rift basalts from eight volcanic centres (Ayelu, Hertali, Dofan, Fantale, Kone, Bosetti and Gedemsa, from NE to SW) in Afar and Main Ethiopian Rift (MER) to assess their source regions and their genetic relationships. These lavas have geochemical

characteristics, i.e., a peak at Nb and troughs at K and Rb in primitive mantle-normalized multielement diagrams, that are consistent with predominant melting of an amphibole-bearing lithospheric mantle. However, the isotopic compositions for these lavas are heterogeneous (87Sr/86Sr = 0.70354-0.70431, 143Nd/144Nd = 0.51280-0.51294, 176Hf/177Hf = 0.28301-0.28315, 206Pb/204Pb = 18.48-19.31, 207Pb/204Pb = 15.53-15.62, 208Pb/204Pb = 38.61- 39.06) and require various mantle reservoirs with distinctive isotopic signatures. The range of isotopic compositions requires the involvement of three distinct source components from the asthenospheric and veined lithospheric mantle. Progressive rifting leads to lithosperic thinning and upwelling of hot asthenospheric mantle, which induces melting of the veined lithospheric mantle. The trace element characteristics of the lavas are dominated by the vein material, which has a higher trace element content than the surrounding mantle. The isotopic composition of the vein material, however, is not very different from the ambient mantle, giving rise of apparent uncoupling of trace element and isotope constraints for the melt source. The uprising basaltic liquids in part inherit a lithospheric trace element signature, while their isotopic compositions are mostly unaffected due to short residence times within the lithosphere in context with progressive rifting and lithospheric thinning. Thus, the geochemical and isotope data are consistent with a multi-component source prevailing beneath the Afar and MER areas in which the basalts are generated during progressive rifting and, thus, passive upwelling of a mantle source.

Keywords/phrases: Ethiopian rift basalts, mixed source, petrogenesis, veined lithosphere

1. Introduction

The Main Ethiopian Rift (MER) constitutes a part of the east African Rift System (EARS) and merges with the oceanic Red Sea and Gulf of Aden rifts in a triple junction located in the Afar depression (Wolfenden et al., 2004). Rifting and volcanism in the MER has been considered to be related to mantle plume activity (e.g., Marty et al., 1996; Pik et al., 1999; Furman, 2007). The impinging Afar plume is assumed to have triggered flood basalt volcanism in the homonymous region ~ 30 Ma ago (Hofmann et al., 1997). There is no consensus on the number(s) of mantle plumes; some authors assume that recent eruptions in the Erta'Ale range in Afar are triggered by the Afar plume (Furman et al., 2006), while others argue that the magmatism in southern Ethiopia is related to a Kenya mantle plume (Rogers et al., 2000) in agreement with plate motion reconstructions from O'Connor et al. (1999) suggesting that the Afar plume is also presently impinging SW of its former location.

At present the MER is in an evolved stage of continental rifting, progressing towards continental break-up (e.g., Wolfenden et al., 2004; Kurz et al., 2007). Recent bimodal volcanic activity takes place exclusively in the axial portions of the rift and is, in most cases, confined to magmatic segments which are aligned en-echelon along the rift axis (Keranen et al., 2004; Kurz et al., 2007). The source of recent basaltic rocks in the MER and their possible interaction with the lithosphere remain controversial. Several authors suggested that at least two isotopically distinct mantle reservoirs contributed to the genesis of Quaternary basaltic lavas in the rift (e.g., Barberi et al., 1980; Hart et al., 1989). Hart et al. (1989) found isotopic evidence for a depleted mid ocean ridge basalt (MORB)-like source and a depleted ocean island basalt (OIB)-like mantle reservoir. Furman et al. (2006) argue that recent Afar lavas originate from a depleted mantle source whereas within-rift basalts show an isotopic signature that resembles enriched mantle. Recently, Rooney et al. (2012) interpreted the isotopic variations observed on modern basalts erupted along the MER as the result of mixing from

three components; the Afar plume, the asthenospheric upper mantle and the continental lithosphere. The influence of the plume signature appears to decrease along the MER away from the Afar to southern Ethiopia (Rooney et al., 2012).

In order to constrain the sources of coeval MER and Afar basalts, we investigated mafic volcanic products from eight volcanic centres in southern Red Sea rift (Ayelu, Hertali and Dofan from NE to SW) and MER (Fantale, Kone, Boseti and Gedemsa from NE to SW; Fig. 1). Based on new major and trace element as well as Sr-Nd-Hf-Pb isotopic data, we discuss the origin of Quarternary rift-related basalts and place constraints on the mantle sources and the melting conditions.

2. Geological background

2.1. Volcanic history

Volcanic activity commenced around 45 Ma ago in southern Ethiopia (George et al., 1998), resulting in bimodal basaltic flows and associated rhyolites covering wide areas. The peak of magmatism occurred ~30 Ma ago, leading to flood basalt eruptions in Ethiopia and Yemen (Hofmann et al., 1997). At ~ 25 Ma, continental rifting commenced in the southern Red Sea (Hart et al., 1989). In southern Ethiopia, extension began ~18 Ma ago (Keranen et al., 2004) and was accompanied by basaltic magmatism, active for about seven to eight million years (George et al., 1998). The Southern Ethiopian Rift propagated northward, reaching the present central Main Ethiopian Rift (MER) ~14 Ma ago (Brotzu et al., 1981) and ultimately joining the southern Red Sea rift ~11 Ma ago (Chernet et al., 1998; Wolfenden et al., 2004). Contemporaneously to the connection between the MER and Red Sea rifts, a flood basalt event occurred in this area (Chernet et al., 1998). Beginning in late Miocene and continuing throughout the Pliocene, silicic volcanic centres emerged from the rift floor (Kurz

et al., 2007). According to Lahitte et al. (2003), the underlying magma chambers represent zones of lithospheric weakness that indicate the preferred direction of future rift propagation. Progressive weakening of the lithosphere in Afar, associated with heating and thermomechanical erosion of the lower crust (Rooney et al., 2007) resulted in the onset of oceanic rifting at 5.3 Ma (Hart et al., 1989). Oceanic rifting is still active in Afar, whereas it has not commenced in the MER yet (Wolfenden et al., 2004).

The end of Pliocene was marked by a change in the orientation of the stress field, giving rise to oblique rifting (Boccaletti et al., 1999). Since then, extension has been localized in narrow (50 km long, maximum 20 km wide) en-echelon arranged segments on the rift floor (Kurz et al., 2007), with a system of bounding faults that are referred to as the Wonji Fault Belt (e.g., Mohr, 1983; Boccaletti et al., 1999; Bonini et al., 2005). Moreover, these segments have been the locus of volcanic activity throughout the Quaternary and thus are referred to as magmatic segments by Kurz et al. (2007). Volcanic activity, associated with the magmatic segment, was initially characterized by large volumes of felsic lavas (Fig. 2). When these faults reached the upper mantle in recent times, basaltic volcanism commenced (Boccaletti et al., 1999). At present, volcanic activity within the magmatic segment is dominated by fissural basalt eruptions (Kurz et al., 2007).

2.2. Nature of the crust and mantle beneath Ethiopia

Based on joint inversion of receiver functions and Rayleigh wave group velocities (Stuart et al., 2006; Dugda et al., 2007; Hammond et al., 2013) and controlled source experiments (Maguire et al., 2006), the crust has a thickness of 35 to 44 km beneath the Ethiopian plateau, and < 18 to 25 km beneath the Afar depression, with thinnest values north of the Afar triple junction. Crust beneath the MER ranges from 25 km in the north to 35 km. Similarly, the lithospheric mantle beneath the Ethiopian plateau extends to a depth of ~70-80 km. Beneath

the MER and Afar the lithospheric mantle extends to a depth of ~50 km. Intensive dyking and numerous magma intrusions lead to a significantly modified lithosphere beneath the MER (Rooney et al., 2007). Zones of high P-wave velocity underneath the magmatic segment, detected by seismic tomography, were interpreted as solidified mafic intrusions (Keranen et al., 2004). Magnetotelluric data revealed melt intrusions in shallow and midcrustal regions (Whaler and Hautot, 2006). Placed beneath the Quaternary extrusions on the surface, this melt is generally associated with the magma chambers feeding the silicic volcanoes in the rift (e.g., Peccerillo et al., 2003; Rooney et al., 2007). Heat input from the anomalously hot asthenosphere (Kendall et al., 2005) further contributes to a weakened lithosphere that is prone to rift propagation (Lahitte et al., 2003; Keranen et al., 2004). Further geophysical measurements demonstrate that crustal thickness in the rift decreases from the south towards Afar, where rifting is most progressed. Moho depths of ~38 km in the southern MER, ~24 km beneath Fantale (Dugda et al., 2005) and ~16 km in Afar were inferred (Hayward and Ebinger, 1996).

3. Alaytical methods

Fresh samples were crushed to 3-5 mm small chips in a steel crusher at the Geologisches Institut/TU Freiberg. The chips were ground to a fine powder (grain size < 63 ^m) in an agate planetary mill. Each sample was divided into representative sub-samples, which were later used for geochemical analysis. Major elements and trace elements for some samples were determined on sample powders using a lithium metaborate/tetraborate fusion procedure by ICP-AES and ICP-MS methods, respectively at Activation Laboratories Ltd. in Ancaster, Canada. Trace elements for some samples (61110F, 61110A, 61110C, 61110D, 61119A, ETH-14, ETH-24, ETH-20, ETH-19, ETH-23, ETH-8, ETH-6, ETH-7 and GMB009) were analysed at the Institut für Geowissenschaften, Universität Kiel. For trace element analysis

250 mg of whole rock powders were disolved in HF-HNO3, diluted and measured on an AGILENT 7500cs ICP-MS instrument. For the procedural details see Garbe-Schönberg (1993).

The isotopic composition of Sr, Nd and Pb was measured at Universität Münster and Deutsches GeoForschungsZentrum (Potsdam) and the isotopic composition of Hf was analyzed at Universität Münster. The "ETH" sample series was prepared according to procedures described by Thirlwall et al. (1997) and were analysed at Royal Holloway University of London. Details of the analytical techniques are found in apeendix A.

4. Results

4.1. Petrography

The distribution of sampled volcanic centres in the MER and Afar is shown in Figure 1. Basic rocks occur as aa-lava flow with mostly unweathered surfaces. The basalts are typically slightly to highly vesicular (vesicles up to 5 cm), but massive basalts are also abundant. Vesicular basalts exhibit brown to reddish crusts when altered. The samples are variably porphyritic with phenocrysts of plagioclase, olivine and clinopyroxene (Ti-augite) with rare microphenocrysts of Fe-Ti oxides. The groundmass exhibits a grey microcrystalline matrix consisting of specular feldspar and in some samples magnetite and glass.

4.2. Major elements

Major and trace element data are reported in appendix B. On the basis of the total alkalis vs. SiO2 classification diagram (TAS, Le Bas et al., 1986), all samples are basalts, with a few samples straddled along the tholeiitic/alkaline boundary (Fig. 2). The MgO contents of the

samples are variable ranging from 4.5 to 9.2 wt.%. The highest MgO contents are found from the Kone and Bosetti volcanic fields within the MER. Variations of major elements against MgO are illustrated in Figure 3. Na2O, TiO2, Fe2O3 and P2O5 exhibit well-defined negative trends with decreasing MgO concentration. Oxides of CaO and Al2O3 do not exhibit the same trends. In general, CaO contents decrease with decreasing MgO, whereas Al2O3 contents slightly increase with decreasing MgO down to ~6 wt%, and then either remain almost constant (Afar) or decrease (MER) with decreasing MgO. K2O contents do not show a systematic variation with MgO, but generally are higher for MER basalts than for Afar basalts. The investigated lavas from Afar systematically show more coherent trends than those of MER.

4.3. Trace elements

Trace element variations as a function of MgO from the northern MER and Afar rifts are shown in Figure 4. Compatible elements, such as Ni and Sc, have low concentrations (Ni < 110 ppm, Sc < 36 ppm). Ni and Cr exhibit strong positive correlation with MgO. Vanadium concentrations tend to increase from 9.15 to 5.90 wt.% MgO and decrease to lower MgO. Large-ion lithophile elements (LILE; e.g., Sr, Ba, Rb) have no clear trend with MgO although on average MER is more enriched in Ba, Rb and La at a given MgO than Afar. The concentrations of high-field strength elements (HFSE; e.g., Nb, Zr, Hf) increase with decreasing MgO for most samples, though some scattering is evident for some samples.

Primitive mantle-normalized multi-element variation diagrams (Fig. 5a) for representative axial rift samples show intraplate volcanic patterns with enrichment in highly to moderately incompatible trace elements similar to OIBs. The patterns are remarkably similar within each magmatic segment, but occasionally display intersecting trace element patterns among different magmatic segments, implying that the basalts might not be co-genetic. The

investigated lavas are characterised by a strong enrichment in Ba and Nb accompanied by a negative anomaly in Rb and K.

Chondrite-normalised Rare Earth Element (REE) patterns of the basalts from individual volcanoes are presented in Figure 5b. The REE patterns of the basalts are parallel to subparallel within eruptive centres, but samples from different volcanic centres commonly intersect each other. Their Lan/Ybn ratios range from 4.1 to 10.6. Heavy Rare Earth Element (HREE) are not fractionated, Tbn/Ybn ratios in the basaltic samples are between 1.48 and 1.95. Furthermore, the rift basalts have high concentrations of HREE (> 10 x chondritic values). The great similarity of trace element pattern for samples within as well as between individual volcanic centers suggests a fairly homogeneous mantle source and/or melting conditions over the entire region.

4.4. Sr, Nd, Hf and Pb isotopes

The isotopic compositions of Sr, Nd, Hf and Pb for rift basalts from Afar and MER are reported in Table 1. Owing to the young age of the examined lavas here (< 0.11 ± 0.01 Ma, Chernet et al., 1998) the measured isotopic ratios (Table 2) were not corrected for in situ

decay and hence they are considered to be representative of the initial values. The 87Sr/86Sr ratios of the rift basalts show little variation and range from 0.7035 to 0.7043. The 143Nd/144Nd ratios show a moderate range from 0.51280 to 0.51294. Basaltic lavas from the MER have significantly lower 143Nd/144Nd and higher 87Sr/86Sr ratios than the Afar lavas. The 176Hf/177Hf ratios of the basalts show a moderate variation ranging from 0.28307 to 0.28315 in Afar and from 0.28301 to 0.28307 in the MER. On the other hand, the rift lavas display a wide range of Pb isotopic ratios (206Pb/204Pb: 18.48-19.32, 207Pb/204Pb: 15.53-15.62, 208Pb/204Pb: 38.61-39.07).

Figure 6 shows isotope data of the lavas along with reference fields for various groups of

basalts from the surrounding region. The axial rift lavas plot in the depleted field relative to Bulk Earth in the Sr-Nd isotope diagram and display the common negative linear correlation which trends from more depleted compositions towards Bulk Earth values. Afar basalts are more radiogenic in Nd and less radiogenic in Sr than MER basalts. Modern rift lavas from MER volcanic fields, including those located offset 20 km to the west of the contemporaneous main rift axis (Trua et al., 1999; Peccerillo et al., 2003; Furman et al., 2006; Rooney et al., 2012; Giordano et al., 2014), display lower 143Nd/144Nd and slightly higher

Sr/ Sr values than the investigated samples. Quarternary basalts from Afar (Vidal et al., 1991; Deniel et al., 1994; Barrat et al., 1998; Daoud et al., 2010) display 87Sr/86Sr and 143Nd/144Nd ratios similar to the samples studied here. Young basalts from the Red Sea and Gulf of Aden spreading centers (Fig. 6) define more depleted Sr-Nd isotopic compositions, commonly displaying higher 143Nd/144Nd ratios. High-Ti flood basalts, ascribed to Afar plume composition (Pik et al., 1999), display a distinct field with higher 143Nd/144Nd values compared to MER basalts. 24 Ma-old shield basalts from the adjacent plateau (Rooney et al., 2014) plot far out of the axial basalt field, but trend toward and beyond the field of intraplate magmatic rocks from Sudan and Egypt, interpreted as lithospheric melts (Lucassen et al., 2008, 2013). Some of the Turkana lavas (Furman et al., 2004) plot within the field of the axial basalts. The Pan-African lithosphere, represented by small volumes of intra-plate magmatic

rocks from Sudan and Egypt (Lucassen et al., 2008, 2013), exhibit a large range of Sr/ Sr

ratios, which overlaps with some of the axial rift basalts, but 143Nd/144Nd is systematically

lower at given Sr/Sr than in Pan-African lithosphere.

In the plots of 206Pb/204Pb versus 207Pb/204Pb and 208Pb/204Pb, the axial rift lavas define

linear arrays that extend from values close to those of depleted mantle towards more radiogenic compositions. Lavas from MER are restricted to less radiogenic Pb isotopic compositions than Afar samples. Axial lavas overlap the field defined by contemporaneous lavas from the remainder of the Ethiopian rift (MER: Trua et al., 1999; Peccerillo et al., 2003;

Furman et al., 2006; Rooney et al., 2012; Giordano et al., 2014; Afar: Vidal et al., 1991; Deniel et al., 1994; Barrat et al., 1998; Daoud et al., 2010), Turkana (Furman et al., 2004) and adjacent high-Ti Oligocene flood basalts (Pik et al., 1999). Young lavas from the Red Sea and Gulf of Aden spreading centers (Altherr et al., 1990; Schilling et al., 1992; Volker et al., 1993; 1997) have characteristic non-radiogenic Pb isotopic compositions. Shield basalts have 206Pb/204Pb ratios comparable with those of Pan-African lithosphere, but are displaced towards higher 207Pb/204Pb values (Rooney et al., 2014).

As usual the 176Hf/177Hf ratios of axial basalts correlate positively with 143Nd/144Nd ratios (Fig. 6). Lavas from the MER including those located offset 20 km to the west of the contemporaneous main rift axis overlap with the axial basalts (Rooney et al., 2012). The Afro-Arabian mantle is characterized by high 176Hf/177Hf and 143Nd/144Nd ratios (Teklay et al., 2010) that overlap with the field of the young rift basalts from Afar.

5. Discussion

5.1. Fractional crystallization

The major element contents of basalts from the Quaternary magmatic segments of the MER and Afar define continuous trends and/or coherent segmented trends marking the liquid line of descent, which represents the changing magma composition during fractional crystallization. All basaltic samples are characterized by low Ni (< 110 ppm) and Cr (< 350 ppm) as well as low average MgO contents (4.5-8.7 wt.%). These values are very low compared to a primary magma (with Ni: > 400-500 ppm, Cr: > 1000 ppm and MgO: 10-15 wt.%, Frey et al., 1978; Hess, 1992) in equilibrium with a typical upper mantle mineral assemblage. Hence, it is concluded that the modern axial rift basalts from Afar and MER have undergone olivine and/or clinopyroxene fractionation, consistent with the observed

phenocrysts assemblage. In addition, the trends of CaO and Al2O3 with MgO indicate fractional crystallization of clinopyroxene and plagioclase.

5.2. Crustal contamination

The lavas are characterized by enrichment in highly and moderately incompatible elements, which may either be a consequence of crustal contamination of mantle-derived magma or derivation from enriched mantle sources. Ratios of some trace elements having a similar degree of incompatibility (i.e., Ce/Pb or Nb/U) are reasonably well constrained in OIB and MORB (Ce/Pb; 25 ± 5; Nb/U = 47 ±10, Hofmann et al., 1986, Fig.7). Most basalts of the Ethiopian rift have Ce/Pb ratios between 20 and 34 similar to those observed in modern OIB and MORB. Three samples from Kone (EK0407b and ETH23) and Gedemsa (M+M2,3), respectively, have lower Ce/Pb (13.0-18.8) and seem to be contaminated by crustal material, which generally has low Ce/Pb (3.3-3.9; Rudnick and Fountain, 1995). However, Ce/Pb displays a negative correlation with MgO (Fig. 7), ruling out significant crustal contamination as the dominant process. Thus, the trace element characteristics of MER lavas can be attributed to an enriched mantle source rather than crustal contamination. Sample 61110F (Ayelu) and ETH24 (Fantale) exhibit elevated Nb/U (112 and 96, respectively), which may reflect U loss during secondary hydrous alteration.

The isotopic composition of Sr and Nd do not form coherent trends with differentiation indexes (e.g., MgO, Fig. 7) , underlining that high level assimilation and fractional crystallization (AFC) processes did not play a major role in the evolution of these rocks. Additionally, the axial rift basalts do not possess negative anomalies in Nb and Ta in the primitive mantle-normalized multi-element variation diagrams (Fig. 6), which otherwise are generally interpreted as an indicator of crustal contamination. Therefore, it is concluded that a significant crustal contribution is absent in the genesis of the axial basalts investigated here.

5.3. Magma generation

The axial basalts from the Ethiopian rift are characterized by low CaO/Al2O3 ratios (0.590.75) and relatively flat HREE patterns (Tbn/Ybn = 1.48-1.95) together with somewhat elevated HREE concentrations higher than 10-times chondritic values which most likely suggest a mantle source containing spinel rather than garnet.

Modern basalts from the Ethiopian rift have well-documented enrichment of Ba and depletion of K and Rb relative to other elements with a similar degree of incompatibility in their primitive mantle-normalised multi-element variation diagram (Fig. 5a). Such enrichments and depletions are thought to be related to amphibole and/or phlogopite in the mantle source (e.g., Furman and Graham, 1999; Ayalew et al., 2006; Jung et al., 2005; 2012; Mayer et al, 2013, Rooney et al., 2014). According to Rosenthal et al. (2009), melting of a mantle source containing phlogopite will result in potassic and ultrapotassic magmas with K2O > Na2O, while magmas with Na2O > K2O, such as observed in the axial basalts, form through melting of amphibole-bearing mantle sources. We therefore suggest that the axial basalts from the Ethiopian rift were derived from melting of amphibole-bearing spinel peridotite sources.

Amphibole is not stable at temperatures of the convecting upper mantle or upwelling thermal plumes from the deep mantle; however, it is stable at pressure-temperature conditions of the lithospheric mantle (up to 3 GPa and 1050-1150 °C, Class and Goldstein, 1997; Mayer et al., 2014). The identification of amphibole in the source region is strong evidence for lithospheric melting and hence we conclude that the modern rift basalts formed by melting of the lithospheric mantle. Ferrando et al. (2008) and Ayalew et al. (2009) describe mantle xenoliths from adjacent plateau that contain amphibole (pargasite), strongly supporting this

argument. The presence of hydrous phases lowers the solidus temperature of the lithospheric mantle making it susceptible to melting at lower temperatures.

Modelling of upper mantle partial melting processes can be illustrated using plots of La/Yb vs. Dy/Yb ratios; such plots can also distinguish between melting in the spinel and garnet stability fields (e.g., Thirlwall et al., 1994; Jung et al., 2012; Mayer et al., 2013). An additional benefit of such plots is that mixing of melts from distinct sources produces linear mixing arrays. Results of non-modal fractional melting calculations using the composition and modal mineralogy of mantle xenoliths sampled beneath the southern MER (Beccaluva et al., 2011) are presented in Figure 8. It is evident that variable but overall small degrees of partial melting of amphibole bearing spinel lherzolite can generate the observed variation in La/Yb, but fail to explain the variation in Dy/Yb. Thus, to account for the small range in Dy/Yb ratios (2.0-2.5) observed in the axial rift samples, melts from garnet-facies mantle (3-4 %) and spinel-facies mantle (4-5 %) have to mix. The MER samples include 0-40% melt from the garnet-facies mantle, whereas the Afar lavas have 0-20% melt fraction from the garnet-facies mantle. A small proportion of melt from garnet-facies mantle in the axial rift suite is consistent with a model of melt generation in response to lithospheric extension that allows for ascent of deeper mantle to low pressure.

Attempts to estimate the temperatures and pressures of melting have been made by employing the compositions of basalts and relate them to the temperatures and pressures of magma generation (Lee et al., 2009). Using this calibration yielded potential temperatures ranging from 1122 to 1158°C and pressures between 0.91 and 1.05 GPa for Afar axial rift basalts and 1125 to 1200 °C and 1.01 and 1.24 GPa for MER basalts. The temperatures and pressures of the magmas plot within the stability field of spinel peridotite falling systematically above the stability curve for pargasitic amphibole according to Green et al. (2010) but below the stability of amphibole after Huckenholz and Gilbert (1984) (Fig. 9). Therefore, the P-T estimates suggest the generation of the parental magmas of the rift basalts

from a predominantly spinel peridotite source that contained small proportions of amphibole This interpretation is compatible with the trace element data; all axial rift lavas are enriched in Ba with depletion in Rb and K, indicating the importance of amphibole melting in their mantle source region.

The suggestion that non-exposed precursor magmas beneath the MER and Afar are likely generated at greater depth is supported by the work of Furman et al. (2006) who inferred that primitive recent Ethiopian rift basalts can be derived by melting of fertile spinel peridotite at pressures of ca. 30 kbar corresponding to a depth of 90-100 km which is significantly greater than the present estimate and the present-day lithosphere thickness. This estimate is similar to the equilibration pressures obtained on mantle xenoliths (35-55 km) from the off-rift Quaternary Debre Zeyt volcanic field (Rooney et al., 2005). An important implication of our proposed depth and pressure of melting (Figs. 8 and 9) is that the lithosphere beneath the MER has undergone significant extension prior to rifting.

5.4. Mantle source characteristics

The variations of trace element and isotopic compositions of nearly all basalts from Afar and MER indicate little or no contamination by crustal material and hence the chemical and Sr, Nd, Hf and Pb isotopic compositions of these rocks reflect the composition of their mantle source(s). Ratios of certain trace elements (e.g., Ce/Pb vs. Nb/U; Fig. 7), considered as tracer of mantle sources, plot within the field of mantle-derived basalts as observed in those of oceanic island and mid oceanic ridge basalt. However, none of the lavas exhibit trace element and isotope signatures typical of depleted mantle material like depletion in incompatible trace elements and elevated Nd isotopic ratios as shown in MORB. This suggests that these basalts most likely originated from an OIB-like mantle source.

The most obvious feature of basalts from the modern axial Ethiopian rift is the existence of

positive Ba and negative Rb and K anomalies in their primitive mantle-normalised multielement variation diagrams (Fig. 5 a). These geochemical features resulted from melting of amphibole-bearing sources. While amphibole is not stable in the sub-lithospheric mantle source (i.e., asthenosphere or mantle plume), it may exist in the lithospheric mantle (Class and Goldstein, 1997). Based on the isotopic compositions (Fig. 10), the axial basalts cannot exclusively be derived from the sub-continental lithospheric mantle characterized by less radiogenic Sr and more radiogenic Nd isotopic compositions compared to the investigated basalts (Baker et al., 2002; Beccaluva et al., 2011; Meshesha et al., 2011). There is also a broad overlap between lithospheric mantle and axial basalts in terms of their Pb isotopic ratios (Baker et al., 2002; Beccaluva et al., 2011; Meshesha et al., 2011). The range of Sr, Nd, Hf and Pb isotope ratios of young rift basalts from Afar and MER (Fig. 10) can be reconciled with the presence of an EM I mantle component that plots somewhere between the depleted mantle (DM) and high |i mantle (HIMU) sources. The input from DM and HIMU sources becomes more apparent in Afar than in MER lavas, which is consistent with the proto-oceanic crust beneath Afar. The chemical and isotopic signatures may be obtained from melting of a veined lithospheric mantle (e.g., Pilet et al., 2008, 2010). Melting is enhanced by higher temperatures in upwelling plume material, as well as the mechanical and, possibly, advective thinning in this rift zone (e.g., Rychert et al., 2012). In such a scenario, veins with EM signatures are hosted by the lithospheric mantle with a DM signature. In the course of progressive rifting and associated lithospheric thinning, the vein material, because of its non-peridotitic composition and hence lower solidus temperature, melts more easily and later induces melting of the surrounding depleted mantle to variable extent. This melting scenario results in variable mixtures depending on the relative proportion of vein material and host material and the mass balance of the elements from the two reservoirs. The signature of veined material is observed in the trace element characteristics (i.e., negative anomalies in Rb and K) because of the contrasting trace element content of a common depleted mantle or

asthenospheric mantle and in the veined lithospheric mantle. In terms of isotopic composition, the signature of the veined mantle is not distinctive due to short residence times of the vein material in the lithospheric mantle and both, host mantle and infiltrated vein material do not have isotopically extreme compositions. It is therefore suggested that melting of metasomatically enriched, amphibole-veined, depleted lithospheric mantle provides a viable mechanism for generating the combined chemical and isotopic signatures of the Afar and MER axial basalts.

The origin of the metasomatic agent that modified the lithospheric mantle is not well constrained. The vein material could have distinct origins, including material from previous subduction of oceanic crust or upwelling mantle material from ascending mantle plumes. There are at least two events that may have delivered the metasomatic component in the lithospheric mantle in this region. The first one is an ancient enrichement that occurred during the Pan-African orogeny including subduction of oceanic crust and arc-derived sedimentary material at about 550 to 900 Ma ago (i.e., Wolde 1996) explaining the range in isotopic compositions of the axial basalts. Wolde (1996) suggested that this signature is now inherited in the lithospheric mantle and likely dates back from Pan-African subduction. However, this type of mantle enrichment should result in negative or at least less positive Nb and strong positive Rb, Sr, Pb and K anomalies in primitive mantle-normalized diagrams and also in low Ce/Pb and low Nb/U ratios; features that are characteristic for subduction zone material. On the other hand, in the axial rift lavas we see positive Nb peaks, Rb and K troughs and Ce/Pb and Nb/U ratios that are characteristic for mantle material unmodified by a subduction zone component. The high 3He/4He ratios of some basalts from the area (Halldorsson et al., 2014) have been used as an argument for the presence of unmodified material at depth. It should be noted, however, that rocks beneath NE Africa predominantly have 3He/4He ratios typically found in tholeitic rocks derived from a depleted asthenospheric source (Sgualdo et al., 2015) and that metasomatism of the lithospheric mantle may uncouple the He isotope signature from

other elements and isotope systems, inducing plume or crustal signatures depending on the source of the metasomatic agent (e.g., Barry et al., 2013, 2015).

The second one is the impingement of a plume, which is ca. 30 Ma older than the age of volcanism. If the plume material delivers the metasomatic component to the lithosphere, no isotopic difference between the metasomes and the plume component will be apparent due to the short residence time between impingement and volcanism. This type of enrichment is compatible with both, trace element and isotope data, and indicate that the axial rift lavas have a distinct OIB signature. It is emphasized here that the observed OIB signature does not imply the existence of a chemical or physical plume but reflects only the presence of a component located deeper than the subcontinental lithospheric mantle. In addition basalts from adjacent plateaus, i.e. the 24 Ma old shield basalts, show isotope systematics that also suggest enriched mantle sources. Here, Rooney et al. (2014) also suggested that the veining of the lithosphere may have happened somewhat earlier than the extraction of the rift melts.

5.5. Regional source signatures

Chemical and isotopic variations of rift basalts erupted on Afar and MER define distinct fields though there are some overlaps. There is a progressive transition from continental to oceanic rifting from MER towards Afar, reflecting that continental break up started in Afar whereas rifting is less advanced in MER. Plots of trace element and isotopic ratios as a function of latitude are shown in Figure 11. La/Yb of Afar and MER basalts appear to increase from NE to SW (i.e. going from Afar towards MER), suggesting either a decrease in the degree of melting in the same direction, a change in the composition of the mantle source or both. Considering that the lithospheric thickness increases from Afar towards the MER (Hayward and Ebinger, 1996), higher La/Yb ratios indicate lower degrees of melting and hence the deeper melting depths in the MER in comparison to Afar, which in turn provides an

explanation why Afar lavas record melting of a slightly different source region than MER rocks. A plot of latitude vs. isotopic composition of the basalts (Fig. 11) reveals that 143Nd/144Nd, 176Hf/177Hf, 87Sr/86Sr and 206Pb/204Pb differ systematically between individual volcanic centres. It appears, however, that samples from Afar were derived from a source more depleted than the source that produced the MER basalts, according to their higher 143Nd/144Nd ratios, higher 176Hf/177Hf ratios and lower 87Sr/86Sr ratios.

As shown in Figure 11, samples from the Hertali (location shown in Fig. 1 and 11) magmatic field display a variation in 206Pb/204Pb without associated changes in La/Yb, 87Sr/86Sr, 143Nd/144Nd and 176Hf/177Hf. This variation is probably a consequence of evolution of the composition of the mantle source with respect to Pb without associated changes in La/Yb, 87Sr/86Sr, 143Nd/144Nd and 176Hf/177Hf. On the other hand, lavas from Dofan exhibit a large variation in 176Hf/177Hf and 143Nd/144Nd , but have nearly identical La/Yb and 87Sr/86Sr ratios. It is important to note that Pb isotopic compositions vary widely among the rift lavas and do not vary together with other element and isotope ratios, presumably indicating that Pb is located mostly in the vein-derived melt which causes different Pb isotope characteristics (from less radiogenic to slightly more radiogenic). The Pb isotopic characteristics require a long-term evolution of a lithospheric component created during a Neoproterozoic subduction event associated with the Pan-African orogeny. Note that the occurrence of vein-derived melts in the mantle is likely connected with mantle metasomatism and there is also a strong link between mantle metasomatism and highly variable composition of hydrous minerals, i.e. phlogopite or amphibole (Rosenbaum, 1993; Mayer et al., 2013). It is also important to note that for the lavas from MER and the Afar field, the slope of 207Pb/204Pb at a given 206Pb/204Pb decrease with increasing 206Pb/204Pb which is in contrast to many arc lavas where the

207nu/204nu 4. • 206nu/204nu • • • 206™/204T»U

Pb/ Pb at a given Pb/ Pb increase with increasing Pb/ Pb.

The Main Ethiopian Rift is ca. 18 Myr younger than the Red Sea and Gulf of Aden rifts and the mantle lithosphere is probably less modified by plume and stretching-related

magmatism (Wolfenden et al., 2004). It has been suggested that the diversity in trace element and isotopic composition of the MER magmas is related to the involvement of one (e.g., Ebinger and Sleep, 1999) or several plume(s) (e.g., Rogers, 2006) although the significance of a mantle plume has recently been questioned (Rychert et al., 2012). In addition, the depleted upper mantle and the continental lithosphere (Barrat et al., 1998; Trua et al., 1999; Rogers et al., 2000; Furman et al., 2006, Furman, 2007, Rooney et al., 2012) played a role in magma generation. Thus, it is reasonable to assume that the complex isotope composition may be the result of mixing processes between these various endmembers. Schilling et al. (1992) defined three mantle endmembers namely the Afar mantle plume, the Pan-African continental lithosphere and the depleted asthenospheric upper mantle. Analyses of oceanic basalts have led to the long-prevailing view that the Earth's mantle consists of a restricted number of large-scale reservoirs: a depleted mantle and about 3-4 enriched reservoirs that mix to produce the isotopic variability observed in oceanic basalts (see Stracke, 2012 for a review). On the other hand, for continental rift basalts it is difficult to distinguish melts from the deep mantle containing recycled lithosphere from shallower melting involving old subcontinental lithosphere (Rooney et al., 2012). Hence, it is often impossible to define a unique source for those basalts.

As pointed out correctly by Rooney et al. (2012), the source of the MER basalts cannot be solely within the subcontinental lithosphere and at least three endmember components are required to account for the Nd, Sr, Pb and Hf isotopic variation. The samples studied here have low 206Pb/204Pb ratios (max. 19.4) which are lower than the characteristic HIMU value (206Pb/204Pb > 20; Zindler & Hart, 1986). This implies an Afar plume composition similar to 'C', the 'common' isotopic composition observed in oceanic basalts (Hanan and Graham, 1996; Rooney et al., 2012). Considering the three-component mixing, the MER data require that the other two endmembers have low 206Pb/204Pb ratios. In addition, Rooney et al. (2012) have shown that one of these two remaining endmembers must have relatively higher

207 204 207Pb/204Pb

ratios than the other, compatible with aged Pan-African continental lithosphere. That aged Pan-African continental lithosphere plays an important role here has also been

considered elsewhere (Rooney, 2010) and is supported by radiogenic Sr/ Sr and unradiogenic 143Nd/144Nd and 176Hf/177Hf ratios of the basalts (Fig. 10). Rooney et al. (2012) presented a multi-isotope modelling that indicates mixtures of depleted mantle material (asthenosphere) of about 49-80% with material from the Afar plume (10-43%) and the Pan-African lithosphere (6-17%). These estimates seem to be broadly compatible with the relative proportions of the involved endmembers presented in Figure 10 (inset). However, the more sophisticated principal component modelling of Rooney et al. (2012) also indicated that a two-stage mixing scenario must be considered. In this model, the first stage involves mixing of asthenospheric mantle and continental lithosphere, followed by a second stage during which this hybrid upper mantle mixes with the Afar plume. This scenario, however, is incompatible with the data presented in Figure 10a (inset) because the data array do not plot along any hypothetical mixing line between a hybrid DM-EM source and the Afar plume source. In 143Nd/144Nd-206Pb/204Pb two-dimensional space (Fig. 10b; inset) a similar distribution of data is observed. Only the Afar data obtained during this study would follow the predicted model of Rooney et al. (2012). We therefore extend and modify the existing models by suggesting that during evolution of the rift, the impinging Afar plume infiltrates the overlying depleted asthenosphere creating a polluted mixed source here. Upon further movement of the plume, the overlying lithosphere is weakened and probably detached and portions of the polluted asthenospheric source are admixed with lithospheric material. This model would also explain why melting probably occured somewhat shallower at 10-15 kbar instead of melting at ca. 30 kbar (Furman et al., 2006). This melting regime is located within the stabilitiy field of spinel peridotite. Note that Furman et al. (2006) invoked a fertile peridotite source which is essentially similar to our metasomatized (= re-fertilized) mantle source; a source that is requested by the isotope data that imply ancient depletion events but

recent trace element enrichment. Effects of mantle metasomatism are apparent by the inferred presence of residual amphibole and enrichment of incompatible trace elements (Rooney et al., 2005; Furman, 2007; this study Fig. 9). One reason for the lack of consensus between the various models is that Pb isotope compositions vary widely among Ethiopian rift lavas and tend not to correlate well with Sr-Nd systematics (Furman et al., 2006).

6. Conclusions

Modern Ethiopian rift basalts were derived from melting of amphibole-bearing spinel peridotite at pressures between 0.91 and 1.05 GPa for Afar lavas and 1.01 and 1.24 GPa for MER basalts, with associated depths of melting of 27-31 km and 30-37 km, respectively. This is consistent with the present-day thickness of the underlying lithosphere (at most 50 km) inferred on the basis of mantle tomography (e.g., Bastow et al., 2005) and from mantle xenolith data (Rooney et al., 2005). The basalts were derived through melting of veined lithospheric sources in response to mantle upwelling, progressive rifting and associated lithospheric thinning. This is apparent from the compositional data for the axial basalts that indicate contrasting trace element contents of the host lithospheric mantle and enclosed enriched veins. The isotopic signature of the mixed mantle source is not very distinctive due to short residence times of the vein material and to the absence of extreme isotopic difference between host mantle and enclosed vein material. Crustal assimilation appears to be negligible throughout the rift. Chemical and isotopic variations observed along rift basalts can either be explained by variable extent of melting or regional source differences due to differnces in the preceeding history.

Acknowledgements

Funding has been provided by the Alexander von Humboldt Foundation through a Research Fellowship for experienced researchers grant to Dereje Ayalew. We thank U. Westernströer for assistance with ICP-MS measurements at U Kiel, Heidi Bayer (U Münster) for help in the isotope laboratory. Barbara Seth is thanked for measuring ETH samples at Royal Holloway, University of London. Lothar Ratschbacher is thanked for financial and logistic support during fieldwork and sampling. Positive and constructive reviews by Paterno R. Castillo and Cindy Ebinger and editorial handling by Sun-Lin Chung helped to improve this article.

References

Altherr, R., Henjes-Kunst, F., Puchelt H., Baumann, A., 1990. Asthenosphere versus lithosphere as possible source for basaltic magmas erupted during formation of the Red Sea: constraints from Sr, Pb and Nd isotopes. Earth and Planetary Science Letters 96, 269286.

Ayalew, D., Arndt, N., Bastien, F., Yirgu, G., Kieffer, B. 2009. A new mantle xenolith

locality from Simien shield volcano, NW Ethiopia. Geological Magazine 146, 144-149. Ayalew, D., Marty, B., Barbey, P., Yirgu, G. and Ketefo, E. 2006. Sub-lithospheric source for

Quaternary alkaline Tepi shield, southwest Ethiopia. Geochemical Journal 40, 47-56. Ayalew, D., Yirgu G., and Pik R. 1999. Geochemical and isotopic (Sr, Nd and Pb) characteristics of volcanic rocks from southwestern Ethiopia. Journal of African Earth Sciences 29, 381-391.

Baker, J., Chazot, G., Menzies, M.,Thirlwall, M., 2002. Lithospheric mantle beneath Arabia: A Pan-African protolith modified by the Afar and older plumes, rather than a source for continental flood volcanism?, in: Menzies, M.A., Klemperer, S.L., Ebinger, C.J., Baker, J., (eds.) Volcanic Rifted Margins. Geological Society of America Special Paper 362, 65-80. Barberi, F., Civetta, L. and Varet, J. 1980. Sr isotopic composition of Afar volcanics and its

implication for mantle evolution. Earth and Planetary Science Letters 50, 247-259.

Barrat, J., Fourcade, S., Jahn, B., Cheminée, J., Capdevila, R., 1998. Isotope (Sr, Nd, Pb, O) and trace-element geochemistry of volcanics from the Erta'Ale range (Ethiopia). Journal of Volcanology and Geothermal Research 80, 85-100.

Barry, P.H., Hilton, D R., Howarth, G.H., Pernet-Fisher, J.F., Day, J.M., Taylor, L.A., 2013. Helium isotope evidence for plume metasomatism of Siberian continental lithosphere. American Geophysical Union, Fall Meeting 2013, abstract #V33A-2728.

Barry, PH., Hilton, DR., Day, J.M.D., Pernet-Fisher, J.F., Howarth, G.H., Magna, T., Agashev, A.M., Pokhilenko, N.P., Pokhilenko, L.N., Taylor, L.A., 2015. Helium isotopic evidence for modification of the cratonic lithosphere during the Permo-Triassic Siberian flood basalt event. Lithos 216-217, 73-80.

Bastow, I., Stuart, G., Kendall, J., Ebinger, C., 2005. Upper-mantle seismic structure in a region of incipient continental breakup: northern Ethiopian rift. Geophysical Journal International 162, 479-493.

Beccaluva, L., Bianchini, G., Ellam, R. M., Natali, C., Santato, A., Siena, F., Stuart, M. F., 2011. Peridotite xenoliths from Ethiopia: inferences on mantle processes from Plume to Rift settings. Geological Society of America Special Paper 478, 77-104.

Boccaletti, M., Mazzuoli, R., Bonini, M., Trua, T., Abebe, B., 1999. Plio-Quaternary volcanotectonic activity in the northern sector of the Main Ethiopian Rift: relationships with oblique rifting. Journal of African Earth Sciences 29, 679-698.

Bonini, M., Corti, G., Innocenti, F., Manetti, P., Mazzarini, F., Abebe, T., Pecskay, Z., 2005. Evolution of the Main Ethiopian Rift in the frame of Afar and Kenya rifts propagation, Tectonics 24, doi:10.1029/2004TC001680.

Brotzu, P., Ganzerli-Valentini, M., Morbidelli, L., Piccirillo, E., Stella, R., Traversa, G., 1981. Basaltic volcanism in the northern sector of the Main Ethiopian Rift. Journal of Volcanology and Geothermal Research 10, 365-382.

Brotzu, P., Morbidelli, L., Piccirillo, E., Traversa, G., 1980. Volcanological and magmatological evidence of the Boseti volcanic complex (Main Ethiopian Rift), Atti dei Convegni Lincei, Accademia Nazionale dei Lincei 47, 317-366.

Chernet, T., Hart, W., Aronson, J., Walter, R., 1998. New age constraints on the timing of volcanism and tectonism in the northern Main Ethiopian Rift-southern Afar transition zone (Ethiopia), Journal of Volcanology and Geothermal Research 80, 267-280.

Class, C., Goldstein, S., 1997. Plume lithosphere interaction in the ocean basins: constraints from source mineralogy. Earth and Planetary Science Letters 150, 245-260.

Daoud, M.A., Maury, R.C., Barrat, J-A., Taylor, R.N., Le Gall, B., Guillou, H., Cotten, J., Rolet , J., 2010. A LREE-depleted component in the Afar plume: further evidence from Quaternary Djibouti basalts. Lithos 114, 327-336.

Daly, E., Keir, D., Ebinger, C. J., Stuart, G. W., Bastow, I. D., Ayele, A., 2008. Crustal tomographic imaging of a transitional continental rift: the Ethiopian rift. Geophysical Journal International 172, 1033-1048.

Deniel, C., Vidal, P., Coulon, C., Vellutini, P., Piguet, P., 1994. Temporal evolution of mantle sources during continental rifting: the volcanism of Djibouti (Afar). Journal of Geophysical Research 99, 2853-2869.

Dugda, M., Nyblade, A., Julia, J., Ammon, C., 2005. Crustal structure in Ethiopia and Kenya from receiver function analysis; implications for rift development in eastern Africa, Journal of Geophysical Research, Solid Earth 110, B01303.

Dugda, M.T., Nyblade, A.A., Julia, J., 2007. Thin lithosphere beneath the Ethiopian plateau revealed by a joint inversion of rayleigh wave group velocities and receiver functions. Journal of Geophysical Research, Solid Earth, 112, B08305.

Ebinger, C. J., Sleep, N. H., 1988. Cenozoic magmatism throughout east Africa resulting from impact of a single plume. Nature 395, 788-791.

Ebinger, C., Casey, M., 2001. Continental breakup in magmatic provinces: An Ethiopian

example. Geology 29, 527-530.

Ferrando, S. Frezzotti, M. L., Neumann, E.-R., De Astis, G., Peccerillo, A., Ayalew, D., Gezahegn, Y., Teklewold, A., 2008. Composition and thermal structure of the lithosphere beneath the Ethiopian plateau: evidence from mantle xenoliths in basanites, Injibara, Lake Tana Province. Mineralogy and Petrology 93, 47-78.

Frey, F., Green, D., Roy, S., 1978. Integrated models of basalt petrogenesis: a study of quartz tholeiites to olivine melilites from South Eastern Australia utilizing geochemical and experimental petrological data. Journal of Petrology 19, 463-513.

Furman, T., 2007. Geochemistry of East African Rift basalts: an overview. Journal of African Earth Sciences 48, 157-160.

Furman, T., Bryce, J., Rooney, T., Hanan, B., Yirgu, G., Ayalew, D., 2006. Heads and tails: 30 million years of the Afar plume, in: Yirgu, G., Ebinger, C. J., Maguire, P. K. H. (eds), The Afar Volcanic Province within the East African Rift System. Geological Society of London Special Publication 259, 95-119.

Furman, T., Graham, D., 1999. Erosion of lithospheric mantle beneath the East African Rift system: geochemical evidence from the Kivu volcanic province. Lithos 48 (1-4), 237-262.

Garbe-Schönberg, C.-D., 1993. Simultaneous determination of 37 trace elements in 28 international rock standards by ICP-MS. Geostandards Newsletter 17, 81-97.

George, R., Rogers, N., Kelley, S., 1998. Earliest magmatism in Ethiopia: Evidence for two mantle plumes in one flood basalt province. Geology 26, 923-926.

Giordano, F., D'Antonio, M., Civetta, L., Tonarini, S., Orsi, G., Ayalew, D., Yirgu, G., Dell'Erba., F., DiVito, M.A. and Isaia, R., 2014. Genesis and evolution of mafic and felsic magmas at Quaternary volcanoes within the Main Ethiopian Rift: insights from Gedemsa and Fanta 'Ale complexes. Lithos 188, 130-144.

Green, D. H., Hibberson, W. O., Kovacs, I., Rosenthal, A., 2010. Water and its influence on the lithosphere/asthenosphere boundary. Nature 467, 448-451.

Halldorsson, S.A., Hilton, D.R., Scarsi, P., Abebe, T., Hopp J., 2014. A common mantle plume source beneath the entire East African Rift System revealed by coupled heliumneon systematics. Geophysical Research Letters 41, 2304-2311.

Hammond, J.O.S., Kendall, J. M., Stuart, G. W., Keir, D., Ebinger, C., Ayele, A., Belachew, M., 2011. The nature of the crust beneath the Afar triple junction: evidence from receiver functions. Geochemistry, Geophysics, Geosystems 2, D0I:10.1029/2011GC003738.

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.

Hart, W., WoldeGabriel, G., Walter, R., Mertzman, S., 1989. Basaltic volcanism in Ethiopia: constraints on continental rifting and mantle interactions. Journal of Geophysical Research 94, 7731-7748.

Hayward, N., Ebinger, C., 1996. Rift kinematics and along-axis segmentation in northern Afar. Tectonics 15, 244-257.

Hess, P., 1992. Phase equilibria constraints on the origin of ocean floor basalts. Geophysical Monograph 71, 67-102.

Hofmann, A.W., Jochum, K.P., Seufert, M., White, W. M. 1986., Nb and Pb in oceanic basalts: new constraints on mantle evolution. Earth and Planetary Science Letters 79, 3345.

Hofmann, C., Courtillot, V., Feraud, G., Rochette, P., Yirgu, G., Ketefo, E., Pik, R., 1997. Timing of the Ethiopian flood basalt event and implications for plume birth and global change. Nature 389, 838-841.

Huckenholz, H. G., Gilbert, M. C., 1984. Stabilität von Ca- Amphibol in Alkalibasalten der Hocheifel. Fortschritte der Mineralogie 62(Bh1), 106-107.

Irvine, T., Baragar, W., 1971. A guide to the chemical classification of the common volcanic rocks. Canadian Journal of Earth Sciences 8, 523-548.

Jung, S., Pfänder, J., Brugmann, G., Stracke, A., 2005. Sources of primitive alkaline volcanic rocks from the Central European Volcanic Province (Rhön, Germany) inferred from Hf, Os and Pb isotopes. Contributions to Mineralogy and Petrology 150, 546-559.

Jung, S., Vieten, K., Romer, R.L., Mezger, K., Hoernes, S., Satir, M., 2012. Petrogenesis of Tertiary alkaline magmas in the Siebengebirge, Germany. Journal of Petrology, 53, 23812409.

Kendall, J., Stuart, G., Ebinger, C., Bastow, I., Keir, D., 2005. Magma-assisted rifting in Ethiopia. Nature 433, 146-148.

Keranen, K., Klemperer, S., Gloaguen, R., 2004. Three-dimensional seismic imaging of a protoridge axis in the Main Ethiopian Rift. Geology 32, 949-952.

Kurz, T., Gloaguen, R., Ebinger, C., Casey, M., Abebe, B., 2007. Deformation distribution in the Main Ethiopian Rift (MER); a remote sensing study. Journal of African Earth Sciences 48, 100-114.

Lahitte, P., Gillot, P., Courtillot, V., 2003. Silicic central volcanoes as precursors to rift propagation: the Afar case. Earth and Planetary Science Letters 207, 103-116.

LeBas, M., LeMaitre, R., Streckeisen, A., Zanettin, B., 1986. A chemical classification of volcanic rocks based on the total alkali-silica diagram. Journal of Petrology 27, 745-750.

Lee, C-T.A., Luf, P., Plank, T., Dalton, H., Leeman, W.P., 2009. Constraints on the depths and temperatures of basaltic magma generation on Earth and other terrestrial planets using new thermobarometers for mafic magmas. Earth and Planetary Science Letters 279, 20-33.

Lucassen, F., Franz, G., Romer, R.L., Pudlo, D., Dulski, P., 2008. Nd, Pb, and Sr isotope composition of Late Mesozoic to Quaternary intra-plate magmatism in NE-Africa (Sudan, Egypt): high-^ signatures from the mantle lithosphere. Contributions to Mineralogy and Petrology 156, 765-784.

Lucassen, F., Pudlo, D., Franz, G., Romer, R.L., Dulski, P., 2013. Cenozoic intra-plate magmatism in the Darfur volcanic province: mantle source, phonolite-trachyte genesis and

relation to other volcanic provinces in NE Africa. International Journal of Earth Sciences 102, 183-205.

Maguire, P.K.H., Keller, G.R., Klemperer, S.L., Mackenzie, G.D., Keranen, K., Harder, S., O'Reilly, B., Thybo, H., Asfaw, L., Khan, M.A. and Amha, M., 2006. Crustal structure of the northern Main Ethiopian Rift from the EAGLE controlled-source survey; a snapshot of incipient lithospheric break-up. In: Yirgu, G., Ebinger, C.J. and Maguire P.K.H. (eds) 2006. The Afar Volcanic Province within the East African Rift System. Geological Society London, Special Publications 259, 269-291.

Marty, B., Pik, R., Yirgu, G., 1996. Helium isotopic variations in Ethiopian plume lavas; nature of magmatic sources and limit on lower mantle contribution. Earth and Planetary Science Letters 144, 223-237.

Mayer, B., Jung, S., Romer, R. L., Pfänder, J. A., Klügel, A. Pack., A., Gröner, E., 2014. Amphibole in alkaline basalts from intraplate settings: implications for the petrogenesis of alkaline lavas from the metasomatised lithospheric mantle. Contributions to Mineralogy and Petrology 167, doi:10.1007/s00410-014-0989-3.

Mayer, B., Jung, S., Romer, R. L., Stracke, A., Haase, K. M. and Garbe-Schönberg, C.-D., 2013. Petrogenesis of Tertiary Hornblende-bearing Lavas in the Rhön, Germany. Journal of Petrology 54, 2095-2123.

Mohr, P., 1983. Perspectives on the Ethiopian Volcanic Province. Bulletin of Volcanology 46, 23-43.

Münker C., Weyer S., Scherer E. E., Mezger K., 2001. Separation of high field strength elements (Nb, Ta, Zr, Hf, and Lu) from rock samples for MC-ICPMS measurements. Geochemistry Geophysics Geosystems 2, doi:10.1029/2001/20011GC000183.

O'Connor, J., Stoffers, P., van den Bogaard, P., McWilliams, M., 1999. Seamount age evidence for sigificantly slower African plate motion since 19 to 30 Ma. Earth and Planetary Science Letters 171, 575-589.

Peccerillo, A., Barberio, M., Yirgu, G., Ayalew, D., Barbieri, M., Wu, T. W., 2003. Relationship between mafic and peralkaline silicic magmatism in continental rift settings: a petrological, geochemical and isotopic study of the Gedemsa volcano, central Ethiopian rift. Journal of Petrology 44, 2003-2032.

Pik, R., Deniel, C., Coulon C., Yirgu, G. Marty, B., 1999. Isotopic and trace element signatures of Ethiopian flood basalts; evidence for plume-lithosphere interactions. Geochimica et Cosmochimica Acta 63, 2263-2279.

Pilet, S., Baker, M.B., Stolper, E.M., 2008. Metasomatized lithosphere and the origin of alkaline lavas. Science 320, 916-919.

Pilet, S., Ulmer, P., Villiger, S., 2010. Liquid line of descent of a basanitic liquid at 1-5 GPa: constraints on the formation of metasomatic veins. Contributions to Mineralogy and Petrology159, 621-64.

Rychert, C.A., Hammond, J.O.S., Harmon, N., Kendall, M.J., Keir, D., Ebinger, C., Bastow, I.D., Ayele, A., Belachew, M., Stuart, G., 2012. Volcanism in the Afar Rift sustained by decompression melting with minimal plume influence. Nature Geoscience 5, 406-409.

Rogers, N. W., 2006. Basaltic magmatism and the geodynamics of the East African Rift System. in: Yirgu, G., Ebinger, C.J., Maguire, P.K.H. (eds). The Afar Volcanic Province within the East African Rift System. Geological Society London Special Publications 259, 77-93.

Rogers, N., Macdonald, R., Fitton, J., George, R., Smith, M., Barreiro, B., 2000. Two mantle plumes beneath the East African rift system: Sr, Nd and Pb isotope evidence from Kenya Rift basalts. Earth and Planetary Science Letters 176, 387-400.

Rooney, T. O., 2010. Geochemical evidence of lithospheric thinning in the southern Main Ethiopian Rift. Lithos 117, 33-48

Rooney, T.O., Hanan, B.B., Graham, D.W., Blichert-Toft, J., Schilling, J-G., 2012. Upper mantle pollution during Afar plume-continental rift interaction. Journal of Petrology 53, 365-389.

Rooney, T., Furman, T., Bastow, I., Ayalew, D., Yirgu, G., 2007. Lithospheric modification during crustal extension in the Main Ethiopian Rift. Journal of Geophysical Research 112, B10201, doi:10.1029/2006JB004916.

Rooney, T., Furman, T., Yirgu, G., Ayalew, D., 2005. Structure of the Ethiopian lithosphere; xenolith evidence in the Main Ethiopian Rift, Geochimica et Cosmochimica Acta 69, 3889-3910.

Rooney, T.O., Nelson, W.R., Dosso, L., Furman, T., Hanan, B., 2014. The role of continental lithosphere metasomes in the production of HIMU-like magmatism on the northeast African and Arabian plates. Geology 42, 419-422.

Rosenbaum, J., 1993. Mantle phlogopite: A significant lead repository? Chemical Geology 106, 475-483.

Rosenthal, A., Foley, S.F., Pearson, D.G., Nowell, B.M., Tappe, S., 2009. Petrogenesis of strongly alkaline primitive volcanic rocks at the propagating tip of the western branch of the East African Rift. Earth and Planetary Science Letters 284, 236-248.

Rychert, C.A., Hammond, J.O.S., Harmon, N., Kendall, M.J., Keir, D., Ebinger, C., Bastow, I.D., Ayele, A., Belachew, M., Stuart, G., 2012. Volcanism in the Afar Rift sustained by decompression melting with minimal plume influence. Nature Geoscience 5, 406-409.

Sgualdo, P., Aviado, K., Beccaluva, L., Bianchini, G., Blichert-Toft, J., Bryce, J.G., Graham, D.W., Natali, C., Siena, F., 2015. Lithospheric mantle evolution in the Afro-Arabian domain: Insights from Bir Ali mantle xenoliths (Yemen). Tectonophysics 650, 3-17.

Shaw, D.M., 1970. Trace element fractionation during anatexis. Geochimica et Cosmochimica Acta 34(2), 237-243.

Schilling, J. G., Kingsley, R. H., Hanan, B. B., Mccully, B. L., 1992. Nd-Sr-Pb isotopic variations along the Gulf of Aden: Evidence for Afar mantle plume continental lithosphere interaction. Journal of Geophysical Research Solid Earth 97, 10927-10966.

Stracke, A., 2012. Earth's heterogeneous mantle: a product of convection-driven interaction between crust and mantle. Chemical Geology 330-331, 274-299.

Stuart, G.W., Bastow, I.D., Ebinger, C.J., 2008. Crustal structure of the northern Main Ethiopian Rift from receiver function studies. In: Yirgu, G., Ebinger, C.J. Maguire, P.K.H. (eds.) The Afar Volcanic Province within the East African Rift System. Geological Society London Special Publications 259, 253-267.

Sun, S., McDonough, W., 1989. Chemical and isotopic systematics of oceanic basalts: implications for mantle compositions and processes. Geological Society Special Publications 42, 313-345.

Thirlwall, M., 1991. Long-term reproducibility of multicollector Sr and Nd isotope ratio analysis. Chemical Geology 94, 85-104.

Thirlwall, M., Jenkins, C., Vroon, P., Mattey, D., 1997. Crustal interaction during construction of oceanic islands; Pb-Sr-Nd-O isotope geochemistry of the shield basalts of Gran Canaria, Canary Islands. Chemical Geology 135, 233-262.

Thompson, R.N., Gibson, S.A., 1994. Magmatic expressions of lithospheric thinning across continental rifts. Tectonophysics 233, 41-68.

Trua, T., Deniel, C., Mazzuoli, R., 1999. Crustal control in the genesis of Plio-Quaternary bimodal magmatism of the Main Ethiopian Rift (MER): geochemical and isotopic (Sr, Nd, Pb) evidence. Chemical Geology 155, 201-231.

Vidal, P., Deniel, C., Vellutini, P. J., Piguet, P., Coulon, C., Vincent, J., Audin, J., 1991. Changes of mantle source in the course of a rift evolution. Geophysical Research Letters 18, 1913-1916.

Volker, F., McCulloch, M.T., Altherr, R., 1993. Submarine basalts from the Red Sea: new Pb, Sr, and Nd isotopic data. Journal of Geophysical Research 20, 927-930.

Volker, F., Altherr, R., Jochum, K. P., McCulloch, M.T., 1997. Quaternary volcanic activity of the southern Red Sea: new data and assessment of models on mantle sources and Afar plume-lithosphere interaction. Tectonophysics 278, 15-29.

Whaler, K.A., Hautot, S., 2006. The electrical resistivity structure of the crust beneath the Main Ethiopian Rift, in: Yirgu, G., Ebinger, C.J., Maguire, P.K.H. (eds), The Afar Volcanic Province within the East African Rift System, The Geological Society of London Special Publications 259, 293-305.

Wolde, B., 1996. Magmatism in the Cenozoic Ethiopian rift zone : the case against the Afar plume hypothesis. Doctoral thesis, Tekniska högskolan i Lulea, 204D.

Wolfenden, E., Ebinger, C., Yirgu, G., Deino, A., Ayalew, D., 2004. Evolution of the northern Main Ethiopian rift: birth of a triple junction. Earth and Planetary Science Letters 224, 213-228.

Zindler, A., Hart, S., 1986. Chemical geodynamics, Annual Reviews Earth and Planetary Science Letters 14, 493-571.

Figure captions

Fig. 1. Distribution of the magmatic segment along the Ethiopian rift (i.e., Afar and MER), after Wolfenden et al. (2004). Inset shows overview of the Afro-Arabian Rift system (east African rift system, and Red Sea and Gulf of Aden rifts).

Fig. 2. Total alkalis (Na2O +K2O) versus SiO2 (TAS; Le Bas et al., 1986) diagram for axial rift basalts from Ethiopian rift (Afar: Ayelu, Hertali and Dofan, and MER: Fantale, Kone, Bosetti and Gedemsa), showing their moderate alkaline affinity. Alkaline-subalkaline division line is from Irvine and Baragar (1971). Grey field denotes literature data from MER (Furman et al., 2006).

Fig. 3. Variations of major elements vs. MgO of recent axial rift basalts along Afar and MER.

Fig. 4. Variations of trace elements vs. MgO for axial rift basalts along Afar and MER.

Fig. 5. (a) Primitive mantle-normalized multi-element variation diagrams and (b) chondrite-normalised Rare Earth Element (REE) patterns for representative axial rift basalts from the MER and Afar. Primitive mantle and chondrite values for normalization are from Sun and McDonough (1989).

Fig. 6. Strontium, Nd, Pb and Hf isotope ratios of axial basalts from Afar and MER in relation to the mantle reservoirs identified for oceanic basalts (Zindler and Hart, 1986; Stracke 2012). Also shown for comparisons are compositional fields from various groups of basalts from adjacent plateau (Pik et al., 1999) and shields (Rooney et al., 2014), Red Sea and Gulf of Aden spreading centers, contemporaneous basalts from Afar and MER, and Pan-African

lithospheric mantle. Data for Afar from Barrat et al. (1998), Daoud et al. (2010), Deniel et al. (1994), and Vidal et al. (1991), for MER from Trua et al. (1999), Peccerillo et al. (2003), Furman et al. (2006), and Rooney et al. (2012) and for Afro-Arabian lithospheric mantle from Baker et al. (2002), Teklay et al. (2010), Beccaluva et al. (2011). DM; depleted mantle, HIMU; high |i, EMI; enriched mantle I, EMII; enriched mantle II, BSE; bulk silicate earth.

Fig. 7. Variations of Ce/Pb vs. Nb/U, and Ce/Pb, 87Sr/86Sr and 143Nd/144Nd vs. MgO of recent axial basalts from Afar and MER. The grey bars denote the range of Ce/Pb and Nb/U in oceanic basalts (Hofmann et al., 1986).

Fig. 8. La/Yb vs. Dy/Yb for axial basalts from Afar and MER. Partial melting curves are calculated using a non-modal batch melting model (Shaw, 1970). Sources are amphibole-garnet and -spinel lherzolites (garnet lherzolite: 0.58 Ol, 0.15 Opx, 0.20 Cpx, 0.02 Gt, 0.05 amph that melts in the proportion 0.10 Ol, 0.20 Opx, 0.40 Cpx, 0.10 Gt, 0.20 amph and spinel lherzolite: 0.58 Ol, 0.15 Opx, 0.20 Cpx, 0.02 sp, 0.05 amph that melts in the proportion 0.10 Ol, 0.20 Opx, 0.40 Cpx, 0.10 sp, 0.20 amph; adopted from Jung et al., 2012). The straight line indicate mixing between melt fractions from garnet-facies mantle and melt fractions from spinel-facies mantle. Adjusted source composition (La 2-7 ppm,Yb 0-19 ppm, Dy 0-45 ppm) is within the range of peridotite xenolith from Mega, southern Ethiopia (Beccaluva et al., 2011). Mineral-melt distribution coefficients were adopted from Jung et al. (2012). Numbers on model curves indicate the per cent melting and mixing.

Fig. 9. Pressure-temperature diagram to illustrate the potential source region of the modern axial rift basalts from Afar and MER (adopted from Jung et al., 2012). It should be noted that, based on geochemical arguments discussed in the text (presence of restitic amphibole,

predominance of melts from spinel peridotite), these temperature and pressure estimates must be viewed as minimum estimates.

Fig. 10. (a) 87Sr/86Sr vs. 206Pb/204Pb and (b) 143Nd/144Nd vs. 206Pb/204Pb for the axial rift basalts to illustrate that the distribution of the isotopes require a mixture of a lithospheric mantle component with a component that itself originated by mixing of the depleted asthenosphere and a plume source (Afar Plume).

Fig. 11. Variations of latitude vs. trace element and isotopic ratios of axial basalts from Afar and MER. The isotopic compositions of the axial basalts indicate that the magmas originated from a lithospheric source with varying contributions from various asthenospheric mantle components. An input from a HIMU ('C') source is more apparent in lavas erupted on Afar, compatible with the presence of proto-oceanic crust in this area.

Figure 1

T—I-1-r

Foidite

• Afar

Rhyolile

J_I_1_L

Si02 (wt. %)

Figure 2

6 8 MgO (wt%)

6 8 MgO (wt%)

6 8 MgO (wt%)

6 8 MgO (wt%)

6 8 MgO (wt%)

6 8 MgO (wt%)

6 8 MgO (wt%)

Figure 3

160 >120

6*bo U

i— 200

• Afar O MER

Q MER (Furman

ei al. 2006]

6 8 MgO (wt%)

6 8 MgO (wt%)

6 8 MgO (wt%)

MgO (wt%)

Figure 4

6110A (Afar) 6110D (Afar) ETH-7 (MER) O ETH-6 {MER)

I MER ¡Furman ei al. 2006)

J_I_I_I_I_I_I_I_I_I_I_I_I_I_I_I_I_I_I_I_L

Thu Nb K LaCePbpr Sr p Nd Sm Zr Eu Ti cd Dy Y Ho Yb

Figure 5a

6110A (Afar) -9- 6110D (Afar) ■Q- ETH-7 (MER) -O ETH-6 (MER) ~| MER (rurman et al. 2006)

La Ce Pr Nd Sm Eu Cd Tb Dy Ho Er Tm Yb Lu

Figure 5b

0.5132-

.. t DM

\ Red Sea and ' ■..Gulf of Aden

0.5130 -

High-T¡ Oligoœne flood basalts

0.5126

0.7026 0.7030 0.7034 0.7038 0.7042 0.7046 87Sr/86Sr

15.7 -

15.6 -

Shield basalts

{ EM II

Afro-Arabian lithospheric mantle

19.0 20.0

206Pb/204Pb

0.2833

• Afar O MER

O MER (Furman el al., 2006) ÍPooneyet al., 2012)

0.2831 -

0.2827

Afro-Arabian -

lithospheric mantle

High-Ti Ollgocene flood basalt

. Red Sea and ■„■■ ' DM Gulf of Aden

J_J_I_I_I_I_l_I_I_I_I_I_I_I_

0.5127 0.5128 0.5129

143Nd/144Nd

0.5130

18.0 19.0 20.0

206Pb/204Pb

Figure 6

0.7043

0.7041

0.7039

6 8 MgO (wt%)

0.7037

0.7035

■ o w O f~\ ("i

_o o U L> o °

: 6> O oo 0

- o o

■ o 0

6 8 MgO (wt%)

0.51294

0.51288

S? 0.51282

0.51276'

6 8 MgO (wt%)

-8 o O

C^J o ooo

1 . • ip .

Figure 7

5% 4% 10%

Amph-grt peridotite melting

Mixing line of 4% melt from amph-grt peridotite and 4 % melt from amph-sp peridotite

Amph-sp peridotite melting X-X-

2% « Afar o MER

O MER (Furman

et ah 2006)

Figure 8

• Afar O MER

O MER (Fiirman

et al. 2COG)

Lithosphère boundary

3 - -a

w s- 1 ""

o 1 ° °

o ° O

1000 1200

Temperature (°C)

Figure 9

0.706'

CD 00^

CO h-00

0.703 -

High-Ti Oligocene flood basalts

Red Sea and Gulf of Aden

Í mfc Afar plume HIMU

17.0 ia.O 19.0 20 0 21.0 206pbi,204pb

shield basalts

NE-Afrlca '-__J-' A mag matic rocks

■ ■ I .

Afar plume I .... I

17.0 18.0 19.0

206pb/204pb

0.5134-

Red Sea and Gulf of Aden

§ 0.5130

0.5126

0.5122

High-Ti Oligocene flood basalts

i " * HIMU Afar plume

* EM 1 . . 1 . . .

17.0 18 0 19.0 20.0 21.0 206pb,20+pb

MER Afar plume EMM

NE-Africa

Shield basalts magmatic rocks

I Afar O MER

O MER (Furman

et al. 2006)

I .... I ■ --.

17.0 18.0 19.0 20.0 206Pb/204Pb

Figure 10

15 L 10 5 7 0 -F-

0.70440

CD 00^

0.70380

0.70320

0.51295

5 0.51285

0.51275

0.28320 -

0.28310 -

0.28300 -

XI CL CD O CM

19.2 18.8 18.4

Figure 11

o • 9 8

8.0 8.4 8.8 9.2

Latitude in degree

Table 1 Sr, Nd, Hf and Pb isotopic compositions of recent axial basalts from Afar and MER, Ethiopian rift.

Volcano Sample Lat. (°) 87Sr/86Sr 2G 143Nd/144Nd 2G 176Hf/177Hf 2G 206Pb/204Pb 207Pb/204Pb 208Pb/204Pb

Afar Ayelu 61110F 10 0.703839 11 0.512888 06 0.283093 06 18.484 15.533 38.613

Hertali 61110B 9.65 0.703541 11 0.512917 06

Hertali 61110D 9.65 0.703565 11 0.512944 06 0.283127 06 19.318 15.625 39.068

Hertali 61110C 9.65 0.703565 11 0.512924 06 0.283129 06 19.280 15.607 39.057

Hertali 61110A 9.65 0.703607 11 0.512904 06 0.283119 06 18.953 15.559 38.814

Dofan 61109A 9.35 0.703716 11 0.512908 06 0.283146 05 18.895 15.546 38.686

Dofan 61109B 9.35 0.703724 11 0.512905 06

Dofan ETH-14 9.35 0.703757 30 0.512893 04 0.283070 04 18.901 15.569 38.807

Fantale ETH-24 9.00 0.703998 03 0.512893 03 0.283041 08 18.668 15.577 38.660

Fantale ETH-20 9.00 0.704006 05 0.512864 04 0.283034 06 18.743 15.592 38.805

Fantale ETH-19 9.00 0.703971 06 0.512849 03 0.283037 06 18.731 15.578 38.757

Kone ETH-23 8.85 0.704352 90 0.512791 04 18.686 15.601 38.864

Boseti ETH-8 8.70 0.703976 04 0.512841 02 0.283042 04 18.890 15.594 38.927

Boseti ETH-6 8.70 0.704248 07 0.512832 04 0.283012 15 18.977 15.605 39.025

Boseti ETH-7 8.70 0.704144 07 0.512825 03 0.283013 05 18.850 15.601 38.976

Gedemsa M+M 2.3 8.25 0.703849 11 0.512871 06

Gedemsa GMB009 8.25 0.703915 11 0.512879 06 0.283070 06 18.889 15.561 38.733

Hignlights:

• We present new major and trace elements and isotope (Sr, Nd, Hf and Pb) data for young Ethiopian rift basalts

• We model the melting conditions of the magmas.

• We examine the nature of the mantle source.

• We compare our findings with the previous results