ARTICLE
Petrogenesis of Butajira-Kibet Quaternary basaltic rocks; Central Main Ethiopian Rift
2023 Volume 57 Issue 3 Pages 100-117
Details
2023 Volume 57 Issue 3 Pages 100-117
Major and trace element data are presented for Quaternary basalts from Butajira-Kibet area, close to the western escarpment of the Central Main Ethiopian Rift to investigate the petrogenetic processes involved and the nature of mantle source compositions. The data indicate that the basalts are mildly alkaline characterized by moderate to high contents of MgO (6.85–9.00 wt.%) and TiO2 (2.40–2.92 wt.%). In the primitive mantle-normalized multi-element diagram, they exhibit troughs at Th, U, Rb and K and peaks at Ba, Nb, Ta, La, Pb and Sr with positive Eu/Eu*(1.01–1.17). Incompatible element ratios e.g., Ba/Nb (8.92–16.25), Rb/Nb (0.42–0.70), K/Nb (233.00–340.00) and Nb/U (47.97–61.05) suggest that the basalts are unaffected by crustal contamination. The acquired data do not evidence fractionation of plagioclase and clinopyroxene. On the other hand, relationships between major elements and their ratios (e.g., Na2O and Ca2O/Al2O3) and trace element ratios (e.g., Zr/Y and (La/Sm)N) indicate partial melting of a clinopyroxene-rich mantle at medium pressure. Chondrite-normalized REE diagram is characterized by relatively flat HREE (TbN/YbN = 1.73–2.13) and significantly enriched HREE (higher than 10-times chondritic values) patterns. LREE/MREE versus MREE/HREE modelling of non-modal batch melting indicates that the basalts are formed by 1.5–3% partial melting of spinel peridotite mantle source with minor contribution from garnet-bearing peridotite. Primitive mantle-normalized multi-element diagram and trace element ratios, such as Nb/Zr, Nb/U, Ba/La, Rb/Zr and Ba/Nb, suggest that the studied basalts are not affected by carbonatite and/or hydrous phase metasomatism. Rather, they are closely similar to the OIB source (mainly EM-I) and the SCLM. They are most likely the result of significant interaction of melts from a mantle plume and the SCLM.
Most of the Earth’s crust is of magmatic origin, attesting to the enormous role that volcanic and related magmatic processes have played in forming the outermost solid crust of our planet. In addition, in the context of plate tectonics, the distribution of volcanoes in the geological past and present can be closely linked to the dynamics of the crust and the mantle. Thus it is barely surprising that many geoscientists work in terrains or on research topics directly or indirectly associated with volcanic rocks (Tilling, 1989).
According to Rooney (2010), the lithosphere is faulted and stretched to accommodate strain in the early stages of continental rifts, which are characterized by relatively large zones of mechanical extension. A famous modern example of continental rifting situated over one or more mantle plumes and surrounded by broad regional plateaus is the East African Rift System (EARS), which includes the Main Ethiopian Rift (MER) (Fig. 1). The MER delivers an excellent basis for studying extensional mafic magmatism (Furman, 2007).
Digital elevation model showing the central sector of the Main Ethiopian Rift and its bounding plateaus.
In the Main Ethiopian Rift, magmatism and rifting are hypothesized to be linked to mantle plume activity (e.g., Pik et al., 1999; Peccerillo et al., 2003). Huge volumes of volcanic products have been erupted in recent stages of the Ethiopian rift opening, with a high prevalence of silicic rocks, minor basalts, and a rarity or absence of intermediate compositions (e.g., Mohr, 1971; Mohr and Zanettin, 1988; Peccerillo et al., 2007). Many volcanoes, particularly the continental rifts, have bimodal distributions, although the origin for this is still a debate (Kampunzu and Mohr, 1991). The MER delivers opportunity to study in a circumscribed area related to variations in volcanism (mainly composition and volumes) as a function of the amount of extension along the rift (Abebe et al., 2007).
A Quaternary volcanic district known as Butajira Volcanic Field (BVF) is located close to the western escarpment of the central sector of the MER (Fig. 1) along the Silti-Debre Zeyt Fault Zone (SDZFZ) (Megerssa et al., 2019). The BVF is part of a linear chain of basaltic scoria cones and lava flows emitted along the SDZFZ (Rooney, 2010). The area of Butajira offers a distinctive opportunity for study of volcanism in an extensional tectonic regime.
The geochemical signatures of rift magmatism can provide details of the magmatic plumbing system, degree of lithospheric thinning, and changes in properties or extents of the magmatic reservoirs that contribute to rift volcanism (Rooney, 2010). The complex geochemical and isotopic compositions of mafic lavas that erupted along the EARS from the Afar triangle in northern Ethiopia to southern Tanzania, show heterogeneity in both source and process (Furman, 2007). The source of the plentiful mafic rocks has been proposed to be the lithospheric mantle (Rogers et al., 2000) or sub-lithospheric reservoirs, either the shallow asthenosphere (e.g., Chorowicz, 2005; Rooney et al., 2005) or deep-seated mantle plume (e.g., Furman et al., 2004).
In the Butajira-Kibet area, exposed volcanic lithologies include ignimbrite, silicic pyroclastic flow deposits, phreatomagmatic deposit, basaltic lava and scoria. Woldegabriel et al. (1990) has determined the ages of the above lithologies as follow: ignimbrite—early-middle Pliocene (4.2–3 Ma), silicic pyroclastic flow deposits—middle-upper Pliocene and phreatomagmatic, basaltic lava, and scoria- Quaternary (<1.6 Ma).
The aim of this contribution is to investigate the petrogenesis of the Butajira-Kibet Quaternary basaltic rocks based on their petrological and geochemical characteristics. For this purpose, a detailed geological map (Fig. 2), of the study area has also been provided.
Geological map of the study area. Sample locations are shown with yellow dots.
Africa has a unique topographic feature as compared to other continents; mountain chains are dominated by basins and highlands (Rogers, 2006). The East African Rift is a classic example of a young intracontinental rift where continental breakup has not been completed (Trua et al., 1999). From the Red Sea and Gulf of Aden southward to Mozambique, the EARS runs for more than 3000 km, and has been documented as a major extensional feature for over a century (e.g., Furman, 2007; Abebe et al., 2007). According to Rooney et al. (2007) magmatism and rifting in East Africa have long been associated with the effect of one or more mantle plumes and this spatio-temporal relationship between magmatism and tectonism along the EARS provides important input for models of lower mantle structure and the effect of one or more mantle plumes on continued rifting in the region. Volcanism in EARS began during the Eocene (~45 Ma; Ebinger et al., 1993) in southern Ethiopia, resulting in ~1 km thick volcanic pile consisting of basaltic flows and associated rhyolites.
The Ethiopian Rift System, which is active volcanically and seismically, is around 1000 km long extending from the Afar depression southwards to a broad zone of basins and ranges near the Ethiopian border with Kenya (e.g., Ebinger et al., 1993; Abebe et al., 2007). It trends approximately NNE-SSW and forms a major graben from about 6°N in southern Ethiopia to about 9°N where it starts to funnel out into the Afar depression (Feyissa et al., 2017). The MER separates the uplifted western and eastern Ethiopian plateaus and the separation between the two plateaus is 80 ± 10 km wide (Mohr, 1983). The rate and total amount of extension of the MER increase northwards, being largest in the Afar triple junction (Abebe et al., 2007). On average MER has an opening rate of 6–7 mm/yr. (e.g., Bonini et al., 2005). There is a general agreement that MER is less evolved than the Red Sea and Gulf of Aden rifts meeting at the Afar triple junction and is about 18 Ma younger than the other two (Wolfenden et al., 2004). Starting from the early episodes of flood-basalt volcanism magmatic activity in the MER seems to have been episodic rather than continuous (e.g., Woldegabriel et al., 1990; Tadesse et al., 2019). According to Corti (2009), volcanic activity in the MER is characterised by a two-phase evolution that is associated with the two-phase evolution of faulting with the transition from large boundary faults to en-echelon Wonji faults.
During the Mio-Pliocene, associated with the activity of the large boundary fault systems, there was widespread volcanic activity encompassing the whole rift depression. Localisation of volcanic centres is mainly controlled by boundary fault systems and by pre-existing fabrics both parallel and transversal to the rift (Corti, 2009). At the end of the Pliocene there was a substantial change in the distribution of volcanic activity, associated with the activation of the Wonji fault belt. Quaternary volcanism is actually focused within the rift depression and strongly localised along the Wonji faults, giving rise to magmatic segments with only minor activity outside the en-echelon deformation belt (e.g., Corti, 2009; Tadesse et al., 2019).
Geographically, the MER is composed of three main segments; northern, central and southern (e.g., Woldegabriel et al., 1990; Hayward and Ebinger, 1996; Corti, 2009; Tadesse et al., 2019) and the magmatic activity shows different characters in these different rift segments. The three segments reflect different stages of the continental extension process that is interpreted from different fault architecture, timing of volcanism-deformation and crustal and lithospheric structure (e.g., Hayward and Ebinger, 1996; Corti, 2009; Tadesse et al., 2019). The MER is characterized by two different systems of normal faults; the border faults which give rise to major fault-escarpments separating the rift depression from the western and eastern plateaus. Commonly, the normal faults are long (≥50 km), widely spaced and characterised by large vertical offset (typically >500 m) (e.g., Boccaletti et al., 1998; Agostini et al., 2011). A set of faults affecting the rift floor, often referred to as the Wonji Fault Belt (WFB) (e.g., Mohr, 1971) and the Silti-Debre Zeyt Fault Zone (SDZFZ) (Woldegabriel et al., 1990; Rooney et al., 2005) are normally short, closely spaced and display relatively small throws (100 m; e.g., Boccaletti et al., 1998; Corti, 2009). In the three MER segments, these fault systems differ in terms of orientation, structural characteristics (length, vertical throw), timing of activation and relation with magmatism (e.g., Hayward and Ebinger, 1996; Corti, 2009; Tadesse et al., 2019).
Constraints on melting and melt emplacement beneath the MER are provided by seismic and geochemical data. P and S-wave seismic tomography models indicate that the lithosphere-asthenosphere boundary lies at around 70 km subsurface with anomalously low velocity zones in the upper asthenosphere attributed to a combination of higher temperatures and the presence of partial melt (Keir et al., 2006). Data from the seismically and volcanically active MER show that it resembles mid-ocean ridges, yet within a continental setting (Casey et al., 2006). Field studies in the MER and Afar depression have shown that dyking and faulting, which are primarily concentrated in magmatic segments about 20 km and less wide, allow for extension (e.g., Hayward and Ebinger, 1996; Casey et al., 2006). The Ethiopian Afar Geoscintific lithospheric Experiment (EAGLE) project results have shown that Quaternary Butajira-Silti and Bishoftu volcanic chains, on the western side of the central MER, are characterized by thinned crust and Vp/Vs~ > 2.00, indicative of partial melt within the crust (Keir et al., 2006).
A total of 13 thin sections from fresh and representative basaltic rock samples were prepared at the Central Geological Laboratory of the Geological Institute of Ethiopia. Then prepared thin sections have been examined under Leica petrographic microscope in the laboratory of Addis Ababa University at College of Natural and Computational Sciences School of Earth Sciences to investigate the mineralogical compositions and textural variations of the basalts.
Analytical techniquesTen representative basaltic rock samples were analysed for major and trace elements. The sample preparation included removal of the weathered parts from the surface of the rock sample, breaking of the sample into desirable sizes, crushing of the broken fresh samples using a Jaw crusher, milling down to micron size particles using agate ball automatic milling machine. After crushing and milling each rock sample (crushed to 70% less than 2 mm and then pulverized split to 85% <75 μm), the Jaw crusher and ball mills were blown out by an air compressor, washed to remove any possible pollutants and dried before going to the next sample to avoid cross-contamination. The powdered samples were finally sealed and packed and submitted to Australian Laboratory Service PLC Addis Ababa Branch, Ethiopia, and then sent to the analytical laboratory in Vancouver, Canada for analysis.
Major elements were determined by Inductively Coupled Plasma Atomic Emission Spectrometry (ICP-AES) and trace elements including REE, by Inductively Coupled Plasma Mass Spectrometry (ICP-MS). Base metals such as Cr, Co, Ni, and V were analysed using multi-element four acid digestions 81 (ME-4ACD81). Loss on Ignition (LOI) at 1000°C was determined by using WST-SEQ. The precision of analytical instruments has been calibrated with six (6) standards from Mineral Sciences Laboratory in Canada including CDN-W-4 and SY-4, African mineral standards (AMIS), Ore-Research and Exploration Pty. Ltd in Australia and Canada (OREAS). The precision of all measured elements fall under the confidence level; the major elements SiO2, Fe2O3, CaO, Al2O3, K2O and TiO2 have 0.1–2% precision whereas, Na2O and MgO have 2–3%. Reproducibility and accuracy were checked using blank and international rock standards (OREAS 906 and MRGeo08). The precision of most trace elements, including REE and transitional metals, is better than 5%. The major (wt.%) and trace element (ppm) data are presented in Table 1.
| Sample | BJ-2 | BJ-4 | BJ-5 | BJ-7 | BJ-8 | BJ-9 | BJ-11 | BJ-14 | BJ-19 | BJ-20 |
|---|---|---|---|---|---|---|---|---|---|---|
| SiO2 | 47.40 | 47.20 | 47.40 | 47.30 | 48.80 | 47.20 | 46.80 | 46.80 | 46.90 | 46.70 |
| Al2O3 | 15.00 | 15.30 | 15.35 | 16.00 | 15.40 | 15.20 | 15.75 | 15.60 | 15.70 | 16.05 |
| Fe2O3 | 12.10 | 12.65 | 12.70 | 12.75 | 12.10 | 11.90 | 12.65 | 13.05 | 13.05 | 13.20 |
| CaO | 10.00 | 9.75 | 9.62 | 9.55 | 9.62 | 9.49 | 9.54 | 9.80 | 10.00 | 9.43 |
| MgO | 9.00 | 8.07 | 7.85 | 7.55 | 8.29 | 7.78 | 7.71 | 8.52 | 8.82 | 6.85 |
| Na2O | 3.03 | 2.95 | 3.23 | 3.32 | 3.25 | 3.37 | 3.27 | 3.13 | 3.03 | 3.37 |
| K2O | 0.99 | 1.24 | 1.16 | 1.16 | 1.27 | 1.36 | 1.14 | 0.94 | 0.86 | 1.12 |
| TiO2 | 2.45 | 2.92 | 2.81 | 2.69 | 2.41 | 2.39 | 2.67 | 2.72 | 2.67 | 2.92 |
| MnO | 0.17 | 0.18 | 0.17 | 0.18 | 0.18 | 0.18 | 0.18 | 0.18 | 0.18 | 0.18 |
| P2O5 | 0.42 | 0.61 | 0.58 | 0.54 | 0.56 | 0.57 | 0.55 | 0.45 | 0.43 | 0.59 |
| LOI | –0.74 | 0.21 | –0.55 | –0.32 | –0.49 | 1.23 | –0.50 | –0.75 | –0.51 | –0.01 |
| Total | 100.01 | 101.28 | 100.50 | 100.89 | 101.58 | 100.85 | 99.94 | 100.62 | 101.30 | 100.54 |
| Zr | 116.00 | 136.00 | 126.00 | 135.00 | 150.00 | 160.00 | 132.00 | 111.00 | 107.00 | 142.00 |
| Ba | 333.00 | 480.00 | 460.00 | 366.00 | 418.00 | 443.00 | 362.00 | 288.00 | 284.00 | 359.00 |
| Ce | 45.00 | 55.70 | 49.10 | 58.40 | 63.00 | 68.50 | 57.10 | 43.90 | 42.70 | 53.80 |
| Cr | 460.00 | 360.00 | 290.00 | 290.00 | 390.00 | 360.00 | 300.00 | 370.00 | 410.00 | 140.00 |
| Cs | 0.12 | 0.10 | 0.10 | 0.17 | 0.18 | 0.18 | 0.17 | 0.11 | 0.14 | 0.09 |
| Dy | 4.14 | 4.44 | 4.67 | 4.87 | 4.54 | 4.69 | 4.45 | 4.15 | 4.06 | 4.51 |
| Er | 2.16 | 2.20 | 2.15 | 2.52 | 2.31 | 2.30 | 2.41 | 2.22 | 2.24 | 2.26 |
| Eu | 1.82 | 2.17 | 2.22 | 2.16 | 2.21 | 2.31 | 2.22 | 1.84 | 1.77 | 2.20 |
| Ga | 16.00 | 17.00 | 17.40 | 16.80 | 17.80 | 17.00 | 17.30 | 16.20 | 16.20 | 17.80 |
| Gd | 5.02 | 5.63 | 5.45 | 5.66 | 5.73 | 5.92 | 6.11 | 4.95 | 5.17 | 5.42 |
| Hf | 3.20 | 3.60 | 3.50 | 3.50 | 4.00 | 4.40 | 3.50 | 3.10 | 3.30 | 3.80 |
| Ho | 0.79 | 0.84 | 0.87 | 0.91 | 0.91 | 0.87 | 0.92 | 0.82 | 0.87 | 0.85 |
| La | 22.70 | 28.20 | 24.80 | 29.70 | 32.40 | 36.20 | 29.50 | 21.70 | 21.40 | 27.20 |
| Lu | 0.26 | 0.24 | 0.26 | 0.29 | 0.29 | 0.30 | 0.28 | 0.27 | 0.31 | 0.27 |
| Nb | 29.90 | 35.70 | 28.30 | 40.80 | 43.30 | 47.40 | 40.60 | 30.00 | 29.20 | 34.80 |
| Nd | 23.90 | 30.00 | 26.70 | 30.00 | 31.30 | 33.40 | 29.70 | 23.40 | 23.70 | 28.90 |
| Pr | 5.56 | 6.85 | 6.09 | 7.06 | 7.42 | 8.06 | 7.14 | 5.45 | 5.40 | 6.60 |
| Rb | 15.20 | 22.10 | 19.90 | 20.70 | 25.30 | 28.70 | 20.90 | 15.10 | 12.30 | 18.70 |
| Sm | 5.11 | 6.33 | 6.22 | 6.21 | 6.50 | 6.85 | 6.48 | 5.29 | 5.57 | 6.40 |
| Sr | 683.00 | 738.00 | 770.00 | 787.00 | 720.00 | 708.00 | 731.00 | 674.00 | 669.00 | 716.00 |
| Ta | 1.90 | 2.20 | 1.70 | 2.50 | 2.70 | 2.90 | 2.40 | 1.90 | 1.90 | 2.20 |
| Tb | 0.71 | 0.81 | 0.79 | 0.85 | 0.83 | 0.79 | 0.84 | 0.72 | 0.75 | 0.82 |
| Th | 2.15 | 2.61 | 1.93 | 2.78 | 3.42 | 3.74 | 2.71 | 2.01 | 1.96 | 2.47 |
| Tm | 0.28 | 0.30 | 0.31 | 0.34 | 0.32 | 0.33 | 0.32 | 0.30 | 0.35 | 0.30 |
| U | 0.53 | 0.74 | 0.59 | 0.77 | 0.84 | 0.89 | 0.76 | 0.57 | 0.55 | 0.57 |
| V | 311.00 | 312.00 | 329.00 | 287.00 | 297.00 | 289.00 | 288.00 | 301.00 | 300.00 | 326.00 |
| Y | 19.20 | 21.30 | 21.50 | 22.40 | 21.90 | 22.50 | 22.90 | 20.20 | 20.70 | 21.50 |
| Yb | 1.64 | 1.73 | 1.93 | 1.98 | 1.95 | 2.06 | 2.08 | 1.89 | 1.92 | 1.93 |
| Co | 49.00 | 46.00 | 46.00 | 44.00 | 45.00 | 44.00 | 45.00 | 49.00 | 49.00 | 47.00 |
| Cu | 63.00 | 35.00 | 28.00 | 41.00 | 59.00 | 58.00 | 36.00 | 48.00 | 49.00 | 42.00 |
| Ni | 170.00 | 123.00 | 89.00 | 95.00 | 138.00 | 125.00 | 99.00 | 129.00 | 140.00 | 70.00 |
| Pb | 4.00 | 4.00 | 3.00 | 3.00 | 3.00 | 4.00 | 3.00 | 3.00 | 2.00 | 4.00 |
| Sc | 28.00 | 26.00 | 26.00 | 25.00 | 25.00 | 25.00 | 25.00 | 26.00 | 26.00 | 26.00 |
| Zn | 89.00 | 95.00 | 93.00 | 93.00 | 88.00 | 90.00 | 93.00 | 88.00 | 87.00 | 96.00 |
Fe2O3 as total iron
Based on modal mineralogy, all studied Butajira-Kibet Quaternary rocks are basaltic in composition. They are predominantly holocrystalline and porphyritic while some of the samples are vesicular. The fabrics of most samples are inequigranular but a wide variety of crystal textures including cumulophyric, seriate and poikilitic are also observed (Fig. 3). Intergrowths of phenocryst phases (plagioclase, olivine and clinopyroxene and subordinate orthopyroxene) are observed (Fig. 3).
Microphotographs of Butajira-Kibet Quaternary basaltic rocks; (A) sample BJ-1, showing large-sized phenocryst of olivine (B) BJ-4 showing relatively small sized phenocrysts (C) sample BJ-14, showing phenocryst assemblage dominated by Plagioclase (D) sample BJ-5, showing cumulophyric and poikilitic textures (E) sample BJ-16, showing poikilitic texture (clinopyroxene embedded in plagioclase phenocryst), cumulophyric texture is also visible (F) sample BJ-9, exhibiting aphyric texture. All microphotographs are XPL views with the red rectangle is 0.07. Cpx-Clinopyroxene, Opx-Orthopyroxene, Oli-Olivine, Plag-plagioclase, Por-pore space, Opa-Opaque oxide.
Porphyritic basalts are predominantly composed of phenocrysts of plagioclase (4–20%) and clinopyroxene (4–15%) with subordinate olivine (2–8%) and orthopyroxene (2–5%). These phenocrysts are set in a fine-grained matrix reaching 54% for porphyritic and up to 84% for aphyric rocks. The mineral phases in the matrix include the above phenocryst minerals, and opaque oxides. The size of phenocryst grains in each sample is almost similar, ranging between 1 and 6 mm, except one sample which has a plagioclase grain measuring a centimetre. The phenocrysts are set in fine-grained holocrystalline matrix consisting of the above phenocryst minerals and opaque oxides.
Phenocrysts of plagioclase, pyroxene and olivine are typically euhedral to subhedral. The presence of euhedral olivine in the samples indicates a primitive parental mafic liquid that has undergone minor olivine fractionation (Rooney et al., 2005). Sometimes, olivine and clinopyroxene phenocrysts have similar sizes and according to Cook et al. (2005), this suggests possible derivation of these minerals from xenolith disaggregation.
Major element geochemistryBefore describing major and trace element data of Butajira-Kibet Quaternary basaltic rocks, it is important to consider the degree of alteration of the analysed samples, if any. Most samples have negative LOI values while two samples (BJ-4 and BJ-9) have positive values of 0.21 and 1.23, respectively. This suggests that the analysed samples are very fresh and least affected by alteration.
In the Total Alkali Silica diagram (TAS) (Irvine and Baragar, 1971; Le Bas et al., 1986, Fig. 4), the analysed samples all plot as basalts with alkaline affinity. These basalts are similar to those of Debre Zeyt, but differ from the MER axial basaltic rocks which arrange from alkaline to sub-alkaline (Fig. 4). Debre Zeyt and the MER axial basaltic rocks range from basalt to trachytic-basalt (Fig. 4).
Total Alkali Silica (TAS) diagram after Le Bas et al. (1986) for Butajira-Kibet Quaternary basaltic rocks. The alkaline-sub-alkaline dividing line is from Irvine and Baragar (1971). MER and Debre Zeyt data are from Ayalew et al. (2016) and Rooney et al. (2005), respectively.
Overall, Butajira-Kibet Quaternary basalts have relatively higher concentrations of MgO and A2O3 and lower contents of TiO2, Fe2O3 and CaO compared with rift axis basaltic rocks. In Fig. 5, selected major element oxides and ratios are plotted against MgO to assess their correlations and infer petrogenetic processes. From the figure, CaO and CaO/Al2O3 exhibit clear positive correlations with MgO, as shown by the shaded region in Fig. 5b and e, respectively, while Na2O shows a well-defined negative trend (Fig. 5f). Fe2O3tot and Al2O3 do not show clear trends until the concentration of MgO decreases to 7.80 wt.% and then their concentrations increase against decreasing MgO; this could be related to partial melting at different pressures rather than fractional crystallization of a parental basaltic magma. On the other hand, the concentrations of K2O and P2O5 increase with decreasing MgO up to 8.52 wt.%; below this value, their relationships with MgO are not well defined (Fig. 5g and h). TiO2 does not have a clear correlation with MgO, though it broadly increases with decreasing MgO. Collectively, the trends of major element variations of Butajira-Kibet Quaternary basalts do not indicate clear characteristics on their evolution process; this means it is not possible to determine if the basaltic rocks are evolved products from primary magma.
Major element variation diagrams plotted against MgO for Butajira-Kibet Quaternary basalts. MER and Debre Zeyt data are from Ayalew et al. (2016) and Rooney et al. (2005), respectively.
Concentrations of selected trace elements as a function of MgO for Butajira-Kibet Quaternary basalts are illustrated in Fig. 6. Compatible trace elements, such as Ni Cr and Co display well-defined positive correlations with increasing MgO. Co shows variable trends (Co is high at 6.85 wt.% MgO) compared to Ni and Cr (Fig. 6a, b and j). High-field strength elements (HFSE), such as Nb, Th and Zr as well as large ion lithophile elements (LIL), such as Rb and Ba, do not form coherent trends with decreasing MgO (Fig. 6c, d and g), while Sr displays a regular increasing (Fig. 6e, f and h). Although these elements do not define well-defined correlations, they exhibit shallow negative slopes against increasing MgO. The moderately compatible trace element (Sc) displays well-defined positive correlation with increasing MgO (Fig. 6i). Butajira-Kibet Quaternary basalts have higher concentrations of Ni, Cr, Nb and Sr and lower concentration of Zr and Sc compared with rift axis basaltic rocks. The above trace element characteristics are more or less comparable with Debre Zeyt Quaternary basaltic rocks.
Selected trace elements variation diagrams plotted against MgO for Butajira-Kibet Quaternary basaltic rocks. MER and Debre Zeyt data are from Ayalew et al. (2016) and Rooney et al. (2005), respectively.
Primitive mantle-normalized multi-element variations for Butajira-Kibet Quaternary basalts (Fig. 7) generally show parallel to sub parallel pattern. The basalts exhibit strong enrichment of highly to moderately incompatible trace elements. All the basaltic samples are enriched in Ba, Nb, Ta, La and Sr. Conversely, the basaltic samples depict Th, U, Rb and K depletions, which is characteristic of most of the Ethiopian basalt province (e.g., Ayalew et al., 2016; 2018). It is to be noted that rift aixs basaltic rocks are highly depleted in K, Th, U, Hf and Pb compared to Butajira-Kibet Quaternary basaltic rocks. On the same variation diagram (Fig. 7), Butajira-Kibet Quaternary basalts show enrichment in Sr and depletion in P compared to rift axis basalts. These basaltic rocks have similar trace element characteristics with Debre Zeyt basaltic rocks (Fig. 7). Overall, Butajira-Kibet Quaternary basaltic rock samples show coherent behaviour throughout the multi-element variation diagram (Fig. 7), having positive Nb-Ta anomalies and higher abundances of LREE. These characteristics are typical of OIB mantle source and indicative of generation of parent magma from a mantle source that has experienced sufficient enrichment in these elements.
Primitive mantle-normalized multi-element variation diagram for Butajira-Kibet Quaternary basalts; normalization values from Sun and McDonough (1989); MER and Debre Zeyt data are from Ayalew et al. (2016) and Rooney et al. (2005), respectively.
In the Chondrite-normalized Rare Earth Element (REE) variation diagram (Fig. 8) Butajira-Kibet Quaternary basaltic rocks are characterized by parallel to sub-parallel patterns. These patterns show Light-Rare Earth Element (LREE) enrichments and properly flat Heavy Rare Earth Element (HREE) distributions (TbN/YbN = 1.73–2.13) which implies that HREE are not fractionated, with HREE abundances greater than ten times (>10.00) chondritic values. There is no observed negative Eu anomaly with increasing total REE contents, although these basaltic rocks possess weak positive Eu anomalies (Eu/Eu* = 1.01–1.17). Note that rift axis basaltic rocks show clear positive Eu-anomaly. Ce does not display any negative anomaly (Fig. 8) in all samples which might be due to insignificant or no crustal contamination of mantle-derived magma.
Chondrite-normalized rare earth element pattern for representative Butajira-Kibet Quaternary basalts; normalization values are from Sun and McDonough (1989); MER and Debre Zeyt data are from Ayalew et al. (2016) and Rooney et al. (2005), respectively.
According to Ayalew et al. (2016) distinct similarity of trace element patterns for samples from different suites or within a similar suite of volcanic centres suggests a fairly homogeneous mantle source and/or melting condition over the entire region. By inference, Butajira-Kibet Quaternary basalts might have a fairly homogenous mantle source and/or melting conditions over that region (Fig. 8).
All analysed Butajira-Kibet Quaternary basalts have moderate to high concentrations of MgO (6.85–9.00 wt.%) with Mg# varying between 53.00 and 62.06. In addition, compatible trace element concentrations, such as Ni: 70.00–170.00 ppm and Cr: 140.00–460.00 ppm, are lower than primary basaltic magma (Ni: 400.00–500.00 ppm, Cr: >1000.00 ppm, MgO: 10.00–15.00 wt.% and Mg# >68.00; e.g., Frey et al., 1978; Hart and Davis, 1978; Ayalew et al., 2018) in equilibrium with a typical upper mantle mineral assemblage. Consequently, these variations with primary magma suggest that, Butajira-Kibet Quaternary basaltic rocks have undergone at least some olivine differentiation that may be taking place in a shallow crustal reservoir and during magma ascent or surface flow.
The plots of CaO and CaO/Al2O3 versus MgO (Fig. 5e and b) show decreasing CaO and CaO/Al2O3 with decreasing MgO; these variations likely indicate the removal of olivine and clinopyroxene from the parental basaltic magma. Na behaves as moderately incompatible with respect to clinopyroxene and the CaO/Al2O3 ratio in clinopyroxene is high (Dungan and Rhodes, 1978). Thus clinopyroxene fractionation would be expected to produce an inverse correlation between CaO/Al2O3 and Na2O and such trend is observed (Fig. 5b and f). On the other hand, Dungan and Rhodes (1978) suggested that clinopyroxene has less Fe2O3t than the magma from which it crystallizes; accordingly clinopyroxene crystallization would lead to progressively higher Fe2O3t contents with decreasing CaO/Al2O3. Consequently lack of clear correlation trend of Fe2O3total with decreasing CaO/Al2O3 together with petrographic observation (Fig. 3) suggests that the Butajira-Kibet Quaternary basalt suite did not undergone significant clinopyroxene fractionation. TiO2 shallowly increases as MgO decreases, indicating no fractionation of titaniferous clinopyroxene. Al2O3 shows somewhat negative relationship with increasing MgO (Fig. 5a) in particular, at low concentrations of MgO; the overall concentration of Al2O3 is also very high, likely indicating absence of plagioclase fractionations in the Butajira-Kibet Quaternary basalts. Furthermore, Sr shows an increasing trend with decreasing MgO and Sr enrichment in multi-element variation diagram (Fig. 7). These features together with Eu having shallow positive anomaly (Eu/Eu* = 1.01–1.17), indicate that there is no significant plagioclase fractionation in the basalts. Moreover, petrographic observations support non-fractionation of plagioclase; where plagioclase phenocrysts dominate almost all basalt samples (Fig. 3).
The plot of (La/Sm)N against Zr/Y shows positive correlation (Fig. 9a) within Butajira-Kibet Quaternary basaltic samples. This relationship is expected during progressive melting of a clinopyroxene-rich source, not during fractionation of clinopyroxene (Furman et al., 2004). Such a trend arises from variations in to the relative compatibilities of these elements (Y > Zr > Nb and Sm > La) in clinopyroxene (e.g., Hart and Dnnn, 1993; Furman et al., 2004). Butajira-Kibet Quaternary basalts are more primitive compared with rift axis basalts have possibly undergone significant fractionation of olivine and clinopyroxene. The two suites also plot in different regions in the (La/Sm)N versus Zr/Y diagram (Fig. 9a). For the Butajira-Kibet Quaternary basalts, (La/Sm)N increases as Zr/Y increases as increases for the Butajira-Kibet Quaternary basalts, while for the rift axis basalts the (La/Sm)N is almost constant as Zr/Y increases.
(a) Primitive mantle-normalized (La/Sm)N versus Zr/Y; (b) ∑REE versus MgO and (c) Chondrite-normalized (La/Yb)N versus LaN (ppm), normalization values are from Sun and McDonough (1989) ; MER data are from Ayalew et al. (2016).
The ratios of REE are further used to constrain the petrogenetic processes involved in the evolution of the analysed rock samples. It is established that the sum of rare earth elements (∑REE) will not show large variations from sample to sample (within suite) when the rocks are formed by fractional crystallization process. It is also known that, during the process of fractional crystallization, incompatible trace element abundances increase while their ratios remain constant or their increments are indistinguishable. On the other hand, during partial melting, both the incompatible trace element ratios and their abundances will increase. Based on these relationships, the La versus (La/Yb)N diagram for Butajira-Kibet Quaternary basaltic rocks (Fig. 9c) suggests that the basaltic rocks are not related to each other by fractional crystallization; instead, they are consistent with their derivation by partial melting of the mantle. As discussed above, rift axis and Debre Zeyt basalts have both similar characters that differ from those of Butajira-Kibet. This petrogenetic difference is probably related to variations in crustal/lithospheric thickness; as we go away from the main rift axis towards the rift margin, the role of fractional crystallization for Quaternary basaltic rocks decrease significantly. It is, therefore, concluded that the geochemical data presented here do not favour fractional crystallization as a significant mechanism for the evolution of Butajira-Kibet Quaternary basalts. The reverse is possibly true for the role of partial melting, which is discussed in a separate section below.
Crustal contaminationButajira-Kibet Quaternary basaltic samples are characterized by enrichment in highly and moderately incompatible elements as shown in Fig. 7, which might be either a consequence of the mantle-derived magma affected by crustal materials or derivation from enriched mantle sources. It is known that LILE/LREE, LILE/HFSE, LREE/HFSE and ratios of trace elements with similar degrees of incompatibility are effective means to evaluate the potential contamination of magmas by crustal materials. These trace element ratios are used to characterize the effect of crustal contamination on the Butajira-Kibet Quaternary basaltic rocks (Table 2). Consideration is made to the accepted fact that the continental crust and crustally derived rocks are enriched in LILE relative to LREE, and depleted in HFSE relative to LREE (Table 2).
| Trace element ratio | Standards | Butajira-Kibet basalt | Debre Zeyt basalt | MER axis basalt | |
|---|---|---|---|---|---|
| Continental crust average | Mantle-derived (OIB) | ||||
| Ba/La | ~25.00 | 6.20–16.90 | 12.00–14.67 | 10.89–17.00 | 10.88–16.49 |
| Ba/Nb | ~54.00 | 4.70–17.80 | 8.92–16.25 | 8.93–14.96 | 8.88–16.51 |
| Rb/Nb | ~4.70 | 0.30–1.23 | 0.42–0.70 | 0.46–0.78 | 0.35–0.71 |
| K/Nb | ~1341.00 | 66.00–432.00 | 233.00–340.00 | 176.00–441.00 | 188.00–303.00 |
| Ce/Pb | <5.00 | >5.00 | 11.25–21.35 | 17.80–27.68 | 13.26–34.30 |
| Nb/U | ~25.00 | 47.00 ± 10.00 | 47.97–61.05 | 47.50–81.16 | 37.76–67.14 |
| La/Nb | ~2.20 | 1.28–0.64 | 0.73–0.88 | 0.69–0.88 | 0.74–1.13 |
Continental crust average and mantle-derived basalt (OIB); Nb/U after Hofmann et al. (1986); Ce/Pb after Rudnick and Fountain (1995); Ba/La, Ba/Nb, Rb/Nb, K/Nb and La/Nb after Weaver (1991); MER axis basalts after Ayalew et al. (2016) and Debre Zeyt basalts after Rooney et al. (2005)
As shown in Table 2, Butajira-Kibet Quaternary basalts have Ba/La, Ba/Nb, Rb/Nb and K/Nb values of 12.00–14.67, 8.92–16.25, 0.42–0.70 and 233.00–340.00 respectively which are below the values for continental crust Weaver (1991). On the other hand Ba/La, Ba/Nb, Rb/Nb and K/Nb values for mantle-derived oceanic basalts (OIB) are 6.20–16.9.00, 4.70–17.80, 0.30–1.23, 66.00–432.00; Weaver (1991) respectively. These ratios suggest that Butajira-Kibet Quaternary basalts are not contaminated by crustal materials; rather they have similarity with OIB. The Nb/U ratios (47.97–61.05) for Butajira-Kibet Quaternary basalt rocks are similar to those of OIB, which indicates that crustal contamination did not play a significant effect in the evolution of studied rocks. In addition to the above ratios; Butajira-Kibet Quaternary basalts have La/Nb values of 0.72–0.88 much lower than average crustal ratios of 2.20 (Weaver, 1991). Crust contaminated basalts have La/Nb ratios around 0.90 and greater (Pik et al., 1998). The above geochemical data hence supports the suggestion that Butajira-Kibet Quaternary basalts are not contaminated by crust materials. Overall, incompatible trace element ratios support the suggestion that crustal contamination has been insignificant in the genesis of Butajira-Kibet Quaternary basaltic rocks (Fig. 10). According to Ayalew et al. (2016), most basalts of the Ethiopian rift are not contaminated by crustal materials to a significant degree.
Nb versus Nb/U plot after Hofmann et al. (1986); the data for MORB and OIB from Hofmann et al. (1986) and for continental crust from Rudnick and Fountain (1995).
The geochemical composition of basalt is significantly impacted by the partial melting of mantle peridotite that produces basaltic magma. The Butajira-Kibet Quaternary basalts are characterised by low CaO/Al2O3 ratios (0.59–0.67) and relatively flat HREE patterns (TbN/YbN = 1.73–2.13), with somewhat elevated HREE concentrations higher than 10-times chondritic values (Fig. 8). According to Ayalew et al. (2016), this is most likely indicator of a mantle source containing spinel rather than garnet.
Experimental studies show that Na2O is perfectly incompatible with respect to olivine, strongly incompatible with respect to orthopyroxene, and at pressures above plagioclase stability (8–9 kbar) the mantle budget of Na2O mainly exists in clinopyroxene (Klein and Langmuir, 1987). Accordingly, Na2O behaves as a moderately incompatible trace element, showing highest concentrations at the smallest extents of melting and decreasing concentration as the extent of melting increases (Jaques and Green, 1980). Thus, if we consider the relative concentration of Na2O at the source is constant at the source; variations in the abundances of Na2O may indicate variations in the extents of melting. CaO and Al2O3 also appear to show systematic behaviour during melting of spinel lherzolite (e.g., Jaques and Green, 1980; Klein and Langmuir, 1987). According to Jaques and Green (1980), Al2O3 abundances are highest in the first increments of melt and decrease with further extents of melting of spinel lherzolite. CaO appears to show somewhat more complicated melting behaviour. CaO increases as clinopyroxene melts, but when clinopyroxene is no longer present as a residual phase, CaO decreases with further melting. Thus CaO/Al2O3 ratio shows the following trends: CaO/Al2O3 is at a minimum as melting begins; it increases then until clinopyroxene has melted out, at which point it has a higher value than the initial source because some Al2O3 remains in orthopyroxene; as melting proceeds and Al2O3 from orthopyroxene is added to the melt, the CaO/Al2O3 ratio decreases slightly. When clinopyroxene is present as a residual phase, increasing the extent of melting causes an increase in the CaO/Al2O3 ratio.
Butajira-Kibet Quaternary basalts have a good characterizing link between Na2O and CaO/Al2O3 ratios; when the concentration of Na2O decreases, CaO/Al2O3 increases (Fig. 11a). This relationship supports the idea of partial melting of clinopyroxene, which is still present as a residual phase and with extent of melting increasing as pressure decreases.
(a) Na2O versus CaO/Al2O3 and (b) SiO2 versus Nb/Y after Greenough et al. (2005); MER and Debre Zeyt data from Ayalew et al. (2016) and Rooney et al. (2005), respectively.
Like Na2O, Al2O3 and CaO, the systematics of SiO2 variations during melting depend on pressure as well as the extent of melting. Experimental studies show that at any one pressure, SiO2 increases with increasing extents of melting, to approximately 40% melting (Jaques and Green, 1980). Thus, Butajira-Kibet Quaternary basalts are most probably formed by moderate extent of melting at a given pressure as shown in Fig. 11b.
The ratios LREE/HREE and MREE/HREE are sensitive to the extent of melting and the amount of residual garnet in the source. Garnet strongly retains in the HREE and consequently, the existence of garnet in the source is the only phase that significantly fractionates the LREE/HREE. To assess melting conditions, non-modal batch melting of garnet-bearing and spinel-bearing peridotite mantle sources using primitive mantle (Sun and McDonough, 1989) as a staring source, has been modelled. On the basis of (La/Sm)N vs. (Sm/Yb)N (Fig. 12), the Butajira-Kibet Quaternary basaltic magmas could be produced by equilibrium melting with 1.5–3% degree partial melting within spinel peridotite mantle sources. Therefore, based on the model calculation; the ranges in LREE/MREE and MREE/HREE values within the Butajira-Kibet Quaternary basaltic rocks may be explained by variable degrees of melting.
(La/Sm)N versus (Sm/Yb)N variation diagram showing non-modal batch melting modelling of a peridotite mantle source for Butajira-Kibet Quaternary basalts. Garnet-bearing mantle source with estimated 60% olivine, 21% orthopyroxene, 8% clinopyroxene, 11% garnet and spinel-bearing lherzolite source with 58% olivine, 27% orthopyroxene, 12% clinopyroxene, 3% spinel. Garnet and spinel peridotite compositions used for modelling are modified from Ayalew et al. (2018) and Peters et al. (2008). Melting curves use a primitive mantle composition with the normalization values are taken from Sun and McDonough (1989) and partition coefficients from McKenzie and O’Nions (1991).
Estimation of the temperatures and pressures of melting have been made by considering the compositions of basalts and relating them to the experimental temperatures and pressures of magma generation (Lee et al., 2009). For the Butajira-Kibet Quaternary basalts, this calibration produced potential temperatures between 1168.05 and 1197.69°C and pressures between 1.22 and 1.36 GPa (Fig. 13 and Table 3). The derived temperatures and pressures of the magmas plot within the stability field of spinel peridotite, which supports the non-modal batch melting model (Fig. 12). These pressure and temperature estimates are comparable to those determined for Quaternary Debre Zeyt basalts (the equilibration pressures at 35–55 km and 1120–1361°C, respectively) (Rooney et al., 2005) and for Quaternary rift basalts (1.01–1.24 GPa. and 1125–1200°C, respectively) (Ayalew et al., 2016). As shown Fig. 11b, Butajira-Kibet Quaternary basalts and Debre Zeyt basalts are formed almost at similar degrees of partial meting and similar pressure conditions. Rift axis basalts, on the other hand, are consistent with their formation by relatively higher extent of partial melting and lower pressure conditions.
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). Note that these temperature and pressure estimates must be viewed as minimum estimates.
| Samples | BJ-2 | BJ-4 | BJ-5 | BJ-7 | BJ-8 | BJ-9 | BJ-11 | BJ-14 | BJ-19 | BJ-20 |
|---|---|---|---|---|---|---|---|---|---|---|
| Temperature (°C) | 1183.20 | 1179.79 | 1176.07 | 1180.89 | 1176.89 | 1180.31 | 1193.60 | 1197.69 | 1168.05 | 1196.41 |
| Pressure (GPa) | 1.30 | 1.28 | 1.31 | 1.22 | 1.33 | 1.33 | 1.36 | 1.36 | 1.30 | 1.29 |
The effect of hydrous metasomatism, such as phlogopite and amphibole and carbonatite on the Butajira-Kibet Quaternary basaltic rocks have been evaluated using major element concentrations and their ratios, as well as incompatible trace element concentrations and their ratios. According to a study by Dasgupta et al. (2007), partial melting of volatile-rich (carbonated) peridotite can produce alkaline melts with—the following major element characteristics: 43–45 wt.% SiO2, 9–12 wt.% Al2O3, 12–16 wt.% CaO, 1–2 wt.% TiO2, 10–11 wt.% FeO, and 2–3 wt.% Na2O. By comparison, Butajira-Kibet Quaternary basalts have quite different major element concentrations implying absence of carbonatite metasomatism. Furthermore, carbonatite-metasomatized mantle is thought to be characterized by very high LREE enrichments and very high HREE depletion (e.g., La/Yb)N ranging between 30.00 and 90.00; Powell et al., 2004), Ti/Eu (<3000; McDonough, 1990; Powell et al., 2004) and Zr/Hf between 45.00–100.00; Rudnick, 1995). Butajira-Kibet Quaternary basalts have Zr/Hf: 32.42–38.57, Ti/Eu: 6226.00–9043.00 and (La/Yb)N 8.24–12.6, comparable with primitive mantle values (Sun and McDonough, 1989; McDonough, 1990). Accordingly, the effect of carbonatitic or hydrous metasomatism on Butajira-Kibet Quaternary basalts is insignificant. Furthermore, considering that melting of a mantle source containing phlogopite will result in potassic and ultrapotassic magmas (e.g., Rosenthal et al., 2009; Ayalew et al., 2016; Foley et al., 2022), the concentrations of K2O (0.86–1.37), Na2O (2.95–3.37) and K2O/Na2O (0.28–0.42) in Butajira-Kibet basalts preclude the presence of phlogopite in the mantle source.
Primitive mantle-normalized incompatible trace elements patterns of anhydrous cumulates are characterized by depletion in highly incompatible elements such as Rb, U, and Th and are relatively flat in the region of the HREE (Pilet et al., 2011). On the other hand, peridotite metasomatized with silicate melts, such as amphibole, shows trace element enrichments in U and Th, negative anomalies for Pb, Sr, Zr, Hf, Nb and Ta (Powell et al., 2004; Kaeser et al., 2007; Pilet et al., 2011). In light of these experimental findings, Butajira-Kibet Quaternary basalts do not require the presence of hydrous phase, as shown in the primitive mantle-normalized multi-element plot (Fig. 7). In contrast, the basaltic rocks have Rb/Sr (0.02–0.04), Ba/Rb (15.44–23.12), La/Nb (0.72–0.88) and incompatible trace element ratios comparable with values for primitive mantle (e.g., Sun and McDonough, 1989).
According to McDonough (1990) the average Ti/Eu ratio is lower in hydrous (amphibole- and/or mica-bearing) peridotites than in anhydrous peridotites (7500.00 ± 3600.00). The study of McDonough (1990) also proposed that hydrous peridotites with low Ti/Eu values (<3000.00) were produced by metasomatic processes capable of fractionating Ti from Eu. Butajira-Kibet Quaternary basalts have high Ti/Eu ratios as shown above, with average value of 7694.00. This indicates that there is no indication of hydrous phase melting and metasomatism is insignificant in the generation of Butajira-Kibet Quaternary basaltic lavas.
Mantle source characteristicsThe variations in trace element composition of almost all Quaternary basaltic samples of Butajira-Kibet indicate no contamination by crustal materials as discussed in crustal contamination part. Low concentrations of K, coupled with high concentrations of Nb and La in some highly undersaturated alkali basalts suggests that K in the melts could be buffered by residual amphibole, phlogopite and clinopyroxene (under very high pressure) during source enrichment processes and/or magma generation (Clague and Frey, 1982; Sun and McDonough, 1989). Such residual potassium-bearing minerals could also cause depletions of Rb, Cs, and Ba in these magmas relative to their adjacent elements on a primitive mantle-normalized multi-element variation diagram (Sun and McDonough, 1989). However, the studied basalts are enriched in Ba and their incompatible trace element patterns are not consistent with residual amphibole and phlogopite melting, as discussed above. Therefore, the source of the basalts is most likely an enriched mantle containing residual clinopyroxene.
Butajira-Kibet Quaternary basalts have low La/Nb (0.72–0.88) relative to chondrite (<0.95). According to Sun and McDonough (1989), this indicative of a mantle source that has undergone small degrees of partial melting and no involvement of hydrous minerals (metasomatism). Furthermore, the basalts exhibit significant enrichment in HFSE (Nb and Ta) relative to the LILE and LREE (La and Ce). It is known that in mantle peridotite, Nb is slightly more incompatible than Ta during melting (Schmidt et al., 2004; Pfänder et al., 2007), resulting in slightly elevated Nb/Ta ratios in primitive mantle melts and island arc rocks (Münker et al., 2003; Pfänder et al., 2007). Low Nb/Ta, however, can be explained by melting of amphiboles as suggested by (Foley et al., 2002; Pfänder et al., 2007). According to Münker et al. (2003), the Nb/Ta ratios for different basalts are: MORB, 11.5–16.6; OIB, 14.6–16.7 and continental basalts, 16.4–18.8. The Butajira-Kibet Quaternary basalts have Nb/Ta ratios in the range 15.37–16.92, which is similar with OIB indicating that primitive mantle melts are free of amphiboles. In the Ba/La versus Ba/Nb diagram (Fig. 14), the Butajira-Kibet Quaternary basalts plot close to or within EM-I OIB. According to Weaver (1991), the EM-I OIB end member is different from the other OIB end member types by its enrichment in Ba relative to other LILE, leading to high Ba/La ratios (13.20–16.90). Enhancement of Ba/Nb ratios is elevated by the absence of Nb enrichment, as shown in Fig. 14 together with the characteristic positive Ba anomaly in the primitive mantle-normalized multi-element diagram (Fig. 7). It has been shown that these basalts are also characterized by higher HFSE/LREE and HFSE/LILE, which is a notable feature in OIB-like patterns (Weaver, 1991)
LILE/HFSE versus LILE/LREE variations for Butajira-Kibet Quaternary basalts; Ba/Nb versus Ba/La, N-MORB, EM-I, EM-II, HIMU and Primordial Mantle data are from Weaver (1991); data for Debre Zeyt basalts are from Rooney et al. (2005).
The Zr/Nb ratios of the investigated samples range from 3.31 to 4.45, suggesting plume type basalts having lower Zr/Nb ratios, in comparison with N-MORB values (Zr/Nb >30). Zr/Ba ratio can be considered as an effective proxy to discriminate between sub-continental lithospheric mantle (SCLM) (Zr/Ba = 0.3–0.5) and asthenospheric sources (Zr/Ba >0.5) (Greenough et al., 1998; Ganguly et al., 2014; Singh et al., 2018). Accordingly, Butajira-Kibet Quaternary basalts, with Zr/Ba ratios of 0.27–0.40 are indicative of considerable contribution from SCLM.
Likewise, Rb/Zr versus Nb/Zr diagram (Fig. 15) is used to investigate the genesis of Butajira-Kibet Quaternary basalts. From the plot at least two mantle components can be distinguished; an OIB-like component similar to plume composition (Sun and McDonough, 1989) and a second component similar to an enriched SCLM (McDonough, 1990). The LILE-HFSE-REE systematics for the Butajira-Kibet Quaternary basalts, therefore, indicate that the parent magmas originated from a heterogeneous mantle source with signatures of (i) from an upwelling plume head, and (ii) from SCLM, with greater contribution from the mantle plume.
LILE/HFSE versus HFSE/HFSE variation for Butajira-Kibet basalts; Rb/Zr versus Nb/Zr, the N-MORB, OIB, E-MORB, and Primitive mantle data are from Sun and McDonough (1989); UC (upper continental crust) and LC (lower continental crust) data are from Taylor and McLennan (1981).
Based on the interpretation of major and trace element data the following conclusions have been reached on the petrogenesis of Butajira-Kibet Quaternary basalts.
a. The basalts are mildly alkaline basalts characterized by moderate to high contents of MgO (6.85–9.00 wt.%), TiO2 (2.40–2.92 wt.%), depletion in Th, U, Rb and K and enrichment in Ba, Nb, Ta, La, Pb and Sr with relatively positive Eu/Eu*(1.01–1.17).
b. Incompatible trace element ratios such as Ba/Nb, Rb/Nb, K/Nb, La/Nb and Nb/U suggest that the basalts are unaffected by crustal contamination.
c. The geochemical data are not indicative of significant fractionation of plagioclase and clinopyroxene; only minor fractionation of olivine has taken place. The basalt is most likely originated by partial melting of clinopyroxene-rich source at medium pressure.
d. Geochemical modelling has revealed that the basalts predominantly formed by 1.5–3% partial melting of a spinel peridotite mantle source with minor contribution amount from garnet peridotite.
e. Interpretation of the geochemical data further suggests that the basaltic magmas are not associated with carbonatite and/or hydrous (amphibole or phlogopite) phase metasomatism.
f. The source regions for Butajira-Kibet basalt magmas are inferred to be a mantle plume with OIB nature (EM-I end member of OIB) and SCLM, to a minor extent. A possible scenario is that an upwelling hot mantle plume encounters and initiates melting of the geochemically enriched mantle component (EM-I) in the asthenosphere. Subsequently, the base of sub-continental lithospheric mantle (SCLM) undergoes partial melting as a result of anomalous heat transferred upwards from the hot mantle plume. Consequently, melts from the SCLM interact with melts from the plume producing magmas that feed the Quaternary Butajira-Kibet basalts.
Our gratitude goes to School of Earth sciences, Addis Ababa University and the Department of Geology, Wachemo University for their financial and logistical support. We are thankful to the Central Laboratory of the Geological Institute of Ethiopia for thin section preparations and to the Australian Laboratory Services for undertaking the whole-rock major and trace element analysis.
