Mafic xenoliths from Egyptian Tertiary basalts and their petrogenetic implications

Mafic xenoliths from Egyptian Tertiary basalts and their petrogenetic implications

Gondwana Research 11 (2007) 516 – 528 www.elsevier.com/locate/gr Mafic xenoliths from Egyptian Tertiary basalts and their petrogenetic implications E...

2MB Sizes 1 Downloads 19 Views

Gondwana Research 11 (2007) 516 – 528 www.elsevier.com/locate/gr

Mafic xenoliths from Egyptian Tertiary basalts and their petrogenetic implications E.S. Farahat a,⁎, M.M. Shaaban b , A.Y. Abdel Aal a a

Geology Department, Faculty of Science, Minia University, El Minia 61519, Egypt b Geology Department, Faculty of Science, Ain Shams University, Cairo, Egypt

Received 6 April 2006; received in revised form 15 July 2006; accepted 18 July 2006 Available online 7 September 2006

Abstract Mineralogical data, coupled with whole-rock major and trace element data of mafic xenoliths from two occurrences of the Egyptian Tertiary basalts, namely Abu Zaabal (AZ) near Cairo and Gabal Mandisha (GM) in the Bahariya Oases, are presented for the first time. Chemically, AZ basalts are sodic transitional, while those of GM are alkaline. In spite of the different petrographic and geochemical features of the host rocks, mafic xenoliths from the two occurrences are broadly similar and composed essentially of clinopyroxene, plagioclase, alkali feldspar, and Fe–Ti oxides. The analytical results of host rocks, xenoliths and their minerals suggest that the xenoliths are cognate to their host magmas rather than basement material. The mafic xenoliths are olivine-free and contain alkali feldspar contrary to the phenocryst assemblage of the host rocks, confirming that they are not cumulates from the host magma. The geochemical and mineralogical characteristics show that the precursor magmas of these xenoliths are more fractionated and possibly contaminated compared to those of the host rocks. Estimated crystallization conditions are ∼ 1–3 kbar for xenoliths from both areas, and temperature of ∼ 950–1100 °C vs. 920–1050 °C for AZ and GM, respectively. These cognate xenoliths probably crystallized from early-formed, highly-fractionated anhydrous magma batches solidified in shallow crustal levels, possibly underwent some AFC during their ascent, and later ripped-up during fresh magma pulses. The xenoliths, although rare, provide an evidence for the importance of crystal fractionation at early evolution of the Egyptian Tertiary basalts. © 2006 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved. Keywords: Tertiary basalts; Egypt; Xenoliths; Fractional crystallization; Petrogenesis

1. Introduction Crustal-derived ultramafic and mafic xenoliths are widely reported from island-arc calc-alkaline suites in addition to continental and oceanic alkaline rocks (e.g. Binns et al., 1970; Munha et al., 1990; Turner et al., 2003). However, they are reported less frequently from continental tholeiitic suites (e.g. Donaldson, 1977; Preston and Bell, 1997). Some continental volcanic fields display compositional variations within single eruptions, attesting the storage and fractionation of early magma batches within the crust which are later transected and ripped-up by magma which erupts at the surface (Németh et al., 2003). Direct information about magma transport, stalling and fractionation in the crust can be obtained from entrained blocks ⁎ Corresponding author. Tel.: +20 86 2378500; fax:+20 86 2363011. E-mail address: [email protected] (E.S. Farahat).

of cognate materials, or nodules (e.g., Tait, 1988; De Silva, 1989; Mattioli et al., 2003). These cognate nodules are commonly interpreted to represent solid fractionates from the host magma (Tait, 1988; Wagner et al., 2003; Holness et al., 2005). However, the mode of occurrence of such xenoliths, frequently at the base of the lava sheets that let them be hidden from direct observations, in addition to possible textural modifications during magma ascent hinders the use of such cognate materials to access directly the solidification history of the host magma. Several basaltic occurrences are sparsely scattered over the whole area of Egypt. Previous studies show the wide distribution of volcanic eruptions in Egypt during the Phanerozoic mainly as a result of periodic reactivation of the fracture system which originated since the late Precambrian. In the light of the available age data, ten episodes were proposed ranging from Mesozoic to Quaternary (Meneisy and Abdel Aal, 1983; Meneisy, 1990). The majority of these episodes occurred in Tertiary times. This

1342-937X/$ - see front matter © 2006 International Association for Gondwana Research. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.gr.2006.07.004

E.S. Farahat et al. / Gondwana Research 11 (2007) 516–528

517

2. Analytical techniques

Fig. 1. Map showing the main basaltic occurrences in northern Egypt and sample localities (modified after Williams and Small, 1984).

Tertiary volcanic activity is intimately associated with the Red Sea rifting. Abdel Aal (1981) and Kamel et al. (1981) classified Phanerozoic volcanic rocks in Egypt into four groups, namely: i) tholeiitic basalts, ii) alkali basalts (hawaiite, sodic basalts, potassic basalts), iii) transitional basalts of the sodic series, and iv) nephelinites (olivine and olivine–melilite nephelinites). In this work we record for the first time (to the best of our knowledge) mafic xenoliths from two occurrences of the Egyptian Tertiary basalts (Fig. 1): At Abu Zaabal (AZ) near Cairo occur sodic transitional basalts, while at Gabal Mandisha (GM) in the Bahariya Oases, central Western Desert, a hawaiite is exposed. The mineralogical and textural properties of these xenoliths along with their mineral chemistry and whole-rock geochemistry were used to constrain their origin and petrogenetic significance.

Quantitative mineral analyses were carried out using a JEOLISM 6310 scanning electron microscope at the Institute of Earth Sciences (Mineralogy and Petrology), Karl Franzens University, Graz, Austria. The analyses were performed at 15 kV acceleration voltage and 5 nA sample current. Na was measured by a wavelength dispersive spectrometer (WDX) with a MicrospecTAP analyzing crystal. The other elements of interest were measured with OXFORD-energy dispersive detector (EDX), model 6687. Natural and synthetic minerals were used as standards. The analyses were normally done at a magnification of 300,000x. Fe3+contents of the analyses were estimated with the general equation of Droop (1987). Whole-rock major- and trace-elements were analyzed at the Institute of Mineralogy, Hannover University, Germany, using a Philips automatic X-Ray fluorescence spectrometer. Major element data were determined on fusion discs with lithium tetraborate flux. FeO concentrations were determined by titration. Trace elements were determined on pressed powder pellets using Rh and W excitation. Results are given with a b 1% error. Loss on ignition (LOI) was determined by heating powder samples for 50 min at 1050 °C. Whole-rock rare-earth elements (REE) in addition to Sc, Ta, Hf, Th and U were determined by Instrumental Neutron Activation Analysis (INAA) at the All-Union Scientific Research Institute of Geology (WNIIG) in Moscow, Russia. The measurement procedure, accuracy and precision of the analyses have been described by Parkhomenko (1980). 3. Field description and relations 3.1. Abu Zaabal The area lies some 30 km to the north-northeast of Cairo (Fig. 1). It is more or less a plateau with an average altitude of about 20 m above sea level. The succession of the different rock units can be arranged chronologically from top to bottom as follows: i) alluvium, ii) basalts, and iii) sandstone (Fig. 2). The basalt occurs in the form of a flow. The lava was most probably erupted to the surface in a relatively quiet fashion, as expected from the absence of any pyroclastic rocks in the area. The thickness of the basalt sheet as revealed from boreholes

Fig. 2. Distribution of basalt in relation to stratigraphic units as revealed from drilled boreholes in Abu Zaabal area (Abdel Aal, 1975).

518

E.S. Farahat et al. / Gondwana Research 11 (2007) 516–528

Table 1 Petrographical summary and modes of host rocks and mafic xenoliths (in vol.%) AZ Host basalt Plagioclase K-feldspar

Host basalt

55% An77–57 3.5–0.3 mm 32% An49–28 6.0–1.0 mm 57% An62–53 1.7–0.4 mm – 15% Anorthoclase – Or50Ab32An18–Or15Ab83An2 6.0–1.0 mm

– b0.5% Fo40–49 0.6– 0.5 mm Clinopyroxene 34% augite, piegonite 4.4–0.2 mm Fe–Ti oxides 6% ilmenite and Ti-magnetite 0.9–0.5 mm Accessory and 5% chlorite calcite apatite secondary minerals basaltic hornblende Feldspathoid Olivine

GM Mafic xenolith

Mafic xenolith

– –

5% nepheline 12% Fo39–70 2.5–0.6 mm

37% An47–32 8.0–1.0 mm 13% Anorthoclase Or45Ab42An13Or18Ab80.5 An1..5 5.0–1.0 mm – –

34% augite, piegonite 10–1.0 mm

20% augite, Ti–augite 2.3–0.5 mm

45% Ti–Augite 15–3.0 mm

8% ilmenite and Ti-magnetite 8.0–1.0 mm 5% ilmenite and Ti-magnetite 1.1–0.4 mm 11% partial melt-fraction chlorite 1% chlorite calcite apatite basaltic hornblende serpentine zeolites

drilled in AZ area ranges from 19.5 to 35.9 m. The thickness decreases rapidly in the western direction where the basalt sheet terminates against the alluvium deposits. Sandstone of Lower Oligocene underlies the basalt sheet as observed in the exposed surface near the bottom of the AZ basalt quarries, and from boreholes. The sandstone, forming a series of varicolored, alternating layers, is fine-grained, clayey, porous and saturated with water, which finds its way through cracks, fissures and joints of the overlying basalt sheet. The AZ basalts have been dated at

5% ilmenite and Ti-magnetite 3.0–0.5 mm b0.5% chlorite

23 ± 0.7 Ma using K/Ar whole-rock method (Meneisy and Abdel Aal, 1983). In order to follow the extension and thickness of the buried basalts in the area, the Egyptian Geological Survey and Mining Authority (EGSMA) drilled fourteen boreholes in 1967. The mafic xenolith was reported in the core samples of borehole No. G2 at depth of 23 m from the ground level (Fig. 2). The xenolith is 25 cm long, and has gradational boundaries with the finer grained parts of the flow.

Fig. 3. Photomicrographs showing the general views of host rocks and mafic xenoliths. Crossed polarizers; scale bars are 1.0 mm long. a) Porphyritic texture of AZ host basalt. b) Intergranular texture of GM host olivine basalt. c) and d) The mafic xenoliths of AZ and GM, respectively. Views dominated by coarse plates of untwinned plagioclase (Plg) and clinopyroxene (CPX).

E.S. Farahat et al. / Gondwana Research 11 (2007) 516–528

3.2. Gabal Mandisha The mountain is one of four basaltic occurrences in the northern part of the Bahariya Oases depression. It lies to the east of Mandisha Village, some 5 km east of Bawiti, the Oases Seat. The GM basalt sheet forms a generally flat oval-shaped hill about 2 km × 1.5 km wide, overlying Cenomanian shale, siltstone and claystone of the Bahariya beds. A smaller isolated hill, capped with basalt, is present east of the main exposure. The basalt sheet attains a thickness of 4 to 10 m. Columnar joints are well

519

developed forming rosettes on the surface of the sheet. Horizontal sheeting subdivides the vertical columns and is more intense near the upper surface of the flow. The basalt sheet is subdividable into two horizons, one upper fresh horizon and a lower moderately weathered unit. El Sharkawi et al. (2002) ascribed this alteration to the effect of paleoaquifers in fractured basalt. A minor dyke cuts the sheet in the western edge of GM with an average thickness of about 140 cm and trending north 80° west. The dyke is highly fractured and intensively weathered. El Etre and Mostafa (1978) reported several blister cones on the surface of the main

Fig. 4. Photomicrographs showing the petrographical features of mafic xenoliths. Crossed polarizers except (e) which is taken in plain-polarized light; scale bars are 1.0 mm long. a) A large blade of plagioclase (Plg) showing irregular extinction poikilitically enclosing clinopyroxene grains similar to graphic texture. A long needle of ilmenite (Ilm) is shown along the right side. AZ. b) Finger-like intergrowths of clinopyroxene (CPX) and plagioclase (Plg). GM. c) A cluster of untwinned plagioclase. GM. d) Large blades of plagioclase showing spongy (sieve) texture due to melting. AZ. e) Fe–Ti oxides (black) intergrown with a feather-like plate of clinopyroxene (CPX). Some of the Fe–Ti oxide grains show irregular or skeletal “fish-bone” like pattern. AZ. f) Fine granophyric intergrowth between plagioclase and clinopyroxene due to eutectic crystallization of the melt fraction on cooling. A long needle of ilmenite (black) is shown. AZ.

520

E.S. Farahat et al. / Gondwana Research 11 (2007) 516–528

basaltic sheet of GM, substantiating its effusive nature. Meneisy and Abdel Aal (1983) assigned a K/Ar whole-rock age of 18 ± 1 Ma to these basalts. The mafic xenolith, encountered at the base of a trench cutting the basalt sheet, forms of a large rounded block up to a meter in diameter. It is often highly friable, presumably due to pervasive alteration, and has sharp contacts with the basalt sheet.

calcite, apatite and basaltic hornblende occur interstitially, replace the early formed minerals or fill cracks, vesicles and cavities. The GM rock is a fine to medium-grained olivine basalt with an intergranular texture as a predominant fabric (Fig. 3b), while subophitic and porphyritic textures, with phenocrysts of olivine are less abundant. It consists of calcic plagioclase, clinopyroxene, olivine and Fe–Ti oxides. Accessory minerals are nepheline and apatite, while secondary ones are chlorite, calcite, serpentine and zeolites (analcite and natrolite).

4. Petrography 4.2. Mafic xenoliths A summary of the petrographic features of the host rocks and xenoliths from the two areas is presented in Table 1. 4.1. Host rocks The AZ basalts are fresh and exhibit porphyritic texture as the predominant fabric (Fig. 3a). Intergranular, ophitic and subophitic textures are also present, but are less abundant. Vesicles and amygdales are well developed, especially at the upper and lower parts of the flow. Veinlets of chlorite and carbonate are found intersecting the fresh basalt. The AZ basalt contains phenocrysts of calcic plagioclase and clinopyroxene in a fine-grained matrix of plagioclase, clinopyroxene and Fe–Ti oxides with minor olivine (Table 1). Chlorite,

Despite the different petrographical features of the host rock, the mafic xenoliths from the two areas are broadly similar. The xenoliths are medium to coarse-grained, dark-green plutonic rocks (Fig. 3c and d) with phaneritic, ophitic to subophitic, poikilitic and in places cumulate textures, typical of igneous rocks; none of them exhibits layering or tectonic fabric. Contrary to the host rocks, the xenoliths are olivine-free and composed mainly of varying proportions of bright green crystals of clinopyroxene, visible in hand specimen, fresh plagioclase, alkali feldspar and Fe–Ti oxides. According to their modal composition, the xenoliths are monzodiorites (after IUGS). The AZ xenolith shows evidence of partial melting by incorporation in the host basalts.

Table 2 Representative electron microprobe analyses of clinopyroxene from the host rocks and mafic xenoliths AZ

GM

Host

SiO2 TiO2 Al2O3 Fe2O3 FeO MnO MgO CaO Na2O Total

Xenolith

Xenolith

ZH-2

ZH-3

ZH-4

ZX-1

ZX-2

ZX-3

ZX-4

MH-1

MH-2

MH-3

MH-4

MX-1

MX-2

MX-3

MX-4

51.05 0.93 1.06 2.11 16.93 0.45 13.81 14.43 0.25 101.02

53.03 0.88 1.28 0.15 9.32 0.36 17.26 17.82 0.20 100.31

52.21 1.01 1.31 1.33 12.11 0.20 16.42 16.01 0.27 100.87

50.08 0.23 0.21 2.92 27.67 0.63 14.82 3.22 0.27 100.04

50.48 0.87 1.20 0.26 17.99 0.44 12.03 15.63 0.27 99.17

49.85 0.61 0.76 0.00 29.35 0.74 11.70 5.94 0.15 99.10

51.45 0.70 1.15 0.30 15.78 0.31 14.34 14.99 0.28 99.3

49.58 1.68 2.78 1.93 11.04 0.36 13.35 18.86 0.31 99.893

49.81 1.31 3.04 0.00 13.52 0.16 8.93 22.41 0.30 99.48

51.39 0.86 2.68 1.66 6.23 0.18 16.25 19.80 0.32 99.38

50.54 1.48 2.20 1.60 8.91 0.37 14.21 20.00 0.33 99.64

52.17 1.16 2.09 0.00 8.19 0.24 15.46 20.40 0.26 99.98

51.69 1.30 1.01 0.00 13.32 0.39 10.77 20.30 0.70 99.48

52.68 1.59 2.41 0.00 8.65 0.23 14.51 20.09 0.54 100.70

49.98 2.03 3.46 0.00 8.37 0.09 14.53 19.69 0.39 98.54

50.80 1.90 3.31 0.16 8.16 0.14 14.99 19.99 0.40 99.85

Formula based on 6 oxygens Si 1.929 1.953 Ti 0.027 0.024 Al 0.047 0.056 Fe3+ 0.061 0.004 Fe2+ 0.541 0.287 Mn 0.014 0.011 Mg 0.778 0.947 Ca 0.584 0.703 Na 0.019 0.014 Total 4.000 4.000 Wo En Fs Mg#

Host

ZH-1

30.72 40.91 28.37 58.98

36.29 48.89 14.82 76.74

Mg# = 100 Mg / (Mg + Fe2+).

1.934 0.028 0.057 0.037 0.375 0.006 0.907 0.635 0.019 4.000 33.12 47.31 19.56 70.75

1.956 0.007 0.010 0.086 0.903 0.021 0.863 0.135 0.020 4.000 7.10 45.40 47.50 48.87

1.954 0.025 0.055 0.008 0.582 0.014 0.694 0.648 0.020 4.000 33.68 36.07 30.25 54.39

1.980 0.018 0.036 0.000 0.975 0.025 0.692 0.253 0.012 3.991 13.18 36.04 50.78 41.51

1.960 0.020 0.052 0.009 0.503 0.010 0.814 0.612 0.021 4.000 31.73 42.20 26.08 61.81

1.874 0.048 0.124 0.055 0.349 0.012 0.752 0.764 0.023 4.000 40.97 40.32 18.71 68.30

1.916 0.038 0.138 0.000 0.435 0.005 0.512 0.923 0.022 3.989 49.36 27.38 23.26 54.07

1.906 0.024 0.117 0.046 0.193 0.006 0.898 0.787 0.023 4.000 41.91 47.82 10.28 82.31

1.899 0.042 0.097 0.045 0.280 0.012 0.796 0.805 0.024 4.000 42.80 42.32 14.89 73.98

1.932 0.032 0.091 0.000 0.254 0.008 0.853 0.810 0.019 3.999 42.25 44.50 13.25 77.06

1.978 0.038 0.046 0.000 0.426 0.013 0.615 0.832 0.052 3.999 44.44 32.81 22.75 59.08

1.943 0.044 0.105 0.000 0.267 0.007 0.798 0.794 0.039 3.997 42.72 42.93 14.35 74.93

1.881 0.057 0.154 0.000 0.264 0.003 0.815 0.794 0.029 3.996 42.40 43.52 14.07 75.53

1.886 0.053 0.145 0.005 0.253 0.004 0.830 0.795 0.029 4.000 42.35 44.16 13.49 76.64

E.S. Farahat et al. / Gondwana Research 11 (2007) 516–528

521

Table 3 Representative electron microprobe analyses of feldspars from the host rocks and mafic xenoliths AZ

GM

Host

Xenolith

ZH-1 SiO2 53.90 TiO2 0.00 Al2O3 28.79 FeO 0.56 MnO 0.00 MgO 0.22 CaO 12.36 Na2O 4.46 K2O 0.17 Total 100.46

Host

Xenolith

ZH-2

ZH-3

ZH-4

ZX-1

ZX-2

ZX-3

ZX-4

MH-1

MH-2

MH-3

MH-4

MX-1

MX-2

MX-3

MX-4

55.03 0.19 28.05 0.86 0.00 0.12 11.04 5.03 0.24 100.56

52.78 0.04 29.35 0.54 0.03 0.12 12.75 3.99 0.20 99.80

53.96 0.05 28.80 0.89 0.06 0.18 11.96 4.45 0.26 100.61

56.66 0.11 27.02 0.52 0.00 0.11 10.18 5.52 0.33 100.45

56.34 0.13 26.57 0.62 0.00 0.09 9.95 5.88 0.39 99.97

57.85 0.12 26.32 0.48 0.00 0.09 9.35 6.22 0.53 100.96

66.35 0.04 18.34 0.25 0.01 0.00 1.42 5.25 8.35 100.01

52.49 0.09 29.02 0.67

52.55 0.15 29.45 0.51 0.06 0.16 13.16 3.94 0.23 100.2

58.32 0.12 26.02 0.63 0.00 0.07 8.96 6.40 0.46 100.98

51.54 0.05 29.88 0.59

60.66 0.05 24.82 0.32 0.00 0.00 6.30 6.85 0.94 99.94

56.50 0.18 27.53 0.42 0.00 0.02 9.33 5.67 0.39 100.04

63.88 0.15 21.45 0.26 0.00 0.00 2.76 7.78 3.12 99.40

65.79 0.17 19.99 0.30 0.00 0.00 0.79 6.82 5.52 99.38

Number of cations based on Si 2.433 2.478 Ti 0.000 0.006 Al 1.532 1.488 Fe2+ 0.021 0.033 Mn 0.000 0.000 Mg 0.015 0.008 Ca 0.598 0.533 Na 0.390 0.439 K 0.010 0.014 Total 4.998 4.998

8 oxygens 2.405 0.002 1.577 0.021 0.001 0.008 0.623 0.353 0.011 5.000

An Ab Or

63.10 35.74 1.16

59.93 39.10 0.97

54.04 44.54 1.41

2.435 0.002 1.532 0.034 0.002 0.012 0.579 0.389 0.015 5.000 58.86 39.62 1.52

2.542 0.004 1.429 0.020 0.000 0.007 0.489 0.480 0.019 4.990 49.49 48.58 1.92

2.55 0.00 1.41 0.02 0.00 0.01 0.48 0.52 0.02 5.011 47.30 50.54 2.16

2.58 0.00 1.39 0.02 0.00 0.01 0.45 0.53 0.03 5.005 43.95 53.10 2.95

Alkali feldspar (mostly anorthoclase with minor sanidine) occurs as euhedral to subhedral untwinned crystals, or shows simple twinning. Plagioclase (andesine) forms large crystals (Fig. 3c and d)), some of which may show complex twinning

2.997 0.001 0.976 0.009 0.000 0.000 0.069 0.459 0.481 4.992 6.84 45.49 47.67

0.16 12.81 4.12 0.22 99.58

2.399 0.003 1.563 0.03 0.011 0.627 0.365 0.013 5.007 62.39 36.32 1.29

2.386 0.005 1.576 0.02 0.002 0.011 0.64 0.347 0.013 4.999 64.00 34.70 1.30

2.599 0.004 1.366 0.023 0.000 0.005 0.428 0.553 0.026 5.004 42.50 54.92 2.58

0.18 13.74 3.5 0.2 99.68

2.356 0.002 1.61 0.02 0.012 0.673 0.31 0.012 4.998 67.64 31.16 1.21

2.721 0.002 1.312 0.012 0.000 0.000 0.303 0.596 0.054 4.999 31.80 62.55 5.66

2.546 0.006 1.462 0.016 0.000 0.001 0.451 0.496 0.022 5.000 46.54 51.18 2.28

2.86 0.01 1.13 0.01 0.00 0.00 0.13 0.68 0.18 5.000 13.42 68.51 18.06

2.967 0.006 1.063 0.011 0.000 0.000 0.038 0.596 0.318 4.999 3.98 62.64 33.38

but usually are untwinned and express irregular extinction. They contain poikilitic inclusions of small crystals of clinopyroxene forming a graphic texture (Fig. 4a) or invade large plates of clinopyroxene forming ophitic to subophitic textures. Sometimes,

Table 4 Representative electron microprobe analyses of Fe–Ti oxides from the host rocks and mafic xenoliths AZ

GM

Titanomagnetite Host

SiO2 TiO2 Al2O3 Fe2O3 FeO MnO MgO Total No. of O Si Ti Al Fe3+ Fe2+ Mn Mg Total

Ilmenite Xenolith

Titanomagnetite

Host

Xenolith

Host

Ilmenite Xenolith

Host

Xenolith

ZH-1

ZH-2

ZX-1

ZX-2

ZH-3

ZH-4

ZX-3

ZX-4

MH-1

MH-2

MX-1

MX-2

MH-3

MH-4

MX-3

MX-4

0.54 9.91 2.09 46.88 40.34 0.16 0.66 100.59

0.31 6.78 1.46 54.19 37.01 0.16 0.66 100.57

0.09 19.56 1.14 28.81 48.43 0.44 0.16 98.63

0.04 27.11 0.80 15.07 54.68 0.69 0.46 98.85

0.37 49.02 0.52 7.82 41.22 0.41 1.62 100.98

0.36 47.69 0.32 9.43 38.70 0.70 2.19 99.40

0.02 51.27 0.09 3.55 44.51 0.58 0.58 100.60

0.07 51.16 0.05 2.53 45.15 0.42 0.29 99.66

0.56 8.71 0.86 49.63 38.11 0.56 0.71 99.14

0.62 6.93 1.84 51.54 35.72 0.34 1.33 98.32

0.32 21.08 2.15 25.78 49.14 0.52 1.19 100.18

0.47 19.13 2.30 28.54 46.32 0.77 1.62 99.15

0.02 50.56 0.11 5.43 42.38 0.41 1.51 100.42

50.48 0.09 5.96 41.92 0.58 1.62 100.65

0.23 52.91 0.43 0.91 40.53 0.37 3.90 99.28

0.22 52.22 0.43 2.07 40.58 0.61 3.38 99.51

4 0.020 0.277 0.092 1.313 1.256 0.005 0.037 3.000

4 0.012 0.191 0.065 1.530 1.161 0.005 0.037 3.000

4 0.003 0.559 0.051 0.824 1.539 0.014 0.009 3.000

4 0.002 0.767 0.035 0.427 1.721 0.022 0.026 3.000

3 0.009 0.911 0.015 0.145 0.852 0.009 0.060 2.000

3 0.009 0.897 0.009 0.178 0.810 0.015 0.082 2.000

3 0.001 0.965 0.003 0.067 0.931 0.012 0.022 2.000

3 0.002 0.973 0.001 0.048 0.955 0.009 0.011 2.000

4 0.021 0.249 0.039 1.421 1.212 0.018 0.040 3.000

4 0.024 0.198 0.082 1.474 1.135 0.011 0.075 3.000

4 0.012 0.584 0.093 0.715 1.514 0.016 0.065 3.000

4 0.017 0.534 0.101 0.797 1.437 0.024 0.090 3.000

3 0.000 0.947 0.003 0.102 0.883 0.009 0.056 2.000

3 0.000 0.943 0.003 0.111 0.871 0.012 0.060 2.000

3 0.006 0.980 0.012 0.017 0.834 0.008 0.143 2.000

3 0.005 0.969 0.013 0.038 0.837 0.013 0.124 2.000

522

E.S. Farahat et al. / Gondwana Research 11 (2007) 516–528

Fig. 5. Clinopyroxene compositions of the host rocks (a) and mafic xenoliths (b) plotted in the En–Wo–Es classification diagram of Morimoto (1988). Open and closed triangles = AZ host rock and xenolith, respectively. Closed and open circles = GM host rock and xenolith, respectively.

particularly in GM xenolith, intergrowth of coarse plagioclase laths and large clinopyroxene plates forms a finger-like pattern (Fig. 4b). In places, cumulus feldspars are also present suggesting adcumulate growth (Fig. 4c). Sometimes, partial melting evidence in AZ xenolith is preserved along plagioclase grain boundaries as a spongy (sieve) texture (Tsuchiyama, 1986) (Fig. 4d). Clinopyroxene occurs either as large (up to 1.5 × 0.5 cm) plates or as small grains enclosed in plagioclase (Fig. 4a). The plates are unzoned and free of exsolution lamellae. The clinopyroxene plates typically enclose grains of Fe–Ti oxides along the rims forming a cedar-like pattern (Fig. 4e). Fine-grained granophyric intergrowths of clinopyroxene and plagioclase commonly observed in the AZ xenoliths (Fig. 4f), probably represent eutectic crystallization of the partial melt fraction on cooling (Shelley, 1993). The Fe–Ti oxides in the investigated xenoliths include titanomagnetite and ilmenite in decreasing order of abundance. Titanomagnetite occurs as skeletal crystals of “fishbone” like patterns (Fig. 4e), blocky euhedral to anhedral crystals and thin to broad plates. Ilmenite occurs as discrete grains of varied morphology. It forms blocky euhedral to subhedral crystals, thin plates, or extremely long (up to 8 mm) needle-like crystals especially in the AZ xenolith (Fig. 4a and f).

Pigeonite is rarely recorded from the xenoliths and host rocks of the AZ area (Fig. 5). It is worth mentioning that the clinopyroxene compositional ranges of the host rocks and xenoliths from each area are broadly similar. TiO2 contents vary from 0.61 to 1.68 wt.% and from 0.8 to 2.03 wt.% for the AZ and GM xenoliths, respectively. This is generally similar to those of the AZ (0.23–1.25 wt.%) and GM (0.86–1.55 wt.%) host rocks (Table 2). The same is true for the Al and Na contents of the xenoliths (0.73–2.78 and 0.97– 3.46 wt.% Al2O3 and 0.14–0.34 and 0.39–0.71 wt.% Na2O, for AZ and GM xenoliths, respectively). This is also correlated with those of their host basalts (0.21–2.74 and 1.31–4.59 wt.% Al2O3, 0.08–0.38 and 0.21–0.46 wt.% Na2O, for AZ and GM host rocks, respectively). Clinopyroxene of the AZ xenolith is relatively low in Mg# [100Mg/(Mg+Fe2+)], 41–68, (Table 2) compared to those of the GM (59–80), implying the more evolved nature of the former. Moreover, the low Al, Ti and Na contents of clinopyroxenes from the AZ host rocks and xenoliths (Table 2) compared to those of GM, suggest that the AZ rock crystallized from a lessalkaline magma.

5. Mineral chemistry The results of electron microprobe analyses of each mineral phase from the host rocks and mafic xenoliths are summarized in Tables 2, 3 and 4. 5.1. Olivine Olivine is encountered in the host rocks only, while the xenoliths are olivine-free. The olivine compositions vary from Fe-rich in the AZ basalts (Fo40–49) to relatively Mg-rich in the GM basalts (Fo39–70). 5.2. Clinopyroxene Clinopyroxene of the investigated host rocks and xenoliths is augite according to the classification of Morimoto (1988).

Fig. 6. Feldspar compositions of the investigated mafic xenoliths. Symbols as in Fig. 5. Light and dark areas are the plagioclase compositions of GM and AZ host rocks, respectively.

E.S. Farahat et al. / Gondwana Research 11 (2007) 516–528 Table 5 Analytical results of major, trace and rare-earth elements from host rocks and xenoliths AZ

GM

Host ZH-1

Xenolith ZH-2

Host

anorthoclase and rarely sanidine of also similar compositional range (Or50Ab32An18 to Or15Ab83An2 and Or45Ab42An13 to Or18Ab80.5An1.5 for AZ and GM, respectively). 5.4. Fe–Ti oxides

Xenolith

ZX-1

ZX-2

MH-1 MH-2 MX-1

MX-2

Major elements (wt.%) 47.97 48.29 SiO2 2.74 2.54 TiO2 Al2O3 15.50 15.40 Fe2O3 5.72 5.59 FeO 6.86 6.98 MnO 0.17 0.16 MgO 5.97 6.21 CaO 9.78 9.52 Na2O 2.60 2.65 0.82 0.79 K2O L.O.I 2.24 2.35 Sum 100.37 100.48 Mg# 47.60 48.70

50.30 2.51 11.10 8.65 5.55 0.11 4.20 7.27 3.81 3.48 3.49 100.47 36.30

50.11 2.47 11.35 8.70 5.90 0.09 3.90 7.18 3.95 3.52 3.37 100.54 34.00

46.95 2.03 14.38 3.11 8.93 0.14 8.03 9.74 3.52 1.40 1.73 99.96 56.90

47.19 1.98 14.20 2.67 8.85 0.14 8.77 9.75 3.44 1.27 1.64 99.90 60.20

47.35 2.09 14.12 4.10 6.94 0.10 6.29 10.49 3.87 2.56 2.66 100.57 52.90

47.49 2.13 13.96 4.00 6.68 0.12 6.19 10.75 4.10 2.48 2.67 100.57 53.30

Trace element (ppm) Cu 67 65 Pb 10 7 Zn 108 124 Ba 245 258 Sr 350 378 Cr 98 115 Co 43 48 Ni 54 60 V 197 317 Y 40 45 Zr 218 219 Hf 4.9 5.3 Th 2.1 2.2 Sc 33.5 33.9 Nb 19 20 Ta 1.2 1.3 La 24.7 22.5 Ce 44.2 42.3 Nd 29 26 Sm 7.1 6.5 Eu 2.2 1.9 Tb 1.2 1 Yb 3.6 3 Lu 0.39 0.34 (La/Y)N 4.92 5.38 (Tb/Yb)N 1.52 1.52 Ba/Zr 1.12 1.18 Nb/Th 9.05 9.09 Nb/La 0.77 0.89

128 8 253 860 280 40 10 9 347 50 550 9.4 4.7 30.7 28 2.4 44.4 83 54 12.3 3.6 1.9 5.8 0.73 5.49 1.49 1.56 5.96 0.63

100 9 252 810 250 38 6 10 353 48 500 8.6 4.2 28 27 2.2 40 74 49 11 3.2 1.7 5 0.62 5.74 1.55 1.62 6.43 0.68

96 8 101 253 410 223 47 178 142 30 116 2.7 1.9 17.4 24 1.1 17.2 29.5 17 4.7 1.7 0.5 1 0.11 12.34 2.27 2.18 12.63 1.40

84 8 98 275 420 276 48 164 145 25 120 2.8 2.3 19.2 22 1.2 20.8 34.8 20 4.9 1.8 0.6 1.1 0.12 13.56 2.48 2.29 9.57 1.06

64 10 107 1110 500 160 20 100 646 26 310 3.6 2.7 14 25 1.7 24.5 43.1 28 6.1 2.3 0.8 1.7 0.22 10.34 2.14 3.58 9.26 1.02

60 12 103 1150 497 162 25 105 596 34 316 3.3 2.8 10.6 33 1.5 23.5 41.2 26 5.5 2.1 0.7 1.5 0.19 11.24 2.12 3.64 11.79 1.40

ol + ne

ol + ne ol + ne ne + ol

ne + ol

Characteristic normative minerals hy + ol ol + ne ol + ne

523

Fe–Ti oxides from host rocks and xenoliths of both areas are titanomagnetite and ilmenite, but with distinctive differences in their chemical composition (Table 4). Ilmenite and titanomagnetite from the AZ xenolith are Mg-poor (0.29–0.62 wt.% MgO for ilmenite and 0.16–0.46 wt.% MgO for titanomagnetite) compared to those of AZ host rock (1.03–2.19 wt.% MgO for ilmenite and 0.27–0.67 wt.% for titanomagnetite). Ilmenite and titanomagnetite of the GM xenolith have higher MgO (2.6–3.94 and 1.19–2.79 wt.% MgO for ilmenite and titanomagnetite, respectively) compared to those of the host basalt (0.63–1.62 and 0.21–1.33 wt.% MgO for ilmenite and titanomagnetite, respectively). On the other hand, titanomagnetite of the xenoliths has a higher Mn content (0.4–0.69 and 0.49– 0.77 wt.% MnO for AZ and GM, respectively) compared to that of ilmenite (0.16–0.79 and 0.37–0.27 wt.% MnO for AZ and GM, respectively), which is not the case in the host rock (0.03–0.16 and 0.34–0.56 wt.% MnO for titanomagnetite and 0.4–0.90 and 0.31–0.92 wt.% MnO for ilmenite from AZ and GM, respectively). According to Neumann (1974) the ratio Mnsp/Mnilm increases with increasing fO2. Thus, the high Mncontent in titanomagnetite from xenoliths of both areas clearly reflects the relatively high fO2 during their crystallization. 6. Geochemistry The analytical results of bulk-rock major, trace and rare-earth elements of host rocks and xenoliths are given in Table 5. The main difference in major element contents between xenoliths and host rocks is that the xenoliths have more SiO2, Na2O, K2O and H2O and low MgO. The generally low Mgnumbers [Mg#= 100Mg/(Mg+Fe2+)] of all samples (34–60) indicate that they do not represent primary melts. On the other hand, the xenoliths have lower Mg# (34–53) compared to those of the host rocks (48–60), suggesting their more evolved nature.

2+

Mg# = 100 Mg / (Mg + Fe ), hy = hypersthene, ol = olivine, ne = nepheline.

5.3. Feldspar Feldspar of the host rocks from both occurrences is Ca-rich plagioclase (An53–77), whereas feldspar of the xenoliths contains plagioclase and alkali feldspar (Table 3; Fig. 6). Plagioclase from both xenoliths is andesine of similar compositional range (An28–49 and An32–47 for AZ and GM xenoliths, respectively). Alkali feldspar from both xenoliths is

Fig. 7. Patterns of incompatible elements normalized to N-MORB (after Pearce, 1983). Symbols as Fig. 5.

524

E.S. Farahat et al. / Gondwana Research 11 (2007) 516–528

Normative mineral compositions were calculated following the CIPW scheme and characteristic normative minerals are listed in Table 5. All host rocks and xenoliths are variable olivine- and nepheline-normative, except AZ host basalts that are hypersthene- and olivine-normative. According to normative composition, the AZ xenolith are classified as “foid-bearing monzodiorite” and GM xenolith as “essexite (foid-bearing monzogabbro)” (after IUGS). The xenoliths are typically poor in the compatible elements Ni, Cr, Co and Sc compared to the host rocks. The AZ xenolith and host rock have low Ni and Cr contents (9–60 and 38– 115 ppm, respectively) compared to those of GM (100–178 and 160–276 ppm, respectively), again implying the more fractionated nature of the former, consistent with the clinopyroxene compositions. Moreover, the Ni and Cr contents of the GM basalts are reasonably high, implying they are close to primary mantle-derived magmas. Contrary to those of AZ, GM xenolith is rich in Sr compared with host rocks. However, the high V concentrations in both AZ and GM xenoliths relative to host rocks indicate that the xenoliths were affected by Fe–Ti oxide and possibly clinopyroxene accumulation. In broad geochemical evolutionary term, the characteristics of the GM host rocks and xenolith are generally more primitive than those of AZ. The trace elements of the xenoliths and host basalts are plotted in the MORB-normalized spidergram (Fig. 7). Both xenoliths and host rocks are enriched in the whole range of incompatible elements, other than Y and Yb, relative to N-type MORB. This suggests their derivation by partial melting of an enriched subcontinental mantle source combined with variable degrees of crustal contamination en route to the surface (Wilson, 1989). MORB-normalized patterns of the xenoliths are generally parallel to those of the host rocks that may reflect their cognate origin. However, the xenolith patterns show higher contents of some elements especially the LILE elements (K and Ba) in addition to Th, Ce, Zr, Hf and Sm, which may be a consequence of crustal contamination and/or crystal fractionation. Y appears to be little varied by the effects of source heterogeneity and crustal contamination. The close similarity in

Fig. 8. Chondrite-normalized REE patterns of the host rocks and xenoliths. Chondrite-normalization values are from Sun and McDonough (1989). Symbols as Fig. 5.

Y and Yb contents of the xenoliths and host rocks from each locality further confirm their cognate origin. The REE patterns of the host rocks and xenoliths from both areas are generally parallel (Fig. 8), providing the most convincing evidence that the xenoliths are cognate in origin to their hosts. The xenolith samples of both areas have enriched patterns of REE with LREE fractionated over HREE [(La/Yb)N = 5.49–5.74 and 10.34– 11.24 for the AZ and GM, respectively], but these patterns are almost parallel to those of the host rocks [(La/Yb)N = 4.92–5.38 and 12.34–13.56 for the AZ and GM, respectively], indicating progressive enrichment by fractional crystallization and/or crustal assimilation. Furthermore, the HREE of xenolith and host rocks of both areas are fractionated and typically parallel for the xenolith and host rock of each area [(Tb/Yb)N = 1.49– 1.55 vs. 1.52 for AZ xenolith and host, respectively, and 2.14– 2.12 vs. 2.27–2.48 for xenolith and host from GM, respectively]. This clearly suggests derivation of the xenolith and host rock of each area from one single parent. 7. Discussions 7.1. Crystallization conditions The complete lack of hydrous phases indicates that the mafic xenoliths crystallized from anhydrous magma. Petrographic study indicates the early crystallization of clinopyroxene, probably preceded by olivine at an early stage, followed by plagioclase, and then Fe–Ti oxides. Alkali-feldspar appeared as the last major phase in the xenoliths. Lindsley's (1983) pyroxene geothermometer yields minimum (because of the absence of orthopyroxene) crystallization temperatures of 950–1150 °C and 920–1050 °C for the studied xenoliths from AZ and GM, respectively. It is evident that pyroxenes preserve liquidus or near-liquidus temperatures. However, the application of two-feldspar thermometer (Haselton et al., 1983) yields very low temperature (b500 °C), probably reflecting inequilibrium crystallization or most likely subsolidus re-equilibration on cooling. A quantitative estimate of pressure can be obtained using the clinopyroxene geobarometer of Nimis (1999). This barometer is based on the regular change of some structural parameters (as cell volume and M1 site volume) with pressure, and can be applied for near-liquidus clinopyroxene. The calculation does not require direct X-ray diffraction analysis of the clinopyroxene crystals because the required structural parameters can be simulated from chemical data. So, the geobarometer only requires clinopyroxene major element compositions and independent estimate of crystallization temperatures. The estimated pressures (computed by the CpxBar program of Nimis, 2000) using the anhydrous melt calibration are negative values. However, using the slightly hydrous melt calibration for a temperature of 1000 °C the highest pressures are less than 4 kbar (± 3.1 kbar) for the xenoliths of both AZ and GM, indicating crystallization at upper crustal levels. The shallow depth of crystallization of the xenoliths is consistent with the presence of phaneritic, subophitic and cumulate textures. Thus, it is reasonable to

E.S. Farahat et al. / Gondwana Research 11 (2007) 516–528

assume that pressures of ∼ 1–3 kbar were prevalent during crystallization of the investigated xenoliths. The regular partitioning of the minor element contents (Mn and Mg) between ilmenite and titanomagnetite of the investigated xenoliths probably reflects their equilibrium crystallization. Accordingly, they can be used to estimate the temperature of equilibrations. By applying the calculations of Spencer and Lindsley (1981), the crystallization conditions is about 925– 980 °C and 1040–1080 °C for AZ and GM xenoliths, respectively, which is generally comparable with those estimated using Lindsley's (1983) clinopyroxene geothermometer. The estimated log fO2 ranges from −12 to −13 for AZ and from −10 to −11 for GM, that lie between NNO and FMQ buffers indicating oxidized conditions. 7.2. Type of the mafic xenoliths Generally, the xenoliths may represent either (1) mantle material, especially depleted melting residues (restites), lifted by the magma, (2) cognate material, representing crystal cumulates from the magma or (3) basement material ripped off the walls of the magma channel or the magma chamber. Although mafic and ultramafic xenoliths are frequently recorded from Tertiary to Quaternary alkali basalts situated on the eastern side of the Red Sea–Dead Sea rift (e.g. El Sharkawi, 1982; McGuire and Stern, 1993; Al-Mishwat and Nasir, 2004), they are unknown from the extensive Tertiary basaltic rocks of Egypt. McGuire (1987) estimated temperature of 830 to 980 °C and pressures of 5 to 9 kbar for gabbroic xenoliths present in alkaline basalts from Saudi Arabia, and suggested that these mafic xenoliths may represent older lower crustal material formed during the Neoproterozoic Pan–African event. Multiple origins for these mafic rocks seem entirely reasonable as some show granulitic metamorphic textures while others have undisturbed igneous cumulate textures (Coleman, 1993). The absence of high-pressure mineral assemblage (e.g. chrome-diopside or garnet) and plastic deformation in the investigated xenoliths precludes them from representing fragments of mantle material, whatever accidental or melting residue. Moreover, enrichments of these xenoliths in the incompatible elements argue against melting residue origin. The host lavas were intruded through Neoproterozoic basement rocks, Paleozoic and Mesozoic sedimentary rocks. As a result, it is possible that the mafic xenoliths represent accidental inclusions ripped-up from pre-existing gabbros formed during the Pan–African event. However, the mineral and chemical compositions of the investigated mafic xenoliths are not similar to those of any type of the Neoproterozoic gabbros in Egypt. The clinopyroxene and whole-rock chemical compositions of the xenoliths suggest that they are cognate to the host lavas. Yet, the lack of olivine and presence of alkali feldspars in the xenoliths contrary to the phenocryst assemblage of the host lavas along with their enrichment in the full range of incompatible elements, indicate that they are not likely to represent simple cumulates from the magma which produced the host rocks. The mafic xenoliths and their host rocks are not derived from primary mantle melts judging from their low MgO, Ni and Cr

525

contents (Table 5). The highest Mg# number of the investigated rocks (61) is lower than the Mg# (70) estimated for primitive mantle-derived basaltic magma by Green et al. (1974). Moreover, they are far from the expected composition of melts in equilibrium with mantle lherzolite mineralogies under water-undersaturated conditions (Prestvik and Goles, 1985). The most obvious explanation for this is that the primary magmas were not basaltic but high-MgO picritic, which subsequently underwent low-pressure fractionation (Cox, 1980). However, the presence of primary picritic magma is still under considerable debate (e.g. Hirschmann et al., 2003). Thompson et al. (1986) have presented a simple model for the storage and ascent of basaltic magma. They suggested that primitive magmas ponded at the Moho, where they underwent fractional crystallization until their densities were reduced sufficiently to regain buoyancy and permit further uprise. These more evolved magmas then penetrated the crust through a plexus of dykes. Thus, this model agree rather well with the investigated mafic xenoliths which may represent early-formed magma batches evolved in a magma chamber at great depth and moved upwards where they solidified as intrusive bodies at shallow crustal levels and were later ripped-up by fresh magmas. 7.3. Crustal contamination versus fractional crystallization The trace and REE element patterns of the investigated xenoliths and host rocks are parallel (Figs. 7 and 8) implying closed system crystallization of xenoliths (i.e. formed in the absence of effective removing of interstitial melts during final stage of crystallization). Consequently, the differences in the mineralogical and geochemical characteristics between xenoliths and host rocks can be explained in terms of crystal fractionation and/or crustal assimilation processes during magma evolution. Compared to the host lavas, mafic xenoliths from the two areas have a marked increase in LILE (e.g. Ba, K, and Th) with corresponding LREE enrichment. These geochemical features are considered indicative of contamination by continental crust material (McDonough, 1990). The MORB-normalized patterns of the investigated xenoliths (Fig. 7) are characterized by distinct Ba and K peaks, implying that crustal contamination could have played a major role in their origin. Kerr et al. (1995) related high Ba contents of the basaltic rocks to assimilation of felsic crustal rocks containing alkali feldspar. However, the silica-undersaturated nature of the xenoliths (Table 5) argues against significant crustal contamination. Distinct negative Nb anomalies in flood basalts were previously explained by crustal contamination (McDonough, 1990), which is not the case of the investigated xenoliths (Fig. 7). Consequently, fractional crystallization rather than crustal contamination is considered as the predominant process for the xenoliths magmatic evolution. Fractional crystallization may have been accompanied by crustal assimilation (AFC process; DePaolo, 1981). Mafic xenoliths, especially those of AZ, have lower Nb/Th and Nb/La together with higher SiO2 and Zr contents relative to their host rocks (Table 5). This can only be explained if the magmas parental to the xenoliths were contaminated by the assimilation of felsic

526

E.S. Farahat et al. / Gondwana Research 11 (2007) 516–528

crustal rocks (Zhou et al., 2000). Hooper (1994) stated that AFC, in which the ability to assimilate is controlled by the degree of exothermal crystallization should result in positive correlation between an index of crystallization (e.g. Zr) and an index of assimilation for which the Ba/Zr ratio is most suitable. The xenoliths of both areas have higher Ba/Zr and Zr contents (Table 5) compared to the host rocks, suggesting that AFC process is probably responsible for some of the geochemical characteristics of the xenoliths and cannot be completely ruled out. In the studied mafic xenoliths many aspects of their major and trace element contents are consistent with crystal fractionation. Compared to host rocks, xenoliths show low abundances of Ni and Co (Table 5), features typically attributed to significant olivine fractionation. The major part of Cr depletion in the xenoliths is most likely related to Cr-spinel rather than clinopyroxene fractionation, taking into consideration the high modal clinopyroxene in the xenoliths. The xenoliths are characterized by relatively high concentrations of Ba, K, LREE and rather Sr. These features probably reflect the importance of olivine over plagioclase in the crystal fractionation. We suggest a polybaric crystallization model for the origin of the xenoliths, i.e. fractionation of more olivine (Crspinel)- and rather clinopyroxene-rich assemblages characteristic of crystallization at higher pressures (e.g. Presnall et al., 1978) followed by shallow level crystallization. Such a model can therefore account for the difference in pressure estimates derived from the clinopyroxene geobarometer, absence of olivine, and presence of cumulate plagioclase. 7.4. Evolution model Based on the above discussion, a number of geochemical and mineralogical characteristics of host rocks and xenoliths suggest a close genetic link between these rocks. Similar patterns of incompatible and REE element abundance in the host rocks and xenoliths (Figs. 7 and 8), combined with the depletion of the xenoliths in many compatible element allow us to conclude that of all possible models of xenoliths magmas generation, fractional crystallization model is the most appropriate. Fractional crystallization occurred possibly at all stages of the xenoliths evolution. In view of the above mentioned arguments, a model for the evolution of the xenoliths is suggested as follows: 1) These xenoliths represent early-formed magma batches that ponded in deep magma chambers at the base of the crust or within the upper mantle. Fractional crystallization processes, dominated by olivine, Cr-spinel and rather clinopyroxene fractionation, occur in such chambers without significant crustal contamination because of the refractory nature of the wall rocks. Due to early crystallization of olivine and to some extent clinopyroxene, the melt become enriched in alkalis, whereas calcium decrease. As a result, plagioclase that crystallized later is impoverished in An component. 2) During their ascent, these evolved magmas may have undergone AFC process. It is important mentioning that, evolved magmas have a low capacity for assimilation (Reiners et al.,

1995; Bohrson and Spera, 2001). Consequently, the effect of AFC, although cannot be ruled out, is probably insignificant 3) These more evolved magmas emplaced at shallow crustal reservoir (most probably as dykes) where solidified. Further fractionation, under low PH2O conditions, may have occurred at these shallow magma storage reservoirs. The shallow depth of emplacement of these magmas is consistent with the pressure estimates and the many textural features of the xenoliths (e.g. phaneritic, subophitic and cumulate textures). Fresh magma pulses have broken these solidified intrusive bodies into pieces and brought onto the surface. Given the fact that these xenoliths represent early-formed magma batches, they would be most susceptible to fractionation and possibly contamination as they establish pathways to the shallow levels. Recently, Holness and Bunbury (2006) ascribed the local occurrence of cognate mafic nodules in the Kula volcanic province to the fact that the plumbing system in this region was used by a series of eruptions which occurred over a short time interval during which solidification of the conduit system was incomplete. Dungan (2005) notes that upper crustal xenoliths are unlikely to survive as intact blocks in mafic magma unless they are incorporated during rapid ascent that leads directly to eruption. The xenolith incorporated in the GM alkali basalt is likely to be transported to the surface more rapidly relative to those found in the less alkalic lavas of AZ. This may account for the small size and evidence for partial melting seen in the mafic xenolith of AZ area (Fig. 4a,d, f) and for the relatively more evolved nature of the AZ rocks. Removal of melt fraction may modify the compositions of xenoliths and host rock. The fine-grained granophyre-like intergrowth of clinopyroxene and plagioclase (Fig. 4f) indicates a degree of undercooling resulting from crystallization under low pressure and PH2O, raising the liquidus temperature of the melt fraction produced. This also may prevent the melt fraction produced from escaping to the host lavas. 8. Conclusions The mafic xenoliths preserved in the Abu Zaabal (AZ) basalts and Gabal Mandisha (GM) alkaline basalts have similar mineral assemblages (plagioclase, clinopyroxene, alkali feldspar and Fe–Ti oxide minerals). Mineralogical and chemical data suggest that they are cognate to their host lavas rather than basement materials. The fact that their mineral assemblages are different from the phenocryst assemblages of the host rocks, suggest that they are not cumulates from the host lavas. Many petrographical and geochemical characteristics show that the xenoliths may represent a more fractionated and possibly slightly contaminated member of the same parent melt as the host rocks. Pyroxene and coexisting ilmenite–titanomagnetite chemistry suggest crystallization at ∼ 1–3 kbar pressure for both xenoliths, ∼950–1150 °C vs. 920–1050 °C, and log fO2 of ∼ −12 to − 13 vs. − 10 to − 11 for AZ and GM, respectively. It is suggested that the mafic xenoliths might represent earlyformed distinctly fractioned magma batches that possibly

E.S. Farahat et al. / Gondwana Research 11 (2007) 516–528

underwent some AFC during ascent, solidified at upper crustal dykes under low PH2O, and were later ripped-up during fresh magma pulses. The AZ xenolith is small in size and shows evidence of partial melting by incorporation in the host lava, contrary to the large unmelted xenolith of GM, implying rapid rise of the GM alkaline magma. This may account for the relatively more evolved nature of the AZ rocks. The xenoliths from both AZ and GM are found near the base of the basalt sheets, suggesting that they are denser than the host magma. This may account for the rare recording of such xenoliths within the Egyptian Tertiary basalts. The xenoliths, although rare, provide direct evidence for the importance of crystal fractionation, and possibly crustal assimilation, during early evolution of the Egyptian Tertiary basalts. Acknowledgements The authors are greatly indebted to the staff members and technicians of the Institute of Earth Science (Mineralogy and Petrology), Karl Franzens University, Graz, Austria; Institute of Mineralogy, Hannover University, Germany; and All-Union Scientific Research Institute of Geology, Moscow, Russia for putting the analytical facilities at their disposal. Profs. G. Hoinkes, A. Mogessie, L. Franz, and M. Okrusch are thanked for their helpful and constructive comments on an earlier version of this paper. We thank the helpful and critical reviews by three anonymous reviewers and Journal editor (M. Santosh), which improved this manuscript. References Abdel Aal, A.Y., 1975. Geological study on Abu Zaabal basalts of Egypt. Unpublished M.Sc. Thesis, Ain Shams University, Egypt. Abdel Aal, A.Y., 1981. Comparative petrological and geochemical studies of post-Cambrian basaltic rocks in Egypt. Unpublished Ph.D. Thesis, Minia University, Egypt. Al-Mishwat, A.T., Nasir, S.J., 2004. Composition of the lower crust of the Arabian Plate: a xenolith perspective. Lithos 72, 45–72. Binns, R.A., Duggan, M.B., Wilkonson, J.F., 1970. High pressure megacrysts in alkaline lavas from north-eastern New South Wales. American Journal of Science 269, 132–168. Bohrson, W.A., Spera, F.J., 2001. Energy-constrained open-system magmatic processes: II. Application of energy-constrained assimilation-fractional crystallization (EC-AFC) model to magmatic systems. Journal of Petrology 42, 1019–1041. Coleman, R.G., 1993. Geological evolution of the Red Sea. Oxford University Press. 186 pp. Cox, K.G., 1980. A model for flood basalt volcanism. Journal of Petrology 21, 629–650. DePaolo, D.J., 1981. Trace element and isotopic effects of combined wallrock assimilation and fractional crystallization. Earth and Planetary Science Letters 53, 189–202. De Silva, S.L., 1989. The origin and significance of crystal rich inclusions in pumices from two Chilean ignimbrites. Geological Magazine 126, 159–175. Donaldson, C.H., 1977. Petrology of anorthite-bearing gabbroic anorthosite dykes in north-west Skye. Journal of Petrology 18, 595–620. Droop, G.T.R., 1987. A general equation for estimating Fe3+ concentrations in ferromagnesian silicates and oxides from microprobe analyses, using stoichiometric criteria. Mineralogical Magazine 51, 431–435. Dungan, M.A., 2005. Partial melting at the earth's surface: implications for assimilation rates and mechanisms in subvolcanic intrusions. Journal of Volcanology and Geothermal Research 140, 193–203.

527

El Etre, H.A., Mostafa, A.R., 1978. Field relations of the main basalt occurrences of the Bahariya region, central Western Desert. Proceedings of the Egyptian Academy of Sciences 31, 191–201. El Sharkawi, M.A., 1982. Lherzolite xenoliths from the Shihan basalts, Jordan. Journal of University of Kuwait (Sciences) 9, 277–300. El Sharkawi, M.A., Sehim, A., Madani, A., 2002. Modes of occurrence of the basaltic rocks in northern Bahariya Oasis, Western Desert, Egypt. Annuls of the Geological Survey of Egypt 25, 83–100. Green, D.H., Edgan, H.D., Beasley, P., Kiss, E., Ware, N.G., 1974. Upper mantle source for some hawaiites, mugearites and benmoreites. Contributions to Mineralogy and Petrology 48, 33–43. Haselton, H.T., Hovis, G.L., Hemingway, B.S., Robie, R.A., 1983. Calorimetric investigation of the excess entropy of mixing in albite– sanidine solid solutions: lack of evidence for Na, K, short-range order and implications for two-feldspar thermometry. American Mineralogist 68, 398–413. Hirschmann, M.M., Kogiso, T., Baker, M.B., Stolper, E.M., 2003. Alkaline magmas generated by partial melting of garnet pyroxenite. Geology 31, 481–484. Holness, M.B., Bunbury, J.M., 2006. Insights into continental rift-related magma chambers: cognate nodules from the Kula Volcanic Province, Western Turkey. Journal of Volcanology and Geothermal Research 153, 241–261. Holness, M.B., Cheadle, M.J., McKenzie, D., 2005. On the use of dihedral angle change to decode late-stage textural evolution in cumulates. Journal of Petrology 46, 1565–1583. Hooper, P.R., 1994. Sources of continental flood basalts: the lithospheric component. In: Subbarao, K.V. (Ed.), Volcanism. Wiley Eastern Ltd., pp. 29–53. Kamel, O.A., Meneisy, M.Y., Abdel Aal, A.Y., 1981. Petrography of Phanerozoic basaltic rocks in Egypt. Bulletin of Faculty of Sciences, vol. 23B. Ain Shams University, pp. 93–124. Kerr, A.C., Kermpton, P.D., Thompson, R.N., 1995. Crustal assimilation during turbulent magma ascent (ATA); new isotopic evidence from the Mull Tertiary lava succession, N.W. Scotland. Contributions to Mineralogy and Petrology 119, 142–154. Lindsley, D.H., 1983. Pyroxene thermometry. American Mineralogist 68, 477–493. Mattioli, M., Serri, G., Salvioli-Mariani, E., Renzulli, A., Holm, P.M., Santi, P., Venturelli, G., 2003. Sub-volcanic infiltration and syneruptive quenching of liquids in cumulate wall-rocks: the example of the gabbroic nodules of Stromboli (Aeolian Islands, Italy). Mineralogy and Petrology 78, 201–230. McGuire, A.V., 1987. Petrology of mantle and crustal inclusions in alkali basalts from western Saudi Arabia: Implications for formation of the Red Sea. Unpublished Ph.D. Thesis, Standford University. McGuire, A.V., Stern, R.J., 1993. Granulite xenoliths from western Saudi Arabia; the lower crust of the late Precambrian Arabian–Nubian shield. Contributions to Mineralogy and Petrology 114, 395–408. McDonough, W.F., 1990. Constraints on the composition of the continental lithospheric mantle. Earth and Planetary Science Letters 101, 1–18. Meneisy, M.Y., 1990. Vulcanicity. In: Said, R. (Ed.), The Geology of Egypt. A.A. Balkema, Rotterdam, pp. 157–174. Meneisy, M.Y., Abdel Aal, A.Y., 1983. Geochronology of Phanerozoic volcanic rocks in Egypt. Bulletin of Faculty of Sciences, vol. 25. Ain Shams University, pp. 163–176. Morimoto, N., 1988. Nomenclature of pyroxenes. Mineralogy and Petrology 39, 55–76. Munha, J., Palacios, T., MacRae, N., Mata, J., 1990. Petrology of ultramafic xenoliths from Madeira Island. Geological Magazine 127, 543–566. Németh, K., White, J.D., Reay, A., Martin, U., 2003. Compositional variation during monogenetic volcano growth and its implications for magma supply to continental volcanic fields. Journal of the Geological Society (London) 160, 523–530. Neumann, E.R., 1974. The distribution of Mn and Fe between ilmenite and magnetites in igneous rocks. American Journal of Sciences 274, 1074–1088. Nimis, P., 1999. Clinopyroxene geobarometery of magmatic rocks. Part 2 — structural geobarometers for basic to acidic, tholeiitic and mildly alkaline magmatic systems. Contributions to Mineralogy and Petrology 135, 62–74.

528

E.S. Farahat et al. / Gondwana Research 11 (2007) 516–528

Nimis, P., 2000. CpxBar-Excel version program (http://dmp.unipd.it/Nimis/ researche.html). Parkhomenko, V.S., 1980. Spektromtricheskiye metody analiza v geokhimii (spectrometric methods of analysis in geochemistry). Novosibirsk 18–30. Pearce, J.A., 1983. Role of subcontinental lithosphere in magma genesis at active continental margins. In: Hawkesworth, C.J., Norry, M.J. (Eds.), Continental basalts and mantle xenoliths. Shiva, Nontwish, pp. 230–249. Presnall, D.C., Dixon, S.A., Dixon, J.R., O;Donnel, T.H., Brenner, N.L., Schrock, R.L., Dycus, D.W., 1978. Liquidus phase relations on the join diopside–forsterite–anorthite from 1 atom to 20 kbar: their bearing on the generation and crystallization of basaltic magma. Contributions to Mineralogy and Petrology 66, 203–220. Preston, R.J., Bell, B.R., 1997. Cognate gabbroic xenoliths from a tholeiitic subvolcanic sill complex: implications for fractional crystallization and crustal contamination processes. Mineralogical Magazine 61, 329–349. Prestvik, T., Goles, P., 1985. Comments on the petrogenesis and tectonic setting of Columbia River Basalts. Earth and Planetary Science Letters 72, 65–73. Reiners, P.W., Nelson, B.K., Ghiorso, M.S., 1995. Assimilation of felsic crust by basaltic magma: thermal limits and extents of crustal contamination of mantle-derived magmas. Geology 23, 563–566. Shelley, D., 1993. Igneous and Metamorphic Rocks under the Microscope. Chapman and Hall, London. Spencer, J.K., Lindsley, D.H., 1981. A solution model for coexisting iron– titanium oxides. American Mineralogist 66, 1189–1201. Sun, S.S., McDonough, W.F., 1989. Chemical and isotopic systematic of ocean basalts: implication for mantle composition and processes. In: Saunders, A.D.,

Norry, M.J. (Eds.), Magmatism in the Ocean Basins, vol. 42. Geolical Society of London, pp. 313–345. Tait, S.R., 1988. Samples form the crystallising boundary layer of a zoned magma chamber. Contributions to Mineralogy and Petrology 100, 470–483. Thompson, R.N., Morrison, M.A., Dickin, A.P., Gibson, I.L., Harmon, R.S., 1986. Two contrasting styles of interaction between basic magmas and continental crust in the British Tertiary Volcanic Province. Journal of Geophysical Research 91, 5985–5997. Tsuchiyama, A., 1986. Melting and dissolution kinetics: application to partial melting and dissolution of xenoliths. Journal of Geophysical Research 91B, 9395–9406. Turner, S., Foden, J., George, R., Evans, P., Varne, R., Elburg, M., Jenner, J., 2003. Rates and processes of potassic magma evolution beneath Sangeang Api volcano, East Sunda arc, Indonesia. Journal of Petrology 44, 491–515. Wagner, C., Mokhtari, A., Deloule, E., Chabaux, F., 2003. Carbonatite and alkaline magmatism in Taourirt (Morocco): petrological, geochemical and Sr–Nd isotope characteristics. Journal of Petrology 44, 937–965. Williams, G.A., Small, J.O., 1984. A study of the Oligo–Miocene basalts in the Western Desert. Proceedings of the 7th Petroleum Exploration Seminar, EGPC, Cairo, pp. 252–268. Wilson, M., 1989. Igneous Petrogenesis, A global tectonic approach. Unwin Hyman. Zhou, M.F., Zhao, T.P., Malpas, J., Sun, M., 2000. Crustal contaminated komatiitic basalts in Southern China: products of a Proterozoic mantle plume beneath the Yangtze block. Precambrian Research 103, 175–189.