The role of HIMU metasomatic components in the North African lithospheric mantle: petrological evidence from the Gharyan lherzolite xenoliths, NW Libya

Abstract

The Neogene–Quaternary alkali-basalt–hawaiite lavas of the Gharyan volcanic field (NW Libya) contain mantle xenoliths. These mostly consist of protogranular spinel lherzolites with superimposed metasomatic textures represented by reaction patches where primary orthopyroxene (opx), clinopyroxene (cpx) and spinel (sp) are the main reacting phases. The secondary parageneses include clinopyroxene (cpx2), olivine (ol2) and feldspar (feld) as reaction rims around opx, spongy-textured clinopyroxene with recrystallized portions (cpx2±feldspar), and brown spinel destabilized in a higher Cr/(Cr+Al) black vermicular aggregate (sp2) generally associated with feldspar microlites. Cpx2 are typically depleted in Na2O and Al2O3 relative to cpx; feldspar includes both alkali-feldspar (Or 17–51) and plagioclase (An 23–64). Bulk rocks have flat heavy rare earth element (HREE) patterns (1.2–2.3 times chondrite) and are variably enriched in light REE (LREE; LaN/YbN up to 6.6). The constituent clinopyroxenes are characterized by flat HREE distributions (8–14.5 times chondrite) and variable LREE enrichment with LaN/YbN up to seven, which generally conform to the bulk-rock chemistry. Samples relatively unaffected by metasomatism have clinopyroxene Sr–Nd isotopic composition (87Sr/86Sr down to 0.7023, 143Nd/144Nd up to 0.5139) that approaches the depleted mantle (DM), suggesting that the lithospheric mantle beneath the area underwent a long-term depletion probably by pre-Palaeozoic extraction of basic melts. The remaining samples approach 87Sr/86Sr c. 0.7030, 143Nd/144Nd c. 0.5130, with 206Pb/204Pb up to 19.66. These data imply that the causative agents of metasomatism were Na-alkali silicate melts with a clear HIMU affinity, in accordance with the isotopic signature of the host lavas (87Sr/86Sr=0.7032, 143Nd/144Nd=0.5130, 206Pb/204Pb=19.60). This prevalent HIMU geochemical signature is comparable with that recorded in Cenozoic alkaline basic lavas and associated mantle xenoliths from other occurrences of the northern–central African lithosphere, suggesting a common regional sub-lithospheric component. The relatively low 3He/4He of the Gharyan xenoliths (5.3–6.5 Ra) indicates that this component originates within the upper mantle and is unrelated to the deep-seated mantle plume source of the Ethiopian–Yemen plateau basalts. Therefore, the Cenozoic volcanic districts of the Saharan belt could be related to smaller-scale shallow mantle upwellings (also referred to as ‘hot fingers’) triggered by intraplate reactivation of regional tectonic lineaments within the Pan-African cratonic basement, as a foreland reaction of the African–Europe collisional system.

Miocene to recent alkaline volcanism between Al Haruj and Waw an Namous (southern Libya)

“…HIMU (high l, l = 238 U/ 204 Pb), EM I (enriched mantle 1), EM II (enriched mantle 2), PM (primitive mantle), N-MORB (mid-oceanic ridge basalt) and CC (continental crust) are from Weaver (1991) Zindler and Hart (1986), FOcal ZOne (FOZO) after Hart et al (1992) and Hauri et al (1994). W2, W3 and W5 (this study); crosses: alkaline lavas from Gharyan, Libya (Beccaluva et al 2008). Fields of alkaline (Halliday et al 1990;Nkouathio et al 2008) and transitional (Moundi et al 2007) lavas from the Cameroon Volcanic Line, tholeiitic and alkaline lavas from Hoggar (in Aït-Hamou et al 2000) and Ethiopian flood basalts (Pik et al 1999) are given for comparison.…”

“…WAN samples display the lowest SiO 2 (Fig. 3) and Al 2 O 3 but highest normative Ne contents and Fe 2 O 3 and among the highest Na 2 O, MgO and P 2 O 5 compositions among all analysed Libyan and Tibestian volcanic rocks (Libya, Almond et al 1974;Busrewil and Wadsworth 1980a, b, 1982-1983Woller and Fediuk 1980;Asran and Aboazom 2004;Farahat et al 2006;Beccaluva et al 2008;Cvetković et al 2010;Tibesti, Chad, Gourgaud and Vincent 2004). Note that the less differentiated Gharyan samples (Almond et al 1974) are similar to the foidite analysed here, however (Fig.…”

“…W9, crosses: clinopyroxene from lherzolite in WAN lavas Wadsworth 1982-1983). Compositional fields of clinopyroxene from Gharyan lherzolites (solid outline) (Beccaluva et al 2008), from Hoggar lherzolites (dashed outline) (Beccaluva et al 2007) and from Batoke volcano cumulative peridotites, Cameroon (dotted outline) (Wandji et al 2009). W1: compositional fields of ferro-aluminous diopside from the Noun Plain basanites and alkali basalts, Cameroon (dashed outline) (Wandji et al 2000) and from various localities (dashed-dotted outline) (compiled in Deer et al 1978).…”

“…It contains olivine (Fo 90), orthopyroxene, clinopyroxene and spinel and is likely to be of the same type as those described by Almond et al (1974) in Gharyan lavas and analysed by Busrewil and Wadsworth (1980b). More recently, Beccaluva et al (2008) analysed in detail these lherzolite mantle xenoliths from Gharyan that contain up to 35 vol.% orthopyroxene and up to 20 vol.% clinopyroxene. A primary paragenesis and superimposed metasomatic textures were also distinguished.…”

“…Note that the extreme trace element ratios are systematically encountered in foidite W2 and basalt W5. The available data on Emi Koussi, As Sawda, Al Haruj and Gharyan basalts (Gourgaud and Vincent 2004;Woller and Fediuk 1980;Asran and Aboazom 2004;Farahat et al 2006;Beccaluva et al 2008;Cvetković et al 2010) are compared in Fig. 7.…”