Journal of Petrology Advance Access published online on June 22, 2007
Journal of Petrology, doi:10.1093/petrology/egm028
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Magma Genesis and Mantle Dynamics at the Harrat Ash Shamah Volcanic Field (Southern Syria)
1Institut Für Geowissenschaften Der Universität Kiel, Olshausenstrasse 40, 24118 Kiel, Germany
2Institut Für Mineralogie Der Universität Münster, Corrensstrasse 24, 48149 Münster, Germany
3Faculty of Civil Engineering, University Of Aleppo, PO Box 5427, Aleppo, Syria
Received August 5, 2005; Revised typescript accepted May 9, 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONGEOLOGICAL SETTINGSAMPLE PREPARATION AND...PETROGRAPHIC AND GEOCHEMICAL...DISCUSSIONCONCLUSIONSREFERENCES |
|---|
A geochemical and petrological study of Miocene to recent alkali basalts, basanites, hawaiites, mugearites, trachytes, and phonolites erupted within the Harrat Ash Shamah volcanic field was performed to reconstruct the magmatic evolution of southern Syria. The major element composition of the investigated lavas is mainly controlled by fractional crystallization of olivine, clinopyroxene, ± Fe–Ti oxides and ± apatite; feldspar fractionation is restricted to the most evolved lavas. Na2O and SiO2 variations within uncontaminated, primitive lavas as well as variably fractionated heavy rare earth element ratios suggest a formation by variable degrees of partial melting of different garnet peridotite sources triggered, probably, by changes in mantle temperature. The isotopic range as well as the variable trace element enrichment observed in the lavas imply derivation from both a volatile- and incompatible element-enriched asthenosphere and from a plume component. In addition, some lavas have been affected by crustal contamination. This effect is most prominent in evolved lavas older than 3·5 Ma, which assimilated 30–40% of crustal material. In general, the periodicity of volcanism in conjunction with temporal changes in lava composition and melting regime suggest that the Syrian volcanism was triggered by a pulsing mantle plume located underneath northwestern Arabia.
KEY WORDS: 40Ar/39Ar ages; intraplate volcanism; mantle plume; partial melting; Syria
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGSAMPLE PREPARATION AND...PETROGRAPHIC AND GEOCHEMICAL...DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Continental mafic volcanic rocks exhibit large variations in chemical and isotopic compositions and their geochemical signatures can be used to constrain dynamic processes within the Earth's interior and the continental lithosphere. Magmatism in continental regions often occurs in areas of thinned lithosphere resulting in decompressional melting of the rising asthenosphere; models suggest that the volume of melt produced depends on the stretching factor of the lithosphere as well as on the temperature and composition of the asthenosphere (McKenzie & Bickle, 1988
). Lithospheric thinning may be initiated either by plate boundary forces or by frictional forces exerted on the base of the lithosphere by the convecting asthenosphere (Sengör & Burke, 1978; Ziegler & Cloetingh, 2004). However, White & McKenzie (1989) suggested that lithospheric stretching without any influence of hot mantle (i.e. a mantle plume) cannot account for extensive magma generation within dry mantle peridotite. In their model continental flood basalt volcanism is restricted to the development of extreme lithospheric thinning, coinciding with the arrival of the head of a mantle plume at the base of the lithosphere and an onset of volcanism by passive upwelling of the mantle as the lithosphere thins. In contrast, Campbell & Griffiths (1990) suggested that extensive volcanism results from a starting mantle plume, but that both continental rifting and an ascending mantle plume are not necessarily required to generate large igneous provinces. Extensive melting of the mantle beneath the continents could also be due to the presence of hydrous mantle, probably in the subcontinental lithosphere. The presence of hydrous phases would lower the solidus of the mantle and melting may occur in response either to lithospheric thinning at average mantle temperatures (Gallagher & Hawkesworth, 1992) or to lithospheric melting by conductive heating of the lithosphere above a mantle plume without melting the plume itself (Turner et al., 1996). Thus, subcontinental lithospheric as well as asthenospheric and plume mantle sources may all contribute to magma genesis in intra-continental regions.
Rifting, faulting and volcanism in Arabia are generally linked to the two-stage evolution of the Red Sea and the Dead Sea rift systems in Miocene and Pliocene times. In addition, volcanism in southwestern Arabia has been associated with the activity of the Afar mantle plume (Schilling, 1973; Debayle et al., 2001). Whereas the petrogenesis and magma source regions (i.e. crust, lithospheric mantle, asthenosphere and plume mantle) in southwestern Arabia are relatively well characterized, the causes and sources of volcanism in northern Arabia are less clear and the following models have been proposed to explain this volcanic activity: (1) lithospheric rifting mobilizing old fossil plume head material beneath the subcontinental lithosphere (Stein & Hofmann, 1992); (2) progressive lithospheric thinning tapping lithospheric (fossil plume material) to asthenospheric sources (Bertrand et al., 2003; Shaw et al., 2003); (3) melting of hydrous mantle lithosphere by heat conduction from sub-lithospheric anomalously hot mantle (Stein et al., 1997; Weinstein et al., 2006); (4) a separate mantle plume existing underneath northern Arabia (Camp & Roobol, 1992); (5) northwestward channelling of hot Afar plume material (Camp & Roobol, 1992).
This study investigates the geochemistry and isotopic signature of lavas from the largest Cenozoic intraplate volcanic field of western Arabia, the Harrat Ash Shamah in Syria, together with published data from the same volcanic field in Jordan (Shaw et al., 2003) and Israel (Weinstein et al., 2006) with the following objectives: (1) to determine the processes controlling the major and trace element chemistry as well as the Sr–Nd–Pb isotopic composition of the volcanic rocks; (2) to determine the geochemical characteristics of the magma sources; (3) to constrain the temporal evolution of the lavas and their mantle sources during at least 24 Myr of magmatic activity.
|
GEOLOGICAL SETTING |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGSAMPLE PREPARATION AND...PETROGRAPHIC AND GEOCHEMICAL...DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Voluminous Cenozoic alkali basaltic volcanism occurs in western Arabia, extending from Yemen over a distance of 2500 km to southeastern Turkey (Fig. 1a). These volcanic fields, the so-called Harrats, cover an area of
200 000 km2, with an average thickness of about 100 m (Coleman et al., 1983; Nasir, 1994; Tarawneh et al., 2000). The Harrat Ash Shamah (or Harrat Ash-Shaam: HAS) is the largest volcanic field on the Arabian Plate and lies at the northwestern margin of the Afro-Arabian dome as outlined by Camp & Roobol (1992). The HAS strikes parallel to the Red Sea and extends some 500 km from northeastern Israel through southern Syria and Jordan to Saudi Arabia, covering more than 50 000 km2 (Ilani et al., 2001; Shaw et al., 2003; Weinstein et al., 2006). The volcanic field varies in width between 200 and 300 km (Tarawneh et al., 2000) and the lava pile reaches a maximum thickness of 1·5 km and an altitude of 1800 m in Syria (Guba & Mustafa, 1988). The Dead Sea transform fault intersects the HAS at its northwestern edge in Israel and two small NE–SW-striking late Cretaceous graben structures lie south and north of the HAS, respectively (Almond, 1986).
|
The HAS mainly consists of massive lava flows, which are often separated by sedimentary horizons and red lateritic weathering crusts (Snyder et al., 1993
; Sharkov et al., 1994); however, feeder dykes, strato- and shield volcanoes, as well as cinder cones, also occur within and on top of the lava piles. Cinder cones with a very young appearance also form the youngest eruption type. A large amount of radiometric age data are available for the HAS (Barberi et al., 1980; Giannérini et al., 1988; Mouty et al., 1992; Mor, 1993; Sharkov et al., 1994; Heimann et al., 1996; Ilani et al., 2001; Shaw et al., 2003; Weinstein et al., 2006) indicating that volcanism started about 24 Myr ago and continued to recent times with a volcanic hiatus of several million years between 13 and 7 Ma in the northern part of the HAS (Fig. 2). We present data from the northern part of the HAS field, also known as the Djebel Druze or Hauran–Druze Plateau in Syria (Ponikarov, 1969), which lies on the western flank of an uplifted region (Best et al., 1990; McBride et al., 1990) and can be divided into four major areas ranging in age from Miocene to historical times (Fig. 1c). The oldest flow units are of early Miocene age (24–16 Ma) and cover an area of about 200 km2 in the northwestern part of the HAS plateau (Fig. 1c). This volcanic sequence is composed of 20–22 lava flows and is up to 400 m thick (Snyder et al., 1993). These lavas unconformably overlie Cretaceous and Paleogene sediments and are covered by younger Pliocene to Quaternary lavas (Tarawneh et al., 2000). Mostly Pliocene (7–3·5 Ma) volcanic rocks are concentrated in the SE, whereas the southwestern area, near the Jordan border, is covered by Pleistocene units (Fig. 1c). The youngest lavas of Holocene age form the northeastern part of the HAS. Volcanic cones up to 100 m high are the products of the youngest volcanic activity of each period and lie on the surfaces of the lava plateaux. These cones often form chains trending NW–SE and seem to trace faults that probably provided the conduits for the magmas (Giannérini et al., 1988). Crustal as well as mantle xenoliths occur in several of the young volcanic cones of the HAS (Sharkov et al., 1989; Nasir, 1992).
|
Precambrian crystalline basement of the Arabian–Nubian shield crops out only in southern Jordan and dips to the NE (Jarrar et al., 2003
). The Precambrian basement consists of gabbroic to granitic rocks that are comparable with the crustal xenoliths occurring in the lavas (Nasir, 1992, 1995). The basement in Syria is overlain by predominantly flat-lying Phanerozoic marine and continental sediments (Sharkov et al., 1994). Seismic reflection data, as well as refraction profiling, have shown that the metamorphic basement lies generally deeper than 6 km (Brew et al.. 2001a) and varies in depth beneath the uplifted region in southern Syria between 6 km and >8 km (McBride et al., 1990). Gravity and teleseismic models of the crust in southern Syria suggest a thickness of 40 ± 4 km, similar to crustal thicknesses interpreted from seismic refraction data in Jordan and Saudi Arabia (El-Isa et al., 1987; Best et al., 1990; McBride et al., 1990; Al-Damegh et al., 2005). An S receiver function study revealed a lithospheric thickness of 70–75 km in the area of the HAS (Mohsen et al., 2006), which is in agreement with earlier estimates based on mantle xenoliths (McGuire & Bohannon, 1989). Seismic tomographic results show the presence of mantle material with relatively low seismic velocities at a depth of about 100 km beneath the northern part of the Arabian Plate (Debayle et al., 2001). |
SAMPLE PREPARATION AND ANALYTICAL METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGSAMPLE PREPARATION AND...PETROGRAPHIC AND GEOCHEMICAL...DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The Syrian part of the HAS volcanic field was extensively sampled and lavas from each volcanic period were collected (Fig. 1, Table 1). Hand specimens were sawn into pieces and weathered surfaces and cracks were removed. After washing with deionized water in an ultrasonic bath the samples were crushed with a screw press and washed again. The crushed material was powdered in an agate ball mill for major and trace element and isotope analyses. Four samples contained fresh glass, which was handpicked under a binocular for chemical analysis.
|
The glass samples and mineral phases were analysed using a JEOL Superprobe 8900 electron microprobe at the Institut für Geowissenschaften, Universität Kiel, using standard wavelength-dispersive techniques. The instrument was operated at an accelerating voltage of 15 kV and a beam current of 15 nA. The beam diameter during calibration and measurement was set at 5 µm for glasses and feldspars and at 1 µm for the remaining phases. Counting times on peaks were set at 15 s, whereas background counting times were always set to half of peak counting times. The quality of the data was checked by repeated measurement of a set of mineral standards and the results of the glass analyses are presented in Table 2.
|
Major and some trace element (Cr, Ni, Zn, Rb, Sr, and Zr) analyses of whole-rock samples were performed at the Universität Kiel using a Philips PW 1400 X-ray fluorescence spectrometer, using international rock standards for calibration and data quality control. Sample powders (0·6 g) were mixed with lithium tetraborate (3·6 g) as flux and fused to glass beads. Replicate measurements of the BHVO-1 standard gave a precision better than 0·30% (SD) for all major elements and generally better than 8% (SD) for trace elements. The accuracy of standard analyses relative to reference values is generally better than 4% for most of the major and trace elements, and only Na2O, P2O5 and Ni (< 7%) show higher deviations (Table 2).
Trace element and rare earth element (REE) analyses were carried out by inductively coupled plasma mass spectrometry (ICP-MS) at the Universität Kiel using an upgraded PlasmaQuad PQ1 system and 100 mg of sample material. Sample preparation was performed following the method of Garbe-Schönberg (1993). Comparison of duplicate analyses of sample SY-234 gave a precision generally better than 1% (SD), except for V, Zr (< 2% SD), Sr and Ba (< 4% SD). The accuracy of the data based on the international rock standard BR is better than 9%, except for Cs, Nb (< 10%) and Li (14%; Table 2).
Most of the Sr, Nd and Pb isotopic analyses were performed at the Zentrallaboratorium für Geochronologie in Münster using a VG Sector 54 multicollector mass spectrometer. A few Sr and Nd isotope analyses were made at the Leibniz-Institut für Meereswissenschaften in Kiel on a Finnigan Triton mass spectrometer and a few Pb isotope analyses on a Finnigan MAT 262 RPQ2+. Isotopic analyses were carried out on 100 mg of powdered sample material and standard ion exchange techniques were used to produce clean Sr, Nd and Pb separates. The samples were not leached because only fresh samples were used, which mostly also show incompatible element ratios such as Nb/U, Ce/Pb or Th/U typical of unaltered mafic rocks. Consequently, our isotope data closely resemble those of leached samples from other studies of the HAS lavas (Shaw et al., 2003; Weinstein et al., 2006). Isotope fractionation was corrected using 86Sr/88Sr = 0·1194 and 146Nd/144Nd = 0·7219. In Münster, standard runs for Sr isotopes gave NBS987 = 0·710299 (n = 16; 2SD = 0·000026) whereas in Kiel, NBS987 = 0·710252 (n = 4; 2SD = 0·000006). All Sr isotope analyses were normalized to NBS987 = 0·710250. Standard runs for 143Nd/144Nd in Münster yielded La Jolla = 0·511862 (n = 14; 2SD = 0·000024) compared with standard runs for the Nd Spex standard in Kiel, which gave values of 0·511711 (n = 7; 2SD = 0·000012) corresponding to a La Jolla value of 0·511855. All Nd isotope analyses were corrected to the La Jolla standard measured in Münster. The NBS982 standard was used to correct Pb isotopic ratios for mass fractionation in Münster. The correction factor in this laboratory is 0·1% per a.m.u. and standard runs (n = 10) gave 206Pb/204Pb = 36·646, 207Pb/204Pb = 17·101 and 208Pb/204Pb = 36·593 with a precision of ±0·022, ± 0·014 and ±0·038 (2SD), respectively. In Kiel, repeat measurements of the NBS981 standard yielded correction factors of 0·1% per a.m.u. Standard runs (n = 28) yielded 206Pb/204Pb = 16·903, 207Pb/204Pb = 15·447 and 208Pb/204Pb = 36·558 with a precision of ±0·018, ± 0·024 and ±0·075 (2SD), respectively. Procedural blanks during the course of the analyses were generally better than 0·2 ng, 0·1 ng and 0·04 ng for Sr, Nd and Pb, respectively (Table 3). K–Ar (Giannerini et al., 1988; Mouty et al., 1992; Sharkov et al., 1994) and 40Ar–39Ar ages (this study) as well as the stratigraphic correlation indicate that the lavas chosen for isotope analyses are not older than 4·5 Ma, and thus age corrections of the measured isotope ratios are insignificant and were not performed.
|
Based on thin-section investigation and on K contents two samples with unaltered groundmass were chosen and prepared for Ar–Ar dating. For analyses the crushed sample material was sieved and the 250–500 µm fraction was used to separate 50 mg of groundmass by picking under a binocular microscope. The samples were irradiated using the FRG2 reactor at the GKSS Research Centre (Geesthacht) for 7 days (Cd-liner). Irradiated samples were reloaded into aluminium trays with multiple pans holding between 4·5 and 7 mg of matrix material. The samples were heated by incrementally increasing laser power output from 0·15 to 20 W from < 350°C to >1800°C until complete fusion. Software was used to control the scanning across the samples with preset patterns and a defocused laser beam, to more evenly heat the material. Automated step-heating was followed by final fusion at
25 W with a focused beam to ensure complete degassing, and essentially evaporating the residual melt or glass sphere.
Incremental heating 40Ar/39Ar age determinations were carried out on microcrystalline groundmass separates using an Ar-ion laser combined with a MAP-216 mass spectrometer at the Leibniz-Institut für Meereswissenschaften in Kiel, Germany. Ion beam currents were measured with the electron multiplier at m/z = 36–40 and half-mass baselines with a 7·5 digit integrating HP multimeter. Peak heights were regressed to inlet time. The peak decay typically was less than 10% during the analyses. Average extraction line blanks are determined as 2 x 1017 mol at m/z = 36 and 4 x 1016 mol at m/z = 40. Mass discrimination was monitored using air-fused zero age basaltic glass and pipette air samples (1· 0083 ± 0·0006 a.m.u.). Correction factors for interfering neutron reactions on Ca and K were determined from co-irradiated CaF2 and K2SO4 salt crystals (36Ca/39Ca = 0·445 ± 0·005, 37Ca/39Ca = 1006 ± 7, 38K/39K = 0·0168, 40K/39K = 0·004 ± 0·002). 40Ar/39Ar ages were measured relative to the flux monitor standard TCR sanidine (27·92 Ma; Duffield & Dalrymple, 1990) and uncertainties for the J-values were estimated as ±0·08% (1
). Incremental heating plateau ages were calculated as weighted mean (Taylor, 1982) and initial 40Ar/36Ar isotope ratios and isochron ages as least-squares fits with correlated errors (York, 1969) applying the decay constants of Steiger & Jäger (1977). The results are compiled in Table 4.
|
|
PETROGRAPHIC AND GEOCHEMICAL CHARACTERISTICS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGSAMPLE PREPARATION AND...PETROGRAPHIC AND GEOCHEMICAL...DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Petrography
Hand specimens of the investigated volcanic rocks are generally fresh, with only a few samples showing evidence of alteration. Almost all samples have porphyritic textures and contain variable amounts of phenocrysts, with olivine and clinopyroxene being the most abundant and feldspars rare. Most olivines show iddingsite rims and some are completely altered. The predominant feldspar is plagioclase but in the trachytic and phonolitic lavas mainly alkali feldspars occur (Table 1). The mostly fine-grained groundmass consists of olivine, clinopyroxene, plagioclase, and Fe–Ti oxides. Rare calcite-filled vesicles are visible in hand specimen. Some lava samples contain agglomerates of clinopyroxene and olivine phenocrysts. Sample SY-272 contains a mantle xenolith
8 mm in diameter, composed of mainly olivine and orthopyroxene with strongly resorbed rims, which is classified as wehrlite.
Classification
In the total alkali vs silica diagram the HAS samples are mainly of basaltic and basanitic–tephritic composition, whereas evolved lavas are rare (Fig. 3). All basaltic rocks fall into the alkaline field using the division line of MacDonald (1968) and, except for four samples, all basalts are nepheline-normative and thus can be classified as alkali basalts. The lavas from the northeastern Syrian part of the HAS resemble those found in the south (Jordan, Shaw et al., 2003) and western edge (Israel, Weinstein et al., 2006) of the HAS (Fig. 3). To study the chemical evolution of the HAS magmatism we grouped the entire dataset into three series based on their relative stratigraphic age, on published K–Ar ages (Giannérini et al., 1988; Mouty et al., 1992; Sharkov et al., 1994) and on 40Ar–39Ar ages (this study; Figs 1c and 3). Series I comprises lavas younger than 1 Ma, volcanic rocks of series II erupted between 1 and 3·5 Ma, whereas lavas of series III generally have ages between 3·5 and 5·5 Ma (Figs 1c and 2). The lavas studied here thus have similar ages to those from the western part of the HAS investigated by Weinstein et al. (2006), whereas many samples from the southern part of the HAS have significantly older ages (Ilani et al., 2001; Shaw et al., 2003). The three lava series that have been mapped also show different chemical compositions. Generally, the youngest series I lavas are alkali basalts and most series II lavas are basanitic and tephritic whereas series III ranges from mainly alkali basalts to basanites, including rare evolved lavas of trachytic and phonolitic composition (Fig. 3).
|
Geochemical composition
Major and trace element abundances
The lavas span a range of MgO concentrations between 11· 4 and 0·2 wt % with Mg-numbers varying from 0·67 to 0·09; however, evolved lavas occur only in series III. The HAS basalts show large variations in major element compositions (Fig. 4), but within each series the relatively primitive lavas show slight increases in SiO2, Na2O and K2O concentrations with decreasing MgO. These elements strongly increase in lavas with MgO contents < 5 wt %. In addition, whereas TiO2 and CaO and also P2O5 concentrations are relatively constant or increase slightly between 12 and 5 wt % MgO they sharply decrease for MgO concentrations below 5 wt %. Al2O3 contents are similar in rocks from different lava series at a given MgO and increase throughout the whole differentiation sequence. The series II lavas with MgO >6 wt % generally have lower SiO2 and higher TiO2, Na2O and K2O as well as high P2O5 contents compared with the other HAS lava series (Fig. 4). However, series III is heterogeneous, with two primitive samples resembling the basanitic–tephritic series II. Series I basalts with more than 8 wt % MgO have slightly lower SiO2 contents than the most primitive series III lavas, and the most primitive lavas of series I have higher TiO2 contents at a given MgO concentration. These chemical differences hint at more than one petrogenetic process controlling the range of variations and indicate that the petrogenesis of the Syrian HAS lavas changed with time.
|
Compatible trace elements such as Ni and Cr are positively correlated with MgO (Fig. 5a and b) and series III lavas with more than 7 wt % MgO have higher Cr concentrations than primitive lavas from the other groups (Fig. 5b). Slightly negative correlations between Sc and MgO can be observed in series I and II lavas, whereas lavas of series III show no correlation (Fig. 5c). Sr correlates negatively with MgO and most of the series II lavas have higher Sr contents than the lavas of series I and II (Fig. 5d), which is in agreement with their more alkaline composition (Figs 3 and 4).
|
In Fig. 6a–c primitive mantle-normalized trace element patterns of the Syrian HAS series are shown. The lavas of all series are enriched in highly incompatible elements and the basanites generally are more enriched than the alkali basalts and hawaiites. The patterns of the three lava series resemble each other and, with the exception of the most evolved rocks, all the lavas show prominent positive anomalies in Ba and Sr, and negative anomalies in Th, U, K and Y. The trachyte patterns have large negative anomalies for Sr, P and Ti, and positive anomalies for Pb, Zr and Hf (Fig. 6c). Slight positive Pb anomalies are also observed for some series III basalts. The HAS samples exhibit a rough positive correlation between Ce/Pb and Nb/U ratios (Fig. 7a), which have been shown to be different and relatively constant for rocks formed from the Earth's mantle and continental crust (Hofmann et al., 1986
). Although a few lavas tend to lower values resembling average continental crust, most samples have Ce/Pb similar to the range suggested for oceanic basalts (Hofmann et al., 1986). In a plot of Th/U vs Rb/Cs most of the HAS lavas fall well within the range of mantle-derived melts, although one sample from series I has a distinct low Th/U ratio (Fig. 7b). The K/La ratios of the Syrian volcanic rocks show a rough negative correlation with (La/Sm)N and the most enriched basalts from series II and III have the lowest K/La (Fig. 7c). Whereas the K/La ratios of series II and III lavas are highly variable, series I samples show a much more restricted range. The trachytes lie outside the basalt trend at high K/La and (La/Sm)N (Fig. 7c). The HAS lavas show a negative correlation between Ce/Pb and K/La similar to samples from Jordan and Israel (Fig. 7d).
|
|
Isotopic compositions
The Syrian series II and III HAS volcanic rocks show a negative trend between Sr and Nd isotope ratios whereas the series I samples have lower 143Nd/144Nd than series III lavas at a 87Sr/86Sr of 0·7033 (Fig. 8a). The series II and III lavas show the same trend of Sr and Nd isotope ratios as the lavas from the Jordanian part of the HAS (Bertrand et al., 2003
; Shaw et al., 2003). In contrast, the series I lavas resemble the lavas from the western edge of the HAS in Israel (Weinstein et al., 2006) and two lavas previously analysed (Bertrand et al., 2003). On a larger scale, the Syrian HAS lavas also mostly overlap with the field of Saudi Arabian lavas, which are highly variable (Fig. 8b). All Syrian HAS basalts have lower 87Sr/86Sr ratios for a given 143Nd/144Nd ratio and higher 206Pb/204Pb than the Afar plume lavas with high 3He/4He (Fig. 8b, g and h). The series II samples are similar to Red Sea mid-ocean ridge basalt (MORB) in terms of their Pb isotope compositions (Fig. 8d and f) and are also relatively MORB-like in Sr and Nd isotope composition (Fig. 8b). In contrast, the lava series I and III show much higher 207Pb/204Pb (Fig. 8d) and 87Sr/86Sr ratios than Red Sea MORB (Fig. 8b and g).
|
Most Syrian HAS lavas lie on a positive trend of 87Sr/86Sr vs Ba/Nb with the series II samples lying at the lower end and two series III samples at the end with high Ba/Nb and Sr isotope ratios (Fig. 9a). Two series III samples have much higher 87Sr/86Sr at intermediate Ba/Nb. In Fig. 9b the Nd isotopic compositions of the HAS lavas are plotted against (La/Sm)N; the data define a rough positive correlation in which series I lavas lie at the low end and series II lavas at the end with high (La/Sm)N and 143Nd/144Nd. Compared with Red Sea lavas and with Afar plume related lavas the investigated samples have generally higher (La/Sm)N, but are similar to the Jordanian HAS lavas.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGSAMPLE PREPARATION AND...PETROGRAPHIC AND GEOCHEMICAL...DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Effects of alteration
During weathering of rocks elements such as K, Rb, Cs, U or Ba can be mobilized by aqueous fluids, leading to both (1) either the addition or subtraction of these elements from the rock and (2) increasing H2O and CO2 contents as expressed by elevated loss on ignition (LOI) values. However, the HAS lavas are usually fresh and show only alteration of olivine to iddingsite and a few calcite amygdules. Except for two samples with high LOI values (2·7 and 3·7 wt %) the LOI content generally lies below 2 wt %, which is typical for alkaline lavas (Table 2). In addition, most of the lavas have Nb/U ratios similar to or lower than those of oceanic basalts (Fig. 7a) and show Th/U and Rb/Cs ratios within the range representative for mantle-derived magmas (Fig. 7b), which indicates that they have not been significantly affected by alteration. Thus, alteration processes appear to have had only a minor effect, even on the mobile element contents of the HAS lavas.
Control of fractional crystallization and crystal accumulation on HAS lavas
The abundance of phenocryst phases, as well as the wide range of MgO contents and Mg-numbers of the samples, implies crystal fractionation during magma ascent (Figs 4 and 5). However, the large variations in the concentrations of incompatible elements such as K (Fig. 4g) and in incompatible element ratios such as (La/Sm)N (Fig. 7c) imply that the lavas within each series (except series I) are not cogenetic. Instead, the magmas of series II and III must have formed from different source compositions, variable degrees of partial melting, or crustal contamination processes. Throughout the differentiation sequence the volcanic series are influenced by crystallization of olivine, clinopyroxene, spinel and, at later stages, apatite, whereas minor feldspar fractionation is restricted to the most evolved lavas. Olivine is one of the major fractionating phases and its extraction is indicated by decreasing Ni and MgO contents (Fig. 5a). In contrast, the most primitive sample SY-272, with a Mg-number of 0·67, has a high Ni content (387 ppm) and a high Ni/Cr ratio (1· 07), but shows comparable trace and major element concentrations to the other lavas (Table 2), suggesting olivine accumulation. This assumption is also confirmed by the occurrence of olivine xenocrysts in thin section and the observed mantle xenolith with strongly resorbed olivines and orthopyroxenes may represent a dissolved peridotite (wehrlite). Decreasing Cr contents at a MgO concentration of about 9 wt % indicate spinel fractionation (Fig. 5b). The negative correlation between Sc and MgO contents and the relatively constant CaO/Al2O3 (not shown) in the lavas with MgO >5 wt % indicate that clinopyroxene fractionation plays no major role in the differentiation of series I and II lavas, but appears to be more prominent in series III lavas as the Sc concentrations and CaO/Al2O3 decrease with decreasing MgO (Fig. 5c). The increasing Al2O3 and Sr concentrations with decreasing MgO as well as the lack of Eu anomalies (Fig. 6) imply that plagioclase fractionation did not play a role in the early stages of differentiation. Between MgO concentrations of 12 and 5 wt % olivine and spinel fractionation clearly dominate over clinopyroxene removal and the inflection of CaO at a MgO of 5 wt % probably indicates a significant increase of the relative proportion of clinopyroxene fractionation as compared with olivine and spinel removal. Concerning the more evolved rocks of series III, the crystallization of titanomagnetite is responsible for the lower FeOT and TiO2 contents in samples with MgO below 5 wt %. The P2O5 decrease starting at 5 wt % MgO is caused by apatite fractionation. Only the most evolved rocks of series III show some evidence for feldspar removal, as indicated by their comparatively low Al2O3 and Sr concentrations and their pronounced negative Eu anomalies (Fig. 6c).
Crustal contamination of the HAS magmas
Magmas rising through the continental lithosphere often stagnate at crustal levels and assimilate crustal material. Mid-ocean ridge and ocean island basalts have average Nb/U ratios of 47 ± 10 and average Ce/Pb ratios of 25 ± 5, reflecting the composition of the Earth's mantle, whereas the continental crust has characteristically low Nb/U and Ce/Pb ratios (Hofmann et al., 1986). The high mantle-like Nb/U and Ce/Pb values of most HAS lavas imply negligible crustal contamination (Fig. 7a). However, several series III lavas have Nb/U and Ce/Pb ratios lower than the range typical for oceanic lavas, which suggest the involvement of crustal contamination (Fig. 7a). The trachytes represent the most evolved lavas and have the lowest Nb/U and Ce/Pb ratios, which imply that assimilation and crystal fractionation plays a major role in their petrogenesis. The HAS lavas from this study as well as those from the parts in Jordan and Israel lie on a negative trend of K/La vs Ce/Pb; the rocks with low Ce/Pb resemble the compositions of some of the Precambrian basement rocks (Fig. 7d). Bulk mixing calculations between the most primitive Syrian samples and the Precambrian basement rocks reveal that contamination of the magmas with about 30–40% of an upper crustal component can account for low Ce/Pb coupled with high K/La ratios as observed in the most evolved rocks of series III (Fig. 7d).
Because isotopic compositions are not influenced by crystal fractionation or the degree of partial melting, the Sr and Nd isotopic array of the HAS lavas implies that at least three isotopically distinct end-members are involved in their petrogenesis (Fig. 8a and b). The lavas (alkali basalts) with the highest 87Sr/86Sr have crust-like Nb/U and Ce/Pb and trend towards the 207Pb/204Pb and 208Pb/204Pb compositions of crustal rocks from Yemen (Fig. 8d and f). Thus, some of the Syrian series III lavas appear to have assimilated crustal material leading to high 87Sr/86Sr, 206Pb/204Pb and 207Pb/204Pb isotope ratios and to low 208Pb/204Pb similar to compositions observed in HAS lavas from Jordan (Shaw et al., 2003) (Fig. 8). Shaw et al. (2003) postulated that the lavas of the Jordan part of HAS have assimilated up to 20% of Late Proterozoic crust during their ascent to the surface; these have low 208Pb/204Pb ratios for a given 206Pb/204Pb when compared with the Syrian lavas (Fig. 8f). On the basis of Sr and Nd isotope compositions, Weinstein et al. (2006) suggested that a crustal contribution to the Israeli magmas is small and probably restricted to a lower crustal component. However, the lavas from the northwestern part of the HAS in Israel (Weinstein et al., 2006) also lie on the negative trend between K/La vs Ce/Pb, with one sample having a Ce/Pb value of 8·5, and thus assimilation of crustal material appears likely (Fig. 7d). We conclude that crustal contamination is an important process affecting the geochemical and isotope compositions of many lavas from the HAS volcanic field, most clearly those from the northern and eastern part of the basaltic plateau (i.e. in Jordan and Syria).
Magma generation in the mantle beneath the HAS volcanic field
Primitive magma compositions and partial melting processes beneath the HAS
Crystal fractionation and crustal assimilation processes have modified the Syrian lavas and, thus, no primary magmas from the mantle have been sampled. To compare the differences in magma genesis we have corrected the effects of fractional crystallization on the major elements SiO2, Na2O and TiO2 to MgO contents of 9 wt % following the approach of Klein & Langmuir (1987) and using the regression lines shown in Fig. 4. Not included in the fractionation correction were crustally contaminated samples; that is, those lavas falling outside the range of mantle compositions in terms of Nb/U and Ce/Pb (Fig. 7a) and lavas with MgO concentrations below 5 wt %, because of their complex fractionation history. We find that lava series II has the highest fractionation-corrected Na2O (Na9·0) and Ti9·0 but the lowest Si9·0 and a negative correlation between Si9·0 and (Dy/Yb)N (Fig. 10). The differences in fractionation-corrected major element and heavy REE (HREE) composition imply that the series II lavas formed by lower degrees of melting than magmas of the other two series, probably also at higher pressures; that is, they may represent melts from the deeper part of the melting column. Depending on the mantle composition, the degree of melting can be determined, and is about 4–6% for series II magmas and 6–12% for the series I and III lavas, with the oldest series III magmas representing the highest degree melts if we assume an enriched garnet peridotite source relative to a primitive mantle composition (Fig. 10a). Similar results can be obtained when the degree of melting is calculated on the basis of REE contents in crustally uncontaminated primitive Syrian lavas. As shown in Fig. 11, the high (La/Sm)N ratios and incompatible element concentrations of the HAS samples cannot be modelled with a source of primitive mantle composition but require melting of an enriched mantle. Because the relatively high Nd isotope ratios of the most primitive magmas suggest that their mantle sources had been depleted for a long period of time (Fig. 8a) we used in our model a re-enriched source composition representing 15% of a 2·5% melt from a primitive mantle that was previously depleted by 0·5% melting. In agreement with Shaw et al. (2003) we find no evidence for melting solely in the spinel stability field because melting of neither a primitive mantle composition nor an enriched mantle source can reproduce the high REE ratios of the Syrian lavas (Fig. 11). However, our model does not require mixing of melts from garnet peridotite with melts from spinel peridotite and instead we suggest that melting occurred only in garnet peridotite facies mantle. This is supported by (1) the nearly constant Yb concentrations indicating buffering by garnet, (2) the very high (Sm/Yb)N and (Dy/Yb)N ratios, and (3) the thickness of the lithosphere of 75–80 km, which is close to the spinel-out reaction boundary for relatively fertile mantle (Fig. 12). We agree with Shaw et al. (2003) that the linear data trends in Fig. 11 could reflect binary mixing, which is also indicated by the radiogenic isotopes (Fig. 8); the two melts probably form by different degrees of melting from two different sources within the garnet stability field. In agreement with our model of the Na and Ti contents, the range of REE ratios described by the uncontaminated Syrian HAS lavas indicates that the most primitive lavas of series II were formed by about 3–6% melting of an enriched garnet lherzolite, whereas most of the primitive lavas of series I and III were generated by larger degrees of partial melting of about 8–12% (Fig. 11).
|
None the less, differences in the degree of melting at different depths can account for variable incompatible element concentrations in both the primitive and fractionation-corrected samples (e.g. Ti, P; Figs 4 and 10). Melting experiments on natural dry lherzolites have shown that the SiO2 contents of primary tholeiitic to alkali basaltic melts mainly reflect the pressure of melt formation; increasing pressures lead to decreasing SiO2 in the melt (Hirose & Kushiro, 1993
). Applying the algorithm developed by Haase (1996) on the basis of experiments on dry peridotite to estimate the average pressures of melting of the primary HAS magmas, as achieved by addition of olivine to primitive magma compositions, suggests that series I and III could be derived from similar depths of about 100 km (Fig. 12). Magma generation at this depth is in accordance with the low seismic velocities observed at about 100 km depth beneath Syria (Debayle et al., 2001; Daradich et al., 2003) and with the seismically and petrologically determined thickness of the Arabian lithosphere (McGuire & Bohannon, 1989; Nasir & Safarjalani, 2000).
Applying the algorithm for pressure estimations to primary series II magmas, melt formation would have taken place at depths of about 140 km (Fig. 12). These abrupt changes in melting depths (from shallow to deep to shallow) over a relative short time span of about 2·5 Myr (i.e. during the eruption of the three HAS series) cannot easily be explained. However, the algorithm developed by Haase (1996) takes into account only those experiments that were carried out under dry conditions (i.e. no volatiles or hydrous minerals were present), and thus the use of the algorithm is not appropriate when melting is assumed to be wet, and would lead to an overestimation of the melting pressure. In this context it has been shown that highly undersaturated magmas with SiO2 <45 wt % similar to series II lavas probably form in the presence of CO2 (e.g. Eggler, 1978; Wyllie, 1978), requiring relatively high CO2 contents in the mantle source of the series II magmas. Hirose (1997), for example, concluded on the basis of diamond aggregate melting experiments that nephelinitic to basanitic magmas form at 3 GPa and >1475°C in the presence of CO2 rather than being the product of high-temperature melting above 3 GPa. Thus, experimental constraints suggest that the basanites of series II could form from a volatile-rich garnet peridotite mantle source at about 100 km depth (Fig. 12), compatible with the strongly fractionated REE patterns observed in the series II magmas (Fig. 11).
|
McKenzie & Bickle (1988
) demonstrated that decompressional melting of the asthenosphere with a potential temperature of 1280°C requires a stretching factor of the lithosphere in the range of 2–5. However, no significant extensional tectonic features (e.g. the formation of graben structures and significant thinning of the lithosphere) can be related to the volcanism at this time and tectonic movements are limited to the Dead Sea fault system (e.g. Heimann & Ron, 1993; Brew et al.. 2001b). Melting of dry mantle peridotite at 90–100 km depth requires a potential mantle temperature increased by about 150°C (Fig. 12). Following the empirical algorithm of Albarède (1992), the temperatures of the primary magmas at about 100 km depth are calculated to be 1457°C, 1504°C and 1429°C for series I, II and III, respectively (Fig. 12). These temperatures are 50–150°C higher than the average potential mantle temperature of 1300°C (McKenzie & Bickle, 1988), and fall into the range of excess temperatures thought to be characteristic for the presence of mantle plumes (e.g. White & McKenzie, 1989). An increased temperature of the upper mantle beneath Syria is likely because (1) seismic tomography shows lower seismic velocities (Debayle et al., 2001; Daradich et al., 2003), (2) mantle xenoliths suggest an increased geothermal gradient in the mantle beneath the Arabian volcanic fields (Camp & Roobol, 1992; Snyder et al., 1993), and (3) the region beneath the HAS volcanic field is uplifted by 2000 m (Brew et al., 2000, 2001a). In agreement with previous workers (Shaw et al., 2003; Weinstein et al., 2006) we conclude that the voluminous melting within the HAS province is a consequence of elevated mantle temperatures beneath this part of the Arabian Plate. The melting conditions of the HAS lavas are consistent with the formation of the silica-undersaturated lavas of series II, with ages between 3·5 and 1 Ma, by lower degrees of melting of a volatile-rich mantle source, and with the derivation of the youngest lavas of series I, as well as the lavas older than 3·5 Ma of series III, by relatively large degrees of melting of mantle material with a significant excess temperature (i.e. from a mantle plume). The process of mantle upwelling is probably young and started in the Miocene, at a time similar to or somewhat before the onset of volcanic activity, because the Paleogene sediments are unconformably overlain by the HAS lavas.
|
Mantle sources of the Syrian HAS magmas
Shallow melting of the lithospheric mantle?
Variable depths of partial melting were also suggested by Shaw et al. (2003
) and Weinstein et al. (2006), who proposed that the early alkali basaltic magmas formed by relatively shallow melting of the lithospheric mantle, whereas the later highly undersaturated magmas may have formed deeper in the asthenosphere. Interestingly, those workers did not find evidence for a late shallow and higher degree melting phase comparable with the series I lavas. The presence of these lavas in the Syrian part of the HAS is difficult to explain by a transition from lithospheric to asthenospheric melting caused by progressive lithospheric thinning, because melting processes in this late stage resemble those for the earliest lavas. Furthermore, Weinstein et al. (2006) attributed the low Ba/Nb and Ba/Sr in the HAS lavas relative to primitive mantle to the presence of residual amphibole in the mantle source; this would be important evidence for lithospheric melting because at pressures below 3 GPa amphibole is stable only at temperatures below 1200°C (Green, 1973; Wallace & Green, 1988). However, Ba/Nb correlates with Sr isotope composition (Fig. 9a) and the crustally uncontaminated series I and III alkali basalts (with high Ce/Pb) have K/La of 100–300, similar to the series II basanites, which in turn closely resemble alkali basalts from islands in the southern Red Sea (Fig. 7d). The Red Sea alkali basalts occur in a region of significant lithospheric thinning (estimated thickness of 30 km) and probably represent melts from a mixture of asthenospheric and Afar plume sources (Rogers, 1993). We find no evidence for the presence of residual amphibole in the source of the HAS magmas and thus no evidence for lithospheric melting. Instead, we suggest that the low Ba/Nb and K/La of the uncontaminated lavas reflects the composition of the mantle source of the HAS magmas. The compositional resemblance of the southern Red Sea alkali basalts and the HAS basanites suggests that they have a common source, which is widely distributed in the upper mantle beneath the Arabian Plate. This source probably occurs in the low-velocity zone outlined by geophysical data (Debayle et al., 2001; Daradich et al., 2003).
The mantle sources of the HAS lavas
Although the HAS lavas have a large range of Sr–Nd–Pb isotopic compositions and high incompatible element ratios, it is possible to define two distinct mantle end-members and one crustal contaminant (Figs 7–9
). The crustal contaminant, discussed above, has high Sr and Pb isotope ratios together with low Ce/Pb and probably represents local Proterozoic upper crustal rocks. One mantle end-member is sampled preferentially by series II lavas and has a high 143Nd/144Nd of 0·5130, 87Sr/86Sr of 0·7031 and relatively unradiogenic Pb isotope ratios, resembling the source of the Red Sea alkali basalts (Fig. 8). This mantle source probably represents the ambient asthenosphere, as it has low highly incompatible trace element ratios (e.g. Ba/Nb 7) similar to MORB (Fig. 9a) and appears to produce only small-degree melts (i.e. it appears to be relatively cold). However, the mantle source of series II lavas is more enriched relative to the Red Sea MORB source in incompatible elements; for example, it has slightly higher Sr isotope ratios implying a higher Rb/Sr ratio compared with the Red Sea MORB source. Thus, the series II magmas probably formed from relatively enriched zones within the heterogeneous asthenosphere, which melt preferentially and dominate the small-degree melts such as the HAS series II and Red Sea islands alkali basalts, but which are more diluted in the Red Sea tholeiitic basalts. Because the Pb isotopic ratios of the HAS series II lavas are similar to those of Red Sea MORB this enrichment must have occurred relatively recently.
The second mantle end-member, most clearly reflected in the series I lavas, has a low 143Nd/144Nd ( 0·51284) and a Sr isotope ratio of 0·7033, but variable Pb isotope ratios with 206Pb/204Pb ranging from about 18·8 to 19·1. The crustally uncontaminated series III lavas resemble the series I samples in terms of Sr and Pb isotopes but have slightly higher 143Nd/144Nd, which probably indicates that they represent mixtures between the series I and II magmas. The series I and III magmas formed by larger degrees of melting than the series II melts and thus from relatively hot mantle material (i.e. from a mantle plume source). This source is more enriched in Ba and K than the series II source because it has Ba/Nb >8 and K/La 300 (Figs 7 and 9).
The Arabian part of the Afro-Arabian dome (Camp & Roobol, 1992) (i.e. the Arabian swell) is a large area of elevated topography, extending from Yemen in the south along western Arabia to Syria in the north, which may have formed by northward flow of material from the Afar plume (Camp & Roobol, 1992); this is consistent with: (1) the low shear-wave velocities in the mantle beneath western and northern Arabia (Debayle et al., 2001); (2) the lack of agreement between mantle and near-surface temperatures, which is taken to be evidence for upwelling mantle beneath western Arabia (McGuire & Bohannon, 1989); (3) the initial ages of volcanism in western Arabia, indicating a northward propagation of the volcanic activity (Ershov & Nikishin, 2004). One component within the Afar mantle plume is reflected in Ethiopian lavas with high 3He/4He ratios (Marty et al., 1996); however, this component has a much higher 87Sr/86Sr for a given 143Nd/144Nd and 206Pb/204Pb than the HAS lavas (Fig. 8b and g). Furthermore, the Ethiopian lavas have lower 206Pb/204Pb and 207Pb/204Pb compared with the HAS lavas. Thus, the Afar plume source with high 3He/4He does not appear to play a noticeable role in Syrian HAS volcanism; this is in agreement with the conclusions of Shaw et al. (2003). However, another mantle source with more radiogenic Pb isotopic compositions is represented in lavas from Afar, southwestern Arabia and from the oceanic crust in the Gulf of Aden (e.g. Schilling et al., 1992; Baker et al., 1996; Stewart & Rogers, 1996) (Fig. 8). The presence of this component beneath the HAS can explain the generation of the series I and III magmas; this component has been suggested to reside either in the Afar plume (Vidal et al., 1991; Deniel et al., 1994; Bertrand et al., 2003) or to represent part of a plume head (Kieffer et al., 2004) or to reflect the composition of the subcontinental lithosphere (Pik et al., 1999; Bertrand et al., 2003). However, in our opinion, the fact that this component also occurs in lavas erupting through oceanic lithosphere in the Gulf of Aden (Schilling et al., 1992) rules out a continental lithospheric origin and we propose that this component must have a relatively high excess temperatures to explain the increased degree of melting at great pressure that produced the series III and I magmas. The repeated tapping of this source, before and after the eruption of the series II magmas, is difficult to explain by models involving lithospheric melting (Stein et al., 1997; Bertrand et al., 2003; Shaw et al., 2003; Weinstein et al., 2006), which typically yields the early magmas in continental rift regimes (e.g. DePaolo & Daley, 2000). Finally, the presence of relatively hot mantle material beneath large parts of Arabia is supported by the available geophysical data indicating a relatively low-velocity linear zone in the form of a pipe or channel at depths of 100–200 km extending from Ethiopia and Yemen through Saudi Arabia to Syria in the north (e.g. Debayle et al., 2001; Daradich et al., 2003; Ershov & Nikishin, 2004). Additionally, seismic tomography data do not support the presence of a separate mantle plume under northern Arabia as proposed by Camp & Roobol (1992). We suggest that this material is probably the mantle source component with high Pb isotope ratios from the Afar mantle plume.
Temporal variations of the HAS magmas
The variable fractionation-corrected major element compositions and REE ratios of the three HAS series, together with the ages of the lavas, suggest that mantle melting processes must have varied considerably during the last 20 Myr underneath the HAS region (Figs 10 and 12). The variable degrees of melting observed in the source of the HAS lavas are probably due to a pulsing influx of mantle plume material from the Afar region in the south. Lateral plume flux has been proposed in the ocean basins, and the occurrence of V-shaped ridges along the Reykjanes spreading axis south of Iceland has been interpreted to reflect the pulsing of the Iceland plume (White et al., 1995). Similarly, the inflowing plume material beneath the HAS may have been pulsing during the last 7 Myr with a decreasing influx between 3·5 and 1 Ma, when mainly asthenospheric material was the dominant source accounting for lower degrees of partial melting. The first period of volcanic activity (oldest series III lavas) occurred between 24 and 3·5 Ma. There was a major break in plume influx between 13 and 7 Ma, when no volcanic activity occurred in the northern HAS (Giannérini et al., 1988; Mouty et al., 1992; Sharkov et al., 1994).
Variation in magma generation conditions over a shorter time period is evident in a stratigraphic section of lava flows sampled along the Wadi Ash-Sham, where lavas belonging to series II and III are exposed (Fig. 1c). These lava flows represent a temporal sequence in which the heights of the lavas above the base correlate with age. The 40Ar/39Ar ages for lavas from the lowermost and uppermost section are 4·2 Ma and 3·3 Ma, respectively (Table 4), suggesting that the about 700 m thick pile of lava erupted within about 900 kyr (Fig. 13). Except for two samples that have low Si9·0 and variable (La/Sm)N, generally suggesting low degrees of partial melting, the series III alkali basalts from the lower part of the section have high silica contents (45·9–47·0 wt %) accompanied by relatively low (La/Sm)N in the range of 1·78–2·69, thus reflecting relatively high degrees of partial melting (Fig. 13). Because most of the lavas show no evidence for crustal contamination, the increasing (La/Sm)N and decreasing Si9·0 towards the top of the sequence indicate a decreasing degree of partial melting from series III to series II. This rapidly changing degree of melting within several hundred thousand years cannot be due to lithospheric thinning, but could be due to the variable influx of mantle with different temperatures and compositions as indicated by the distinct isotope compositions of the two series.
|
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGSAMPLE PREPARATION AND...PETROGRAPHIC AND GEOCHEMICAL...DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The investigation of the Miocene to recent lavas from the Harrat Ash Shamah volcanic field in southern Syria leads to the following conclusions.
(1) The lava suite comprises alkali basalts, basanites, tephrites, hawaiites, mugearites, rare trachytes and phonolites.
(2) The major and compatible trace element composition of the lavas is controlled by the fractional crystallization of olivine, clinopyroxene, ± Fe–Ti oxides.
(3) The major element chemistry of the highly differentiated lavas is additionally controlled by apatite and minor feldspar fractionation.
(4) Some rocks have been affected by olivine accumulation as shown by their high Ni concentrations and high Ni/Cr ratios.
(5) The low Ce/Pb (15·9–5·99) and Nb/U (34·6–22·0) and high K/La (> 350) ratios as well as the high 87Sr/86Sr (> 0·7039) and low 143Nd/144Nd (< 0·5216) ratios displayed by a number of samples indicate the assimilation of continental crustal material.
(6) The major element variations and the variably fractionated HREE ratios of the uncontaminated, primitive lavas could be produced by different degrees of partial melting of garnet peridotite mantle sources at depths of about 100 km.
(7) As indicated by the isotopic and incompatible trace element ratios, the parental magmas of the Harrat Ash Shamah volcanic field are derived from volatile- and incompatible element-enriched asthenosphere and plume mantle sources.
(8) The discontinuous volcanism found throughout the HAS volcanic field in conjunction with the temporal variations of lava composition and melting regime are consistent with the activity of a pulsing mantle plume underneath southern Syria that can be related to the Afar mantle plume.
|
ACKNOWLEDGEMENTS |
|---|
The ‘field team’ is deeply indebted to the University of Aleppo for the support in Syria. H. Blaschek, D. Garbe-Schönberg and K. Kißling are thanked for help with the ICP-MS analyses. P. Appel, H. Baier, F. Hauff, S. Hauff, B. Mader, P. Rengers, P. van den Bogaard and A. Weinkauf are acknowledged for their help during analytical work. M.-S.K. would like to thank N.A. Stroncik for helpful comments on an earlier version of the manuscript. The constructive reviews and comments by N.W. Rogers, J. Baker, Y. Weinstein and the editor M. Wilson helped to improve the manuscript significantly. This study was funded by the Deutsche Forschungsgemeinschaft (grant HA 2568/5) and is part of M.-S.K.'s dissertation.
|
FOOTNOTES |
|---|
Present address: Department of Earth Sciences, University of Aarhus, Høegh-Guldbergs Gade 2, 8000 Aarhus C, Denmark. 
*Corresponding author. Present address: GeoForschungsZentrum Potsdam, Telegrafenberg, 14473 Potsdam, Germany. Telephone: +49 (0)331 288 1468. Fax: +49 (0)331 288 1474. E-mail: krieni{at}gfz-potsdam.de
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONGEOLOGICAL SETTINGSAMPLE PREPARATION AND...PETROGRAPHIC AND GEOCHEMICAL...DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Albarède F. How deep do common basaltic magmas form and differentiate? Journal of Geophysical Research (1992) 97:10997–11009.[CrossRef]
Al-Damegh K, Sandvol E, Barazangi M. Crustal structure of the Arabian plate: New constraints from the analysis of teleseismic receiver functions. Earth and Planetary Science Letters (2005) 231:177–196. doi: 110.1016/j.epsl.2004.1012.1020.[CrossRef][Web of Science]
Almond DC. Geological evolution of the Afro-Arabian Dome. Tectonophysics (1986) 131:301–332.[CrossRef][Web of Science]
Altherr R, Henjes-Kunst F, Baumann A. Asthenosphere versus lithosphere as possible sources for basaltic magmas erupted during formation of the Red Sea; constraints from Sr, Pb and Nd isotopes. Earth and Planetary Science Letters (1990) 96:269–286.[CrossRef][Web of Science]
Baker JA, Thirlwall MF, Menzies MA. Sr–Nd–Pb isotopic and trace element evidence for crustal contamination of plume-derived flood basalts; Oligocene flood volcanism in western Yemen. Geochimica et Cosmochimica Acta (1996) 60:2559–2581.[CrossRef][Web of Science]
Baker JA, Macpherson CG, Menzies MA, Thirlwall MF, Al Kadasi M, Mattey DP. Resolving crustal and mantle contributions to continental flood volcanism, Yemen; constraints from mineral oxygen isotope data. Journal of Petrology (2000) 41:1805–1820.
Barberi F, Capaldi G, Gasperini P, Marinelli G, Santacroce R, Scandone R, Treuil M, Varet J. Recent basaltic volcanism of Jordan and its implications on the geodynamic history of the Dead Sea shear zone. In: Geodynamic Evolution of the Afro-Arabian Rift System—Carrelli A, ed. (1980) Rome: Accademia Nazionale dei Lincei. 667–683.
Bertrand H, Chazot G, Blichert-Toft J, Thoral S. Implications of widespread high-µ volcanism on the Arabian Plate for Afar mantle plume and lithosphere composition. Chemical Geology (2003) 198:47–61.[CrossRef][Web of Science]
Best JA, Barazangi M, Al Saad D, Sawaf T, Gebran A. Bouguer gravity trends and crustal structure of the Palmyride mountain belt and surrounding northern Arabian Platform in Syria. Geology (1990) 18:1235–1239.
Brew G, Barazangi M, Sawaf T, Al-Maleh K. Tectonic map and geologic evolution of Syria: The role of GIS. Leading Edge (2000) 19:176–182.
Brew G, Barazangi M, Al-Maleh AK, Sawaf T. Tectonic and geologic evolution of Syria. GeoArabia (2001a) 6:573–616.
Brew G, Lupa J, Barazangi M, Sawaf T, Al-Imam A, Zaza T. Structure and tectonic development of the Ghab basin and the Dead Sea fault system, Syria. Journal of the Geological Society, London (2001b) 158:665–674.
Camp VE, Roobol MJ. Upwelling asthenosphere beneath western Arabia and its regional implications. Journal of Geophysical Research (1992) 97:15255–15271.[CrossRef]
Campbell IH, Griffiths RW. Implications of mantle plume structure for the evolution of flood basalts. Earth and Planetary Science Letters (1990) 99:79–93.[CrossRef][Web of Science]
Coleman RG, Gregory RT, Brown GF. Cenozoic volcanic rocks of Saudi Arabia. US Geological Survey, Open-File Report (1983) OF 83-0788:86.
Daradich A, Mitrovica JX, Pysklywec RN, Willett SD, Forte AM. Mantle flow, dynamic topography, and rift-flank uplift of Arabia. Geology (2003) 31:901–904.
Debayle E, Leveque JJ, Cara M. Seismic evidence for a deeply rooted low-velocity anomaly in the upper mantle beneath the northeastern Afro/Arabian continent. Earth and Planetary Science Letters (2001) 193:423–436.[CrossRef][Web of Science]
Deniel C, Vidal P, Coulon C, Vellutini P.-J, Piguet P. Temporal evolution of mantle sources during continental rifting: The volcanism of Djibouti (Afar). Journal of Geophysical Research (1994) 99:2853–2869.[CrossRef]
DePaolo DJ, Daley EE. Neodymium isotopes in basalts of the southwest Basin and Range and lithospheric thinning during continental extension. Chemical Geology (2000) 169:157–185.[CrossRef][Web of Science]
Duffield WA, Dalrymple GB. The Taylor Creek Rhyolite of New Mexico; a rapidly emplaced field of lava domes and flows. Bulletin of Volcanology (1990) 52:475–487.[CrossRef][Web of Science]
Eggler DH. The effect of CO2 upon partial melting of peridotite in the system Na2O–CaO–Al2O3–MgO–SiO2–CO2 to 35 kb, with an analysis of melting in a peridotite–H2O–CO2 system. American Journal of Science (1978) 278:305–343.
Eissen JP, Juteau T, Joron JL, Dupré B, Humler E, Al’Mukhamedov A. Petrology and geochemistry of basalts from the Red Sea axial rift at 18 degrees North. Journal of Petrology (1989) 30:791–839.
El-Isa ZH, Mechie J, Prodehl C, Makris J, Rihm R. A crustal structure study of Jordan derived from seismic refraction data. Tectonophysics (1987) 138:235–253.[CrossRef][Web of Science]
Ershov AV, Nikishin AM. Recent geodynamics of the Caucasus–Arabia–East Africa region. Geotectonics (2004) 38:123–136.[Web of Science]
Falloon TJ, Green DH. Solidus of carbonated fertile peridotite under fluid-saturated conditions. Geology (1990) 18:195–199.
Gallagher K, Hawkesworth CJ. Dehydration melting and the generation of continental flood basalts. Nature (1992) 358:57–59.[CrossRef]
Garbe-Schönberg C.-D. Simultaneous determination of thirty-seven trace elements in twenty-eight international rock standards by ICP-MS. Geostandards Newsletter (1993) 17:81–97.[Web of Science]
Giannérini G, Campredon R, Féraud G, Abou Zakhem B. Déformations intraplaques et volcanisme associé; exemple de la bordure NW de la plaque Arabique au Cenozoique. Bulletin de la Société Géologique de France, Huitieme Série (1988) 4:937–947.
Green DH. Experimental melting studies on a model upper mantle composition at high pressure under water-saturated and water-undersaturated conditions. Earth and Planetary Science Letters (1973) 19:37–53.[CrossRef][Web of Science]
Guba I, Mustafa H. Structural control of young basaltic fissure eruptions in the plateau basalt area of the Arabian Plate, northeastern Jordan. Journal of Volcanology and Geothermal Research (1988) 35:319–334.[CrossRef][Web of Science]
Haase KM. The relationship between the age of the lithosphere and the composition of oceanic magmas; constraints on partial melting, mantle sources and the thermal structure of the plates. Earth and Planetary Science Letters (1996) 144:75–92.[CrossRef][Web of Science]
Haase KM, Muehe R, Stoffers P. Magmatism during extension of the lithosphere; geochemical constraints from lavas of the Shaban Deep, northern Red Sea. Chemical Geology (2000) 166:225–239.[CrossRef][Web of Science]
Hart SR, Dunn T. Experimental cpx/melt partitioning of 24 trace elements. Contributions to Mineralogy and Petrology (1993) 113:1–8.[CrossRef][Web of Science]
Hegner E, Pallister JS. Pb, Sr, and Nd isotopic characteristics of Tertiary Red Sea Rift volcanics from the central Saudi Arabian coastal plain. Journal of Geophysical Research (1989) 94:7749–7755.
Heimann A, Ron H. Geometric changes of plate boundaries along part of the northern Dead Sea Transform; geochronologic and paleomagnetic evidence. Tectonics (1993) 12:477–491.[Web of Science]
Heimann A, Steinitz G, Mor D, Shaliv G. The Cover Basalt Formation, its age and its regional and tectonic setting; implications from K–Ar and 40Ar/39Ar geochronology. Israel Journal of Earth-Sciences (1996) 45:55–71.
Hirose K. Partial melt compositions of carbonated peridotite at 3 GPa and role of CO2 in alkali-basalt magma generation. Geophysical Research Letters (1997) 24:2837–2840.[CrossRef][Web of Science]
Hirose K, Kushiro I. Partial melting of dry peridotites at high pressures; determination of compositions of melts segregated from peridotite using aggregates of diamond. Earth and Planetary Science Letters (1993) 114:477–489.[CrossRef][Web of Science]
Hofmann AW, Jochum KP, Seufert M, White WM. Nb and Pb in oceanic basalts; new constraints on mantle evolution. Earth and Planetary Science Letters (1986) 79:33–45.[CrossRef][Web of Science]
Ilani S, Harlavan Y, Tarawneh K, Rabba I, Weinberger R, Ibrahim K, Peltz S, Steinitz G. New K–Ar ages of basalts from the Harrat Ash Shaam volcanic field in Jordan: Implications for the span and duration of the upper-mantle upwelling beneath the western Arabian plate. Geology (2001) 29:171–174.
Jarrar G, Stern RJ, Saffarini G, Al-Zubi H. Late- and post-orogenic Neoproterozoic intrusions of Jordan: implications for crustal growth in the northernmost segment of the East African Orogen. Precambrian Reseach (2003) 123:295–319.[CrossRef]
Johnson K. TM. Experimental determination of partition coefficients for rare earth and high-field-strength elements between clinopyroxene, garnet, and basaltic melt at high pressure. Contributions to Mineralogy and Petrology (1998) 133:60–68.[CrossRef][Web of Science]
Kelemen PB, Shimizu N, Dunn T. Relative depletion of niobium in some arc magmas and the continental crust; partitioning of K, Nb, La and Ce during melt/rock reaction in the upper mantle. Earth and Planetary Science Letters (1993) 120:111–133.[CrossRef][Web of Science]
Kieffer B, Arndt N, Lapierre H, Bastien F, Bosch D, Pecher A, Yirgu G, Ayalew D, Weis D, Jerram DA, Keller F, Meugniot C. Flood and shield basalts fromEthiopia: magmas from the African Superswell. Journal of Petrology (2004) 45:793–834. doi: 10.1093/petrology/egg1112.
Klein EM, Langmuir CH. Global correlations of ocean ridge basalt chemistry with axial depth and crustal thickness. Journal of Geophysical Research (1987) 92:8089–8115.
Le Bas MJ, Le Maitre RW, Streckeisen A, Zanettin BA. Chemical classification of volcanic rocks based on the total alkali–silica diagram. Journal of Petrology (1986) 27:745–750.
Leeman WP, Scheidegger KF. Olivine/liquid distribution coefficients and a test for crystal–liquid equilibrium. Earth and Planetary Science Letters (1977) 35:247–257.[CrossRef][Web of Science]
Coats RR, Hay RL, Anderson CA. Composition and origin of Hawaiian lavas. Studies in Volcanology: A Memoir in Honour of Howel Wiliams (1968) 116:477–522. Geological Society of America, Memoirs.
Marty B, Pik R, Gezahegn Y. Helium isotopic variations in Ethiopian plume lavas; nature of magmatic sources and limit on lower mantle contribution. Earth and Planetary Science Letters (1996) 144:223–237.[CrossRef][Web of Science]
McBride JH, Barazangi M, Best J, Al Saad D, Sawaf T, Al Otri M, Gebran A. Seismic reflection structure of intracratonic Palmyride fold–thrust belt and surrounding Arabian Platform, Syria. AAPG Bulletin (1990) 74:238–259.[Abstract]
McDonough WF, Sun S.-s. The composition of the Earth. Chemical Geology (1995) 120:223–253.[CrossRef][Web of Science]
McGuire AV, Bohannon RG. Timing of mantle upwelling; evidence for a passive origin for the Red Sea Rift. Journal of Geophysical Research (1989) 94:1677–1682.
McKenzie D, Bickle MJ. The volume and composition of melt generated by extension of the lithosphere. Journal of Petrology (1988) 29:625–679.
Mohsen A, Kind R, Sobolev SV, Weber M, Group D. Thickness of the lithosphere east of the Dead Sea Transform. Geophysical Journal International (2006) 167:845–852.[CrossRef][Web of Science]
Mor D. A time-table for the Levant volcanic province, according to K–Ar dating in the Golan Heights, Israel. Journal of African Earth Sciences (1993) 16:223–234.[CrossRef]
Mouty M, Delaloye M, Fontignie D, Piskin O, Wagner JJ. The volcanic activity in Syria and Lebanon between Jurassic and Actual. Schweizerische Mineralogische und Petrographische Mitteilungen (1992) 72:91–105.
Mysen BO, Kushiro I. Compositional variations of coexisting phases with degree of melting of peridotite in the upper mantle. American Mineralogist (1977) 62:843–865.[Abstract]
Nasir S. The lithosphere beneath the northwestern part of the Arabian Plate (Jordan); evidence from xenoliths and geophysics. Tectonophysics (1992) 201:357–370.[CrossRef][Web of Science]
Nasir S. Geochemistry and petrogenesis of Cenozoic volcanic rocks from the northwestern part of the Arabian continental alkali basalt province, Jordan. Africa Geoscience Review (1994) 1:455–467.
Nasir S. Mafic lower crustal xenoliths from the northwestern part of the Arabian Plate. European Journal of Mineralogy (1995) 7:217–230.
Nasir S, Safarjalani A. Lithospheric petrology beneath the northern part of the Arabian Plate in Syria: evidence from xenoliths in alkali basalts. Journal of African Earth Sciences (2000) 30:149–168.[CrossRef]
Onuma N, Higuchi H, Wakita H, Nagasawa H. Trace element partition between two pyroxenes and the host lava. Earth and Planetary Science Letters (1968) 5:47–51.[CrossRef][Web of Science]
Pik R, Deniel C, Coulon C, Yirgu G, Marty B. Isotopic and trace element signatures of Ethiopian flood basalts; evidence for plume–lithosphere interactions. Geochimica et Cosmochimica Acta (1999) 63:2263–2279.[CrossRef][Web of Science]
Ponikarov VP. Syria (1969) Moscow: Nedra (in Russian).
Ponikarov VP, Protasevich L, Maximov A, Tkachev G. Geological Map of Syria, 1:200 000 (1963) Moscow: V.O. Technoexport.
Robinson J. AC, Wood BJ. The depth of the spinel to garnet transition at the peridotite solidus. Earth and Planetary Science Letters (1998) 164:277–284.[CrossRef][Web of Science]
Rogers NW. The isotope and trace element geochemistry of basalts from the volcanic islands of the southern Red Sea. Journal of the Geological Society, London (1993) 76:455–467.
Rudnick RL, Fountain DM. Nature and composition of the continental crust; a lower crustal perspective. Reviews of Geophysics (1995) 33:267–309.[CrossRef][Web of Science]
Scarrow JH, Cox KG. Basalts generated by decompressive adiabatic melting of a mantle plume; a case study from the Isle of Skye, NW Scotland. Journal of Petrology (1995) 36:3–22.
Schilling JG. Afar mantle plume: Rare earth evidence. Nature (1973) 242:2–5.[CrossRef]
Schilling JG, Kingsley RH, Hanan BB, McCully BL. Nd–Sr–Pb isotopic variations along the Gulf of Aden; evidence for Afar mantle plume–continental lithosphere interaction. Journal of Geophysical Research (1992) 97:10927–10966.[CrossRef]
Sengör A. MC, Burke K. Relative timing of rifting and volcanism on Earth and its tectonic implications. Geophysical Research Letters (1978) 5:419–421.[Web of Science]
Sharkov YV, Lazko YY, Fedosova SP, Hanna S. Depth-derived inclusions of the Quaternary volcano Tel-Danun, southern Syria, in relation to the problem of intraplate basaltic magmatism. Geokhimiya (1989) 11:1609–1623. (in Russian).
Sharkov YV, Chernyshev IV, Devyatkin YV, Dodonov AY, Ivanenko VV, Karpenko MI, Leonov YG, Novikov VM, Khanna S, Khatib K. Geochronology of late Cenozoic basalts in western Syria. Petrologiya (1994) 2:385–394. (in Russian).
Shaw JE, Baker JA, Menzies MA, Thirlwall MF, Ibrahim KM. Petrogenesis of the largest intraplate volcanic field on the Arabian Plate (Jordan): a mixed lithosphere–asthenosphere source activated by lithospheric extension. Journal of Petrology (2003) 44:1657–1679.
Snyder GA, Taylor LA, Jerde EA, Sharkov Y, Laz KY, Hanna S. Petrogenesis of garnet pyroxenite and spinel peridotite xenoliths of the Tell-Danun alkali basalt volcano, Harrat as Shamah, Syria. International Geology Review (1993) 35:1104–1120.
Steiger RH, Jäger E. Subcommission on geochronology; convention on the use of decay constants in geo- and cosmochronology. Earth and Planetary Science Letters (1977) 36:359–362.[CrossRef][Web of Science]
Stein M, Hofmann AW. Fossil plume head beneath the Arabian lithosphere? Earth and Planetary Science Letters (1992) 114:193–209.[CrossRef][Web of Science]
Stein M, Navon O, Kessel R. Chromatographic metasomatism of the Arabian–Nubian lithosphere. Earth and Planetary Science Letters (1997) 152:75–91.[CrossRef][Web of Science]
Stewart K, Rogers N. Mantle plume and lithosphere contributions to basalts from southern Ethiopia. Earth and Planetary Science Letters (1996) 139:195–211.[CrossRef][Web of Science]
Sun SS, McDonough WF. Chemical and isotopic systematics of oceanic basalts; implications for mantle composition and processes. Magmatism in the Ocean Basins. Geological Society, London, Special Publications—Saunders AD, Norry MJ, eds. (1989) 42:313–345.
Suzuki T, Akaogi M, Nakamura E. Partitioning of major elements between garnet-structured minerals and silicate melt at pressure of 3–15 GPa. Physics of the Earth and Planetary Interiors (2000) 120:79–92.[CrossRef]
Tarawneh K, Ilani S, Rabba I, Harlavan Y, Peltz S, Ibrahim K, Weinberger R, Steinitz G. Dating of the Harrat Ash Shaam Basalts Northeast Jordan (Phase 1). In: Hashemite Kingdom of Jordan, Natural Resources Authority; Geological Survey of Israel Report GSI/2/2000 (2000) 59.
Taylor JR. An Introduction to Error Analysis (1982) Mill Valley, CA: University Science Books.
Turner S, Hawkesworth C, Gallagher K, Stewart K, Peate D, Mantovani M. Mantle plumes, flood basalts, and thermal models for melt generation beneath continents: Assessment of a conductive heating model and application to the Parana. Journal of Geophysical Research (1996) 101:11503–11518.[CrossRef]
Vidal P, Deniel C, Vellutini PJ, Piguet P, Coulon C, Vincent J, Audin J. Changes of mantle sources in the course of a rift evolution: The Afar case. Geophysical Research Letters (1991) 18:1913–1916.[Web of Science]
Wallace ME, Green DH. An experimental determination of primary carbonatite magma composition. Nature (1988) 335:343–346.[CrossRef]
Weinstein Y. Spatial and temporal geochemical variability in basin-related volcanism, northern Israel. Journal of African Earth Sciences (2000) 30:865–886.[CrossRef]
Weinstein Y, Navon O, Altherr R, Stein M. The role of lithospheric mantle heterogeneity in the generation of Plio-Pleistocene alkali basaltic suites from NW Harrat Ash Shaam (Israel). Journal of Petrology (2006) 47:1017–1050. doi:10.1093/petrology/egl1003.
White R, McKenzie D. Magmatism at rift zones; the generation of volcanic continental margins and flood basalts. In: Special Section on Magmatism Associated with Lithospheric Extension—Leeman WP, Fitton JG, eds. (1989) Washington, DC: American Geophysical Union. 7685–7729.
White RS, Bown JW, Smallwood JR. The temperature of the Iceland plume and origin of outward-propagating V-shaped ridges. Journal of the Geological Society, London (1995) 152:1039–1045.
Wyllie PJ. Mantle fluid compositions buffered in peridotite–CO2–H2O by carbonates, amphibole, and phlogopite. Journal of Geology (1978) 86:687–713.[Web of Science]
York D. Least squares fitting of a straight line with correlated errors. Earth and Planetary Science Letters (1969) 5:320–324.[CrossRef][Web of Science]
Ziegler PA, Cloetingh S. Dynamic processes controlling evolution of rifted basins. Earth-Science Reviews (2004) 64:1–50.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||