Keywords

1 Introduction

The Arabian Shield (AS, Fig. 6.1) consists of Neoproterozoic juvenile tectonostratigraphic island-arc terranes formed by subduction within and around the Mozambique Ocean between 850 and 550 Ma in the framework of the Gondwana supercontinent assembly (Stern 1994; Nehlig et al. 2002; Stern and Johnson 2010). These terranes are joined by ophiolite-decorated sutures (Stoeser and Camp 1985; Johnson and Woldehaimanot 2003). The AS is divided into at least five tectonostratigraphic terranes (Fig. 6.1; Stoeser and Camp 1985): Midyan, Hijaz, Asir, Ar Ryan and Afif with some workers (e.g., Johnson and Woldehaimanot 2003) discussed the possibility of three more terranes (Jiddah, Ad Dawadimi and Khida). All terranes (except Khida; Stacey and Agar 1985; Stoeser and Frost 2006) are regarded as Neoproterozoic oceanic arcs, including fore-arc and/or backarc crust and ophiolites (Dilek and Ahmed 2003; Stoeser and Frost 2006).

Fig. 6.1
figure 1

a General geologic map of the Arabian Shield showing major tectonostratigraphic terranes, ophiolite belts, sutures, fault zones and post-accretionary basins in the Arabian Shield of western Saudi Arabia (modified after Nehlig et al. 2002; Johnson and Woldenhaimanot 2003; Stern and Johnson 2010). Numbers are locations of the I-type intrusive rocks represented in geochemical Figures as follows: (1) Bulghah and Humaymah, (2) Sukhaybarat, (3) Jabal Ghadarah, and (4) Makkah Suite. b Geological sketch map of the Afif Terrane and adjacent parts of the Arabian Shield showing location of the study area (modified after Johnson and Kattan 1999)

Full size image

The evolution of the AS terranes records three main tectonic stages of intra-oceanic subduction (850–700 Ma), collision and terrain amalgamation (700–635 Ma), and tectonic escape, strike-slip faulting and extension (635–550 Ma) of the newly formed continental crust (Genna et al. 2002; Johnson and Woldehaimanot 2003; Be’eri-Shlevin et al. 2009; Eyal et al. 2010; Stern and Johnson 2010; Johnson et al. 2011). The volcanic arc crust range in composition from tholeiitic, through calc-alkaline to late high-K suites (Robool et al. 1983). Syn-to Post-collisional calc-alkaline to alkaline I-type granites intruded by Post-collisional K-rich alkaline to peralkaline A-type granites (e.g., Ali et al. 2009, 2012; Be’eri-Shlevin et al. 2009; Eyal et al. 2010; Moreno et al. 2014; Moghazi et al. 2012).

I-type granitoids (Cryogenian and Early Ediacaran) are major components of the ANS (Moussa et al. 2008; Johnson et al. 2011). Geochemical studies reveal that they are metaluminous to slightly peraluminous, calc-alkaline, subduction-related intrusives (Pearce et al. 1984; Maniar and Piccoli 1989; Jarrar et al. 2003; Moussa et al. 2008; Be’eri-Shlevin et al. 2009; Ali et al. 2015; Robinson et al. 2015), formed synchronously with the formation of large molasse basins (Genna et al. 2002; Johnson 1998; Nehlig et al. 2002; Stern and Johnson 2010). The orogenic phase was then followed by continued convergence and the development of regionally extensive strike-slip fault (escape tectonics) from 610 to 525 Ma (Greiling et al. 1994; Stern 1994; Genna et al. 2002; Johnson et al. 2011; Robinson et al. 2014).

Intrusion of post-collisional alkaline to peralkaline rocks (A-type granites) and development of extensive pull-apart basins during the final suturing of the Arabian-Nubian Shield and the formation of subsequent extensional basins (Greiling et al. 1994; Johnson 1998, 2003; Wilde and Youssef 2000; Nehlig et al. 2002; Johnson et al. 2011). The alkaline granites constitute about 2% of the Arabian–Nubian Shield and represent one of the largest fields of alkaline granites on Earth (Stoeser 1986; Liégeois and Stern 2010; Johnson et al. 2011). Petrogenetic models for the generation of A-type alkaline granites range from partial melting of the lower crust to extreme differentiation of mantle-derived tholeiitic or alkaline basaltic magma (Bonin and Giret 1990; Turner et al. 1992; Collins et al. 1992; Frost and Frost 1997; Patiño Douce 1997; Mushkin et al. 2003; Huang et al, 2011). Assimilation of older crustal material into mantle derived granitic magma is plausible model to account for the petrogenesis of such rocks (Eby 1990, 1992; Kemp et al. 2006; Zhang et al. 2012).

This paper will present results of U–Pb zircon geochronology and whole-rock geochemical data of some plutonic rocks from Bulghah and Humaymah areas (west of the Afif terrane) of Saudi Arabia. While gold is mined from Bulghah area since 2001, it is still under evaluation and development by Ma’aden Co. in the Humaymah area. In these areas, gold is spatially associated with some intrusive rocks intruding the layered volanosedimentary the Hulayfah/Siham and Murdamah group rocks. These intrusive rocks can be grouped into two groups; older mafic to intermediate gabbro-diorite and younger intermediate to felsic tonalite-granodiorite intrusive bodies. These data will be used to clarify the magma sources for these rocks, and to assess fractional crystallization as a possible petrogenetic mechanism for the formation of these rocks (Miller 1985; Secchi et al. 1991; Teixeira et al. 2012; Wang et al. 2014).

2 Geology Background and Petrography

The two selected areas (Bulghah and Humaymah) for the present study are located in the Afif terrane (Figs. 6.1 and 6.2). The Afif terrane is one of the largest terranes in the Arabian Shield. Much of the recent researches on the shield in the last 15 years, Johnson and Kattan (1999) addressed the composite nature of the Afif terrane in age and provenance, mainly based on detailed mapping, Pb-isotope systematics and U–Pb zircon geochronology (e.g., Stacey and Hedge 1984; Stacey and Agar 1985; Agar 1985, 1988; Agar et al. 1992). Based on these studies, the Afif terrane was divided into four possible subterranes including the Khida, Nuqrah, Siham, and Suwaj (Fig. 6.2). While the Khida subterrane represents remnants of the Archean to Paleoproterozoic continental rocks, the Nuqrah, Siham, and Suwaj subterranes and the post-amalgamation assemblages (Murdama and Bani Ghayy groups) represent the Neoproterozoic volcanic/magmatic arc-subterranes (Johnson and Kattan 1999). Agar et al. (1992) reported 1.7 Ga zircon ages for pre-Neoproterozoic sources in the Khida subterrane. The ophiolite assemblages (mafic/ultramafic rocks) represent the oldest Neoproterozoic rocks in the study areas. A plagiogranite dyke in serpentinized peridotites from a near vicinity to the study area in the Bir Tuluha ophiolite (Fig. 6.2) yielded U–Pb zircon ages from 843 to 821 Ma (Pallister et al. 1988). The mafic-ultramafic rocks are represented by gabbro, metabasalt, serpentinites, listvenites and amphibolites. The Siham group (Fig. 6.2) is unconformably overlain by the volcanosedimentary rocks of the Bani Ghayy group (Stacey and Agar 1985). The Siham group is metamorphosed to greenschist facies and composed mainly of volcanosedimentary assemblages. It is composed mainly of basaltic lava range from tholeiitic to calc-alkaline and has affinity with lavas in active continental margins (Agar 1986). The sedimentary rocks of Siham group are represented by lithic sandstone, shale and conglomerate. The age of the Siham group is constrained at ~746 ± 10 Ma (Agar et al. 1992) from a granodiorite sample from Naim complex which intruded the Siham group. The Bani Ghayy group in the study area consists of unmetamorphosed sandstone, conglomerate, limestone, basalt and rhyolite. In the western part of the study area (Fig. 6.2), the Bani Ghayy and Murdama groups are separated by a thrust fault. Some workers considered the two groups to be equivalent because they are similar in age and lithology (Brown et al. 1989). The two groups suggested to be deposited during and soon after the Nabitah orogeny (680–640 Ma, Johnson et al. 2011). Stacey and Agar (1985) reported a U–Pb zircon age of 620 ± 5 Ma for a rhyolite sample from the Bani Ghayy group and a volcanic sample from Murdama group yielded a U–Pb zircon age of 625 ± 4 (Kusky and Matsah 2003).. The volcanosedimentary rocks and intermediate to felsic intrusions of the Afif terrane were intruded by gabbros to highly evolved granites between ~750 and 570 Ma. Based on their composition, textures and/or absolute ages, these rocks are classified from the oldest to the youngest as the Jidh, Suwaj, Fahud, Khishaybi, Humaymah, Haml and syn- to post-Shammar intrusions of the Idah and Abanat suites. The post-collision monzogranite rocks of Haml batholith were dated at 609 ± 8 Ma (Robinson et al. 2014).

Fig. 6.2
figure 2

Geological map showing the study areas (Bulghah and Humaymah) modified from the Geological maps 1:250000 of the Al Hissu quadrangle, sheet 24 E (Delfour 1981) and the Nuqrah quadrangle, sheet 25 E (Delfour 1977)

Full size image

3 Analytical Techniques

Four representative samples of the intrusive rocks were selected for dating using U–Pb zircon geochronology. Bulk samples of these rocks (each weight ≈ 3 kg) were crushed and pulverised to a fine fraction and sieved. The fractions of these samples that fall between 250 and +75 µm were washed by distilled water and dried. These initial sample preparations were carried out in the laboratory facilities at the department of Mineral Resources and Rocks at the Faculty of Earth Sciences, King Abdulaziz University (KAU) in Jeddah (Saudi Arabia). Heavy liquids were used to separate the heavy fraction from these samples using Methylene Iodide (specific gravity 3.3 gm/ml). The non-magnetic fraction of these samples which may contain the zircon grains were separated using a Franz Isodynamic separator. Finally, zircon grains were collected by handpicking under a binocular stereoscopic microscope. Zircons were analysed for their U–Pb age at John de Laeter Centre, Curtin University of Technology (Australia) and at the University of Oslo (Norway).

Prior to analysis, zircon grains were imaged on the polished puck by cathodoluminecence (CL) to examine the growth structure of individual grains using a scanning electron microscope. U–Pb isotope data on zircons were obtained by laser-ablation inductively coupled plasma source mass spectrometry (LA-ICP-MS) using a Nu Plasma HR mass spectrometer and a New Wave LUV213 laser microprobe at University of Oslo (Norway) and by sensitive high resolution ion microprobe (SHRIMP II) at Curtin University (Australia). The analytical protocols of the LA-ICP-MS are described in detail by Rosa et al. (2009), and SHRIMP analytical procedure is described by Compston et al. (1984) and Kennedy and de Laeter (1994). The 1065 ± 0.6 Ma (2σ) Geostandards zircon 91500 (Wiedenbeck et al. 1995) and 600 ± 4.5 Ma (2σ) Geostandards GJ-1 (Jackson et al. 2004) were both used for Pb/U calibration and U concentration estimates and were analysed on a regular basis during the analytical sessions by LA-ICP-MS, and BR266 (559 Ma; 903 ppm U) was used for SHRIMP. Calculations used the routines of Isoplot (Ludwig 2001a, b). U–Th–Pb concentrations and isotopic compositions are listed in Table 6.1 (SHRIMP II) and Table 6.2 (LA-ICP-MS), and plotted as two-sigma error ellipses on Concordia diagrams (Wetherill 1956).

Table 6.1 SHRIMP Th-U-Pb zircon data for samples from Saudi Arabia
Full size table
Table 6.2 LA-ICP-MS isotopic data for zircons from Saudi Arabia
Full size table

A total of 53 intrusive samples were selected for major, trace and rare earth elements analyses. Major element compositions and Sc, Ba, and Ni abundances were determined by inductively coupled plasma-atomic emission spectrometry (ICP-AES). All the others trace and rare earth elements (REE) were determined by inductively coupled plasma-mass spectrometry (ICP-MS). All the analyses were carried out at the ACME Analytical Laboratories Ltd., Canada. Analytical precision, as calculated from replicate analyses, is 0.5% for major elements and varies from 2 to 20% for trace elements. Analytical results are listed in the Table 6.3.

Table 6.3 Major and trace elements of the intrusive rocks from Saudi Arabia
Full size table

4 Results

4.1 U–Pb Geochronology

Sample B139 (N25° 01′ 37.0″, E41° 36′ 32.9″) is a dark grey, coarse-grained diorite collected from Bulghah area. Zircons separated from this sample are euhedral, acicular, and yellowish brown in color. A total of 21 measurements were made on the twenty one zircon crystals (Table 6.1). The U content varies from 65 to 161 ppm and Th from 18 to 70 ppm. One zircon analysis produced a concordant data point with 206Pb/238U age of 707 ± 10 Ma, interpreted as that as xenocryst derived from older source material. Omitting three analyses with high common Pb and/or discordant leaves seventeen tightly grouped and concordant analyses that define a concordia age of 677 ± 8 Ma (95% conf.; MSWD = 0.91; Fig. 6.3a). We interpret this as the age of intrusion and crystallization of the diorite.

Fig. 6.3
figure 3

U–Pb concordia diagrams of ion microprobe (SHRIMP) and laser-ablation inductively coupled plasma source mass spectrometry (LA-ICP-MS) data for zircons studied from the Bulghah and Humaymah areas. Dashed ellipses indicate zircon analyses that were excluded from age calculations. Errors for data ellipses are 2σ. Analytical data are given in Tables 6.1 and 6.2

Full size image

Sample B-mine (N24° 59′ 21.5″, E41° 35′ 51.6″) is a quartz-diorite collected from Bulghah mine area (Fig. 6.2). Zircon extracted from this sample is mostly euhedral, clear to yellow-brown, prismatic (~200 μm × 50 μm) and show well-developed oscillatory zoning consistent with a magmatic origin. One analysis was made on each of 10 zircons (Table 6.1) and these are presented on a concordia plot (Fig. 6.3b). Zircon U contents are low (58–163 ppm) and Th/U is low (0.3–0.4; Table 6.1). One analysis is concordant and has low Th content (16 ppm) and yields a 206Pb/238U age of 746 ± 14 Ma, higher than the 9 analyses, interpreted to represent an inherited grain from the older rocks in the area. The remaining nine analyses are concordant, defining a concordia age of 667 ± 6 Ma (2σ; MSWD = 2.3; Fig. 6.3b). This age is identical to the age of the previous diorite sample (B139) and is interpreted as the crystallization age of the quartz-diorite.

Sample Hm19 (N24° 42′ 11.3″, E41° 42′ 57.8″) is medium to coarse-grained tonalite from Humaymah area (Fig. 6.2). Zircon recovered from this sample is subhedral to euhedral and yellow to pale brown. Zircon grains are idiomorphic, slender, and needle shaped, and exhibit well-preserved oscillatory growth zoning. These zircons contain moderate U contents (76–325 ppm) and have Th/U in the range expected for igneous zircons (0.2–0.5). Out of twenty-one zircon grains analyzed (Table 6.1), eleven analyses yield a concordia age of 683 ± 4 Ma (2σ, MSWD = 0.03), interpreted to represent inheritance from older rocks in the area. Three analyses are discordant and excluded from age calculation (Fig. 6.3c). Seven analyses were done on the rims of the same zircon grains produced a concordant data points and yield a concordia age of 661 ± 5 Ma (2σ, MSWD = 2.7; Fig. 6.3c), interpreted to represent the crystallization age of the tonalite intrusion.

Sample HC196 (N24° 43′ 01.5″, E41° 43′ 2.5″) is coarse-grained tonalite from Humaymah (Fig. 6.2). Zircon recovered from this sample is subhedral to euhedral and yellow to pale brown. Zircon grains are idiomorph, slender, and needle shaped, and exhibit well-preserved oscillatory growth zoning. Out of fourteen zircon grains analyzed (Table 6.2), twelve analyses yielded a concordia age of 643 ± 4 Ma (2σ, MSWD = 1.1), interpreted to represent the age of the tonalite. Two analyses are discordant and excluded from age calculation (Fig. 6.3d). This age is indistinguishable from the age of a quartz diorite sample (629 ± 6 Ma; Harbi et al. 2018) in the Sukhaybarat area (70 km north of the study area) and the age of a monzogranite sample (631 ± 1 Ma) from Bulghah area (Bakhsh, unpublished data).

4.2 Whole-Rock Geochemistry

The results of chemical composition of 53 whole-rock samples from Bulghah and Humaymah areas are listed in Table 6.3. We compared our results with previous geochemical data of the I-type intrusive rocks (Figs. 6.5, 6.6, 6.7, 6.8, 6.9, 6.10, 6.11, 6.12 and 6.13) from Harbi et al. (2016, 2018) and Robinson et al. (2015). The intrusive samples from Bulghah show wide variations (in wt%) of SiO2 (51.4–74.7), Al2O3 (12.39–19.95), CaO (1.05–9.95), Na2O (3.10–5.85), MgO (0.30–6.50), Fe2O3t (1.06–7.41), and TiO2 (0.18–1.02) (Table 6.3, Fig. 6.4). The K2O content indicates that the studied samples are high-K calc-alkaline (1.07–4.11 wt%; Fig. 6.4d). The intrusive rocks from Humaymah are more mafic than samples from Bulghah area. They exhibit variable variations (in wt%) of SiO2 (48.4–64.3), Al2O3 (14.0–19.33), CaO (4.26–12.22), Na2O (1.05–4.09), and high TiO2 (0.45–1.60), MgO (1.78–10.64) and Fe2O3t (5.35–9.89) (Table 6.3, Fig. 6.4). The intrusive rocks have been classified using the total alkalis versus silica (Cox et al. 1979), Q′–ANOR diagram of Streckeisen and Le Maitre (1979) and R1-R2 diagram of de la Roche et al. (1980). The intrusive samples of Bulghah plot in the monzogranite–synogranite, granodiorite-tonalite and syenodiorite fields (Fig. 6.5), whereas the Humaymah intrusive samples fall in the diorite, quartz diorite, gabbro and quartz gabbro fields, and one sample (Hm 19) plots in the tonalite field (Fig. 6.5).

Fig. 6.4
figure 4

Major and trace element variations diagrams (Harker diagrams) for the gabbro, diorite, tonalite, granodiorite and monzogranites define two different trends and showing compositional gap between the mafic intrusive rocks (gabbro and diorite) and granitoids (granodiorite-tonalite and monzogranite)

Full size image
Fig. 6.5
figure 5

Major and trace elements chemical classification diagrams. a Total alkalis (Na2O + K2O) versus SiO2 (Cox et al. 1979), b The Q’–ANOR plutonic rocks classification diagram using their molecular normative compositions (after Streckeisen and Le Maitre 1979), c R1-R2 diagram (de la Roche et al. 1980). The field of Sukhaybarat I-type granitoids from Harbi et al. (2018) and the field of Jabal Ghadarah I-type granitoids from Harbi et al. (2016). Analytical data are given in Table 6.2

Full size image

According to Shand’s index (Maniar and Piccoli 1989) classification diagram, the studied mafic intrusive samples (gabbro and diorite-quartz diorite) from Bulghah and Humaymah are strongly metaluminous, whereas the granitoid rocks (granodiorite, tonalite and monzogranite) are strongly metaluminous to slightly peraluminous (Fig. 6.6a). On the major element discrimination diagram of Sylvester (1998), the studied intrusive rocks are calk-alkaline (Fig. 6.6b). They also fall in the calc-alkaline field (AI < 0.87; Fig. 6.6c) according to Liégeois and Black (1987). Three monzogranite samples (B46, B49 and B50) fall in the alkaline and highly fractionated calc alkaline field in both diagrams (Fig. 6.6b, c).

Fig. 6.6
figure 6

a Shand’s index (Maniar and Piccoli 1989) classification diagram, showing the intrusive studied samples exhibit metaluminous to slightly peraluminous geochemical features. b (Al2O3 + CaO/FeOt + Na2O + K2O) versus 100(MgO + FeOt + TiO2 + SiO2) discrimination diagram (Sylvester 1998) showing the predominantly calc-alkaline to highly fractionated calc-alkaline characteristics of the studied samples. c Agpaitic index (AI = Na + K/Al) versus SiO2 diagram showing the calc-alkaline characters of the studied intrusive samples, the line with AI = 0.87 (Liégeois and Black 1987) separates alkaline and calc-alkaline granite series. The field of Sukhaybarat I-type granitoids from Harbi et al. (2018) and the field of Jabal Ghadarah I-type granitoids from Harbi et al. (2016)

Full size image

Primitive mantle normalized trace-element patterns for all intrusive samples from Bulghah and Humaymah are enriched in incompatible elements, showing negative anomalies in Nb–Ta, Ti and P and a positive Pb anomaly (Fig. 6.7). The mafic and granitoid rocks in the studied areas have variable Al2O3/(CaO + Na2O + K2O) molar ratios (A/CNK), mostly less than 1.1 (Fig. 6.6a), which is typical of I-type magmatism (Chappell 1999). The trace element patterns (Fig. 6.7) are typical of those developed in island arc settings (Elliott 2003; Ali et al. 2015).

Fig. 6.7
figure 7

Primitive mantle-normalized trace element diagrams for the studied Bulghah and Humaymah samples. Normalizing values from Sun and McDonough (1989)

Full size image

The granodiorite-tonalite and monzogranite samples are both LREE and MREE enriched [(La/Yb)N = 6.12 − 21.70], with slightly fractionated HREE patterns [(Dy/Yb)N = 0.93 − 2.13] and minor negative to slightly positive Eu (Eu/Eu* = 0.35 − 1.25) and positive Sr anomalies (Figs. 6.7 and 6.8) which are indicative of feldspar fractionation. The diorite-quartz diorite samples are LREE and MREE enriched [(La/Yb)N = 3.38 − 11.24], with flat to slightly fractionated HREE patterns [(Dy/Yb)N = 1.00 − 1.63] and minor negative to slightly positive Eu (Eu/Eu* = 0.75 − 1.29) and positive Sr anomalies (Figs. 6.7 and 6.8). However, the Humaymah gabbro samples show almost flat REE patterns [(La/Yb)N = 2.06 − 4.67 and (Dy/Yb)N = 1.14 − 1.28], indicating that garnet did not control elemental partitioning during melting or fractionation, with positive Eu (Eu/Eu* = 1.08 − 1.43) and Sr anomalies (Figs. 6.7 and 6.8), and overall REE abundance is lower.

Fig. 6.8
figure 8

Chondrite normalized REE patterns for the studied Bulghah and Humaymah samples. Normalizing values from Sun and McDonough (1989)

Full size image

The variation of Th and U show strong positive correlations (Fig. 6.9a) which indicates magmatic behavior of Th and U during magmatic differentiation (Moghazi et al. 2011). Moreover, the variation of Ta versus Nb, and Hf versus Zr (Fig. 6.9b, c) show positive correlation, suggesting that the behavior and enrichment of Nb and Ta are largely controlled by magmatic process (Lehmann and Mahawat 1989; Charoy and Nornoha 1991).

Fig. 6.9
figure 9

a Th versus U, b Ta versus Nb, and c Hf versus Zr variation diagrams of the studied intrusive samples from the Bulghah and Humaymah areas, showing geochemical positive correlation between the different granitic types, suggesting that the behavior of these trace elements are largely controlled by magmatic processes. The field of Sukhaybarat I-type granitoids from Harbi et al. (2018)

Full size image

5 Discussion

New U–Pb zircon ages and geochemical data that characterize the Bulghah and Humaymah areas, when combined with available geochemical data from previous studies (Harbi et al. 2016, 2018; Robinson et al. 2015), allow us to discuss their tectonic setting and their petrogenesis.

5.1 Tectonic Stetting

Granitic rocks are divided into I-, S- and A-types (Whalen et al. 1987; Chappell and White 1992, 2001). A-type granites contain high temperature hydrous phases such as pyroxene and fayalite, and late crystallization biotite and alkali amphibole (Wormald and Price 1988; King et al. 1997). However, petrographic investigation indicates that Bulghah and Humaymah granites show no pyroxene or fayalite, and contain hornblende which is inconsistent with alkali amphibole in typical A-type granites. Moreover, the slight enrichment in high field strength element (HFSE) and REE concentrations, in combination with the petrographic characteristics and the slight low Zr, Nb, Ce and Y concentrations in most granitic samples, suggest that the studied samples are not A-type granites. The studied granitoids are not also S-type granites because they are metaluminous to slightly peraluminous, and have A/CNK < 1.1 and A/NK > 1 (Fig. 6.6a). It show also that P2O5 decreases with increasing SiO2 (Fig. 6.4h). Additionally, the studied samples contain no Al-rich minerals such as cordierite, muscovite or garnet which are considered the common features of S-type granites (Chappell and White 1992; Huang et al. 2013).

Except for three samples, the granitoid rocks of Bulghah and Humaymah fall in the magnesian field on FeOt/FeOt + MgO versus SiO2 diagram (Frost et al. 2001) (Figure not shown). They are classified as I-type granites on the 10,000 × Ga/Al versus Zr diagram (Whalen et al. 1987) (Fig. 6.10a). Rb, Ta, Nb and Y, among other trace elements, have been used to discriminate the different tectonic setting of granitoid rocks (Pearce et al. 1984). The studied granitoids fall in the volcanic-arc/syn-collisional fields (VAG + syn-COLG) in the Y + Nb versus Rb, Y versus Nb, and Yb versus Ta diagrams (Fig. 6.10b–c) of Pearce et al. (1984).

Fig. 6.10
figure 10

Major and trace elements tectonic discrimination diagrams. a Rb versus Y + Nb, b Nb versus Y and c Ta versus Yb diagrams (after Pearce et al. 1984), showing the studied granitoids plot in the VAG field. d Zr versus 10,000 Ga/Al diagram (Whalen et al. 1987), showing the studied intrusive fall in the I-type granite field. However, few monzogranite samples fall in the A-type field, perhaps due to the mobility of Al (major element). Note diorite intrusive samples are also plotted to highlight the compositional range within these suites, but are classified in Fig. 6.11. VAG = volcanic arc granite, syn-COLG = syn-collision granite, ORG = ocean ridge granite, WPG = within plate granite and post-COLG = post-collisional granite. The field of Sukhaybarat I-type granitoids from Harbi et al. (2018) and the field of Jabal Ghadarah I-type granitoids from Harbi et al. (2016)

Full size image

Mafic intrusive rocks (gabbro and diorite) have been classified using the Cr versus Y diagram (Fig. 6.10a; Pearce et al. 1984) and Cr versus Ce/Sr diagram (Fig. 6.10b; Pearce 1982), they plot in the island arc tholeiites (IAT) field. However, V versus Ti/1000 discrimination diagram ((Fig. 6.7c; Shervais 1982) classify the gabbro and diorite samples as mid-ocean ridge (MORB)/back-arc basin (BAB) mafic components. In summary, variation in multi-elements diagrams (Fig. 6.7) and REE discrimination diagram (Figs. 6.10 and 6.11) indicates that the studied rocks are related to calc-alkaline I-type magmatism formed in an island-arc setting.

Fig. 6.11
figure 11

Mafic intrusive classification schemes applied to gabbroic and dioritic samples from Bulghah and Humaymah areas. a Cr versus Y and b Cr versus Ce/Sr (after Pearce et al. 1984), showing the studied samples fall in the IAT field. c V versus Ti/1000 (after Shervais 1982), showing the studied samples plot in the IAT, MORB and BAB fields. The field of Makkah Suite mafic intrusive rocks (Jeddah terrane) from Robinson et al. (2015). Results are discussed in text. MORB = mid-ocean ridge basalt, BAB = back-arc basalt, and IAT = island-arc tholeiitic

Full size image

5.2 Petrogenetic Evaluation

The main compositional trends of the studied intrusive rocks are shown on Harker variation diagrams (Fig. 6.4). The figure shows that many elements do not have straight-line variations (e.g., Al2O3, MgO, CaO, TiO2, Na2O, P2O5, Nb, Ni and Y) but exhibiting compositional gap between different intrusive types. This suggest two different compositional trends in Bulghah and Humaymah intrusive rocks; one for gabbro and diorite and the other for granodiorite-tonalite and monzogranite. Overall, the mafic intrusive rocks are characterized by slightly LREE enriched patterns and relatively flat HREE patterns [(Dy/Yb)N = 1.00 − 1.33, except for three diorite samples show higher values [(Dy/Yb)N = 1.45 − 1.63) (Fig. 6.8a, d, e). Most samples show Cs, Sr, Rb and K enrichments and Nb, Ta and Ti depletions. The gabbroic samples have La/Nb ratios range from 0.9 to 2.3 and slightly LREE enriched [(La/Yb)N = 2.06 − 4.67] (Fig. 6.8), although all mafic samples (gabbro and diorite) are outside this range [(La/Yb)N = 2.06 − 11.24]. All mafic intrusive samples show negative Ce and positive or slightly negative Eu anomalies (Eu/Eu* = 0.75 − 1.43) (Figs. 6.7 and 6.8). These geochemical features are comparable to those of the arc-metavolcanic calc-alkaline rocks of the Arabian Shield (Ali et al. 2010), which were produced by partial melting of plagioclase- or spinel-peridotite in the upper most mantle <80 km deep in an intra-oceanic island arc, as indicated from the flat HREE patterns (Rudnick et al. 2004) and other trace element characteristics (Ali et al. 2010). This suggests that the mafic intrusive rocks of Bulghah and Humaymah represent the plutonic equivalents of the Arabian Shield arc metavolcanic calc-alkaline rocks.

I-type granitoids can be formed through number of magmatic processes (e.g., Miller 1985; Jiang et al. 2007; Hassanen et al. 1996; Best and Christiansen 2001; Roberts and Clements 1993; Skjerlie and Johnston 1992; Frost and Frost 1997; Clemens et al. 2011; Chappell et al. 2012; Huang et al. 2013; Zhang et al. 2015), including Derivation from magmas generated by partial melting of mafic and intermediate igneous rocks has been proposed for some I-type granitoids (see Best and Christiansen 2001). Others have suggested that the I-type granitoids derived from partial melting of ancient lower crust (e.g., Jiang et al. 2007) or hydrous, calc-alkaline to high-K calk-alkaline old metamorphic rocks (Roberts and Clemens 1993). Several studies proposed fractional crystallization of mafic/intermediate magmas as an important process to generate felsic magmas (e.g., Miller and Mittlefehlt 1984; Miller 1985; Secchi et al. 1991; Teixeira et al. 2012). Partial melting of eclogite or garnet amphibolite at mantle depth (Hassanen et al. 1996) and dehydration melting of amphibole-bearing tonalite (Creaser et al. 1991; Skjerlie and Johnston 1992; Frost and Frost 1997) are other potential sources to generate I-type granites. Differentiation (i.e. partial melting and/or fractional crystallization) of crust and then the magma may have been modified to a degree by fractional crystallization (Miller 1985; Turpin et al. 1990; Barbarin 1996; Champion and Chappell 1992; Chappell et al. 2012; Zhang et al. 2015) could be another source for the I-type granites.

The I-type granitoids (granodiorite-tonalite to monzogranite) of the current study cannot account as a product of mafic magma fractionation since Harker diagrams exhibiting compositional gap between different intrusive types (Fig. 6.4). This exclude the possibility that the granitoids were produced from magmas generated by partial melting of mafic igneous rocks. This is further supported by the U–Pb geochronology data, the dioritic rocks yielded U–Pb ages of 677 ± 8 Ma and 667 ± 6 Ma (Fig. 6.3), whereas the U–Pb zircon age of the monzogranite intrusion is 631 ± 1 Ma (Bakhsh, unpublished data). A large span in U–Pb ages from 677 (mafic rocks) to 631 (intermediate and felsic rocks) implies that the granodiorite and monzogranite originated from a different magma source than that of the mafic magma (gabbro and diorite).

The compositional variations from granodiorite-tonalite to monzogranite of Bulghah and Humaymah perhaps caused by varying degree of partial melting suggest various degree of fractional crystallization. Varying degree of partial melting of a source could effectively produce melts that have diverse concentration of both major and trace elements (Huang et al. 2013). The most felsic rocks (monzogranite) representative of the lowest degree of partial melting of a source should have the highest La/Yb but lowest Dy/Yb (Huang et al. 2013). However, monzogranite and granodiorite samples show scattered ratios (Fig. 6.12a), inconsistent with the partial melting trend. Thus, it is unlikely that varying degree of partial melting is responsible for the diverse elemental concentrations in Bulghah grantoid rocks. Partial melting model of garnet amphibolite at mantle depth (Hassanen et al. 1996) generates a peraluminous melt (Beaard and Lofgen 1991). However, the studied granitoid samples are metaluminous to weakly peraluminous, indicating water undersaturated partial melting (Huang et al. 2013). Garnet with plagioclase and orthopyroxene would be major residual phases at high pressure (Wolf and Wyllie 1994). If garnet is a residual phase in the source, the HREE patterns will show strong depletion, but the studied granitoid samples show flat to slightly depleted HREE (Fig. 6.8b, c, f). Therefore, the possibility of a high pressure magma source can be excluded as a source for the studied granitoid samples. Contrasting REE patterns between the granodiorite-tonalite and monzogranite samples may have resulted from various degree of fractional crystallization. K-feldspar separation is responsible for the depletion of Ba (Wu et al. 2002; Fig. 6.7c, e), whereas Plagioclase fractionation depleted Eu and Sr in the melts (Figs. 6.7 and 6.8). As shown by Sr, Sr/Y versus Eu/Eu* (Fig. 6.12b and c), the correlations observed in Bulghah granite samples are indicative of plagioclase fractionation. Fractionation of amphibole will lower the Nb/Ta and Dy/Yb ratios in the remaining melt because of its D(Nb)/D(Ta) > 1 (Tiepolo et al. 2001) and D(Dy)/D(Yb) > 1 (Sisson 1994). A positive correlation between Nb/Ta and Dy/Yb in the Bulghah I-type granitoids strongly suggests amphibole fractionation (Fig. 6.12d). This is further confirmed by the positive correlation between Er and Dy corresponds to the fractionation of hornblende from the parental magma (Drummond et al. 1996; Fig. 6.12e). Biotite have a low value for Th, but high partition coefficients for Sc and V (Bea et al. 1994), therefore biotite fractionation will increase SiO2/Al2O3 but decrease Sc/Th and V/Th ratios in residual melts (Huang et al. 2013). Thus, negative correlations between V/Th and Sc/Th and SiO2/Al2O3 (Fig. 6.12f, g), perhaps indicate biotite fractionation between granodiorite and monzogranite samples of Bulghah. Ti-bearing minerals, such as titanite and ilmenite, might be other fractionated phases as suggested by increasing TiO2 with decreasing SiO2 (Huang et al. 2008; Table 6.3, Fig. 6.4c). The fractional crystallization of titanium-rich minerals commonly generates negative Nb–Ta and Ti anomalies in basalts (Xiong et al. 2005; Huang et al. 2008). However, this cannot be the reason for low Nb–Ta (Fig. 6.8b) in the studied granitoids because there are lack of correlations between TiO2/Yb, Eu/Eu* and Nb/La (Fig. 6.12h, i), indicating a minor role of Fe–Ti oxide fractionation in the genesis of the studied granitoids. Therefore, various degrees of plagioclase, hornblende and biotite fractional crystallization would be important mechanism for variations in chemical composition of the studied granitoids (Fig. 6.12a–i). The wide range of Mg# (14.5–40.2), Ni (<1–36 ppm) and V (11–89 ppm) contents of the studied granitoids (Table 6.3), perhaps suggest that two or more different magma sources are likely for the granodiorites and monzogranites (Kelemen 1995; Taylor and Mclennan 1985; Wang et al. 2014). This is supported by the relationships between Y/Nb with Th/Nb, Th/Ta and Ce/Pb (Fig. 6.13) which are sensitive to mantle and continental crust magma sources (Hofmann et al. 1986; Miller et al. 1994; Montero et al. 2009; Rudnick and Gao 2003; Rudnick et al. 2004; Moreno et al. 2014). The relationships of the Y/Nb with Th/Nb and Th/Ta (Fig. 6.13a, b) provide an efficient discrimination between oceanic island (OIB) and convergent margin (ACM) rock types. This is because that any change in incompatible trace-element ratios resulting from magmatic differentiation (e.g., crystal fractionation or partial melting) should be small in comparison to that which may result from two magma sources. The studied samples lie within the field of convergent margin magmatism and inside, or close to, the continental crust field (Fig. 6.13a and b). The (Y/Nb)N versus (Ce/Pb)N diagram also show that the granodiorite and the monzogranite samples have the features of a convergent margin and continental crust component (Fig. 6.13c). Therefore, these relationships (Fig. 6.13) suggest the significant involvement of a continental crust component in the granodiorites and monzogranites.

Fig. 6.12
figure 12

a Dy/Yb versus La/Yb compositional variation diagram, indicating that the studied granitoid samples inconsistent with the partial melting trend. b Sr versus Eu/Eu* and c Sr/Y versus Eu/Eu* variation diagrams, showing that fractionation of plagioclase played an important role in the differentiation of the studied I-type granitoids. d Nb/Ta versus Dy/Yb and e Dy versus Er variation diagrams, showing that fractionation of hornblende played an important role in the differentiation of the studied I-type granitoids. f V/Th versus SiO2/Al2O3 and g Sc/Th versus SiO2/Al2O3 variation diagrams, suggesting biotite fractionation between granodiorite and monzogranite samples. h TiO2/Yb versus Nb/La and i Eu/Eu* versus Nb/La diagrams for the granitoid samples, showing no correlation indicate a minor role of Fe–Ti oxides fractionation in the genesis of the granodiorites and monzogranites

Full size image
Fig. 6.13
figure 13

Relationships of (Th/Nb)N, (Th/Ta)N, and (Ce/Pb)N versus (Y/Nb)N for the studied granitoid samples from the Bulghah and Humaymah areas with compositional fields of active continental margin (ACM), ocean island basalt (OIB) and continental crust (CC) from Moreno et al. (2014). The field of Sukhaybarat I-type granitoids from Harbi et al. (2018) and the field of Jabal Ghadarah I-type granitoids from Harbi et al. (2016). Normalization values are from McDonough and Sun (1995)

Full size image

These rocks also show crustal-like trace-element patterns with negative Nb, Ta and Ti, and positive Pb anomalies (Fig. 6.7c, e, f). This is supported by Nb/Ta ratios ranging from (7.6 to 21.0, average of 14.5) that are close to the composition of magmas derived from crust and mantle (17.5; Green 1995). However, mixing model between mafic and felsic magmas needs to be confirmed by whole-rock Sr-Nd isotopes.

6 Conclusions

The following are the conclusions from our study:

  1. 1.

    Zircon U–Pb dating indicates that the mafic intrusive rocks from Bulghah and Humaymah, Saudi Arabia were formed at ~670 Ma, whereas the granitoid I-type intrusions were formed between 661 ± 5 and 643 ± 4 Ma, confirming the importance of the 700–640 Ma crustal forming event in Saudi Arabia.

  2. 2.

    Our analyses of the intrusive samples revealed no evidence of pre-Neoproterozoic zircons, further indicating that ANS crust in this region is mostly juvenile.

  3. 3.

    The studied intrusive rocks range in composition from gabbro-diorite to monzogranite. The gabbro-diorites and granodiorite-tonalite rocks are calk-alkaline, metaluminous to slightly peraluminous, whereas the monzogranite samples are classified as calc-alkaline to highly fractionated calc-alkaline.

  4. 4.

    The granodiorite-tonalite and monzogranite samples are I-type granites and classified as magnesian formed in a volcanic arc setting.

  5. 5.

    The mafic intrusive rocks (gabbro and diorite) formed in island arc setting, perhaps represent the plutonic equivalents of the Arabian Shield arc metavolcanic calc-alkaline rocks which were produced by partial melting of plagioclase- or spinel-peridotite in the upper most mantle <80 km deep in an intra-oceanic island arc.

  6. 6.

    Y/Nb with Th/Ta, Th/Nb Nb and Ce/Pb relationships indicated that the granodiorites-tonalies and monzogranites generated by a combination of mafic parental magma contaminated with crustal materials and controlled by fractional crystallization.