Neoproterozoic subduction-related metavolcanic and metasedimentary rocks from the Rey Bouba Greenstone Belt of north-central Cameroon in the Central African Fold Belt: New insights into a continental arc geodynamic setting

Geological sketch map of Cameroon

This paper presents new geochemical and geochronological data for the low to medium grade Rey Bouba Greenstone Belt (RBGB), located in northern Cameroon at the northern margin of the Central African Fold Belt (CAFB), and discusses the maximum age of volcanic activity, the maximum depositional age of metasedimentary rocks and geotectonic implications. Geochemical data on volcanic rocks highlight the predominance of transitional to calc-alkaline magma compositions (Zr/Y = 3.62–26.38) with mostly andesite to basaltic andesite with an unusually high Mg concentration (>5 wt.%, high-Mg andesite), but also basalt and trachy-andesite protoliths. Moreover, chondrite-normalized REE patterns and primitive mantle-normalized spidergrams show enrichment of LREE relative to HREE with flat to depleted elements, and moderate to slight negative Nb–Ta, Ti and Eu anomalies respectively; which are consistent with a continental arc setting related to a subduction zone. U–Pb zircon LA-ICP-MS geochronology on felsic metavolcanic or metatuff fix the maximum age for the volcanic activity in the RBGB at ca 670 Ma. Detrital zircon grains indicate that Neoproterozoic zircon (Ediacaran to Cryogenian) are the main source for the detritus that fed the basin, combined with minor Paleoproterozoic and Mesoproterozoic inputs. The maximum depositional age, corresponding to the youngest graphical age peak controlled by multiple grain ages is consistently constrained between 645 and 630 Ma, whereas the age of low grade metamorphism weakly recorded by overgrowths on detrital zircon in the RBGB basin is around 600 Ma. These results provide new insights into the geodynamic processes during the Neoproterozoic along the northern margin of the CAFB of northern Cameroon, suggesting that the RBGB, where high-Mg andesite magmatism has taken place is consistent with a continental arc-related basin

M. Houketchang Bouyoa,b,∗, Y. Zhaoa,∗∗, J. Penaye b,∗∗, S.H. Zhanga, U.O. Njel b

( note: please, you can find all details on the orginal document. see in the end of the text)

1. Introduction 

The assembly of the Gondwana Supercontinent during the Late Neoproterozoic–Cambrianinvolved closure ofthe intervening Neoproterozoic ocean basins and subduction of a substantial volume of oceanic lithosphere along a number of convergent margins (Collins et al., 2007; Santosh et al., 2009, 2012; Boger, 2011; Ngako and Njonfang, 2011). Major accretionary processes contributing to continental growth in Africa during the Neoproterozoic have also been  identified (Condie, 2003; Penaye et al., 2006; Pouclet et al., 2006; Tchameni et al., 2006; Isseini et al., 2012). The CAFB is a collage of Paleoproterozoic microcontinents and Neoproterozoic plutonic and volcanic arcs attached to the Archaean Congo craton during a Pan-African continental collision at ca 600 Ma (Toteu et al., 2004, 2006; Van Schmus et al., 2008; Bouyo Houketchang et al., 2009, 2013; Nkoumbou et al., 2013). Well-preserved Neoproterozoic magmatic arcs bounded by narrow low to medium grade volcanosedimentary schist belts have been described in northern Cameroon and in southwestern Chad Republic (Fig. 1). These volcanosedimentary sequences include the Poli, Bibemi-Zalbi and Rey Bouba greenstone belts in northern Cameroon, and Zalbi and Goueygoudoum greenstone belts in southwest Chad. In previous works, these sequences were generally interpreted as pre-tectonic back-arc basins intruded by or associated withthe calc-alkaline TTG

∗ Corresponding author at: Centre for Geological and Mining Research, PO Box 333, Garoua, Cameroon. Tel.: +237 696 215 566. ∗∗ Corresponding authors. E-mail address: mbouyo2100@yahoo.fr (M.H. Bouyo).

suite of the Sinassi and Mayo Kebbi Batholiths (Toteu et al., 1984, 1987, 2006; Pinna et al., 1994; Pouclet et al., 2006). In the Cameroonian part, where most of the geological, geochronological and isotopic data have been well-documented in the Poli, but also in the Bibemi-Zalbi belts, the Rey Bouba belt has been poorly studied and very little is known about its geochemistry, age and provenance of detrital and volcanic material. The ages and geochemistry of volcanic and sedimentary rocks from the Rey Bouba basin are therefore crucial to better understand the tectonic setting and geodynamic evolution of the CAFB along its northern margin. We present and discuss here new geochemical and laser inductively coupled plasma mass spectrometry (LA-ICP-MS) U–Pb data for zircon from metavolcanic and metasedimentary rocks to show that the Rey Bouba Greenstone Belt, was likely deposited in a continental arc setting related to a subduction zone.

2. Geological setting

 The NE–SW trending Rey Bouba Greenstone Belt is one of the volcanosedimentary belts of the CAFB in northern Cameroon (Fig. 1 and Table 1). Located in the Western Cameroonian Domain, it defines a narrow greenschist belt extending about 80 × 16 km, closely related to the low to high grade Poli Belt and continues beyond the Chadian border by the greenschist Goueygoudoum Belt and the Bibemi-Zalbi Belt. Furthermore, it is located along the Tcholliré-Banyo shear zone (TBF) interpreted as a major terrain boundary separating the younger Neoproterozoic to Mesoproterozoic Western Cameroonian Domain on the west side from the older reworked Paleoproterozoic Adamawa-Yadé Domain on the east side (Penaye et al., 1989; Pinna et al., 1994; Toteu et al., 2001, 2004; Van Schmus et al., 2008; Bouyo Houketchang et al., 2009).

Fig. 2. Geological map of the Sinassi region showing the volcanosedimentary Rey Bouba Greenstone Belt and the associated granitoids. A schematic stratigraphic column for the RBGB is inserted.

The RBGB mainly consists of greenschist-facies mafic to felsic volcanic, volcanosedimentary and sedimentary rocks associated with a set of pre-, syn- and post-tectonic granitoids and dykes. It has been defined as a back-arc basin related with an oceanic plate subduction below the southeastern continental margin of the Adamaoua-Yadé Domain (Pouclet et al., 2006; Ngako and Njonfang, 2011). However, only very few geochronological data obtained by Pb–Pb minimum ages on single zircon are available for the RBGB indicating ages of 557 ± 17 Ma for the post-tectonic Vaimba granite and 750 ± 20 Ma for the Gatougel dacitic tuff (Pinna et al., 1994).
The Poli Belt defines a pre- to syn-collisional basin developed upon, or in the vicinity of young magmatic arcs. The filling of the basin occurred in a docking-arc/back-arc context (Toteu et al., 2006). It consists of low, medium- to high-grade Neoproterozoic schists and gneisses of volcanic, volcano-sedimentary and sedimentary origin. Metavolcanics are tholeiitic basalt and calcalkaline rhyolite emplaced in an extensional crustal environment (Njel, 1986; Toteu, 1990). The depositional age is constrained between 700–665 Ma; detrital sources comprise ca. 920, 830, 780 and 736 Ma magmatic rocks (Toteu et al., 1987, 2006).
The Bibemi-Zalbi Belt extends continuously within both the Cameroonian and Chadian territories, and is locally called Bibemi and Zalbi Greenstone Belt, respectively. It is well studied in SW Chad, where it is dated around 700 ± 10 Ma on metabasalt (Isseini, 2011) and 777 ± 5 Ma on epiclastite (Doumnang, 2006). Following the assumption of Pinna et al. (1994) of an arc and back-arc basin systemlinked to a subduction beneath theAdamawa-Yadé Domain, Pouclet et al. (2006) interpreted the region in terms of fore-arc basin – volcanic arc – back-arc basin that were accreted eastward to the Adamawa-Yadé Domain, along the Tcholliré-Banyo shear zone. This model implies the subduction and closure of a western oceanic basin that was located between the Western Cameroonian Domain that could have belonged to the Central Saharan Ghost Craton (Black and Liégeois, 1993) and the remobilized Paleoproterozoic Adamawa-Yadé Domain. All these terranes are now included in the southern part of the Saharan Metacraton (Abdelsalam et al., 2002).
At the scale of the CAFB in Cameroon, the tectonic history is complex and summarized by Ngako et al. (2008) and Ngako and Njonfang (2011) in three main tectonic events related to PanAfrican collision and post-collision evolution: (i) crustal thickening (ca 630–620 Ma, and even 600 Ma, Bouyo Houketchang et al., 2009, 2013); (ii) left lateral wrench movements (613–585 Ma); and (iii) right lateral wrench movement (585–540 Ma), successively; the latter being related at the global scale to the final amalgamation of the Gondwana Supercontinent (Alkmim et al., 2001; Meert, 2003; Collins and Pisarevsky, 2005) during Latest Neoproterozoic-Earliest Cambrian.

3. General sample descriptions and procedures for geochronological and geochemical analyses 

At the local scale, the geological framework of northern Cameroon is dominated by a NE–SW extensive magmatic arc province which includes a heterogeneous and complex plutonic (TTG) and volcanosedimentary sequences (Fig. 2) that underwent a polyphase deformation characterized by sub-vertical foliation, sub-horizontal stretching lineation and folds associated with greenschist to amphibolite facies metamorphism. In this study, we focused on greenschist facies volcanic and sedimentary components of the RBGB from which forty five samples were collected (Fig. 2). From metric to multi-metric thickness, mafic metavolcanic rocks are massive, in elongated bands, flagstones, or blocks, which commonly alternate with felsic metavolcanic, metasandstone, metasiltstone, quartzite, schist and slaty shale locally highly distorted. Conglomeratic layers consisting of angular to rounded clasts of mafic volcanic, granitoid, quartzite and mineral  fragments such as quartz within a finer-grained chlorite-rich matrix are observed in Baba Sara area (Fig. 3a–d). Thirty samples were selected for thin sections, 18 for geochemistry and four for geochronology. In thin section, most of the rocks examined display poorly preserved primary magmatic and sedimentary textures superimposed by metamorphic recrystallisation fabrics (Fig. 3e–h). Major elements except FeO were analyzed on fused glass discs by X-ray fluorescence spectrometry and FeO contents by classical wet chemical analysis at the Analytical Laboratory of the Beijing Research Institute of Uranium Geology. Trace element concentrations were determined using inductively coupled plasma-mass spectrometry (ICP-MS) at the Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, Guangzhou, China. Detailed analytical methods for ICP-MS were described by Chen et al. (2010). For most of the trace elements, analytical precision and accuracy are better than 10%. Major element analyses were recalculated to 100 wt.% anhydrous basis for intercomparisons, where Fe2+ is assumed to be 80% total Fe, with the prior conversion of reported Fe concentration (FeO or Fe2O3) as Fe2O3t from appropriate atomic or molecular weights (Vocke, 1999; Verma and Armstrong-Altrin, 2013). Chondrite and primitive mantle reservoir compositions are those of Sun and McDonough (1989). Mg-number indicating the level of evolution of volcanic rock was defined as 100MgO/(MgO + FeO) in mol per cent. Being highly resistant to chemical and physical influences, zircon is a particularly usefulmineralfor petro-chronological investigations (Corfu et al., 2003; Sircombe, 2004; Vermeesch, 2004; Andersen, 2005). In order to constrain the timing of volcanic activity and depositional age of the RBGB, one felsic metavolcanic tuff, VA-R 127, and three representative metasedimentary samples including, VA-R 193, VA-R 204 and VA-R 222, were collected for detailed study and dating. Zircon grains from each sample were separated from 2 to 5 kg crushed rock samples by conventional magnetic and methylene iodide liquid separation. The separated zircon grains were handpicked and mounted in epoxy resin. The epoxy mounts were polished to expose the mounted minerals, carbon coated, photographed in both transmitted and reflected light, and imaged via Cathodoluminescence (CL) using the JSM- 6510 scanning electron microscope (SEM) at Beijing Geoanalysis Limited Company. U–Pb dating of zircon was conducted by LA-MCICP-MS (Gehrels et al., 2008; Johnston et al., 2009; Jackson et al., 2004) at the Isotopic Laboratory of Tianjin Centre of Geological Survey-China. Laser sampling was performed using a UP193-FX, and a Neptune MC-ICP-MS instrument was used to acquire ionsignal intensities. Heliumwas applied as a carrier gas for the ablated material. Trace element compositions of zircon were calibrated against Nist610 using Si as an internal standard. GJ-1 standard gem zircon from Sydney (Australia), with 207Pb/206Pb age based on 8 TIMS determinations of 609 Ma (Jackson et al., 2004), used as an external standard for U–Pb dating was mounted with samples cleaned in 1 N nitric acid immediately prior to analysis to remove surface Pb contamination, and then measured twice every eight analyses. The counting time during the analyses is 60 s (including the first 20 s for background). Fractionation and instrumental mass bias are corrected by direct calibration against a zircon standard analyzed under carefully matched conditions, using He as the ablation gas to increase the reproducibility of the Pb/U fractionation. Time-resolved data acquisition is employed to evaluate zircon homogeneity and to allow selective integration of signals to minimize common Pb contributions and Pb loss, and thus to maximize concordance. LA-ICP-MS measured values for GJ-1show that zircon is relatively low in Th, with mean U and Th contents of 230 and 15 ppm, respectively. The Calibration errors (2) on the 206Pb/238U, 207Pb/235U and 207Pb/206Pb ratio were 1.9%, 3.0%

Fig. 3. Field outcrop photographs and thin section microphotographs of various rock types of the RBGB. (a) Mafic metavolcanic rock at Mayo Rian-Mayo Vaïmba confluence; (b) Fine-grained metasandstone in Mayo Vaïmba river; (c) Metasandstone at Baba Sara; (d) Metaconglomerate around Baba Sara; (e) Felsic metavolcanic rock (VA-R 235) with abundant very small crystals (suggesting recrystallisation has occurred) of quartz ± plagioclase in the lighter bands, but also few epidote ± sericite ± calcite in darker bands; (f) Felsic metatuff (VA-R 127) with quartz + feldspar embedded in argillaceous matrix; (g) Medium metasandstone (VA-R 191) showing mature sediments with well sorted grains; (h) Metasiltstone (VA-R 166) showing a very fine-grained matrix and concentration of small muscovite flake + quartz intergrown after clasts of feldspar.
Fig. 4. CaO/Al2O3–MgO–SiO2 diagram showing geochemically unaltered nature of most of the metavolcanic rock samples of the RBGB. 

and 2.4%, respectively and propagated through the error analysis. During the experiment, sites for dating were selected on the basis of CL and photomicrograph images, in order to obtain as representative a population as possible. Laser energy of 10–11 J/cm2, a frequency of 8 Hz and a spot size of 35 m were used for the instrument. Off-line selection, integration of background, analyte signals, time-drift correction and quantitative calibration for trace element analyses were performed by ICPMSDataCal (Liu et al., 2008). Concordia diagrams, histograms overlain by probability density distributions and weighted mean calculations were made using Isoplot (Ver3.23) (Ludwig, 2003) to show the distribution of the zircon populations over the full age range of all grains (see text, Tables 3–6 and Figs. 6–8 for details on individual samples).

 4. Geochemical results 

Major and trace element data of the investigated rocks from the RBGB are presented in Table 2 (complete data set may be accessed from the online data repository (see Appendix A)). Postdepositional processes as well as low-temperature metamorphism (metasomatism) are known to affect the mobility of certain elements (Cs, Rb, Li, B, Ba, Sr, Si,K, etc.). Index of alteration also includes the degree of hydration (i.e. Loss on Ignition which it is generally less than 1%). Some of the analyzed samples in the present study have Loss on Ignition (LOI) values higher than expected (Table 2). The data were, therefore, plotted in the CaO/Al2O3–MgO–SiO2 diagram, proposed by Schweitzer and Kröner (1985) to test alteration effects in metavolcanic rocks. As seen, majority of the samples plot in the field of “unaltered rocks” except few samples with highest LOI (VA-R 223 and VA-R 127), which also show lowest (0.35 wt.%) and highest CaO (6.39 wt.%) contents respectively (Fig. 4; Table 2). Both mafic to intermediate and felsic metavolcanic rocks of the RBGB are characterized by magmatic compositions ranging from andesite/basaltic andesite to trachy-andesite. Only one sample clearly shows basaltic composition in the bivariate Zr/Ti versus Nb/Y diagram (Fig. 5). In addition, they show (Zr/Y = 3.62–26.38) ratios generally comparable to those of transitional (Zr/Y = 4.5–7.0) and calc-alkaline volcanic compositions. On the basis of petrographic characteristics, major elements, and chondrite and primitive mantle normalized rare earth and trace element patterns, they have been subdivided into three major groups (Fig. 6; Table 2): mafic to intermediate metavolcanic, felsic metavolcanic and metasedimentary rocks.

Fig. 5. Zr/Ti versus Nb/Y classification diagram for metavolcanic rocks of the RBGB (Table 2). Metasedimentary rocks are also plotted for intercomparisons. Compositional fields revised by Pearce (1996) after Winchester and Floyd (1977).

 4.1. Mafic to intermediate metavolcanic rocks

 Mafic to intermediate metavolcanic rocks (Fig. 5) are enriched in MgO (2.89 and 6.19 wt.%), moderate to strong fractioned with a variable Mg-number (Mg#) between 38 and 65% (Table 2). They have low to medium SiO2 (50.95–63.00 wt.%) contents and relatively high TiO2% (0.65–1.03), MgO% (2.89–6.19), and CaO% (0.35–8.86) respectively with an average of (0.83, 4.50, and 4.64 wt.%; Table 2). The relatively high concentrations of transition metals V (up to 332 ppm), Cr (249 ppm), Ni (162 ppm), Cu (236 ppm), Sc (28 ppm) and total REE (313 ppm) are noticeable (Table 2). On chondrite and primitive mantle normalized diagrams, they display enrichment in LREE and slight depletion in HREE with no Eu anomaly (Fig. 6a and Table 2); but also erratic distribution in the mobile LILE (Rb, Ba, K) and negative anomalies in Nb-Ta and Ti (Fig. 6b). Inter-element ratios Nb/Ta and (La/Yb)N give values of 0.18–0.58 and 1.92–39.00 (with an average of 10.62), respectively.

4.2. Felsic metavolcanic rocks

 Felsic metavolcanic rocks display the highest SiO2 (69.51–75.37 wt.%) and lowest MgO (0.41–1.46 wt.%) contents (Table 2). They are also characterized by low concentrations oftransition metals: Ti (0.11–0.49 wt.%), Sc (4–13 ppm), V (14–62 ppm), Cr (2–22 ppm), Ni (1–12 ppm) and total REE (33–68 ppm). Felsic metavolcanic rocks exhibit enriched LREE and flat to depleted HREE patterns (Fig. 6c). The trace element patterns of this group are relatively enriched in LILE (Rb, Ba, K) and have negative anomalies in Nb-Ta, P and Ti (Fig. 6d); but variable Eu anomalies (Eu/Eu* = 0.60–1.31; Table 2). Nb/Ta and (La/Yb)N ratios are between (0.13–0.38) and (1.45–28.35 with an average of 12.48) respectively.

4.3. Metasedimentary rocks

Metasedimentary rocks, compared with the previous groups show variable contents of SiO2 (58.61–66.86 wt.%), Mg (2.36–3.71 wt.%); transition metals Ti (0.57–1.00 wt.%), Sc (13–20 ppm), V (88–131 ppm), Cr (65–114 ppm), Ni (34–64 ppm) and total REE (65–157 ppm) (Table 2) between the highest values for mafic metavolcanic rocks and the lowest for felsic metavolcanic rocks and vice versa. However, Ba and Zr have the highest contents, reaching respectively up to 1739 ppm and 212 ppm. The chondrite normalized plots show enrichment in LREE relative to HREE which have a flat pattern and slight fractionated or negative Eu anomaly (Fig. 6e; Table 2). Regarding the primitive mantle normalized diagram, they show enrichment in LILE (Rb, Ba, K) and negative anomalies in Nb-Ta and Ti (Fig. 6f). Their Nb/Ta and (La/Yb)N ratios are respectively (0.33–0.78) and (2.66–8.35 with an average of 5.61).

 Fig. 6. Chondrite and primitive mantle normalized diagrams for (a and b) Mafic to intermediate metavolcanic rocks, (c and d) Felsic metavolcanic rocks and (e and f) Metasedimentary rocks.

5. Geochronological results 

In order to give an adequate, accurate and precise feature of the age distribution ofthe volcanic and sedimentary rocks of RBGB, several hundred grains of zircon were selected randomly from each sample. Among the four samples selected for dating, 318 zircon crystals were individually analyzed in three samples, and only 26 crystals saved for the fourth; while a maximum of 120 representative target spots from each sample were located using CL images. In total, 394 analytical spots were performed on 344 zircon grains.

5.1. Sample VA-R 127: Felsic metatuff 

This sample is interpreted as a felsic metavolcanic rock or metatuff comprising a mixture of fragments and euhedral to subeuhedral fine- to medium-grained quartz and feldspar crystals embedded in a fine-grained matrix. Collected near Vaïmba village (Fig. 2), and generally displaying a grey colour or grey yellowish to brownish weathered surfaces, the rock has very fine-grained illite, sericite and chlorite associated with sub-angular to sub-rounded quartz, and minor feldspar, epidote and oxides. Late stage microfractures or veinlets filled by calcite are also locally observed. Zircon grains from this sample are light-pink to brownish in colour. They are commonly prismatic, but occasionally rounded or fragmentary in shape, possibly related to different populations. The grain size varies from ca 80 to 230 m, the majority being between ca 105 and 115 m. On CL images (Fig. 7a), all the grains display igneous features with oscillatory or sector zoning. In most cases, grains show different colours in grey, with faint, broad zoning, and in fewer cases, they are unzoned. Overgrowths are very rare to absent. A number of 110 points were analyzed in situ on 107 grains. Most of the analyzed zircon grains (92%) show 232Th/238U ratios higher than 0.2 (Table 3, complete data set may be accessed from the online data repository (see Appendix A)), pointing to a magmatic origin (Williams and Claesson, 1987). From the data, two main groups of concordant and discordant zircon (Fig. 8a and Table 3) are observed. The concordant group is very clustered (inset on Fig. 8a) and display three distinct sub-groups or populations on the probability density distribution and histogram diagrams (Fig. 8b), with a high concentration of 206Pb/238U ages at ∼670 Ma (33 analyses; median = 668 ± 8.5 [1] Ma), at ∼700 Ma (27 analyses; median = 697 ± 3.1 [1] Ma) and at ∼760 Ma (5 analyses; median = 758 ± 2.8 [1] Ma). The discordant group shows signifi- cant episodic loss of lead which may be related either to analytical issues or to relatively high U contents of analyzed zircon grains (dark domains, Fig. 7a: 127–3 and Table 3: 127–3: 713 ppm, 22: 371 ppm, 85: 476 ppm, 94: 253 ppm, 98: 451 ppm, etc.). However they yield an upper intercept of 1431 Ma and lower trajectory to ca 0 Ma (Fig. 8a).

5.2. Sample VA-R 193: Schistose sandstone 

RockVA-R193 is a light- tomedium-grey, schistose sandstone or metasiltstone from Baba Sara, showing a very fine-grained groundmass of silica and clayey minerals in association with clasts of quartz, pyrite and iron oxides. The unit is locally affected by recrystallisationfabrics due to intense deformation, and interbedded with felsic metavolcanic rock and quartzite bands, and commonly contains lithic fragments giving it the appearance of conglomeratic volcanosedimentary schist. Detrital grains of zircon range in size mostly from ca 110 to 200 m, with pale pink to yellowish colour. Grains are euhedral to sub-euhedral, mostly displaying oscillatory or sector zoning typical of igneous rocks (Fig. 7b). Moreover, some individual crystals show strong variations in the development of zoned domains, where one large uniform central zone is succeeded by much finer oscillatoryzoned bands. A total of 120 spots on 105 zircon grains were analyzed, with 94% indicating 232Th/238U ratios higher than 0.2 (Table 4, complete data set may be accessed from the online data repository (see Appendix A)). The U–Pb isotope data obtained from the zircon grains show a wide range of ages almost all concordant (115 analyses) between 529 ± 3 Ma and 906 ± 5 Ma (Fig. 8c) with a peak at 645 Ma representing a dominantly Neoproterozoic source (Fig. 8d). The presence of Late Mesoproterozoic and Paleoproterozoic components is also indicated by two zircon ages at 1016 ± 6 Ma (193–32), and 2077 ± 12 Ma (193–119) respectively. Five discordant data (193–10, 12, 48, 52 and 104; Table 4) are not considered.

5.3. Sample VA-R 204: Chl-schistose sandstone 

Sample VA-R 204 is a greenish grey to dark green, fine-grained chlorite-bearing schistose sandstone or volcanoclastic schist, collected near the river Mayo Godi, about 6 km NW from Baba Sara. The rock consists essentially of quartz and chlorite in a cryptocrystalline matrix. It is locally strongly deformed and interbedded with well-foliated slaty schists. Zircon grains are euhedral to sub-euhedral, and commonly vary in size from ca 135 to 180 m and up to 235 m in places. They are mostly pale pink and display morphologies and internal structures characterized by more or less well-preserved magmatic growth zoning (Fig. 7c). These grains for the most part have less blunted or eroded terminations, suggesting a short distance of transportation prior to deposition, while very few show thin overgrowths. Some 44 laser ablation analyses were performed on 26 crystals. 232Th/238U ratios range from 0.03 to 1.67 with the majority (80%) higher than 0.2 (Table 5, complete data set may be accessed from the online data repository (see Appendix A)). Data are plotted in Fig. 8e. Most analyses are concordant defining three main groups with 206Pb/238U ages ranging from 585 ± 4 to 876 ± 6 Ma (36 analyses, insert Fig. 8e) for the first group, one analysis at 1259 ± 10 Ma for the second group, and the third group at ca. 2000 Ma (four analyses). A predominant Neoproterozoic source is shown by a broad probability density distribution from Ediacaran to Tonian with maximum peaks at 663 and 700 Ma, but also at 682, 630 and 600 Ma (Fig. 8f). This sample also confirms the presence of older Mesoproterozoic and Paleoproterozoic sources for the sediments.

5.4. Sample VA-R 222: Ferruginous metasandstone

 The rock specimen is brown to ocher, probably related to its dominantly ferruginous cementingmaterial.Itis afine- tomediumgrained ferruginous metasandstone, collected upstream of the Mayo Beoulpir river. Minerals identified in thin section include principally sub-angular to sub-rounded quartz, but also plagioclase, calcite, pyrite, iron oxides, epidote, illite, chlorite and indistinct grains of brownish altered ferromagnesian mineral. Folded veinlets of quartz and calcite are also observed. Detrital crystals of zircon are prismatic to round, and pink, yellowish, or brown in colour. Crystals are sub-euhedral to euhedral, range in size from ca 75 to 200 m, with the majority between 110 and 150 m, and mostly show a well-developed oscillatory to sector zoning (Fig. 7d). The differences in morphology and colour among zircon may reflect distinct sources, and the small degree of abrasion suggests a proximal provenance. A whole of 120 measurements were performed in situ on 106 zircon grains, with 99% displaying 232Th/238U ratios higher than 0.2 (Table 6, complete data set may be accessed from the online data repository (see Appendix A)). Concordia diagrams (Fig. 8g) show scattered discordant points with relatively high U contents (dark domains, Fig. 7d: 222–13 and 26; Table 6: 222–13: 172 ppm, 26:

Fig. 7. Cathodoluminescence images of zircon from representative metavolcanic and metasedimentary samples of the RBGB basin: (a) felsic metatuff, VA-R 127; (b) schistose sandstone, VA-R 193; (c) chlorite-bearing schistose sandstone, VA-R 204 and (d) ferruginous metasandstone, VA-R 222. The figure also shows circles indicating analyzed spots (not at scale), with numbers representing 206Pb–238U age and spot position listed respectively in Tables 3–6.

227 ppm, 29: 156 ppm, 30: 240 ppm, 65: 327 ppm, 117: 326 ppm, etc.) possibly due either to analytical issues or to common lead or extreme lead loss; and concordant points (56 analyses, diagram insert) all yielding Neoproterozoic 206Pb/238U ages from 574 ± 4 to 893 ± 6 Ma. The combined probability density distribution and histogram diagrams on concordant points show a maximum peak at 660 Ma, but also at 740, 795 and 848 Ma (Fig. 8h).

6. Discussion

6.1. Age of the volcanic activity

U–Pb zircon LA-ICP-MS geochronology on a felsic metavolcanic rock or metatuff (sample VA-R 127; insert Fig. 8a) is used to constrain the timing of magmatic activity within the RBGB. On CL images (Fig. 7a), zircon grains from VA-R 127 show typical oscillatory growth or sector zoning and are characterized by high 232Th/238U ratios (Table 3) indicating a magmatic origin. Concordant zircon grains (65 analyses) reported on Fig. 8 a (insert) and b define three peaks (ca 760, 700 and 670 Ma) which we interpreted as the times of different magmatic sources in the region. They indicate that the magmatic activity in the RBGB is prominent and took place in the Neoproterozoic. The peak at ∼760 Ma reflects an earlier inherited magmatic event which is consistent with the Pb/Pb age of 750 ± 20 Ma on Gatouguel dacitic tuff (Pinna et al., 1994), whereas the youngest peak at ∼670 Ma fix the maximum age for the volcanic activity in the RBGB. From previous studies in the Poli Belt by Toteu (1990) and Toteu et al. (2001, 2006), a wide range of Neoproterozoic detrital zircon (Tonian–Cryogenian) at 700, 780, 830, and 920 Ma have been recorded (Table 1), some of which correspond to well identified magmatic rocks originating from neighbouring areas (e.g., Goldyna metarhyolite, medium-grade schists and underlying volcanic sequence). In the Chadian part of the Bibémi-Zalbi Belt, Isseini (2011) constrains a crystallization U–Pb age on metabasalt at 700 ± 10 Ma, quite different from the previous U–Pb age of 777 ± 5 Ma derived from an epiclastite by Doumnang (2006). The distribution of ages implies at least three major periods or episodes, representing formation times of volcanic rocks at the northern margin of the CAFB: (1) an early volcanic activity from ca 920 to 830 reported only in the Poli Belt,(2) a second from ca 780 to 700 reported in the Poli and Bibémi-Zalbi belts, and (3) a third volcanic event around 670 Ma or younger reported in the RBGB belt. These results show that the volcanic activity is older in the Poli and Bibémi-Zalbi belts and younger in the RBGB.

 6.2. Provenance of the sediments

 LA-ICP-MS U–Pb detrital zircon dating presented here provides new constraints onthe geological evolutionoftheRBGB, dominated by Neoproterozoic zircon with some Paleoproterozoic, Mesoproterozoic, and Early Cambrian inputs. Over 70% of zircon grains have Neoproterozoic (Ediacaran to Tonian) ages ranging from ca 574 to ca 906 Ma (samples VA-R 193, VA-R 204 and VA-R 222), with major

Fig. 8. U–Pb concordia and probability density distribution-histogram diagrams for detrital zircon analyses from representative metavolcanic and metasedimentary samples of the RBGB basin: (a and b) felsic metatuff, VA-R 127; (c and d) schistose sandstone, VA-R 193; (e and f) chlorite-bearing schistose sandstone, VA-R 204 and (g and h) ferruginous metasandstone, VA-R 222.

peaks at 645, ca 660 and 700 Ma (Fig. 8d, f and h) but also at 600, 630, ca 680, 740, ca 800 and ca 850 Ma (Fig. 8f and h); indicating that the Neoproterozoic (Ediacaran to Cryogenian) was the main source for the Rey Bouba basin. These ages correspond well with the U–Pb and Ndmodel ages ofthe juxtaposed NeoproterozoicWestern Cameroonian Domain (granitoids and volcanosedimentary sources at ca 595, 600, 615, 630, 660, 693, 736, 750, 780, 830, and 920 Ma; Table 1; Toteu et al., 2001, 2004, 2006; Bouyo Houketchang et al., 2009; Dawaï et al., 2013) as well as the Pan-African Mayo Kebbi Batholith or magmatic arc of SW Chad (ca 640, 665 and 740 Ma; Table 1; Penaye et al., 2006; Isseini et al., 2012). Looking at the high proportion (more than 90%) of 232Th/238U ratios higher than 0.2, this implies that the predominant sources of Neoproterozoic detritus that fed the RBGB basin were derived from erosion of the adjacent magmatic arc and Western Cameroon Domain or Block. In this regard, our data clearly demonstrate that northern Cameroon volcanosedimentary basins at the northern margin of CAFB are younger than the magmatic arc, and not intruded by the latter as suggested by previous works (Pinna et al., 1994; Penaye et al., 2006; Pouclet et al., 2006). However, there is also minor input of Paleoproterozoic (Orosirian to Rhyacian) zircon at ca 2000 Ma and 2077 ± 12 Ma, recorded on samples VA-R 193 (Fig. 8c) and VA-R 204 (Fig. 8e). Rare Mesoproterozoic (Stenian to Ectasian) zircon at 1016 ± 6 Ma (sample VA-R 193; Fig. 8c) and 1259 ± 10 Ma (sample VA-R 204; Fig. 8e), and very rare (less than 1%) Early Cambrian (Fortunian) zircon at 529 ± 3 Ma (sample VA-R 193; insert Fig. 8c) also occur. Paleoproterozoicdetrital zirconreportedfrommetavolcanosedimentary schist of the RBGB may be correlated with the nearby gneisses of Adamawa-Yadé Domain (e.g. Mayo Makat and Mayo Kout gneisses at 2218 ± 14 Ma, and also Mbé gneiss at 2014 ± 7 Ma, Penaye et al., 1989; Mayo Kout orthogneiss at 2039 ± 26 Ma, Bouyo Houketchang et al., 2009; Table 1). The Mesoproterozoic zircon may be regarded as evidence of eroded rocks that remain unknown or hidden, since the area is not mapped in detail. The Neoproterozoic ages at ca 600 Ma (sample VA-R 204, Fig. 8f) agree with tectonometamorphic ages previously reported from volcanosedimentary basins within the CAFB (zircon U–Pb at 620 ± 10 Ma, Yaoundé: Penaye et al., 1993; composite Sm–Nd isochron at 616 Ma, Bivouba-Yaoundé: Toteu et al., 1994; Pb–Pb zircon evaporation at 611 ± 20 Ma, Yaoundé: Stendal et al., 2006; zircon U–Pb at ca 600 Ma, Lom: Toteu et al., 2006; Sm–Nd garnet-whole rock at 628 ± 68 Ma, Bafia: Numbem Tchakounté et al., 2007; zircon U–Pb and Sm–Nd garnet-whole rock at ca 600 Ma, Banyo and Tcholliré: Bouyo Houketchang et al., 2009, 2013; monazite U–Th–Pb at 622 ± 43 Ma, Boumnyebel-Yaoundé: Yonta-Ngouné et al., 2010; at 613 ± 33 Ma, Yaoundé: Owona et al., 2011). The Early Cambrian ages are likely to be minimum estimates for the formation of RBGB or may represent post-metamorphic cooling.

6.3. Age of metamorphism 

Although very rare to absent on CL images (Fig. 7), younger metamorphic overgrowth rims around older magmatic cores are locally present and characterized by 232Th/238U ratios lower than 0.2 (less than 9% of the entire samples; Tables 3–6). Except for a few very young ages (e.g. 288 Ma or 432 Ma, VA-R 127–3 and 127-24 respectively; Table 3), which may be due to calibration problems or lead loss after deposition, about 67% of this metamorphic population yield Late Neoproterozoic ages, most often discordant. However, the youngest and best example to constrain the age of metamorphism is given by sample VA-R 204-24 with a concordant 206Pb/238U age of 600 ± 4 Ma, and 232Th/238U ratio of 0.03 (Table 5). This result is consistent with the ca 600 Ma age of metamorphism obtained from high pressure granulites (Bouyo Houketchang et al., 2009, 2013) in Banyo and Tcholliré (Poli Belt) 0 10 20 30 40 50 60 500 600 700 800 900 1000 1100 Age (Ma) Number All of the metasedimentary rocks; n=207 645

 Fig. 9. Synthesis diagram displaying the combined probability density distributionhistogram for the geochronological data from the three metasedimentary samples. within the CAFB of north-central Cameroon (Table 1). This is the most direct evidence for the timing of continental collision. 

6.4. Maximum depositional age of the RBGB basin

The maximum depositional age for Precambrian stratigraphic units lacking preserved biostratigraphic age control such as fossils is commonly constrained by the youngest U–Pb ages of zircon grains in populations of detrital zircon (Jones et al., 2009). As mentioned in previous sections, Neoproterozoic ages around 600 Ma and Early Cambrian ages are best interpreted as tectonometamorphic ages and post metamorphic cooling respectively, and therefore, they were not considered in the detrital zircon populations. The maximum depositional age for each sample is defined here as the 206Pb/208U age of the youngest dated 99.5% to 100.5% concordant zircon at 621 ± 5 Ma in the case of sample VA-R 204- 23 (Table 5; Fig. 8e); 631 ± 4 Ma in the case of sample VA-R 222-16 (Table 6; Fig. 8g), and 626 ± 4 Ma in the case of sample VA-R 193-13 (Table 4; Fig. 8c). Our results show somewhat different youngest ages from one sample to another, and so individually are not yet considered to represent actual depositional ages. Dickinson and Gehrels (2009) tested the research strategy of using youngest U–Pb ages of detrital zircon to constrain the maximum depositional ages of strata containing zircon grains by comparing a total of 5365 concordant or nearly concordant U–Pb ages of detrital zircon in 58 samples of Mesozoic sandstone from the Colorado Plateau and adjacent areas with depositional ages known independently from biostratigraphy. Their analysis confirms that using youngest detrital zircon ages to constrain maximum depositional age is a valid procedure, but indicates that results vary somewhat depending upon how the youngest grain age is specified. Moreover, they add that in general, the youngest-age measurements based on multiple grain ages are more consistently compatible with depositional ages. In our approach to constrain the maximum depositional age, we consider the geochronological data from the three metasedimentary samples as a composite sample of detrital zircon, in order to do statistical comparisons of similarity between zircon populations within the RBGB. From a total of 284 spots on 237 grains analyzed, 207 are considered significant in terms of concordance and have been used to construct a probability density distribution-histogram diagram (Fig. 9; Tables 4–6). From Fig. 9, a single peak controlled by multiple ages is defined at 645 Ma,representing the main provenance source of the detrital zircon, as well as the youngest probability density distribution peak. In light of Figs. 8f and 9, the maximum depositional age can be consistently constrained between 645–630 Ma, whereas 600 Ma represents the age of low grade metamorphism recorded by the detrital zircon overgrowths from the RBGB basin.

 6.5. Petrogenesis of volcanic and sedimentary rocks 

Knowledge of the tectonic setting of an ancient basin is very important for understanding the geodynamic evolution and associated mineral resources, as well as the paleogeographic environment. Although there are many discrimination diagrams for ultrabasic and basic tectono-magmatic, and sedimentary (Bhatia, 1983; Roser and Korsch, 1986; Armstrong-Altrin and Verma, 2005) settings in the literature with success rate sometimes controversial for a specific geodynamic context, very few or almost none exist for intermediate composition rocks. In this work, we constrain the RBGB basin tectonic setting based on new sets of multidimensional tectonic discrimination diagrams for intermediate magmas from five tectonic settings (island arc-IA, continental arc-CA, continental rift-CRand ocean island-OI combined together, and collision-Col) of Verma andVerma (2013), but also for sedimentary rocks from three tectonic settings (arc, continental rift, and collision) of Verma and Armstrong-Altrin (2013); all based on worldwide examples with high success rates of about 63–100%. In two different sets of five discrimination diagrams for intermediate rocks designed to discriminate among five tectonic settings based on major-elements (Fig. 10A, accessible from the online data repository (see Appendix A)) and immobile trace-elements (Fig. 10B, accessible from the online data repository (see Appendix A)), about 80% of our samples (mafic to intermediate metavolcanic rocks) fall in the arc tectonic setting field consistent with the dominantly transitional to calc-alkaline andesite to basaltic andesite geochemical signatures of the studied rocks (Fig. 5). Similarly, in the new multi-dimensional major-element based diagrams for tectonic discrimination of siliciclastic sediments from three main tectonic settings, the RBGB metasedimentary rocks mostly plot in the arc tectonic setting field for high-silica (Fig. 11a, accessible from the online data repository (see Appendix A)) and low-silica (Fig. 11b, accessible from the online data repository (see Appendix A)) contents. Moreover, the chondrite and primitive mantle normalized diagrams show respectively enrichment in LREE relative to HREE and moderate to slight negative Nb-Ta, Ti and Eu anomalies for all analyzed samples (Fig. 6); compatible with a subduction zone. Thus an arc setting related to convergence and subduction can be clearly inferred for the northern margin of the CAFB during the Neoproterozoic (645–630 Ma), and therefore indicates the depositional environment of the RBGB basin prior to collision at ca 600 Ma.

6.6. Geodynamic interpretation 

As mentioned above, mafic to intermediate metavolcanic rocks from the RBGB show geochemical features similar to those of wellstudied high-Mg andesite (HMA) from Mitsuiyama (Okoppe area of North Hokkaido in Japan, with MgO contents ranging from 4.9 to 7.9 wt.%, Ayabe et al., 2012). These include: andesitic composition (Fig. 5) with high contents of MgO (2.89–6.19% or Mg-number between 38–65% with an average of 52; Table 2), high concentrations of transition metals V (up to 332 ppm), Cr (249 ppm), Ni (162 ppm), Cu (236 ppm) and Sc (28 ppm), relatively low (La/Yb)N ratios between 1.92–39.00 with an average of 10.62, depleted rare earth elements (Fig. 6), and also high field strength elements (Nb, Ta, Ti), which are indicators of arc volcanic rocks developed in continental setting (Figs. 5, 10 and 11). This is supported by the detrital sources that fed the RBGB which are derived predominantly from erosion of the adjacent Neoproterozoic Mayo-Kebbi and Sinassi juvenile magmatic arcs of the Western Cameroonian Domain and from the Paleoproterozoic Adamawa-Yade Domain. Therefore, the RBGB may correspond to an extensional basin developed upon or behind a Neoproterozoic magmatic arc. Fig. 12 (accessible from the online data repository (see Appendix A) illustrates a subductionrelated geodynamic setting model for the RBGB (Fig. 12A) in the network of Neoproterozoic magmatic arcs (Fig. 12B) of northcentral Cameroon in the CAFB. Although the andesitic rocks from the RBGB are Mg-rich, their geotectonic setting isquitedifferentfromthat oftheprimitivehighMg andesite as defined by Wood and Turner (2009) or by Ayabe et al. (2012) in the Okoppe volcanic field of North Hokkaido, Japan. Finally, it seems evident that there are different types of HMA in subduction-related settings; some are primitive and result from interaction of a melt derived from subducted oceanic basaltic crust and the overlying mantle wedge peridotite (Shiraki et al., 1978; Meijer, 1980; Crawford et al., 1981; Tatsumi and Ishizaka, 1982; Xu et al., 2000) and others aremore evolved and resultfrominteraction of melt derived from subducted oceanic crust and the continental magmatic arc.

 7. Conclusions 

The main objective of our work was to acquire new geochemical and geochronological data from the low to medium grade volcanosedimentary Rey Bouba Greenstone Belt (RBGB) of northern Cameroon at the northern margin of the Central African Fold Belt (CAFB), in order to contribute for a better knowledge and understanding of tectonic settings and geodynamic evolution during its Precambrian history. The overall results of our work show that: - Petrogenesis of mafic to intermediate metavolcanic rocks from the RBGB is consistent with the dominantly transitional to calcalkaline andesite to basaltic andesite, indicating geochemical features that are similar to typical high-Mg andesite including: andesitic composition with high contents of MgO (2.89–6.19% or Mg-number between 38 and 65%), high concentrations of transition metals V (up to 332 ppm), Cr (249 ppm), Ni (162 ppm), Cu (236 ppm) and Sc (28 ppm), relatively low (La/Yb)N ratios between 1.92 and 39.00 with an average of 10.62, depleted rare earth elements, and also in high field strength elements (Nb, Ta, Ti), which are suitable signatures of arc volcanic rocks; - U–Pb zircon LA-ICP-MS geochronology on felsic metavolcanic or metatuff fix the maximum age for the volcanic activity in the RBGB at ∼670 Ma, thus younger than in the neighbouring Poli and Bibémi-Zalbi belts; - the dating of detrital zircon indicate that the main source for the detritus that fed the RBGB is Neoproterozoic (Ediacaran to Cryogenian) with minor Paleoproterozoic and Mesoproterozoic inputs from the neighbouring Mayo Kebbi-Sinassi magmatic arc and Adamawa-Yade Domain; - the age of low grade metamorphism weakly recorded by younger metamorphic overgrowth rims around older magmatic cores, characterized by 232Th/238U ratios lower than 0.2 is 600 Ma; - the maximum depositional age of the RBGB, corresponding to the youngest graphical age peak controlled by multiple grain ages is consistently constrained between 645 and 630 Ma. Our data therefore provide new insights into the geodynamic processes during the late Neoproterozoic, suggesting thattheRBGB, where high-Mg andesite magmatism has taken place is consistent with a continental arc setting related to a subduction zone.

Source:

M. Houketchang Bouyo, Y. Zhao, J. Penaye, S.H. Zhang, U.O. Njel, 2015.Neoproterozoic subduction-related metavolcanic and metasedimentary rocks from the Rey Bouba Greenstone Belt of north-central Cameroon in the Central African Fold Belt: New insights into a continental arc geodynamic setting. 

Precambrian Research 261:40-51.