1 Introduction

The emplacement of mafic dykes in any plutonic or metamorphic basement is very important for the understanding of geological events over time as they act as an important time marker in the geological record [1]. The geochemistry of mafic dykes also provides important information for the proper understanding of the origin of magmas, their melting conditions and tectonic events in subduction zones [2, 3]. The sources of magma for mafic or intermediate dykes emplaced within the crust are multiple and include among others, the sub-continental lithospheric mantle, plume–lithosphere interaction and the contamination of mantle magma by the crust [4,5,6,7]. Moreover, it is also demonstrated that fluids and magma generated from subducted plates and subducted sediments can also participate in the generation of mafic dykes’ magma in a convergent zone [8]. Several geodynamic processes have been shown to explain the formation of dykes. These include: (1) lithospheric distension or plate opening [9,10,11,12]; (2) subduction of oceanic plate [13, 14]; (3) arc collision [15, 16]; and (4) continental collision [17].

Unlike the volcanic edifices of the Cameroon Volcanic Line (CVL), the Central Cameroon Shear Zone (CCSZ) consisting mainly of granites, gneisses and mylonites hosts very few mafic rocks. Despite this rarity, some mafic dykes have been discovered in the Magba [18] and Njimom [19] areas in the CCSZ and their origin and tectonic setting have been discussed. However, in the Mfengou-Manki area, columnar jointed basalts were not observed. Apart from the work of Kouamo Keutchafo et al. [19] who studied only the mafic dykes situated along the Foumban-Njimom highway, no author has ever studied the columnar jointed basalts of the Mfengou area because it was situated far from the main highway and hidden by a relative soil cover and vegetation. The outcrop was discovered in 2016 and was used as quarry for the Foumban-Manki road construction. Moreover, the petrology, and the extension of the mafic dykes from the Mfengou-Manki were not yet well known. This work aims firstly to study the petrogenesis and tectonic setting of the mafic dykes and the columnar jointed basalts, and secondly to study their probable rock source using the ICP AES, ICPMS and electron microprobe analytical methods.

2 Geological context

The geology of Cameroon is dominated by two main geological groups, namely the Central African Fold Belt in Cameroon (CAFB) and the Congo Craton (CC). The CAFB in Cameroon hosts the current study area and is subdivided into three domains namely, the Northwest Cameroon Domain (NWCD), the Yaounde Domain (YD), and the Adamawa Yade Domain (AYD) (Fig. 1). Located in the south of Cameroon, the Yaounde Domain is bounded in the south by the Congo Craton and in its northern sector by the Sanaga Fault (SF). The geological formations found in this area are pre- to syn-tectonic intrusive granitoids and mafic dykes with alkaline and transitional affinities and metasediments [20]. Its southern side is marked by the predominance of mafic dykes and/or serpentinized ultramafic rocks, gabbros and diorites [21]. The Adamawa Yade domain lies between the Sanaga Fault in the south and the Tchollire-Banyo Fault (TBF) in the north (Fig. 1). It is the domain that hosts the study area and consists of large Pan-African intrusions and metamorphosed Palaeoproterozoic relics emplaced during the Pan-African tectonic evolution. The Northwest Cameroon Domain (NWCD) also known as the Poli Group [22, 23] is located in the North Cameroon. It is limited in the south by the Tchollire-Banyo Fault and consists of Meso to Neo-Proterozoic volcano-sedimentary basins variably metamorphosed into gneisses and schists [24]. In places, it is covered by pre to syn- or late-tectonic intrusive calc-alkaline formations (granites, granodiorites and diorites) and metamorphosed volcanic and detrital deposits [23,24,25].

Fig. 1 
figure 1

Geological map of Cameroon modified after Toteu et al. [83] and Ntieche et al. [27]. The Central African Shear Zone is defined by a system of NE-trending faults comprising Tchollire-Banyo Fault (TBF), Adamawa Fault (AF), Sanaga Fault (SF), and Kribi-Campo Fault (KCF). The inset is the map of the African continent, showing the location of Cameroon relative to the distribution of cratons and mobile belts

Full size image

Located in the southern part of the Adamawa Yade domain (Central domain) (Fig. 1), precisely within the Central Cameroon Shear Zone (CCSZ), the Foumban-Magba area (hosting the study zone) is made up of granitoids subdivided into granites, gneisses, mylonites and migmatites and few mafic to intermediate rocks [18, 26,27,28,29]. The Central Cameroon Shear Zone is a Pan-African mega structure made up of lineaments oriented N70E and extending from Sudan into northeastern Brazil [29]. This fault network allowed the upwelling of the mafic rocks magma following the C2 shear axes and their emplacement within the granite and gneiss substratum [18]. The mafic rocks are high-K calc-alkaline to shoshonitic and are situated in the Magba area within the Foumban-Magba Zone. They are also found in the Kekem and Nyos area at the south of the CCSZ. In Magba and Kekem, they were emplaced in a subduction environment and were formed from the sub-continental lithospheric mantle with the participation of a crustal component [18]. The Mfengou-Manki area consists of granites, migmatites, gneisses and mylonites which locally host mafic to intermediate rocks not yet studied in detail (Fig. 2).

Fig. 2
figure 2

Geological sketch map of the study area

Full size image

3 Analytical methods

Representative fresh rock samples were collected during fieldwork for petrography, geochemistry and mineral chemistry studies. From the fresh samples collected in the field, thirty (30) thin sections were made. Twenty-four (24) samples (17 columnar jointed basalts, 07 mafic dykes), were selected for the whole rock geochemistry at the commercial BVM, ACME Lab, Vancouver (Canada) using inductively coupled plasma-atomic emission (ICP-AES) for major elements and inductively coupled plasma-mass spectrometry (ICP-MS) for trace elements and REE. A mixture of 0.2 g of each sample and 0.9 g of lithium metaborate/lithium tetraborate flux was prepared and melted in an oven at 1000 °C. The resulting melt was cooled and dissolved in 100 ml of 4% nitric acid and 2% hydrochloric acid. After dissolution, the solution was then analyzed by ICP-MS and the results were corrected for inter-element spectral interferences. For the Loss On Ignition (LOI) process, 1.0 g of each sample powder is placed in an oven at 1000 °C for 1 h, then cooled and weighed. The LOI (in percentage) is then calculated from the difference in weight. Different standards were used for data quality assurance.

For mineral chemistry, four samples (02 mafic columnar joints and 02 mafic dykes) were selected. The observed mineral assemblages were imaged, and mineral compositions were measured using a CAMECA SX 100 (15 kV, 10 nA) electron microprobe at the Pierre et Marie Curie University, Paris VI, France. Standards used were diopside for Mg, Ca and Si, Fe2O3 for Fe, MnTiO3 for Ti and Mn, Cr2O3 for Cr, albite for Na, and orthoclase for K and Al. Counting times were 10 s for both peaks and background, with a 5 µm defocused beam.

4 Results

4.1 Field relationship

Columnar jointed basalts are located in the NW of the study area, at about 4 km from the main road joining Foumban to Magba (Fig. 2). They occur at the hilltop within the granite substratum and are oriented E-W to NE-SW. The outcrop appears as a crater (about 800 m long over 500 m wide). The columns are vertical, inclined, or horizontal and display mostly sections showing pentagonal and hexagonal shape with a side measuring about 20 to 50 cm (Fig. 3a, b). Some columns faces display erosion crust with orange to yellowish color due to alteration. The rock is melanocratic and display aphyric to microlitic texture.

Fig. 3
figure 3

Field photographs of the studied mafic rocks. a Inclined columnar jointed basalts at Mfengou, b Horizontal columnar jointed basalts, c Mafic dyke crosscutting ultramylonite host rock at Nchinsap village, d Mafic dyke occurring in the Vuh-Ngouo cliff. Note the thickness of the dyke and its sharp contact with the host mylonite

Full size image

Mafic dykes are exposed only on the shear band belonging to the Foumban-Bankim Shear Zone. The shear band measures about 10 km in width and 18 km long. The dykes are melanocratic displaying fine to medium-grained texture and crosscut the granites, mylonites and gneisses. They are exposed in the Chinsap, Chinpouet, Mayouom, Mayap and Vuh-Ngouo areas. At the Vuh-Ngouo hill and Mayap area, the mafic dykes crosscut the gneisses and mylonite, while at the Chinsap, Mayouom and Chinpouot, they crosscut the mylonite and granite.

All the dykes strike NW–SE to NNW-SSE. The dykes’ outcrops present variable sizes varying from 20 cm to 3 m width and from 10 to 20 m length (Fig. 3c, d). They present mostly a sharp contact with their host rock.

4.2 Petrography

4.2.1 Columnar jointed basalts

Columnar jointed basalts have microlitic textures and consist mainly of plagioclase (20–40 vol.%), pyroxene (20–30 vol.%), chlorite (5–15 vol.%), olivine (5–10 vol.%), opaques (1–15 vol.%), K-feldspar (2–7 vol.%) and calcite (1–3 vol.%) (Fig. 4a). Plagioclases occur as elongated microslats together with K-fedspar in the groundmass surrounding olivine and pyroxene phenocrysts. In some sections prepared from the inclined columnar, the plagioclase crystals are well aligned, indicating the magma flow direction (Fig.  4a). Pyroxenes form isolated globular or elongated phenocrysts and microcrysts in the groundmass. Some crystals are altered into iddingsite or chlorite. In places, numerous pyroxene microcryts form a rim around quartz xenocrysts testifying to the fact that the quartz crystals were taken from the host granite-gneissic substratum by the mafic magma during its ascent to the surface (Fig.  4b). Olivines are less abundant in the rock and present globular shape. They are in place altered into serpentine (Fig.  4a). Calcite occurs as subhedral phenocrysts and displays its characteristic twining although attacked at its edges by some alteration reaction (Fig.  4b).

Fig. 4
figure 4

Photomicrographs of mafic rocks’ thin sections. a Columnar jointed basalts displaying a microlitic texture. Note the orientation of the plagioclase laths indicating the magma flow direction, b Columnar jointed basalts presenting quartz xenocryst surrounded by pyroxene microcrystals forming a reaction rim testifying that the quartz crystal was ripped from the ganito-gneissic substratum by the mafic magma during its ascent to the surface, c Medium-grained texture in mafic dykes, d Plagioclase laths surrounding relics of olivine crystals that underwent the serpentinization process

Full size image

4.2.2 Mafic dykes

Mafic dykes are fine to medium-grained (Fig. 4c, d). They consist of plagioclase (20–40 vol.%), pyroxene (15–35 vol.%), olivine (5–10 vol.%), opaque (5–8 vol.%), K-feldspar (2–8 vol.%) and chlorite (1–5 vol.%) (Fig. 4c, d). K-feldspars are mainly sanidine and are in place crosscutting plagioclases laths. In places, they show their characteristic clear appearance and simple twining. Plagioclases occur as elongated laths radially oriented while pyroxene appears only as microcrystals in the groundmass. Olivines are in the form of subhedral and anhedral crystals locally associated with pyroxenes and plagioclases. Some crystals show serpentinization reaction (Fig. 4d).

4.3 Mineral chemistry

4.3.1 Feldspars

Feldspar analyses are shown in Table 1. In mafic dykes, K-feldspars are orthoclase, sodic sanidine and potassic albite with orthoclase content ranging from Or11 to Or85 (Fig. 5a). Plagioclases display a wide composition range, varying from potassic oligoclase to potassic labradorite (An32-55) (Fig. 5a). The columnar jointed basalts show a wide K-feldspar composition ranging from sanidine to calcio-potassic albite (Or06 to Or64) and the plagioclase ranging from labradorite to oligoclase (An23 to An58). The feldspar present very high FeO content (18.24- 30.07 wt%) and low Al2O3 (0.13–0.58 wt%) and MgO (0.03–0.31 wt%) testifying to the substitution of Al3+ and Mg2+ by the Fe3+ [30].The mafic dyke’s feldspars globally present similar chemical characteristic as the Njimom dykes [19] (within this study area), but the chemistry of the feldspar of the mafic columnar is different from that of the mafic dykes and the Njimom dykes (Fig. 5a).

Table 1 Representative chemical composition (wt%) and structural formulae of feldspar from the Mfengou-Manki mafic rocks
Full size table
Fig. 5
figure 5

a Feldspar composition for the Mfengou-Manki mafic rocks plotted on the An-Ab-Or diagram [88], b Pyroxene composition of the studied rocks on a Wo–En–Fs diagram [31]

Full size image

4.3.2 Pyroxene

Representative microprobe analyses of the pyroxenes and their calculated structural formulae based on six oxygen end-members are presented in Table 2. Pyroxenes are essentially clinopyroxenes and show a restricted compositional range, with Al2O3 (2.33–4.86), FeOt (10.01–17.98), MgO (8.02–13.08), CaO (19.81–21.98) and TiO2 (0.08–2.45) concentrations in mafic dykes and Al2O3 (3.98–6.52), FeOt (10.22–12.18), MgO (11.45–12.97), CaO (19.89–21.86) and TiO2 (0.97–3.65) content in columnar jointed basalts (Table 2). The studied pyroxenes’ mineral composition falls within Wo43-50, En25-39 and Fs15-24 for the mafic dykes and Wo43-49, En36-42 and Fs09-17 for the columnar jointed basalts. On the Wo-En-Fs ternary diagram after Morimoto et al. [31], most of the samples fall within the augite field and few in the diopside field (Fig. 5b). The chemistry of pyroxenes of the mafic dykes and mafic columnar joints are similar to those of the Njimom dykes [19] (Fig. 5b).

Table 2 Representative chemical composition (wt%) and structural formulae of clinopyroxenes from the Mfengou-Manki mafic rocks
Full size table

4.4 Thermobarometry and oxygen fugacity of the magma

Temperature and pressure of the crystallization of numerous minerals are generally used to understand the emplacement conditions of igneous or metamorphic rocks. Several thermo-barometers were used to estimate the pressure and temperature of the crystallization of the mafic dykes and columnar jointed basalts from the Mfengou-Manki area, but three of them have been retained for this work because they don’t compulsorily need the simultaneous presence of two pyroxenes (orthopyroxene and clinopyroxene) and can also be used for both Mg-Ca-Fe and Fe–Mg pyroxenes. The first one is the method used by Soesoo [32] for the formation temperature of pyroxenes. In that method, XPT and YPT indices (Fig. 6a) are used, and are based on the following equations:

  1. a)

    XPT = 0.446 SiO2 + 0.187 TiO2–0.404 Al2O3 + 0.346 FeO–0.052 MnO + 0.309 MgO + 0.446 CaO–0.446 Na2O.

  2. b)

    YPT = − 0.369 SiO2 + 0.535 TiO2 – 0.317 Al2O3 + 0/232 FeO + 0/235 MnO–0.516 MgO–0/167 CaO–0/153 Na2O, where XPT and YPT are respectively the scores of the first and second eigen- vectors [32].

Fig. 6
figure 6

a Crystallization temperature of pyroxenes of the studied mafic rocks [32], (b) Estimated pressure of the crystallization of the pyroxenes using the diagram after Soesoo [32], c Clinopyroxene barometry diagram [34] of the studied mafic rocks, d Temperature of crystallization of the feldspars of the Mfengou-Manki mafic rocks using the triangular thermometry diagram after Kroll et al. [37]

Full size image

This method allows distinction of clinopyroxenes crystallizing from basaltic magmas over a large range of temperature (1100–1300 °C) and pressure (0–20 kbar) [32].

According to that method, the crystallization temperature of clinopyroxenes from the studied rocks ranges from 1100 to 1150 °C for the mafic dykes and from 1150 to 1200 °C for the columnar jointed basalts under the respective pressure of crystallization varying from 2 to 6 Kbar for the mafic dykes and from 6 to 15 kbar for the columnar jointed basalts (Fig. 6a, b).

Using the thermobarometric calculation after Nimis and Taylor [33] on clinopyroxenes at 3 kbar, the temperature of crystallization of clinopyroxenes ranges from 1091 to 1233 °C for the mafic dykes and from 1120 to 1229 °C for the columnar jointed basalts. When plotted on the clinopyroxene barometry diagram after Helz [34], all samples plot within the low to medium-pressure igneous rock fields (Fig. 6c). Putirka [35] established the thermometers and barometers specific for volcanic systems. This thermobarometer is most suitable for this work because the rocks are volcanic (columnar jointed basalts) and hypabyssal (mafic dykes) in origin. The temperature and pressure calculations using those thermometers and barometers give a wide range of temperature of clinopyroxene crystallization varying from 1064 to 1152 °C for the columnar jointed basalts and from 1071 to 1193 °C for the mafic dykes under the respective pressure of 2 to 13 kbar for the columnar jointed basalts and 4 to 12 kbar for the mafic dykes. This wide range of pressure corresponds to a wide range of clinopyroxene crystallization depth as mostly founded in mafic intrusions of the Cameroun Volcanic Line [36].

Moreover, the feldspar triangular thermometry diagram [37] gives the temperature ranging from 700 to 1000 °C for the mafic dykes and from 800 to 1100 °C for the columnar jointed basalts with one sample exceeding 1100 °C (Fig. 6d).

Evaluating the oxygen fugacity of the magma of the studied mafic rocks is important because oxygen fugacity plays an essential role in changing the liquidus temperature, melt and crystals composition, and the magmatic processes in general [38]. Plotted on the AlVI + 2Ti + Cr versus AlVI + Na diagram [39], the studied samples fall in the high oxygen fugacity area (Fig. 7).

Fig. 7
figure 7

Oxygen fugacity of the studied mafic rocks using the AlVI + 2Ti + Cr versus Al.VI + Na diagram after Schweitzer et al. [39]

Full size image

4.5 Geochemistry

4.5.1 Columnar jointed basalts

The columnar jointed basalts are basic to intermediate and present low to moderate content of SiO2 (46.89–52.71 wt.%) (Table 3). They plot in the basalt, trachy-basalt and basaltic trachy-andesite field of the TAS (total alkali-silica) classification diagram after Lebas et al. [40] (Fig. 8a). They display moderate to high Al2O3 (12.25–15.85wt.%), TiO2 (1.69–2.29 wt.%) and low MgO (4.15–7.9 wt%) contents. The columnar jointed basalts present low to moderate Mg# value (40.16–53.05) and the A/CNK molar ratio ranging from 0.75 to 1.41. The LOI (1–6 wt.%) of the studied columnar jointed basalts are low to high, testifying to the variable degree of weathering of the rock. All the columnar jointed basalts samples plot within the Calc-alkaline domain on the AFM diagram of Irvine and Baragar [41] (Fig. 8b). On the plot of SiO2 (wt%) versus other major and trace elements, Al2O3 and K2O show a positive correlation with SiO2 while MgO, CaO and TiO2 display a negative correlation with SiO2 (Fig. 9). The trace elements such as Ni and Cr present negative correlation with SiO2 while Y and Ba present positive correlation with SiO2 (Fig. 9).

Table 3 Major (wt%) and trace (ppm) element contents in mafic columnar joints and mafic dykes from the Mfengou-Manki area
Full size table
Fig. 8
figure 8

a Classification diagram of the Mfengou-Manki mafic rocks after Le Bas et al. [40], b AFM diagram of the studied mafic rocks after Irvigne and Baragar [41]

Full size image
Fig. 9
figure 9

Harker diagram of some major and trace elements of the studied mafic rocks against SiO2

Full size image

The studied columnar jointed basalts show low concentrations of Ni (35- 173 ppm) and low to moderate concentrations of Cr (87–268 ppm) and Co (29–53 ppm). On the multi-element diagram normalized to primitive mantle abundances given by McDonough and Sun [42], the columnar jointed basalts display slight depletion in Nb and Ta, pronounced depletion in Th, U, Pb and Zr, and enrichment in Cs, Ba and Rb (Fig. 10a). The REE diagram normalized to the primitive mantle [42] shows enrichment in LREE (LaN/SmN = 2.39–3.65) compared to HREE (GdN/YbN = 1.80–2.94) and negative to slight positive europium anomalies ((Eu/Eu*)N = 0.6 to 1.03) (Fig. 10b).

Fig. 10
figure 10

Primitive mantle-normalized patterns for the studied mafic rocks [42]. a Trace element patterns of the columnar jointed basalts, b REE patterns of the columnar jointed basalts, c Trace element patterns of the mafic dykes, d REE patterns of the mafic dykes

Full size image

4.5.2 Mafic dykes

The mafic dykes are essentially basic, with a narrow SiO2 content ranging from 46.89 wt% to 50.44 wt%. Plotted on the TAS (total alkali silica) classification diagram after Lebas et al. [40], the mafic dyke samples fall in the field of basalt, trachy-basalt and basaltic trachy-andesite (Fig. 8a). They present high MgO (8–12.85 wt.%) and TiO2 (1.87–2.64 wt.%) with moderate to high Mg# (50–65) and the A/CNK molar ratio varying from 0.77–1.18. The mafic dykes display a wide range of LOI values (2.89–7.30 wt.%) (Table 3). On the AFM diagram after Irvine and Baragar [41], they plot on the calc-alkaline fields (Fig. 8b). Plotted on the SiO2 versus other major and trace elements variation diagrams, Al2O3, and K2O present a positive correlation with SiO2 while TiO2 and CaO present negative correlation with SiO2 (Fig. 9). Moreover, Cr and Ni show negative correlation with SiO2 while Rb, Ba, and Y are showing positive correlation with SiO2 (Fig. 9).

The mafic dykes show high Ni (86–240 ppm) and Cr (133 to 295 ppm) concentrations and low to moderate Co (29–52 ppm) contents. On the multi-element diagram normalized to primitive mantle abundances given by McDonough and Sun [42], the mafic dyke samples show positive anomalies in Rb, Ba, Nb, Ta, Ti and Pb, and negative anomalies in Th, U, Ce, and Zr (Fig. 10c). The REE diagram normalized to the primitive mantle [42] shows a slight fractionation, and is enriched in LREE (LaN/SmN = 2.07–3.38) compared to HREE (GdN/YbN = 1.87–2.19) (Fig. 10d). They also show negative to slight positive europium anomalies ((Eu/Eu*)N = 0.76 to 1.01).

5 Discussion

5.1 Alteration

After magma emplacement and crystallization, some alteration processes can take place in the rock. The effect of that alteration can be seen either in thin sections or in LOI values or trace elements ratios. Some indications of alteration are seen in few thin sections such as chloritisation of pyroxene and serpentinization of olivine. However, most samples remain largely unaltered. The LOI (1–7 wt%) of all the studied rocks indicates that the samples were subjected to variable degrees of alteration (Table 3). Moreover, the Y/Ho ratios can also be used for evaluating the role of alteration. The Y/Ho ratio is varying from 22 to 36 in the columnar jointed basalts (with three samples presenting the value > 44) and from 24 to 31 in mafic dykes. The mean value (Y/Ho = 31) of all the columnar jointed basalts and mafic dykes are slightly over the value of 28 that is observed in fresh basaltic rocks [43] and below the values observed in altered rocks (between 44 and 74) [44, 45] indicating globally that secondary processes were not significant on the studied rocks.

5.1.1 Crustal contamination

Understanding the nature of the mantle source and the petrogenetic processes requires a good understanding of the role of crustal contamination. In the wide range, the major oxides and LILE compositions generally change during the post-magmatic processes. The overall variations in concentrations of major oxides, LILE, HFSE and LREEs can be caused by crustal contamination [46].

Trace elements chemistry is a good proxy to assess crustal contamination. Mafic magma affected by crustal contamination generally presents Ti and P depletion and enrichment of Th and LREEs (e.g., [47]). The studied mafic dykes and columnar do not have large P and Ti troughs on primitive mantle-normalized multi-element diagrams (Fig. 10a, c), testifying to insignificant or minor crustal input in the magma. The low concentrations of U and Th in the studied dykes may also point to a lack of significant crustal contamination [48]. The mafic dyke samples show low Ba contents (mean = 483 ppm) and low Lu/Yb (0.13 -0.14) ratio while the columnar jointed basalts display slightly high Ba (mean = 647) contents and Lu/Yb (0.13–0.16) ratio compared to the mid-continental crust [Ba = 259–628 ppm; and Lu/Yb (0.16–0.18)] [49]. These values may suggest possible small degree of crustal contamination on the columnar jointed basalts or the enrichment of those elements in the source. This low degree of contamination is also supported by the rapid crystallization of the columnar jointed basalts magma during quenching (highlighted by the high FeO content (18.24- 30.07 wt%) and low Al2O3 ( 0.13–0.58 wt%) and MgO (0.03–0.31 wt%) contents testifying to the substitution of Al3+ and Mg2+ by the Fe3+) [30]. On the thin section, the low degree of contamination is evidenced by the presence of only a few occurrences of quartz xenoliths (probably taken from the crustal substratum) reacting with microcrysts of pyroxenes (Fig. 4b).

5.2 Magma processes

Some authors [50, 51] have demonstrated that high field strength elements (HFSE) such as Zr, Hf, Nb, Ta, Th, Y, P and Ti are stable even when they have undergone high-grade metamorphism. Large ion lithophile elements (LILE) such as Ba, Cs and Rb, on the other hand, have a strong crustal-mantle fractionation and may be unstable during metamorphism or crustal contamination during magma ascension, thus altering the physicochemical composition of the magma [52, 53]. Due to their relative stability during crustal contamination or metamorphism, transition metals and HFSEs are mainly used to study the tectonic settings and magma sources of rocks [54, 55]. The major element geochemistry of the studied mafic rocks demonstrates the calc-alkaline nature of the columnar jointed basalts and the mafic dykes (Fig. 8b). Plotted on the Ti versus (Ca + Na) diagram of pyroxenes after Leterrier et al. [56], the studied pyroxenes plot on the boundary between tholeiitic and alkali basalt fields, but mostly in the alkali basalt field (Fig. 11a). The columnar jointed basalts are basic to intermediate (SiO2 = 46–52 wt%), and display low to moderate Mg# (40–53), Ni(35–173 ppm), Co (29–53 ppm) and Cr (87–268 ppm) while the mafic dykes are essentially basic (SiO2 = 46–50 wt%), and display moderate to high Mg# (50–65), Ni (86–240 ppm), Co (29–52 ppm and Cr (133–295 ppm) (Table 3). These chemical compositions are lower than those of the average composition of primary mantle (Ni = 300–400 ppm, Co = 50–70 ppm, Cr = 300–500 ppm) [57]. These major elements and trace element characteristics indicate that the studied rocks had undergone fractionation, with the mafic dykes being the least fractionated.

Fig. 11
figure 11

a Ti versus (Ca + Na) diagram of pyroxenes after Leterrier et al. [56], b La/Sm versus Sm/Yb diagram for the studied dykes. Melting curves for garnet and spinel lherzolite are after Aldanmaz et al. [61]. Numbers along the curves represent the percentage of partial melting for different proportions of garnet (Gt) and spinel (Sp), c Th/Yb versus Nb/Yb diagram [62], d Nb/La versus La/Yb variation diagram for the studied mafic rocks

Full size image

Moreover, on the plot of SiO2 (wt%) versus other major and trace elements, Al2O3, K2O, Y and Ba show a positive correlation with SiO2, while MgO, CaO, TiO2, Ni and Cr display a negative correlation with SiO2, demonstrating the crystallization process especially the fractionation of olivine and clinopyroxene [57] (Fig. 9). The studied mafic rocks present negative to slightly positive Eu anomalies (Eu/Eu* = 06–1.03) indicating plagioclase fractionation.

5.3 Partial melting and magma source

Culler [58] has shown that the profiles of the REE patterns are important to better understand the melting conditions of the rock. Thus, a low melting percentage of the magma source can produce slightly tilted REE patterns, while a higher degree of mantle melting produces flat REE patterns [59]. Furthermore, the inclined profile of REE can also indicate the origin of an enriched mantle source, whereas the flat profiles may demonstrate a depleted mantle source. The REE profiles of the columnar jointed basalts and mafic dykes exhibit a relatively tilted pattern, suggesting a low degree of partial melting of an enriched mantle source. The REE pattern of the mafic dykes and columnar jointed basalts are quite similar and may indicate the same source for their magmas. This assumption is also strongly supported by the relatively high Dy/Yb ratios (~ 2, mean = 2.14 for the mafic dykes and mean = 2.09 for the columnar jointed basalts) of the studied rocks, implying that partial melting could have taken place in the spinel-garnet transition zone [60]. This hypothesis is also supported by the (La/Sm) versus (Sm/Yb) diagram after Aldanmaz et al. [61], which shows the columnar jointed basalts and the mafic dyke samples falling near the E-MORB domain following the spinel + garnet line (Fig. 11b). The studied samples also plot mostly along the mantle array of the Nb/Yb versus Th/Yb diagram [62], with few columnar jointed basalts samples extending on the arc related magma area, consolidating the mantle source for the studied mafic rocks (Fig. 11c). Moreover, the Ce/Pb ratios of the studied samples vary from 7 to 23 for the columnar jointed basalts and from 5 to 10 for the mafic dykes (except one sample with Ce/Pb value of 2.78). These values are lower than the average values obtained from the basaltic lavas from the Cameroon Volcanic Line (25–32) and also lower than MORB and OIB values (25 ± 5, [63], but they are higher than the average value found in the upper-crust (3.2, [64]). The trace element compositions of the columnar jointed basalts and mafic dykes show some visible differences. The columnar jointed basalts show a relative depletion in Nb, Ta and Pb on the primitive mantle normalized multi-element plot (Fig. 10a) while the mafic dykes display pronounced enrichment in Nb, Ta and Pb (Fig. 10c). These features suggest a certain degree of evolution or contamination of the columnar jointed basalts compared to the mafic dyke samples.

Considering the characteristics of the trace elements composition of the columnar jointed basalts, the following deduction could be made, firstly of (1) a mixture of a plume-like melt with the continental crust or; (2) plume-like melt with the sub-continental lithospheric mantle to produce the columnar jointed basalts. However, when the mixing involves the continental crust, very significant Nb–Ta negative anomalies appear on the PM-normalized multi-element patterns for the studied rocks, which is not the case here.

Moreover, the plume-derived melts generally have a low La/Ta ratio (8–15). When this value is much higher (> 25 in general), it indicates the involvement of the Sub-continental lithospheric mantle (SCLM). Also, the melts contaminated by crustal materials generally exhibit high La/Sm ratio (> 5) [65]. The studied mafic dykes display low La/Ta ratios (mostly < 12) and low La/Sm ratios (< 5) testifying to its plume source while the columnar jointed basalts present varying La/Ta ratios (10 to 20 with one sample with the value of 37) and low La/Sm ratio (3 to 5) suggesting the participation of the SCLM to their formation.

Furthermore, the studied samples from the Mfengou-Manki area are depleted in incompatible element and have the La/Nb ratios (0.6–1.7), Th/Nb (0.04–0.25), Zr/Nb (2.56–9.27) and Ba/Nb (5.31–25.86). These ratios are lower than those of the continental crust and very close to those of the OIB, indicating the OIB-type geochemical signatures for the studied mafic rocks [66].

The columnar jointed basalts display low Nb/La (< 1) ratio while the mafic dykes display high Nb/La (> 1) ratio suggesting the participation of the lithospheric mantle source to the genesis of the columnar jointed basalts and exclusively the OIB-like asthenospheric mantle for the mafic dykes [67, 68]. This result is also supported by the Nb/La versus La/Yb diagram (Fig. 11d), in which the mafic dykes plot mainly on the HIMU-OIB asthenospheric mantle and the columnar jointed basalts samples plotting mainly on the mixed lithospheric-asthenospheric mantle area.

Also, studied mafic rocks display moderate TiO2 contents (between 1.69 and 2.29 wt.% for the columnar jointed basalts and 1.87–2.64 wt.% for the mafic dykes) and characterize magmas generated from a slightly shallower mantle source compared to high-Ti magmas. It has been demonstrated that high-Ti content in basalts are evidence for the involvement of asthenospheric mantle source components in the genesis of the basaltic rocks [69, 70]. It is therefore very likely that the studied columnar jointed basalts were generated by shallow melting of asthenospheric mantle source mixed with different amounts of melts from SCLM.

The geochemical composition of the plume-derived melts may have been imparted into SCLM through the plume- lithosphere interaction especially during the delamination process [70, 71]. So, the contamination of asthenospheric magma by lithosphere-derived melt probably played an important role in the genesis of the studied columnar jointed basalts as demonstrated on the Soltan Maidan basalts in Iran [72].

This geological phenomenon is not new in Cameroon. The mixing of source material from the asthenospheric and the lithospheric mantle has also been previously mentioned in some areas of the Cameroon Volcanic Line such as the Bioko island [73], the Mts Bambouto and Oku, the Ngaoundere Plateau [74], and the Bamenda Mountains [75].

5.4 Comparison with the other mafic dykes in the Cameroon Central Shear Zone

The geochemical and mineralogical characteristics of the Mfengou-Manki columnar jointed basalts and mafic dykes compared to those of other dykes in the Central African Fold Belt in Cameroon reveal some differences and similarities. The studied mafic rocks present silica content (46% to 52%) similar to those from the Njimom area (46–51%), but high Ni (86–240 ppm) and Cr (133–295 ppm) content compared to those of Njimom dykes (Ni = 30–110 ppm; Cr = 70–180 ppm) [19].

The mafic dykes from the Tikar Plain and Kekem areas derived from the sub-continental lithospheric mantle source [26, 76] and are indeed different from the Mfengou-Manki mafic dykes and columnar jointed basalts that originated from asthenospheric and mixed asthenospheric-lithospheric mantle respectively. The studied mafic dykes, compared to the columnar jointed basalts, are more enriched in Nb, Ta Ti and Sr, but depleted in Zr compared to the Tikar plain and Kekem dykes, (Fig. 12a, b).

Fig. 12
figure 12

a Primitive mantle-normalized trace elements for the Mfengou-Manki columnar jointed basalts compared to the Tikar plain, Kekem and Njimom dykes, b Primitive mantle-normalized trace elements for the Mfengou-Manki mafic dykes compared to the Tikar plain, Kekem and Njimom dykes

Full size image

From the field relationship, the mafic dykes are oriented parallel to the Tikar plain dykes and Kekem dykes (NW–SE direction) and are perpendicular to the direction of the Cameroon Volcanic Line (NE-SW). That direction is similar to that of the second deformation phase (D2) in the CCSZ dominated by the shearing movement [26]. The C2 shear axes (NW–SE) of the second deformation phase in the Central Cameroon Shear Zone may have favored the emplacement of the studied dykes as it is the case for the Tikar plain mafic dykes.

5.5 Tectonic setting and geodynamic implication in the Cameroon Volcanic Line and Central Cameroon Shear Zone

The geochemistry of a specific type of magmatic rock may reflect or indicate its tectonic setting [77]. Trace elements such as Nb, Y, Ta, Th, Sm, Zr, Hf, and Yb are generally used for the discrimination of the tectonic environment in which the rock has been formed. This is because of their relative stability and low ion energy potential during metamorphism and weathering processes. Using some of those elements (Nb, Zr, Y, Th, Hf), for the tectonic discrimination triangular diagrams (Hf/3-Th-Nb/16) after Wood [78] (Fig. 13), the studied samples fall within the E-MORB and within-plate tectonic domains. According to Rudnick and Gao [79], negative Ti, Ta and Nb anomalies are typical of continental crust and rocks whose chemical composition has been modified by subduction [80]. The enrichment in Nb, Ti and Ta of the mafic dykes exclude a contribution of the subducted slab to their formation [81, 82]. In contrast, the columnar jointed basalts show relatively negative anomalies in Nb and Ta, testifying to the little participation of the sub-continental lithospheric mantle in their formation.

Fig. 13
figure 13

Hf/3-Th-Nb/16 tectonic classification diagram of the studied mafic rocks [78]

Full size image

The dykes may have intruded along lower to mid-crustal sub-vertical feeder channels and were emplaced along the shear axes during the C2 shearing stage in the Adamawa Yade domain of the Central African Fold Belt in Cameroon [26]. This hypothesis is also supported by several authors indicating that the Pan African orogeny that affected the Neoproterozoic granitic basement of the Cameroon Line favored the emplacement of fractures and zones of weaknesses [24, 74, 83, 84] that have facilitated the emplacement of extrusive (lava flow and columnar) and intrusive magmatic (mafic dykes and anorogenic complexes) rocks [85]. Moreover, recent geophysical studies on the Central African Fold Belt in Cameroon demonstrated the existence of both the subduction, slab detachment and the delamination of the sub-continental lithospheric mantle under the Adamawa Yade domain and the West Cameroon domain ([86], Fig. 14).

Fig. 14
figure 14

Schematic model modified from Goussi Galamo et al. [87] showing the subduction involving the Congo Craton and the Adamawa-Yade domain and the delamination of the SCLM under the Central Cameroon Shear Zone. The SCLM delamination favoured the mixing between the plume and the SCLM magmas leading to the formation of columnar jointed basalts. The mafic dykes followed the fracture of the shear zone to be emplaced at the surface and sub-surface without significant contamination

Full size image

6 Conclusion

The aim of this work was to study the petrography, geochemistry and mineral chemistry of the mafic dykes and the columnar jointed basalts, in order to constrain their petrogenesis and tectonic settings and the role of lithospheric or asthenospheric source in their genesis. Two hypotheses have been proposed for the origin of the columnar jointed basalts: (1) a mixture of a plume-like melt with the continental crust or (2) plume-like melt with the sub-continental lithospheric mantle.

The petrography and geochemistry of the Mfengou-Manki mafic rocks permitted to highlight two main mafic rock types, namely the mafic dykes and the columnar jointed basalts. The mafic dykes present fine to medium-grained texture while the columnar jointed basalts display the microlitic texture. The mafic dykes are calc-alkaline while the columnar jointed basalts are calc-alkaline to tholeiitic in nature. The geochemistry of the analyzed samples indicates that the mafic dykes are from the asthenospheric mantle source while the columnar jointed basalts are from the mixed lithospheric-asthenospheric mantle.

The mafic dykes emplaced at a temperature of 1071 to 1193 °C and a pressure of 4 to 12 kbar while the columnar jointed basalts emplaced at a temperature of 1064 to 1152 °C and 2 to 13 kbar pressure.

The studied rocks emplaced in the within-plate domain following the C2 shear axes in the Cameroon Central Shear Zone. The delamination of the sub-continental lithospheric mantle under the Central Cameroon Shear Zone in the Adamawa Yade domain and the West Cameroon domain may have favored the ascend of the plume-derived melts which imparted into SCLM through the plume-lithosphere interaction. However, the petrography, whole-rock geochemistry and mineral chemistry alone are not sufficient to better explain the magma sources and processes leading to the formation of such complex geochemical features. Isotopic and geochronological studies should be done respectively in the future work to better constrain the magma sources, the mixing processes and the emplacement ages of the studied mafic rocks.