Petrology of Mafic Dykes from the Njimom Area (West-Cameroon): A Contribution to the Characterization of Late Paleozoic and Mesozoic Magmatism in the Southern Continental Part of the Cameroon Volcanic Line

In the western Cameroon, crop out several dyke swarms of Paleozoic–Mesozoic age. These dykes intrude the Precambrian basement in the southern continental part of the Cretaceous Cameroon Volcanic Line. In the Njimom area, two groups of mafic dykes that crosscut the Neoproterozoic basement rocks have been observed. A first group intrudes the mylonites whereas the second group intrudes the granites. The dykes are alkaline basalts and hawaiites. The mineralogical assemblage of both groups of dykes consists of plagioclase, clinopyroxene, altered olivine, and opaque oxides. The dykes that cross-cut the Precambrian mylonitic gneisses show moderate TiO2 (1.7–2.0 wt.%), low MgO (4.4–7.1 wt.%), and compatible trace element concentrations (e.g., Cr = 70–180 ppm; Ni = 30–110 ppm). The dykes that intrude the granites have TiO2 contents between 2.3 and 2.5 wt.% and moderate compatible trace element concentrations (e.g., Cr = 260–280 ppm; Ni = 170–230 ppm). MgO varies from 5.9 to 9.2 wt.%. All mafic dykes are enriched in light lanthanide element and show moderate Zr/Nb and high Zr/Y, Nb/Yb, and Ti/V ratios similar to those of average ocean island basalt (OIB)-type magmas. Some dykes that intrude the mylonites show evidence of contamination by continental crust. The composition of the clinopyroxenes of the dykes that intrude the mylonites clearly indicate different and unrelated parental magmas from dykes that intrude the granites. Contents and fractionation of the least and the most incompatible elements suggest low degrees of partial melting (3–5%) of heterogeneous source slightly enriched in incompatible elements in the spinel stability field. The geochemical features of Njimom dykes (in particular the dykes that intrude the granites) are similar to those of Paleozoic and Mesozoic dykes recorded in the southern continental part of the Cameroon Volcanic Line, suggesting multiple reactivations of pre-existing fractures that resulted in the fragmentation of western Gondwana and the opening of the South Atlantic Ocean.


Introduction
Volcanic activity in Cameroon is concentrated along the well-known Cameroon Volcanic Line (CVL), an alignment of oceanic and continental magmatic centers that straddle the boundary continental/oceanic crust in Central Africa (e.g., [1]). The CVL extends (with a N30 • E direction) over more than 1000 km in Cameroon, Equatorial Guinea, Sao Tome, and Principe. According to some authors, the CVL extends also in Nigeria in the Biu Plateau and in Chad (e.g., [2,3]). Volcanic rocks of the Cameroon Line cover the Precambrian basement of Neoproterozoic age (e.g., [4]). In the southern continental part of the CVL, the Paleozoic-Mesozoic dyke swarms show mineralogical and geochemical characteristics different from the common CVL rocks (e.g., [5,6]). While in CVL the rocks are mainly alkaline, the dykes are essentially transitional and/or tholeiitic. The oldest basalts of the CVL are of Tertiary age while the dykes span over Paleozoic and Mesozoic eras with youngest know age (192.10 ± 7.45 Ma) being found for the Kendem dyke which is the only NW-SE dyke dated [7]. On a structural basis, orientations of basaltic dykes are aggregate to the range NNE-ENE with a predominance of the ENE orientation which is locally known as the Adamawa Shear Zone (ASZ) which runs from the Gulf of Guinea to Sudan and in a pre-drift reconstitution, is thought to prolong to Brazil as the Pernambuco Shear Zone. Tectonic studies evidence a Riedel fracture model with a dextral shearing along the ASZ (e.g., [8,9]) with a NW-SE shortening (σ1) associated to the formation of the Benue aulacogen, a NE-SW stretching (σ3) corresponding to the CVL. Our field works along the ASZ in the Njimom area indicate the existence of NW-SE oriented basaltic dyke swarms.
Here we present a study of the Njimom dykes. With field, petrographic and whole-rock geochemical (major and trace element) data, we decipher the petrogenesis of this dyke suite and the signification of the NW-SE dykes group in the framework of the opening of the Southern Atlantic Ocean.

Geological Setting
The studied dykes are located to the West of the city of Njimom, in the southern continental part of the CVL (Figures 1 and 2). The area lies on a~0.6 Ga granitic-gneissic basement (e.g., [10,11]) strongly mylonitized along the Central Cameroonian Shear Zone, a segment of the regional Adamawa Shear Zone. The rocks that compose the Pan-African basement in the area are coarse grained granitoids, granoblastic orthogneisses, migmatites, metagabbros, and mylonites [12]. Structural features and kinematic indicators point to a syntectonic emplacement of the basement rocks and provide detailed information on the relative timing of deformation [13,14]. Younger mafic dykes and quartz-feldspathic veins crosscut mylonites and granites while amphibolitic enclaves are widespread in gneisses. Migmatites are associated with orthogneisses and display high K to shoshonitic characters [12]. The spatial distinction in pre-to syn-orogenic magmatism allows recognition of a north to south potassium increasing trend, compatible with the existence of a northwestern trending Pan-African subduction (e.g., [15]). This sheared margin is marked by the superposition of two mylonitic foliations operating in opposing sense (at constant direction), under high and low metamorphic conditions (e.g., [13]). Volcanic activity in the studied area consists of relics of basaltic plateau (ca. 51 Ma; [16]) of the CVL in the Foumban area (60 km to the SW of the dykes studied in this work).  [17]. CCSZ = Central Cameroon shear zone, SF = Sanaga Fault, TBF = Tcholliré-Banyo fault, RLSZ = Rocher du Loup shear zone, ASZ = Adamawa shear zone, GGSZ = Godé-Gormaya shear zone, MNZ = Mayo Nolti shear zone. The insert is the position of the studied area in pre-drift reconstruction [11]. PA = Patos shear zone, KF = Kandi Fault, PSZ = Pernambuco shear zone. The locations of the mafic dyke swarms (Njimom, Kendem, Dschang, Bangangte, Manjo, Nyos) are also shown.
The Njimom dykes that intrude the mylonitic basement show a weakly porphyritic to aphyric texture (Figure 4a-c). The groundmass of the dykes NK0 and NK2A is strongly altered (Figure 4a,b). Alteration effects are seen mostly in the presence of clay minerals, chlorite, quartz and calcite in the interstices, and iddingsitization of olivine phenocrysts in the sample NK2A (Figure 4b). The dyke NK2C shows an intergranular texture and contain altered olivine, clinopyroxene, plagioclase, and Fe-Ti oxides (Figure 4c). The dyke NK6 that intrudes the granites has phenocrysts and microlites of corroded olivine, with plagioclase, clinopyroxene, and Fe-Ti oxides in the groundmass. The mesostasis is altered and also has scarce alkali feldspar. Calcite-filled amygdales are common (Figure 4d).

Field Occurrence and Petrography
The Njimom dykes strike mainly 135 • N and 155 • N and cross-cut respectively mylonitic gneisses and granites of the Precambrian basement of the area. The dykes are well exposed along the Foumban-Magba newly tarred road. The dykes in mylonites (5 •     The Njimom dykes that intrude the mylonitic basement show a weakly porphyritic to aphyric texture (Figure 4a-c). The groundmass of the dykes NK0 and NK2A is strongly altered (Figure 4a,b). Alteration effects are seen mostly in the presence of clay minerals, chlorite, quartz and calcite in the interstices, and iddingsitization of olivine phenocrysts in the sample NK2A (Figure 4b). The dyke NK2C shows an intergranular texture and contain altered olivine, clinopyroxene, plagioclase, and Fe-Ti oxides (Figure 4c). The dyke NK6 that intrudes the granites has phenocrysts and microlites of corroded olivine, with plagioclase, clinopyroxene, and Fe-Ti oxides in the groundmass. The mesostasis is altered and also has scarce alkali feldspar. Calcite-filled amygdales are common (Figure 4d).  (d) porphyric texture in basalt NK6. The olivine is completely iddingsitized. The dykes in (a-c) cross-cut the mylonitic gneiss. The dyke NK6 intrudes (d) the granitic basement. Mineral names are abbreviated as cpx clinopyroxene; pl, plagioclase; ol, olivine; cc, calcite.

Analytical Methods
Four dykes of the Njimom area, were chosen for mineral chemical analysis. The mineral compositions (Tables 1-3) were obtained at the University of Naples, using an Oxford Instruments Microanalysis Unit equipped with an INCA X-act detector and a JEOL JSM- sub-ophitic texture in basalt NK2A. The olivine phenocrysts are completely altered; (c) intergranular texture in aphyric basalt NK2C with plagioclase, altered olivine, clinopyroxene and opaque oxides; (d) porphyric texture in basalt NK6. The olivine is completely iddingsitized. The dykes in (a-c) cross-cut the mylonitic gneiss. The dyke NK6 intrudes (d) the granitic basement. Mineral names are abbreviated as cpx clinopyroxene; pl, plagioclase; ol, olivine; cc, calcite.

Feldspars
Plagioclase in the dyke NK6 is labradorite (An53-62) and is accompanied by groundmass anorthoclase and minor sodic sanidine (Table 2; Figure 6). In the dykes NK2A and NK2C, plagioclase ranges from An62 to An33 (Figure 6). Anorthoclase and sodic sanidine have also been found in the groundmass.

Feldspars
Plagioclase in the dyke NK6 is labradorite (An 53-62 ) and is accompanied by groundmass anorthoclase and minor sodic sanidine (Table 2; Figure 6). In the dykes NK2A and NK2C, plagioclase ranges from An 62 to An 33 ( Figure 6). Anorthoclase and sodic sanidine have also been found in the groundmass.

Whole-Rock Geochemistry
The Njimom dykes described above range from alkali basalt through hawaiite to mugearite on the total alkali vs silica diagram from [18] (Figure 8). The variability in major element chemistry of the Njimom dykes could be the result of secondary or deuteric alteration of the rocks. Some care must be taken in interpreting the chemical data, for alteration has variably but visibly affected all the samples. Like hand specimen and petrographic observations indicate that some Njimom samples are altered indicating that their compositions may have been modified by late-or post-crystallization fluid-rock interactions. Alteration is reflected in a very general way in LOI (loss on ignition) values, which range widely from 3.34 to 9.21 wt.% in the mafic dykes. Correlations between Ca and more

Whole-Rock Geochemistry
The Njimom dykes described above range from alkali basalt through hawaiite to mugearite on the total alkali vs silica diagram from [18] (Figure 8). The variability in major element chemistry of the Njimom dykes could be the result of secondary or deuteric alteration of the rocks. Some care must be taken in interpreting the chemical data, for alteration has variably but visibly affected all the samples. Like hand specimen and petrographic observations indicate that some Njimom samples are altered indicating that their compositions may have been modified by late-or post-crystallization fluid-rock interactions. Alteration is reflected in a very general way in LOI (loss on ignition) values, which range widely from 3.34 to 9.21 wt.% in the mafic dykes. Correlations between Ca and more

Whole-Rock Geochemistry
The Njimom dykes described above range from alkali basalt through hawaiite to mugearite on the total alkali vs silica diagram from [18] (Figure 8). The variability in major element chemistry of the Njimom dykes could be the result of secondary or deuteric alteration of the rocks. Some care must be taken in interpreting the chemical data, for alteration has variably but visibly affected all the samples. Like hand specimen and petrographic observations indicate that some Njimom samples are altered indicating that their compositions may have been modified by late-or post-crystallization fluid-rock interactions. Alteration is reflected in a very general way in LOI (loss on ignition) values, which range widely from 3.34 to 9.21 wt.% in the mafic dykes. Correlations between Ca and more alteration-resistant elements such as Ti or Zr are poor, suggesting that Ca contents have been affected significantly in some dykes. However, the alteration typically has little effect on the immobile trace elements (e.g., [19,20]). For these reasons, in this study we use only the alteration-resistant elements (Ti, Zr, Nb, Y, Th, and REE) and their ratios for geochemical interpretations. To classify the mafic volcanic rocks, we use the incompatible trace-element classification scheme for altered volcanic rocks (Figure 9: [21]), which indicates that Njimom samples are alkaline basalts and hawaiites. Major and trace element contents of dykes from the Njimom area are listed in Table 4. The major element analyses are recalculated to 100 wt.% LOI-free. The dykes that cross-cut the Precambrian granites have SiO 2 contents between 46.0 and 47.5 wt.%, TiO 2 between 2.3 and 2.5 wt.% and moderate compatible trace element concentrations (e.g., Cr = 260-280 ppm; Ni = 170-230 ppm). MgO varies from 5.9 to 9.2 wt.% and Mg# (Mg# = Mg × 100/(Mg + Fe 2+ )) from 53 to 60 indicating that the samples moderately evolved. In contrast, the dykes that cross-cut the Precambrian mylonitic gneisses are more evolved. They have higher SiO 2 contents (49.1-54.8 wt.%) and lower TiO 2 (1.7-2.0 wt.%), MgO (4.4-7.1 wt.%; Mg# = 45-58), and compatible trace element concentrations (e.g., Cr = 70-180 ppm; Ni = 30-110 ppm). Major and trace element variations, using Zr as a differentiation index, are shown in Figure 10. The data for the two groups of dykes identified in the Njimom area plot in distinct fields in most major and trace element variation diagrams, particularly in the SiO 2 , TiO 2 , Nb, Y, and REE ( Figure 10). The dykes that cross-cut the Precambrian granites (Figure 11a) are moderately light lanthanide element (LREE) enriched (La n = 58 to 70 and La n /Yb n from 7.6 to 8.7, the subscript "n" means chondrite normalized). The dykes that intrude the Precambrian mylonitic gneisses are variably LREE enriched (La n /Yb n = 7.2-10.6) and show small negative anomalies in Eu (Eu n /Eu* = 0.67-0.92, where Eu n is chondrite normalized europium and Eu* = (Sm n × Gd n ) 0.5 ; Figure 11b). Primitive mantle-normalized multielement patterns for dykes that intrude the granites have marked peaks in Sr are and smoothly decreasing normalized abundances from Nb to Lu (Figure 11c). Primitive mantle-normalized incompatible element patterns for dykes that intrude the mylonites show marked troughs at Th and Ti and weak troughs at Nb and Sr (Figure 11d). alteration-resistant elements such as Ti or Zr are poor, suggesting that Ca contents have been affected significantly in some dykes. However, the alteration typically has little effect on the immobile trace elements (e.g., [19,20]). For these reasons, in this study we use only the alteration-resistant elements (Ti, Zr, Nb, Y, Th, and REE) and their ratios for geochemical interpretations. To classify the mafic volcanic rocks, we use the incompatible traceelement classification scheme for altered volcanic rocks (Figure 9: [21]), which indicates that Njimom samples are alkaline basalts and hawaiites. Major and trace element contents of dykes from the Njimom area are listed in Table 4. The major element analyses are recalculated to 100 wt.% LOI-free. The dykes that cross-cut the Precambrian granites have  Figure 11b). Primitive mantle-normalized multielement patterns for dykes that intrude the granites have marked peaks in Sr are and smoothly decreasing normalized abundances from Nb to Lu (Figure 11c). Primitive mantle-normalized incompatible element patterns for dykes that intrude the mylonites show marked troughs at Th and Ti and weak troughs at Nb and Sr (Figure 11d). Figure 8. TAS classification diagram [18] for the Njimom dykes. The subalkaline-alkaline line is from [22]. Data of mafic dykes (Bangoua, Dschang, Maham, and Kendem) are from [7], Kekem dykes from [5], Manjo dykes from [6], Nyos dykes from [23]. Figure 8. TAS classification diagram [18] for the Njimom dykes. The subalkaline-alkaline line is from [22]. Data of mafic dykes (Bangoua, Dschang, Maham, and Kendem) are from [7], Kekem dykes from [5], Manjo dykes from [6], Nyos dykes from [23].

Crustal Contamination
The parental melts of the Njimom dykes must have experienced some differentiation before intruding the upper crust, as reflected in the evolved compositions of all the samples of this study. The Njimom dykes were derived by fractional crystallization of olivine, clinopyroxene, and plagioclase from Mg-rich basaltic melts, consistent with the petrographic assemblages observed in these dykes and their low magnesium numbers (45-60). Figure 11. (a,b) Chondrite-normalized REE diagrams for Njimom dykes. The chondrite values used for normalization are those of [24]. Promethium is interpolated. N-MORB and E-MORB and OIB patterns are plotted with the values of [25]. (c,d) Primitive mantle-normalized incompatible element diagrams for Njimom dykes. Primitive mantle values are from [26]. Data from Manjo [6]; Kekem [5]; Bangoua, Maham, Dschang, and Kendem [7] are shown for comparison.

Crustal Contamination
The parental melts of the Njimom dykes must have experienced some differentiation before intruding the upper crust, as reflected in the evolved compositions of all the samples of this study. The Njimom dykes were derived by fractional crystallization of olivine, clinopyroxene, and plagioclase from Mg-rich basaltic melts, consistent with the petrographic assemblages observed in these dykes and their low magnesium numbers (45-60). Trace element ratios such as Nb/U, Th/Nb, and La/Nb have been used to assess the role of crustal contamination in Njimom dykes because they are not strongly modified from their source material by partial melting or fractional crystallization processes. The continental crust has low Nb/U (4.4-25) and high La/Nb (1.6-2.6) and Th/Nb (0.24-0.88) ratios (e.g., [27]). Ocean island basalt (OIB) and mid-ocean ridge basalt (e.g., E-MORB and N-MORB) are both characterized by high Nb/U (>45) and low La/Nb (0.8-1.1) and Th/Nb (<0.1) ratios (e.g., [25,28]). In Njimom area, four dykes that intrude mylonites have Nb/U (19.4-26.7), Th/Nb (0.19-0.26), and La/Nb (1.2-1.5) ratios suggesting small crustal input. Furthermore, these dykes show small negative Nb anomalies on the primitive mantle-normalized multi-element patterns (Figure 11d), reflecting small crustal contamination. In contrast, the dykes that cross-cut the granites show higher Nb/U (43.3-50.0) and lower Th/Nb (~0.08) and La/Nb (0.7-0.8) (Figure 12). These values are close to those of MORB and OIB suggesting no crustal contamination. Interestingly, the same observations can be made for other mafic dykes coming from the southern continental part of the CVL [5][6][7], thus suggesting that these mafic dykes remained preserved from crustal influence during their evolution.  [27]. Data for oceanic basalts (OIB) and MORBs are from [25,28]. Data from Manjo [6]; Kekem [5]; Bangoua, Maham, Dschang, and Kendem [7] are shown for comparison.

Petrogenesis
Two different dykes groups are present in the Njimom area. There are sufficient geochemical and mineralogical evidences to deduce that the dykes that cross-cut the granites and mylonites are not comagmatic. For example, the different TiO2 and Al2O3 contents in the clinopyroxenes of the Njimom dykes, as well as their different incompatible element patterns ( Figure 11) cannot be produced by closed-system crystal fractionation of olivine, Figure 12. La/Nb vs Nb/U for the Njimom dykes. Data for upper continental crust (UCC) and lower continental (LCC) (star symbols) are from [27]. Data for oceanic basalts (OIB) and MORBs are from [25,28]. Data from Manjo [6]; Kekem [5]; Bangoua, Maham, Dschang, and Kendem [7] are shown for comparison.

Petrogenesis
Two different dykes groups are present in the Njimom area. There are sufficient geochemical and mineralogical evidences to deduce that the dykes that cross-cut the granites and mylonites are not comagmatic. For example, the different TiO 2 and Al 2 O 3 contents in the clinopyroxenes of the Njimom dykes, as well as their different incompatible element patterns (Figure 11) cannot be produced by closed-system crystal fractionation of olivine, plagioclase, clinopyroxene, and opaque oxides.
Because none of the compositions observed in Njimom area can be considered as a mantle-derived primary magma, it is likely that multiple saturation of Cr-spinel, olivine, clinopyroxene, and plagioclase occurred before the emplacement of all the magmas to shallow crustal levels. Therefore, we have modelled the extent of partial melting utilizing ratios of elements that are not modified by low to moderate fractionation of the phases mentioned in the preceding. In the Nd/Sm versus Tb/Yb diagram (Figure 13), the Njimom appear to have formed by small-degree melts (2-5%) of sources slightly enriched in incompatible elements in the spinel stability field. The melts from garnet-bearing sources could be present only in very minor amounts. The mildly enriched geochemical characteristics of the Njimom dykes are evident from ratios of Zr/Nb (5-8), Zr/Y (6-9), Nb/Yb (9)(10)(11)(12)(13)(14)(15)(16)(17), and Ti/V (68-114) that are similar to the values of (OIB) and E-MORB worldwide (e.g., [25]). Njimom dykes plot in the OIB field in the Ti-V diagram ( Figure 14). The chemical composition of the Nijmom dykes is generally similar to that observed in Kendem, Dschang, Bangangte, and Manjo dyke swarms, particularly in the abundance of relatively immobile incompatible elements. These similarities warrant that broadly similar mantle sources, possibly melted in different degrees, and petrogenetic histories were involved in the petrogenesis of the Paleozoic-Mesozoic Cameroon dykes. From the genetic point of view, the geochemical characteristics of the Paleozoic-Mesozoic Cameroon dykes indicate that they cannot be considered as typical of plume-derived melts. We therefore argue that the genesis of these dykes could be referred to as passive rifting and melting of the shallow lithosphere, possibly followed by crustal contamination of the mafic magmas during the ascent through the Precambrian crust. This intracontinental extensional setting could have been caused by the opening of the South Atlantic Ocean.
Geosciences 2022, 12, x FOR PEER REVIEW 29 of 32 that they cannot be considered as typical of plume-derived melts. We therefore argue that the genesis of these dykes could be referred to as passive rifting and melting of the shallow lithosphere, possibly followed by crustal contamination of the mafic magmas during the ascent through the Precambrian crust. This intracontinental extensional setting could have been caused by the opening of the South Atlantic Ocean.  [29]. Mode of the source, eutectics, and partition coefficients after [30].  [29]. Mode of the source, eutectics, and partition coefficients after [30]. Figure 13. (Nd/Sm)pm vs. (Tb/Yb)pm diagram illustrating non-modal fractional melting models for the Njimom dykes. The values along curves are the degrees of partial melting. The composition of the source is the lithospheric mantle of [29]. Mode of the source, eutectics, and partition coefficients after [30].

Geodynamic Implications
The Njimom mafic dykes have a strong NW-SE preferred orientation. In the Riedel fracture model for lineaments of Cameroon [8,9], it corresponds to the direction of major σ1 regional constraints. Previous studies on basaltic dykes in the southern continental part Figure 14. Ti-V diagram (after [31]) for Njimom dykes.

Geodynamic Implications
The Njimom mafic dykes have a strong NW-SE preferred orientation. In the Riedel fracture model for lineaments of Cameroon [8,9], it corresponds to the direction of major σ1 regional constraints. Previous studies on basaltic dykes in the southern continental part of the CVL record measurements of ca. 30 dykes (e.g., [9,32,33]). Two dykes at Kendem (near Mamfe) show a similar orientation. The Ar/Ar age of 192.10 ± 7.45 Ma recorded in Kendem (the lonely existing age for basaltic dykes of the southern continental part of the CVL) is older compared to the age of 125 million years considered as the beginning of the opening of the South Atlantic Ocean. Ar/Ar age significantly older (421.3 ± 3.5 Ma) has been obtained for a basaltic dyke outcropping along the ''Dschang's cliff' road at ca. 10 km to the west of the city of Dschang [7]. On a regional basis, the NW-SE direction is also known as the Benue direction in Cameroon which is linked to the Benue Through aulacogen initiated during the opening of the southern Atlantic (e.g., [34]). Alkaline magmatism and abundance of crustal xenoliths in some studied dykes can thus be linked to the paroxysmal stage preceding the opening of the Atlantic Ocean while transitional/tholeiitic affinities recorded for older (Paleozoic) basaltic dykes may better indicate simple reactivations of Precambrian fractures. The Njimom dykes could represent the transition between an older tholeiitic magmatism and the dominant Cretaceous alkaline magmatism of the Cameroon Volcanic Line.

Conclusions
On the basis of field, mineralogical and whole-rock geochemical data of the Njimom dykes, we present the following conclusions: (1) Two groups of mafic dykes that crosscut the Neoproterozoic basement rocks have been observed. A first group intrudes the mylonites whereas the second group intrudes the granites; (2) The Njimom dykes are alkaline basalts and hawaiites with a weakly porphyritic to aphyric texture and contain clinopyroxene, plagioclase, ± altered olivine and opaque oxides; (3) Geochemical variations at each dyke group are compatible with fractional crystallization with no or little crustal contamination, whereas the two dyke groups represent distinctive magma sources in the mantle; (4) Geochemical characteristics of the Njimom dykes can be modelled by partial melting (2-5%) of lherzolite slightly enriched in incompatible elements in the spinel stability field; (5) The geochemical features of Njimom dykes are similar to those observed in the Paleozoic and Mesozoic dykes recorded in the southern continental part of the Cameroon Volcanic Line, suggesting a similar mantle source evolution; (6) The magmatic activity in the Njimom area probably was synchronous with the initial phase of the opening of the southern Atlantic Ocean.

Data Availability Statement:
The whole-rock geochemical (major and trace element) and mineralogical data used in this manuscript are original and reported in the tables.