July 2024 LIP of the Month

Emplacement dynamics and geochemistry of the ca. 2.08 Ga Devarabanda dyke swarm, Dharwar craton: exploring sub-swarm variations

Srinjoy Datta¹, Sayandeep Banerjee¹, Amiya K. Samal¹*, Rajesh K. Srivastava¹, Richard E. Ernst², Aishwarya Mohan¹

¹Department of Geology, Banaras Hindu University, Varanasi- 221005, India

²Department of Earth Sciences, Carleton University, Ottawa, Canada K1S 5B6

Extracted and updated from:

Introduction:

Large Igneous Provinces (LIPs) essentially comprise volcanic flood basalt provinces, underlying plumbing architecture and magmatic underplating (Ernst, 2014). The flood basalt exposures of Precambrian LIPs are generally lost due to denudation, leaving behind the plumbing architecture as the only representative exposures (Ernst, 2014; Ernst et al., 2019; Srivastava et al., 2022). Therefore, only plumbing system of a LIP (dyke swarms and sill complexes) is available for its identification and additional studies. Further, it is accepted that dyke exposures close to the plume centre are generally emplaced under the stronger influence of the primary magmatic stress field, whereas those at greater distances are controlled by the regional stress regimes (e.g., Baragar et al., 1996; Ernst et al., 2001; Hou et al., 2010). However, emplacement mechanisms of dyke swarms can be more complicated than previously perceived (cf. Macdonald et al., 2010). Linear dyke swarms are generally emplaced through rifting and are perpendicular to the spreading direction. However, the stress fields associated with a radiating dyke swarm are usually more complex. The nature of influence, and the interplay between the local magmatic stress fields, regional stress regimes, and pre-existing structures in the emplacement of a radiating dyke swarm is still poorly constrained.

The Dharwar craton hosts several generations of Paleoproterozoic mafic dyke swarms that span in age from ca. 2.37 to ca. 1.79 Ga (Söderlund et al., 2019). Among these, the ca. 2.08 Ga Devarabanda dyke swarm stands out due to its remarkable radiating geometry, which is evident in three distinct sub-swarms: SS-1 (N- to NNE-trending), SS-2 (NW-trending), and SS-3 (NE-trending) (Fig. 1). This swarm is exposed around the Proterozoic Cudappah basin, and is suggested to have a proximal plume centre located beneath the basin (Kumar et al., 2015; Parashuramulu et al., 2022). The unique geometry of the swarm, and its proximity to the plume centre, deemed it as an appropriate candidate for investigating the influence of the regional and local (magmatic) stress regimes in the dyke emplacement. Further, the chemistry and emplacement mechanisms may often vary within a large radiating dyke swarm (Baragar et al., 1996). Therefore, investigating the three sub-swarms of the ca. 2.08 Ga Devarabanda dyke swarm offers an opportunity to explore potential variability in geochemistry and emplacement systems. We further integrated constraints from geochemistry, geophysics and anisotropy of magnetic susceptibility (AMS) to develop a comprehensive understanding about this dyke swarm. Geochemistry was primarily used to characterize the source and to verify the genetic relationship between the three sub-swarms. AMS characteristics were employed to study the magma flow dynamics and infer the emplacement direction. Further geophysical evidence was used to identify the sub-surface features which might have played a role in the dyke emplacement processes. Our study indicates the presence of multiple magma pulses and independent emplacement events for the three sub-swarms. This dyke swarm is also linked with opening of the Proterozoic Cudappah basin, further enhancing its tectonic significance. In the present work, the new geochemical and AMS data were also compared with available geochemical data (Samal et al., 2021; Parashuramulu et al., 2022) and magnetic fabric data (Kumar et al., 2015), substantially increasing the confidence in our inferences.

Geochemical variation and mantle source characteristics

The initial observations were made from petrographic investigation, where the SS-3 samples hosted substantial primary olivine phenocrysts, and a coarser average grain size than the other two sub-swarms (Fig. 2). The primitive mantle normalized multielement spider diagram and chondrite normalized rare earth element (REE) plot suggests that the SS-3 samples were more depleted than the SS-1 and SS-2 dykes (Fig. 3). Furthermore, the SS-3 dykes had Mg# (50 - 68.4), Ni (150 – 298 ppm) and Cr (159 – 316 ppm) concentrations, which indicate their more primitive nature. This was also substantiated by available geochronological data, which suggest SS-3 to have comparatively older age (2087 ± 5 Ma) (Söderlund et al., 2019), than the other two subswarms (2080 ± 6 to 2083 ± 1 Ma) (Kumar et al., 2015; Söderlund et al., 2019). The differences in the major and trace element compositions between the three sub-swarms is shown in Table 1.

Although there were some subtle geochemical differences in the major and trace element geochemistry, all the three sub-swarms displayed similar trends in the multielement and REE plots (Fig. 3). All the samples had inclined REE slope, with distinct LREE enrichment and flat HREE pattern (Fig. 3a). This suggested a limited contribution of garnet in the mantle source. Further, the multielement spider diagram displayed strong Nb-Ta-Ti negative anomaly, flat Zr-Hf pattern and positive Eu and Sr anomaly, which were dignified as distinct crustal signatures (Fig. 3b). All the three sub-swarms followed a similar differentiation trend of the major oxides and had overlapping HFSE values (Fig. 4 a-d) suggesting a common parental melt for all three sub-swarms. Figure 4e and f suggest that all samples were evolved under influence of olivine, clinopyroxene and Cr-spinel fractionation. However, the varying fractionation trends suggest differences in the degree of magma evolution in the magma chambers. The low Gd/Yb values and TiO₂/Yb values for all the three sub-swarms suggested their derivation from a shallow spinel dominated lherzolite source. This was better displayed graphically in the (La/Sm)_CN vs. (Gd/Yb)_CN plot, and the Nb/Yb vs. TiO₂/Yb plot (Fig. 5 a, b). Further, the robust LIP printing diagram (Pearce et al., 2021) attributed observed crustal signatures in the studied samples to subduction related metasomatism of the mantle source (Fig. 5c). Corroborating the geochemical evidences, we suggest that the three subswarms were derived from a common shallow mantle that was probably metasomatised by an earlier subduction event, prior to the dyke emplacement. Thereafter, they underwent magmatic differentiation in crustal magma chambers. The SS-3 was then emplaced as an earlier primitive pulse, and the other two subswarms were emplaced as later differentiated pulses.

Magnetic fabrics and structural framework

The three sub-swarms also show distinct differences in their magnetic fabrics. The SS-3 samples were characterised by sub-horizontal to inclined magnetic lineation suggesting horizontal magma injection, whereas the SS-1 and SS-2 samples dominantly displayed sub-vertical to inclined magnetic lineation indicating vertical magma injection (Fig. 6; Table 1). Further, the AMS fabrics in the host granites were compared with the dyke geometry to test the influence of the pre-existing strain fabrics in the dyke emplacement events (Fig. 7). The SS-1 and SS-2 dyke geometries have a low angle relationship with the AMS fabrics of the associated granite samples, indicating a favorable influence of the pre-existing strain fabrics in the dyke emplacement systems. On the contrary, the SS-3 dykes have a high angle relationship with the granite fabrics suggesting its emplacement under a stronger influence of magmatic stress regime related to the mantle plume, probably through primary fracture opening. Additionally, we analysed the gravity anomaly map of the Dharwar craton (Fig. 8; reference), to identify any sub-surface structures that might have affected the dyke emplacement events. We observed good spatial correlation of a linear gravity low zone (GL-1) with the SS-2 dykes. It was inferred as a pre-existing shear zone which might have facilitated the SS-2 emplacement. Earlier workers have related this zone as the Kurduwadi Lineament (Kale and Peshwa, 1988), which is also suggested as a major tectonic boundary in recent studies (Vasanthi and Santosh, 2021). Moreover, a gravity high zone (GH-1) was identified around the SS-1 exposures which might be related to probable crustal thinning and active rifting during the dyke emplacement. On the contrary, no such sub-surface feature was associated with the SS-3 dykes, indicating minor or no influence of pre-existing structural framework in its emplacement.

Emplacement model

Corroborating our evidence, an emplacement model (Fig. 9) has been proposed for this dyke swarm. We suggest that a plume head located underneath the present-day Cudappah basin, might have resulted in a domal uplift, that was accommodated by concentric fractures and normal faults. The magma generated from the metasomatised lithospheric mantle was guided by these normal faults and stored in crustal magma chambers of different configuration, surrounding the central domal uplift. The SS-3 was emplaced in the pre-rift phase, as a primitive pulse by lateral magma injection into radial fractures related to the domal uplift. The SS-2 and SS-1 were emplaced as subsequent magma pulses in the later rift phase primarily by vertical magma injection. The SS-2 emplaced along a pre-existing Kurduwadi lineament, probably reclaiming it as a rift zone, whereas the SS-1 was emplaced by active rifting and crustal thinning. The spatial variation and multiple pulses in the dyke swarm indicate diverse emplacement systems having a combined influence of pre-existing strain fabrics and the magmatic stress regime related to the plume centre. The emplacement model supports the apical graben concept (Baragar et al., 1996), suggesting subsidence of the domal uplift as a probable precursor to the opening of the Proterozoic Cudappah basin.

Conclusions

• The studied dyke swarm display distinct intra-swarm geochemical variation; the SS-3 having more primitive composition as compared to the SS-1 and SS-2 sub-swarms. All the three sub-swarms were derived from a common parental melt generated from a metasomatised spinel lherzolite mantle source.

• AMS studies indicate the SS-3 to be emplaced by lateral magma flow, whereas the SS-1 and SS-2 were emplaced by vertical magma injection. Comparison with the granite fabrics indicate a favorable relationship of the pre-existing strain fabrics in the host granites in the emplacement of the SS-1 and SS-2 samples, but the SS-3 was emplaced under the influence of the primary magmatic stress regime.

• An emplacement model was proposed, which suggest the magma to have mobilized to crustal magma chambers surrounding a central domal uplift due to the underlying plume, where the magma evolved independently. The SS-3 was emplaced as a primitive magma pulse by lateral magma injection, during an earlier pre-rift stage. Whereas the SS-1 and SS-2 were emplaced by vertical magma injection in a later rift stage. The SS-1 was emplaced by active rifting and crustal thinning, and the SS-2 was emplaced along a pre-existing Kurduwadi lineament. Eventually the domal uplift subsided to form an apical graben that acted as a precursor to the Cudappah basin.

References