April 2023 LIP of the Month

The Mutare-Fingeren Dyke Swarm of the eastern Kalahari Craton

Ashley P. Gumsley

Institute of Earth Sciences, University of Silesia in Katowice, 41-205, Sosnowiec, Poland

Email: ashley.gumsley@us.edu.pl

This has been extracted and modified from Gumsley, AP, et al. (2023). The Mutare–Fingeren dyke swarm: the enigma of the Kalahari Craton's exit from supercontinent Rodinia. In: Geological Society of London Special Publications 537, 1-22. https://doi.org/10.1144/SP537-2022-206 (Open Access)

Introduction

New U–Pb isotope dilution–thermal ionisation mass spectrometry (ID-TIMS) baddeleyite dating by Gumsley et al. (2023) identified remnants of a new ca. 724–712 Ma LIP on the eastern Kalahari Craton in southern Africa and East Antarctica (Figure 1): the combined Mutare–Fingeren Dyke Swarm. This dyke swarm occurs in north-eastern Zimbabwe (Mutare Dyke Swarm) and western Dronning Maud Land (Fingeren Dyke Swarm). During this time, the Rodinia supercontinent broke apart (Li et al., 2008). Rodinia break-up is associated with widespread intraplate magmatism on many cratons, including the ca. 720–719 Ma Franklin large igneous province (LIP) of Laurentia (Heaman et al., 1992; Denyszyn et al., 2009; Pu et al., 2022). Coeval magmatism has also been identified recently in Siberia (Ernst et al., 2016) and South China (Lu et al., 2022). This extensive magmatism terminates ∼1 Myr before the onset of the Sturtian Snowball Earth (Pu et al., 2022). However, LIP-scale magmatism and global glaciation are probably related.

The Mutare Dyke Swarm:

In eastern Zimbabwe of southern Africa, different trends of mafic dyke swarms exist on the Archaean to Mesoproterozoic Kalahari Craton (de Kock et al., 2019). One such dyke swarm, first described in Wilson et al. (1987), is the Mutare Dyke Swarm (Figure 1). This dyke swarm was described as having an arcuate shape, ranging from northwest-trending near the northern margin of the Zimbabwe Craton sector of the Kalahari Craton to north-northwest-trending near the eastern margin of the Zimbabwe Craton south of Mutare and into Mozambique, a distance of over ⁓400 km. The larger dykes in the swarm are well spaced, up to ⁓28 m in width, and can be traced discontinuously along strike for more than ⁓150 km. Approximately ⁓80 km to the west of Chimanimani on the eastern margin of the Zimbabwe Craton, this previously undated swarm merges with the sub-parallel Sebanga Dyke Swarm, which has a range of ages from ca. 2512 Ma to ca. 2470 Ma and ca. 2408 Ma (Söderlund et al., 2010). However, ca. 1112-1106 Ma Umkondo-aged dykes are also present in the region with a similar trend (de Kock et al., 2014). Further, the Mutare Dyke Swarm cuts the Sebanga Dyke Swarm in northern Zimbabwe and the ca. 1886-1872 Ma Mashonaland Sill Province (of the Mashonaland LIP; Söderlund et al., 2010; Hanson et al., 2011; Stidolph, 1977). The Mutare Dyke Swarm was originally considered part of the ca. 1112-1106 Ma Umkondo LIP (Wilson et al., 1987). However, based on a limited whole-rock Rb-Sr date and palaeomagnetism on two dykes north-east of Harare (i.e., the Mahumi and Chenjera dykes; Stidolph, 1977), Wilson et al. (1987) suggested the age of the swarm was at least ca. 500 Ma, but not more than ca. 700 Ma. This interpretation agrees with Leitner and Phaup (1974), who stated that these dykes pre-date late Neoproterozoic to Cambrian metamorphism. Mukwakwami (2004) presented an unpublished U-Pb ID-TIMS baddeleyite weighted mean ²⁰⁶Pb-²³⁸U date of 724.0 ± 2.1 Ma on this swarm in the Chimanimani area. This linear sub-vertical dolerite dyke trends at ⁓156° and is at least ⁓40 m wide. Itis intrusive into the low-grade metamorphosed Mesoproterozoic Umkondo Group on the eastern margin of the Zimbabwe Craton. This dolerite was interpreted to be older than ca. 180 Ma Karoo LIP magmatism. The sample was taken from what was originally interpreted as a feeder into the voluminous ca. 1112-1106 Ma Umkondo LIP (e.g., Watson, 1969; de Kock et al., 2014). However, Mukwakwami (2004) has shown that this dyke belongs to the younger the Mutare Dyke Swarm. The sampling site of Mukwakwami (2004) was re-sampled by Gumsley et al. (2023). The new age agrees with the unpublished 724.0 ± 2.1 U-Pb ID-TIMS age on baddeleyite by Mukwakwami (2004), with the ca. 712-724 Ma age obtained from the same sample site of Mukwakwami (2004). Further north on the eastern margin of the Zimbabwe Craton, in the Nyanga area, dolerites were generally classified as belonging to two types, based mostly on the palaeomagnetic studies of McElhinny and Opdyke (1964), according to Stocklmayer (1978). These two dolerite types are now included within the ca. 1886-1872 Ma Mashonaland LIP (e.g., Söderlund et al., 2010) and the ca. 1112-1106 Ma Umkondo LIP (e.g., de Kock et al., 2014), the existence of which was verified by Wingate (2001) in the dolerite sills of the area. The Mutare Dyke Swarm was also interpreted to be in the area according to Wilson et al. (1987), as was shown from geochemical evidence by Ward et al. (2000). In Gumsley et al. (2023), a U-Pb ID-TIMS baddeleyite age of 715 ± 22 Ma was obtained on a dykes in the Nyanga areas of eastern Zimbabwe (Figure 2a). Two K-Ar whole-rock ages of 715 ± 70 Ma and 757 ± 20 Ma presented by Vail and Dodson (1969) on dolerites from the same region are likely related to the same magmatic event.

The Fingeren Dyke Swarm

In the Grunehogna Craton sector of the Kalahari Craton in western Donning Maud Land of East Antarctica, north-northeast-trending dykes intrude into the Mesoproterozoic Ritscherflya Supergroup (Figure 1). They also cut across the ca. 1112-1106 Ma Umkondo LIP-aged Borgmassivet sills. The north-northeast-trending dykes were assigned an Umkondo LIP-age in the Borgmassivet area. In contrast, further to the north in the Ahlmanryggen area, both ca. 1112-1106 Ma Umkondo LIP-aged dykes and ca. 190-178 Ma Karoo-aged dykes, following the same trend, have been described more intensively (Riley et al., 2005; Riley and Millar, 2014). The older north-northeast-trending dykes received only cursory study until the work of Gumsley et al. (2023), where they are termed the Fingeren Dyke Swarm and dated by U-Pb ID-TIMS on baddeleyite to 716.6 ± 1.2 Ma and 718.9 ± 5.8 Ma. These two dyke samples were taken in the Borgmassivet area, both from linear sub-vertical dolerite dykes trending ⁓011° that are ⁓40-80 m wide on the Fingeren and Veten nunataks (Figure 2b). The dykes are intrusive into the low-grade metamorphosed Ritschtersflya Supergroup (Veten Member, Högfonna Formation, Ahlmannryggen Group) and a Umkondo-aged Borgmassivet sill. These dykes were generally considered to be feeders of the Borgmassivet sills in this area, according to the mapping of Wolmarans and Kent (1982).

Petrography

Both dyke swarms display variable amounts of deuteric alteration, as was recorded by Vail (1966), together with low greenschist facies metamorphism, especially closer to the margin of the Kalahari Craton in both the Zimbabwe Craton sector and the Grunehogna Craton sector. This alteration is documented by the partial replacement of pyroxenes and plagioclase feldspar by amphibole, epidote, chlorite and sericite. Likely, Ediacaran to Cambrian deformation and metamorphism on the eastern margin of Kalahari from the Mozambique/Maud Belt is responsible (e.g., Jacobs et al., 2003). However, whereas samples of the Mutare Dyke Swarm are usually low- to medium-grade greenschist facies, samples of the Fingeren Dyke Swarm are very low- to low-greenschist facies.

Geochemistry

Ward et al. (2000) considered that the dykes of the Mutare Dyke Swarm were composed of three geochemical groupings (i.e., M1, M2 and M3), with modelling failing to indicate any genetic relationship between them. In de Kock et al. (2014), the M1 grouping was linked to the ca. 724 Ma magmatic event based on the unpublished age of Mukwakwami (2004). The work of Gumsley et al. (2023) confirmed this hypothesis. The samples from Gumsley et al. (2023) of both the Mutare Dyke Swarm and the Fingeren Dyke Swarm have high-Fe tholeiitic gabbroic/basaltic EMORB-like compositions (Figure 4; 5), along with other members of the Mutare Dyke Swarm (the M1 geochemical group in Ward et al. (2000). Therefore, of the three compositions identified by Ward et al. (2000), only M1 is further considered to belong to the Mutare Dyke Swarm dated by Gumsley et al. (2023) to ca. 724-712 Ma. Regarding the other two compositional groups identified by Ward et al. (2000), one (M2) likely belongs to the Sebanga Dyke Swarm (ca. 2512-2408 Ma; Söderlund et al., 2010), which merges with the Mutare Dyke Swarm to the west. The other (M3) is likely ca. 1112-1106 Ma Umkondo age, based on a dated dyke at Devuli Ranch at 1110 ± 19 Ma (de Kock et al., 2014), which has proximity and a similar trend to the Mutare Dyke Swarm. The coeval Mutare Dyke Swarm and the Fingeren Dyke Swarm appears to be the product of melting in the asthenosphere (Figure 5).

Although dykes with such EMORB-like compositions have not been reported on the Grunehogna Craton, ⁓300 km to the southwest in the Maud Belt, such dykes exist in the Heimefrontfjella (Figure 1; Bauer et al., 2003). These dykes have been assigned an age of ca. 586 Ma based on a single zircon grain (Bauer et al., 2003), which is inconclusive. Dykes have also been mapped ⁓100 km south of the Grunehogna Craton in the Kirwanveggan area of the Maud Belt (Grantham et al., 1995), but they are north-east-trending although they may be of similar age. These dykes were noted to intrude the late Meso- to early Neoproterozoic Maud Belt and are heavily deformed and metamorphosed in places by Ediacaran to Cambrian orogenesis (Grantham et al., 1995; Bauer et al., 2003). Approximately ⁓200 km east of the Grunehogna Craton, further dykes of similar geochemistry were noted in the Maud Belt from H.U. Sverdrupfjella (Grosch et al., 2007) and were assigned an age of ca. 800 Ma. These dykes, however, have been metamorphosed at amphibolite facies. Further U-Pb geochronology and geochemical studies are required to determine the full extent of the Fingeren Dyke Swarm in Grunehogna and the surrounding Maud Belt of Kalahari.

A new large igneous province

The combined Mutare–Fingeren Dyke Swarm on the Kalahari Craton is interpreted as a LIP, as defined by Ernst (2014). Definitions vary, but an estimate of >0.1 Mkm² and a time interval of <10 Myr can be diagnostic, and the combined Mutare–Fingeren Dyke Swarm meets these criteria. LIPs are typically assigned to mantle plume events (e.g., Ernst 2014). However, the Mutare–Fingeren Dyke Swarm, in large parts, occurs relatively close to the margins of the Zimbabwe and Grunehogna cratons and may even intrude into the Mesoproterozoic Maud Belt of Kalahari. East of the Grunehogna Craton in central Dronning Maud Land, near the Schirmacher Oasis, East Antarctica, remnants of a continental magmatic arc have yielded U–Pb zircon ages of ca. 785–770 Ma, as shown by Jacobs et al. (2020). Jacobs et al. (2020) inferred that subduction of old oceanic crust beneath the Kalahari Craton margin resulting in slab roll-back leading to back-arc extension. There is no evidence, however, for possible back-arc extension along this part of the margin until ca. 650–600 Ma, which involved ultra-high temperature metamorphism and the intrusion of syntectonic granites interpreted to record asthenospheric upwelling, as suggested by Baba et al. (2010). The discrepancy in timing between arc magmatism/back-arc extension in the Schirmacher Oasis region and intrusion of the Mutare–Fingeren Dyke Swarm at ca. 724–712 Ma does not support a direct relationship between these events. Further north along the Kalahari margin, Mesoproterozoic arc rocks within the Mozambique Belt outboard of the eastern margin of the Zimbabwe Craton show generally strong Ediacaran–Cambrian overprinting similar to that documented in the Maud Belt (e.g. Fritz et al. 2013; Chaúque et al. 2019; Thomas et al. 2022). The only documented Neoproterozoic arc-related rocks in this large area are ca. 820–700 Ma meta-volcanic and meta-plutonic rocks present within the far travelled Cabo Delgado Nappe Complex in northeastern Mozambique, which is interpreted to have formed in oceanic arcs or as part of small continental blocks juxtaposed against older Mesoproterozoic arc crust during Gondwana assembly (Viola et al. 2008; Boyd et al. 2010). The available data from Gumsley et al. (2023) thus provide no clear evidence for a causal relationship between the emplacement of the Mutare–Fingeren Dyke Swarm and tectonomagmatic events along the entire eastern Kalahari margin. We conclude that the dyke swarm results from the decompression melting of upwelling asthenosphere mantle, possibly a mantle plume, unrelated to plate tectonic processes along the eastern Kalahari margin.

Acknowledgments

APG acknowledges financial support through a grant from the National Science Centre (Naradowe Centrum Nauki; NCN), Poland (SONATINA 3 grant no. UMO-2019/32/C/ST10/00238).

References