Evidence for participation of microbial mats in the deposition of the siliciclastic ‘ore formation’ in the Copperbelt of Zambia

Abstract

The Copperbelt of Zambia is the world’s largest sediment-hosted stratiform copper province, hosted in siliciclastic sediments of the Roan Group, which forms the basal part of the Neoproterozoic–Paleozoic Katanga Supergroup. Much of the ore deposition occurred between 880 Ma and 780 Ma, on a rimmed platform consisting of a carbonate barrier, a lagoonal basin and tidal flats grading into sabkhas in the hinterland. Various sedimentary structures developed in the ore formation at the Mindola Open Pit mine, are herein considered to be microbially induced and are identified as microbial shrinkage cracks, wrinkle structures, mat deformation structures, petees, concentric microfaults, and microbial mat chips. The occurrence of these structures in all ore formation units at the Mindola Mine suggests microbial mats grew on the paleo-sediment surface throughout deposition of the cupriferous succession. As these structures require cohesive layers, the mats were likely of the cyanobacterial type, that grew in the well aerated intertidal to lower supratidal zones. Cyanobacterial mats typically consist of a surface layer of filamentous cyanobacteria underlain by anaerobic, heterotrophic sulfate reducing bacteria (SRB). A distinct sulfide mineral zonation, developed in all major deposits of the Copperbelt, ranges from barren supratidal (sabkha) sediments, through chalcocite in the lower supratidal zone, to bornite followed by chalcopyrite in the intertidal zone, and pyrite in the subtidal zone and anoxic lagoonal depotcentre. This sequence of minerals can be modelled as a paragenetic sequence of mineralization resulting from the progressive reduction of a source fluid, indicating that geochemical conditions of ore formation, at least, are produced by the activity of SRB.

Introduction

The Copperbelt of Zambia (Fig. 1) and the adjoining Katanga province (Democratic Republic of Congo; DRC) contain the largest sediment-hosted stratiform Cu–Co deposits on Earth, with altogether more than 190 million tonnes measured production and reserves (Kirkham, 1989). In the Zambian Copperbelt, the main mineralized horizon is the ‘ore formation’, which lithologically ranges from arkosic sandstone to siltstone and argillite, locally with a considerable carbonate component. The main part of the ‘ore formation’ is the ‘ore shale’, which is a true argillite or, where slightly coarser grained, a siltstone. The ‘ore shale’ is in places carbonaceous, ranging from a slightly carbonaceous siltstone/argillite to a carbonaceous shale (commonly containing pyrite) with free carbon up to ca. 1.5% (Mendelsohn, 1961).

After decades of discussion about the syn- or epigenetic origin of the ore deposits (see reviews by Sweeney et al., 1991, Cailteux et al., 2005), it has been proposed that mineralization occurred “during early diagenesis, primarily by bacterial action on metal-rich solutions derived from the pre-Katangan basement rocks” (Sweeney and Binda, 1994, p. 303). Replacement of primary sulfate by carbonates and concomitant formation of sulfide was considered to have been the major mineralizing process, the underlying bio-chemical reactions being bacterial reduction of seawater sulfate and formation of carbonates as a by-product (Sweeney et al., 1986, Sweeney et al., 1991, Sweeney and Binda, 1989). This is supported by δ¹³C values from −8.77 to −20.52 per mil (PDB) in the carbonates, indicating a dominantly organic source (Sweeney et al., 1986). In the last years, however, field and isotopic data have been presented for the Nchanga deposit in Zambia, which clearly indicate an epigenetic, replacive origin of this deposit, without any linkage to bacterial sulfide formation (McGowan et al., 2003, McGowan et al., 2006). According to this model, mineralization resulted from fluid mixing during basin inversion at the onset of Lufilian deformation. Epigenetic fluid mixing has also been suggested for a stratiform copper mineralization in the Lufilian foreland (DRC) by El Desouky et al. (2008). Nevertheless, and to continue the debate, further evidence has also been collected in favour of early diagenetic mineralization, particularly in the DRC part of the Copperbelt (e.g. Muchez et al., 2008). Considering all the observations made and evidence collected over almost 100 years of research in the Copperbelt, a multistage formation of the ore deposits appears likely, starting during early diagenesis and ending during the Lufilian Orogeny and subsequent local supergene enrichment (see also Selley et al., 2005, Dewaele et al., 2006). The present paper contributes to the debate and elucidates the possible role of microbial mats in the accretion and early diagenetic mineralization of the siliciclastic ‘ore formation’.

Stromatolitic buildups have been observed in carbonate rocks laterally adjoining to and locally overstepping the ‘ore formation’. For instance, on the SW side of the Kafue anticline (Fig. 2), individual biohermal dolomite reefs are developed on paleotopographic highs fringing the Chambishi and Roan basins (Annels, 1984). These stromatolitic reefs are usually not mineralized and appear as ‘barren gaps’ in the ‘ore formation’. From the siliciclastic ‘ore formation’, however, where bacterial activity is claimed to have been responsible for the mineralization during early diagenesis, signatures of microbial communities and constructions have not been reported yet. This is not surprising, because in siliciclastic deposits stromatolitic buildups are usually not developed (Seilacher, 1999) and the organic matter is easily destroyed (Krumbein and Cohen, 1977, Park, 1977).

Despite this, it has been recognized in the last years that a range of unusual sedimentary structures in siliciclastic rocks are likely related to ancient microbial mats and their responses to physical processes acting upon them, like tractional currents and subaerial exposition and desiccation. It has been proposed to group these structures under the terms ‘microbially induced sedimentary structures’ (MISS; Noffke et al., 2001b, Noffke et al., 2008) or ‘mat-related structures’ (MRS). A catalogue of such structures related to modern microbial mats in siliciclastic peritidal zones was presented by Gerdes et al. (2000a), whereas ancient MRS have been described, among others, by Schieber, 1986, Schieber, 1998, Schieber, 1999, Pflüger and Gresse, 1996, Hagadorn and Bottjer, 1997, Hagadorn and Bottjer, 1999, Pflüger, 1999, Gehling, 1999, Gehling, 2000, Seilacher, 1999, Noffke, 2000, Porada and Löffler, 2000, Bouougri and Porada, 2002, Prave, 2002, Noffke et al., 2003. An atlas illustrating both modern and ancient MRS has recently been presented by Schieber et al. (2007).

Studies of modern cyanobacterial mats in siliciclastic peritidal domains (e.g. Krumbein et al., 1994, Noffke et al., 1996, Noffke et al., 2001a, Noffke et al., 2001b, Noffke et al., 2003, Gerdes et al., 2000a, Gerdes et al., 2000b) revealed that microbes, if occurring abundantly in peritidal systems, tend to form “compact microbial layers that project above the siliciclastic sediment” (Gerdes et al., 2000a). Such organic layers are complex microbial communities which characteristically develop a vertical lamination dominated by “a few functional groups of microbes: cyanobacteria, colorless sulfur bacteria, purple sulfur bacteria, and sulfate reducing bacteria” which successfully cooperate in a kind of “joint venture” (van Gemerden, 1993, p. 3). In intertidal to lower supratidal zones, the top portion of a mat is usually developed as a strongly cohesive, felty to leathery layer, 2–5 mm thick, consisting of interwoven bacterial filaments, clusters of coccoid bacteria and abundant sticky EPS (extracellular polymeric substances). In the portion below, fermentation and degradation processes prevail that favour taxic and functional diversity of the joint venture community. Finally, in the bottom portion, anaerobic sulfate-reducing bacteria (SRB) benefit from the continuity of production and degradation of organic matter and the presence of sulphate dissolved in pore water and produce, among others, hydrogen sulfide.

Metal binding and ore-forming processes are to be expected in the anoxic, iron sulfide-rich layer of SRB, whereas structure forming processes will mainly affect the cohesive surface layer, due to its steady subaqueous or subaerial exposure. For the purpose of this paper, five groups of structures are distinguished: (i) structures reflecting mat morphology, e.g., “wrinkle structures” (see review by Porada and Bouougri (2007)); (ii) structures related to mat distortion and deformation (crumpling, folding, slumping); (iii) structures related to mat erosion, e.g. “microbial mat chips” (Gerdes et al., 2000a, Gerdes et al., 2000b), “microbial sand chips” (Pflüger and Gresse, 1996); (iv) structures related to mat desiccation and shrinkage, collectively termed “microbial shrinkage cracks” (Porada and Löffler, 2000, Bouougri and Porada, 2002); (v) structures related to lateral mat growth/expansion or to trapping of gases uprising from decaying buried mats, e.g. ’petees‘ (see Reineck et al. (1990)) for definition of term), bulges or growth domes (Bouougri et al., 2007).

In the first part of this paper, sedimentary structures observed at Mindola Open Pit (MOP) are presented and discussed, which may have been caused by ancient microbial mats that colonized on siliciclastic sediment surfaces in peritidal zones of the Zambian ‘ore formation’. In the second part, implications are discussed with respect to mineralization and the mineralizing process potentially shaped by the presence of these microbial mat communities. Strictly, inferences drawn from observations at MOP are valid for this locality only, but may be extended to other parts of the Copperbelt, in which similar depositional conditions prevailed.