Focussed hydrothermal alteration in upper crustal oceanic faults on Macquarie Island

• Main Author: Lewis, SJ
• Format: Thesis
• Language: English
• Published: 2007
• Subjects: Hydrothermal alteration; Geology; Structural; Faults (Geology); Caroline Creek; Pacific; Southern Ocean; Wall Rocks; Macquarie Island
• Online Access: https://eprints.utas.edu.au/20655/; https://eprints.utas.edu.au/20655/1/whole_LewisStevenJohn2007_thesis.pdf; https://eprints.utas.edu.au/20655/2/whole_LewisStevenJohn2007Vol2_thesis.pdf

Macquarie Island is an uplifted exposure of oceanic lithosphere lying deep within the Southern Ocean, less than ten kilometres east of the dextral transpressional boundary which separates the Indo-Australian and Pacific plates (54° 30' S, 158° 55' E). The island is a geological oddity; essentially an intra-oceanic ophiolite residing in its primary marine basin. Macquarie Island hosts a diverse spectrum of igneous rock types from all stratigraphic levels of the ocean crust, with pillow basalts and sheeted dolerite dykes most abundant. These upper crustal rocks were formed in the Late Miocene during slow crustal accretion and rifting (spreading rate ~ 20 mm per year) at the relict Proto-Macquarie Spreading Ridge (PMSR). Typical of many slow-spreading (magma-poor) mid-ocean ridges, tectonic activity played a dominant but episodic role in the geological evolution of Macquarie Island. Faults, fractures, and brittle shear zones are widespread across the island, and discrete igneous rock domains commonly have faulted margins. Many discontinuities were originally formed during arnagmatic extension at the PMSR, although more recent transpressive tectonism (< 10 Ma) has further modified or reactivated most of these structures. An extensive array of neotectonic faults have also formed during ongoing uplift and deformation (post-spreading). The Major Lake, Caroline Cove, and Sellick Bay Faults are prominent structural zones which cut sharply across upper crustal rocks in central and southern Macquarie Island. These steeply dipping fault systems are oriented subparallel (NW to NNE-strike) and are regularly spaced across the island (~ 10 km apart). Each fault forms a major geological boundary separating distinct lithologic domains which host disparate rock types (pillow basalts, sheeted dykes, and their transitional volcanic—intrusive packages) and regional alteration assemblages (recharge-related). Variably influenced by multiple episodes of extensional and strike-slip tectonic activity, all of these significant crustal discontinuities initially developed during seafloor-spreading at the PMSR. However, their orientations are largely oblique to the ENE trend of the palaeo-rift axis, suggesting that they may have formed structural accommodation zones during the waning stages of slow extension, or that they were related to active tectonism at a proximal ridge—transform intersection. Structurally focussed zones of intensely altered volcanic rocks and sheeted dolerite dykes are intimately associated (spatially and genetically) with the Major Lake, Caroline Cove, and Sellick Bay Faults. Consistent geological and hydrothermal relationships attest to their critical role as crustal pathways and fracture conduits for hydrothermal systems. These major oceanic structures host highly distinctive hydrothermal assemblages which mainly consist of semi- to pervasively altered basaltic wall rocks, abundant hydrothermal veins and breccias, and patchy domains of massive sulfide mineralisation. The alteration zones comprise (in total) six focussed hydrothermal facies, which are here defined as: (1) the vein and breccia, quartz-chlorite (VQC) facies; (2) the massive and veined, chlorite-quartz-pyrite (CQP) fades; (3) the vein-dominated, prehnite-zeolite (VPZ) fades; (4) the foliated, massive chlorite (FMC) fades; (5) the pervasive, Fe-oxyhydroxide overprint (PFO) facies; and (6) the narrow, focussed quartz vein (NQV) fades. The formation of a single diagnostic facies dominated peak hydrothermal conditions in each major structure; the VQC fades in the Major Lake Fault, the CQP facies in the Caroline Cove Fault, and the VPZ fades in the Sellick Bay Fault. In addition, both the VQC and CQP fades are partially overprinted by smaller-scale alteration assemblages that post-dated peak hydrothermal activity, i.e., the FMC and PFO fades. The dominant fault-hosted alteration facies are highly anomalous, multi-component hydrothermal assemblages. They each have significantly different alteration minerals, hydrothermal textures, and physical attributes compared to regional igneous rock domains, i.e., non-fault zone crustal blocks. Multiple episodes of superimposed fluid—rock interaction produced these distinctive fault-hosted assemblages, with focussed hydrothermal activity mostly post-dating the main period of axial magmatism at the PMSR (crustal accretion). However, high-temperature fluid circulation was strongly influenced (initiated and sustained) by small-scale dyke pulses, mostly injected at deeper crustal levels than the present fault exposures. Rapid fluid discharge from the basement reservoir was probably also initiated (in part) by discrete tectonic events, such as oblique extension localised along specific fault segments. The physical manifestations of these relict fluid flow events are now well exposed at six key sites on Macquarie Island, where they afford an unparalleled opportunity to study the processes and products of structurally focussed hydrothermal activity in the ocean crust. The Major Lake Fault Zone (MLFZ) is one of the best preserved and most extensive oceanic discontinuities on Macquarie Island. Mainly oriented NNW along its ~ 4 km-long strike, the MLFZ separates upper greenschist and lower amphibolite facies rocks of the Sandell Bay and Lusitania Bay Dyke Swarms (in the southern footwall) from zeolite facies pillow basalts in the northern hangingwall. The fault is the main structural host of quartz + chlorite-dominated alteration (VQC facies) and is also associated with late-stage foliated, massive 'chlorite' (FMC) fades. Furthermore, five planar alteration zones representing the narrow (< 1-2 m-wide), focussed quartz vein (NQV) facies are situated in the regional footwall package ~ 1-1.5 km south of the MLFZ. There are four discrete VQC facies sites at ~ 750 m-spaced intervals along the length of the Major Lake Fault, forming 1-15 m-wide and 80-150 m-long alteration zones. The distinctive spatial segmentation of VQC fades zones likely reflects small-scale hydrothermal upflow systems, locally controlled by basement structures such as obliquely intersecting faults or slight variations in the orientation of the MLFZ, e.g., structural step-overs, kinked zones or offsets. In contrast, the FMC fades is only well exposed at a single 10 m-wide and 55 m-long outcrop segment in the central deformation corridor of the MLFZ (although its true strike extent and physical dimensions are uncertain due to the effects of surface cover and erosion). The VQC fades comprises a multi-stage paragenetic association of hydrothermally derived alteration minerals. Quartz, chlorite and albite are dominant components, whereas epidote and pyrite occur sporadically in altered basalts and dolerites. The main VQC phases are reflected by moderate to strong enrichments in whole-rock Si and Fe concentrations, consistently depleted levels of Ca (due to igneous plagioclase alteration, e.g., the conversion of labradorite to albite), and highly elevated loss-on-ignition values (enhanced volatile contents). The earliest alteration stage is characterised by selectively pervasive chlorite alteration, which is widespread throughout the primary igneous groundmass. Chlorite also forms narrow veinlets and inEtlls many vuggy cavities. Importantly, chlorites in the VQC fades are compositionally diverse and extend from Mg-rich to Fe-rich varieties (Fe # range = 0.29 to 0.64), with many also containing > 1 wt. % Mn (strongly suggestive of a fluid mixing origin). Hydrothermal fluids mostly ranged from 220°-260° C (mean = 237° C) during the precipitation of chlorite (paragenetic Stage I). In contrast, fluids were up to ~ 50° C hotter during the later formation of discrete quartz veins, intricate quartz stockwork arrays, and irregularly shaped patches of massive quartz alteration (Stage II). However, this diagnostic quartz-bearing stage precipitated across a much broader thermal range (176°-309° C), with fluid temperature variations of up to ~ 80° C occurring between discrete sites and different quartz sub-stages along the MLFZ. The distribution and abundance of pyrite trace elements in the VQC facies also varies at different outcrop sites, and are moderately well correlated with quartz precipitation temperatures. The most significant trace element enrichments are for Co, Ni, Cu, Zn, and Pb. Highly elevated levels of these elements mainly occur in pyrites associated with the lower temperature range of VQC quartz (≤220° C), whereas only Se concentrations are consistently enriched at higher temperature sites on the MLFZ. In contrast, sulfur isotope compositions of VQC facies pyrite are not site-specific. Their 834S values encompass the known range of seafloor-hosted sulfides (-1 ‰ to +11.9 ‰) although most are within +1 to +4 ‰, indicating that sulfur was mainly derived from leached magmatic crust. The Caroline Cove Fault Zone (CCFZ) is variably exposed for ~ lkm along-strike at the south-western extremity of Macquarie Island. This NNW to NNE-oriented structure is dominated by abundant small-scale normal- and sinistral-oblique faults. Clay-bearing gouge and breccia zones (neotectonic) occur commonly in the central deformation corridor, clearly post-dating the main period of hydrothermal alteration. The CCFZ cuts across regional pillow basalts with zeolite and lower greenschist facies assemblages, and also partly forms the north-eastern boundary of the massive and veined, chlorite-quartz-pyrite (CQP) facies. The diagnostic CQP facies outcrops exclusively at Caroline Cove (northern end of the CCFZ), where it forms a fault-bound wedge of intensely altered volcanic rocks ~ 200 m-long and up to 80 m-wide (strike extent limited by neotectonic faulting and surface cover). Sporadic occurrences of the foliated, massive outcrop segment in the central deformation corridor of the MLFZ (although its true strike extent and physical dimensions are uncertain due to the effects of surface cover and erosion). The VQC facies comprises a multi-stage paragenetic association of hydrothermally derived alteration minerals. Quartz, chlorite and albite are dominant components, whereas epidote and pyrite occur sporadically in altered basalts and dolerites. The main VQC phases are reflected by moderate to strong enrichments in whole-rock Si and Fe concentrations, consistently depleted levels of Ca (due to igneous plagioclase alteration, e.g., the conversion of labradorite to albite), and highly elevated loss-on-ignition values (enhanced volatile contents). The earliest alteration stage is characterised by selectively pervasive chlorite alteration, which is widespread throughout the primary igneous groundmass. Chlorite also forms narrow veinlets and infills many vuggy cavities. Importantly, chlorites in the VQC fades are compositionally diverse and extend from Mg-rich to Fe-rich varieties (Fe # range = 0.29 to 0.64), with many also containing > 1 wt. % Mn (strongly suggestive of a fluid mixing origin). Hydrothermal fluids mostly ranged from 220°-260° C (mean = 237° C) during the precipitation of chlorite (paragenetic Stage I). In contrast, fluids were up to ~ 50° C hotter during the later formation of discrete quartz veins, intricate quartz stockwork arrays, and irregularly shaped patches of massive quartz alteration (Stage II). However, this diagnostic quartz-bearing stage precipitated across a much broader thermal range (176°-309° C), with fluid temperature variations of up to ~ 80° C occurring between discrete sites and different quartz sub-stages along the MLFZ. The distribution and abundance of pyrite trace elements in the VQC facies also varies at different outcrop sites, and are moderately well correlated with quartz precipitation temperatures. The most significant trace element enrichments are for Co, Ni, Cu, Zn, and Pb. Highly elevated levels of these elements mainly occur in pyrites associated with the lower temperature range of VQC quartz (≤ 220° C), whereas only Se concentrations are consistently enriched at higher temperature sites on the MLFZ. In contrast, sulfur isotope compositions of VQC facies pyrite are not site-specific. Their `δ^34S` values encompass the known range of seafloor-hosted sulfides (-1 ‰ to +11.9 ‰ ) although most are within +1 to +4 ‰ , indicating that sulfur was mainly derived from leached magmatic crust. The Caroline Cove Fault Zone (CCFZ) is variably exposed for ~ 1 km along-strike at the south-western extremity of Macquarie Island. This NNW to NNE-oriented structure is dominated by abundant small-scale normal- and sinistral-oblique faults. Clay-bearing gouge and breccia zones (neotectonic) occur commonly in the central deformation corridor, clearly post-dating the main period of hydrothermal alteration. The CCFZ cuts across regional pillow basalts with zeolite and lower greenschist facies assemblages, and also partly forms the north-eastern boundary of the massive and veined, chlorite-quartz-pyrite (CQP) facies. The diagnostic CQP facies outcrops exclusively at Caroline Cove (northern end of the CCFZ), where it forms a fault-bound wedge of intensely altered volcanic rocks ~ 200 m-long and up to 80 m-wide (strike extent limited by neotectonic faulting and surface cover). Sporadic occurrences of the foliated, massive chlorite (FMC) facies and the pervasive, Fe-oxyhydroxide overprint (PFO) fades also occur in the CCFZ. Both of these late hydrothermal assemblages are spatially restricted to the 1-2 m-wide, highly tectonised core of the fault zone (well exposed in the lower Caroline Creek valley), where they have partially overprinted intensely altered CQP facies basalts and some regional (less altered) volcanic rocks in the hangingwall package. Similar to intensely altered rocks in the MLFZ, the CQP facies consists of a multi-stage paragenetic association with abundant chlorite, quartz, and albite. However, the CQP assemblage is more diverse; both pyrite and epidote are also dominant alteration minerals which form massive groundmass patches or interconnected vein arrays. Minor chakopyrite is an important component of the sulfide-bearing assemblage, and late-stage barite infills some quartz vein cavities. The suite of alteration minerals strongly influences whole-rock geochemical compositions in the CQP facies. Moderate to strong enrichments in Si, Fe, S, Zn, Cu, Ba, and Mg are characteristic of these altered basalts, whereas only Ca is consistently depleted. Significant elemental additions are associated with net mass gains which range from 50-500 %, and the intense hydrothermal effects are further reflected by 5-30 % enrichments in the main volatile components, i.e., elevated loss-on-ignition values. Early stage chlorites are texturally comparable to those in the VQC facies, although all are Mg-rich (mean Fe # = 0.33) and have relatively lower Mn contents (mean Mn = 0.6 wt. %). Stage I CQP chlorites precipitated from slightly lower temperature hydrothermal fluids than those in the MLFZ (mean = 221° C), whereas massive and vein quartz in paragenetic Stage II formed between ~ 230-305° C (mostly ~ 300° C). Sulfides and epidote also precipitated during the peak quartz-forming stage, although the parent fluids of the CQP facies were not as highly enriched in trace elements as their VQC counterparts. Pyrites in the CQP facies have relatively low concentrations of metallic trace elements, with only Co, Ni, and Se consistently elevated above regional background levels. Pyrite sulfur isotope compositions are mostly +5 to +9 ‰, although some are highly `δ^34S`—enriched (up to +11.9 ‰ ). The presence of Mg-rich chlorites and late-stage barite, the consistent addition (mass gain) of whole-rock Mg, the low pyrite trace element loads, and the `δ^34S`—enriched sulfur isotope compositions all strongly suggest that significant volumes of entrained seawater (with relatively unmodified chemical compositions) played an important role in the evolution of the CQP facies. Variations in the volume of fault-entrained seawater between the Major Lake and Caroline Cove Faults were likely influenced by contrasting host rock permeabilities (pillow basalts versus sheeted dykes), different crustal depths, and heterogeneous fluid circulation patterns, i.e., differences in the mechanism of along-fault and up-fault focussed flow. The diagnostic alteration facies in the NW-striking Sellick Bay Fault Zone (SBFZ) is highly distinctive, and differs significantly from those of fault-hosted rocks at Major Lake and Caroline Cove. The vein-dominated, prehnite-zeolite (VPZ) facies outcrops in a ~ 50 m-wide and ~ 200 m-long fault-parallel zone of strongly altered pillow basalts at a single locality in the upper Sellick Bay escarpment. The structurally focussed alteration package is dominated by the heterogeneous development of Ca-rich zeolites (laumontite), prehnite, pumpellyite, and Fe-oxyhydroxides. These hydrothermal minerals, and their associated textural features, contrast sharply with those in the partly oxidised and relatively low grade volcanic rocks of surrounding regional domains. The VPZ facies comprises an unusual multi-component paragenesis (one of very few documented pumpellyite occurrences in oceanic crust) which formed during five main stages of focussed hydrothermal activity. The alteration minerals mainly occur as discrete veins > 10 mm-wide, intensive veinlet stockworks, and massively altered pillow basalt cores and inter-pillow margins. Relict hydrothermal fluid temperatures in the SBFZ were consistently < 200° C, and the low pressure conditions (0.5-1.0 kbar) suggest relatively shallow subseafloor depths (< 200 m) during pervasive fluid—rock interaction. Parent fluids of the VPZ facies were highly enriched in Ca and Fe, but contained lower concentrations of aqueous silica and `H_2S` compared to those which formed the VQC and CQP facies. In addition, the VPZ facies is not overprinted by late-stage, low temperature alteration assemblages (i.e., no FMC or PFO facies occur in the SBFZ), suggesting that hydrothermal activity was not as long-lived as in the Major Lake and Caroline Cove Fault Zones. The formation of the VPZ facies may be related to large-scale fluid recharge of the basement reservoir, or it may reflect the effects of compositionally diverse hydrothermal flow systems (discharge-related?) at the PMSR. Subseafloor hydrothermal activity has clearly played an important role in the geological evolution of upper crustal rocks on Macquarie Island. The complex and dynamic palaeo-hydrothermal system was closely linked to episodic bouts of magmatism and tectonism during the waning stages of slow-spreading (crustal extension) at the PMSR. The most anomalous and intense fluid—rock interactions were strongly focussed in major oceanic fault zones, which provided highly permeable fluid pathways at these crustal levels. Pervasive hydrothermal processes mainly involved partial to complete wall rock (mineral) replacement, intense veining, and hydraulic brecciation of basalts and dolerites at depths of up to several hundred metres below the seafloor. Spatial and temporal differences in physico-chemical fluid parameters and host rock reactions characterise the discrete hydrothermal systems which formed each alteration facies. Rapidly discharging fluid up flow systems were segmented in some faults due to underlying structural influences in the basement, e.g., along the MLFZ. In addition, variations in the relative size, intensity, and crustal level of hydrothermal flow patterns are reflected by heterogeneous mineral and geochernical compositions; these are particularly influenced by chemically distinct fluid reservoirs, i.e., the variable influence of deep crustal fluids (evolved) and less modified seawater. The structurally focussed alteration zones have also been affected by episodic tectonic activity (syn- and post-rift) which further modified their distribution and architecture. Also, the sporadic and spatially restricted effects of late hydrothermal activity (FMC and PFO facies overprint zones) indicate that Macquarie Island's major oceanic faults were long-lived fluid conduits. However, the widespread precipitation of hydrothermal cements significantly decreased fracture penneabilities over time, requiring ongoing structural renewal in each fault zone to maintain hydrothermal flow patterns and associated wall rock alteration processes in the upper crust.