Age and geochemistry of the Boucaut Volcanics in the Neoproterozoic Adelaide Rift Complex, South Australia

Tectonics and Earth Systems (TES) Group and the Mineral Exploration CRC, Department of Earth Sciences, The University of Adelaide, SA 5005, Australia Geological Survey of Canada, 601 Booth Street, Ottawa, Ontario, Canada Metal Earth, Harquail School of Earth Sciences, Laurentian University, Sudbury, Ontario, Canada School of Earth, Atmospheric and Life Sciences, University of Wollongong, Wollongong Australia


Introduction
The Adelaide Rift Complex in South Australia preserves Tonian to Middle Cambrian sedimentary and minor volcanic rocks. They preserve some of the best evidence for the evolving, and sometimes tumultuous, events that characterise this time in Earth's climatic, biological and geological systems. For example, the earliest known complex multicellular lifeforms are preserved within the Ediacara Hills of the Flinders Ranges, and extensive tillites provide evidence for Earth's global Cryogenian glaciation events (Hoffman et al., 2017;Le Heron et al., 2011). Of significance to paleogeographic reconstructions, the Adelaide Rift Complex also contains rocks that have been interpreted as forming during the breakup of supercontinent Rodinia (Merdith et al., 2017a;Powell et al., 1994;Preiss, 2000) (Figure 1).
Understanding the tectonic and geological evolution of the Adelaide Rift Complex underpins our understanding of these significant Neoproterozoic events, not only in Australia, but globally.
One of the major challenges in reconstructing the evolution of the Adelaide Rift Complex, is the lack of datable volcanic units and/or fossil assemblages that can provide quantitative age constraints on rifting and sedimentation. These challenges limit our ability to confidently correlate sequences in the Adelaide Rift Complex with those in other regions.

Tectonic overview
The Neoproterozoic to middle Cambrian stratigraphy within the Adelaide Rift Complex formed during at least five major successive rift cycles that led to the breakup of supercontinent Rodinia (Preiss, 2000). In the Adelaide Rift Complex, initiation of the breakup of Rodinia is marked by the 827 ± 6 Ma Gairdner Dyke Swarm (Wingate et al., 1998), which is interpreted to be coeval with the poorly dated Wooltana Volcanics (Compston et al., 1966). The second phase of rifting in the Adelaide Rift Complex is marked by the 802 ± 10 Ma Rook Tuff within the Callanna Group (pers. comm. Fanning 1994in Preiss 2000. The third phase of rifting is marked by the Boucaut Volcanics. This rift phase marks the beginning of extensive syn-rift facies within the Adelaidean, yet, it has resisted attempts at dating and forms the focus of this study. According to both the SWEAT (south-west US -East Antarctica; Dalziel, 1991;Moores, 1991) and AUSWUS (Australia-Western US; Burrett and Berry, 2000) hypotheses, the Laurentian and Valley have been proposed (e.g. Dehler et al., 2017;Mahon et al., 2014). An analysis of kinematic data for the different reconstructions in Figure 1 showed that models that put Australia adjacent to southern Laurentia (e.g. AUSWUS, and a more extreme version with Australia adjacent to Mexico -AUSMEX, Wingate et al. 2002) are the easiest to reconcile with Phanerozoic plate kinematic norms (Merdith et al., 2017b).
On a smaller scale, correlations between the Adelaide Rift Complex and northwest Tasmania have also been proposed, for example, between the c. 790 Ma Black River Dolomite of northwest Tasmania (Calver, 1998)

Constraining correlations has important implications for paleogeographic reconstructions of
Laurentia-Australia in the Rodinia supercontinent. Unfortunately, many of these correlations rely on old and/or unreliable age data, particularly for the Adelaide Rift Complex ( Figure 2).

Figure 1: GPlates tectonic reconstructions of the different Australia-Laurentia models at 788 Ma (the crystallisation age of the Boucaut Volcanics), made using the models in Merdith et al. (2017b). South Australia is fixed in its present-day position with all other blocks rotated relative to it at 788 Ma. SWEAT reconstruction based on Dalziel (1991); Hoffman (1991); Moores (1991). AUSWUS reconstruction based on Karlstrom et al. (1999). AUSMEX reconstruction based on Wingate et al. (2002). Missing Link reconstruction based on Li et al. (1995). NA = North Australia, SA = South
Australia.

The Boucaut Volcanics
The Boucaut Volcanics lie at the base of the Burra Group and provide an important maximum age constraint for this package. They also constrain the maximum age for the underlying Callanna Group. The age of the Boucaut Volcanics has been most widely reported as 777 ± 7 Ma (pers. comm. Fanning 1994in Preiss 2000, however, no isotopic data are published for this associated age. Confusingly, another source (Drexel et al., 1993) mentions that Fanning (1989) derived an upper intercept age of 783 ± 42 Ma for the Boucaut Volcanics, however the original source of these data are obscure. Regardless, robust isotopic age determinations are needed to constrain the age of this significant unit.
The Boucaut Volcanics are dominated by pale pink to grey rhyolite, with amygdaloidal andesite and basalt also present (Forbes, 1978). These rocks have undergone several phases of deformation and have been metamorphosed to 'biotite grade' (Forbes, 1978). The Boucaut Volcanics occur within the southeastern part of the Nackara Arc, and the majority of outcrops are isolated and many are sheared along the northeast-trending Anabama Shear Zone (Preiss, 2000). It has been suggested that the Boucaut Volcanics mark the onset of early Torrensian rifting in the Adelaide Rift Complex (Preiss, 2000).
In this contribution, we have collected new U-Pb zircon data from a rhyolite within the Boucaut Volcanics, to provide a robust age constraint on the timing of eruption. Significantly, this new age constrains the base of the Burra Group and the onset of early rifting within the Adelaide Rift Complex, providing important constraints on plate reconstructions for the breakup of supercontinent Rodinia (e.g. Merdith et al., 2017a;Merdith et al., 2017b).

Sample descriptions
The Sample descriptions and locations are provided in Table 1 and shown in Figure 2. Basalts were collected from the type section whilst rhyolites were collected from surrounding outcrops on the tops of small hills. Examples of outcrop textures are shown in Figure 3.   (Vermeesch, 2018) Trace element profiles from analyses that are within 10% of concordance are shown in Figure 6 along with their Th/U ratios. Zircons show Th/U values between 0.5 and 1.2 that are consistent with igneous zircons (Belousova et al., 2002). The majority of near concordant zircon divide into two coupled Th/U and REE populations ( Figure 6). One population has Th/U ratios >0.8, elevated rare earth elements and moderate positive Ce anomalies. The second population has Th/U ratios <0.8 and a pronounced positive Ce anomaly. Both populations have moderate negative Eu anomalies and positive medium to high rare earth element gradients. The negative

Figure 5: a) Conordia plot of U-Pb data, data within 5% of concordance are included in age calculations (orange ellipses), and excluded data are interpreted as Pb-loss (white ellipses); b) Weighted average plot of the same data shown in a. Plots and data produced using IsoplotR
Eu anomaly can be caused by the presence of plagioclase in the magma that the zircon grew in, and/or by a reducing magma. The latter possibility is discounted as a positive Ce anomaly is a sign of an oxidising magma (Trail et al., 2012). Additionally, Kirkland et al. (2015) showed that Th/U ratios positively correlate with temperature in a cooling fractionating magma due to the preferential magma depletion of U as the magma cools. We use these observations to suggest that our analysed zircons reflect growth in a cooling fractionating magma chamber that was becoming progressively more oxidized as it cooled.

Whole rock geochemistry
Rock samples from the Boucaut Volcanics range from basaltic to rhyolitic compositions, with SiO2 ranging from 45% to 79% (Figure 7a). Around half of the samples plot within the Rhyolite field on a total alkali silica (TAS) diagram (Le Bas et al., 1986), with the remaining samples plotting within the Basalt, Andesite and Dacite fields. Samples range from Ferroan to Magnesian.
UNPUBLISHED MANUSCRIPT -Corresponding author contact: Sheree Armistead -sarmistead@laurentian.ca On the REE diagram of sample/chondrite (Figure 7c), samples are enriched in LREE over HREE, with some exhibiting a negative Eu anomaly indicating plagioclase fractionation.
On the sample/primitive mantle REE diagram, samples show a strong negative Sr anomaly, which is more pronounced for the felsic samples. Samples show a strong positive Pb anomaly, with this signature being more pronounced for mafic samples.