July 2016 LIP of the Month

Importance of geochemistry as a fingerprinting tool for LIPs: Example from Proterozoic mafic magmatism in the Belt-Purcell Basin and Wyoming province of western Laurentia

Chris Rogers
PhD Candidate Ottawa-Carleton Geoscience Centre, Department. of Earth Sciences, Carleton University, 1125 Colonel By Drive, Ottawa, Canada K1S 5B6
Email: Chris.Rogers@Carleton.ca

Richard Ernst
Ottawa-Carleton Geoscience Centre, Department. of Earth Sciences, Carleton University, 1125 Colonel By Drive, Ottawa, Canada K1S 5B6;
Faculty of Geology and Geography, Tomsk State University, 36 Lenin Ave, Tomsk, 634050, Russia

Email: Richard.Ernst@ErnstGeosciences.com

Brian Cousens
Ottawa Carleton Geoscience Centre, Department. of Earth Sciences, Carleton University, 1125 Colonel By Drive, Ottawa, Canada K1S 5B6
Email: Brian.Cousens@Carleton.ca

Ulf Söderlund
Department of Earth & Ecosystem Science, Lund University, Sweden.
Email: Ulf.Soderlund@geol.lu.se

Steve S. Harlan
9209 Fairway Court, Manassas Park, VA 20111, USA.
Email: sharlan92@gmail.com

Alana Mackinder
Ottawa Carleton Geoscience Centre, Department. of Earth Sciences, Carleton University, 1125 Colonel By Drive, Ottawa, Canada K1S 5B6
Email: AlanaMackinder@cmail.carleton.ca

Based on:

Rogers, C., Mackinder, A., Ernst, R., Cousens, B. (2016a). Mafic Magmatism in the Belt-Purcell Basin and Wyoming Province of Western Laurentia. Geological Society of America Special Papers 522 (in press).

and

Rogers, C., Ernst, R. E., Cousens, B., Soderlund, U., Hamilton, M., Harlan, S. S. (2016b). Newly Recognized 1550-1590 Ma Tobacco Root -Western Channel Lip, and Further Evidence for Bringing South Australia Back Home to Western Laurentia. GAC-MAC. Whitehorse, YT: GAC-MAC.

Introduction

Large Igneous Provinces (LIPs) are large volume, mainly short duration, intraplate events of dominantly mafic composition (Ernst 2014). The constituent magmatic units of LIPs (primarily dykes, sills, layered intrusions, and volcanic rocks) are recognized mainly on the basis of matching ages (ideally based on precise U-Pb dating of zircon or baddeleyite), matching primary paleomagnetic directions, and/or a simple geometric relationship of dykes (e.g. a radiating dyke swarm) (e.g. Bleeker and Ernst 2006). One important tool that has been underutilized until recent years is geochemistry. Trace element and isotope geochemistry can be used to recognize intermixed intraplate magmatic units with distinct compositions that can either represent separate events or distinct geochemical packages within a LIP (e.g. Ernst, 2014 and references therein).

Proterozoic intraplate mafic magmatism in western Laurentia represents a useful example in which geochemical discriminants can demonstrably resolve intermixed events (Rogers et al. 2016a). This magmatism was originally thought to be related to either the ca. 780 Ma Gunbarrel LIP (Feb. 2004, LOTM http://www.largeigneousprovinces.org/04feb) or the ca. 1460 Ma Moyie-Purcell LIP (Figure 1; published dates located in Figure 2 and summarized in Table 1a and 1b). A comprehensive regional geochemical analysis involving 204 samples of dolerite dykes and sills (Rogers et al. 2016a), augmented with 103 analyses from Anderson and Goodfellow (2000) shows that there are in fact numerous distinct geochemical groups (at least 17) that argue for the presence of more than just these two previously recognized LIPs (Figure 3a-c). The geochemical evidence led to a test by U-Pb dating which confirmed the presence of entirely new, previously unrecognized magmatic provinces of different age.

Geochemical Signatures of Western Laurentia

Key trace element parameters used to distinguish separate compositional groups in this study include the following: rare earth (Figure 4) and multi-element (Figure 5) plot patterns, Ti-V plots (Figure 6), and Th/Yb-Nb/Yb plots (Figure 7) (summarized in Table 2); these collectively highlight differences in depth of melting, degree of melting, and interaction between these melts and the continental lithosphere. Full details of this approach are explored in Rogers et al. (2016a). The mafic magmatism, originally grouped into either the ca. 780 Ma Gunbarrel LIP or ca. 1460 Moyie-Purcell LIP, can now be divided into at least 17 geochemical groups. Some of these geochemical groups are expected to represent new intraplate events, potentially of LIP scale. The specifics of new geochemical groupings are shown below, and more fully detailed in Rogers et al. (2016a).

Of the 17 geochemical groups, seven of them can be correlated to the Moyie-Purcell LIP which can be further subdivided into subgroups of ca. 1450, ca. 1460 Ma and ca. 1470 Ma pulses (Rogers et al. 2016a). Two other groups belong to a newly recognized 1550-1590 Ma Tobacco Root Mountain event (highlighted more below), and another belongs to the well-known ca. 780 Ma Gunbarrel LIP (Mackinder, 2014; Sandeman et al. 2014, and references therein). Three additional distinct geochemical groups are in the process of being precisely dated by U-Pb methods, and preliminary indications are that these groups also have different ages. The final four geochemical groups are each expressed by three or fewer samples, and though localized to specific areas, may characterize important unrecognized intraplate events.

Geochemical Signatures (GS) GS 7 and GS 8: Originally, geochemical groups GS 7 and GS 8, in the Tobacco Root Mountains were correlated to the Moyie-Purcell LIP (Ernst and Buchan 2010) via imprecise Rb-Sr and Sm-Nd ages (Wooden 1975; Harlan et al. 2008). However, Rogers et al. (2016a), noted that GS 7 and GS 8 represented distinct geochemical groups, and his prediction of their different ages is confirmed by new U-Pb dating of ca. 1550 and 1590 Ma, respectively (Rogers et al. 2015; Rogers et al. 2016b). The two groups are now named the ca.1550 Ma Ramshorn Creek (GS 7) and ca. 1590 Ma Mammoth (GS 8) groups, and are considered parts of the Tobacco Root Mountain event (Rogers et al., 2015, Rogers et al., 2016a). The geochemical differences between the 1590 Ma Mammoth and 1550 Ma Ramshorn Creek groups appear to reflect moderate levels of mantle lithospheric contamination in the former, but minimal influence in the latter (e.g. Figures 5, 7, 8).

Ages of 1590 Ma, that match with the Mammoth group age, have been obtained for the Western Channel Diabase (2000 km north in the Great Bear Lake area of NW Laurentia) and also in the Gawler Range mafic volcanics (Gawler craton, Australia), which was reconstructed adjacent to NW Laurentia at this time (Hamilton and Buchan, 2010).

A geochemical comparison of the 1550-1590 Ma units in the Tobacco Root Mountains with the 1590 Ma Western Channel Diabase and Gawler Range volcanics is shown in Figure 8. Three Western Channel samples plot with the Mammoth Group, but one plots in the 1550 Ma Ramshorn Creek grouping. The Gawler Range mafic magmatism mostly plots with the 1590 Ma Mammoth Group data but some plots to even more contaminated levels. Overall, as shown in Figure 8 there is a broad linkage of all these 1550-1590 Ma data (in the Tobacco Root Mountains of western Laurentia, northwestern Laurentia, and formerly attached Gawler craton) to variable lithospheric contamination of an EMORB type source.

The new ca. 1550 and 1590 Ma ages of the Ramshorn Creek and Mammoth groups also help solve an aspect of the Meso- to Neoproterozoic Laurentian apparent polar wander path (APWP), previously constructed assuming a Moyie-Purcell (1460 Ma) age (Figure 9; APWP not shown). The mean paleomagnetic directions for dykes from the Tobacco Root Mountains from Harlan et al. (2008), now recognized to have 1550 and1590 Ma ages (Ramshorn Creek and Mammoth groups), are compared with the direction for the ca. 1590 Ma Western Channel Diabase pole (Hamilton and Buchan, 2010) after extrapolation into the coordinates for the Tobacco Root Mountains. The difference between Tobacco Root Mountain and Western Channel paleomagnetic directions is explained by block tilting of the Tobacco Root Mountain sites due to the Laramide and younger deformation (yielding a tectonic correction of strike / dip of 002º / 19º). A similar correction explains the offset in the 780 Ma paleomagnetic direction in the Tobacco Root Mountains with respect to a reference paleomagnetic direction for the 780 Ma Gunbarrel event (Rogers et al. 2015; Rogers et al. 2016b).

This project has demonstrated the potential of geochemical studies to identify anomalous magmatic suites, to predict distinct igneous events, and target units for follow-up geochronological investigations. In this case, the combination of geochemistry and U-Pb geochronology has exposed the presence of a previously unrecognized 1590 - 1550 Ma Tobacco Root Mountain event (ongoing studies are aimed at understanding whether this is a single long-lived event, or two distinct periods of mafic magmatism). The orientation of dykes in the Tobacco Root Mountain event is indistinguishable from the regional ca. 1460 and 780 Ma Proterozoic mafic dykes. Therefore, dyke orientations, whether they are radial or linear, cannot be used in isolation to determine LIP associations.

For example, in a paleomagnetic study of dykes in the Beartooth Mountains, Montana, a dyke thought to be a splay of late Paleoproterozoic Sourdough swarm was sampled for geochemistry and paleomagnetism (Kilian et al. in revision). However, its εNd value yielded a value that was anomalously high compared to other Sourdough dykes. When recalculated to 780 Ma, the age of the Gunbarrel LIP, the recalculated εNd for this lone dyke was an exact match to that of other Gunbarrel intrusions (Cousens, B. and Kilian, T. pers. comm., 2016).

Another example is an interpretation linking the Irene and Huckleberry volcanics in northern Washington State (USA) with the ca. 780 Ma Gunbarrel LIP by Harlan et al. (2003), based on an Sm-Nd age (Devlin et al. 1988) and proximity to the Gunbarrel LIP. However Mackinder (2014) was able to show that the Irene and Huckleberry volcanics are geochemically and isotopically distinct from the Gunbarrel LIP. She interpreted that the Irene and Huckleberry volcanic rocks were not fed from Gunbarrel intrusions and they may be rift volcanics associated with the break up of Rodinia, and could correspond to a later pulse of the Gunbarrel LIP.

Conclusions

Geochronological, paleomagnetic and dyke orientation studies have proven important for determining the overall extent of LIP events within a crustal block and for reconstruction of matching LIP units between formerly neighboring crustal blocks. However, geochemical studies can furnish equally important constraints on the distribution of individual LIPs. The additional evidence that geochemistry provides can allow LIPs to be parsed into distinct compositional groups that represent either magma batches with distinct differentiation histories within a single LIP, or intermixed magmatic units that belong to different LIPs. The application of our geochemical studies to intraplate Proterozoic mafic igneous units in western Laurentia has provided compositional ‘fingerprints’ that allow recognition of 17 individual geochemical groups of which 9 appear distinct from the ca. 1460 Ma Moyie-Purcell and ca. 780 Ma Gunbarrel LIPs, and likely represent distinct events of different age(s). This hypothesis has been confirmed through subsequent U-Pb dating of two of these groups, yielding discrete ages at ca. 1550 and ca. 1590 Ma.

Acknowledgement

Helpful comments on this LOTM by Mike Hamilton are gratefully appreciated.