Magnetic susceptibility as a high-resolution correlation and paleoenvironmental tool in Palaeozoic records: merits and pitfalls.

Magnetic susceptibility (MS) had been used for decades on sedimentary rocks as a paleoclimatic tool and as a tool for high resolution correlations on the basis of the link between MS and detrital inputs (highly magnetic minerals). MS is a very interesting tool for correlation and paleoenvironmental research because data acquisition is fast and straightforward, providing the high amount of samples needed for such study. Although, mostly in the case of Palaeozoic rocks, the original paleoenvironmental signal can be affected by different syn- or post-depositional processes.
In this work, we synthesized 10 years of our research on the application and on the origin of the magnetic susceptibility signal through the Devonian record of Belgium. This record is of peculiar interest, because the biostratigraphy is well known and there is a relatively continuous record over 20 My and through various paleoenvironments (platform, ramp, atoll) with different diagenetic maturity.
The first step was to compare MS with facies and to try to correlate the different sections. On the Frasnian platform sections, there is a clear link between MS and facies, with MS increasing towards shallower facies. This increase of MS towards shallower facies seems logical since these facies were deposited closer to the detrital inputs sources. When considering ramp (Eifelian-Givetian) and atoll sections (Frasnian), the MS relationship with facies is opposite and MS increases towards the deepest facies. This was interpreted as related to the water agitation and carbonate production rate being lower in the deeper environments and allowing concentration of magnetic particles. Although, correlations were possible between the different atoll sections, but not with the carbonate platform sections, since the MS signals are opposite in both settings. So the depositional setting can be a key parameter influencing the way the original magnetic susceptibility signal is recorded.
In order to identify some of the external parameters influencing the magnetic susceptibility signal; a MS curve from the Eiflian-Givetian ramp was selected for time-series analysis. This spectral analysis highlight persistent high-frequency meter scale cycles which are interpreted as reflecting changes in flux of magnetic minerals, most likely controlled by monsoon rainfall-intensity. By combining chrono- and biostratigraphic information with theoretical knowledge of sedimentation rates in different depositional environments, these cycles are interpreted as astronomically driven (precession-dominated). So this show that Magnetic susceptibility signal in some cases can be affected by astronomical forcing, through detrital inputs.
To get a better understanding about the nature and origin of the magnetic susceptibility signal, it is crucial to identify the magnetic minerals which are carrying this MS signal. Geochemistry and magnetic analyses (Hysteresis and acquisition curves of the Isothermal Remanent Magnetic Saturation IRM) were performed on three sections (2 Frasnian platform and one Eifelian givetian platform sections). Magnetic parameters show that the MS signal is mostly influenced by small grains magnetites which were formed through remagnetization event. This remagnetization happened during the Carboniferous and is related to the transformation of smectite into illite (realizing iron available for magnetite formation). On 2 sections (1 Frasnian platform and 1 Eifelian givetian platform), there is a strong link between the MS signal and TiO2, Al2O3 and Zr which are detrital proxies. This shows that despite the remagnetization, the MS signal is still related to detrital inputs, indicating that the newly formed magnetite probably remained associated with the original clay minerals, leading to a globally increased signal still reflecting the primary trends. In one section (Frasnian platform), there is no strong link between MS signal and detrital proxies, indicating that in this case the MS signal was affected by the remagnetization event (probably in relation with fluid circulation).
In conclusion, MS appears as a convolved signal, influenced on one hand by primary parameters such as detrital inputs and orbital cyclicities. Although, these paleoenvironmental trends can also be affected by depositional parameters such as water agitation and carbonate production. Furthermore, the diagenesis can also have a strong impact on the final signal. This clearly highlights the need to use magnetic susceptibility under the control of other techniques (comparison with other paleoenvironmental proxies or magnetic or geochemical analysis).