Detrital Controls on Magnetosusceptibility and Cyclostratigraphy Records

Magnetic susceptibility (MS) is a rapid, simple, and nondestructive measurement frequently made on marine sediments and sedimentary rocks. The practical utility of MS is for correlation purposes, however many records demonstrate Milankovitch climatic cycles and are also used for the astronomical tuning of cyclostratigraphic records. To correctly interpret long-term cyclostratigraphic MS records, it is important to understand the factors influencing MS values. It is often argued (and occasionally documented) that MS values exhibit an inverse relationship with calcium carbonate concentrations in marine sediment samples. As a result, MS records are often interpreted as being controlled by calcium carbonate concentrations, and therefore, that MS may be a reliable indication of past carbonate productivity. However, other studies demonstrate strong detrital controls on MS values and conclude that MS variations are a function of changes in terrigenous input. To test these two models, we present new experimental results and extensive data sets that demonstrate that changes in mineralogy and/or detrital input are the main controls on MS variability, not carbonate productivity. To simulate the effect of variations in biological productivity on marine sediments, we prepared three separate samples of iron-rich minerals found in marine sediments and then incrementally diluted them with powdered diamagnetic calcium carbonate (MS ~ -3x10-9 m3/kg) and measured the new MS value. Prepared samples were designed to cover the common range of MS observed for lithified marine sediments. In order of decreasing initial MS, the samples are: 1) ~1 g of powdered Iceland basalt (MS = 7.58x10-6 m3/kg); 2) a ~1 g paramagnetic illite standard (MS = 1.20x10-7 m3/kg); and 3) a ~2 g 50/50 mixture of illite and quartz sand (MS = 4.53x10-8 m3/kg). The calcium carbonate dilution required to reduce initial sample susceptibilities by one order of magnitude is over 700% for all samples, which is far beyond natural variations in carbonate productivity. Even the mixed illite-quartz sample required ~880% dilution by calcite to reduce the susceptibility one order of magnitude. A nonlinear regression line (R=0.99) fits all experimental data. To further test whether MS is controlled by calcium carbonate concentrations in marine sediments, we present MS and calcium carbonate data for a 210 m section of the Upper Cretaceous Niobrara Chalk, Kansas. Overall, MS and calcium carbonate are not significantly correlated (R=0.14, n=1840).