Geochemistry of low-grade clastic rocks in the Acatlán Complex of southern Mexico: Evidence for local provenance in felsic–intermediate igneous rocks

Geochemical data of sedimentary basins have been used to determine their plate tectonic setting. Here we test geochemical data for paired psammite–pelite samples from two greywacke packages that underlie 60% of the Paleozoic Acatlán Complex of southern Mexico: one an Ordovician rift-passive margin suite and the other a Carboniferous active margin suite. Both units consist of low-grade, polydeformed psammites and pelites associated with rare tholeiitic flows and dikes. Analyses of paired psammite–pelite samples from a number of different units in the eastern part of the Acatlán Complex indicate that they all have a similar geochemistry with their major components dependent on the proportion of sand and clay. The chemical index of alteration yielded values between 55 and 70 recording a moderate degree of weathering of plagioclase to illite/kaolinite in the source area. The REE patterns suggest provenance in felsic–intermediate igneous rocks, and the wide range of TiO2–Zr ratios indicates that this was followed by extensive sorting. The mantle-normalized trace element patterns of the psammites generally display positive Zr and Hf anomalies reflecting elevated concentration of zircon in these rocks. Detrital zircon ages from some of these units show that the major source for these clastic rocks probably lay in the adjacent Mesoproterozoic Oaxacan Complex and in the Neoproterozoic basement underlying the Acatlán Complex and/or the Yucatan Peninsula. Geological parameters point to a rift-passive margin setting for the Ordovician rocks and an active, subduction-related margin for the Carboniferous units. However, the commonly used tectonic discrimination, bi- to tri-variant diagrams cannot readily distinguish between the two suites on geochemical grounds. Their geochemistry probably reflects the general felsic–intermediate compositions of the rocks in the Precambrian source regions, rather than the contemporaneous tectonic setting. Future work involving a more advanced statistical approach (e.g. multivariant methods, multidimensional scaling and classification trees) may yield more refined results, particularly in combination with detrital zircon geochronology.