Sediment provenance in a tropical fluvial and marine context by magnetic ‘fingerprinting’ of transportable sand fractions

Abstract

AbstractThe sources and fluxes of sediment to the Great Barrier Reef lagoon from north‐eastern Australian rivers have been the subject of much concern and study, with the large catchments of the Burdekin and Fitzroy Rivers thought to be the key sources at present. Here, the utility of newly developed magnetic ‘fingerprinting’ methods for identifying sediment provenance, both onshore and offshore, and in association with individual large flood events, is investigated. Within the Burdekin catchment, sediments are mobilized from different subcatchments by runoff generated by intense, localized rainfall events. Magnetic measurements were made on untreated and acid‐treated samples of river channel sediments within the Burdekin River subcatchments and from the estuarine and inner shelf depocentres of Burdekin River sediments. The acid treatment removes all discrete magnetic particles and coatings, and leaves magnetic inclusions (protected within host silicate grains) as the basis of the measured magnetic signature of a sample. The magnetic properties of the acid‐treated samples display statistically distinct sediment provenance groupings. Sand samples from the Upper Burdekin River appear magnetically distinct from samples from tributaries of the Burdekin (e.g. Hann Creek, Fanning River) and also from nearby coastal rivers, including the Haughton. Suspended sand samples from a Burdekin flood event in 2000 appear to have a different source compared with those from floods in 1998 and 1999. Comparisons of the terrestrial, acid‐treated sand fractions with the same, acid‐treated, sand‐size fractions from transects taken offshore suggest that the surface sediments in Upstart Bay and Bowling Green Bay have different sources. Some of these sources are as yet unidentified but may represent the unsampled, lower‐discharge south‐western Burdekin subcatchments, and/or along‐shore drift of sand from the south, perhaps even from the Fitzroy River, over millennial timescales of cyclone pumping. The magnetic inclusion method precludes any obfuscation or confounding of sediment source, which might arise from hydraulic sorting and/or post‐depositional magnetic diagenesis or authigenesis.