Construction of detrital mineral populations: insights from mixing of U–Pb zircon ages in Himalayan rivers


William H. Amidon, Geological and Planetary Sciences Division, California Institute of Technology, Pasadena, CA 91125, USA. E-mail:


Fission-track, U–Pb and Pb–Pb analyses of detrital heavy mineral populations in depositional basins and modern river sediments are widely used to infer the exhumational history of mountain belts. However, relatively few studies address the underlying assumption that detrital mineral populations provide an accurate representation of their entire source region. Implicit in this assumption is the idea that all units have equal potential to contribute heavy minerals in proportion to their exposure area in the source region. In reality, the detrital mineral population may be biased by variable concentrations of minerals in bedrock and differential erosion rates within the source region. This study evaluates the relative importance of these two variables by using mixing of U–Pb zircon ages to trace zircon populations from source units, through the fluvial system, and into the foreland.

The first part of the study focuses on the Marsyandi drainage in central Nepal, using tributaries that drain single formations to define the U–Pb age distributions of individual units and using trunk river samples to evaluate the relative contributions from each lithology. Observed mixing proportions are compared with proportions predicted by a simple model incorporating lithologic exposure area and zircon concentration. The relative erosion rates that account for the discrepancy between the observed and predicted mixing proportions are then modelled and compared with independent erosional proxies. The study also compares U–Pb age distributions from four adjacent drainages spanning ∼250 km along the Himalayan front using the Kolmogorov–Smirnov statistic and statistical estimates of the proportion of zircon derived from each upstream lithology. Results show that, along this broad swath of rugged mountains, the U–Pb age distributions are remarkably similar, thereby allowing data from more localized sources to be extrapolated along strike.