The distribution of sediment residence times at the foot of mountains and its implications for proxies recorded in sedimentary basins

The geochemical and physical properties of terrigenous sediment stacked in sedimentary basins are used as proxies for the paleo-environmental conditions that prevailed during their period of deposition. Nevertheless, sediment grains have a stochastic transit from mountain sources to sedimentary basins: a fraction of grains are stored for a long time while others are recycled from old deposits. Consequently, the temporal representativity of a population of grains in a sedimentary stratum is uncertain. The potential recycling of old material is a major concern in the reconstruction of paleo-environments and this recycling is usually difficult to evaluate. In particular, the distribution of grain residence times in basins, between sources and sinks, is out of reach. Here we use a landscape evolution model that traces grains to analyse the distribution of residence times in an alluvial apron at the foot of a mountain relief. We study an end-member scenario that is the least favourable for the storage of grains: when the mountain is eroding at the same rate as rock is uplifting. In this case, the alluvial apron behaves as a by-pass zone, when averaging sediment flux over Ma, and the storage of grains of any size should be minimal. Yet, the model predicts that some grains are stored for hundreds of thousands of years before exiting the alluvial apron. Consequently, the mean residence time of sediment grains is much higher than the observed residence time of 95% of the grains exported by the alluvial apron rivers. This process may explain very long residence times found in fluvial systems by geochemical methods based on bulk measurements of sediment. Furthermore, it suggests that grains stored for a very long time, although a minority, can bias time-dependent proxies.