A tale of two riffles: Using multidimensional, multifractional, timevarying sediment transport to assess selfmaintenance in poolriffle sequences

Pool-riffle sequences play a central role in providing habitat diversity conditions both in terms of flow and substrate in gravel bed streams. Understanding their capacity to self-maintain has been the focus of research for many years, starting with the velocity reversal hypothesis. This hypothesis relied only on cross sectional averaged flow information, but its limited success prompted extensions of the hypothesis and alternative explanations for self-maintenance. Significant advances beyond the velocity reversal hypothesis have been achieved by incorporating more information either on flow or sediment transport characteristics. However, this has been done in a compartmentalized way, with studies either focusing on one or the other aspect. This work bridges the gap between these two aspects by using an approximate methodology that combines observed characteristic stage-dependent 3-D flow patterns with time-varying cross sectional information on bed shear stresses, sediment distribution, and sediment bed changes during a 1 year record of continuous discharges from a real stream. This methodology allows us to track the behavior of different sediment size fractions along flow streamlines over time and identify self-maintenance conditions due to the combined effect of both flow multidimensionality and sediment transport. We apply this approximate methodology to two contiguous pools and riffles and demonstrate that, unexpectedly, they may rely on different mechanisms for self-maintenance due to differences in geometry and sediment size distribution. We also demonstrate that our methodology is potentially overarching and integrative of previous partial approaches based on flow multidimensionality or sediment transport, which tend to underestimate the occurrence of self-maintenance.