The appearance of the regular vegetated ridge patterns observed in some ephemeral rivers of semi-arid regions (Nanson, Tooth, & Knigthon 2002) has previously been explained by hydraulic arguments (optimization of the bed load transport capacity, see Huang & Nanson (2007)) without including the role of vegetation in the process. Those arguments provide neither the conditions under which the more efficient anabranching system can be realized nor a description of the dynamics leading to anabranching. As an alternative, we propose a simplified model accounting for the flow-mediated interactions between riparian vegetation located at different points of the river channel. Classically, the appearance of river morphologic features is explained by the equations of morphodynamics. However, due to the complexity of the action of vegetation on the flow and on sediment transport, a complete physically-based set of coupled ecomorphodynamic equations is not available yet. We propose an effective model for the interactions between vegetated obstacles. Depending on the relative position of the obstacles, one observes either positive or negative feedbacks: on the lee-side of a permeable obstacle, flood sheltering may occur and favor deposition that helps in turn the establishment of biomass, conversely scouring is increased laterally due to obstacle-induced flow diversion (Zong & Nepf 2010). In the situation where the hydrological timescale (flooding frequency for a given effective magnitude) and the biological timescale (vegetation development rate) are comparable, the spatially inhomogeneous feedbacks can result in the appearance of organized regular structures. Aerial photographs give us the characteristic morphological scale of patterns. We then perform a stability analysis of our model and derive a set of conditions under which the combination of hydrological, ecological and pedological factors allows the formation of anabranching patterns. We discuss the role of the different ingredients in relation to the fluvial environments in which such patterns typically emerge. Finally, we conjecture on the relevance of the proposed mechanism to explain ubiquitous vegetated scroll bars that are observed on the interior of meander bends.

Several research investigations have explored the interaction between morphodynamic and vegetation growth processes from both the modelling and the experimental viewpoints. Results have mainly been concerned with morphologic analyses of the effects of vegetation on long term riverbed evolution without addressing the relative role of the timescales between such processes. This paper presents for the first time the statistics of uprooted biomass obtained while perturbing the vegetation growing in the river bed with periodic disturbances of constant magnitude. That is, we force the biological and hydrological processes to interact and study the related timescales in order to shed light on the role of flood disturbances in selecting the component of the biomass that has a higher chance of survival in relation to its growth stage. A simple interpretative stochastic model is then presented and thoroughly discussed in a companion paper (Biomass selection by floods and related timescales: Part 2. Stochastic modelling).