Biomass selection by floods and related timescales. Part 2: Stochastic modeling

We analyze the results from a flume experiment presented in a companion paper by Perona et al. [1] exploring the selective action of floods on pioneer riparian vegetation. We study the way seedlings react to periodic flow disturbances in the early stages of their growth and propose a conjecture to describe their statistical distribution in the alluvial bare sediment. In order to access the regime of competition between growth and uprooting by the floods, the experiment focuses on the situation where the flooding frequency is comparable with the plant germination and growth rates and gives a detailed insight into the statistics of the riverbed and uprooted vegetation. We develop a stochastic description for the growth-uprooting process and use our model to analyze the collected data. Results confirm the conjecture that the flooding events remove young vegetation with characteristics (we focus on the distribution of the main root length) remaining constant over time and allow to define the scale over which vegetation becomes resilient to uprooting.