Previous studies have shown that the gastropod Hydrobia ulvae destabilizes the top layers of fine-grained sediments. This process is mediated by the formation of a "biogenic" fluff layer that includes tracks, faecal pellets and mucus. This fluff layer has been shown to be easily resuspended before general bed erosion. In order to examine how fluff layer and bed. erosion interact, flume experiments were performed with fluid sediments of varying water contents. Ten thousand snails were placed and allowed to crawl for 5 h on the sediment surface, and then the resuspended sediment mass was measured in response to step-wise shear stress increases. Two distinct erosion phases were observed: (1) initial resuspension of the fluff layer and (2) the subsequent bed erosion. Both the bioturbation by snails and sediment water content interacted positively to increase erosion rates during the phase of fluff layer erosion. The presence of a fluff layer due to the snail's activities did not affect rates of subsequent bed erosion. A vertical model (1DV) was developed to simulate the succession in time of the two distinct erosion phases. Within this deterministic model, erosion rates of the fluff layer depend on the quantity of sediment that is present in the fluff layer. Previous behavioral observations of track formation mechanisms were integrated into model equations to account for the snail density and the water content dependence. The observed bed erosion was fairly well reproduced by considering the variation with depth of the sediment density as measured in the experiments. This model suggests a new approach for assessing the erosion of natural sediments under the influence of H. ulvae population density, water content and tidal currents.