Geophones were used as a proxy for determining the behavior of sediment transport in boulder mountain river. Field observations aggregated by a statistical analysis (PCA) allowed to identifying periods in which the behaviour between the water discharge $(Q_{w})$, the sediment flux $(Q_{s})$ and the meteorologic conditions was correlated. Within each period a relationship between $(Q_{s})$ and Qw could be found and summarized by two parameters: the critical drag discharge for motion initiation $((Q_{d})$ and the mean transport rate $(T_{mean})$. The coefficients of determination were close to 0.8, which showed the consistency of the linear relationship. The variability in the relationship $(Q_{s})$ $(Q_{w})$ could be related to the availability to motion of sediments. In contrast with laboratory experiments, in which sediment storage is indefinitely large, natural systems can suffer from sediment shortage under certain conditions. This may explain why bedload transport equations determined from laboratory experiments significantly overestimate the bedload transport rate.